diff options
| author | André Fabian Silva Delgado <emulatorman@parabola.nu> | 2016-04-16 15:30:54 -0300 |
|---|---|---|
| committer | André Fabian Silva Delgado <emulatorman@parabola.nu> | 2016-04-16 15:30:54 -0300 |
| commit | bdcfd44fb5b5fb8fd660e7f93f1095c507481024 (patch) | |
| tree | e423b07154d422b711ddfadedb87c43317d3c4f6 | |
| parent | 4a327fcef90ba27150a3e8741441b68c605ae248 (diff) | |
Linux-libre 4.5.1-gnupck-4.5.1-gnu
374 files changed, 44306 insertions, 1327 deletions
diff --git a/Documentation/ABI/testing/debugfs-aufs b/Documentation/ABI/testing/debugfs-aufs new file mode 100644 index 000000000..99642d105 --- /dev/null +++ b/Documentation/ABI/testing/debugfs-aufs @@ -0,0 +1,50 @@ +What: /debug/aufs/si_<id>/ +Date: March 2009 +Contact: J. R. Okajima <hooanon05g@gmail.com> +Description: + Under /debug/aufs, a directory named si_<id> is created + per aufs mount, where <id> is a unique id generated + internally. + +What: /debug/aufs/si_<id>/plink +Date: Apr 2013 +Contact: J. R. Okajima <hooanon05g@gmail.com> +Description: + It has three lines and shows the information about the + pseudo-link. The first line is a single number + representing a number of buckets. The second line is a + number of pseudo-links per buckets (separated by a + blank). The last line is a single number representing a + total number of psedo-links. + When the aufs mount option 'noplink' is specified, it + will show "1\n0\n0\n". + +What: /debug/aufs/si_<id>/xib +Date: March 2009 +Contact: J. R. Okajima <hooanon05g@gmail.com> +Description: + It shows the consumed blocks by xib (External Inode Number + Bitmap), its block size and file size. + When the aufs mount option 'noxino' is specified, it + will be empty. About XINO files, see the aufs manual. + +What: /debug/aufs/si_<id>/xino0, xino1 ... xinoN +Date: March 2009 +Contact: J. R. Okajima <hooanon05g@gmail.com> +Description: + It shows the consumed blocks by xino (External Inode Number + Translation Table), its link count, block size and file + size. + When the aufs mount option 'noxino' is specified, it + will be empty. About XINO files, see the aufs manual. + +What: /debug/aufs/si_<id>/xigen +Date: March 2009 +Contact: J. R. Okajima <hooanon05g@gmail.com> +Description: + It shows the consumed blocks by xigen (External Inode + Generation Table), its block size and file size. + If CONFIG_AUFS_EXPORT is disabled, this entry will not + be created. + When the aufs mount option 'noxino' is specified, it + will be empty. About XINO files, see the aufs manual. diff --git a/Documentation/ABI/testing/sysfs-aufs b/Documentation/ABI/testing/sysfs-aufs new file mode 100644 index 000000000..82f951849 --- /dev/null +++ b/Documentation/ABI/testing/sysfs-aufs @@ -0,0 +1,31 @@ +What: /sys/fs/aufs/si_<id>/ +Date: March 2009 +Contact: J. R. Okajima <hooanon05g@gmail.com> +Description: + Under /sys/fs/aufs, a directory named si_<id> is created + per aufs mount, where <id> is a unique id generated + internally. + +What: /sys/fs/aufs/si_<id>/br0, br1 ... brN +Date: March 2009 +Contact: J. R. Okajima <hooanon05g@gmail.com> +Description: + It shows the abolute path of a member directory (which + is called branch) in aufs, and its permission. + +What: /sys/fs/aufs/si_<id>/brid0, brid1 ... bridN +Date: July 2013 +Contact: J. R. Okajima <hooanon05g@gmail.com> +Description: + It shows the id of a member directory (which is called + branch) in aufs. + +What: /sys/fs/aufs/si_<id>/xi_path +Date: March 2009 +Contact: J. R. Okajima <hooanon05g@gmail.com> +Description: + It shows the abolute path of XINO (External Inode Number + Bitmap, Translation Table and Generation Table) file + even if it is the default path. + When the aufs mount option 'noxino' is specified, it + will be empty. About XINO files, see the aufs manual. diff --git a/Documentation/cgroup-v2.txt b/Documentation/cgroup-v2.txt index ff49cf901..81eb37821 100644 --- a/Documentation/cgroup-v2.txt +++ b/Documentation/cgroup-v2.txt @@ -1368,6 +1368,12 @@ system than killing the group. Otherwise, memory.max is there to limit this type of spillover and ultimately contain buggy or even malicious applications. +Setting the original memory.limit_in_bytes below the current usage was +subject to a race condition, where concurrent charges could cause the +limit setting to fail. memory.max on the other hand will first set the +limit to prevent new charges, and then reclaim and OOM kill until the +new limit is met - or the task writing to memory.max is killed. + The combined memory+swap accounting and limiting is replaced by real control over swap space. diff --git a/Documentation/filesystems/aufs/README b/Documentation/filesystems/aufs/README new file mode 100644 index 000000000..ed1bafe08 --- /dev/null +++ b/Documentation/filesystems/aufs/README @@ -0,0 +1,391 @@ + +Aufs4 -- advanced multi layered unification filesystem version 4.x +http://aufs.sf.net +Junjiro R. Okajima + + +0. Introduction +---------------------------------------- +In the early days, aufs was entirely re-designed and re-implemented +Unionfs Version 1.x series. Adding many original ideas, approaches, +improvements and implementations, it becomes totally different from +Unionfs while keeping the basic features. +Recently, Unionfs Version 2.x series begin taking some of the same +approaches to aufs1's. +Unionfs is being developed by Professor Erez Zadok at Stony Brook +University and his team. + +Aufs4 supports linux-4.0 and later, and for linux-3.x series try aufs3. +If you want older kernel version support, try aufs2-2.6.git or +aufs2-standalone.git repository, aufs1 from CVS on SourceForge. + +Note: it becomes clear that "Aufs was rejected. Let's give it up." + According to Christoph Hellwig, linux rejects all union-type + filesystems but UnionMount. +<http://marc.info/?l=linux-kernel&m=123938533724484&w=2> + +PS. Al Viro seems have a plan to merge aufs as well as overlayfs and + UnionMount, and he pointed out an issue around a directory mutex + lock and aufs addressed it. But it is still unsure whether aufs will + be merged (or any other union solution). +<http://marc.info/?l=linux-kernel&m=136312705029295&w=1> + + +1. Features +---------------------------------------- +- unite several directories into a single virtual filesystem. The member + directory is called as a branch. +- you can specify the permission flags to the branch, which are 'readonly', + 'readwrite' and 'whiteout-able.' +- by upper writable branch, internal copyup and whiteout, files/dirs on + readonly branch are modifiable logically. +- dynamic branch manipulation, add, del. +- etc... + +Also there are many enhancements in aufs, such as: +- test only the highest one for the directory permission (dirperm1) +- copyup on open (coo=) +- 'move' policy for copy-up between two writable branches, after + checking free space. +- xattr, acl +- readdir(3) in userspace. +- keep inode number by external inode number table +- keep the timestamps of file/dir in internal copyup operation +- seekable directory, supporting NFS readdir. +- whiteout is hardlinked in order to reduce the consumption of inodes + on branch +- do not copyup, nor create a whiteout when it is unnecessary +- revert a single systemcall when an error occurs in aufs +- remount interface instead of ioctl +- maintain /etc/mtab by an external command, /sbin/mount.aufs. +- loopback mounted filesystem as a branch +- kernel thread for removing the dir who has a plenty of whiteouts +- support copyup sparse file (a file which has a 'hole' in it) +- default permission flags for branches +- selectable permission flags for ro branch, whether whiteout can + exist or not +- export via NFS. +- support <sysfs>/fs/aufs and <debugfs>/aufs. +- support multiple writable branches, some policies to select one + among multiple writable branches. +- a new semantics for link(2) and rename(2) to support multiple + writable branches. +- no glibc changes are required. +- pseudo hardlink (hardlink over branches) +- allow a direct access manually to a file on branch, e.g. bypassing aufs. + including NFS or remote filesystem branch. +- userspace wrapper for pathconf(3)/fpathconf(3) with _PC_LINK_MAX. +- and more... + +Currently these features are dropped temporary from aufs4. +See design/08plan.txt in detail. +- nested mount, i.e. aufs as readonly no-whiteout branch of another aufs + (robr) +- statistics of aufs thread (/sys/fs/aufs/stat) + +Features or just an idea in the future (see also design/*.txt), +- reorder the branch index without del/re-add. +- permanent xino files for NFSD +- an option for refreshing the opened files after add/del branches +- light version, without branch manipulation. (unnecessary?) +- copyup in userspace +- inotify in userspace +- readv/writev + + +2. Download +---------------------------------------- +There are three GIT trees for aufs4, aufs4-linux.git, +aufs4-standalone.git, and aufs-util.git. Note that there is no "4" in +"aufs-util.git." +While the aufs-util is always necessary, you need either of aufs4-linux +or aufs4-standalone. + +The aufs4-linux tree includes the whole linux mainline GIT tree, +git://git.kernel.org/.../torvalds/linux.git. +And you cannot select CONFIG_AUFS_FS=m for this version, eg. you cannot +build aufs4 as an external kernel module. +Several extra patches are not included in this tree. Only +aufs4-standalone tree contains them. They are described in the later +section "Configuration and Compilation." + +On the other hand, the aufs4-standalone tree has only aufs source files +and necessary patches, and you can select CONFIG_AUFS_FS=m. +But you need to apply all aufs patches manually. + +You will find GIT branches whose name is in form of "aufs4.x" where "x" +represents the linux kernel version, "linux-4.x". For instance, +"aufs4.0" is for linux-4.0. For latest "linux-4.x-rcN", use +"aufs4.x-rcN" branch. + +o aufs4-linux tree +$ git clone --reference /your/linux/git/tree \ + git://github.com/sfjro/aufs4-linux.git aufs4-linux.git +- if you don't have linux GIT tree, then remove "--reference ..." +$ cd aufs4-linux.git +$ git checkout origin/aufs4.0 + +Or You may want to directly git-pull aufs into your linux GIT tree, and +leave the patch-work to GIT. +$ cd /your/linux/git/tree +$ git remote add aufs4 git://github.com/sfjro/aufs4-linux.git +$ git fetch aufs4 +$ git checkout -b my4.0 v4.0 +$ (add your local change...) +$ git pull aufs4 aufs4.0 +- now you have v4.0 + your_changes + aufs4.0 in you my4.0 branch. +- you may need to solve some conflicts between your_changes and + aufs4.0. in this case, git-rerere is recommended so that you can + solve the similar conflicts automatically when you upgrade to 4.1 or + later in the future. + +o aufs4-standalone tree +$ git clone git://github.com/sfjro/aufs4-standalone.git aufs4-standalone.git +$ cd aufs4-standalone.git +$ git checkout origin/aufs4.0 + +o aufs-util tree +$ git clone git://git.code.sf.net/p/aufs/aufs-util aufs-util.git +- note that the public aufs-util.git is on SourceForge instead of + GitHUB. +$ cd aufs-util.git +$ git checkout origin/aufs4.0 + +Note: The 4.x-rcN branch is to be used with `rc' kernel versions ONLY. +The minor version number, 'x' in '4.x', of aufs may not always +follow the minor version number of the kernel. +Because changes in the kernel that cause the use of a new +minor version number do not always require changes to aufs-util. + +Since aufs-util has its own minor version number, you may not be +able to find a GIT branch in aufs-util for your kernel's +exact minor version number. +In this case, you should git-checkout the branch for the +nearest lower number. + +For (an unreleased) example: +If you are using "linux-4.10" and the "aufs4.10" branch +does not exist in aufs-util repository, then "aufs4.9", "aufs4.8" +or something numerically smaller is the branch for your kernel. + +Also you can view all branches by + $ git branch -a + + +3. Configuration and Compilation +---------------------------------------- +Make sure you have git-checkout'ed the correct branch. + +For aufs4-linux tree, +- enable CONFIG_AUFS_FS. +- set other aufs configurations if necessary. + +For aufs4-standalone tree, +There are several ways to build. + +1. +- apply ./aufs4-kbuild.patch to your kernel source files. +- apply ./aufs4-base.patch too. +- apply ./aufs4-mmap.patch too. +- apply ./aufs4-standalone.patch too, if you have a plan to set + CONFIG_AUFS_FS=m. otherwise you don't need ./aufs4-standalone.patch. +- copy ./{Documentation,fs,include/uapi/linux/aufs_type.h} files to your + kernel source tree. Never copy $PWD/include/uapi/linux/Kbuild. +- enable CONFIG_AUFS_FS, you can select either + =m or =y. +- and build your kernel as usual. +- install the built kernel. + Note: Since linux-3.9, every filesystem module requires an alias + "fs-<fsname>". You should make sure that "fs-aufs" is listed in your + modules.aliases file if you set CONFIG_AUFS_FS=m. +- install the header files too by "make headers_install" to the + directory where you specify. By default, it is $PWD/usr. + "make help" shows a brief note for headers_install. +- and reboot your system. + +2. +- module only (CONFIG_AUFS_FS=m). +- apply ./aufs4-base.patch to your kernel source files. +- apply ./aufs4-mmap.patch too. +- apply ./aufs4-standalone.patch too. +- build your kernel, don't forget "make headers_install", and reboot. +- edit ./config.mk and set other aufs configurations if necessary. + Note: You should read $PWD/fs/aufs/Kconfig carefully which describes + every aufs configurations. +- build the module by simple "make". + Note: Since linux-3.9, every filesystem module requires an alias + "fs-<fsname>". You should make sure that "fs-aufs" is listed in your + modules.aliases file. +- you can specify ${KDIR} make variable which points to your kernel + source tree. +- install the files + + run "make install" to install the aufs module, or copy the built + $PWD/aufs.ko to /lib/modules/... and run depmod -a (or reboot simply). + + run "make install_headers" (instead of headers_install) to install + the modified aufs header file (you can specify DESTDIR which is + available in aufs standalone version's Makefile only), or copy + $PWD/usr/include/linux/aufs_type.h to /usr/include/linux or wherever + you like manually. By default, the target directory is $PWD/usr. +- no need to apply aufs4-kbuild.patch, nor copying source files to your + kernel source tree. + +Note: The header file aufs_type.h is necessary to build aufs-util + as well as "make headers_install" in the kernel source tree. + headers_install is subject to be forgotten, but it is essentially + necessary, not only for building aufs-util. + You may not meet problems without headers_install in some older + version though. + +And then, +- read README in aufs-util, build and install it +- note that your distribution may contain an obsoleted version of + aufs_type.h in /usr/include/linux or something. When you build aufs + utilities, make sure that your compiler refers the correct aufs header + file which is built by "make headers_install." +- if you want to use readdir(3) in userspace or pathconf(3) wrapper, + then run "make install_ulib" too. And refer to the aufs manual in + detail. + +There several other patches in aufs4-standalone.git. They are all +optional. When you meet some problems, they will help you. +- aufs4-loopback.patch + Supports a nested loopback mount in a branch-fs. This patch is + unnecessary until aufs produces a message like "you may want to try + another patch for loopback file". +- vfs-ino.patch + Modifies a system global kernel internal function get_next_ino() in + order to stop assigning 0 for an inode-number. Not directly related to + aufs, but recommended generally. +- tmpfs-idr.patch + Keeps the tmpfs inode number as the lowest value. Effective to reduce + the size of aufs XINO files for tmpfs branch. Also it prevents the + duplication of inode number, which is important for backup tools and + other utilities. When you find aufs XINO files for tmpfs branch + growing too much, try this patch. +- lockdep-debug.patch + Because aufs is not only an ordinary filesystem (callee of VFS), but + also a caller of VFS functions for branch filesystems, subclassing of + the internal locks for LOCKDEP is necessary. LOCKDEP is a debugging + feature of linux kernel. If you enable CONFIG_LOCKDEP, then you will + need to apply this debug patch to expand several constant values. + If don't know what LOCKDEP, then you don't have apply this patch. + + +4. Usage +---------------------------------------- +At first, make sure aufs-util are installed, and please read the aufs +manual, aufs.5 in aufs-util.git tree. +$ man -l aufs.5 + +And then, +$ mkdir /tmp/rw /tmp/aufs +# mount -t aufs -o br=/tmp/rw:${HOME} none /tmp/aufs + +Here is another example. The result is equivalent. +# mount -t aufs -o br=/tmp/rw=rw:${HOME}=ro none /tmp/aufs + Or +# mount -t aufs -o br:/tmp/rw none /tmp/aufs +# mount -o remount,append:${HOME} /tmp/aufs + +Then, you can see whole tree of your home dir through /tmp/aufs. If +you modify a file under /tmp/aufs, the one on your home directory is +not affected, instead the same named file will be newly created under +/tmp/rw. And all of your modification to a file will be applied to +the one under /tmp/rw. This is called the file based Copy on Write +(COW) method. +Aufs mount options are described in aufs.5. +If you run chroot or something and make your aufs as a root directory, +then you need to customize the shutdown script. See the aufs manual in +detail. + +Additionally, there are some sample usages of aufs which are a +diskless system with network booting, and LiveCD over NFS. +See sample dir in CVS tree on SourceForge. + + +5. Contact +---------------------------------------- +When you have any problems or strange behaviour in aufs, please let me +know with: +- /proc/mounts (instead of the output of mount(8)) +- /sys/module/aufs/* +- /sys/fs/aufs/* (if you have them) +- /debug/aufs/* (if you have them) +- linux kernel version + if your kernel is not plain, for example modified by distributor, + the url where i can download its source is necessary too. +- aufs version which was printed at loading the module or booting the + system, instead of the date you downloaded. +- configuration (define/undefine CONFIG_AUFS_xxx) +- kernel configuration or /proc/config.gz (if you have it) +- behaviour which you think to be incorrect +- actual operation, reproducible one is better +- mailto: aufs-users at lists.sourceforge.net + +Usually, I don't watch the Public Areas(Bugs, Support Requests, Patches, +and Feature Requests) on SourceForge. Please join and write to +aufs-users ML. + + +6. Acknowledgements +---------------------------------------- +Thanks to everyone who have tried and are using aufs, whoever +have reported a bug or any feedback. + +Especially donators: +Tomas Matejicek(slax.org) made a donation (much more than once). + Since Apr 2010, Tomas M (the author of Slax and Linux Live + scripts) is making "doubling" donations. + Unfortunately I cannot list all of the donators, but I really + appreciate. + It ends Aug 2010, but the ordinary donation URL is still available. + <http://sourceforge.net/donate/index.php?group_id=167503> +Dai Itasaka made a donation (2007/8). +Chuck Smith made a donation (2008/4, 10 and 12). +Henk Schoneveld made a donation (2008/9). +Chih-Wei Huang, ASUS, CTC donated Eee PC 4G (2008/10). +Francois Dupoux made a donation (2008/11). +Bruno Cesar Ribas and Luis Carlos Erpen de Bona, C3SL serves public + aufs2 GIT tree (2009/2). +William Grant made a donation (2009/3). +Patrick Lane made a donation (2009/4). +The Mail Archive (mail-archive.com) made donations (2009/5). +Nippy Networks (Ed Wildgoose) made a donation (2009/7). +New Dream Network, LLC (www.dreamhost.com) made a donation (2009/11). +Pavel Pronskiy made a donation (2011/2). +Iridium and Inmarsat satellite phone retailer (www.mailasail.com), Nippy + Networks (Ed Wildgoose) made a donation for hardware (2011/3). +Max Lekomcev (DOM-TV project) made a donation (2011/7, 12, 2012/3, 6 and +11). +Sam Liddicott made a donation (2011/9). +Era Scarecrow made a donation (2013/4). +Bor Ratajc made a donation (2013/4). +Alessandro Gorreta made a donation (2013/4). +POIRETTE Marc made a donation (2013/4). +Alessandro Gorreta made a donation (2013/4). +lauri kasvandik made a donation (2013/5). +"pemasu from Finland" made a donation (2013/7). +The Parted Magic Project made a donation (2013/9 and 11). +Pavel Barta made a donation (2013/10). +Nikolay Pertsev made a donation (2014/5). +James B made a donation (2014/7 and 2015/7). +Stefano Di Biase made a donation (2014/8). +Daniel Epellei made a donation (2015/1). +OmegaPhil made a donation (2016/1). + +Thank you very much. +Donations are always, including future donations, very important and +helpful for me to keep on developing aufs. + + +7. +---------------------------------------- +If you are an experienced user, no explanation is needed. Aufs is +just a linux filesystem. + + +Enjoy! + +# Local variables: ; +# mode: text; +# End: ; diff --git a/Documentation/filesystems/aufs/design/01intro.txt b/Documentation/filesystems/aufs/design/01intro.txt new file mode 100644 index 000000000..5d0121439 --- /dev/null +++ b/Documentation/filesystems/aufs/design/01intro.txt @@ -0,0 +1,157 @@ + +# Copyright (C) 2005-2016 Junjiro R. Okajima + +Introduction +---------------------------------------- + +aufs [ei ju: ef es] | [a u f s] +1. abbrev. for "advanced multi-layered unification filesystem". +2. abbrev. for "another unionfs". +3. abbrev. for "auf das" in German which means "on the" in English. + Ex. "Butter aufs Brot"(G) means "butter onto bread"(E). + But "Filesystem aufs Filesystem" is hard to understand. + +AUFS is a filesystem with features: +- multi layered stackable unification filesystem, the member directory + is called as a branch. +- branch permission and attribute, 'readonly', 'real-readonly', + 'readwrite', 'whiteout-able', 'link-able whiteout', etc. and their + combination. +- internal "file copy-on-write". +- logical deletion, whiteout. +- dynamic branch manipulation, adding, deleting and changing permission. +- allow bypassing aufs, user's direct branch access. +- external inode number translation table and bitmap which maintains the + persistent aufs inode number. +- seekable directory, including NFS readdir. +- file mapping, mmap and sharing pages. +- pseudo-link, hardlink over branches. +- loopback mounted filesystem as a branch. +- several policies to select one among multiple writable branches. +- revert a single systemcall when an error occurs in aufs. +- and more... + + +Multi Layered Stackable Unification Filesystem +---------------------------------------------------------------------- +Most people already knows what it is. +It is a filesystem which unifies several directories and provides a +merged single directory. When users access a file, the access will be +passed/re-directed/converted (sorry, I am not sure which English word is +correct) to the real file on the member filesystem. The member +filesystem is called 'lower filesystem' or 'branch' and has a mode +'readonly' and 'readwrite.' And the deletion for a file on the lower +readonly branch is handled by creating 'whiteout' on the upper writable +branch. + +On LKML, there have been discussions about UnionMount (Jan Blunck, +Bharata B Rao and Valerie Aurora) and Unionfs (Erez Zadok). They took +different approaches to implement the merged-view. +The former tries putting it into VFS, and the latter implements as a +separate filesystem. +(If I misunderstand about these implementations, please let me know and +I shall correct it. Because it is a long time ago when I read their +source files last time). + +UnionMount's approach will be able to small, but may be hard to share +branches between several UnionMount since the whiteout in it is +implemented in the inode on branch filesystem and always +shared. According to Bharata's post, readdir does not seems to be +finished yet. +There are several missing features known in this implementations such as +- for users, the inode number may change silently. eg. copy-up. +- link(2) may break by copy-up. +- read(2) may get an obsoleted filedata (fstat(2) too). +- fcntl(F_SETLK) may be broken by copy-up. +- unnecessary copy-up may happen, for example mmap(MAP_PRIVATE) after + open(O_RDWR). + +In linux-3.18, "overlay" filesystem (formerly known as "overlayfs") was +merged into mainline. This is another implementation of UnionMount as a +separated filesystem. All the limitations and known problems which +UnionMount are equally inherited to "overlay" filesystem. + +Unionfs has a longer history. When I started implementing a stackable +filesystem (Aug 2005), it already existed. It has virtual super_block, +inode, dentry and file objects and they have an array pointing lower +same kind objects. After contributing many patches for Unionfs, I +re-started my project AUFS (Jun 2006). + +In AUFS, the structure of filesystem resembles to Unionfs, but I +implemented my own ideas, approaches and enhancements and it became +totally different one. + +Comparing DM snapshot and fs based implementation +- the number of bytes to be copied between devices is much smaller. +- the type of filesystem must be one and only. +- the fs must be writable, no readonly fs, even for the lower original + device. so the compression fs will not be usable. but if we use + loopback mount, we may address this issue. + for instance, + mount /cdrom/squashfs.img /sq + losetup /sq/ext2.img + losetup /somewhere/cow + dmsetup "snapshot /dev/loop0 /dev/loop1 ..." +- it will be difficult (or needs more operations) to extract the + difference between the original device and COW. +- DM snapshot-merge may help a lot when users try merging. in the + fs-layer union, users will use rsync(1). + +You may want to read my old paper "Filesystems in LiveCD" +(http://aufs.sourceforge.net/aufs2/report/sq/sq.pdf). + + +Several characters/aspects/persona of aufs +---------------------------------------------------------------------- + +Aufs has several characters, aspects or persona. +1. a filesystem, callee of VFS helper +2. sub-VFS, caller of VFS helper for branches +3. a virtual filesystem which maintains persistent inode number +4. reader/writer of files on branches such like an application + +1. Callee of VFS Helper +As an ordinary linux filesystem, aufs is a callee of VFS. For instance, +unlink(2) from an application reaches sys_unlink() kernel function and +then vfs_unlink() is called. vfs_unlink() is one of VFS helper and it +calls filesystem specific unlink operation. Actually aufs implements the +unlink operation but it behaves like a redirector. + +2. Caller of VFS Helper for Branches +aufs_unlink() passes the unlink request to the branch filesystem as if +it were called from VFS. So the called unlink operation of the branch +filesystem acts as usual. As a caller of VFS helper, aufs should handle +every necessary pre/post operation for the branch filesystem. +- acquire the lock for the parent dir on a branch +- lookup in a branch +- revalidate dentry on a branch +- mnt_want_write() for a branch +- vfs_unlink() for a branch +- mnt_drop_write() for a branch +- release the lock on a branch + +3. Persistent Inode Number +One of the most important issue for a filesystem is to maintain inode +numbers. This is particularly important to support exporting a +filesystem via NFS. Aufs is a virtual filesystem which doesn't have a +backend block device for its own. But some storage is necessary to +keep and maintain the inode numbers. It may be a large space and may not +suit to keep in memory. Aufs rents some space from its first writable +branch filesystem (by default) and creates file(s) on it. These files +are created by aufs internally and removed soon (currently) keeping +opened. +Note: Because these files are removed, they are totally gone after + unmounting aufs. It means the inode numbers are not persistent + across unmount or reboot. I have a plan to make them really + persistent which will be important for aufs on NFS server. + +4. Read/Write Files Internally (copy-on-write) +Because a branch can be readonly, when you write a file on it, aufs will +"copy-up" it to the upper writable branch internally. And then write the +originally requested thing to the file. Generally kernel doesn't +open/read/write file actively. In aufs, even a single write may cause a +internal "file copy". This behaviour is very similar to cp(1) command. + +Some people may think it is better to pass such work to user space +helper, instead of doing in kernel space. Actually I am still thinking +about it. But currently I have implemented it in kernel space. diff --git a/Documentation/filesystems/aufs/design/02struct.txt b/Documentation/filesystems/aufs/design/02struct.txt new file mode 100644 index 000000000..783328a75 --- /dev/null +++ b/Documentation/filesystems/aufs/design/02struct.txt @@ -0,0 +1,245 @@ + +# Copyright (C) 2005-2016 Junjiro R. Okajima + +Basic Aufs Internal Structure + +Superblock/Inode/Dentry/File Objects +---------------------------------------------------------------------- +As like an ordinary filesystem, aufs has its own +superblock/inode/dentry/file objects. All these objects have a +dynamically allocated array and store the same kind of pointers to the +lower filesystem, branch. +For example, when you build a union with one readwrite branch and one +readonly, mounted /au, /rw and /ro respectively. +- /au = /rw + /ro +- /ro/fileA exists but /rw/fileA + +Aufs lookup operation finds /ro/fileA and gets dentry for that. These +pointers are stored in a aufs dentry. The array in aufs dentry will be, +- [0] = NULL (because /rw/fileA doesn't exist) +- [1] = /ro/fileA + +This style of an array is essentially same to the aufs +superblock/inode/dentry/file objects. + +Because aufs supports manipulating branches, ie. add/delete/change +branches dynamically, these objects has its own generation. When +branches are changed, the generation in aufs superblock is +incremented. And a generation in other object are compared when it is +accessed. When a generation in other objects are obsoleted, aufs +refreshes the internal array. + + +Superblock +---------------------------------------------------------------------- +Additionally aufs superblock has some data for policies to select one +among multiple writable branches, XIB files, pseudo-links and kobject. +See below in detail. +About the policies which supports copy-down a directory, see +wbr_policy.txt too. + + +Branch and XINO(External Inode Number Translation Table) +---------------------------------------------------------------------- +Every branch has its own xino (external inode number translation table) +file. The xino file is created and unlinked by aufs internally. When two +members of a union exist on the same filesystem, they share the single +xino file. +The struct of a xino file is simple, just a sequence of aufs inode +numbers which is indexed by the lower inode number. +In the above sample, assume the inode number of /ro/fileA is i111 and +aufs assigns the inode number i999 for fileA. Then aufs writes 999 as +4(8) bytes at 111 * 4(8) bytes offset in the xino file. + +When the inode numbers are not contiguous, the xino file will be sparse +which has a hole in it and doesn't consume as much disk space as it +might appear. If your branch filesystem consumes disk space for such +holes, then you should specify 'xino=' option at mounting aufs. + +Aufs has a mount option to free the disk blocks for such holes in XINO +files on tmpfs or ramdisk. But it is not so effective actually. If you +meet a problem of disk shortage due to XINO files, then you should try +"tmpfs-ino.patch" (and "vfs-ino.patch" too) in aufs4-standalone.git. +The patch localizes the assignment inumbers per tmpfs-mount and avoid +the holes in XINO files. + +Also a writable branch has three kinds of "whiteout bases". All these +are existed when the branch is joined to aufs, and their names are +whiteout-ed doubly, so that users will never see their names in aufs +hierarchy. +1. a regular file which will be hardlinked to all whiteouts. +2. a directory to store a pseudo-link. +3. a directory to store an "orphan"-ed file temporary. + +1. Whiteout Base + When you remove a file on a readonly branch, aufs handles it as a + logical deletion and creates a whiteout on the upper writable branch + as a hardlink of this file in order not to consume inode on the + writable branch. +2. Pseudo-link Dir + See below, Pseudo-link. +3. Step-Parent Dir + When "fileC" exists on the lower readonly branch only and it is + opened and removed with its parent dir, and then user writes + something into it, then aufs copies-up fileC to this + directory. Because there is no other dir to store fileC. After + creating a file under this dir, the file is unlinked. + +Because aufs supports manipulating branches, ie. add/delete/change +dynamically, a branch has its own id. When the branch order changes, +aufs finds the new index by searching the branch id. + + +Pseudo-link +---------------------------------------------------------------------- +Assume "fileA" exists on the lower readonly branch only and it is +hardlinked to "fileB" on the branch. When you write something to fileA, +aufs copies-up it to the upper writable branch. Additionally aufs +creates a hardlink under the Pseudo-link Directory of the writable +branch. The inode of a pseudo-link is kept in aufs super_block as a +simple list. If fileB is read after unlinking fileA, aufs returns +filedata from the pseudo-link instead of the lower readonly +branch. Because the pseudo-link is based upon the inode, to keep the +inode number by xino (see above) is essentially necessary. + +All the hardlinks under the Pseudo-link Directory of the writable branch +should be restored in a proper location later. Aufs provides a utility +to do this. The userspace helpers executed at remounting and unmounting +aufs by default. +During this utility is running, it puts aufs into the pseudo-link +maintenance mode. In this mode, only the process which began the +maintenance mode (and its child processes) is allowed to operate in +aufs. Some other processes which are not related to the pseudo-link will +be allowed to run too, but the rest have to return an error or wait +until the maintenance mode ends. If a process already acquires an inode +mutex (in VFS), it has to return an error. + + +XIB(external inode number bitmap) +---------------------------------------------------------------------- +Addition to the xino file per a branch, aufs has an external inode number +bitmap in a superblock object. It is also an internal file such like a +xino file. +It is a simple bitmap to mark whether the aufs inode number is in-use or +not. +To reduce the file I/O, aufs prepares a single memory page to cache xib. + +As well as XINO files, aufs has a feature to truncate/refresh XIB to +reduce the number of consumed disk blocks for these files. + + +Virtual or Vertical Dir, and Readdir in Userspace +---------------------------------------------------------------------- +In order to support multiple layers (branches), aufs readdir operation +constructs a virtual dir block on memory. For readdir, aufs calls +vfs_readdir() internally for each dir on branches, merges their entries +with eliminating the whiteout-ed ones, and sets it to file (dir) +object. So the file object has its entry list until it is closed. The +entry list will be updated when the file position is zero and becomes +obsoleted. This decision is made in aufs automatically. + +The dynamically allocated memory block for the name of entries has a +unit of 512 bytes (by default) and stores the names contiguously (no +padding). Another block for each entry is handled by kmem_cache too. +During building dir blocks, aufs creates hash list and judging whether +the entry is whiteouted by its upper branch or already listed. +The merged result is cached in the corresponding inode object and +maintained by a customizable life-time option. + +Some people may call it can be a security hole or invite DoS attack +since the opened and once readdir-ed dir (file object) holds its entry +list and becomes a pressure for system memory. But I'd say it is similar +to files under /proc or /sys. The virtual files in them also holds a +memory page (generally) while they are opened. When an idea to reduce +memory for them is introduced, it will be applied to aufs too. +For those who really hate this situation, I've developed readdir(3) +library which operates this merging in userspace. You just need to set +LD_PRELOAD environment variable, and aufs will not consume no memory in +kernel space for readdir(3). + + +Workqueue +---------------------------------------------------------------------- +Aufs sometimes requires privilege access to a branch. For instance, +in copy-up/down operation. When a user process is going to make changes +to a file which exists in the lower readonly branch only, and the mode +of one of ancestor directories may not be writable by a user +process. Here aufs copy-up the file with its ancestors and they may +require privilege to set its owner/group/mode/etc. +This is a typical case of a application character of aufs (see +Introduction). + +Aufs uses workqueue synchronously for this case. It creates its own +workqueue. The workqueue is a kernel thread and has privilege. Aufs +passes the request to call mkdir or write (for example), and wait for +its completion. This approach solves a problem of a signal handler +simply. +If aufs didn't adopt the workqueue and changed the privilege of the +process, then the process may receive the unexpected SIGXFSZ or other +signals. + +Also aufs uses the system global workqueue ("events" kernel thread) too +for asynchronous tasks, such like handling inotify/fsnotify, re-creating a +whiteout base and etc. This is unrelated to a privilege. +Most of aufs operation tries acquiring a rw_semaphore for aufs +superblock at the beginning, at the same time waits for the completion +of all queued asynchronous tasks. + + +Whiteout +---------------------------------------------------------------------- +The whiteout in aufs is very similar to Unionfs's. That is represented +by its filename. UnionMount takes an approach of a file mode, but I am +afraid several utilities (find(1) or something) will have to support it. + +Basically the whiteout represents "logical deletion" which stops aufs to +lookup further, but also it represents "dir is opaque" which also stop +further lookup. + +In aufs, rmdir(2) and rename(2) for dir uses whiteout alternatively. +In order to make several functions in a single systemcall to be +revertible, aufs adopts an approach to rename a directory to a temporary +unique whiteouted name. +For example, in rename(2) dir where the target dir already existed, aufs +renames the target dir to a temporary unique whiteouted name before the +actual rename on a branch, and then handles other actions (make it opaque, +update the attributes, etc). If an error happens in these actions, aufs +simply renames the whiteouted name back and returns an error. If all are +succeeded, aufs registers a function to remove the whiteouted unique +temporary name completely and asynchronously to the system global +workqueue. + + +Copy-up +---------------------------------------------------------------------- +It is a well-known feature or concept. +When user modifies a file on a readonly branch, aufs operate "copy-up" +internally and makes change to the new file on the upper writable branch. +When the trigger systemcall does not update the timestamps of the parent +dir, aufs reverts it after copy-up. + + +Move-down (aufs3.9 and later) +---------------------------------------------------------------------- +"Copy-up" is one of the essential feature in aufs. It copies a file from +the lower readonly branch to the upper writable branch when a user +changes something about the file. +"Move-down" is an opposite action of copy-up. Basically this action is +ran manually instead of automatically and internally. +For desgin and implementation, aufs has to consider these issues. +- whiteout for the file may exist on the lower branch. +- ancestor directories may not exist on the lower branch. +- diropq for the ancestor directories may exist on the upper branch. +- free space on the lower branch will reduce. +- another access to the file may happen during moving-down, including + UDBA (see "Revalidate Dentry and UDBA"). +- the file should not be hard-linked nor pseudo-linked. they should be + handled by auplink utility later. + +Sometimes users want to move-down a file from the upper writable branch +to the lower readonly or writable branch. For instance, +- the free space of the upper writable branch is going to run out. +- create a new intermediate branch between the upper and lower branch. +- etc. + +For this purpose, use "aumvdown" command in aufs-util.git. diff --git a/Documentation/filesystems/aufs/design/03atomic_open.txt b/Documentation/filesystems/aufs/design/03atomic_open.txt new file mode 100644 index 000000000..741ad6d66 --- /dev/null +++ b/Documentation/filesystems/aufs/design/03atomic_open.txt @@ -0,0 +1,72 @@ + +# Copyright (C) 2015-2016 Junjiro R. Okajima + +Support for a branch who has its ->atomic_open() +---------------------------------------------------------------------- +The filesystems who implement its ->atomic_open() are not majority. For +example NFSv4 does, and aufs should call NFSv4 ->atomic_open, +particularly for open(O_CREAT|O_EXCL, 0400) case. Other than +->atomic_open(), NFSv4 returns an error for this open(2). While I am not +sure whether all filesystems who have ->atomic_open() behave like this, +but NFSv4 surely returns the error. + +In order to support ->atomic_open() for aufs, there are a few +approaches. + +A. Introduce aufs_atomic_open() + - calls one of VFS:do_last(), lookup_open() or atomic_open() for + branch fs. +B. Introduce aufs_atomic_open() calling create, open and chmod. this is + an aufs user Pip Cet's approach + - calls aufs_create(), VFS finish_open() and notify_change(). + - pass fake-mode to finish_open(), and then correct the mode by + notify_change(). +C. Extend aufs_open() to call branch fs's ->atomic_open() + - no aufs_atomic_open(). + - aufs_lookup() registers the TID to an aufs internal object. + - aufs_create() does nothing when the matching TID is registered, but + registers the mode. + - aufs_open() calls branch fs's ->atomic_open() when the matching + TID is registered. +D. Extend aufs_open() to re-try branch fs's ->open() with superuser's + credential + - no aufs_atomic_open(). + - aufs_create() registers the TID to an internal object. this info + represents "this process created this file just now." + - when aufs gets EACCES from branch fs's ->open(), then confirm the + registered TID and re-try open() with superuser's credential. + +Pros and cons for each approach. + +A. + - straightforward but highly depends upon VFS internal. + - the atomic behavaiour is kept. + - some of parameters such as nameidata are hard to reproduce for + branch fs. + - large overhead. +B. + - easy to implement. + - the atomic behavaiour is lost. +C. + - the atomic behavaiour is kept. + - dirty and tricky. + - VFS checks whether the file is created correctly after calling + ->create(), which means this approach doesn't work. +D. + - easy to implement. + - the atomic behavaiour is lost. + - to open a file with superuser's credential and give it to a user + process is a bad idea, since the file object keeps the credential + in it. It may affect LSM or something. This approach doesn't work + either. + +The approach A is ideal, but it hard to implement. So here is a +variation of A, which is to be implemented. + +A-1. Introduce aufs_atomic_open() + - calls branch fs ->atomic_open() if exists. otherwise calls + vfs_create() and finish_open(). + - the demerit is that the several checks after branch fs + ->atomic_open() are lost. in the ordinary case, the checks are + done by VFS:do_last(), lookup_open() and atomic_open(). some can + be implemented in aufs, but not all I am afraid. diff --git a/Documentation/filesystems/aufs/design/03lookup.txt b/Documentation/filesystems/aufs/design/03lookup.txt new file mode 100644 index 000000000..5b6b000b5 --- /dev/null +++ b/Documentation/filesystems/aufs/design/03lookup.txt @@ -0,0 +1,100 @@ + +# Copyright (C) 2005-2016 Junjiro R. Okajima + +Lookup in a Branch +---------------------------------------------------------------------- +Since aufs has a character of sub-VFS (see Introduction), it operates +lookup for branches as VFS does. It may be a heavy work. But almost all +lookup operation in aufs is the simplest case, ie. lookup only an entry +directly connected to its parent. Digging down the directory hierarchy +is unnecessary. VFS has a function lookup_one_len() for that use, and +aufs calls it. + +When a branch is a remote filesystem, aufs basically relies upon its +->d_revalidate(), also aufs forces the hardest revalidate tests for +them. +For d_revalidate, aufs implements three levels of revalidate tests. See +"Revalidate Dentry and UDBA" in detail. + + +Test Only the Highest One for the Directory Permission (dirperm1 option) +---------------------------------------------------------------------- +Let's try case study. +- aufs has two branches, upper readwrite and lower readonly. + /au = /rw + /ro +- "dirA" exists under /ro, but /rw. and its mode is 0700. +- user invoked "chmod a+rx /au/dirA" +- the internal copy-up is activated and "/rw/dirA" is created and its + permission bits are set to world readable. +- then "/au/dirA" becomes world readable? + +In this case, /ro/dirA is still 0700 since it exists in readonly branch, +or it may be a natively readonly filesystem. If aufs respects the lower +branch, it should not respond readdir request from other users. But user +allowed it by chmod. Should really aufs rejects showing the entries +under /ro/dirA? + +To be honest, I don't have a good solution for this case. So aufs +implements 'dirperm1' and 'nodirperm1' mount options, and leave it to +users. +When dirperm1 is specified, aufs checks only the highest one for the +directory permission, and shows the entries. Otherwise, as usual, checks +every dir existing on all branches and rejects the request. + +As a side effect, dirperm1 option improves the performance of aufs +because the number of permission check is reduced when the number of +branch is many. + + +Revalidate Dentry and UDBA (User's Direct Branch Access) +---------------------------------------------------------------------- +Generally VFS helpers re-validate a dentry as a part of lookup. +0. digging down the directory hierarchy. +1. lock the parent dir by its i_mutex. +2. lookup the final (child) entry. +3. revalidate it. +4. call the actual operation (create, unlink, etc.) +5. unlock the parent dir + +If the filesystem implements its ->d_revalidate() (step 3), then it is +called. Actually aufs implements it and checks the dentry on a branch is +still valid. +But it is not enough. Because aufs has to release the lock for the +parent dir on a branch at the end of ->lookup() (step 2) and +->d_revalidate() (step 3) while the i_mutex of the aufs dir is still +held by VFS. +If the file on a branch is changed directly, eg. bypassing aufs, after +aufs released the lock, then the subsequent operation may cause +something unpleasant result. + +This situation is a result of VFS architecture, ->lookup() and +->d_revalidate() is separated. But I never say it is wrong. It is a good +design from VFS's point of view. It is just not suitable for sub-VFS +character in aufs. + +Aufs supports such case by three level of revalidation which is +selectable by user. +1. Simple Revalidate + Addition to the native flow in VFS's, confirm the child-parent + relationship on the branch just after locking the parent dir on the + branch in the "actual operation" (step 4). When this validation + fails, aufs returns EBUSY. ->d_revalidate() (step 3) in aufs still + checks the validation of the dentry on branches. +2. Monitor Changes Internally by Inotify/Fsnotify + Addition to above, in the "actual operation" (step 4) aufs re-lookup + the dentry on the branch, and returns EBUSY if it finds different + dentry. + Additionally, aufs sets the inotify/fsnotify watch for every dir on branches + during it is in cache. When the event is notified, aufs registers a + function to kernel 'events' thread by schedule_work(). And the + function sets some special status to the cached aufs dentry and inode + private data. If they are not cached, then aufs has nothing to + do. When the same file is accessed through aufs (step 0-3) later, + aufs will detect the status and refresh all necessary data. + In this mode, aufs has to ignore the event which is fired by aufs + itself. +3. No Extra Validation + This is the simplest test and doesn't add any additional revalidation + test, and skip the revalidation in step 4. It is useful and improves + aufs performance when system surely hide the aufs branches from user, + by over-mounting something (or another method). diff --git a/Documentation/filesystems/aufs/design/04branch.txt b/Documentation/filesystems/aufs/design/04branch.txt new file mode 100644 index 000000000..e68f4d3df --- /dev/null +++ b/Documentation/filesystems/aufs/design/04branch.txt @@ -0,0 +1,61 @@ + +# Copyright (C) 2005-2016 Junjiro R. Okajima + +Branch Manipulation + +Since aufs supports dynamic branch manipulation, ie. add/remove a branch +and changing its permission/attribute, there are a lot of works to do. + + +Add a Branch +---------------------------------------------------------------------- +o Confirm the adding dir exists outside of aufs, including loopback + mount, and its various attributes. +o Initialize the xino file and whiteout bases if necessary. + See struct.txt. + +o Check the owner/group/mode of the directory + When the owner/group/mode of the adding directory differs from the + existing branch, aufs issues a warning because it may impose a + security risk. + For example, when a upper writable branch has a world writable empty + top directory, a malicious user can create any files on the writable + branch directly, like copy-up and modify manually. If something like + /etc/{passwd,shadow} exists on the lower readonly branch but the upper + writable branch, and the writable branch is world-writable, then a + malicious guy may create /etc/passwd on the writable branch directly + and the infected file will be valid in aufs. + I am afraid it can be a security issue, but aufs can do nothing except + producing a warning. + + +Delete a Branch +---------------------------------------------------------------------- +o Confirm the deleting branch is not busy + To be general, there is one merit to adopt "remount" interface to + manipulate branches. It is to discard caches. At deleting a branch, + aufs checks the still cached (and connected) dentries and inodes. If + there are any, then they are all in-use. An inode without its + corresponding dentry can be alive alone (for example, inotify/fsnotify case). + + For the cached one, aufs checks whether the same named entry exists on + other branches. + If the cached one is a directory, because aufs provides a merged view + to users, as long as one dir is left on any branch aufs can show the + dir to users. In this case, the branch can be removed from aufs. + Otherwise aufs rejects deleting the branch. + + If any file on the deleting branch is opened by aufs, then aufs + rejects deleting. + + +Modify the Permission of a Branch +---------------------------------------------------------------------- +o Re-initialize or remove the xino file and whiteout bases if necessary. + See struct.txt. + +o rw --> ro: Confirm the modifying branch is not busy + Aufs rejects the request if any of these conditions are true. + - a file on the branch is mmap-ed. + - a regular file on the branch is opened for write and there is no + same named entry on the upper branch. diff --git a/Documentation/filesystems/aufs/design/05wbr_policy.txt b/Documentation/filesystems/aufs/design/05wbr_policy.txt new file mode 100644 index 000000000..1726d5d06 --- /dev/null +++ b/Documentation/filesystems/aufs/design/05wbr_policy.txt @@ -0,0 +1,51 @@ + +# Copyright (C) 2005-2016 Junjiro R. Okajima + +Policies to Select One among Multiple Writable Branches +---------------------------------------------------------------------- +When the number of writable branch is more than one, aufs has to decide +the target branch for file creation or copy-up. By default, the highest +writable branch which has the parent (or ancestor) dir of the target +file is chosen (top-down-parent policy). +By user's request, aufs implements some other policies to select the +writable branch, for file creation several policies, round-robin, +most-free-space, and other policies. For copy-up, top-down-parent, +bottom-up-parent, bottom-up and others. + +As expected, the round-robin policy selects the branch in circular. When +you have two writable branches and creates 10 new files, 5 files will be +created for each branch. mkdir(2) systemcall is an exception. When you +create 10 new directories, all will be created on the same branch. +And the most-free-space policy selects the one which has most free +space among the writable branches. The amount of free space will be +checked by aufs internally, and users can specify its time interval. + +The policies for copy-up is more simple, +top-down-parent is equivalent to the same named on in create policy, +bottom-up-parent selects the writable branch where the parent dir +exists and the nearest upper one from the copyup-source, +bottom-up selects the nearest upper writable branch from the +copyup-source, regardless the existence of the parent dir. + +There are some rules or exceptions to apply these policies. +- If there is a readonly branch above the policy-selected branch and + the parent dir is marked as opaque (a variation of whiteout), or the + target (creating) file is whiteout-ed on the upper readonly branch, + then the result of the policy is ignored and the target file will be + created on the nearest upper writable branch than the readonly branch. +- If there is a writable branch above the policy-selected branch and + the parent dir is marked as opaque or the target file is whiteouted + on the branch, then the result of the policy is ignored and the target + file will be created on the highest one among the upper writable + branches who has diropq or whiteout. In case of whiteout, aufs removes + it as usual. +- link(2) and rename(2) systemcalls are exceptions in every policy. + They try selecting the branch where the source exists as possible + since copyup a large file will take long time. If it can't be, + ie. the branch where the source exists is readonly, then they will + follow the copyup policy. +- There is an exception for rename(2) when the target exists. + If the rename target exists, aufs compares the index of the branches + where the source and the target exists and selects the higher + one. If the selected branch is readonly, then aufs follows the + copyup policy. diff --git a/Documentation/filesystems/aufs/design/06fhsm.txt b/Documentation/filesystems/aufs/design/06fhsm.txt new file mode 100644 index 000000000..84b46dc5b --- /dev/null +++ b/Documentation/filesystems/aufs/design/06fhsm.txt @@ -0,0 +1,105 @@ + +# Copyright (C) 2011-2016 Junjiro R. Okajima + +File-based Hierarchical Storage Management (FHSM) +---------------------------------------------------------------------- +Hierarchical Storage Management (or HSM) is a well-known feature in the +storage world. Aufs provides this feature as file-based with multiple +writable branches, based upon the principle of "Colder, the Lower". +Here the word "colder" means that the less used files, and "lower" means +that the position in the order of the stacked branches vertically. +These multiple writable branches are prioritized, ie. the topmost one +should be the fastest drive and be used heavily. + +o Characters in aufs FHSM story +- aufs itself and a new branch attribute. +- a new ioctl interface to move-down and to establish a connection with + the daemon ("move-down" is a converse of "copy-up"). +- userspace tool and daemon. + +The userspace daemon establishes a connection with aufs and waits for +the notification. The notified information is very similar to struct +statfs containing the number of consumed blocks and inodes. +When the consumed blocks/inodes of a branch exceeds the user-specified +upper watermark, the daemon activates its move-down process until the +consumed blocks/inodes reaches the user-specified lower watermark. + +The actual move-down is done by aufs based upon the request from +user-space since we need to maintain the inode number and the internal +pointer arrays in aufs. + +Currently aufs FHSM handles the regular files only. Additionally they +must not be hard-linked nor pseudo-linked. + + +o Cowork of aufs and the user-space daemon + During the userspace daemon established the connection, aufs sends a + small notification to it whenever aufs writes something into the + writable branch. But it may cost high since aufs issues statfs(2) + internally. So user can specify a new option to cache the + info. Actually the notification is controlled by these factors. + + the specified cache time. + + classified as "force" by aufs internally. + Until the specified time expires, aufs doesn't send the info + except the forced cases. When aufs decide forcing, the info is always + notified to userspace. + For example, the number of free inodes is generally large enough and + the shortage of it happens rarely. So aufs doesn't force the + notification when creating a new file, directory and others. This is + the typical case which aufs doesn't force. + When aufs writes the actual filedata and the files consumes any of new + blocks, the aufs forces notifying. + + +o Interfaces in aufs +- New branch attribute. + + fhsm + Specifies that the branch is managed by FHSM feature. In other word, + participant in the FHSM. + When nofhsm is set to the branch, it will not be the source/target + branch of the move-down operation. This attribute is set + independently from coo and moo attributes, and if you want full + FHSM, you should specify them as well. +- New mount option. + + fhsm_sec + Specifies a second to suppress many less important info to be + notified. +- New ioctl. + + AUFS_CTL_FHSM_FD + create a new file descriptor which userspace can read the notification + (a subset of struct statfs) from aufs. +- Module parameter 'brs' + It has to be set to 1. Otherwise the new mount option 'fhsm' will not + be set. +- mount helpers /sbin/mount.aufs and /sbin/umount.aufs + When there are two or more branches with fhsm attributes, + /sbin/mount.aufs invokes the user-space daemon and /sbin/umount.aufs + terminates it. As a result of remounting and branch-manipulation, the + number of branches with fhsm attribute can be one. In this case, + /sbin/mount.aufs will terminate the user-space daemon. + + +Finally the operation is done as these steps in kernel-space. +- make sure that, + + no one else is using the file. + + the file is not hard-linked. + + the file is not pseudo-linked. + + the file is a regular file. + + the parent dir is not opaqued. +- find the target writable branch. +- make sure the file is not whiteout-ed by the upper (than the target) + branch. +- make the parent dir on the target branch. +- mutex lock the inode on the branch. +- unlink the whiteout on the target branch (if exists). +- lookup and create the whiteout-ed temporary name on the target branch. +- copy the file as the whiteout-ed temporary name on the target branch. +- rename the whiteout-ed temporary name to the original name. +- unlink the file on the source branch. +- maintain the internal pointer array and the external inode number + table (XINO). +- maintain the timestamps and other attributes of the parent dir and the + file. + +And of course, in every step, an error may happen. So the operation +should restore the original file state after an error happens. diff --git a/Documentation/filesystems/aufs/design/06mmap.txt b/Documentation/filesystems/aufs/design/06mmap.txt new file mode 100644 index 000000000..991c0b1fa --- /dev/null +++ b/Documentation/filesystems/aufs/design/06mmap.txt @@ -0,0 +1,59 @@ + +# Copyright (C) 2005-2016 Junjiro R. Okajima + +mmap(2) -- File Memory Mapping +---------------------------------------------------------------------- +In aufs, the file-mapped pages are handled by a branch fs directly, no +interaction with aufs. It means aufs_mmap() calls the branch fs's +->mmap(). +This approach is simple and good, but there is one problem. +Under /proc, several entries show the mmapped files by its path (with +device and inode number), and the printed path will be the path on the +branch fs's instead of virtual aufs's. +This is not a problem in most cases, but some utilities lsof(1) (and its +user) may expect the path on aufs. + +To address this issue, aufs adds a new member called vm_prfile in struct +vm_area_struct (and struct vm_region). The original vm_file points to +the file on the branch fs in order to handle everything correctly as +usual. The new vm_prfile points to a virtual file in aufs, and the +show-functions in procfs refers to vm_prfile if it is set. +Also we need to maintain several other places where touching vm_file +such like +- fork()/clone() copies vma and the reference count of vm_file is + incremented. +- merging vma maintains the ref count too. + +This is not a good approach. It just fakes the printed path. But it +leaves all behaviour around f_mapping unchanged. This is surely an +advantage. +Actually aufs had adopted another complicated approach which calls +generic_file_mmap() and handles struct vm_operations_struct. In this +approach, aufs met a hard problem and I could not solve it without +switching the approach. + +There may be one more another approach which is +- bind-mount the branch-root onto the aufs-root internally +- grab the new vfsmount (ie. struct mount) +- lazy-umount the branch-root internally +- in open(2) the aufs-file, open the branch-file with the hidden + vfsmount (instead of the original branch's vfsmount) +- ideally this "bind-mount and lazy-umount" should be done atomically, + but it may be possible from userspace by the mount helper. + +Adding the internal hidden vfsmount and using it in opening a file, the +file path under /proc will be printed correctly. This approach looks +smarter, but is not possible I am afraid. +- aufs-root may be bind-mount later. when it happens, another hidden + vfsmount will be required. +- it is hard to get the chance to bind-mount and lazy-umount + + in kernel-space, FS can have vfsmount in open(2) via + file->f_path, and aufs can know its vfsmount. But several locks are + already acquired, and if aufs tries to bind-mount and lazy-umount + here, then it may cause a deadlock. + + in user-space, bind-mount doesn't invoke the mount helper. +- since /proc shows dev and ino, aufs has to give vma these info. it + means a new member vm_prinode will be necessary. this is essentially + equivalent to vm_prfile described above. + +I have to give up this "looks-smater" approach. diff --git a/Documentation/filesystems/aufs/design/06xattr.txt b/Documentation/filesystems/aufs/design/06xattr.txt new file mode 100644 index 000000000..7bfa94f7b --- /dev/null +++ b/Documentation/filesystems/aufs/design/06xattr.txt @@ -0,0 +1,81 @@ + +# Copyright (C) 2014-2016 Junjiro R. Okajima + +Listing XATTR/EA and getting the value +---------------------------------------------------------------------- +For the inode standard attributes (owner, group, timestamps, etc.), aufs +shows the values from the topmost existing file. This behaviour is good +for the non-dir entries since the bahaviour exactly matches the shown +information. But for the directories, aufs considers all the same named +entries on the lower branches. Which means, if one of the lower entry +rejects readdir call, then aufs returns an error even if the topmost +entry allows it. This behaviour is necessary to respect the branch fs's +security, but can make users confused since the user-visible standard +attributes don't match the behaviour. +To address this issue, aufs has a mount option called dirperm1 which +checks the permission for the topmost entry only, and ignores the lower +entry's permission. + +A similar issue can happen around XATTR. +getxattr(2) and listxattr(2) families behave as if dirperm1 option is +always set. Otherwise these very unpleasant situation would happen. +- listxattr(2) may return the duplicated entries. +- users may not be able to remove or reset the XATTR forever, + + +XATTR/EA support in the internal (copy,move)-(up,down) +---------------------------------------------------------------------- +Generally the extended attributes of inode are categorized as these. +- "security" for LSM and capability. +- "system" for posix ACL, 'acl' mount option is required for the branch + fs generally. +- "trusted" for userspace, CAP_SYS_ADMIN is required. +- "user" for userspace, 'user_xattr' mount option is required for the + branch fs generally. + +Moreover there are some other categories. Aufs handles these rather +unpopular categories as the ordinary ones, ie. there is no special +condition nor exception. + +In copy-up, the support for XATTR on the dst branch may differ from the +src branch. In this case, the copy-up operation will get an error and +the original user operation which triggered the copy-up will fail. It +can happen that even all copy-up will fail. +When both of src and dst branches support XATTR and if an error occurs +during copying XATTR, then the copy-up should fail obviously. That is a +good reason and aufs should return an error to userspace. But when only +the src branch support that XATTR, aufs should not return an error. +For example, the src branch supports ACL but the dst branch doesn't +because the dst branch may natively un-support it or temporary +un-support it due to "noacl" mount option. Of course, the dst branch fs +may NOT return an error even if the XATTR is not supported. It is +totally up to the branch fs. + +Anyway when the aufs internal copy-up gets an error from the dst branch +fs, then aufs tries removing the just copied entry and returns the error +to the userspace. The worst case of this situation will be all copy-up +will fail. + +For the copy-up operation, there two basic approaches. +- copy the specified XATTR only (by category above), and return the + error unconditionally if it happens. +- copy all XATTR, and ignore the error on the specified category only. + +In order to support XATTR and to implement the correct behaviour, aufs +chooses the latter approach and introduces some new branch attributes, +"icexsec", "icexsys", "icextr", "icexusr", and "icexoth". +They correspond to the XATTR namespaces (see above). Additionally, to be +convenient, "icex" is also provided which means all "icex*" attributes +are set (here the word "icex" stands for "ignore copy-error on XATTR"). + +The meaning of these attributes is to ignore the error from setting +XATTR on that branch. +Note that aufs tries copying all XATTR unconditionally, and ignores the +error from the dst branch according to the specified attributes. + +Some XATTR may have its default value. The default value may come from +the parent dir or the environment. If the default value is set at the +file creating-time, it will be overwritten by copy-up. +Some contradiction may happen I am afraid. +Do we need another attribute to stop copying XATTR? I am unsure. For +now, aufs implements the branch attributes to ignore the error. diff --git a/Documentation/filesystems/aufs/design/07export.txt b/Documentation/filesystems/aufs/design/07export.txt new file mode 100644 index 000000000..c23930b49 --- /dev/null +++ b/Documentation/filesystems/aufs/design/07export.txt @@ -0,0 +1,45 @@ + +# Copyright (C) 2005-2016 Junjiro R. Okajima + +Export Aufs via NFS +---------------------------------------------------------------------- +Here is an approach. +- like xino/xib, add a new file 'xigen' which stores aufs inode + generation. +- iget_locked(): initialize aufs inode generation for a new inode, and + store it in xigen file. +- destroy_inode(): increment aufs inode generation and store it in xigen + file. it is necessary even if it is not unlinked, because any data of + inode may be changed by UDBA. +- encode_fh(): for a root dir, simply return FILEID_ROOT. otherwise + build file handle by + + branch id (4 bytes) + + superblock generation (4 bytes) + + inode number (4 or 8 bytes) + + parent dir inode number (4 or 8 bytes) + + inode generation (4 bytes)) + + return value of exportfs_encode_fh() for the parent on a branch (4 + bytes) + + file handle for a branch (by exportfs_encode_fh()) +- fh_to_dentry(): + + find the index of a branch from its id in handle, and check it is + still exist in aufs. + + 1st level: get the inode number from handle and search it in cache. + + 2nd level: if not found in cache, get the parent inode number from + the handle and search it in cache. and then open the found parent + dir, find the matching inode number by vfs_readdir() and get its + name, and call lookup_one_len() for the target dentry. + + 3rd level: if the parent dir is not cached, call + exportfs_decode_fh() for a branch and get the parent on a branch, + build a pathname of it, convert it a pathname in aufs, call + path_lookup(). now aufs gets a parent dir dentry, then handle it as + the 2nd level. + + to open the dir, aufs needs struct vfsmount. aufs keeps vfsmount + for every branch, but not itself. to get this, (currently) aufs + searches in current->nsproxy->mnt_ns list. it may not be a good + idea, but I didn't get other approach. + + test the generation of the gotten inode. +- every inode operation: they may get EBUSY due to UDBA. in this case, + convert it into ESTALE for NFSD. +- readdir(): call lockdep_on/off() because filldir in NFSD calls + lookup_one_len(), vfs_getattr(), encode_fh() and others. diff --git a/Documentation/filesystems/aufs/design/08shwh.txt b/Documentation/filesystems/aufs/design/08shwh.txt new file mode 100644 index 000000000..ad58ebe15 --- /dev/null +++ b/Documentation/filesystems/aufs/design/08shwh.txt @@ -0,0 +1,39 @@ + +# Copyright (C) 2005-2016 Junjiro R. Okajima + +Show Whiteout Mode (shwh) +---------------------------------------------------------------------- +Generally aufs hides the name of whiteouts. But in some cases, to show +them is very useful for users. For instance, creating a new middle layer +(branch) by merging existing layers. + +(borrowing aufs1 HOW-TO from a user, Michael Towers) +When you have three branches, +- Bottom: 'system', squashfs (underlying base system), read-only +- Middle: 'mods', squashfs, read-only +- Top: 'overlay', ram (tmpfs), read-write + +The top layer is loaded at boot time and saved at shutdown, to preserve +the changes made to the system during the session. +When larger changes have been made, or smaller changes have accumulated, +the size of the saved top layer data grows. At this point, it would be +nice to be able to merge the two overlay branches ('mods' and 'overlay') +and rewrite the 'mods' squashfs, clearing the top layer and thus +restoring save and load speed. + +This merging is simplified by the use of another aufs mount, of just the +two overlay branches using the 'shwh' option. +# mount -t aufs -o ro,shwh,br:/livesys/overlay=ro+wh:/livesys/mods=rr+wh \ + aufs /livesys/merge_union + +A merged view of these two branches is then available at +/livesys/merge_union, and the new feature is that the whiteouts are +visible! +Note that in 'shwh' mode the aufs mount must be 'ro', which will disable +writing to all branches. Also the default mode for all branches is 'ro'. +It is now possible to save the combined contents of the two overlay +branches to a new squashfs, e.g.: +# mksquashfs /livesys/merge_union /path/to/newmods.squash + +This new squashfs archive can be stored on the boot device and the +initramfs will use it to replace the old one at the next boot. diff --git a/Documentation/filesystems/aufs/design/10dynop.txt b/Documentation/filesystems/aufs/design/10dynop.txt new file mode 100644 index 000000000..49afc5899 --- /dev/null +++ b/Documentation/filesystems/aufs/design/10dynop.txt @@ -0,0 +1,34 @@ + +# Copyright (C) 2010-2016 Junjiro R. Okajima + +Dynamically customizable FS operations +---------------------------------------------------------------------- +Generally FS operations (struct inode_operations, struct +address_space_operations, struct file_operations, etc.) are defined as +"static const", but it never means that FS have only one set of +operation. Some FS have multiple sets of them. For instance, ext2 has +three sets, one for XIP, for NOBH, and for normal. +Since aufs overrides and redirects these operations, sometimes aufs has +to change its behaviour according to the branch FS type. More importantly +VFS acts differently if a function (member in the struct) is set or +not. It means aufs should have several sets of operations and select one +among them according to the branch FS definition. + +In order to solve this problem and not to affect the behaviour of VFS, +aufs defines these operations dynamically. For instance, aufs defines +dummy direct_IO function for struct address_space_operations, but it may +not be set to the address_space_operations actually. When the branch FS +doesn't have it, aufs doesn't set it to its address_space_operations +while the function definition itself is still alive. So the behaviour +itself will not change, and it will return an error when direct_IO is +not set. + +The lifetime of these dynamically generated operation object is +maintained by aufs branch object. When the branch is removed from aufs, +the reference counter of the object is decremented. When it reaches +zero, the dynamically generated operation object will be freed. + +This approach is designed to support AIO (io_submit), Direct I/O and +XIP (DAX) mainly. +Currently this approach is applied to address_space_operations for +regular files only. diff --git a/Documentation/scheduler/sched-BFS.txt b/Documentation/scheduler/sched-BFS.txt new file mode 100644 index 000000000..77a1e37b7 --- /dev/null +++ b/Documentation/scheduler/sched-BFS.txt @@ -0,0 +1,357 @@ +BFS - The Brain Fuck Scheduler by Con Kolivas. + +Goals. + +The goal of the Brain Fuck Scheduler, referred to as BFS from here on, is to +completely do away with the complex designs of the past for the cpu process +scheduler and instead implement one that is very simple in basic design. +The main focus of BFS is to achieve excellent desktop interactivity and +responsiveness without heuristics and tuning knobs that are difficult to +understand, impossible to model and predict the effect of, and when tuned to +one workload cause massive detriment to another. + + +Design summary. + +BFS is best described as a single runqueue, O(n) lookup, earliest effective +virtual deadline first design, loosely based on EEVDF (earliest eligible virtual +deadline first) and my previous Staircase Deadline scheduler. Each component +shall be described in order to understand the significance of, and reasoning for +it. The codebase when the first stable version was released was approximately +9000 lines less code than the existing mainline linux kernel scheduler (in +2.6.31). This does not even take into account the removal of documentation and +the cgroups code that is not used. + +Design reasoning. + +The single runqueue refers to the queued but not running processes for the +entire system, regardless of the number of CPUs. The reason for going back to +a single runqueue design is that once multiple runqueues are introduced, +per-CPU or otherwise, there will be complex interactions as each runqueue will +be responsible for the scheduling latency and fairness of the tasks only on its +own runqueue, and to achieve fairness and low latency across multiple CPUs, any +advantage in throughput of having CPU local tasks causes other disadvantages. +This is due to requiring a very complex balancing system to at best achieve some +semblance of fairness across CPUs and can only maintain relatively low latency +for tasks bound to the same CPUs, not across them. To increase said fairness +and latency across CPUs, the advantage of local runqueue locking, which makes +for better scalability, is lost due to having to grab multiple locks. + +A significant feature of BFS is that all accounting is done purely based on CPU +used and nowhere is sleep time used in any way to determine entitlement or +interactivity. Interactivity "estimators" that use some kind of sleep/run +algorithm are doomed to fail to detect all interactive tasks, and to falsely tag +tasks that aren't interactive as being so. The reason for this is that it is +close to impossible to determine that when a task is sleeping, whether it is +doing it voluntarily, as in a userspace application waiting for input in the +form of a mouse click or otherwise, or involuntarily, because it is waiting for +another thread, process, I/O, kernel activity or whatever. Thus, such an +estimator will introduce corner cases, and more heuristics will be required to +cope with those corner cases, introducing more corner cases and failed +interactivity detection and so on. Interactivity in BFS is built into the design +by virtue of the fact that tasks that are waking up have not used up their quota +of CPU time, and have earlier effective deadlines, thereby making it very likely +they will preempt any CPU bound task of equivalent nice level. See below for +more information on the virtual deadline mechanism. Even if they do not preempt +a running task, because the rr interval is guaranteed to have a bound upper +limit on how long a task will wait for, it will be scheduled within a timeframe +that will not cause visible interface jitter. + + +Design details. + +Task insertion. + +BFS inserts tasks into each relevant queue as an O(1) insertion into a double +linked list. On insertion, *every* running queue is checked to see if the newly +queued task can run on any idle queue, or preempt the lowest running task on the +system. This is how the cross-CPU scheduling of BFS achieves significantly lower +latency per extra CPU the system has. In this case the lookup is, in the worst +case scenario, O(n) where n is the number of CPUs on the system. + +Data protection. + +BFS has one single lock protecting the process local data of every task in the +global queue. Thus every insertion, removal and modification of task data in the +global runqueue needs to grab the global lock. However, once a task is taken by +a CPU, the CPU has its own local data copy of the running process' accounting +information which only that CPU accesses and modifies (such as during a +timer tick) thus allowing the accounting data to be updated lockless. Once a +CPU has taken a task to run, it removes it from the global queue. Thus the +global queue only ever has, at most, + + (number of tasks requesting cpu time) - (number of logical CPUs) + 1 + +tasks in the global queue. This value is relevant for the time taken to look up +tasks during scheduling. This will increase if many tasks with CPU affinity set +in their policy to limit which CPUs they're allowed to run on if they outnumber +the number of CPUs. The +1 is because when rescheduling a task, the CPU's +currently running task is put back on the queue. Lookup will be described after +the virtual deadline mechanism is explained. + +Virtual deadline. + +The key to achieving low latency, scheduling fairness, and "nice level" +distribution in BFS is entirely in the virtual deadline mechanism. The related +tunable in BFS is the rr_interval, or "round robin interval". This is the +maximum time two SCHED_OTHER (or SCHED_NORMAL, the common scheduling policy) +tasks of the same nice level will be running for, or looking at it the other +way around, the longest duration two tasks of the same nice level will be +delayed for. When a task requests cpu time, it is given a quota (time_slice) +equal to the rr_interval and a virtual deadline. The virtual deadline is +offset from the current time in jiffies by this equation: + + jiffies + (prio_ratio * rr_interval) + +The prio_ratio is determined as a ratio compared to the baseline of nice -20 +and increases by 10% per nice level. The deadline is a virtual one only in that +no guarantee is placed that a task will actually be scheduled by this time, but +it is used to compare which task should go next. There are three components to +how a task is next chosen. First is time_slice expiration. If a task runs out +of its time_slice, it is descheduled, the time_slice is refilled, and the +deadline reset to that formula above. Second is sleep, where a task no longer +is requesting CPU for whatever reason. The time_slice and deadline are _not_ +adjusted in this case and are just carried over for when the task is next +scheduled. Third is preemption, and that is when a newly waking task is deemed +higher priority than a currently running task on any cpu by virtue of the fact +that it has an earlier virtual deadline than the currently running task. The +earlier deadline is the key to which task is next chosen for the first and +second cases. Once a task is descheduled, it is put back on the queue, and an +O(n) lookup of all queued-but-not-running tasks is done to determine which has +the earliest deadline and that task is chosen to receive CPU next. + +The CPU proportion of different nice tasks works out to be approximately the + + (prio_ratio difference)^2 + +The reason it is squared is that a task's deadline does not change while it is +running unless it runs out of time_slice. Thus, even if the time actually +passes the deadline of another task that is queued, it will not get CPU time +unless the current running task deschedules, and the time "base" (jiffies) is +constantly moving. + +Task lookup. + +BFS has 103 priority queues. 100 of these are dedicated to the static priority +of realtime tasks, and the remaining 3 are, in order of best to worst priority, +SCHED_ISO (isochronous), SCHED_NORMAL, and SCHED_IDLEPRIO (idle priority +scheduling). When a task of these priorities is queued, a bitmap of running +priorities is set showing which of these priorities has tasks waiting for CPU +time. When a CPU is made to reschedule, the lookup for the next task to get +CPU time is performed in the following way: + +First the bitmap is checked to see what static priority tasks are queued. If +any realtime priorities are found, the corresponding queue is checked and the +first task listed there is taken (provided CPU affinity is suitable) and lookup +is complete. If the priority corresponds to a SCHED_ISO task, they are also +taken in FIFO order (as they behave like SCHED_RR). If the priority corresponds +to either SCHED_NORMAL or SCHED_IDLEPRIO, then the lookup becomes O(n). At this +stage, every task in the runlist that corresponds to that priority is checked +to see which has the earliest set deadline, and (provided it has suitable CPU +affinity) it is taken off the runqueue and given the CPU. If a task has an +expired deadline, it is taken and the rest of the lookup aborted (as they are +chosen in FIFO order). + +Thus, the lookup is O(n) in the worst case only, where n is as described +earlier, as tasks may be chosen before the whole task list is looked over. + + +Scalability. + +The major limitations of BFS will be that of scalability, as the separate +runqueue designs will have less lock contention as the number of CPUs rises. +However they do not scale linearly even with separate runqueues as multiple +runqueues will need to be locked concurrently on such designs to be able to +achieve fair CPU balancing, to try and achieve some sort of nice-level fairness +across CPUs, and to achieve low enough latency for tasks on a busy CPU when +other CPUs would be more suited. BFS has the advantage that it requires no +balancing algorithm whatsoever, as balancing occurs by proxy simply because +all CPUs draw off the global runqueue, in priority and deadline order. Despite +the fact that scalability is _not_ the prime concern of BFS, it both shows very +good scalability to smaller numbers of CPUs and is likely a more scalable design +at these numbers of CPUs. + +It also has some very low overhead scalability features built into the design +when it has been deemed their overhead is so marginal that they're worth adding. +The first is the local copy of the running process' data to the CPU it's running +on to allow that data to be updated lockless where possible. Then there is +deference paid to the last CPU a task was running on, by trying that CPU first +when looking for an idle CPU to use the next time it's scheduled. Finally there +is the notion of "sticky" tasks that are flagged when they are involuntarily +descheduled, meaning they still want further CPU time. This sticky flag is +used to bias heavily against those tasks being scheduled on a different CPU +unless that CPU would be otherwise idle. When a cpu frequency governor is used +that scales with CPU load, such as ondemand, sticky tasks are not scheduled +on a different CPU at all, preferring instead to go idle. This means the CPU +they were bound to is more likely to increase its speed while the other CPU +will go idle, thus speeding up total task execution time and likely decreasing +power usage. This is the only scenario where BFS will allow a CPU to go idle +in preference to scheduling a task on the earliest available spare CPU. + +The real cost of migrating a task from one CPU to another is entirely dependant +on the cache footprint of the task, how cache intensive the task is, how long +it's been running on that CPU to take up the bulk of its cache, how big the CPU +cache is, how fast and how layered the CPU cache is, how fast a context switch +is... and so on. In other words, it's close to random in the real world where we +do more than just one sole workload. The only thing we can be sure of is that +it's not free. So BFS uses the principle that an idle CPU is a wasted CPU and +utilising idle CPUs is more important than cache locality, and cache locality +only plays a part after that. + +When choosing an idle CPU for a waking task, the cache locality is determined +according to where the task last ran and then idle CPUs are ranked from best +to worst to choose the most suitable idle CPU based on cache locality, NUMA +node locality and hyperthread sibling business. They are chosen in the +following preference (if idle): + +* Same core, idle or busy cache, idle threads +* Other core, same cache, idle or busy cache, idle threads. +* Same node, other CPU, idle cache, idle threads. +* Same node, other CPU, busy cache, idle threads. +* Same core, busy threads. +* Other core, same cache, busy threads. +* Same node, other CPU, busy threads. +* Other node, other CPU, idle cache, idle threads. +* Other node, other CPU, busy cache, idle threads. +* Other node, other CPU, busy threads. + +This shows the SMT or "hyperthread" awareness in the design as well which will +choose a real idle core first before a logical SMT sibling which already has +tasks on the physical CPU. + +Early benchmarking of BFS suggested scalability dropped off at the 16 CPU mark. +However this benchmarking was performed on an earlier design that was far less +scalable than the current one so it's hard to know how scalable it is in terms +of both CPUs (due to the global runqueue) and heavily loaded machines (due to +O(n) lookup) at this stage. Note that in terms of scalability, the number of +_logical_ CPUs matters, not the number of _physical_ CPUs. Thus, a dual (2x) +quad core (4X) hyperthreaded (2X) machine is effectively a 16X. Newer benchmark +results are very promising indeed. Benchmark contributions are most welcome. + + +Features + +As the initial prime target audience for BFS was the average desktop user, it +was designed to not need tweaking, tuning or have features set to obtain benefit +from it. Thus the number of knobs and features has been kept to an absolute +minimum and should not require extra user input for the vast majority of cases. +There are precisely 2 tunables, and 2 extra scheduling policies. The rr_interval +and iso_cpu tunables, and the SCHED_ISO and SCHED_IDLEPRIO policies. In addition +to this, BFS also uses sub-tick accounting. What BFS does _not_ now feature is +support for CGROUPS. The average user should neither need to know what these +are, nor should they need to be using them to have good desktop behaviour. + +There are two "scheduler" tunables, the round robin interval and the +interactive flag. These can be accessed in + + /proc/sys/kernel/rr_interval + /proc/sys/kernel/interactive + +rr_interval value + +The value is in milliseconds, and the default value is set to 6ms. Valid values +are from 1 to 1000. Decreasing the value will decrease latencies at the cost of +decreasing throughput, while increasing it will improve throughput, but at the +cost of worsening latencies. The accuracy of the rr interval is limited by HZ +resolution of the kernel configuration. Thus, the worst case latencies are +usually slightly higher than this actual value. BFS uses "dithering" to try and +minimise the effect the Hz limitation has. The default value of 6 is not an +arbitrary one. It is based on the fact that humans can detect jitter at +approximately 7ms, so aiming for much lower latencies is pointless under most +circumstances. It is worth noting this fact when comparing the latency +performance of BFS to other schedulers. Worst case latencies being higher than +7ms are far worse than average latencies not being in the microsecond range. +Experimentation has shown that rr intervals being increased up to 300 can +improve throughput but beyond that, scheduling noise from elsewhere prevents +further demonstrable throughput. + +interactive flag + +This is a simple boolean that can be set to 1 or 0, set to 1 by default. This +sacrifices some of the interactive performance by giving tasks a degree of +soft affinity for logical CPUs when it will lead to improved throughput, but +enabling it also sacrifices the completely deterministic nature with respect +to latency that BFS otherwise normally provides, and subsequently leads to +slightly higher latencies and a noticeably less interactive system. + + +Isochronous scheduling. + +Isochronous scheduling is a unique scheduling policy designed to provide +near-real-time performance to unprivileged (ie non-root) users without the +ability to starve the machine indefinitely. Isochronous tasks (which means +"same time") are set using, for example, the schedtool application like so: + + schedtool -I -e amarok + +This will start the audio application "amarok" as SCHED_ISO. How SCHED_ISO works +is that it has a priority level between true realtime tasks and SCHED_NORMAL +which would allow them to preempt all normal tasks, in a SCHED_RR fashion (ie, +if multiple SCHED_ISO tasks are running, they purely round robin at rr_interval +rate). However if ISO tasks run for more than a tunable finite amount of time, +they are then demoted back to SCHED_NORMAL scheduling. This finite amount of +time is the percentage of _total CPU_ available across the machine, configurable +as a percentage in the following "resource handling" tunable (as opposed to a +scheduler tunable): + + /proc/sys/kernel/iso_cpu + +and is set to 70% by default. It is calculated over a rolling 5 second average +Because it is the total CPU available, it means that on a multi CPU machine, it +is possible to have an ISO task running as realtime scheduling indefinitely on +just one CPU, as the other CPUs will be available. Setting this to 100 is the +equivalent of giving all users SCHED_RR access and setting it to 0 removes the +ability to run any pseudo-realtime tasks. + +A feature of BFS is that it detects when an application tries to obtain a +realtime policy (SCHED_RR or SCHED_FIFO) and the caller does not have the +appropriate privileges to use those policies. When it detects this, it will +give the task SCHED_ISO policy instead. Thus it is transparent to the user. +Because some applications constantly set their policy as well as their nice +level, there is potential for them to undo the override specified by the user +on the command line of setting the policy to SCHED_ISO. To counter this, once +a task has been set to SCHED_ISO policy, it needs superuser privileges to set +it back to SCHED_NORMAL. This will ensure the task remains ISO and all child +processes and threads will also inherit the ISO policy. + +Idleprio scheduling. + +Idleprio scheduling is a scheduling policy designed to give out CPU to a task +_only_ when the CPU would be otherwise idle. The idea behind this is to allow +ultra low priority tasks to be run in the background that have virtually no +effect on the foreground tasks. This is ideally suited to distributed computing +clients (like setiathome, folding, mprime etc) but can also be used to start +a video encode or so on without any slowdown of other tasks. To avoid this +policy from grabbing shared resources and holding them indefinitely, if it +detects a state where the task is waiting on I/O, the machine is about to +suspend to ram and so on, it will transiently schedule them as SCHED_NORMAL. As +per the Isochronous task management, once a task has been scheduled as IDLEPRIO, +it cannot be put back to SCHED_NORMAL without superuser privileges. Tasks can +be set to start as SCHED_IDLEPRIO with the schedtool command like so: + + schedtool -D -e ./mprime + +Subtick accounting. + +It is surprisingly difficult to get accurate CPU accounting, and in many cases, +the accounting is done by simply determining what is happening at the precise +moment a timer tick fires off. This becomes increasingly inaccurate as the +timer tick frequency (HZ) is lowered. It is possible to create an application +which uses almost 100% CPU, yet by being descheduled at the right time, records +zero CPU usage. While the main problem with this is that there are possible +security implications, it is also difficult to determine how much CPU a task +really does use. BFS tries to use the sub-tick accounting from the TSC clock, +where possible, to determine real CPU usage. This is not entirely reliable, but +is far more likely to produce accurate CPU usage data than the existing designs +and will not show tasks as consuming no CPU usage when they actually are. Thus, +the amount of CPU reported as being used by BFS will more accurately represent +how much CPU the task itself is using (as is shown for example by the 'time' +application), so the reported values may be quite different to other schedulers. +Values reported as the 'load' are more prone to problems with this design, but +per process values are closer to real usage. When comparing throughput of BFS +to other designs, it is important to compare the actual completed work in terms +of total wall clock time taken and total work done, rather than the reported +"cpu usage". + + +Con Kolivas <kernel@kolivas.org> Tue, 5 Apr 2011 diff --git a/Documentation/sysctl/kernel.txt b/Documentation/sysctl/kernel.txt index 87119dc9b..f77a2514e 100644 --- a/Documentation/sysctl/kernel.txt +++ b/Documentation/sysctl/kernel.txt @@ -39,6 +39,7 @@ show up in /proc/sys/kernel: - hung_task_timeout_secs - hung_task_warnings - kexec_load_disabled +- iso_cpu - kptr_restrict - kstack_depth_to_print [ X86 only ] - l2cr [ PPC only ] @@ -67,6 +68,7 @@ show up in /proc/sys/kernel: - randomize_va_space - real-root-dev ==> Documentation/initrd.txt - reboot-cmd [ SPARC only ] +- rr_interval - rtsig-max - rtsig-nr - sem @@ -396,6 +398,16 @@ kernel stack. ============================================================== +iso_cpu: (BFS CPU scheduler only). + +This sets the percentage cpu that the unprivileged SCHED_ISO tasks can +run effectively at realtime priority, averaged over a rolling five +seconds over the -whole- system, meaning all cpus. + +Set to 70 (percent) by default. + +============================================================== + l2cr: (PPC only) This flag controls the L2 cache of G3 processor boards. If @@ -750,6 +762,20 @@ rebooting. ??? ============================================================== +rr_interval: (BFS CPU scheduler only) + +This is the smallest duration that any cpu process scheduling unit +will run for. Increasing this value can increase throughput of cpu +bound tasks substantially but at the expense of increased latencies +overall. Conversely decreasing it will decrease average and maximum +latencies but at the expense of throughput. This value is in +milliseconds and the default value chosen depends on the number of +cpus available at scheduler initialisation with a minimum of 6. + +Valid values are from 1-1000. + +============================================================== + rtsig-max & rtsig-nr: The file rtsig-max can be used to tune the maximum number diff --git a/MAINTAINERS b/MAINTAINERS index ef7947adf..cd862e1df 100644 --- a/MAINTAINERS +++ b/MAINTAINERS @@ -228,13 +228,13 @@ F: kernel/sys_ni.c ABIT UGURU 1,2 HARDWARE MONITOR DRIVER M: Hans de Goede <hdegoede@redhat.com> -L: lm-sensors@lm-sensors.org +L: linux-hwmon@vger.kernel.org S: Maintained F: drivers/hwmon/abituguru.c ABIT UGURU 3 HARDWARE MONITOR DRIVER M: Alistair John Strachan <alistair@devzero.co.uk> -L: lm-sensors@lm-sensors.org +L: linux-hwmon@vger.kernel.org S: Maintained F: drivers/hwmon/abituguru3.c @@ -386,14 +386,14 @@ F: Documentation/devicetree/bindings/net/ieee802154/adf7242.txt ADM1025 HARDWARE MONITOR DRIVER M: Jean Delvare <jdelvare@suse.com> -L: lm-sensors@lm-sensors.org +L: linux-hwmon@vger.kernel.org S: Maintained F: Documentation/hwmon/adm1025 F: drivers/hwmon/adm1025.c ADM1029 HARDWARE MONITOR DRIVER M: Corentin Labbe <clabbe.montjoie@gmail.com> -L: lm-sensors@lm-sensors.org +L: linux-hwmon@vger.kernel.org S: Maintained F: drivers/hwmon/adm1029.c @@ -438,7 +438,7 @@ F: drivers/video/backlight/adp8860_bl.c ADS1015 HARDWARE MONITOR DRIVER M: Dirk Eibach <eibach@gdsys.de> -L: lm-sensors@lm-sensors.org +L: linux-hwmon@vger.kernel.org S: Maintained F: Documentation/hwmon/ads1015 F: drivers/hwmon/ads1015.c @@ -451,7 +451,7 @@ F: drivers/macintosh/therm_adt746x.c ADT7475 HARDWARE MONITOR DRIVER M: Jean Delvare <jdelvare@suse.com> -L: lm-sensors@lm-sensors.org +L: linux-hwmon@vger.kernel.org S: Maintained F: Documentation/hwmon/adt7475 F: drivers/hwmon/adt7475.c @@ -628,7 +628,7 @@ F: include/linux/ccp.h AMD FAM15H PROCESSOR POWER MONITORING DRIVER M: Huang Rui <ray.huang@amd.com> -L: lm-sensors@lm-sensors.org +L: linux-hwmon@vger.kernel.org S: Supported F: Documentation/hwmon/fam15h_power F: drivers/hwmon/fam15h_power.c @@ -786,7 +786,7 @@ F: drivers/input/mouse/bcm5974.c APPLE SMC DRIVER M: Henrik Rydberg <rydberg@bitmath.org> -L: lm-sensors@lm-sensors.org +L: linux-hwmon@vger.kernel.org S: Odd fixes F: drivers/hwmon/applesmc.c @@ -1825,7 +1825,7 @@ F: include/media/i2c/as3645a.h ASC7621 HARDWARE MONITOR DRIVER M: George Joseph <george.joseph@fairview5.com> -L: lm-sensors@lm-sensors.org +L: linux-hwmon@vger.kernel.org S: Maintained F: Documentation/hwmon/asc7621 F: drivers/hwmon/asc7621.c @@ -1918,7 +1918,7 @@ F: drivers/net/wireless/ath/carl9170/ ATK0110 HWMON DRIVER M: Luca Tettamanti <kronos.it@gmail.com> -L: lm-sensors@lm-sensors.org +L: linux-hwmon@vger.kernel.org S: Maintained F: drivers/hwmon/asus_atk0110.c @@ -2082,6 +2082,19 @@ F: include/linux/audit.h F: include/uapi/linux/audit.h F: kernel/audit* +AUFS (advanced multi layered unification filesystem) FILESYSTEM +M: "J. R. Okajima" <hooanon05g@gmail.com> +L: linux-unionfs@vger.kernel.org +L: aufs-users@lists.sourceforge.net (members only) +W: http://aufs.sourceforge.net +T: git://github.com/sfjro/aufs4-linux.git +S: Supported +F: Documentation/filesystems/aufs/ +F: Documentation/ABI/testing/debugfs-aufs +F: Documentation/ABI/testing/sysfs-aufs +F: fs/aufs/ +F: include/uapi/linux/aufs_type.h + AUXILIARY DISPLAY DRIVERS M: Miguel Ojeda Sandonis <miguel.ojeda.sandonis@gmail.com> W: http://miguelojeda.es/auxdisplay.htm @@ -3037,7 +3050,7 @@ F: mm/swap_cgroup.c CORETEMP HARDWARE MONITORING DRIVER M: Fenghua Yu <fenghua.yu@intel.com> -L: lm-sensors@lm-sensors.org +L: linux-hwmon@vger.kernel.org S: Maintained F: Documentation/hwmon/coretemp F: drivers/hwmon/coretemp.c @@ -3625,7 +3638,7 @@ T: git git://git.infradead.org/users/vkoul/slave-dma.git DME1737 HARDWARE MONITOR DRIVER M: Juerg Haefliger <juergh@gmail.com> -L: lm-sensors@lm-sensors.org +L: linux-hwmon@vger.kernel.org S: Maintained F: Documentation/hwmon/dme1737 F: drivers/hwmon/dme1737.c @@ -4322,7 +4335,7 @@ F: include/video/exynos_mipi* F71805F HARDWARE MONITORING DRIVER M: Jean Delvare <jdelvare@suse.com> -L: lm-sensors@lm-sensors.org +L: linux-hwmon@vger.kernel.org S: Maintained F: Documentation/hwmon/f71805f F: drivers/hwmon/f71805f.c @@ -4401,7 +4414,7 @@ F: fs/* FINTEK F75375S HARDWARE MONITOR AND FAN CONTROLLER DRIVER M: Riku Voipio <riku.voipio@iki.fi> -L: lm-sensors@lm-sensors.org +L: linux-hwmon@vger.kernel.org S: Maintained F: drivers/hwmon/f75375s.c F: include/linux/f75375s.h @@ -4958,8 +4971,8 @@ F: drivers/media/usb/hackrf/ HARDWARE MONITORING M: Jean Delvare <jdelvare@suse.com> M: Guenter Roeck <linux@roeck-us.net> -L: lm-sensors@lm-sensors.org -W: http://www.lm-sensors.org/ +L: linux-hwmon@vger.kernel.org +W: http://hwmon.wiki.kernel.org/ T: quilt http://jdelvare.nerim.net/devel/linux/jdelvare-hwmon/ T: git git://git.kernel.org/pub/scm/linux/kernel/git/groeck/linux-staging.git S: Maintained @@ -5484,7 +5497,7 @@ F: drivers/usb/atm/ueagle-atm.c INA209 HARDWARE MONITOR DRIVER M: Guenter Roeck <linux@roeck-us.net> -L: lm-sensors@lm-sensors.org +L: linux-hwmon@vger.kernel.org S: Maintained F: Documentation/hwmon/ina209 F: Documentation/devicetree/bindings/i2c/ina209.txt @@ -5492,7 +5505,7 @@ F: drivers/hwmon/ina209.c INA2XX HARDWARE MONITOR DRIVER M: Guenter Roeck <linux@roeck-us.net> -L: lm-sensors@lm-sensors.org +L: linux-hwmon@vger.kernel.org S: Maintained F: Documentation/hwmon/ina2xx F: drivers/hwmon/ina2xx.c @@ -5985,7 +5998,7 @@ F: drivers/isdn/hardware/eicon/ IT87 HARDWARE MONITORING DRIVER M: Jean Delvare <jdelvare@suse.com> -L: lm-sensors@lm-sensors.org +L: linux-hwmon@vger.kernel.org S: Maintained F: Documentation/hwmon/it87 F: drivers/hwmon/it87.c @@ -6021,7 +6034,7 @@ F: drivers/media/dvb-frontends/ix2505v* JC42.4 TEMPERATURE SENSOR DRIVER M: Guenter Roeck <linux@roeck-us.net> -L: lm-sensors@lm-sensors.org +L: linux-hwmon@vger.kernel.org S: Maintained F: drivers/hwmon/jc42.c F: Documentation/hwmon/jc42 @@ -6071,14 +6084,14 @@ F: drivers/tty/serial/jsm/ K10TEMP HARDWARE MONITORING DRIVER M: Clemens Ladisch <clemens@ladisch.de> -L: lm-sensors@lm-sensors.org +L: linux-hwmon@vger.kernel.org S: Maintained F: Documentation/hwmon/k10temp F: drivers/hwmon/k10temp.c K8TEMP HARDWARE MONITORING DRIVER M: Rudolf Marek <r.marek@assembler.cz> -L: lm-sensors@lm-sensors.org +L: linux-hwmon@vger.kernel.org S: Maintained F: Documentation/hwmon/k8temp F: drivers/hwmon/k8temp.c @@ -6605,27 +6618,27 @@ F: net/llc/ LM73 HARDWARE MONITOR DRIVER M: Guillaume Ligneul <guillaume.ligneul@gmail.com> -L: lm-sensors@lm-sensors.org +L: linux-hwmon@vger.kernel.org S: Maintained F: drivers/hwmon/lm73.c LM78 HARDWARE MONITOR DRIVER M: Jean Delvare <jdelvare@suse.com> -L: lm-sensors@lm-sensors.org +L: linux-hwmon@vger.kernel.org S: Maintained F: Documentation/hwmon/lm78 F: drivers/hwmon/lm78.c LM83 HARDWARE MONITOR DRIVER M: Jean Delvare <jdelvare@suse.com> -L: lm-sensors@lm-sensors.org +L: linux-hwmon@vger.kernel.org S: Maintained F: Documentation/hwmon/lm83 F: drivers/hwmon/lm83.c LM90 HARDWARE MONITOR DRIVER M: Jean Delvare <jdelvare@suse.com> -L: lm-sensors@lm-sensors.org +L: linux-hwmon@vger.kernel.org S: Maintained F: Documentation/hwmon/lm90 F: Documentation/devicetree/bindings/hwmon/lm90.txt @@ -6633,7 +6646,7 @@ F: drivers/hwmon/lm90.c LM95234 HARDWARE MONITOR DRIVER M: Guenter Roeck <linux@roeck-us.net> -L: lm-sensors@lm-sensors.org +L: linux-hwmon@vger.kernel.org S: Maintained F: Documentation/hwmon/lm95234 F: drivers/hwmon/lm95234.c @@ -6700,7 +6713,7 @@ F: drivers/scsi/sym53c8xx_2/ LTC4261 HARDWARE MONITOR DRIVER M: Guenter Roeck <linux@roeck-us.net> -L: lm-sensors@lm-sensors.org +L: linux-hwmon@vger.kernel.org S: Maintained F: Documentation/hwmon/ltc4261 F: drivers/hwmon/ltc4261.c @@ -6870,28 +6883,28 @@ F: include/uapi/linux/matroxfb.h MAX16065 HARDWARE MONITOR DRIVER M: Guenter Roeck <linux@roeck-us.net> -L: lm-sensors@lm-sensors.org +L: linux-hwmon@vger.kernel.org S: Maintained F: Documentation/hwmon/max16065 F: drivers/hwmon/max16065.c MAX20751 HARDWARE MONITOR DRIVER M: Guenter Roeck <linux@roeck-us.net> -L: lm-sensors@lm-sensors.org +L: linux-hwmon@vger.kernel.org S: Maintained F: Documentation/hwmon/max20751 F: drivers/hwmon/max20751.c MAX6650 HARDWARE MONITOR AND FAN CONTROLLER DRIVER M: "Hans J. Koch" <hjk@hansjkoch.de> -L: lm-sensors@lm-sensors.org +L: linux-hwmon@vger.kernel.org S: Maintained F: Documentation/hwmon/max6650 F: drivers/hwmon/max6650.c MAX6697 HARDWARE MONITOR DRIVER M: Guenter Roeck <linux@roeck-us.net> -L: lm-sensors@lm-sensors.org +L: linux-hwmon@vger.kernel.org S: Maintained F: Documentation/hwmon/max6697 F: Documentation/devicetree/bindings/i2c/max6697.txt @@ -7455,7 +7468,7 @@ F: drivers/scsi/NCR_D700.* NCT6775 HARDWARE MONITOR DRIVER M: Guenter Roeck <linux@roeck-us.net> -L: lm-sensors@lm-sensors.org +L: linux-hwmon@vger.kernel.org S: Maintained F: Documentation/hwmon/nct6775 F: drivers/hwmon/nct6775.c @@ -8235,7 +8248,7 @@ F: drivers/video/logo/logo_parisc* PC87360 HARDWARE MONITORING DRIVER M: Jim Cromie <jim.cromie@gmail.com> -L: lm-sensors@lm-sensors.org +L: linux-hwmon@vger.kernel.org S: Maintained F: Documentation/hwmon/pc87360 F: drivers/hwmon/pc87360.c @@ -8247,7 +8260,7 @@ F: drivers/char/pc8736x_gpio.c PC87427 HARDWARE MONITORING DRIVER M: Jean Delvare <jdelvare@suse.com> -L: lm-sensors@lm-sensors.org +L: linux-hwmon@vger.kernel.org S: Maintained F: Documentation/hwmon/pc87427 F: drivers/hwmon/pc87427.c @@ -8601,8 +8614,8 @@ F: drivers/rtc/rtc-puv3.c PMBUS HARDWARE MONITORING DRIVERS M: Guenter Roeck <linux@roeck-us.net> -L: lm-sensors@lm-sensors.org -W: http://www.lm-sensors.org/ +L: linux-hwmon@vger.kernel.org +W: http://hwmon.wiki.kernel.org/ W: http://www.roeck-us.net/linux/drivers/ T: git git://git.kernel.org/pub/scm/linux/kernel/git/groeck/linux-staging.git S: Maintained @@ -8807,7 +8820,7 @@ F: drivers/media/usb/pwc/* PWM FAN DRIVER M: Kamil Debski <k.debski@samsung.com> -L: lm-sensors@lm-sensors.org +L: linux-hwmon@vger.kernel.org S: Supported F: Documentation/devicetree/bindings/hwmon/pwm-fan.txt F: Documentation/hwmon/pwm-fan @@ -10113,28 +10126,28 @@ F: Documentation/devicetree/bindings/media/i2c/nokia,smia.txt SMM665 HARDWARE MONITOR DRIVER M: Guenter Roeck <linux@roeck-us.net> -L: lm-sensors@lm-sensors.org +L: linux-hwmon@vger.kernel.org S: Maintained F: Documentation/hwmon/smm665 F: drivers/hwmon/smm665.c SMSC EMC2103 HARDWARE MONITOR DRIVER M: Steve Glendinning <steve.glendinning@shawell.net> -L: lm-sensors@lm-sensors.org +L: linux-hwmon@vger.kernel.org S: Maintained F: Documentation/hwmon/emc2103 F: drivers/hwmon/emc2103.c SMSC SCH5627 HARDWARE MONITOR DRIVER M: Hans de Goede <hdegoede@redhat.com> -L: lm-sensors@lm-sensors.org +L: linux-hwmon@vger.kernel.org S: Supported F: Documentation/hwmon/sch5627 F: drivers/hwmon/sch5627.c SMSC47B397 HARDWARE MONITOR DRIVER M: Jean Delvare <jdelvare@suse.com> -L: lm-sensors@lm-sensors.org +L: linux-hwmon@vger.kernel.org S: Maintained F: Documentation/hwmon/smsc47b397 F: drivers/hwmon/smsc47b397.c @@ -11067,7 +11080,7 @@ F: include/linux/mmc/sh_mobile_sdhi.h TMP401 HARDWARE MONITOR DRIVER M: Guenter Roeck <linux@roeck-us.net> -L: lm-sensors@lm-sensors.org +L: linux-hwmon@vger.kernel.org S: Maintained F: Documentation/hwmon/tmp401 F: drivers/hwmon/tmp401.c @@ -11819,14 +11832,14 @@ F: Documentation/networking/vrf.txt VT1211 HARDWARE MONITOR DRIVER M: Juerg Haefliger <juergh@gmail.com> -L: lm-sensors@lm-sensors.org +L: linux-hwmon@vger.kernel.org S: Maintained F: Documentation/hwmon/vt1211 F: drivers/hwmon/vt1211.c VT8231 HARDWARE MONITOR DRIVER M: Roger Lucas <vt8231@hiddenengine.co.uk> -L: lm-sensors@lm-sensors.org +L: linux-hwmon@vger.kernel.org S: Maintained F: drivers/hwmon/vt8231.c @@ -11845,21 +11858,21 @@ F: drivers/w1/ W83791D HARDWARE MONITORING DRIVER M: Marc Hulsman <m.hulsman@tudelft.nl> -L: lm-sensors@lm-sensors.org +L: linux-hwmon@vger.kernel.org S: Maintained F: Documentation/hwmon/w83791d F: drivers/hwmon/w83791d.c W83793 HARDWARE MONITORING DRIVER M: Rudolf Marek <r.marek@assembler.cz> -L: lm-sensors@lm-sensors.org +L: linux-hwmon@vger.kernel.org S: Maintained F: Documentation/hwmon/w83793 F: drivers/hwmon/w83793.c W83795 HARDWARE MONITORING DRIVER M: Jean Delvare <jdelvare@suse.com> -L: lm-sensors@lm-sensors.org +L: linux-hwmon@vger.kernel.org S: Maintained F: drivers/hwmon/w83795.c @@ -1,6 +1,6 @@ VERSION = 4 PATCHLEVEL = 5 -SUBLEVEL = 0 +SUBLEVEL = 1 EXTRAVERSION = -gnu NAME = Blurry Fish Butt diff --git a/arch/arc/boot/dts/axs10x_mb.dtsi b/arch/arc/boot/dts/axs10x_mb.dtsi index 44a578c10..ab5d5701e 100644 --- a/arch/arc/boot/dts/axs10x_mb.dtsi +++ b/arch/arc/boot/dts/axs10x_mb.dtsi @@ -47,6 +47,14 @@ clocks = <&apbclk>; clock-names = "stmmaceth"; max-speed = <100>; + mdio0 { + #address-cells = <1>; + #size-cells = <0>; + compatible = "snps,dwmac-mdio"; + phy1: ethernet-phy@1 { + reg = <1>; + }; + }; }; ehci@0x40000 { diff --git a/arch/arc/include/asm/bitops.h b/arch/arc/include/asm/bitops.h index 57c1f3384..0352fb8d2 100644 --- a/arch/arc/include/asm/bitops.h +++ b/arch/arc/include/asm/bitops.h @@ -35,21 +35,6 @@ static inline void op##_bit(unsigned long nr, volatile unsigned long *m)\ \ m += nr >> 5; \ \ - /* \ - * ARC ISA micro-optimization: \ - * \ - * Instructions dealing with bitpos only consider lower 5 bits \ - * e.g (x << 33) is handled like (x << 1) by ASL instruction \ - * (mem pointer still needs adjustment to point to next word) \ - * \ - * Hence the masking to clamp @nr arg can be elided in general. \ - * \ - * However if @nr is a constant (above assumed in a register), \ - * and greater than 31, gcc can optimize away (x << 33) to 0, \ - * as overflow, given the 32-bit ISA. Thus masking needs to be \ - * done for const @nr, but no code is generated due to gcc \ - * const prop. \ - */ \ nr &= 0x1f; \ \ __asm__ __volatile__( \ diff --git a/arch/arc/include/asm/io.h b/arch/arc/include/asm/io.h index 694ece8a0..27b17adea 100644 --- a/arch/arc/include/asm/io.h +++ b/arch/arc/include/asm/io.h @@ -129,15 +129,23 @@ static inline void __raw_writel(u32 w, volatile void __iomem *addr) #define writel(v,c) ({ __iowmb(); writel_relaxed(v,c); }) /* - * Relaxed API for drivers which can handle any ordering themselves + * Relaxed API for drivers which can handle barrier ordering themselves + * + * Also these are defined to perform little endian accesses. + * To provide the typical device register semantics of fixed endian, + * swap the byte order for Big Endian + * + * http://lkml.kernel.org/r/201603100845.30602.arnd@arndb.de */ #define readb_relaxed(c) __raw_readb(c) -#define readw_relaxed(c) __raw_readw(c) -#define readl_relaxed(c) __raw_readl(c) +#define readw_relaxed(c) ({ u16 __r = le16_to_cpu((__force __le16) \ + __raw_readw(c)); __r; }) +#define readl_relaxed(c) ({ u32 __r = le32_to_cpu((__force __le32) \ + __raw_readl(c)); __r; }) #define writeb_relaxed(v,c) __raw_writeb(v,c) -#define writew_relaxed(v,c) __raw_writew(v,c) -#define writel_relaxed(v,c) __raw_writel(v,c) +#define writew_relaxed(v,c) __raw_writew((__force u16) cpu_to_le16(v),c) +#define writel_relaxed(v,c) __raw_writel((__force u32) cpu_to_le32(v),c) #include <asm-generic/io.h> diff --git a/arch/arm/boot/dts/at91-sama5d3_xplained.dts b/arch/arm/boot/dts/at91-sama5d3_xplained.dts index ff888d21c..f3e2b96c0 100644 --- a/arch/arm/boot/dts/at91-sama5d3_xplained.dts +++ b/arch/arm/boot/dts/at91-sama5d3_xplained.dts @@ -303,6 +303,7 @@ regulator-name = "mmc0-card-supply"; regulator-min-microvolt = <3300000>; regulator-max-microvolt = <3300000>; + regulator-always-on; }; gpio_keys { diff --git a/arch/arm/boot/dts/at91-sama5d4_xplained.dts b/arch/arm/boot/dts/at91-sama5d4_xplained.dts index 569026e8f..da84e65b5 100644 --- a/arch/arm/boot/dts/at91-sama5d4_xplained.dts +++ b/arch/arm/boot/dts/at91-sama5d4_xplained.dts @@ -268,5 +268,6 @@ regulator-min-microvolt = <3300000>; regulator-max-microvolt = <3300000>; vin-supply = <&vcc_3v3_reg>; + regulator-always-on; }; }; diff --git a/arch/arm64/include/asm/pgtable.h b/arch/arm64/include/asm/pgtable.h index 819aff5d5..727321078 100644 --- a/arch/arm64/include/asm/pgtable.h +++ b/arch/arm64/include/asm/pgtable.h @@ -279,7 +279,7 @@ extern void __sync_icache_dcache(pte_t pteval, unsigned long addr); static inline void set_pte_at(struct mm_struct *mm, unsigned long addr, pte_t *ptep, pte_t pte) { - if (pte_valid(pte)) { + if (pte_present(pte)) { if (pte_sw_dirty(pte) && pte_write(pte)) pte_val(pte) &= ~PTE_RDONLY; else @@ -649,6 +649,7 @@ extern pgd_t idmap_pg_dir[PTRS_PER_PGD]; * bits 0-1: present (must be zero) * bits 2-7: swap type * bits 8-57: swap offset + * bit 58: PTE_PROT_NONE (must be zero) */ #define __SWP_TYPE_SHIFT 2 #define __SWP_TYPE_BITS 6 diff --git a/arch/ia64/include/asm/io.h b/arch/ia64/include/asm/io.h index a865d2a04..5de673ac9 100644 --- a/arch/ia64/include/asm/io.h +++ b/arch/ia64/include/asm/io.h @@ -433,6 +433,7 @@ static inline void __iomem * ioremap_cache (unsigned long phys_addr, unsigned lo return ioremap(phys_addr, size); } #define ioremap_cache ioremap_cache +#define ioremap_uc ioremap_nocache /* diff --git a/arch/powerpc/platforms/cell/spufs/sched.c b/arch/powerpc/platforms/cell/spufs/sched.c index 998f632e7..5c80a0f9d 100644 --- a/arch/powerpc/platforms/cell/spufs/sched.c +++ b/arch/powerpc/platforms/cell/spufs/sched.c @@ -64,11 +64,6 @@ static struct timer_list spusched_timer; static struct timer_list spuloadavg_timer; /* - * Priority of a normal, non-rt, non-niced'd process (aka nice level 0). - */ -#define NORMAL_PRIO 120 - -/* * Frequency of the spu scheduler tick. By default we do one SPU scheduler * tick for every 10 CPU scheduler ticks. */ diff --git a/arch/s390/include/asm/pci.h b/arch/s390/include/asm/pci.h index c873e682b..2b2ced9dc 100644 --- a/arch/s390/include/asm/pci.h +++ b/arch/s390/include/asm/pci.h @@ -45,7 +45,7 @@ struct zpci_fmb { u64 rpcit_ops; u64 dma_rbytes; u64 dma_wbytes; -} __packed __aligned(16); +} __packed __aligned(64); enum zpci_state { ZPCI_FN_STATE_RESERVED, diff --git a/arch/s390/kernel/entry.S b/arch/s390/kernel/entry.S index cd5a19138..c920b81be 100644 --- a/arch/s390/kernel/entry.S +++ b/arch/s390/kernel/entry.S @@ -1199,114 +1199,12 @@ cleanup_critical: .quad .Lpsw_idle_lpsw .Lcleanup_save_fpu_regs: - TSTMSK __LC_CPU_FLAGS,_CIF_FPU - bor %r14 - clg %r9,BASED(.Lcleanup_save_fpu_regs_done) - jhe 5f - clg %r9,BASED(.Lcleanup_save_fpu_regs_fp) - jhe 4f - clg %r9,BASED(.Lcleanup_save_fpu_regs_vx_high) - jhe 3f - clg %r9,BASED(.Lcleanup_save_fpu_regs_vx_low) - jhe 2f - clg %r9,BASED(.Lcleanup_save_fpu_fpc_end) - jhe 1f - lg %r2,__LC_CURRENT - aghi %r2,__TASK_thread -0: # Store floating-point controls - stfpc __THREAD_FPU_fpc(%r2) -1: # Load register save area and check if VX is active - lg %r3,__THREAD_FPU_regs(%r2) - TSTMSK __LC_MACHINE_FLAGS,MACHINE_FLAG_VX - jz 4f # no VX -> store FP regs -2: # Store vector registers (V0-V15) - VSTM %v0,%v15,0,%r3 # vstm 0,15,0(3) -3: # Store vector registers (V16-V31) - VSTM %v16,%v31,256,%r3 # vstm 16,31,256(3) - j 5f # -> done, set CIF_FPU flag -4: # Store floating-point registers - std 0,0(%r3) - std 1,8(%r3) - std 2,16(%r3) - std 3,24(%r3) - std 4,32(%r3) - std 5,40(%r3) - std 6,48(%r3) - std 7,56(%r3) - std 8,64(%r3) - std 9,72(%r3) - std 10,80(%r3) - std 11,88(%r3) - std 12,96(%r3) - std 13,104(%r3) - std 14,112(%r3) - std 15,120(%r3) -5: # Set CIF_FPU flag - oi __LC_CPU_FLAGS+7,_CIF_FPU - lg %r9,48(%r11) # return from save_fpu_regs + larl %r9,save_fpu_regs br %r14 -.Lcleanup_save_fpu_fpc_end: - .quad .Lsave_fpu_regs_fpc_end -.Lcleanup_save_fpu_regs_vx_low: - .quad .Lsave_fpu_regs_vx_low -.Lcleanup_save_fpu_regs_vx_high: - .quad .Lsave_fpu_regs_vx_high -.Lcleanup_save_fpu_regs_fp: - .quad .Lsave_fpu_regs_fp -.Lcleanup_save_fpu_regs_done: - .quad .Lsave_fpu_regs_done .Lcleanup_load_fpu_regs: - TSTMSK __LC_CPU_FLAGS,_CIF_FPU - bnor %r14 - clg %r9,BASED(.Lcleanup_load_fpu_regs_done) - jhe 1f - clg %r9,BASED(.Lcleanup_load_fpu_regs_fp) - jhe 2f - clg %r9,BASED(.Lcleanup_load_fpu_regs_vx_high) - jhe 3f - clg %r9,BASED(.Lcleanup_load_fpu_regs_vx) - jhe 4f - lg %r4,__LC_CURRENT - aghi %r4,__TASK_thread - lfpc __THREAD_FPU_fpc(%r4) - TSTMSK __LC_MACHINE_FLAGS,MACHINE_FLAG_VX - lg %r4,__THREAD_FPU_regs(%r4) # %r4 <- reg save area - jz 2f # -> no VX, load FP regs -4: # Load V0 ..V15 registers - VLM %v0,%v15,0,%r4 -3: # Load V16..V31 registers - VLM %v16,%v31,256,%r4 - j 1f -2: # Load floating-point registers - ld 0,0(%r4) - ld 1,8(%r4) - ld 2,16(%r4) - ld 3,24(%r4) - ld 4,32(%r4) - ld 5,40(%r4) - ld 6,48(%r4) - ld 7,56(%r4) - ld 8,64(%r4) - ld 9,72(%r4) - ld 10,80(%r4) - ld 11,88(%r4) - ld 12,96(%r4) - ld 13,104(%r4) - ld 14,112(%r4) - ld 15,120(%r4) -1: # Clear CIF_FPU bit - ni __LC_CPU_FLAGS+7,255-_CIF_FPU - lg %r9,48(%r11) # return from load_fpu_regs + larl %r9,load_fpu_regs br %r14 -.Lcleanup_load_fpu_regs_vx: - .quad .Lload_fpu_regs_vx -.Lcleanup_load_fpu_regs_vx_high: - .quad .Lload_fpu_regs_vx_high -.Lcleanup_load_fpu_regs_fp: - .quad .Lload_fpu_regs_fp -.Lcleanup_load_fpu_regs_done: - .quad .Lload_fpu_regs_done /* * Integer constants diff --git a/arch/s390/kernel/setup.c b/arch/s390/kernel/setup.c index 9220db5c9..93fc63ef6 100644 --- a/arch/s390/kernel/setup.c +++ b/arch/s390/kernel/setup.c @@ -327,6 +327,7 @@ static void __init setup_lowcore(void) + PAGE_SIZE - STACK_FRAME_OVERHEAD - sizeof(struct pt_regs); lc->current_task = (unsigned long) init_thread_union.thread_info.task; lc->thread_info = (unsigned long) &init_thread_union; + lc->lpp = LPP_MAGIC; lc->machine_flags = S390_lowcore.machine_flags; lc->stfl_fac_list = S390_lowcore.stfl_fac_list; memcpy(lc->stfle_fac_list, S390_lowcore.stfle_fac_list, diff --git a/arch/s390/pci/pci.c b/arch/s390/pci/pci.c index 8f19c8f9d..8f75edc99 100644 --- a/arch/s390/pci/pci.c +++ b/arch/s390/pci/pci.c @@ -864,8 +864,11 @@ static inline int barsize(u8 size) static int zpci_mem_init(void) { + BUILD_BUG_ON(!is_power_of_2(__alignof__(struct zpci_fmb)) || + __alignof__(struct zpci_fmb) < sizeof(struct zpci_fmb)); + zdev_fmb_cache = kmem_cache_create("PCI_FMB_cache", sizeof(struct zpci_fmb), - 16, 0, NULL); + __alignof__(struct zpci_fmb), 0, NULL); if (!zdev_fmb_cache) goto error_fmb; diff --git a/arch/sh/mm/kmap.c b/arch/sh/mm/kmap.c index ec29e14ec..bf25d7c79 100644 --- a/arch/sh/mm/kmap.c +++ b/arch/sh/mm/kmap.c @@ -36,6 +36,7 @@ void *kmap_coherent(struct page *page, unsigned long addr) BUG_ON(!test_bit(PG_dcache_clean, &page->flags)); + preempt_disable(); pagefault_disable(); idx = FIX_CMAP_END - @@ -64,4 +65,5 @@ void kunmap_coherent(void *kvaddr) } pagefault_enable(); + preempt_enable(); } diff --git a/arch/um/drivers/mconsole_kern.c b/arch/um/drivers/mconsole_kern.c index b821b13d3..8a6b57108 100644 --- a/arch/um/drivers/mconsole_kern.c +++ b/arch/um/drivers/mconsole_kern.c @@ -133,7 +133,7 @@ void mconsole_proc(struct mc_request *req) ptr += strlen("proc"); ptr = skip_spaces(ptr); - file = file_open_root(mnt->mnt_root, mnt, ptr, O_RDONLY); + file = file_open_root(mnt->mnt_root, mnt, ptr, O_RDONLY, 0); if (IS_ERR(file)) { mconsole_reply(req, "Failed to open file", 1, 0); printk(KERN_ERR "open /proc/%s: %ld\n", ptr, PTR_ERR(file)); diff --git a/arch/x86/Kconfig b/arch/x86/Kconfig index c46662f64..b971ea4be 100644 --- a/arch/x86/Kconfig +++ b/arch/x86/Kconfig @@ -911,10 +911,26 @@ config SCHED_SMT depends on SMP ---help--- SMT scheduler support improves the CPU scheduler's decision making - when dealing with Intel Pentium 4 chips with HyperThreading at a + when dealing with Intel P4/Core 2 chips with HyperThreading at a cost of slightly increased overhead in some places. If unsure say N here. +config SMT_NICE + bool "SMT (Hyperthreading) aware nice priority and policy support" + depends on SCHED_BFS && SCHED_SMT + default y + ---help--- + Enabling Hyperthreading on Intel CPUs decreases the effectiveness + of the use of 'nice' levels and different scheduling policies + (e.g. realtime) due to sharing of CPU power between hyperthreads. + SMT nice support makes each logical CPU aware of what is running on + its hyperthread siblings, maintaining appropriate distribution of + CPU according to nice levels and scheduling policies at the expense + of slightly increased overhead. + + If unsure say Y here. + + config SCHED_MC def_bool y prompt "Multi-core scheduler support" @@ -1160,22 +1176,23 @@ config MICROCODE bool "CPU microcode loading support" default y depends on CPU_SUP_AMD || CPU_SUP_INTEL - depends on BLK_DEV_INITRD select FW_LOADER ---help--- - If you say Y here, you will be able to update the microcode on - certain Intel and AMD processors. The Intel support is for the - IA32 family, e.g. Pentium Pro, Pentium II, Pentium III, Pentium 4, - Xeon etc. The AMD support is for families 0x10 and later. You will - obviously need the actual microcode binary data itself which is not - shipped with the Linux kernel. + Intel and AMD processors. The Intel support is for the IA32 family, + e.g. Pentium Pro, Pentium II, Pentium III, Pentium 4, Xeon etc. The + AMD support is for families 0x10 and later. You will obviously need + the actual microcode binary data itself which is not shipped with + the Linux kernel. - This option selects the general module only, you need to select - at least one vendor specific module as well. + The preferred method to load microcode from a detached initrd is described + in Documentation/x86/early-microcode.txt. For that you need to enable + CONFIG_BLK_DEV_INITRD in order for the loader to be able to scan the + initrd for microcode blobs. - To compile this driver as a module, choose M here: the module - will be called microcode. + In addition, you can build-in the microcode into the kernel. For that you + need to enable FIRMWARE_IN_KERNEL and add the vendor-supplied microcode + to the CONFIG_EXTRA_FIRMWARE config option. config MICROCODE_INTEL bool "Intel microcode loading support" @@ -1995,7 +2012,7 @@ config HOTPLUG_CPU config BOOTPARAM_HOTPLUG_CPU0 bool "Set default setting of cpu0_hotpluggable" default n - depends on HOTPLUG_CPU + depends on HOTPLUG_CPU && !SCHED_BFS ---help--- Set whether default state of cpu0_hotpluggable is on or off. @@ -2024,7 +2041,7 @@ config BOOTPARAM_HOTPLUG_CPU0 config DEBUG_HOTPLUG_CPU0 def_bool n prompt "Debug CPU0 hotplug" - depends on HOTPLUG_CPU + depends on HOTPLUG_CPU && !SCHED_BFS ---help--- Enabling this option offlines CPU0 (if CPU0 can be offlined) as soon as possible and boots up userspace with CPU0 offlined. User diff --git a/arch/x86/entry/common.c b/arch/x86/entry/common.c index 03663740c..1a4477ced 100644 --- a/arch/x86/entry/common.c +++ b/arch/x86/entry/common.c @@ -268,6 +268,7 @@ static void exit_to_usermode_loop(struct pt_regs *regs, u32 cached_flags) /* Called with IRQs disabled. */ __visible inline void prepare_exit_to_usermode(struct pt_regs *regs) { + struct thread_info *ti = pt_regs_to_thread_info(regs); u32 cached_flags; if (IS_ENABLED(CONFIG_PROVE_LOCKING) && WARN_ON(!irqs_disabled())) @@ -275,12 +276,22 @@ __visible inline void prepare_exit_to_usermode(struct pt_regs *regs) lockdep_sys_exit(); - cached_flags = - READ_ONCE(pt_regs_to_thread_info(regs)->flags); + cached_flags = READ_ONCE(ti->flags); if (unlikely(cached_flags & EXIT_TO_USERMODE_LOOP_FLAGS)) exit_to_usermode_loop(regs, cached_flags); +#ifdef CONFIG_COMPAT + /* + * Compat syscalls set TS_COMPAT. Make sure we clear it before + * returning to user mode. We need to clear it *after* signal + * handling, because syscall restart has a fixup for compat + * syscalls. The fixup is exercised by the ptrace_syscall_32 + * selftest. + */ + ti->status &= ~TS_COMPAT; +#endif + user_enter(); } @@ -332,14 +343,6 @@ __visible inline void syscall_return_slowpath(struct pt_regs *regs) if (unlikely(cached_flags & SYSCALL_EXIT_WORK_FLAGS)) syscall_slow_exit_work(regs, cached_flags); -#ifdef CONFIG_COMPAT - /* - * Compat syscalls set TS_COMPAT. Make sure we clear it before - * returning to user mode. - */ - ti->status &= ~TS_COMPAT; -#endif - local_irq_disable(); prepare_exit_to_usermode(regs); } diff --git a/arch/x86/include/asm/apic.h b/arch/x86/include/asm/apic.h index c80f6b6f3..e8c4fba52 100644 --- a/arch/x86/include/asm/apic.h +++ b/arch/x86/include/asm/apic.h @@ -644,8 +644,8 @@ static inline void entering_irq(void) static inline void entering_ack_irq(void) { - ack_APIC_irq(); entering_irq(); + ack_APIC_irq(); } static inline void ipi_entering_ack_irq(void) diff --git a/arch/x86/include/asm/hw_irq.h b/arch/x86/include/asm/hw_irq.h index 1815b7362..84b3d194a 100644 --- a/arch/x86/include/asm/hw_irq.h +++ b/arch/x86/include/asm/hw_irq.h @@ -141,6 +141,7 @@ struct irq_alloc_info { struct irq_cfg { unsigned int dest_apicid; u8 vector; + u8 old_vector; }; extern struct irq_cfg *irq_cfg(unsigned int irq); diff --git a/arch/x86/include/asm/microcode.h b/arch/x86/include/asm/microcode.h index 1e1b07a5a..9d3a96c4d 100644 --- a/arch/x86/include/asm/microcode.h +++ b/arch/x86/include/asm/microcode.h @@ -3,6 +3,7 @@ #include <asm/cpu.h> #include <linux/earlycpio.h> +#include <linux/initrd.h> #define native_rdmsr(msr, val1, val2) \ do { \ @@ -143,4 +144,29 @@ static inline void reload_early_microcode(void) { } static inline bool get_builtin_firmware(struct cpio_data *cd, const char *name) { return false; } #endif + +static inline unsigned long get_initrd_start(void) +{ +#ifdef CONFIG_BLK_DEV_INITRD + return initrd_start; +#else + return 0; +#endif +} + +static inline unsigned long get_initrd_start_addr(void) +{ +#ifdef CONFIG_BLK_DEV_INITRD +#ifdef CONFIG_X86_32 + unsigned long *initrd_start_p = (unsigned long *)__pa_nodebug(&initrd_start); + + return (unsigned long)__pa_nodebug(*initrd_start_p); +#else + return get_initrd_start(); +#endif +#else /* CONFIG_BLK_DEV_INITRD */ + return 0; +#endif +} + #endif /* _ASM_X86_MICROCODE_H */ diff --git a/arch/x86/include/asm/perf_event.h b/arch/x86/include/asm/perf_event.h index 7bcb861a0..5a2ed3ed2 100644 --- a/arch/x86/include/asm/perf_event.h +++ b/arch/x86/include/asm/perf_event.h @@ -165,6 +165,7 @@ struct x86_pmu_capability { #define GLOBAL_STATUS_ASIF BIT_ULL(60) #define GLOBAL_STATUS_COUNTERS_FROZEN BIT_ULL(59) #define GLOBAL_STATUS_LBRS_FROZEN BIT_ULL(58) +#define GLOBAL_STATUS_TRACE_TOPAPMI BIT_ULL(55) /* * IBS cpuid feature detection diff --git a/arch/x86/include/asm/xen/hypervisor.h b/arch/x86/include/asm/xen/hypervisor.h index 8b2d4bea9..39171b364 100644 --- a/arch/x86/include/asm/xen/hypervisor.h +++ b/arch/x86/include/asm/xen/hypervisor.h @@ -62,4 +62,6 @@ void xen_arch_register_cpu(int num); void xen_arch_unregister_cpu(int num); #endif +extern void xen_set_iopl_mask(unsigned mask); + #endif /* _ASM_X86_XEN_HYPERVISOR_H */ diff --git a/arch/x86/kernel/apic/vector.c b/arch/x86/kernel/apic/vector.c index 3b670df4b..ad59d70bc 100644 --- a/arch/x86/kernel/apic/vector.c +++ b/arch/x86/kernel/apic/vector.c @@ -213,6 +213,7 @@ update: */ cpumask_and(d->old_domain, d->old_domain, cpu_online_mask); d->move_in_progress = !cpumask_empty(d->old_domain); + d->cfg.old_vector = d->move_in_progress ? d->cfg.vector : 0; d->cfg.vector = vector; cpumask_copy(d->domain, vector_cpumask); success: @@ -655,46 +656,97 @@ void irq_complete_move(struct irq_cfg *cfg) } /* - * Called with @desc->lock held and interrupts disabled. + * Called from fixup_irqs() with @desc->lock held and interrupts disabled. */ void irq_force_complete_move(struct irq_desc *desc) { struct irq_data *irqdata = irq_desc_get_irq_data(desc); struct apic_chip_data *data = apic_chip_data(irqdata); struct irq_cfg *cfg = data ? &data->cfg : NULL; + unsigned int cpu; if (!cfg) return; - __irq_complete_move(cfg, cfg->vector); - /* * This is tricky. If the cleanup of @data->old_domain has not been * done yet, then the following setaffinity call will fail with * -EBUSY. This can leave the interrupt in a stale state. * - * The cleanup cannot make progress because we hold @desc->lock. So in - * case @data->old_domain is not yet cleaned up, we need to drop the - * lock and acquire it again. @desc cannot go away, because the - * hotplug code holds the sparse irq lock. + * All CPUs are stuck in stop machine with interrupts disabled so + * calling __irq_complete_move() would be completely pointless. */ raw_spin_lock(&vector_lock); - /* Clean out all offline cpus (including ourself) first. */ + /* + * Clean out all offline cpus (including the outgoing one) from the + * old_domain mask. + */ cpumask_and(data->old_domain, data->old_domain, cpu_online_mask); - while (!cpumask_empty(data->old_domain)) { + + /* + * If move_in_progress is cleared and the old_domain mask is empty, + * then there is nothing to cleanup. fixup_irqs() will take care of + * the stale vectors on the outgoing cpu. + */ + if (!data->move_in_progress && cpumask_empty(data->old_domain)) { raw_spin_unlock(&vector_lock); - raw_spin_unlock(&desc->lock); - cpu_relax(); - raw_spin_lock(&desc->lock); + return; + } + + /* + * 1) The interrupt is in move_in_progress state. That means that we + * have not seen an interrupt since the io_apic was reprogrammed to + * the new vector. + * + * 2) The interrupt has fired on the new vector, but the cleanup IPIs + * have not been processed yet. + */ + if (data->move_in_progress) { /* - * Reevaluate apic_chip_data. It might have been cleared after - * we dropped @desc->lock. + * In theory there is a race: + * + * set_ioapic(new_vector) <-- Interrupt is raised before update + * is effective, i.e. it's raised on + * the old vector. + * + * So if the target cpu cannot handle that interrupt before + * the old vector is cleaned up, we get a spurious interrupt + * and in the worst case the ioapic irq line becomes stale. + * + * But in case of cpu hotplug this should be a non issue + * because if the affinity update happens right before all + * cpus rendevouz in stop machine, there is no way that the + * interrupt can be blocked on the target cpu because all cpus + * loops first with interrupts enabled in stop machine, so the + * old vector is not yet cleaned up when the interrupt fires. + * + * So the only way to run into this issue is if the delivery + * of the interrupt on the apic/system bus would be delayed + * beyond the point where the target cpu disables interrupts + * in stop machine. I doubt that it can happen, but at least + * there is a theroretical chance. Virtualization might be + * able to expose this, but AFAICT the IOAPIC emulation is not + * as stupid as the real hardware. + * + * Anyway, there is nothing we can do about that at this point + * w/o refactoring the whole fixup_irq() business completely. + * We print at least the irq number and the old vector number, + * so we have the necessary information when a problem in that + * area arises. */ - data = apic_chip_data(irqdata); - if (!data) - return; - raw_spin_lock(&vector_lock); + pr_warn("IRQ fixup: irq %d move in progress, old vector %d\n", + irqdata->irq, cfg->old_vector); } + /* + * If old_domain is not empty, then other cpus still have the irq + * descriptor set in their vector array. Clean it up. + */ + for_each_cpu(cpu, data->old_domain) + per_cpu(vector_irq, cpu)[cfg->old_vector] = VECTOR_UNUSED; + + /* Cleanup the left overs of the (half finished) move */ + cpumask_clear(data->old_domain); + data->move_in_progress = 0; raw_spin_unlock(&vector_lock); } #endif diff --git a/arch/x86/kernel/cpu/microcode/intel.c b/arch/x86/kernel/cpu/microcode/intel.c index de8a164a6..f9c7f3302 100644 --- a/arch/x86/kernel/cpu/microcode/intel.c +++ b/arch/x86/kernel/cpu/microcode/intel.c @@ -551,10 +551,14 @@ scan_microcode(struct mc_saved_data *mc_saved_data, unsigned long *initrd, cd.data = NULL; cd.size = 0; - cd = find_cpio_data(p, (void *)start, size, &offset); - if (!cd.data) { + /* try built-in microcode if no initrd */ + if (!size) { if (!load_builtin_intel_microcode(&cd)) return UCODE_ERROR; + } else { + cd = find_cpio_data(p, (void *)start, size, &offset); + if (!cd.data) + return UCODE_ERROR; } return get_matching_model_microcode(0, start, cd.data, cd.size, @@ -690,7 +694,7 @@ int __init save_microcode_in_initrd_intel(void) if (count == 0) return ret; - copy_initrd_ptrs(mc_saved, mc_saved_in_initrd, initrd_start, count); + copy_initrd_ptrs(mc_saved, mc_saved_in_initrd, get_initrd_start(), count); ret = save_microcode(&mc_saved_data, mc_saved, count); if (ret) pr_err("Cannot save microcode patches from initrd.\n"); @@ -728,16 +732,20 @@ void __init load_ucode_intel_bsp(void) struct boot_params *p; p = (struct boot_params *)__pa_nodebug(&boot_params); - start = p->hdr.ramdisk_image; size = p->hdr.ramdisk_size; - _load_ucode_intel_bsp( - (struct mc_saved_data *)__pa_nodebug(&mc_saved_data), - (unsigned long *)__pa_nodebug(&mc_saved_in_initrd), - start, size); + /* + * Set start only if we have an initrd image. We cannot use initrd_start + * because it is not set that early yet. + */ + start = (size ? p->hdr.ramdisk_image : 0); + + _load_ucode_intel_bsp((struct mc_saved_data *)__pa_nodebug(&mc_saved_data), + (unsigned long *)__pa_nodebug(&mc_saved_in_initrd), + start, size); #else - start = boot_params.hdr.ramdisk_image + PAGE_OFFSET; size = boot_params.hdr.ramdisk_size; + start = (size ? boot_params.hdr.ramdisk_image + PAGE_OFFSET : 0); _load_ucode_intel_bsp(&mc_saved_data, mc_saved_in_initrd, start, size); #endif @@ -748,20 +756,14 @@ void load_ucode_intel_ap(void) struct mc_saved_data *mc_saved_data_p; struct ucode_cpu_info uci; unsigned long *mc_saved_in_initrd_p; - unsigned long initrd_start_addr; enum ucode_state ret; #ifdef CONFIG_X86_32 - unsigned long *initrd_start_p; - mc_saved_in_initrd_p = - (unsigned long *)__pa_nodebug(mc_saved_in_initrd); + mc_saved_in_initrd_p = (unsigned long *)__pa_nodebug(mc_saved_in_initrd); mc_saved_data_p = (struct mc_saved_data *)__pa_nodebug(&mc_saved_data); - initrd_start_p = (unsigned long *)__pa_nodebug(&initrd_start); - initrd_start_addr = (unsigned long)__pa_nodebug(*initrd_start_p); #else - mc_saved_data_p = &mc_saved_data; mc_saved_in_initrd_p = mc_saved_in_initrd; - initrd_start_addr = initrd_start; + mc_saved_data_p = &mc_saved_data; #endif /* @@ -773,7 +775,7 @@ void load_ucode_intel_ap(void) collect_cpu_info_early(&uci); ret = load_microcode(mc_saved_data_p, mc_saved_in_initrd_p, - initrd_start_addr, &uci); + get_initrd_start_addr(), &uci); if (ret != UCODE_OK) return; diff --git a/arch/x86/kernel/cpu/perf_event.c b/arch/x86/kernel/cpu/perf_event.c index 1b443db2d..6532f5b40 100644 --- a/arch/x86/kernel/cpu/perf_event.c +++ b/arch/x86/kernel/cpu/perf_event.c @@ -596,6 +596,19 @@ void x86_pmu_disable_all(void) } } +/* + * There may be PMI landing after enabled=0. The PMI hitting could be before or + * after disable_all. + * + * If PMI hits before disable_all, the PMU will be disabled in the NMI handler. + * It will not be re-enabled in the NMI handler again, because enabled=0. After + * handling the NMI, disable_all will be called, which will not change the + * state either. If PMI hits after disable_all, the PMU is already disabled + * before entering NMI handler. The NMI handler will not change the state + * either. + * + * So either situation is harmless. + */ static void x86_pmu_disable(struct pmu *pmu) { struct cpu_hw_events *cpuc = this_cpu_ptr(&cpu_hw_events); diff --git a/arch/x86/kernel/cpu/perf_event.h b/arch/x86/kernel/cpu/perf_event.h index 7bb61e32f..98be6d6d3 100644 --- a/arch/x86/kernel/cpu/perf_event.h +++ b/arch/x86/kernel/cpu/perf_event.h @@ -586,6 +586,7 @@ struct x86_pmu { pebs_broken :1, pebs_prec_dist :1; int pebs_record_size; + int pebs_buffer_size; void (*drain_pebs)(struct pt_regs *regs); struct event_constraint *pebs_constraints; void (*pebs_aliases)(struct perf_event *event); @@ -904,6 +905,8 @@ void intel_pmu_lbr_init_skl(void); void intel_pmu_lbr_init_knl(void); +void intel_pmu_pebs_data_source_nhm(void); + int intel_pmu_setup_lbr_filter(struct perf_event *event); void intel_pt_interrupt(void); diff --git a/arch/x86/kernel/cpu/perf_event_intel.c b/arch/x86/kernel/cpu/perf_event_intel.c index fed2ab1f1..760aec1e8 100644 --- a/arch/x86/kernel/cpu/perf_event_intel.c +++ b/arch/x86/kernel/cpu/perf_event_intel.c @@ -1502,7 +1502,15 @@ static __initconst const u64 knl_hw_cache_extra_regs }; /* - * Use from PMIs where the LBRs are already disabled. + * Used from PMIs where the LBRs are already disabled. + * + * This function could be called consecutively. It is required to remain in + * disabled state if called consecutively. + * + * During consecutive calls, the same disable value will be written to related + * registers, so the PMU state remains unchanged. hw.state in + * intel_bts_disable_local will remain PERF_HES_STOPPED too in consecutive + * calls. */ static void __intel_pmu_disable_all(void) { @@ -1884,6 +1892,16 @@ again: if (__test_and_clear_bit(62, (unsigned long *)&status)) { handled++; x86_pmu.drain_pebs(regs); + /* + * There are cases where, even though, the PEBS ovfl bit is set + * in GLOBAL_OVF_STATUS, the PEBS events may also have their + * overflow bits set for their counters. We must clear them + * here because they have been processed as exact samples in + * the drain_pebs() routine. They must not be processed again + * in the for_each_bit_set() loop for regular samples below. + */ + status &= ~cpuc->pebs_enabled; + status &= x86_pmu.intel_ctrl | GLOBAL_STATUS_TRACE_TOPAPMI; } /* @@ -1929,7 +1947,10 @@ again: goto again; done: - __intel_pmu_enable_all(0, true); + /* Only restore PMU state when it's active. See x86_pmu_disable(). */ + if (cpuc->enabled) + __intel_pmu_enable_all(0, true); + /* * Only unmask the NMI after the overflow counters * have been reset. This avoids spurious NMIs on @@ -3396,6 +3417,7 @@ __init int intel_pmu_init(void) intel_perfmon_event_map[PERF_COUNT_HW_STALLED_CYCLES_BACKEND] = X86_CONFIG(.event=0xb1, .umask=0x3f, .inv=1, .cmask=1); + intel_pmu_pebs_data_source_nhm(); x86_add_quirk(intel_nehalem_quirk); pr_cont("Nehalem events, "); @@ -3459,6 +3481,7 @@ __init int intel_pmu_init(void) intel_perfmon_event_map[PERF_COUNT_HW_STALLED_CYCLES_BACKEND] = X86_CONFIG(.event=0xb1, .umask=0x3f, .inv=1, .cmask=1); + intel_pmu_pebs_data_source_nhm(); pr_cont("Westmere events, "); break; diff --git a/arch/x86/kernel/cpu/perf_event_intel_ds.c b/arch/x86/kernel/cpu/perf_event_intel_ds.c index 10602f0a4..955140140 100644 --- a/arch/x86/kernel/cpu/perf_event_intel_ds.c +++ b/arch/x86/kernel/cpu/perf_event_intel_ds.c @@ -51,7 +51,8 @@ union intel_x86_pebs_dse { #define OP_LH (P(OP, LOAD) | P(LVL, HIT)) #define SNOOP_NONE_MISS (P(SNOOP, NONE) | P(SNOOP, MISS)) -static const u64 pebs_data_source[] = { +/* Version for Sandy Bridge and later */ +static u64 pebs_data_source[] = { P(OP, LOAD) | P(LVL, MISS) | P(LVL, L3) | P(SNOOP, NA),/* 0x00:ukn L3 */ OP_LH | P(LVL, L1) | P(SNOOP, NONE), /* 0x01: L1 local */ OP_LH | P(LVL, LFB) | P(SNOOP, NONE), /* 0x02: LFB hit */ @@ -70,6 +71,14 @@ static const u64 pebs_data_source[] = { OP_LH | P(LVL, UNC) | P(SNOOP, NONE), /* 0x0f: uncached */ }; +/* Patch up minor differences in the bits */ +void __init intel_pmu_pebs_data_source_nhm(void) +{ + pebs_data_source[0x05] = OP_LH | P(LVL, L3) | P(SNOOP, HIT); + pebs_data_source[0x06] = OP_LH | P(LVL, L3) | P(SNOOP, HITM); + pebs_data_source[0x07] = OP_LH | P(LVL, L3) | P(SNOOP, HITM); +} + static u64 precise_store_data(u64 status) { union intel_x86_pebs_dse dse; @@ -269,7 +278,7 @@ static int alloc_pebs_buffer(int cpu) if (!x86_pmu.pebs) return 0; - buffer = kzalloc_node(PEBS_BUFFER_SIZE, GFP_KERNEL, node); + buffer = kzalloc_node(x86_pmu.pebs_buffer_size, GFP_KERNEL, node); if (unlikely(!buffer)) return -ENOMEM; @@ -286,7 +295,7 @@ static int alloc_pebs_buffer(int cpu) per_cpu(insn_buffer, cpu) = ibuffer; } - max = PEBS_BUFFER_SIZE / x86_pmu.pebs_record_size; + max = x86_pmu.pebs_buffer_size / x86_pmu.pebs_record_size; ds->pebs_buffer_base = (u64)(unsigned long)buffer; ds->pebs_index = ds->pebs_buffer_base; @@ -1319,6 +1328,7 @@ void __init intel_ds_init(void) x86_pmu.bts = boot_cpu_has(X86_FEATURE_BTS); x86_pmu.pebs = boot_cpu_has(X86_FEATURE_PEBS); + x86_pmu.pebs_buffer_size = PEBS_BUFFER_SIZE; if (x86_pmu.pebs) { char pebs_type = x86_pmu.intel_cap.pebs_trap ? '+' : '-'; int format = x86_pmu.intel_cap.pebs_format; @@ -1327,6 +1337,14 @@ void __init intel_ds_init(void) case 0: printk(KERN_CONT "PEBS fmt0%c, ", pebs_type); x86_pmu.pebs_record_size = sizeof(struct pebs_record_core); + /* + * Using >PAGE_SIZE buffers makes the WRMSR to + * PERF_GLOBAL_CTRL in intel_pmu_enable_all() + * mysteriously hang on Core2. + * + * As a workaround, we don't do this. + */ + x86_pmu.pebs_buffer_size = PAGE_SIZE; x86_pmu.drain_pebs = intel_pmu_drain_pebs_core; break; diff --git a/arch/x86/kernel/cpu/perf_event_intel_uncore_snbep.c b/arch/x86/kernel/cpu/perf_event_intel_uncore_snbep.c index 33acb884c..4547b2cca 100644 --- a/arch/x86/kernel/cpu/perf_event_intel_uncore_snbep.c +++ b/arch/x86/kernel/cpu/perf_event_intel_uncore_snbep.c @@ -2875,11 +2875,13 @@ static struct intel_uncore_type bdx_uncore_sbox = { .format_group = &hswep_uncore_sbox_format_group, }; +#define BDX_MSR_UNCORE_SBOX 3 + static struct intel_uncore_type *bdx_msr_uncores[] = { &bdx_uncore_ubox, &bdx_uncore_cbox, - &bdx_uncore_sbox, &hswep_uncore_pcu, + &bdx_uncore_sbox, NULL, }; @@ -2888,6 +2890,10 @@ void bdx_uncore_cpu_init(void) if (bdx_uncore_cbox.num_boxes > boot_cpu_data.x86_max_cores) bdx_uncore_cbox.num_boxes = boot_cpu_data.x86_max_cores; uncore_msr_uncores = bdx_msr_uncores; + + /* BDX-DE doesn't have SBOX */ + if (boot_cpu_data.x86_model == 86) + uncore_msr_uncores[BDX_MSR_UNCORE_SBOX] = NULL; } static struct intel_uncore_type bdx_uncore_ha = { diff --git a/arch/x86/kernel/cpu/perf_event_knc.c b/arch/x86/kernel/cpu/perf_event_knc.c index 5b0c232d1..b931095e8 100644 --- a/arch/x86/kernel/cpu/perf_event_knc.c +++ b/arch/x86/kernel/cpu/perf_event_knc.c @@ -263,7 +263,9 @@ again: goto again; done: - knc_pmu_enable_all(0); + /* Only restore PMU state when it's active. See x86_pmu_disable(). */ + if (cpuc->enabled) + knc_pmu_enable_all(0); return handled; } diff --git a/arch/x86/kernel/ioport.c b/arch/x86/kernel/ioport.c index 37dae792d..435466fbd 100644 --- a/arch/x86/kernel/ioport.c +++ b/arch/x86/kernel/ioport.c @@ -28,8 +28,18 @@ asmlinkage long sys_ioperm(unsigned long from, unsigned long num, int turn_on) if ((from + num <= from) || (from + num > IO_BITMAP_BITS)) return -EINVAL; +#ifdef CONFIG_SCHED_BFS_AUTOISO + if (turn_on) { + struct sched_param param = { .sched_priority = 0 }; + if (!capable(CAP_SYS_RAWIO)) + return -EPERM; + /* Start X as SCHED_ISO */ + sched_setscheduler_nocheck(current, SCHED_ISO, ¶m); + } +#else if (turn_on && !capable(CAP_SYS_RAWIO)) return -EPERM; +#endif /* * If it's the first ioperm() call in this thread's lifetime, set the @@ -96,18 +106,31 @@ asmlinkage long sys_ioperm(unsigned long from, unsigned long num, int turn_on) SYSCALL_DEFINE1(iopl, unsigned int, level) { struct pt_regs *regs = current_pt_regs(); - unsigned int old = (regs->flags >> 12) & 3; struct thread_struct *t = ¤t->thread; + /* + * Careful: the IOPL bits in regs->flags are undefined under Xen PV + * and changing them has no effect. + */ + unsigned int old = t->iopl >> X86_EFLAGS_IOPL_BIT; + if (level > 3) return -EINVAL; /* Trying to gain more privileges? */ if (level > old) { +#ifdef CONFIG_SCHED_BFS_AUTOISO + struct sched_param param = { .sched_priority = 0 }; +#endif if (!capable(CAP_SYS_RAWIO)) return -EPERM; +#ifdef CONFIG_SCHED_BFS_AUTOISO + /* Start X as SCHED_ISO */ + sched_setscheduler_nocheck(current, SCHED_ISO, ¶m); +#endif } - regs->flags = (regs->flags & ~X86_EFLAGS_IOPL) | (level << 12); - t->iopl = level << 12; + regs->flags = (regs->flags & ~X86_EFLAGS_IOPL) | + (level << X86_EFLAGS_IOPL_BIT); + t->iopl = level << X86_EFLAGS_IOPL_BIT; set_iopl_mask(t->iopl); return 0; diff --git a/arch/x86/kernel/process_64.c b/arch/x86/kernel/process_64.c index b9d99e0f8..9f7518760 100644 --- a/arch/x86/kernel/process_64.c +++ b/arch/x86/kernel/process_64.c @@ -48,6 +48,7 @@ #include <asm/syscalls.h> #include <asm/debugreg.h> #include <asm/switch_to.h> +#include <asm/xen/hypervisor.h> asmlinkage extern void ret_from_fork(void); @@ -411,6 +412,17 @@ __switch_to(struct task_struct *prev_p, struct task_struct *next_p) task_thread_info(prev_p)->flags & _TIF_WORK_CTXSW_PREV)) __switch_to_xtra(prev_p, next_p, tss); +#ifdef CONFIG_XEN + /* + * On Xen PV, IOPL bits in pt_regs->flags have no effect, and + * current_pt_regs()->flags may not match the current task's + * intended IOPL. We need to switch it manually. + */ + if (unlikely(static_cpu_has(X86_FEATURE_XENPV) && + prev->iopl != next->iopl)) + xen_set_iopl_mask(next->iopl); +#endif + if (static_cpu_has_bug(X86_BUG_SYSRET_SS_ATTRS)) { /* * AMD CPUs have a misfeature: SYSRET sets the SS selector but diff --git a/arch/x86/kvm/i8254.c b/arch/x86/kvm/i8254.c index b0ea42b78..ab5318727 100644 --- a/arch/x86/kvm/i8254.c +++ b/arch/x86/kvm/i8254.c @@ -245,7 +245,7 @@ static void kvm_pit_ack_irq(struct kvm_irq_ack_notifier *kian) * PIC is being reset. Handle it gracefully here */ atomic_inc(&ps->pending); - else if (value > 0) + else if (value > 0 && ps->reinject) /* in this case, we had multiple outstanding pit interrupts * that we needed to inject. Reinject */ @@ -288,7 +288,9 @@ static void pit_do_work(struct kthread_work *work) * last one has been acked. */ spin_lock(&ps->inject_lock); - if (ps->irq_ack) { + if (!ps->reinject) + inject = 1; + else if (ps->irq_ack) { ps->irq_ack = 0; inject = 1; } @@ -317,10 +319,10 @@ static enum hrtimer_restart pit_timer_fn(struct hrtimer *data) struct kvm_kpit_state *ps = container_of(data, struct kvm_kpit_state, timer); struct kvm_pit *pt = ps->kvm->arch.vpit; - if (ps->reinject || !atomic_read(&ps->pending)) { + if (ps->reinject) atomic_inc(&ps->pending); - queue_kthread_work(&pt->worker, &pt->expired); - } + + queue_kthread_work(&pt->worker, &pt->expired); if (ps->is_periodic) { hrtimer_add_expires_ns(&ps->timer, ps->period); diff --git a/arch/x86/kvm/vmx.c b/arch/x86/kvm/vmx.c index 9bd8f44ba..539062e24 100644 --- a/arch/x86/kvm/vmx.c +++ b/arch/x86/kvm/vmx.c @@ -2702,8 +2702,15 @@ static void nested_vmx_setup_ctls_msrs(struct vcpu_vmx *vmx) } else vmx->nested.nested_vmx_ept_caps = 0; + /* + * Old versions of KVM use the single-context version without + * checking for support, so declare that it is supported even + * though it is treated as global context. The alternative is + * not failing the single-context invvpid, and it is worse. + */ if (enable_vpid) vmx->nested.nested_vmx_vpid_caps = VMX_VPID_INVVPID_BIT | + VMX_VPID_EXTENT_SINGLE_CONTEXT_BIT | VMX_VPID_EXTENT_GLOBAL_CONTEXT_BIT; else vmx->nested.nested_vmx_vpid_caps = 0; @@ -7398,6 +7405,7 @@ static int handle_invept(struct kvm_vcpu *vcpu) if (!(types & (1UL << type))) { nested_vmx_failValid(vcpu, VMXERR_INVALID_OPERAND_TO_INVEPT_INVVPID); + skip_emulated_instruction(vcpu); return 1; } @@ -7456,6 +7464,7 @@ static int handle_invvpid(struct kvm_vcpu *vcpu) if (!(types & (1UL << type))) { nested_vmx_failValid(vcpu, VMXERR_INVALID_OPERAND_TO_INVEPT_INVVPID); + skip_emulated_instruction(vcpu); return 1; } @@ -7472,12 +7481,17 @@ static int handle_invvpid(struct kvm_vcpu *vcpu) } switch (type) { + case VMX_VPID_EXTENT_SINGLE_CONTEXT: + /* + * Old versions of KVM use the single-context version so we + * have to support it; just treat it the same as all-context. + */ case VMX_VPID_EXTENT_ALL_CONTEXT: __vmx_flush_tlb(vcpu, to_vmx(vcpu)->nested.vpid02); nested_vmx_succeed(vcpu); break; default: - /* Trap single context invalidation invvpid calls */ + /* Trap individual address invalidation invvpid calls */ BUG_ON(1); break; } diff --git a/arch/x86/kvm/x86.c b/arch/x86/kvm/x86.c index eaf6ee8c2..d47d231e0 100644 --- a/arch/x86/kvm/x86.c +++ b/arch/x86/kvm/x86.c @@ -2752,6 +2752,7 @@ void kvm_arch_vcpu_load(struct kvm_vcpu *vcpu, int cpu) } kvm_make_request(KVM_REQ_STEAL_UPDATE, vcpu); + vcpu->arch.switch_db_regs |= KVM_DEBUGREG_RELOAD; } void kvm_arch_vcpu_put(struct kvm_vcpu *vcpu) diff --git a/arch/x86/mm/tlb.c b/arch/x86/mm/tlb.c index 8f4cc3dfa..5fb6adaaa 100644 --- a/arch/x86/mm/tlb.c +++ b/arch/x86/mm/tlb.c @@ -106,8 +106,6 @@ static void flush_tlb_func(void *info) if (f->flush_mm != this_cpu_read(cpu_tlbstate.active_mm)) return; - if (!f->flush_end) - f->flush_end = f->flush_start + PAGE_SIZE; count_vm_tlb_event(NR_TLB_REMOTE_FLUSH_RECEIVED); if (this_cpu_read(cpu_tlbstate.state) == TLBSTATE_OK) { @@ -135,12 +133,20 @@ void native_flush_tlb_others(const struct cpumask *cpumask, unsigned long end) { struct flush_tlb_info info; + + if (end == 0) + end = start + PAGE_SIZE; info.flush_mm = mm; info.flush_start = start; info.flush_end = end; count_vm_tlb_event(NR_TLB_REMOTE_FLUSH); - trace_tlb_flush(TLB_REMOTE_SEND_IPI, end - start); + if (end == TLB_FLUSH_ALL) + trace_tlb_flush(TLB_REMOTE_SEND_IPI, TLB_FLUSH_ALL); + else + trace_tlb_flush(TLB_REMOTE_SEND_IPI, + (end - start) >> PAGE_SHIFT); + if (is_uv_system()) { unsigned int cpu; diff --git a/arch/x86/pci/fixup.c b/arch/x86/pci/fixup.c index e58565556..0ae7e9fa3 100644 --- a/arch/x86/pci/fixup.c +++ b/arch/x86/pci/fixup.c @@ -540,3 +540,10 @@ static void twinhead_reserve_killing_zone(struct pci_dev *dev) } } DECLARE_PCI_FIXUP_HEADER(PCI_VENDOR_ID_INTEL, 0x27B9, twinhead_reserve_killing_zone); + +static void pci_bdwep_bar(struct pci_dev *dev) +{ + dev->non_compliant_bars = 1; +} +DECLARE_PCI_FIXUP_EARLY(PCI_VENDOR_ID_INTEL, 0x6fa0, pci_bdwep_bar); +DECLARE_PCI_FIXUP_EARLY(PCI_VENDOR_ID_INTEL, 0x6fc0, pci_bdwep_bar); diff --git a/arch/x86/xen/enlighten.c b/arch/x86/xen/enlighten.c index d09e4c9d7..e3679db17 100644 --- a/arch/x86/xen/enlighten.c +++ b/arch/x86/xen/enlighten.c @@ -961,7 +961,7 @@ static void xen_load_sp0(struct tss_struct *tss, tss->x86_tss.sp0 = thread->sp0; } -static void xen_set_iopl_mask(unsigned mask) +void xen_set_iopl_mask(unsigned mask) { struct physdev_set_iopl set_iopl; diff --git a/arch/xtensa/kernel/head.S b/arch/xtensa/kernel/head.S index 9ed55649a..05e1df943 100644 --- a/arch/xtensa/kernel/head.S +++ b/arch/xtensa/kernel/head.S @@ -128,7 +128,7 @@ ENTRY(_startup) wsr a0, icountlevel .set _index, 0 - .rept XCHAL_NUM_DBREAK - 1 + .rept XCHAL_NUM_DBREAK wsr a0, SREG_DBREAKC + _index .set _index, _index + 1 .endr diff --git a/arch/xtensa/mm/cache.c b/arch/xtensa/mm/cache.c index d75aa1476..1a804a2f9 100644 --- a/arch/xtensa/mm/cache.c +++ b/arch/xtensa/mm/cache.c @@ -97,11 +97,11 @@ void clear_user_highpage(struct page *page, unsigned long vaddr) unsigned long paddr; void *kvaddr = coherent_kvaddr(page, TLBTEMP_BASE_1, vaddr, &paddr); - pagefault_disable(); + preempt_disable(); kmap_invalidate_coherent(page, vaddr); set_bit(PG_arch_1, &page->flags); clear_page_alias(kvaddr, paddr); - pagefault_enable(); + preempt_enable(); } void copy_user_highpage(struct page *dst, struct page *src, @@ -113,11 +113,11 @@ void copy_user_highpage(struct page *dst, struct page *src, void *src_vaddr = coherent_kvaddr(src, TLBTEMP_BASE_2, vaddr, &src_paddr); - pagefault_disable(); + preempt_disable(); kmap_invalidate_coherent(dst, vaddr); set_bit(PG_arch_1, &dst->flags); copy_page_alias(dst_vaddr, src_vaddr, dst_paddr, src_paddr); - pagefault_enable(); + preempt_enable(); } #endif /* DCACHE_WAY_SIZE > PAGE_SIZE */ diff --git a/arch/xtensa/platforms/iss/console.c b/arch/xtensa/platforms/iss/console.c index 70cb408bc..92d785fef 100644 --- a/arch/xtensa/platforms/iss/console.c +++ b/arch/xtensa/platforms/iss/console.c @@ -100,21 +100,23 @@ static void rs_poll(unsigned long priv) { struct tty_port *port = (struct tty_port *)priv; int i = 0; + int rd = 1; unsigned char c; spin_lock(&timer_lock); while (simc_poll(0)) { - simc_read(0, &c, 1); + rd = simc_read(0, &c, 1); + if (rd <= 0) + break; tty_insert_flip_char(port, c, TTY_NORMAL); i++; } if (i) tty_flip_buffer_push(port); - - - mod_timer(&serial_timer, jiffies + SERIAL_TIMER_VALUE); + if (rd) + mod_timer(&serial_timer, jiffies + SERIAL_TIMER_VALUE); spin_unlock(&timer_lock); } diff --git a/block/blk-core.c b/block/blk-core.c index 64d137b04..f98581f9e 100644 --- a/block/blk-core.c +++ b/block/blk-core.c @@ -2203,7 +2203,7 @@ int blk_insert_cloned_request(struct request_queue *q, struct request *rq) if (q->mq_ops) { if (blk_queue_io_stat(q)) blk_account_io_start(rq, true); - blk_mq_insert_request(rq, false, true, true); + blk_mq_insert_request(rq, false, true, false); return 0; } diff --git a/crypto/asymmetric_keys/x509_cert_parser.c b/crypto/asymmetric_keys/x509_cert_parser.c index 021d39c0b..13c4e5a5f 100644 --- a/crypto/asymmetric_keys/x509_cert_parser.c +++ b/crypto/asymmetric_keys/x509_cert_parser.c @@ -494,7 +494,7 @@ int x509_decode_time(time64_t *_t, size_t hdrlen, unsigned char tag, const unsigned char *value, size_t vlen) { - static const unsigned char month_lengths[] = { 31, 29, 31, 30, 31, 30, + static const unsigned char month_lengths[] = { 31, 28, 31, 30, 31, 30, 31, 31, 30, 31, 30, 31 }; const unsigned char *p = value; unsigned year, mon, day, hour, min, sec, mon_len; @@ -540,9 +540,9 @@ int x509_decode_time(time64_t *_t, size_t hdrlen, if (year % 4 == 0) { mon_len = 29; if (year % 100 == 0) { - year /= 100; - if (year % 4 != 0) - mon_len = 28; + mon_len = 28; + if (year % 400 == 0) + mon_len = 29; } } } diff --git a/crypto/keywrap.c b/crypto/keywrap.c index b1d106ce5..72014f963 100644 --- a/crypto/keywrap.c +++ b/crypto/keywrap.c @@ -212,7 +212,7 @@ static int crypto_kw_decrypt(struct blkcipher_desc *desc, SEMIBSIZE)) ret = -EBADMSG; - memzero_explicit(&block, sizeof(struct crypto_kw_block)); + memzero_explicit(block, sizeof(struct crypto_kw_block)); return ret; } @@ -297,7 +297,7 @@ static int crypto_kw_encrypt(struct blkcipher_desc *desc, /* establish the IV for the caller to pick up */ memcpy(desc->info, block->A, SEMIBSIZE); - memzero_explicit(&block, sizeof(struct crypto_kw_block)); + memzero_explicit(block, sizeof(struct crypto_kw_block)); return 0; } diff --git a/drivers/acpi/resource.c b/drivers/acpi/resource.c index d02fd5304..56241eb34 100644 --- a/drivers/acpi/resource.c +++ b/drivers/acpi/resource.c @@ -27,8 +27,20 @@ #ifdef CONFIG_X86 #define valid_IRQ(i) (((i) != 0) && ((i) != 2)) +static inline bool acpi_iospace_resource_valid(struct resource *res) +{ + /* On X86 IO space is limited to the [0 - 64K] IO port range */ + return res->end < 0x10003; +} #else #define valid_IRQ(i) (true) +/* + * ACPI IO descriptors on arches other than X86 contain MMIO CPU physical + * addresses mapping IO space in CPU physical address space, IO space + * resources can be placed anywhere in the 64-bit physical address space. + */ +static inline bool +acpi_iospace_resource_valid(struct resource *res) { return true; } #endif static bool acpi_dev_resource_len_valid(u64 start, u64 end, u64 len, bool io) @@ -127,7 +139,7 @@ static void acpi_dev_ioresource_flags(struct resource *res, u64 len, if (!acpi_dev_resource_len_valid(res->start, res->end, len, true)) res->flags |= IORESOURCE_DISABLED | IORESOURCE_UNSET; - if (res->end >= 0x10003) + if (!acpi_iospace_resource_valid(res)) res->flags |= IORESOURCE_DISABLED | IORESOURCE_UNSET; if (io_decode == ACPI_DECODE_16) diff --git a/drivers/acpi/sleep.c b/drivers/acpi/sleep.c index 9cb975200..f054cadf3 100644 --- a/drivers/acpi/sleep.c +++ b/drivers/acpi/sleep.c @@ -714,6 +714,7 @@ static int acpi_hibernation_enter(void) static void acpi_hibernation_leave(void) { + pm_set_resume_via_firmware(); /* * If ACPI is not enabled by the BIOS and the boot kernel, we need to * enable it here. diff --git a/drivers/block/brd.c b/drivers/block/brd.c index cb27190e9..f7ecc287d 100644 --- a/drivers/block/brd.c +++ b/drivers/block/brd.c @@ -341,7 +341,7 @@ static blk_qc_t brd_make_request(struct request_queue *q, struct bio *bio) if (unlikely(bio->bi_rw & REQ_DISCARD)) { if (sector & ((PAGE_SIZE >> SECTOR_SHIFT) - 1) || - bio->bi_iter.bi_size & PAGE_MASK) + bio->bi_iter.bi_size & ~PAGE_MASK) goto io_error; discard_from_brd(brd, sector, bio->bi_iter.bi_size); goto out; diff --git a/drivers/block/loop.c b/drivers/block/loop.c index 423f4ca7d..0b816b24b 100644 --- a/drivers/block/loop.c +++ b/drivers/block/loop.c @@ -706,6 +706,24 @@ static inline int is_loop_device(struct file *file) return i && S_ISBLK(i->i_mode) && MAJOR(i->i_rdev) == LOOP_MAJOR; } +/* + * for AUFS + * no get/put for file. + */ +struct file *loop_backing_file(struct super_block *sb) +{ + struct file *ret; + struct loop_device *l; + + ret = NULL; + if (MAJOR(sb->s_dev) == LOOP_MAJOR) { + l = sb->s_bdev->bd_disk->private_data; + ret = l->lo_backing_file; + } + return ret; +} +EXPORT_SYMBOL_GPL(loop_backing_file); + /* loop sysfs attributes */ static ssize_t loop_attr_show(struct device *dev, char *page, diff --git a/drivers/block/mtip32xx/mtip32xx.c b/drivers/block/mtip32xx/mtip32xx.c index 9b180dbbd..1c330b61f 100644 --- a/drivers/block/mtip32xx/mtip32xx.c +++ b/drivers/block/mtip32xx/mtip32xx.c @@ -173,7 +173,13 @@ static struct mtip_cmd *mtip_get_int_command(struct driver_data *dd) { struct request *rq; + if (mtip_check_surprise_removal(dd->pdev)) + return NULL; + rq = blk_mq_alloc_request(dd->queue, 0, BLK_MQ_REQ_RESERVED); + if (IS_ERR(rq)) + return NULL; + return blk_mq_rq_to_pdu(rq); } @@ -233,15 +239,9 @@ static void mtip_async_complete(struct mtip_port *port, "Command tag %d failed due to TFE\n", tag); } - /* Unmap the DMA scatter list entries */ - dma_unmap_sg(&dd->pdev->dev, cmd->sg, cmd->scatter_ents, cmd->direction); - rq = mtip_rq_from_tag(dd, tag); - if (unlikely(cmd->unaligned)) - up(&port->cmd_slot_unal); - - blk_mq_end_request(rq, status ? -EIO : 0); + blk_mq_complete_request(rq, status); } /* @@ -581,6 +581,8 @@ static void mtip_completion(struct mtip_port *port, dev_warn(&port->dd->pdev->dev, "Internal command %d completed with TFE\n", tag); + command->comp_func = NULL; + command->comp_data = NULL; complete(waiting); } @@ -618,8 +620,6 @@ static void mtip_handle_tfe(struct driver_data *dd) port = dd->port; - set_bit(MTIP_PF_EH_ACTIVE_BIT, &port->flags); - if (test_bit(MTIP_PF_IC_ACTIVE_BIT, &port->flags)) { cmd = mtip_cmd_from_tag(dd, MTIP_TAG_INTERNAL); dbg_printk(MTIP_DRV_NAME " TFE for the internal command\n"); @@ -628,7 +628,7 @@ static void mtip_handle_tfe(struct driver_data *dd) cmd->comp_func(port, MTIP_TAG_INTERNAL, cmd, PORT_IRQ_TF_ERR); } - goto handle_tfe_exit; + return; } /* clear the tag accumulator */ @@ -701,7 +701,7 @@ static void mtip_handle_tfe(struct driver_data *dd) fail_reason = "thermal shutdown"; } if (buf[288] == 0xBF) { - set_bit(MTIP_DDF_SEC_LOCK_BIT, &dd->dd_flag); + set_bit(MTIP_DDF_REBUILD_FAILED_BIT, &dd->dd_flag); dev_info(&dd->pdev->dev, "Drive indicates rebuild has failed. Secure erase required.\n"); fail_all_ncq_cmds = 1; @@ -771,11 +771,6 @@ static void mtip_handle_tfe(struct driver_data *dd) } } print_tags(dd, "reissued (TFE)", tagaccum, cmd_cnt); - -handle_tfe_exit: - /* clear eh_active */ - clear_bit(MTIP_PF_EH_ACTIVE_BIT, &port->flags); - wake_up_interruptible(&port->svc_wait); } /* @@ -1007,6 +1002,7 @@ static bool mtip_pause_ncq(struct mtip_port *port, (fis->features == 0x27 || fis->features == 0x72 || fis->features == 0x62 || fis->features == 0x26))) { clear_bit(MTIP_DDF_SEC_LOCK_BIT, &port->dd->dd_flag); + clear_bit(MTIP_DDF_REBUILD_FAILED_BIT, &port->dd->dd_flag); /* Com reset after secure erase or lowlevel format */ mtip_restart_port(port); clear_bit(MTIP_PF_SE_ACTIVE_BIT, &port->flags); @@ -1021,12 +1017,14 @@ static bool mtip_pause_ncq(struct mtip_port *port, * * @port Pointer to port data structure * @timeout Max duration to wait (ms) + * @atomic gfp_t flag to indicate blockable context or not * * return value * 0 Success * -EBUSY Commands still active */ -static int mtip_quiesce_io(struct mtip_port *port, unsigned long timeout) +static int mtip_quiesce_io(struct mtip_port *port, unsigned long timeout, + gfp_t atomic) { unsigned long to; unsigned int n; @@ -1037,16 +1035,21 @@ static int mtip_quiesce_io(struct mtip_port *port, unsigned long timeout) to = jiffies + msecs_to_jiffies(timeout); do { if (test_bit(MTIP_PF_SVC_THD_ACTIVE_BIT, &port->flags) && - test_bit(MTIP_PF_ISSUE_CMDS_BIT, &port->flags)) { + test_bit(MTIP_PF_ISSUE_CMDS_BIT, &port->flags) && + atomic == GFP_KERNEL) { msleep(20); continue; /* svc thd is actively issuing commands */ } - msleep(100); + if (atomic == GFP_KERNEL) + msleep(100); + else { + cpu_relax(); + udelay(100); + } + if (mtip_check_surprise_removal(port->dd->pdev)) goto err_fault; - if (test_bit(MTIP_DDF_REMOVE_PENDING_BIT, &port->dd->dd_flag)) - goto err_fault; /* * Ignore s_active bit 0 of array element 0. @@ -1099,6 +1102,7 @@ static int mtip_exec_internal_command(struct mtip_port *port, struct mtip_cmd *int_cmd; struct driver_data *dd = port->dd; int rv = 0; + unsigned long start; /* Make sure the buffer is 8 byte aligned. This is asic specific. */ if (buffer & 0x00000007) { @@ -1107,6 +1111,10 @@ static int mtip_exec_internal_command(struct mtip_port *port, } int_cmd = mtip_get_int_command(dd); + if (!int_cmd) { + dbg_printk(MTIP_DRV_NAME "Unable to allocate tag for PIO cmd\n"); + return -EFAULT; + } set_bit(MTIP_PF_IC_ACTIVE_BIT, &port->flags); @@ -1119,7 +1127,7 @@ static int mtip_exec_internal_command(struct mtip_port *port, if (fis->command != ATA_CMD_STANDBYNOW1) { /* wait for io to complete if non atomic */ if (mtip_quiesce_io(port, - MTIP_QUIESCE_IO_TIMEOUT_MS) < 0) { + MTIP_QUIESCE_IO_TIMEOUT_MS, atomic) < 0) { dev_warn(&dd->pdev->dev, "Failed to quiesce IO\n"); mtip_put_int_command(dd, int_cmd); @@ -1162,6 +1170,8 @@ static int mtip_exec_internal_command(struct mtip_port *port, /* Populate the command header */ int_cmd->command_header->byte_count = 0; + start = jiffies; + /* Issue the command to the hardware */ mtip_issue_non_ncq_command(port, MTIP_TAG_INTERNAL); @@ -1170,10 +1180,12 @@ static int mtip_exec_internal_command(struct mtip_port *port, if ((rv = wait_for_completion_interruptible_timeout( &wait, msecs_to_jiffies(timeout))) <= 0) { + if (rv == -ERESTARTSYS) { /* interrupted */ dev_err(&dd->pdev->dev, - "Internal command [%02X] was interrupted after %lu ms\n", - fis->command, timeout); + "Internal command [%02X] was interrupted after %u ms\n", + fis->command, + jiffies_to_msecs(jiffies - start)); rv = -EINTR; goto exec_ic_exit; } else if (rv == 0) /* timeout */ @@ -2890,6 +2902,42 @@ static int mtip_ftl_rebuild_poll(struct driver_data *dd) return -EFAULT; } +static void mtip_softirq_done_fn(struct request *rq) +{ + struct mtip_cmd *cmd = blk_mq_rq_to_pdu(rq); + struct driver_data *dd = rq->q->queuedata; + + /* Unmap the DMA scatter list entries */ + dma_unmap_sg(&dd->pdev->dev, cmd->sg, cmd->scatter_ents, + cmd->direction); + + if (unlikely(cmd->unaligned)) + up(&dd->port->cmd_slot_unal); + + blk_mq_end_request(rq, rq->errors); +} + +static void mtip_abort_cmd(struct request *req, void *data, + bool reserved) +{ + struct driver_data *dd = data; + + dbg_printk(MTIP_DRV_NAME " Aborting request, tag = %d\n", req->tag); + + clear_bit(req->tag, dd->port->cmds_to_issue); + req->errors = -EIO; + mtip_softirq_done_fn(req); +} + +static void mtip_queue_cmd(struct request *req, void *data, + bool reserved) +{ + struct driver_data *dd = data; + + set_bit(req->tag, dd->port->cmds_to_issue); + blk_abort_request(req); +} + /* * service thread to issue queued commands * @@ -2902,7 +2950,7 @@ static int mtip_ftl_rebuild_poll(struct driver_data *dd) static int mtip_service_thread(void *data) { struct driver_data *dd = (struct driver_data *)data; - unsigned long slot, slot_start, slot_wrap; + unsigned long slot, slot_start, slot_wrap, to; unsigned int num_cmd_slots = dd->slot_groups * 32; struct mtip_port *port = dd->port; @@ -2917,9 +2965,7 @@ static int mtip_service_thread(void *data) * is in progress nor error handling is active */ wait_event_interruptible(port->svc_wait, (port->flags) && - !(port->flags & MTIP_PF_PAUSE_IO)); - - set_bit(MTIP_PF_SVC_THD_ACTIVE_BIT, &port->flags); + (port->flags & MTIP_PF_SVC_THD_WORK)); if (kthread_should_stop() || test_bit(MTIP_PF_SVC_THD_STOP_BIT, &port->flags)) @@ -2929,6 +2975,8 @@ static int mtip_service_thread(void *data) &dd->dd_flag))) goto st_out; + set_bit(MTIP_PF_SVC_THD_ACTIVE_BIT, &port->flags); + restart_eh: /* Demux bits: start with error handling */ if (test_bit(MTIP_PF_EH_ACTIVE_BIT, &port->flags)) { @@ -2939,6 +2987,32 @@ restart_eh: if (test_bit(MTIP_PF_EH_ACTIVE_BIT, &port->flags)) goto restart_eh; + if (test_bit(MTIP_PF_TO_ACTIVE_BIT, &port->flags)) { + to = jiffies + msecs_to_jiffies(5000); + + do { + mdelay(100); + } while (atomic_read(&dd->irq_workers_active) != 0 && + time_before(jiffies, to)); + + if (atomic_read(&dd->irq_workers_active) != 0) + dev_warn(&dd->pdev->dev, + "Completion workers still active!"); + + spin_lock(dd->queue->queue_lock); + blk_mq_all_tag_busy_iter(*dd->tags.tags, + mtip_queue_cmd, dd); + spin_unlock(dd->queue->queue_lock); + + set_bit(MTIP_PF_ISSUE_CMDS_BIT, &dd->port->flags); + + if (mtip_device_reset(dd)) + blk_mq_all_tag_busy_iter(*dd->tags.tags, + mtip_abort_cmd, dd); + + clear_bit(MTIP_PF_TO_ACTIVE_BIT, &dd->port->flags); + } + if (test_bit(MTIP_PF_ISSUE_CMDS_BIT, &port->flags)) { slot = 1; /* used to restrict the loop to one iteration */ @@ -2971,10 +3045,8 @@ restart_eh: } if (test_bit(MTIP_PF_REBUILD_BIT, &port->flags)) { - if (mtip_ftl_rebuild_poll(dd) < 0) - set_bit(MTIP_DDF_REBUILD_FAILED_BIT, - &dd->dd_flag); - clear_bit(MTIP_PF_REBUILD_BIT, &port->flags); + if (mtip_ftl_rebuild_poll(dd) == 0) + clear_bit(MTIP_PF_REBUILD_BIT, &port->flags); } } @@ -3089,7 +3161,7 @@ static int mtip_hw_get_identify(struct driver_data *dd) if (buf[288] == 0xBF) { dev_info(&dd->pdev->dev, "Drive indicates rebuild has failed.\n"); - /* TODO */ + set_bit(MTIP_DDF_REBUILD_FAILED_BIT, &dd->dd_flag); } } @@ -3263,20 +3335,25 @@ out1: return rv; } -static void mtip_standby_drive(struct driver_data *dd) +static int mtip_standby_drive(struct driver_data *dd) { - if (dd->sr) - return; + int rv = 0; + if (dd->sr || !dd->port) + return -ENODEV; /* * Send standby immediate (E0h) to the drive so that it * saves its state. */ if (!test_bit(MTIP_PF_REBUILD_BIT, &dd->port->flags) && - !test_bit(MTIP_DDF_SEC_LOCK_BIT, &dd->dd_flag)) - if (mtip_standby_immediate(dd->port)) + !test_bit(MTIP_DDF_REBUILD_FAILED_BIT, &dd->dd_flag) && + !test_bit(MTIP_DDF_SEC_LOCK_BIT, &dd->dd_flag)) { + rv = mtip_standby_immediate(dd->port); + if (rv) dev_warn(&dd->pdev->dev, "STANDBY IMMEDIATE failed\n"); + } + return rv; } /* @@ -3289,10 +3366,6 @@ static void mtip_standby_drive(struct driver_data *dd) */ static int mtip_hw_exit(struct driver_data *dd) { - /* - * Send standby immediate (E0h) to the drive so that it - * saves its state. - */ if (!dd->sr) { /* de-initialize the port. */ mtip_deinit_port(dd->port); @@ -3334,8 +3407,7 @@ static int mtip_hw_shutdown(struct driver_data *dd) * Send standby immediate (E0h) to the drive so that it * saves its state. */ - if (!dd->sr && dd->port) - mtip_standby_immediate(dd->port); + mtip_standby_drive(dd); return 0; } @@ -3358,7 +3430,7 @@ static int mtip_hw_suspend(struct driver_data *dd) * Send standby immediate (E0h) to the drive * so that it saves its state. */ - if (mtip_standby_immediate(dd->port) != 0) { + if (mtip_standby_drive(dd) != 0) { dev_err(&dd->pdev->dev, "Failed standby-immediate command\n"); return -EFAULT; @@ -3596,6 +3668,28 @@ static int mtip_block_getgeo(struct block_device *dev, return 0; } +static int mtip_block_open(struct block_device *dev, fmode_t mode) +{ + struct driver_data *dd; + + if (dev && dev->bd_disk) { + dd = (struct driver_data *) dev->bd_disk->private_data; + + if (dd) { + if (test_bit(MTIP_DDF_REMOVAL_BIT, + &dd->dd_flag)) { + return -ENODEV; + } + return 0; + } + } + return -ENODEV; +} + +void mtip_block_release(struct gendisk *disk, fmode_t mode) +{ +} + /* * Block device operation function. * @@ -3603,6 +3697,8 @@ static int mtip_block_getgeo(struct block_device *dev, * layer. */ static const struct block_device_operations mtip_block_ops = { + .open = mtip_block_open, + .release = mtip_block_release, .ioctl = mtip_block_ioctl, #ifdef CONFIG_COMPAT .compat_ioctl = mtip_block_compat_ioctl, @@ -3664,10 +3760,9 @@ static int mtip_submit_request(struct blk_mq_hw_ctx *hctx, struct request *rq) rq_data_dir(rq))) { return -ENODATA; } - if (unlikely(test_bit(MTIP_DDF_SEC_LOCK_BIT, &dd->dd_flag))) + if (unlikely(test_bit(MTIP_DDF_SEC_LOCK_BIT, &dd->dd_flag) || + test_bit(MTIP_DDF_REBUILD_FAILED_BIT, &dd->dd_flag))) return -ENODATA; - if (test_bit(MTIP_DDF_REBUILD_FAILED_BIT, &dd->dd_flag)) - return -ENXIO; } if (rq->cmd_flags & REQ_DISCARD) { @@ -3779,11 +3874,33 @@ static int mtip_init_cmd(void *data, struct request *rq, unsigned int hctx_idx, return 0; } +static enum blk_eh_timer_return mtip_cmd_timeout(struct request *req, + bool reserved) +{ + struct driver_data *dd = req->q->queuedata; + int ret = BLK_EH_RESET_TIMER; + + if (reserved) + goto exit_handler; + + if (test_bit(req->tag, dd->port->cmds_to_issue)) + goto exit_handler; + + if (test_and_set_bit(MTIP_PF_TO_ACTIVE_BIT, &dd->port->flags)) + goto exit_handler; + + wake_up_interruptible(&dd->port->svc_wait); +exit_handler: + return ret; +} + static struct blk_mq_ops mtip_mq_ops = { .queue_rq = mtip_queue_rq, .map_queue = blk_mq_map_queue, .init_request = mtip_init_cmd, .exit_request = mtip_free_cmd, + .complete = mtip_softirq_done_fn, + .timeout = mtip_cmd_timeout, }; /* @@ -3850,7 +3967,6 @@ static int mtip_block_initialize(struct driver_data *dd) mtip_hw_debugfs_init(dd); -skip_create_disk: memset(&dd->tags, 0, sizeof(dd->tags)); dd->tags.ops = &mtip_mq_ops; dd->tags.nr_hw_queues = 1; @@ -3860,12 +3976,13 @@ skip_create_disk: dd->tags.numa_node = dd->numa_node; dd->tags.flags = BLK_MQ_F_SHOULD_MERGE; dd->tags.driver_data = dd; + dd->tags.timeout = MTIP_NCQ_CMD_TIMEOUT_MS; rv = blk_mq_alloc_tag_set(&dd->tags); if (rv) { dev_err(&dd->pdev->dev, "Unable to allocate request queue\n"); - goto block_queue_alloc_init_error; + goto block_queue_alloc_tag_error; } /* Allocate the request queue. */ @@ -3880,6 +3997,7 @@ skip_create_disk: dd->disk->queue = dd->queue; dd->queue->queuedata = dd; +skip_create_disk: /* Initialize the protocol layer. */ wait_for_rebuild = mtip_hw_get_identify(dd); if (wait_for_rebuild < 0) { @@ -3976,8 +4094,9 @@ kthread_run_error: read_capacity_error: init_hw_cmds_error: blk_cleanup_queue(dd->queue); - blk_mq_free_tag_set(&dd->tags); block_queue_alloc_init_error: + blk_mq_free_tag_set(&dd->tags); +block_queue_alloc_tag_error: mtip_hw_debugfs_exit(dd); disk_index_error: spin_lock(&rssd_index_lock); @@ -3994,6 +4113,22 @@ protocol_init_error: return rv; } +static void mtip_no_dev_cleanup(struct request *rq, void *data, bool reserv) +{ + struct driver_data *dd = (struct driver_data *)data; + struct mtip_cmd *cmd; + + if (likely(!reserv)) + blk_mq_complete_request(rq, -ENODEV); + else if (test_bit(MTIP_PF_IC_ACTIVE_BIT, &dd->port->flags)) { + + cmd = mtip_cmd_from_tag(dd, MTIP_TAG_INTERNAL); + if (cmd->comp_func) + cmd->comp_func(dd->port, MTIP_TAG_INTERNAL, + cmd, -ENODEV); + } +} + /* * Block layer deinitialization function. * @@ -4025,12 +4160,23 @@ static int mtip_block_remove(struct driver_data *dd) } } - if (!dd->sr) - mtip_standby_drive(dd); + if (!dd->sr) { + /* + * Explicitly wait here for IOs to quiesce, + * as mtip_standby_drive usually won't wait for IOs. + */ + if (!mtip_quiesce_io(dd->port, MTIP_QUIESCE_IO_TIMEOUT_MS, + GFP_KERNEL)) + mtip_standby_drive(dd); + } else dev_info(&dd->pdev->dev, "device %s surprise removal\n", dd->disk->disk_name); + blk_mq_freeze_queue_start(dd->queue); + blk_mq_stop_hw_queues(dd->queue); + blk_mq_all_tag_busy_iter(dd->tags.tags[0], mtip_no_dev_cleanup, dd); + /* * Delete our gendisk structure. This also removes the device * from /dev @@ -4040,7 +4186,8 @@ static int mtip_block_remove(struct driver_data *dd) dd->bdev = NULL; } if (dd->disk) { - del_gendisk(dd->disk); + if (test_bit(MTIP_DDF_INIT_DONE_BIT, &dd->dd_flag)) + del_gendisk(dd->disk); if (dd->disk->queue) { blk_cleanup_queue(dd->queue); blk_mq_free_tag_set(&dd->tags); @@ -4081,7 +4228,8 @@ static int mtip_block_shutdown(struct driver_data *dd) dev_info(&dd->pdev->dev, "Shutting down %s ...\n", dd->disk->disk_name); - del_gendisk(dd->disk); + if (test_bit(MTIP_DDF_INIT_DONE_BIT, &dd->dd_flag)) + del_gendisk(dd->disk); if (dd->disk->queue) { blk_cleanup_queue(dd->queue); blk_mq_free_tag_set(&dd->tags); @@ -4426,7 +4574,7 @@ static void mtip_pci_remove(struct pci_dev *pdev) struct driver_data *dd = pci_get_drvdata(pdev); unsigned long flags, to; - set_bit(MTIP_DDF_REMOVE_PENDING_BIT, &dd->dd_flag); + set_bit(MTIP_DDF_REMOVAL_BIT, &dd->dd_flag); spin_lock_irqsave(&dev_lock, flags); list_del_init(&dd->online_list); @@ -4443,12 +4591,17 @@ static void mtip_pci_remove(struct pci_dev *pdev) } while (atomic_read(&dd->irq_workers_active) != 0 && time_before(jiffies, to)); + if (!dd->sr) + fsync_bdev(dd->bdev); + if (atomic_read(&dd->irq_workers_active) != 0) { dev_warn(&dd->pdev->dev, "Completion workers still active!\n"); } - blk_mq_stop_hw_queues(dd->queue); + blk_set_queue_dying(dd->queue); + set_bit(MTIP_DDF_REMOVE_PENDING_BIT, &dd->dd_flag); + /* Clean up the block layer. */ mtip_block_remove(dd); diff --git a/drivers/block/mtip32xx/mtip32xx.h b/drivers/block/mtip32xx/mtip32xx.h index 327478400..7617888f7 100644 --- a/drivers/block/mtip32xx/mtip32xx.h +++ b/drivers/block/mtip32xx/mtip32xx.h @@ -134,16 +134,24 @@ enum { MTIP_PF_EH_ACTIVE_BIT = 1, /* error handling */ MTIP_PF_SE_ACTIVE_BIT = 2, /* secure erase */ MTIP_PF_DM_ACTIVE_BIT = 3, /* download microcde */ + MTIP_PF_TO_ACTIVE_BIT = 9, /* timeout handling */ MTIP_PF_PAUSE_IO = ((1 << MTIP_PF_IC_ACTIVE_BIT) | (1 << MTIP_PF_EH_ACTIVE_BIT) | (1 << MTIP_PF_SE_ACTIVE_BIT) | - (1 << MTIP_PF_DM_ACTIVE_BIT)), + (1 << MTIP_PF_DM_ACTIVE_BIT) | + (1 << MTIP_PF_TO_ACTIVE_BIT)), MTIP_PF_SVC_THD_ACTIVE_BIT = 4, MTIP_PF_ISSUE_CMDS_BIT = 5, MTIP_PF_REBUILD_BIT = 6, MTIP_PF_SVC_THD_STOP_BIT = 8, + MTIP_PF_SVC_THD_WORK = ((1 << MTIP_PF_EH_ACTIVE_BIT) | + (1 << MTIP_PF_ISSUE_CMDS_BIT) | + (1 << MTIP_PF_REBUILD_BIT) | + (1 << MTIP_PF_SVC_THD_STOP_BIT) | + (1 << MTIP_PF_TO_ACTIVE_BIT)), + /* below are bit numbers in 'dd_flag' defined in driver_data */ MTIP_DDF_SEC_LOCK_BIT = 0, MTIP_DDF_REMOVE_PENDING_BIT = 1, @@ -153,6 +161,7 @@ enum { MTIP_DDF_RESUME_BIT = 6, MTIP_DDF_INIT_DONE_BIT = 7, MTIP_DDF_REBUILD_FAILED_BIT = 8, + MTIP_DDF_REMOVAL_BIT = 9, MTIP_DDF_STOP_IO = ((1 << MTIP_DDF_REMOVE_PENDING_BIT) | (1 << MTIP_DDF_SEC_LOCK_BIT) | diff --git a/drivers/bluetooth/ath3k.c b/drivers/bluetooth/ath3k.c index cda867ada..79129bf08 100644 --- a/drivers/bluetooth/ath3k.c +++ b/drivers/bluetooth/ath3k.c @@ -82,6 +82,7 @@ static const struct usb_device_id ath3k_table[] = { { USB_DEVICE(0x0489, 0xe05f) }, { USB_DEVICE(0x0489, 0xe076) }, { USB_DEVICE(0x0489, 0xe078) }, + { USB_DEVICE(0x0489, 0xe095) }, { USB_DEVICE(0x04c5, 0x1330) }, { USB_DEVICE(0x04CA, 0x3004) }, { USB_DEVICE(0x04CA, 0x3005) }, @@ -92,6 +93,7 @@ static const struct usb_device_id ath3k_table[] = { { USB_DEVICE(0x04CA, 0x300d) }, { USB_DEVICE(0x04CA, 0x300f) }, { USB_DEVICE(0x04CA, 0x3010) }, + { USB_DEVICE(0x04CA, 0x3014) }, { USB_DEVICE(0x0930, 0x0219) }, { USB_DEVICE(0x0930, 0x021c) }, { USB_DEVICE(0x0930, 0x0220) }, @@ -113,10 +115,12 @@ static const struct usb_device_id ath3k_table[] = { { USB_DEVICE(0x13d3, 0x3362) }, { USB_DEVICE(0x13d3, 0x3375) }, { USB_DEVICE(0x13d3, 0x3393) }, + { USB_DEVICE(0x13d3, 0x3395) }, { USB_DEVICE(0x13d3, 0x3402) }, { USB_DEVICE(0x13d3, 0x3408) }, { USB_DEVICE(0x13d3, 0x3423) }, { USB_DEVICE(0x13d3, 0x3432) }, + { USB_DEVICE(0x13d3, 0x3472) }, { USB_DEVICE(0x13d3, 0x3474) }, /* Atheros AR5BBU12 with sflash firmware */ @@ -144,6 +148,7 @@ static const struct usb_device_id ath3k_blist_tbl[] = { { USB_DEVICE(0x0489, 0xe05f), .driver_info = BTUSB_ATH3012 }, { USB_DEVICE(0x0489, 0xe076), .driver_info = BTUSB_ATH3012 }, { USB_DEVICE(0x0489, 0xe078), .driver_info = BTUSB_ATH3012 }, + { USB_DEVICE(0x0489, 0xe095), .driver_info = BTUSB_ATH3012 }, { USB_DEVICE(0x04c5, 0x1330), .driver_info = BTUSB_ATH3012 }, { USB_DEVICE(0x04ca, 0x3004), .driver_info = BTUSB_ATH3012 }, { USB_DEVICE(0x04ca, 0x3005), .driver_info = BTUSB_ATH3012 }, @@ -154,6 +159,7 @@ static const struct usb_device_id ath3k_blist_tbl[] = { { USB_DEVICE(0x04ca, 0x300d), .driver_info = BTUSB_ATH3012 }, { USB_DEVICE(0x04ca, 0x300f), .driver_info = BTUSB_ATH3012 }, { USB_DEVICE(0x04ca, 0x3010), .driver_info = BTUSB_ATH3012 }, + { USB_DEVICE(0x04ca, 0x3014), .driver_info = BTUSB_ATH3012 }, { USB_DEVICE(0x0930, 0x0219), .driver_info = BTUSB_ATH3012 }, { USB_DEVICE(0x0930, 0x021c), .driver_info = BTUSB_ATH3012 }, { USB_DEVICE(0x0930, 0x0220), .driver_info = BTUSB_ATH3012 }, @@ -175,10 +181,12 @@ static const struct usb_device_id ath3k_blist_tbl[] = { { USB_DEVICE(0x13d3, 0x3362), .driver_info = BTUSB_ATH3012 }, { USB_DEVICE(0x13d3, 0x3375), .driver_info = BTUSB_ATH3012 }, { USB_DEVICE(0x13d3, 0x3393), .driver_info = BTUSB_ATH3012 }, + { USB_DEVICE(0x13d3, 0x3395), .driver_info = BTUSB_ATH3012 }, { USB_DEVICE(0x13d3, 0x3402), .driver_info = BTUSB_ATH3012 }, { USB_DEVICE(0x13d3, 0x3408), .driver_info = BTUSB_ATH3012 }, { USB_DEVICE(0x13d3, 0x3423), .driver_info = BTUSB_ATH3012 }, { USB_DEVICE(0x13d3, 0x3432), .driver_info = BTUSB_ATH3012 }, + { USB_DEVICE(0x13d3, 0x3472), .driver_info = BTUSB_ATH3012 }, { USB_DEVICE(0x13d3, 0x3474), .driver_info = BTUSB_ATH3012 }, /* Atheros AR5BBU22 with sflash firmware */ diff --git a/drivers/bluetooth/btusb.c b/drivers/bluetooth/btusb.c index d897e7454..154b9ecc9 100644 --- a/drivers/bluetooth/btusb.c +++ b/drivers/bluetooth/btusb.c @@ -196,6 +196,7 @@ static const struct usb_device_id blacklist_table[] = { { USB_DEVICE(0x0489, 0xe05f), .driver_info = BTUSB_ATH3012 }, { USB_DEVICE(0x0489, 0xe076), .driver_info = BTUSB_ATH3012 }, { USB_DEVICE(0x0489, 0xe078), .driver_info = BTUSB_ATH3012 }, + { USB_DEVICE(0x0489, 0xe095), .driver_info = BTUSB_ATH3012 }, { USB_DEVICE(0x04c5, 0x1330), .driver_info = BTUSB_ATH3012 }, { USB_DEVICE(0x04ca, 0x3004), .driver_info = BTUSB_ATH3012 }, { USB_DEVICE(0x04ca, 0x3005), .driver_info = BTUSB_ATH3012 }, @@ -206,6 +207,7 @@ static const struct usb_device_id blacklist_table[] = { { USB_DEVICE(0x04ca, 0x300d), .driver_info = BTUSB_ATH3012 }, { USB_DEVICE(0x04ca, 0x300f), .driver_info = BTUSB_ATH3012 }, { USB_DEVICE(0x04ca, 0x3010), .driver_info = BTUSB_ATH3012 }, + { USB_DEVICE(0x04ca, 0x3014), .driver_info = BTUSB_ATH3012 }, { USB_DEVICE(0x0930, 0x0219), .driver_info = BTUSB_ATH3012 }, { USB_DEVICE(0x0930, 0x021c), .driver_info = BTUSB_ATH3012 }, { USB_DEVICE(0x0930, 0x0220), .driver_info = BTUSB_ATH3012 }, @@ -227,10 +229,12 @@ static const struct usb_device_id blacklist_table[] = { { USB_DEVICE(0x13d3, 0x3362), .driver_info = BTUSB_ATH3012 }, { USB_DEVICE(0x13d3, 0x3375), .driver_info = BTUSB_ATH3012 }, { USB_DEVICE(0x13d3, 0x3393), .driver_info = BTUSB_ATH3012 }, + { USB_DEVICE(0x13d3, 0x3395), .driver_info = BTUSB_ATH3012 }, { USB_DEVICE(0x13d3, 0x3402), .driver_info = BTUSB_ATH3012 }, { USB_DEVICE(0x13d3, 0x3408), .driver_info = BTUSB_ATH3012 }, { USB_DEVICE(0x13d3, 0x3423), .driver_info = BTUSB_ATH3012 }, { USB_DEVICE(0x13d3, 0x3432), .driver_info = BTUSB_ATH3012 }, + { USB_DEVICE(0x13d3, 0x3472), .driver_info = BTUSB_ATH3012 }, { USB_DEVICE(0x13d3, 0x3474), .driver_info = BTUSB_ATH3012 }, /* Atheros AR5BBU12 with sflash firmware */ diff --git a/drivers/char/tpm/tpm-chip.c b/drivers/char/tpm/tpm-chip.c index 45cc39aab..252142524 100644 --- a/drivers/char/tpm/tpm-chip.c +++ b/drivers/char/tpm/tpm-chip.c @@ -136,11 +136,13 @@ struct tpm_chip *tpmm_chip_alloc(struct device *dev, chip->cdev.owner = chip->pdev->driver->owner; chip->cdev.kobj.parent = &chip->dev.kobj; + devm_add_action(dev, (void (*)(void *)) put_device, &chip->dev); + return chip; } EXPORT_SYMBOL_GPL(tpmm_chip_alloc); -static int tpm_dev_add_device(struct tpm_chip *chip) +static int tpm_add_char_device(struct tpm_chip *chip) { int rc; @@ -151,7 +153,6 @@ static int tpm_dev_add_device(struct tpm_chip *chip) chip->devname, MAJOR(chip->dev.devt), MINOR(chip->dev.devt), rc); - device_unregister(&chip->dev); return rc; } @@ -162,16 +163,17 @@ static int tpm_dev_add_device(struct tpm_chip *chip) chip->devname, MAJOR(chip->dev.devt), MINOR(chip->dev.devt), rc); + cdev_del(&chip->cdev); return rc; } return rc; } -static void tpm_dev_del_device(struct tpm_chip *chip) +static void tpm_del_char_device(struct tpm_chip *chip) { cdev_del(&chip->cdev); - device_unregister(&chip->dev); + device_del(&chip->dev); } static int tpm1_chip_register(struct tpm_chip *chip) @@ -222,7 +224,7 @@ int tpm_chip_register(struct tpm_chip *chip) tpm_add_ppi(chip); - rc = tpm_dev_add_device(chip); + rc = tpm_add_char_device(chip); if (rc) goto out_err; @@ -274,6 +276,6 @@ void tpm_chip_unregister(struct tpm_chip *chip) sysfs_remove_link(&chip->pdev->kobj, "ppi"); tpm1_chip_unregister(chip); - tpm_dev_del_device(chip); + tpm_del_char_device(chip); } EXPORT_SYMBOL_GPL(tpm_chip_unregister); diff --git a/drivers/char/tpm/tpm_crb.c b/drivers/char/tpm/tpm_crb.c index 8342cf51f..26bab5a29 100644 --- a/drivers/char/tpm/tpm_crb.c +++ b/drivers/char/tpm/tpm_crb.c @@ -302,11 +302,11 @@ static int crb_acpi_remove(struct acpi_device *device) struct device *dev = &device->dev; struct tpm_chip *chip = dev_get_drvdata(dev); - tpm_chip_unregister(chip); - if (chip->flags & TPM_CHIP_FLAG_TPM2) tpm2_shutdown(chip, TPM2_SU_CLEAR); + tpm_chip_unregister(chip); + return 0; } diff --git a/drivers/char/tpm/tpm_eventlog.c b/drivers/char/tpm/tpm_eventlog.c index bd72fb042..4e6940acf 100644 --- a/drivers/char/tpm/tpm_eventlog.c +++ b/drivers/char/tpm/tpm_eventlog.c @@ -232,7 +232,7 @@ static int tpm_binary_bios_measurements_show(struct seq_file *m, void *v) { struct tcpa_event *event = v; struct tcpa_event temp_event; - char *tempPtr; + char *temp_ptr; int i; memcpy(&temp_event, event, sizeof(struct tcpa_event)); @@ -242,10 +242,16 @@ static int tpm_binary_bios_measurements_show(struct seq_file *m, void *v) temp_event.event_type = do_endian_conversion(event->event_type); temp_event.event_size = do_endian_conversion(event->event_size); - tempPtr = (char *)&temp_event; + temp_ptr = (char *) &temp_event; - for (i = 0; i < sizeof(struct tcpa_event) + temp_event.event_size; i++) - seq_putc(m, tempPtr[i]); + for (i = 0; i < (sizeof(struct tcpa_event) - 1) ; i++) + seq_putc(m, temp_ptr[i]); + + temp_ptr = (char *) v; + + for (i = (sizeof(struct tcpa_event) - 1); + i < (sizeof(struct tcpa_event) + temp_event.event_size); i++) + seq_putc(m, temp_ptr[i]); return 0; diff --git a/drivers/clk/bcm/clk-bcm2835.c b/drivers/clk/bcm/clk-bcm2835.c index 015e687ff..9f4df8f64 100644 --- a/drivers/clk/bcm/clk-bcm2835.c +++ b/drivers/clk/bcm/clk-bcm2835.c @@ -1107,13 +1107,15 @@ static int bcm2835_pll_divider_set_rate(struct clk_hw *hw, struct bcm2835_pll_divider *divider = bcm2835_pll_divider_from_hw(hw); struct bcm2835_cprman *cprman = divider->cprman; const struct bcm2835_pll_divider_data *data = divider->data; - u32 cm; - int ret; + u32 cm, div, max_div = 1 << A2W_PLL_DIV_BITS; - ret = clk_divider_ops.set_rate(hw, rate, parent_rate); - if (ret) - return ret; + div = DIV_ROUND_UP_ULL(parent_rate, rate); + + div = min(div, max_div); + if (div == max_div) + div = 0; + cprman_write(cprman, data->a2w_reg, div); cm = cprman_read(cprman, data->cm_reg); cprman_write(cprman, data->cm_reg, cm | data->load_mask); cprman_write(cprman, data->cm_reg, cm & ~data->load_mask); diff --git a/drivers/clk/rockchip/clk-rk3188.c b/drivers/clk/rockchip/clk-rk3188.c index 7f7444cbf..05263571c 100644 --- a/drivers/clk/rockchip/clk-rk3188.c +++ b/drivers/clk/rockchip/clk-rk3188.c @@ -748,6 +748,7 @@ static const char *const rk3188_critical_clocks[] __initconst = { "hclk_peri", "pclk_cpu", "pclk_peri", + "hclk_cpubus" }; static void __init rk3188_common_clk_init(struct device_node *np) diff --git a/drivers/clk/rockchip/clk-rk3368.c b/drivers/clk/rockchip/clk-rk3368.c index 21f3ea909..57acb625c 100644 --- a/drivers/clk/rockchip/clk-rk3368.c +++ b/drivers/clk/rockchip/clk-rk3368.c @@ -165,7 +165,7 @@ static const struct rockchip_cpuclk_reg_data rk3368_cpuclkb_data = { .core_reg = RK3368_CLKSEL_CON(0), .div_core_shift = 0, .div_core_mask = 0x1f, - .mux_core_shift = 15, + .mux_core_shift = 7, }; static const struct rockchip_cpuclk_reg_data rk3368_cpuclkl_data = { @@ -218,29 +218,29 @@ static const struct rockchip_cpuclk_reg_data rk3368_cpuclkl_data = { } static struct rockchip_cpuclk_rate_table rk3368_cpuclkb_rates[] __initdata = { - RK3368_CPUCLKB_RATE(1512000000, 2, 6, 6), - RK3368_CPUCLKB_RATE(1488000000, 2, 5, 5), - RK3368_CPUCLKB_RATE(1416000000, 2, 5, 5), - RK3368_CPUCLKB_RATE(1200000000, 2, 4, 4), - RK3368_CPUCLKB_RATE(1008000000, 2, 4, 4), - RK3368_CPUCLKB_RATE( 816000000, 2, 3, 3), - RK3368_CPUCLKB_RATE( 696000000, 2, 3, 3), - RK3368_CPUCLKB_RATE( 600000000, 2, 2, 2), - RK3368_CPUCLKB_RATE( 408000000, 2, 2, 2), - RK3368_CPUCLKB_RATE( 312000000, 2, 2, 2), + RK3368_CPUCLKB_RATE(1512000000, 1, 5, 5), + RK3368_CPUCLKB_RATE(1488000000, 1, 4, 4), + RK3368_CPUCLKB_RATE(1416000000, 1, 4, 4), + RK3368_CPUCLKB_RATE(1200000000, 1, 3, 3), + RK3368_CPUCLKB_RATE(1008000000, 1, 3, 3), + RK3368_CPUCLKB_RATE( 816000000, 1, 2, 2), + RK3368_CPUCLKB_RATE( 696000000, 1, 2, 2), + RK3368_CPUCLKB_RATE( 600000000, 1, 1, 1), + RK3368_CPUCLKB_RATE( 408000000, 1, 1, 1), + RK3368_CPUCLKB_RATE( 312000000, 1, 1, 1), }; static struct rockchip_cpuclk_rate_table rk3368_cpuclkl_rates[] __initdata = { - RK3368_CPUCLKL_RATE(1512000000, 2, 7, 7), - RK3368_CPUCLKL_RATE(1488000000, 2, 6, 6), - RK3368_CPUCLKL_RATE(1416000000, 2, 6, 6), - RK3368_CPUCLKL_RATE(1200000000, 2, 5, 5), - RK3368_CPUCLKL_RATE(1008000000, 2, 5, 5), - RK3368_CPUCLKL_RATE( 816000000, 2, 4, 4), - RK3368_CPUCLKL_RATE( 696000000, 2, 3, 3), - RK3368_CPUCLKL_RATE( 600000000, 2, 3, 3), - RK3368_CPUCLKL_RATE( 408000000, 2, 2, 2), - RK3368_CPUCLKL_RATE( 312000000, 2, 2, 2), + RK3368_CPUCLKL_RATE(1512000000, 1, 6, 6), + RK3368_CPUCLKL_RATE(1488000000, 1, 5, 5), + RK3368_CPUCLKL_RATE(1416000000, 1, 5, 5), + RK3368_CPUCLKL_RATE(1200000000, 1, 4, 4), + RK3368_CPUCLKL_RATE(1008000000, 1, 4, 4), + RK3368_CPUCLKL_RATE( 816000000, 1, 3, 3), + RK3368_CPUCLKL_RATE( 696000000, 1, 2, 2), + RK3368_CPUCLKL_RATE( 600000000, 1, 2, 2), + RK3368_CPUCLKL_RATE( 408000000, 1, 1, 1), + RK3368_CPUCLKL_RATE( 312000000, 1, 1, 1), }; static struct rockchip_clk_branch rk3368_clk_branches[] __initdata = { @@ -384,10 +384,10 @@ static struct rockchip_clk_branch rk3368_clk_branches[] __initdata = { * Clock-Architecture Diagram 3 */ - COMPOSITE(0, "aclk_vepu", mux_pll_src_cpll_gpll_usb_p, 0, + COMPOSITE(0, "aclk_vepu", mux_pll_src_cpll_gpll_npll_usb_p, 0, RK3368_CLKSEL_CON(15), 6, 2, MFLAGS, 0, 5, DFLAGS, RK3368_CLKGATE_CON(4), 6, GFLAGS), - COMPOSITE(0, "aclk_vdpu", mux_pll_src_cpll_gpll_usb_p, 0, + COMPOSITE(0, "aclk_vdpu", mux_pll_src_cpll_gpll_npll_usb_p, 0, RK3368_CLKSEL_CON(15), 14, 2, MFLAGS, 8, 5, DFLAGS, RK3368_CLKGATE_CON(4), 7, GFLAGS), @@ -442,7 +442,7 @@ static struct rockchip_clk_branch rk3368_clk_branches[] __initdata = { GATE(SCLK_HDMI_HDCP, "sclk_hdmi_hdcp", "xin24m", 0, RK3368_CLKGATE_CON(4), 13, GFLAGS), GATE(SCLK_HDMI_CEC, "sclk_hdmi_cec", "xin32k", 0, - RK3368_CLKGATE_CON(5), 12, GFLAGS), + RK3368_CLKGATE_CON(4), 12, GFLAGS), COMPOSITE_NODIV(0, "vip_src", mux_pll_src_cpll_gpll_p, 0, RK3368_CLKSEL_CON(21), 15, 1, MFLAGS, diff --git a/drivers/cpufreq/cpufreq.c b/drivers/cpufreq/cpufreq.c index e979ec78b..da8b7b9f4 100644 --- a/drivers/cpufreq/cpufreq.c +++ b/drivers/cpufreq/cpufreq.c @@ -25,6 +25,7 @@ #include <linux/kernel_stat.h> #include <linux/module.h> #include <linux/mutex.h> +#include <linux/sched.h> #include <linux/slab.h> #include <linux/suspend.h> #include <linux/syscore_ops.h> @@ -1900,6 +1901,12 @@ int __cpufreq_driver_target(struct cpufreq_policy *policy, } out: + if (likely(retval != -EINVAL)) { + if (target_freq == policy->max) + cpu_nonscaling(policy->cpu); + else + cpu_scaling(policy->cpu); + } return retval; } EXPORT_SYMBOL_GPL(__cpufreq_driver_target); diff --git a/drivers/cpufreq/cpufreq_conservative.c b/drivers/cpufreq/cpufreq_conservative.c index 606ad74ab..ce8a7e14d 100644 --- a/drivers/cpufreq/cpufreq_conservative.c +++ b/drivers/cpufreq/cpufreq_conservative.c @@ -15,8 +15,14 @@ #include "cpufreq_governor.h" /* Conservative governor macros */ +#ifdef CONFIG_SCHED_BFS +#define DEF_FREQUENCY_UP_THRESHOLD (63) +#define DEF_FREQUENCY_DOWN_THRESHOLD (26) +#else #define DEF_FREQUENCY_UP_THRESHOLD (80) #define DEF_FREQUENCY_DOWN_THRESHOLD (20) +#endif + #define DEF_FREQUENCY_STEP (5) #define DEF_SAMPLING_DOWN_FACTOR (1) #define MAX_SAMPLING_DOWN_FACTOR (10) diff --git a/drivers/cpufreq/cpufreq_ondemand.c b/drivers/cpufreq/cpufreq_ondemand.c index 9e6fe9e86..38fbd8771 100644 --- a/drivers/cpufreq/cpufreq_ondemand.c +++ b/drivers/cpufreq/cpufreq_ondemand.c @@ -19,7 +19,12 @@ #include "cpufreq_governor.h" /* On-demand governor macros */ +#ifdef CONFIG_SCHED_BFS +#define DEF_FREQUENCY_UP_THRESHOLD (63) +#else #define DEF_FREQUENCY_UP_THRESHOLD (80) +#endif + #ifdef CONFIG_PCK_INTERACTIVE #define DEF_SAMPLING_DOWN_FACTOR (10) #else diff --git a/drivers/cpufreq/intel_pstate.c b/drivers/cpufreq/intel_pstate.c index cd83d477e..0248282d4 100644 --- a/drivers/cpufreq/intel_pstate.c +++ b/drivers/cpufreq/intel_pstate.c @@ -575,8 +575,13 @@ static void atom_set_pstate(struct cpudata *cpudata, int pstate) vid_fp = clamp_t(int32_t, vid_fp, cpudata->vid.min, cpudata->vid.max); vid = ceiling_fp(vid_fp); - if (pstate > cpudata->pstate.max_pstate) - vid = cpudata->vid.turbo; + if (pstate < cpudata->pstate.max_pstate) + cpu_scaling(cpudata->cpu); + else { + if (pstate > cpudata->pstate.max_pstate) + vid = cpudata->vid.turbo; + cpu_nonscaling(cpudata->cpu); + } val |= vid; diff --git a/drivers/crypto/atmel-aes.c b/drivers/crypto/atmel-aes.c index 3eb3f1279..7de007abe 100644 --- a/drivers/crypto/atmel-aes.c +++ b/drivers/crypto/atmel-aes.c @@ -2085,9 +2085,9 @@ static int atmel_aes_probe(struct platform_device *pdev) } aes_dd->io_base = devm_ioremap_resource(&pdev->dev, aes_res); - if (!aes_dd->io_base) { + if (IS_ERR(aes_dd->io_base)) { dev_err(dev, "can't ioremap\n"); - err = -ENOMEM; + err = PTR_ERR(aes_dd->io_base); goto res_err; } diff --git a/drivers/crypto/atmel-sha.c b/drivers/crypto/atmel-sha.c index 8bf9914d4..68d47a2da 100644 --- a/drivers/crypto/atmel-sha.c +++ b/drivers/crypto/atmel-sha.c @@ -1404,9 +1404,9 @@ static int atmel_sha_probe(struct platform_device *pdev) } sha_dd->io_base = devm_ioremap_resource(&pdev->dev, sha_res); - if (!sha_dd->io_base) { + if (IS_ERR(sha_dd->io_base)) { dev_err(dev, "can't ioremap\n"); - err = -ENOMEM; + err = PTR_ERR(sha_dd->io_base); goto res_err; } diff --git a/drivers/crypto/atmel-tdes.c b/drivers/crypto/atmel-tdes.c index 2c7a628d0..bf467d7be 100644 --- a/drivers/crypto/atmel-tdes.c +++ b/drivers/crypto/atmel-tdes.c @@ -1417,9 +1417,9 @@ static int atmel_tdes_probe(struct platform_device *pdev) } tdes_dd->io_base = devm_ioremap_resource(&pdev->dev, tdes_res); - if (!tdes_dd->io_base) { + if (IS_ERR(tdes_dd->io_base)) { dev_err(dev, "can't ioremap\n"); - err = -ENOMEM; + err = PTR_ERR(tdes_dd->io_base); goto res_err; } diff --git a/drivers/crypto/ccp/ccp-crypto-aes-cmac.c b/drivers/crypto/ccp/ccp-crypto-aes-cmac.c index d89f20c04..3d9acc53d 100644 --- a/drivers/crypto/ccp/ccp-crypto-aes-cmac.c +++ b/drivers/crypto/ccp/ccp-crypto-aes-cmac.c @@ -220,6 +220,39 @@ static int ccp_aes_cmac_digest(struct ahash_request *req) return ccp_aes_cmac_finup(req); } +static int ccp_aes_cmac_export(struct ahash_request *req, void *out) +{ + struct ccp_aes_cmac_req_ctx *rctx = ahash_request_ctx(req); + struct ccp_aes_cmac_exp_ctx state; + + state.null_msg = rctx->null_msg; + memcpy(state.iv, rctx->iv, sizeof(state.iv)); + state.buf_count = rctx->buf_count; + memcpy(state.buf, rctx->buf, sizeof(state.buf)); + + /* 'out' may not be aligned so memcpy from local variable */ + memcpy(out, &state, sizeof(state)); + + return 0; +} + +static int ccp_aes_cmac_import(struct ahash_request *req, const void *in) +{ + struct ccp_aes_cmac_req_ctx *rctx = ahash_request_ctx(req); + struct ccp_aes_cmac_exp_ctx state; + + /* 'in' may not be aligned so memcpy to local variable */ + memcpy(&state, in, sizeof(state)); + + memset(rctx, 0, sizeof(*rctx)); + rctx->null_msg = state.null_msg; + memcpy(rctx->iv, state.iv, sizeof(rctx->iv)); + rctx->buf_count = state.buf_count; + memcpy(rctx->buf, state.buf, sizeof(rctx->buf)); + + return 0; +} + static int ccp_aes_cmac_setkey(struct crypto_ahash *tfm, const u8 *key, unsigned int key_len) { @@ -352,10 +385,13 @@ int ccp_register_aes_cmac_algs(struct list_head *head) alg->final = ccp_aes_cmac_final; alg->finup = ccp_aes_cmac_finup; alg->digest = ccp_aes_cmac_digest; + alg->export = ccp_aes_cmac_export; + alg->import = ccp_aes_cmac_import; alg->setkey = ccp_aes_cmac_setkey; halg = &alg->halg; halg->digestsize = AES_BLOCK_SIZE; + halg->statesize = sizeof(struct ccp_aes_cmac_exp_ctx); base = &halg->base; snprintf(base->cra_name, CRYPTO_MAX_ALG_NAME, "cmac(aes)"); diff --git a/drivers/crypto/ccp/ccp-crypto-sha.c b/drivers/crypto/ccp/ccp-crypto-sha.c index d14b3f28e..8ef06fad8 100644 --- a/drivers/crypto/ccp/ccp-crypto-sha.c +++ b/drivers/crypto/ccp/ccp-crypto-sha.c @@ -207,6 +207,43 @@ static int ccp_sha_digest(struct ahash_request *req) return ccp_sha_finup(req); } +static int ccp_sha_export(struct ahash_request *req, void *out) +{ + struct ccp_sha_req_ctx *rctx = ahash_request_ctx(req); + struct ccp_sha_exp_ctx state; + + state.type = rctx->type; + state.msg_bits = rctx->msg_bits; + state.first = rctx->first; + memcpy(state.ctx, rctx->ctx, sizeof(state.ctx)); + state.buf_count = rctx->buf_count; + memcpy(state.buf, rctx->buf, sizeof(state.buf)); + + /* 'out' may not be aligned so memcpy from local variable */ + memcpy(out, &state, sizeof(state)); + + return 0; +} + +static int ccp_sha_import(struct ahash_request *req, const void *in) +{ + struct ccp_sha_req_ctx *rctx = ahash_request_ctx(req); + struct ccp_sha_exp_ctx state; + + /* 'in' may not be aligned so memcpy to local variable */ + memcpy(&state, in, sizeof(state)); + + memset(rctx, 0, sizeof(*rctx)); + rctx->type = state.type; + rctx->msg_bits = state.msg_bits; + rctx->first = state.first; + memcpy(rctx->ctx, state.ctx, sizeof(rctx->ctx)); + rctx->buf_count = state.buf_count; + memcpy(rctx->buf, state.buf, sizeof(rctx->buf)); + + return 0; +} + static int ccp_sha_setkey(struct crypto_ahash *tfm, const u8 *key, unsigned int key_len) { @@ -403,9 +440,12 @@ static int ccp_register_sha_alg(struct list_head *head, alg->final = ccp_sha_final; alg->finup = ccp_sha_finup; alg->digest = ccp_sha_digest; + alg->export = ccp_sha_export; + alg->import = ccp_sha_import; halg = &alg->halg; halg->digestsize = def->digest_size; + halg->statesize = sizeof(struct ccp_sha_exp_ctx); base = &halg->base; snprintf(base->cra_name, CRYPTO_MAX_ALG_NAME, "%s", def->name); diff --git a/drivers/crypto/ccp/ccp-crypto.h b/drivers/crypto/ccp/ccp-crypto.h index 76a96f0f4..a326ec20b 100644 --- a/drivers/crypto/ccp/ccp-crypto.h +++ b/drivers/crypto/ccp/ccp-crypto.h @@ -129,6 +129,15 @@ struct ccp_aes_cmac_req_ctx { struct ccp_cmd cmd; }; +struct ccp_aes_cmac_exp_ctx { + unsigned int null_msg; + + u8 iv[AES_BLOCK_SIZE]; + + unsigned int buf_count; + u8 buf[AES_BLOCK_SIZE]; +}; + /***** SHA related defines *****/ #define MAX_SHA_CONTEXT_SIZE SHA256_DIGEST_SIZE #define MAX_SHA_BLOCK_SIZE SHA256_BLOCK_SIZE @@ -171,6 +180,19 @@ struct ccp_sha_req_ctx { struct ccp_cmd cmd; }; +struct ccp_sha_exp_ctx { + enum ccp_sha_type type; + + u64 msg_bits; + + unsigned int first; + + u8 ctx[MAX_SHA_CONTEXT_SIZE]; + + unsigned int buf_count; + u8 buf[MAX_SHA_BLOCK_SIZE]; +}; + /***** Common Context Structure *****/ struct ccp_ctx { int (*complete)(struct crypto_async_request *req, int ret); diff --git a/drivers/crypto/marvell/cesa.c b/drivers/crypto/marvell/cesa.c index c0656e7f3..80239ae69 100644 --- a/drivers/crypto/marvell/cesa.c +++ b/drivers/crypto/marvell/cesa.c @@ -420,7 +420,7 @@ static int mv_cesa_probe(struct platform_device *pdev) res = platform_get_resource_byname(pdev, IORESOURCE_MEM, "regs"); cesa->regs = devm_ioremap_resource(dev, res); if (IS_ERR(cesa->regs)) - return -ENOMEM; + return PTR_ERR(cesa->regs); ret = mv_cesa_dev_dma_init(cesa); if (ret) diff --git a/drivers/crypto/ux500/cryp/cryp_core.c b/drivers/crypto/ux500/cryp/cryp_core.c index 4c243c1ff..790f7cadc 100644 --- a/drivers/crypto/ux500/cryp/cryp_core.c +++ b/drivers/crypto/ux500/cryp/cryp_core.c @@ -1440,9 +1440,9 @@ static int ux500_cryp_probe(struct platform_device *pdev) device_data->phybase = res->start; device_data->base = devm_ioremap_resource(dev, res); - if (!device_data->base) { + if (IS_ERR(device_data->base)) { dev_err(dev, "[%s]: ioremap failed!", __func__); - ret = -ENOMEM; + ret = PTR_ERR(device_data->base); goto out; } diff --git a/drivers/crypto/ux500/hash/hash_core.c b/drivers/crypto/ux500/hash/hash_core.c index d6fdc583c..574e87c7f 100644 --- a/drivers/crypto/ux500/hash/hash_core.c +++ b/drivers/crypto/ux500/hash/hash_core.c @@ -1659,9 +1659,9 @@ static int ux500_hash_probe(struct platform_device *pdev) device_data->phybase = res->start; device_data->base = devm_ioremap_resource(dev, res); - if (!device_data->base) { + if (IS_ERR(device_data->base)) { dev_err(dev, "%s: ioremap() failed!\n", __func__); - ret = -ENOMEM; + ret = PTR_ERR(device_data->base); goto out; } spin_lock_init(&device_data->ctx_lock); diff --git a/drivers/edac/amd64_edac.c b/drivers/edac/amd64_edac.c index 9eee13ef8..d87a47547 100644 --- a/drivers/edac/amd64_edac.c +++ b/drivers/edac/amd64_edac.c @@ -1452,7 +1452,7 @@ static u64 f1x_get_norm_dct_addr(struct amd64_pvt *pvt, u8 range, u64 chan_off; u64 dram_base = get_dram_base(pvt, range); u64 hole_off = f10_dhar_offset(pvt); - u64 dct_sel_base_off = (pvt->dct_sel_hi & 0xFFFFFC00) << 16; + u64 dct_sel_base_off = (u64)(pvt->dct_sel_hi & 0xFFFFFC00) << 16; if (hi_rng) { /* diff --git a/drivers/edac/sb_edac.c b/drivers/edac/sb_edac.c index f5c6b97c8..93f0d4120 100644 --- a/drivers/edac/sb_edac.c +++ b/drivers/edac/sb_edac.c @@ -1839,8 +1839,8 @@ static void get_memory_layout(const struct mem_ctl_info *mci) edac_dbg(0, "TAD#%d: up to %u.%03u GB (0x%016Lx), socket interleave %d, memory interleave %d, TGT: %d, %d, %d, %d, reg=0x%08x\n", n_tads, gb, (mb*1000)/1024, ((u64)tmp_mb) << 20L, - (u32)TAD_SOCK(reg), - (u32)TAD_CH(reg), + (u32)(1 << TAD_SOCK(reg)), + (u32)TAD_CH(reg) + 1, (u32)TAD_TGT0(reg), (u32)TAD_TGT1(reg), (u32)TAD_TGT2(reg), @@ -2118,7 +2118,7 @@ static int get_memory_error_data(struct mem_ctl_info *mci, } ch_way = TAD_CH(reg) + 1; - sck_way = TAD_SOCK(reg) + 1; + sck_way = 1 << TAD_SOCK(reg); if (ch_way == 3) idx = addr >> 6; @@ -2175,7 +2175,7 @@ static int get_memory_error_data(struct mem_ctl_info *mci, n_tads, addr, limit, - (u32)TAD_SOCK(reg), + sck_way, ch_way, offset, idx, @@ -2190,18 +2190,12 @@ static int get_memory_error_data(struct mem_ctl_info *mci, offset, addr); return -EINVAL; } - addr -= offset; - /* Store the low bits [0:6] of the addr */ - ch_addr = addr & 0x7f; - /* Remove socket wayness and remove 6 bits */ - addr >>= 6; - addr = div_u64(addr, sck_xch); -#if 0 - /* Divide by channel way */ - addr = addr / ch_way; -#endif - /* Recover the last 6 bits */ - ch_addr |= addr << 6; + + ch_addr = addr - offset; + ch_addr >>= (6 + shiftup); + ch_addr /= ch_way * sck_way; + ch_addr <<= (6 + shiftup); + ch_addr |= addr & ((1 << (6 + shiftup)) - 1); /* * Step 3) Decode rank diff --git a/drivers/firmware/broadcom/bcm47xx_nvram.c b/drivers/firmware/broadcom/bcm47xx_nvram.c index 0c2f0a61b..0b631e5b5 100644 --- a/drivers/firmware/broadcom/bcm47xx_nvram.c +++ b/drivers/firmware/broadcom/bcm47xx_nvram.c @@ -94,15 +94,14 @@ static int nvram_find_and_copy(void __iomem *iobase, u32 lim) found: __ioread32_copy(nvram_buf, header, sizeof(*header) / 4); - header = (struct nvram_header *)nvram_buf; - nvram_len = header->len; + nvram_len = ((struct nvram_header *)(nvram_buf))->len; if (nvram_len > size) { pr_err("The nvram size according to the header seems to be bigger than the partition on flash\n"); nvram_len = size; } if (nvram_len >= NVRAM_SPACE) { pr_err("nvram on flash (%i bytes) is bigger than the reserved space in memory, will just copy the first %i bytes\n", - header->len, NVRAM_SPACE - 1); + nvram_len, NVRAM_SPACE - 1); nvram_len = NVRAM_SPACE - 1; } /* proceed reading data after header */ diff --git a/drivers/gpio/gpio-pca953x.c b/drivers/gpio/gpio-pca953x.c index 23196c5fc..99b375c95 100644 --- a/drivers/gpio/gpio-pca953x.c +++ b/drivers/gpio/gpio-pca953x.c @@ -367,9 +367,11 @@ static void pca953x_gpio_set_multiple(struct gpio_chip *gc, memcpy(reg_val, chip->reg_output, NBANK(chip)); mutex_lock(&chip->i2c_lock); for(bank=0; bank<NBANK(chip); bank++) { - unsigned bankmask = mask[bank/4] >> ((bank % 4) * 8); + unsigned bankmask = mask[bank / sizeof(*mask)] >> + ((bank % sizeof(*mask)) * 8); if(bankmask) { - unsigned bankval = bits[bank/4] >> ((bank % 4) * 8); + unsigned bankval = bits[bank / sizeof(*bits)] >> + ((bank % sizeof(*bits)) * 8); reg_val[bank] = (reg_val[bank] & ~bankmask) | bankval; } } diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_atpx_handler.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_atpx_handler.c index 3c895863f..81dc6b654 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_atpx_handler.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_atpx_handler.c @@ -63,6 +63,10 @@ bool amdgpu_has_atpx(void) { return amdgpu_atpx_priv.atpx_detected; } +bool amdgpu_has_atpx_dgpu_power_cntl(void) { + return amdgpu_atpx_priv.atpx.functions.power_cntl; +} + /** * amdgpu_atpx_call - call an ATPX method * @@ -142,10 +146,6 @@ static void amdgpu_atpx_parse_functions(struct amdgpu_atpx_functions *f, u32 mas */ static int amdgpu_atpx_validate(struct amdgpu_atpx *atpx) { - /* make sure required functions are enabled */ - /* dGPU power control is required */ - atpx->functions.power_cntl = true; - if (atpx->functions.px_params) { union acpi_object *info; struct atpx_px_params output; diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c index 51bfc1145..d6c68d00c 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c @@ -62,6 +62,12 @@ static const char *amdgpu_asic_name[] = { "LAST", }; +#if defined(CONFIG_VGA_SWITCHEROO) +bool amdgpu_has_atpx_dgpu_power_cntl(void); +#else +static inline bool amdgpu_has_atpx_dgpu_power_cntl(void) { return false; } +#endif + bool amdgpu_device_is_px(struct drm_device *dev) { struct amdgpu_device *adev = dev->dev_private; @@ -1511,7 +1517,7 @@ int amdgpu_device_init(struct amdgpu_device *adev, if (amdgpu_runtime_pm == 1) runtime = true; - if (amdgpu_device_is_px(ddev)) + if (amdgpu_device_is_px(ddev) && amdgpu_has_atpx_dgpu_power_cntl()) runtime = true; vga_switcheroo_register_client(adev->pdev, &amdgpu_switcheroo_ops, runtime); if (runtime) diff --git a/drivers/gpu/drm/amd/amdgpu/sdma_v2_4.c b/drivers/gpu/drm/amd/amdgpu/sdma_v2_4.c index ed90e3b64..8c01e1565 100644 --- a/drivers/gpu/drm/amd/amdgpu/sdma_v2_4.c +++ b/drivers/gpu/drm/amd/amdgpu/sdma_v2_4.c @@ -32,8 +32,8 @@ #include "oss/oss_2_4_d.h" #include "oss/oss_2_4_sh_mask.h" -#include "gmc/gmc_8_1_d.h" -#include "gmc/gmc_8_1_sh_mask.h" +#include "gmc/gmc_7_1_d.h" +#include "gmc/gmc_7_1_sh_mask.h" #include "gca/gfx_8_0_d.h" #include "gca/gfx_8_0_enum.h" diff --git a/drivers/gpu/drm/amd/powerplay/hwmgr/cz_hwmgr.c b/drivers/gpu/drm/amd/powerplay/hwmgr/cz_hwmgr.c index cf01177ca..2ea012e88 100644 --- a/drivers/gpu/drm/amd/powerplay/hwmgr/cz_hwmgr.c +++ b/drivers/gpu/drm/amd/powerplay/hwmgr/cz_hwmgr.c @@ -241,6 +241,11 @@ static int cz_initialize_dpm_defaults(struct pp_hwmgr *hwmgr) phm_cap_set(hwmgr->platform_descriptor.platformCaps, PHM_PlatformCaps_DynamicUVDState); + phm_cap_set(hwmgr->platform_descriptor.platformCaps, + PHM_PlatformCaps_UVDDPM); + phm_cap_set(hwmgr->platform_descriptor.platformCaps, + PHM_PlatformCaps_VCEDPM); + cz_hwmgr->cc6_settings.cpu_cc6_disable = false; cz_hwmgr->cc6_settings.cpu_pstate_disable = false; cz_hwmgr->cc6_settings.nb_pstate_switch_disable = false; diff --git a/drivers/gpu/drm/radeon/atombios_encoders.c b/drivers/gpu/drm/radeon/atombios_encoders.c index 01b20e14a..6104d7d74 100644 --- a/drivers/gpu/drm/radeon/atombios_encoders.c +++ b/drivers/gpu/drm/radeon/atombios_encoders.c @@ -892,8 +892,6 @@ atombios_dig_encoder_setup2(struct drm_encoder *encoder, int action, int panel_m else args.v1.ucLaneNum = 4; - if (ENCODER_MODE_IS_DP(args.v1.ucEncoderMode) && (dp_clock == 270000)) - args.v1.ucConfig |= ATOM_ENCODER_CONFIG_DPLINKRATE_2_70GHZ; switch (radeon_encoder->encoder_id) { case ENCODER_OBJECT_ID_INTERNAL_UNIPHY: args.v1.ucConfig = ATOM_ENCODER_CONFIG_V2_TRANSMITTER1; @@ -910,6 +908,10 @@ atombios_dig_encoder_setup2(struct drm_encoder *encoder, int action, int panel_m args.v1.ucConfig |= ATOM_ENCODER_CONFIG_LINKB; else args.v1.ucConfig |= ATOM_ENCODER_CONFIG_LINKA; + + if (ENCODER_MODE_IS_DP(args.v1.ucEncoderMode) && (dp_clock == 270000)) + args.v1.ucConfig |= ATOM_ENCODER_CONFIG_DPLINKRATE_2_70GHZ; + break; case 2: case 3: diff --git a/drivers/gpu/drm/radeon/radeon_atpx_handler.c b/drivers/gpu/drm/radeon/radeon_atpx_handler.c index c4b4f298a..9bc408c9f 100644 --- a/drivers/gpu/drm/radeon/radeon_atpx_handler.c +++ b/drivers/gpu/drm/radeon/radeon_atpx_handler.c @@ -62,6 +62,10 @@ bool radeon_has_atpx(void) { return radeon_atpx_priv.atpx_detected; } +bool radeon_has_atpx_dgpu_power_cntl(void) { + return radeon_atpx_priv.atpx.functions.power_cntl; +} + /** * radeon_atpx_call - call an ATPX method * @@ -141,10 +145,6 @@ static void radeon_atpx_parse_functions(struct radeon_atpx_functions *f, u32 mas */ static int radeon_atpx_validate(struct radeon_atpx *atpx) { - /* make sure required functions are enabled */ - /* dGPU power control is required */ - atpx->functions.power_cntl = true; - if (atpx->functions.px_params) { union acpi_object *info; struct atpx_px_params output; diff --git a/drivers/gpu/drm/radeon/radeon_device.c b/drivers/gpu/drm/radeon/radeon_device.c index 4197ca1bb..e2396336f 100644 --- a/drivers/gpu/drm/radeon/radeon_device.c +++ b/drivers/gpu/drm/radeon/radeon_device.c @@ -103,6 +103,12 @@ static const char radeon_family_name[][16] = { "LAST", }; +#if defined(CONFIG_VGA_SWITCHEROO) +bool radeon_has_atpx_dgpu_power_cntl(void); +#else +static inline bool radeon_has_atpx_dgpu_power_cntl(void) { return false; } +#endif + #define RADEON_PX_QUIRK_DISABLE_PX (1 << 0) #define RADEON_PX_QUIRK_LONG_WAKEUP (1 << 1) @@ -1433,7 +1439,7 @@ int radeon_device_init(struct radeon_device *rdev, * ignore it */ vga_client_register(rdev->pdev, rdev, NULL, radeon_vga_set_decode); - if (rdev->flags & RADEON_IS_PX) + if ((rdev->flags & RADEON_IS_PX) && radeon_has_atpx_dgpu_power_cntl()) runtime = true; vga_switcheroo_register_client(rdev->pdev, &radeon_switcheroo_ops, runtime); if (runtime) diff --git a/drivers/gpu/drm/radeon/radeon_display.c b/drivers/gpu/drm/radeon/radeon_display.c index 2d9196a44..bfcef4db8 100644 --- a/drivers/gpu/drm/radeon/radeon_display.c +++ b/drivers/gpu/drm/radeon/radeon_display.c @@ -1683,10 +1683,8 @@ int radeon_modeset_init(struct radeon_device *rdev) /* setup afmt */ radeon_afmt_init(rdev); - if (!list_empty(&rdev->ddev->mode_config.connector_list)) { - radeon_fbdev_init(rdev); - drm_kms_helper_poll_init(rdev->ddev); - } + radeon_fbdev_init(rdev); + drm_kms_helper_poll_init(rdev->ddev); /* do pm late init */ ret = radeon_pm_late_init(rdev); diff --git a/drivers/gpu/drm/radeon/radeon_dp_mst.c b/drivers/gpu/drm/radeon/radeon_dp_mst.c index df7a1719c..9d210bbca 100644 --- a/drivers/gpu/drm/radeon/radeon_dp_mst.c +++ b/drivers/gpu/drm/radeon/radeon_dp_mst.c @@ -525,17 +525,9 @@ static bool radeon_mst_mode_fixup(struct drm_encoder *encoder, drm_mode_set_crtcinfo(adjusted_mode, 0); { struct radeon_connector_atom_dig *dig_connector; - int ret; - dig_connector = mst_enc->connector->con_priv; - ret = radeon_dp_get_dp_link_config(&mst_enc->connector->base, - dig_connector->dpcd, adjusted_mode->clock, - &dig_connector->dp_lane_count, - &dig_connector->dp_clock); - if (ret) { - dig_connector->dp_lane_count = 0; - dig_connector->dp_clock = 0; - } + dig_connector->dp_lane_count = drm_dp_max_lane_count(dig_connector->dpcd); + dig_connector->dp_clock = drm_dp_max_link_rate(dig_connector->dpcd); DRM_DEBUG_KMS("dig clock %p %d %d\n", dig_connector, dig_connector->dp_lane_count, dig_connector->dp_clock); } diff --git a/drivers/gpu/drm/radeon/radeon_fb.c b/drivers/gpu/drm/radeon/radeon_fb.c index d2e628eea..d17959633 100644 --- a/drivers/gpu/drm/radeon/radeon_fb.c +++ b/drivers/gpu/drm/radeon/radeon_fb.c @@ -292,7 +292,8 @@ out_unref: void radeon_fb_output_poll_changed(struct radeon_device *rdev) { - drm_fb_helper_hotplug_event(&rdev->mode_info.rfbdev->helper); + if (rdev->mode_info.rfbdev) + drm_fb_helper_hotplug_event(&rdev->mode_info.rfbdev->helper); } static int radeon_fbdev_destroy(struct drm_device *dev, struct radeon_fbdev *rfbdev) @@ -325,6 +326,10 @@ int radeon_fbdev_init(struct radeon_device *rdev) int bpp_sel = 32; int ret; + /* don't enable fbdev if no connectors */ + if (list_empty(&rdev->ddev->mode_config.connector_list)) + return 0; + /* select 8 bpp console on RN50 or 16MB cards */ if (ASIC_IS_RN50(rdev) || rdev->mc.real_vram_size <= (32*1024*1024)) bpp_sel = 8; @@ -377,11 +382,15 @@ void radeon_fbdev_fini(struct radeon_device *rdev) void radeon_fbdev_set_suspend(struct radeon_device *rdev, int state) { - fb_set_suspend(rdev->mode_info.rfbdev->helper.fbdev, state); + if (rdev->mode_info.rfbdev) + fb_set_suspend(rdev->mode_info.rfbdev->helper.fbdev, state); } bool radeon_fbdev_robj_is_fb(struct radeon_device *rdev, struct radeon_bo *robj) { + if (!rdev->mode_info.rfbdev) + return false; + if (robj == gem_to_radeon_bo(rdev->mode_info.rfbdev->rfb.obj)) return true; return false; @@ -389,12 +398,14 @@ bool radeon_fbdev_robj_is_fb(struct radeon_device *rdev, struct radeon_bo *robj) void radeon_fb_add_connector(struct radeon_device *rdev, struct drm_connector *connector) { - drm_fb_helper_add_one_connector(&rdev->mode_info.rfbdev->helper, connector); + if (rdev->mode_info.rfbdev) + drm_fb_helper_add_one_connector(&rdev->mode_info.rfbdev->helper, connector); } void radeon_fb_remove_connector(struct radeon_device *rdev, struct drm_connector *connector) { - drm_fb_helper_remove_one_connector(&rdev->mode_info.rfbdev->helper, connector); + if (rdev->mode_info.rfbdev) + drm_fb_helper_remove_one_connector(&rdev->mode_info.rfbdev->helper, connector); } void radeon_fbdev_restore_mode(struct radeon_device *rdev) diff --git a/drivers/gpu/drm/vc4/vc4_bo.c b/drivers/gpu/drm/vc4/vc4_bo.c index 22278bcfc..ac8eafea6 100644 --- a/drivers/gpu/drm/vc4/vc4_bo.c +++ b/drivers/gpu/drm/vc4/vc4_bo.c @@ -499,11 +499,12 @@ vc4_create_shader_bo_ioctl(struct drm_device *dev, void *data, if (IS_ERR(bo)) return PTR_ERR(bo); - ret = copy_from_user(bo->base.vaddr, + if (copy_from_user(bo->base.vaddr, (void __user *)(uintptr_t)args->data, - args->size); - if (ret != 0) + args->size)) { + ret = -EFAULT; goto fail; + } /* Clear the rest of the memory from allocating from the BO * cache. */ diff --git a/drivers/hid/hid-core.c b/drivers/hid/hid-core.c index 7e89288b1..99446ffd7 100644 --- a/drivers/hid/hid-core.c +++ b/drivers/hid/hid-core.c @@ -1891,6 +1891,7 @@ static const struct hid_device_id hid_have_special_driver[] = { { HID_USB_DEVICE(USB_VENDOR_ID_LOGITECH, USB_DEVICE_ID_LOGITECH_ELITE_KBD) }, { HID_USB_DEVICE(USB_VENDOR_ID_LOGITECH, USB_DEVICE_ID_LOGITECH_CORDLESS_DESKTOP_LX500) }, { HID_USB_DEVICE(USB_VENDOR_ID_LOGITECH, USB_DEVICE_ID_LOGITECH_EXTREME_3D) }, + { HID_USB_DEVICE(USB_VENDOR_ID_LOGITECH, USB_DEVICE_ID_LOGITECH_DUAL_ACTION) }, { HID_USB_DEVICE(USB_VENDOR_ID_LOGITECH, USB_DEVICE_ID_LOGITECH_WHEEL) }, { HID_USB_DEVICE(USB_VENDOR_ID_LOGITECH, USB_DEVICE_ID_LOGITECH_RUMBLEPAD_CORD) }, { HID_USB_DEVICE(USB_VENDOR_ID_LOGITECH, USB_DEVICE_ID_LOGITECH_RUMBLEPAD) }, @@ -2615,9 +2616,10 @@ int hid_add_device(struct hid_device *hdev) /* * Scan generic devices for group information */ - if (hid_ignore_special_drivers || - (!hdev->group && - !hid_match_id(hdev, hid_have_special_driver))) { + if (hid_ignore_special_drivers) { + hdev->group = HID_GROUP_GENERIC; + } else if (!hdev->group && + !hid_match_id(hdev, hid_have_special_driver)) { ret = hid_scan_report(hdev); if (ret) hid_warn(hdev, "bad device descriptor (%d)\n", ret); diff --git a/drivers/hid/hid-multitouch.c b/drivers/hid/hid-multitouch.c index 296d49915..a20fc604f 100644 --- a/drivers/hid/hid-multitouch.c +++ b/drivers/hid/hid-multitouch.c @@ -396,6 +396,11 @@ static void mt_feature_mapping(struct hid_device *hdev, td->is_buttonpad = true; break; + case 0xff0000c5: + /* Retrieve the Win8 blob once to enable some devices */ + if (usage->usage_index == 0) + mt_get_feature(hdev, field->report); + break; } } diff --git a/drivers/hid/i2c-hid/i2c-hid.c b/drivers/hid/i2c-hid/i2c-hid.c index b9216938a..bb897497f 100644 --- a/drivers/hid/i2c-hid/i2c-hid.c +++ b/drivers/hid/i2c-hid/i2c-hid.c @@ -283,17 +283,21 @@ static int i2c_hid_set_or_send_report(struct i2c_client *client, u8 reportType, u16 dataRegister = le16_to_cpu(ihid->hdesc.wDataRegister); u16 outputRegister = le16_to_cpu(ihid->hdesc.wOutputRegister); u16 maxOutputLength = le16_to_cpu(ihid->hdesc.wMaxOutputLength); + u16 size; + int args_len; + int index = 0; + + i2c_hid_dbg(ihid, "%s\n", __func__); + + if (data_len > ihid->bufsize) + return -EINVAL; - /* hid_hw_* already checked that data_len < HID_MAX_BUFFER_SIZE */ - u16 size = 2 /* size */ + + size = 2 /* size */ + (reportID ? 1 : 0) /* reportID */ + data_len /* buf */; - int args_len = (reportID >= 0x0F ? 1 : 0) /* optional third byte */ + + args_len = (reportID >= 0x0F ? 1 : 0) /* optional third byte */ + 2 /* dataRegister */ + size /* args */; - int index = 0; - - i2c_hid_dbg(ihid, "%s\n", __func__); if (!use_data && maxOutputLength == 0) return -ENOSYS; diff --git a/drivers/idle/intel_idle.c b/drivers/idle/intel_idle.c index cd4510a63..146eed70b 100644 --- a/drivers/idle/intel_idle.c +++ b/drivers/idle/intel_idle.c @@ -65,7 +65,7 @@ #include <asm/mwait.h> #include <asm/msr.h> -#define INTEL_IDLE_VERSION "0.4" +#define INTEL_IDLE_VERSION "0.4.1" #define PREFIX "intel_idle: " static struct cpuidle_driver intel_idle_driver = { @@ -994,36 +994,92 @@ static void intel_idle_cpuidle_devices_uninit(void) } /* - * intel_idle_state_table_update() - * - * Update the default state_table for this CPU-id + * ivt_idle_state_table_update(void) * - * Currently used to access tuned IVT multi-socket targets + * Tune IVT multi-socket targets * Assumption: num_sockets == (max_package_num + 1) */ -void intel_idle_state_table_update(void) +static void ivt_idle_state_table_update(void) { /* IVT uses a different table for 1-2, 3-4, and > 4 sockets */ - if (boot_cpu_data.x86_model == 0x3e) { /* IVT */ - int cpu, package_num, num_sockets = 1; - - for_each_online_cpu(cpu) { - package_num = topology_physical_package_id(cpu); - if (package_num + 1 > num_sockets) { - num_sockets = package_num + 1; - - if (num_sockets > 4) { - cpuidle_state_table = ivt_cstates_8s; - return; - } + int cpu, package_num, num_sockets = 1; + + for_each_online_cpu(cpu) { + package_num = topology_physical_package_id(cpu); + if (package_num + 1 > num_sockets) { + num_sockets = package_num + 1; + + if (num_sockets > 4) { + cpuidle_state_table = ivt_cstates_8s; + return; } } + } + + if (num_sockets > 2) + cpuidle_state_table = ivt_cstates_4s; + + /* else, 1 and 2 socket systems use default ivt_cstates */ +} +/* + * sklh_idle_state_table_update(void) + * + * On SKL-H (model 0x5e) disable C8 and C9 if: + * C10 is enabled and SGX disabled + */ +static void sklh_idle_state_table_update(void) +{ + unsigned long long msr; + unsigned int eax, ebx, ecx, edx; + + + /* if PC10 disabled via cmdline intel_idle.max_cstate=7 or shallower */ + if (max_cstate <= 7) + return; + + /* if PC10 not present in CPUID.MWAIT.EDX */ + if ((mwait_substates & (0xF << 28)) == 0) + return; + + rdmsrl(MSR_NHM_SNB_PKG_CST_CFG_CTL, msr); + + /* PC10 is not enabled in PKG C-state limit */ + if ((msr & 0xF) != 8) + return; + + ecx = 0; + cpuid(7, &eax, &ebx, &ecx, &edx); + + /* if SGX is present */ + if (ebx & (1 << 2)) { - if (num_sockets > 2) - cpuidle_state_table = ivt_cstates_4s; - /* else, 1 and 2 socket systems use default ivt_cstates */ + rdmsrl(MSR_IA32_FEATURE_CONTROL, msr); + + /* if SGX is enabled */ + if (msr & (1 << 18)) + return; + } + + skl_cstates[5].disabled = 1; /* C8-SKL */ + skl_cstates[6].disabled = 1; /* C9-SKL */ +} +/* + * intel_idle_state_table_update() + * + * Update the default state_table for this CPU-id + */ + +static void intel_idle_state_table_update(void) +{ + switch (boot_cpu_data.x86_model) { + + case 0x3e: /* IVT */ + ivt_idle_state_table_update(); + break; + case 0x5e: /* SKL-H */ + sklh_idle_state_table_update(); + break; } - return; } /* @@ -1063,6 +1119,14 @@ static int __init intel_idle_cpuidle_driver_init(void) if (num_substates == 0) continue; + /* if state marked as disabled, skip it */ + if (cpuidle_state_table[cstate].disabled != 0) { + pr_debug(PREFIX "state %s is disabled", + cpuidle_state_table[cstate].name); + continue; + } + + if (((mwait_cstate + 1) > 2) && !boot_cpu_has(X86_FEATURE_NONSTOP_TSC)) mark_tsc_unstable("TSC halts in idle" diff --git a/drivers/infiniband/ulp/isert/ib_isert.c b/drivers/infiniband/ulp/isert/ib_isert.c index f121e6129..0e1a802c3 100644 --- a/drivers/infiniband/ulp/isert/ib_isert.c +++ b/drivers/infiniband/ulp/isert/ib_isert.c @@ -65,6 +65,7 @@ isert_rdma_accept(struct isert_conn *isert_conn); struct rdma_cm_id *isert_setup_id(struct isert_np *isert_np); static void isert_release_work(struct work_struct *work); +static void isert_wait4flush(struct isert_conn *isert_conn); static inline bool isert_prot_cmd(struct isert_conn *conn, struct se_cmd *cmd) @@ -820,12 +821,31 @@ isert_put_conn(struct isert_conn *isert_conn) kref_put(&isert_conn->kref, isert_release_kref); } +static void +isert_handle_unbound_conn(struct isert_conn *isert_conn) +{ + struct isert_np *isert_np = isert_conn->cm_id->context; + + mutex_lock(&isert_np->mutex); + if (!list_empty(&isert_conn->node)) { + /* + * This means iscsi doesn't know this connection + * so schedule a cleanup ourselves + */ + list_del_init(&isert_conn->node); + isert_put_conn(isert_conn); + complete(&isert_conn->wait); + queue_work(isert_release_wq, &isert_conn->release_work); + } + mutex_unlock(&isert_np->mutex); +} + /** * isert_conn_terminate() - Initiate connection termination * @isert_conn: isert connection struct * * Notes: - * In case the connection state is FULL_FEATURE, move state + * In case the connection state is BOUND, move state * to TEMINATING and start teardown sequence (rdma_disconnect). * In case the connection state is UP, complete flush as well. * @@ -837,23 +857,19 @@ isert_conn_terminate(struct isert_conn *isert_conn) { int err; - switch (isert_conn->state) { - case ISER_CONN_TERMINATING: - break; - case ISER_CONN_UP: - case ISER_CONN_FULL_FEATURE: /* FALLTHRU */ - isert_info("Terminating conn %p state %d\n", - isert_conn, isert_conn->state); - isert_conn->state = ISER_CONN_TERMINATING; - err = rdma_disconnect(isert_conn->cm_id); - if (err) - isert_warn("Failed rdma_disconnect isert_conn %p\n", - isert_conn); - break; - default: - isert_warn("conn %p teminating in state %d\n", - isert_conn, isert_conn->state); - } + if (isert_conn->state >= ISER_CONN_TERMINATING) + return; + + isert_info("Terminating conn %p state %d\n", + isert_conn, isert_conn->state); + isert_conn->state = ISER_CONN_TERMINATING; + err = rdma_disconnect(isert_conn->cm_id); + if (err) + isert_warn("Failed rdma_disconnect isert_conn %p\n", + isert_conn); + + isert_info("conn %p completing wait\n", isert_conn); + complete(&isert_conn->wait); } static int @@ -887,35 +903,27 @@ static int isert_disconnected_handler(struct rdma_cm_id *cma_id, enum rdma_cm_event_type event) { - struct isert_np *isert_np = cma_id->context; - struct isert_conn *isert_conn; - bool terminating = false; - - if (isert_np->cm_id == cma_id) - return isert_np_cma_handler(cma_id->context, event); - - isert_conn = cma_id->qp->qp_context; + struct isert_conn *isert_conn = cma_id->qp->qp_context; mutex_lock(&isert_conn->mutex); - terminating = (isert_conn->state == ISER_CONN_TERMINATING); - isert_conn_terminate(isert_conn); - mutex_unlock(&isert_conn->mutex); - - isert_info("conn %p completing wait\n", isert_conn); - complete(&isert_conn->wait); - - if (terminating) - goto out; - - mutex_lock(&isert_np->mutex); - if (!list_empty(&isert_conn->node)) { - list_del_init(&isert_conn->node); - isert_put_conn(isert_conn); - queue_work(isert_release_wq, &isert_conn->release_work); + switch (isert_conn->state) { + case ISER_CONN_TERMINATING: + break; + case ISER_CONN_UP: + isert_conn_terminate(isert_conn); + isert_wait4flush(isert_conn); + isert_handle_unbound_conn(isert_conn); + break; + case ISER_CONN_BOUND: + case ISER_CONN_FULL_FEATURE: /* FALLTHRU */ + iscsit_cause_connection_reinstatement(isert_conn->conn, 0); + break; + default: + isert_warn("conn %p teminating in state %d\n", + isert_conn, isert_conn->state); } - mutex_unlock(&isert_np->mutex); + mutex_unlock(&isert_conn->mutex); -out: return 0; } @@ -934,12 +942,16 @@ isert_connect_error(struct rdma_cm_id *cma_id) static int isert_cma_handler(struct rdma_cm_id *cma_id, struct rdma_cm_event *event) { + struct isert_np *isert_np = cma_id->context; int ret = 0; isert_info("%s (%d): status %d id %p np %p\n", rdma_event_msg(event->event), event->event, event->status, cma_id, cma_id->context); + if (isert_np->cm_id == cma_id) + return isert_np_cma_handler(cma_id->context, event->event); + switch (event->event) { case RDMA_CM_EVENT_CONNECT_REQUEST: ret = isert_connect_request(cma_id, event); @@ -985,13 +997,10 @@ isert_post_recvm(struct isert_conn *isert_conn, u32 count) rx_wr--; rx_wr->next = NULL; /* mark end of work requests list */ - isert_conn->post_recv_buf_count += count; ret = ib_post_recv(isert_conn->qp, isert_conn->rx_wr, &rx_wr_failed); - if (ret) { + if (ret) isert_err("ib_post_recv() failed with ret: %d\n", ret); - isert_conn->post_recv_buf_count -= count; - } return ret; } @@ -1007,12 +1016,9 @@ isert_post_recv(struct isert_conn *isert_conn, struct iser_rx_desc *rx_desc) rx_wr.num_sge = 1; rx_wr.next = NULL; - isert_conn->post_recv_buf_count++; ret = ib_post_recv(isert_conn->qp, &rx_wr, &rx_wr_failed); - if (ret) { + if (ret) isert_err("ib_post_recv() failed with ret: %d\n", ret); - isert_conn->post_recv_buf_count--; - } return ret; } @@ -1132,12 +1138,9 @@ isert_rdma_post_recvl(struct isert_conn *isert_conn) rx_wr.sg_list = &sge; rx_wr.num_sge = 1; - isert_conn->post_recv_buf_count++; ret = ib_post_recv(isert_conn->qp, &rx_wr, &rx_wr_fail); - if (ret) { + if (ret) isert_err("ib_post_recv() failed: %d\n", ret); - isert_conn->post_recv_buf_count--; - } return ret; } @@ -1633,7 +1636,6 @@ isert_rcv_completion(struct iser_rx_desc *desc, ib_dma_sync_single_for_device(ib_dev, rx_dma, rx_buflen, DMA_FROM_DEVICE); - isert_conn->post_recv_buf_count--; } static int @@ -2048,7 +2050,8 @@ is_isert_tx_desc(struct isert_conn *isert_conn, void *wr_id) void *start = isert_conn->rx_descs; int len = ISERT_QP_MAX_RECV_DTOS * sizeof(*isert_conn->rx_descs); - if (wr_id >= start && wr_id < start + len) + if ((wr_id >= start && wr_id < start + len) || + (wr_id == isert_conn->login_req_buf)) return false; return true; @@ -2072,10 +2075,6 @@ isert_cq_comp_err(struct isert_conn *isert_conn, struct ib_wc *wc) isert_unmap_tx_desc(desc, ib_dev); else isert_completion_put(desc, isert_cmd, ib_dev, true); - } else { - isert_conn->post_recv_buf_count--; - if (!isert_conn->post_recv_buf_count) - iscsit_cause_connection_reinstatement(isert_conn->conn, 0); } } @@ -3214,6 +3213,7 @@ accept_wait: conn->context = isert_conn; isert_conn->conn = conn; + isert_conn->state = ISER_CONN_BOUND; isert_set_conn_info(np, conn, isert_conn); diff --git a/drivers/infiniband/ulp/isert/ib_isert.h b/drivers/infiniband/ulp/isert/ib_isert.h index 8d50453ee..1aa019ab9 100644 --- a/drivers/infiniband/ulp/isert/ib_isert.h +++ b/drivers/infiniband/ulp/isert/ib_isert.h @@ -84,6 +84,7 @@ enum iser_ib_op_code { enum iser_conn_state { ISER_CONN_INIT, ISER_CONN_UP, + ISER_CONN_BOUND, ISER_CONN_FULL_FEATURE, ISER_CONN_TERMINATING, ISER_CONN_DOWN, @@ -179,7 +180,6 @@ struct isert_device; struct isert_conn { enum iser_conn_state state; - int post_recv_buf_count; u32 responder_resources; u32 initiator_depth; bool pi_support; diff --git a/drivers/infiniband/ulp/srpt/ib_srpt.c b/drivers/infiniband/ulp/srpt/ib_srpt.c index 0c37fee36..4328679a6 100644 --- a/drivers/infiniband/ulp/srpt/ib_srpt.c +++ b/drivers/infiniband/ulp/srpt/ib_srpt.c @@ -1670,47 +1670,6 @@ send_sense: return -1; } -/** - * srpt_rx_mgmt_fn_tag() - Process a task management function by tag. - * @ch: RDMA channel of the task management request. - * @fn: Task management function to perform. - * @req_tag: Tag of the SRP task management request. - * @mgmt_ioctx: I/O context of the task management request. - * - * Returns zero if the target core will process the task management - * request asynchronously. - * - * Note: It is assumed that the initiator serializes tag-based task management - * requests. - */ -static int srpt_rx_mgmt_fn_tag(struct srpt_send_ioctx *ioctx, u64 tag) -{ - struct srpt_device *sdev; - struct srpt_rdma_ch *ch; - struct srpt_send_ioctx *target; - int ret, i; - - ret = -EINVAL; - ch = ioctx->ch; - BUG_ON(!ch); - BUG_ON(!ch->sport); - sdev = ch->sport->sdev; - BUG_ON(!sdev); - spin_lock_irq(&sdev->spinlock); - for (i = 0; i < ch->rq_size; ++i) { - target = ch->ioctx_ring[i]; - if (target->cmd.se_lun == ioctx->cmd.se_lun && - target->cmd.tag == tag && - srpt_get_cmd_state(target) != SRPT_STATE_DONE) { - ret = 0; - /* now let the target core abort &target->cmd; */ - break; - } - } - spin_unlock_irq(&sdev->spinlock); - return ret; -} - static int srp_tmr_to_tcm(int fn) { switch (fn) { @@ -1745,7 +1704,6 @@ static void srpt_handle_tsk_mgmt(struct srpt_rdma_ch *ch, struct se_cmd *cmd; struct se_session *sess = ch->sess; uint64_t unpacked_lun; - uint32_t tag = 0; int tcm_tmr; int rc; @@ -1761,25 +1719,10 @@ static void srpt_handle_tsk_mgmt(struct srpt_rdma_ch *ch, srpt_set_cmd_state(send_ioctx, SRPT_STATE_MGMT); send_ioctx->cmd.tag = srp_tsk->tag; tcm_tmr = srp_tmr_to_tcm(srp_tsk->tsk_mgmt_func); - if (tcm_tmr < 0) { - send_ioctx->cmd.se_tmr_req->response = - TMR_TASK_MGMT_FUNCTION_NOT_SUPPORTED; - goto fail; - } unpacked_lun = srpt_unpack_lun((uint8_t *)&srp_tsk->lun, sizeof(srp_tsk->lun)); - - if (srp_tsk->tsk_mgmt_func == SRP_TSK_ABORT_TASK) { - rc = srpt_rx_mgmt_fn_tag(send_ioctx, srp_tsk->task_tag); - if (rc < 0) { - send_ioctx->cmd.se_tmr_req->response = - TMR_TASK_DOES_NOT_EXIST; - goto fail; - } - tag = srp_tsk->task_tag; - } rc = target_submit_tmr(&send_ioctx->cmd, sess, NULL, unpacked_lun, - srp_tsk, tcm_tmr, GFP_KERNEL, tag, + srp_tsk, tcm_tmr, GFP_KERNEL, srp_tsk->task_tag, TARGET_SCF_ACK_KREF); if (rc != 0) { send_ioctx->cmd.se_tmr_req->response = TMR_FUNCTION_REJECTED; diff --git a/drivers/input/misc/ati_remote2.c b/drivers/input/misc/ati_remote2.c index cfd58e87d..1c5914cae 100644 --- a/drivers/input/misc/ati_remote2.c +++ b/drivers/input/misc/ati_remote2.c @@ -817,26 +817,49 @@ static int ati_remote2_probe(struct usb_interface *interface, const struct usb_d ar2->udev = udev; + /* Sanity check, first interface must have an endpoint */ + if (alt->desc.bNumEndpoints < 1 || !alt->endpoint) { + dev_err(&interface->dev, + "%s(): interface 0 must have an endpoint\n", __func__); + r = -ENODEV; + goto fail1; + } ar2->intf[0] = interface; ar2->ep[0] = &alt->endpoint[0].desc; + /* Sanity check, the device must have two interfaces */ ar2->intf[1] = usb_ifnum_to_if(udev, 1); + if ((udev->actconfig->desc.bNumInterfaces < 2) || !ar2->intf[1]) { + dev_err(&interface->dev, "%s(): need 2 interfaces, found %d\n", + __func__, udev->actconfig->desc.bNumInterfaces); + r = -ENODEV; + goto fail1; + } + r = usb_driver_claim_interface(&ati_remote2_driver, ar2->intf[1], ar2); if (r) goto fail1; + + /* Sanity check, second interface must have an endpoint */ alt = ar2->intf[1]->cur_altsetting; + if (alt->desc.bNumEndpoints < 1 || !alt->endpoint) { + dev_err(&interface->dev, + "%s(): interface 1 must have an endpoint\n", __func__); + r = -ENODEV; + goto fail2; + } ar2->ep[1] = &alt->endpoint[0].desc; r = ati_remote2_urb_init(ar2); if (r) - goto fail2; + goto fail3; ar2->channel_mask = channel_mask; ar2->mode_mask = mode_mask; r = ati_remote2_setup(ar2, ar2->channel_mask); if (r) - goto fail2; + goto fail3; usb_make_path(udev, ar2->phys, sizeof(ar2->phys)); strlcat(ar2->phys, "/input0", sizeof(ar2->phys)); @@ -845,11 +868,11 @@ static int ati_remote2_probe(struct usb_interface *interface, const struct usb_d r = sysfs_create_group(&udev->dev.kobj, &ati_remote2_attr_group); if (r) - goto fail2; + goto fail3; r = ati_remote2_input_init(ar2); if (r) - goto fail3; + goto fail4; usb_set_intfdata(interface, ar2); @@ -857,10 +880,11 @@ static int ati_remote2_probe(struct usb_interface *interface, const struct usb_d return 0; - fail3: + fail4: sysfs_remove_group(&udev->dev.kobj, &ati_remote2_attr_group); - fail2: + fail3: ati_remote2_urb_cleanup(ar2); + fail2: usb_driver_release_interface(&ati_remote2_driver, ar2->intf[1]); fail1: kfree(ar2); diff --git a/drivers/input/misc/ims-pcu.c b/drivers/input/misc/ims-pcu.c index 43d082530..f47851868 100644 --- a/drivers/input/misc/ims-pcu.c +++ b/drivers/input/misc/ims-pcu.c @@ -1663,6 +1663,8 @@ static int ims_pcu_parse_cdc_data(struct usb_interface *intf, struct ims_pcu *pc pcu->ctrl_intf = usb_ifnum_to_if(pcu->udev, union_desc->bMasterInterface0); + if (!pcu->ctrl_intf) + return -EINVAL; alt = pcu->ctrl_intf->cur_altsetting; pcu->ep_ctrl = &alt->endpoint[0].desc; @@ -1670,6 +1672,8 @@ static int ims_pcu_parse_cdc_data(struct usb_interface *intf, struct ims_pcu *pc pcu->data_intf = usb_ifnum_to_if(pcu->udev, union_desc->bSlaveInterface0); + if (!pcu->data_intf) + return -EINVAL; alt = pcu->data_intf->cur_altsetting; if (alt->desc.bNumEndpoints != 2) { diff --git a/drivers/input/misc/powermate.c b/drivers/input/misc/powermate.c index 63b539d3d..84909a12f 100644 --- a/drivers/input/misc/powermate.c +++ b/drivers/input/misc/powermate.c @@ -307,6 +307,9 @@ static int powermate_probe(struct usb_interface *intf, const struct usb_device_i int error = -ENOMEM; interface = intf->cur_altsetting; + if (interface->desc.bNumEndpoints < 1) + return -EINVAL; + endpoint = &interface->endpoint[0].desc; if (!usb_endpoint_is_int_in(endpoint)) return -EIO; diff --git a/drivers/input/mouse/synaptics.c b/drivers/input/mouse/synaptics.c index e678212f1..a5c2eb291 100644 --- a/drivers/input/mouse/synaptics.c +++ b/drivers/input/mouse/synaptics.c @@ -862,8 +862,9 @@ static void synaptics_report_ext_buttons(struct psmouse *psmouse, if (!SYN_CAP_MULTI_BUTTON_NO(priv->ext_cap)) return; - /* Bug in FW 8.1, buttons are reported only when ExtBit is 1 */ - if (SYN_ID_FULL(priv->identity) == 0x801 && + /* Bug in FW 8.1 & 8.2, buttons are reported only when ExtBit is 1 */ + if ((SYN_ID_FULL(priv->identity) == 0x801 || + SYN_ID_FULL(priv->identity) == 0x802) && !((psmouse->packet[0] ^ psmouse->packet[3]) & 0x02)) return; diff --git a/drivers/md/bcache/super.c b/drivers/md/bcache/super.c index 8d0ead98e..a296425a7 100644 --- a/drivers/md/bcache/super.c +++ b/drivers/md/bcache/super.c @@ -1015,8 +1015,12 @@ int bch_cached_dev_attach(struct cached_dev *dc, struct cache_set *c) */ atomic_set(&dc->count, 1); - if (bch_cached_dev_writeback_start(dc)) + /* Block writeback thread, but spawn it */ + down_write(&dc->writeback_lock); + if (bch_cached_dev_writeback_start(dc)) { + up_write(&dc->writeback_lock); return -ENOMEM; + } if (BDEV_STATE(&dc->sb) == BDEV_STATE_DIRTY) { bch_sectors_dirty_init(dc); @@ -1028,6 +1032,9 @@ int bch_cached_dev_attach(struct cached_dev *dc, struct cache_set *c) bch_cached_dev_run(dc); bcache_device_link(&dc->disk, c, "bdev"); + /* Allow the writeback thread to proceed */ + up_write(&dc->writeback_lock); + pr_info("Caching %s as %s on set %pU", bdevname(dc->bdev, buf), dc->disk.disk->disk_name, dc->disk.c->sb.set_uuid); @@ -1366,6 +1373,9 @@ static void cache_set_flush(struct closure *cl) struct btree *b; unsigned i; + if (!c) + closure_return(cl); + bch_cache_accounting_destroy(&c->accounting); kobject_put(&c->internal); @@ -1828,11 +1838,12 @@ static int cache_alloc(struct cache_sb *sb, struct cache *ca) return 0; } -static void register_cache(struct cache_sb *sb, struct page *sb_page, +static int register_cache(struct cache_sb *sb, struct page *sb_page, struct block_device *bdev, struct cache *ca) { char name[BDEVNAME_SIZE]; - const char *err = "cannot allocate memory"; + const char *err = NULL; + int ret = 0; memcpy(&ca->sb, sb, sizeof(struct cache_sb)); ca->bdev = bdev; @@ -1847,27 +1858,35 @@ static void register_cache(struct cache_sb *sb, struct page *sb_page, if (blk_queue_discard(bdev_get_queue(ca->bdev))) ca->discard = CACHE_DISCARD(&ca->sb); - if (cache_alloc(sb, ca) != 0) + ret = cache_alloc(sb, ca); + if (ret != 0) goto err; - err = "error creating kobject"; - if (kobject_add(&ca->kobj, &part_to_dev(bdev->bd_part)->kobj, "bcache")) - goto err; + if (kobject_add(&ca->kobj, &part_to_dev(bdev->bd_part)->kobj, "bcache")) { + err = "error calling kobject_add"; + ret = -ENOMEM; + goto out; + } mutex_lock(&bch_register_lock); err = register_cache_set(ca); mutex_unlock(&bch_register_lock); - if (err) - goto err; + if (err) { + ret = -ENODEV; + goto out; + } pr_info("registered cache device %s", bdevname(bdev, name)); + out: kobject_put(&ca->kobj); - return; + err: - pr_notice("error opening %s: %s", bdevname(bdev, name), err); - goto out; + if (err) + pr_notice("error opening %s: %s", bdevname(bdev, name), err); + + return ret; } /* Global interfaces/init */ @@ -1965,7 +1984,8 @@ static ssize_t register_bcache(struct kobject *k, struct kobj_attribute *attr, if (!ca) goto err_close; - register_cache(sb, sb_page, bdev, ca); + if (register_cache(sb, sb_page, bdev, ca) != 0) + goto err_close; } out: if (sb_page) diff --git a/drivers/md/dm-cache-metadata.c b/drivers/md/dm-cache-metadata.c index f6543f3a9..27f2ef300 100644 --- a/drivers/md/dm-cache-metadata.c +++ b/drivers/md/dm-cache-metadata.c @@ -867,19 +867,40 @@ static int blocks_are_unmapped_or_clean(struct dm_cache_metadata *cmd, return 0; } -#define WRITE_LOCK(cmd) \ - if (cmd->fail_io || dm_bm_is_read_only(cmd->bm)) \ +#define WRITE_LOCK(cmd) \ + down_write(&cmd->root_lock); \ + if (cmd->fail_io || dm_bm_is_read_only(cmd->bm)) { \ + up_write(&cmd->root_lock); \ return -EINVAL; \ - down_write(&cmd->root_lock) + } #define WRITE_LOCK_VOID(cmd) \ - if (cmd->fail_io || dm_bm_is_read_only(cmd->bm)) \ + down_write(&cmd->root_lock); \ + if (cmd->fail_io || dm_bm_is_read_only(cmd->bm)) { \ + up_write(&cmd->root_lock); \ return; \ - down_write(&cmd->root_lock) + } #define WRITE_UNLOCK(cmd) \ up_write(&cmd->root_lock) +#define READ_LOCK(cmd) \ + down_read(&cmd->root_lock); \ + if (cmd->fail_io || dm_bm_is_read_only(cmd->bm)) { \ + up_read(&cmd->root_lock); \ + return -EINVAL; \ + } + +#define READ_LOCK_VOID(cmd) \ + down_read(&cmd->root_lock); \ + if (cmd->fail_io || dm_bm_is_read_only(cmd->bm)) { \ + up_read(&cmd->root_lock); \ + return; \ + } + +#define READ_UNLOCK(cmd) \ + up_read(&cmd->root_lock) + int dm_cache_resize(struct dm_cache_metadata *cmd, dm_cblock_t new_cache_size) { int r; @@ -1015,22 +1036,20 @@ int dm_cache_load_discards(struct dm_cache_metadata *cmd, { int r; - down_read(&cmd->root_lock); + READ_LOCK(cmd); r = __load_discards(cmd, fn, context); - up_read(&cmd->root_lock); + READ_UNLOCK(cmd); return r; } -dm_cblock_t dm_cache_size(struct dm_cache_metadata *cmd) +int dm_cache_size(struct dm_cache_metadata *cmd, dm_cblock_t *result) { - dm_cblock_t r; + READ_LOCK(cmd); + *result = cmd->cache_blocks; + READ_UNLOCK(cmd); - down_read(&cmd->root_lock); - r = cmd->cache_blocks; - up_read(&cmd->root_lock); - - return r; + return 0; } static int __remove(struct dm_cache_metadata *cmd, dm_cblock_t cblock) @@ -1188,9 +1207,9 @@ int dm_cache_load_mappings(struct dm_cache_metadata *cmd, { int r; - down_read(&cmd->root_lock); + READ_LOCK(cmd); r = __load_mappings(cmd, policy, fn, context); - up_read(&cmd->root_lock); + READ_UNLOCK(cmd); return r; } @@ -1215,18 +1234,18 @@ static int __dump_mappings(struct dm_cache_metadata *cmd) void dm_cache_dump(struct dm_cache_metadata *cmd) { - down_read(&cmd->root_lock); + READ_LOCK_VOID(cmd); __dump_mappings(cmd); - up_read(&cmd->root_lock); + READ_UNLOCK(cmd); } int dm_cache_changed_this_transaction(struct dm_cache_metadata *cmd) { int r; - down_read(&cmd->root_lock); + READ_LOCK(cmd); r = cmd->changed; - up_read(&cmd->root_lock); + READ_UNLOCK(cmd); return r; } @@ -1276,9 +1295,9 @@ int dm_cache_set_dirty(struct dm_cache_metadata *cmd, void dm_cache_metadata_get_stats(struct dm_cache_metadata *cmd, struct dm_cache_statistics *stats) { - down_read(&cmd->root_lock); + READ_LOCK_VOID(cmd); *stats = cmd->stats; - up_read(&cmd->root_lock); + READ_UNLOCK(cmd); } void dm_cache_metadata_set_stats(struct dm_cache_metadata *cmd, @@ -1312,9 +1331,9 @@ int dm_cache_get_free_metadata_block_count(struct dm_cache_metadata *cmd, { int r = -EINVAL; - down_read(&cmd->root_lock); + READ_LOCK(cmd); r = dm_sm_get_nr_free(cmd->metadata_sm, result); - up_read(&cmd->root_lock); + READ_UNLOCK(cmd); return r; } @@ -1324,9 +1343,9 @@ int dm_cache_get_metadata_dev_size(struct dm_cache_metadata *cmd, { int r = -EINVAL; - down_read(&cmd->root_lock); + READ_LOCK(cmd); r = dm_sm_get_nr_blocks(cmd->metadata_sm, result); - up_read(&cmd->root_lock); + READ_UNLOCK(cmd); return r; } @@ -1417,7 +1436,13 @@ int dm_cache_write_hints(struct dm_cache_metadata *cmd, struct dm_cache_policy * int dm_cache_metadata_all_clean(struct dm_cache_metadata *cmd, bool *result) { - return blocks_are_unmapped_or_clean(cmd, 0, cmd->cache_blocks, result); + int r; + + READ_LOCK(cmd); + r = blocks_are_unmapped_or_clean(cmd, 0, cmd->cache_blocks, result); + READ_UNLOCK(cmd); + + return r; } void dm_cache_metadata_set_read_only(struct dm_cache_metadata *cmd) @@ -1440,10 +1465,7 @@ int dm_cache_metadata_set_needs_check(struct dm_cache_metadata *cmd) struct dm_block *sblock; struct cache_disk_superblock *disk_super; - /* - * We ignore fail_io for this function. - */ - down_write(&cmd->root_lock); + WRITE_LOCK(cmd); set_bit(NEEDS_CHECK, &cmd->flags); r = superblock_lock(cmd, &sblock); @@ -1458,19 +1480,17 @@ int dm_cache_metadata_set_needs_check(struct dm_cache_metadata *cmd) dm_bm_unlock(sblock); out: - up_write(&cmd->root_lock); + WRITE_UNLOCK(cmd); return r; } -bool dm_cache_metadata_needs_check(struct dm_cache_metadata *cmd) +int dm_cache_metadata_needs_check(struct dm_cache_metadata *cmd, bool *result) { - bool needs_check; + READ_LOCK(cmd); + *result = !!test_bit(NEEDS_CHECK, &cmd->flags); + READ_UNLOCK(cmd); - down_read(&cmd->root_lock); - needs_check = !!test_bit(NEEDS_CHECK, &cmd->flags); - up_read(&cmd->root_lock); - - return needs_check; + return 0; } int dm_cache_metadata_abort(struct dm_cache_metadata *cmd) diff --git a/drivers/md/dm-cache-metadata.h b/drivers/md/dm-cache-metadata.h index 2ffee21f3..852874419 100644 --- a/drivers/md/dm-cache-metadata.h +++ b/drivers/md/dm-cache-metadata.h @@ -66,7 +66,7 @@ void dm_cache_metadata_close(struct dm_cache_metadata *cmd); * origin blocks to map to. */ int dm_cache_resize(struct dm_cache_metadata *cmd, dm_cblock_t new_cache_size); -dm_cblock_t dm_cache_size(struct dm_cache_metadata *cmd); +int dm_cache_size(struct dm_cache_metadata *cmd, dm_cblock_t *result); int dm_cache_discard_bitset_resize(struct dm_cache_metadata *cmd, sector_t discard_block_size, @@ -137,7 +137,7 @@ int dm_cache_write_hints(struct dm_cache_metadata *cmd, struct dm_cache_policy * */ int dm_cache_metadata_all_clean(struct dm_cache_metadata *cmd, bool *result); -bool dm_cache_metadata_needs_check(struct dm_cache_metadata *cmd); +int dm_cache_metadata_needs_check(struct dm_cache_metadata *cmd, bool *result); int dm_cache_metadata_set_needs_check(struct dm_cache_metadata *cmd); void dm_cache_metadata_set_read_only(struct dm_cache_metadata *cmd); void dm_cache_metadata_set_read_write(struct dm_cache_metadata *cmd); diff --git a/drivers/md/dm-cache-target.c b/drivers/md/dm-cache-target.c index 5780accff..bb9b92ebb 100644 --- a/drivers/md/dm-cache-target.c +++ b/drivers/md/dm-cache-target.c @@ -984,9 +984,14 @@ static void notify_mode_switch(struct cache *cache, enum cache_metadata_mode mod static void set_cache_mode(struct cache *cache, enum cache_metadata_mode new_mode) { - bool needs_check = dm_cache_metadata_needs_check(cache->cmd); + bool needs_check; enum cache_metadata_mode old_mode = get_cache_mode(cache); + if (dm_cache_metadata_needs_check(cache->cmd, &needs_check)) { + DMERR("unable to read needs_check flag, setting failure mode"); + new_mode = CM_FAIL; + } + if (new_mode == CM_WRITE && needs_check) { DMERR("%s: unable to switch cache to write mode until repaired.", cache_device_name(cache)); @@ -3510,6 +3515,7 @@ static void cache_status(struct dm_target *ti, status_type_t type, char buf[BDEVNAME_SIZE]; struct cache *cache = ti->private; dm_cblock_t residency; + bool needs_check; switch (type) { case STATUSTYPE_INFO: @@ -3583,7 +3589,9 @@ static void cache_status(struct dm_target *ti, status_type_t type, else DMEMIT("rw "); - if (dm_cache_metadata_needs_check(cache->cmd)) + r = dm_cache_metadata_needs_check(cache->cmd, &needs_check); + + if (r || needs_check) DMEMIT("needs_check "); else DMEMIT("- "); diff --git a/drivers/md/dm-snap.c b/drivers/md/dm-snap.c index 376638608..e4d1bafe7 100644 --- a/drivers/md/dm-snap.c +++ b/drivers/md/dm-snap.c @@ -1105,6 +1105,7 @@ static int snapshot_ctr(struct dm_target *ti, unsigned int argc, char **argv) int i; int r = -EINVAL; char *origin_path, *cow_path; + dev_t origin_dev, cow_dev; unsigned args_used, num_flush_bios = 1; fmode_t origin_mode = FMODE_READ; @@ -1135,11 +1136,19 @@ static int snapshot_ctr(struct dm_target *ti, unsigned int argc, char **argv) ti->error = "Cannot get origin device"; goto bad_origin; } + origin_dev = s->origin->bdev->bd_dev; cow_path = argv[0]; argv++; argc--; + cow_dev = dm_get_dev_t(cow_path); + if (cow_dev && cow_dev == origin_dev) { + ti->error = "COW device cannot be the same as origin device"; + r = -EINVAL; + goto bad_cow; + } + r = dm_get_device(ti, cow_path, dm_table_get_mode(ti->table), &s->cow); if (r) { ti->error = "Cannot get COW device"; diff --git a/drivers/md/dm-table.c b/drivers/md/dm-table.c index 061152a43..cb5d0daf5 100644 --- a/drivers/md/dm-table.c +++ b/drivers/md/dm-table.c @@ -365,6 +365,26 @@ static int upgrade_mode(struct dm_dev_internal *dd, fmode_t new_mode, } /* + * Convert the path to a device + */ +dev_t dm_get_dev_t(const char *path) +{ + dev_t uninitialized_var(dev); + struct block_device *bdev; + + bdev = lookup_bdev(path); + if (IS_ERR(bdev)) + dev = name_to_dev_t(path); + else { + dev = bdev->bd_dev; + bdput(bdev); + } + + return dev; +} +EXPORT_SYMBOL_GPL(dm_get_dev_t); + +/* * Add a device to the list, or just increment the usage count if * it's already present. */ @@ -372,23 +392,15 @@ int dm_get_device(struct dm_target *ti, const char *path, fmode_t mode, struct dm_dev **result) { int r; - dev_t uninitialized_var(dev); + dev_t dev; struct dm_dev_internal *dd; struct dm_table *t = ti->table; - struct block_device *bdev; BUG_ON(!t); - /* convert the path to a device */ - bdev = lookup_bdev(path); - if (IS_ERR(bdev)) { - dev = name_to_dev_t(path); - if (!dev) - return -ENODEV; - } else { - dev = bdev->bd_dev; - bdput(bdev); - } + dev = dm_get_dev_t(path); + if (!dev) + return -ENODEV; dd = find_device(&t->devices, dev); if (!dd) { diff --git a/drivers/md/dm-thin-metadata.c b/drivers/md/dm-thin-metadata.c index f962d6453..185010d9c 100644 --- a/drivers/md/dm-thin-metadata.c +++ b/drivers/md/dm-thin-metadata.c @@ -1981,5 +1981,8 @@ bool dm_pool_metadata_needs_check(struct dm_pool_metadata *pmd) void dm_pool_issue_prefetches(struct dm_pool_metadata *pmd) { - dm_tm_issue_prefetches(pmd->tm); + down_read(&pmd->root_lock); + if (!pmd->fail_io) + dm_tm_issue_prefetches(pmd->tm); + up_read(&pmd->root_lock); } diff --git a/drivers/md/dm.c b/drivers/md/dm.c index dd834927b..c338aebb4 100644 --- a/drivers/md/dm.c +++ b/drivers/md/dm.c @@ -1109,12 +1109,8 @@ static void rq_completed(struct mapped_device *md, int rw, bool run_queue) * back into ->request_fn() could deadlock attempting to grab the * queue lock again. */ - if (run_queue) { - if (md->queue->mq_ops) - blk_mq_run_hw_queues(md->queue, true); - else - blk_run_queue_async(md->queue); - } + if (!md->queue->mq_ops && run_queue) + blk_run_queue_async(md->queue); /* * dm_put() must be at the end of this function. See the comment above @@ -1214,9 +1210,9 @@ static void dm_requeue_original_request(struct mapped_device *md, { int rw = rq_data_dir(rq); + rq_end_stats(md, rq); dm_unprep_request(rq); - rq_end_stats(md, rq); if (!rq->q->mq_ops) old_requeue_request(rq); else { @@ -1336,7 +1332,10 @@ static void dm_complete_request(struct request *rq, int error) struct dm_rq_target_io *tio = tio_from_request(rq); tio->error = error; - blk_complete_request(rq); + if (!rq->q->mq_ops) + blk_complete_request(rq); + else + blk_mq_complete_request(rq, error); } /* diff --git a/drivers/md/multipath.c b/drivers/md/multipath.c index 0a72ab6e6..dd483bb2e 100644 --- a/drivers/md/multipath.c +++ b/drivers/md/multipath.c @@ -129,7 +129,9 @@ static void multipath_make_request(struct mddev *mddev, struct bio * bio) } multipath = conf->multipaths + mp_bh->path; - mp_bh->bio = *bio; + bio_init(&mp_bh->bio); + __bio_clone_fast(&mp_bh->bio, bio); + mp_bh->bio.bi_iter.bi_sector += multipath->rdev->data_offset; mp_bh->bio.bi_bdev = multipath->rdev->bdev; mp_bh->bio.bi_rw |= REQ_FAILFAST_TRANSPORT; diff --git a/drivers/md/raid1.c b/drivers/md/raid1.c index 4e3843f7d..bb5bce059 100644 --- a/drivers/md/raid1.c +++ b/drivers/md/raid1.c @@ -2274,6 +2274,7 @@ static void handle_write_finished(struct r1conf *conf, struct r1bio *r1_bio) if (fail) { spin_lock_irq(&conf->device_lock); list_add(&r1_bio->retry_list, &conf->bio_end_io_list); + conf->nr_queued++; spin_unlock_irq(&conf->device_lock); md_wakeup_thread(conf->mddev->thread); } else { @@ -2391,8 +2392,10 @@ static void raid1d(struct md_thread *thread) LIST_HEAD(tmp); spin_lock_irqsave(&conf->device_lock, flags); if (!test_bit(MD_CHANGE_PENDING, &mddev->flags)) { - list_add(&tmp, &conf->bio_end_io_list); - list_del_init(&conf->bio_end_io_list); + while (!list_empty(&conf->bio_end_io_list)) { + list_move(conf->bio_end_io_list.prev, &tmp); + conf->nr_queued--; + } } spin_unlock_irqrestore(&conf->device_lock, flags); while (!list_empty(&tmp)) { diff --git a/drivers/md/raid10.c b/drivers/md/raid10.c index 1c1447dd3..e3fd725d5 100644 --- a/drivers/md/raid10.c +++ b/drivers/md/raid10.c @@ -2664,6 +2664,7 @@ static void handle_write_completed(struct r10conf *conf, struct r10bio *r10_bio) if (fail) { spin_lock_irq(&conf->device_lock); list_add(&r10_bio->retry_list, &conf->bio_end_io_list); + conf->nr_queued++; spin_unlock_irq(&conf->device_lock); md_wakeup_thread(conf->mddev->thread); } else { @@ -2691,8 +2692,10 @@ static void raid10d(struct md_thread *thread) LIST_HEAD(tmp); spin_lock_irqsave(&conf->device_lock, flags); if (!test_bit(MD_CHANGE_PENDING, &mddev->flags)) { - list_add(&tmp, &conf->bio_end_io_list); - list_del_init(&conf->bio_end_io_list); + while (!list_empty(&conf->bio_end_io_list)) { + list_move(conf->bio_end_io_list.prev, &tmp); + conf->nr_queued--; + } } spin_unlock_irqrestore(&conf->device_lock, flags); while (!list_empty(&tmp)) { diff --git a/drivers/md/raid5.c b/drivers/md/raid5.c index b4f02c995..32d52878f 100644 --- a/drivers/md/raid5.c +++ b/drivers/md/raid5.c @@ -340,8 +340,7 @@ static void release_inactive_stripe_list(struct r5conf *conf, int hash) { int size; - unsigned long do_wakeup = 0; - int i = 0; + bool do_wakeup = false; unsigned long flags; if (hash == NR_STRIPE_HASH_LOCKS) { @@ -362,19 +361,15 @@ static void release_inactive_stripe_list(struct r5conf *conf, !list_empty(list)) atomic_dec(&conf->empty_inactive_list_nr); list_splice_tail_init(list, conf->inactive_list + hash); - do_wakeup |= 1 << hash; + do_wakeup = true; spin_unlock_irqrestore(conf->hash_locks + hash, flags); } size--; hash--; } - for (i = 0; i < NR_STRIPE_HASH_LOCKS; i++) { - if (do_wakeup & (1 << i)) - wake_up(&conf->wait_for_stripe[i]); - } - if (do_wakeup) { + wake_up(&conf->wait_for_stripe); if (atomic_read(&conf->active_stripes) == 0) wake_up(&conf->wait_for_quiescent); if (conf->retry_read_aligned) @@ -687,15 +682,14 @@ raid5_get_active_stripe(struct r5conf *conf, sector_t sector, if (!sh) { set_bit(R5_INACTIVE_BLOCKED, &conf->cache_state); - wait_event_exclusive_cmd( - conf->wait_for_stripe[hash], + wait_event_lock_irq( + conf->wait_for_stripe, !list_empty(conf->inactive_list + hash) && (atomic_read(&conf->active_stripes) < (conf->max_nr_stripes * 3 / 4) || !test_bit(R5_INACTIVE_BLOCKED, &conf->cache_state)), - spin_unlock_irq(conf->hash_locks + hash), - spin_lock_irq(conf->hash_locks + hash)); + *(conf->hash_locks + hash)); clear_bit(R5_INACTIVE_BLOCKED, &conf->cache_state); } else { @@ -720,9 +714,6 @@ raid5_get_active_stripe(struct r5conf *conf, sector_t sector, } } while (sh == NULL); - if (!list_empty(conf->inactive_list + hash)) - wake_up(&conf->wait_for_stripe[hash]); - spin_unlock_irq(conf->hash_locks + hash); return sh; } @@ -2089,6 +2080,14 @@ static int resize_chunks(struct r5conf *conf, int new_disks, int new_sectors) unsigned long cpu; int err = 0; + /* + * Never shrink. And mddev_suspend() could deadlock if this is called + * from raid5d. In that case, scribble_disks and scribble_sectors + * should equal to new_disks and new_sectors + */ + if (conf->scribble_disks >= new_disks && + conf->scribble_sectors >= new_sectors) + return 0; mddev_suspend(conf->mddev); get_online_cpus(); for_each_present_cpu(cpu) { @@ -2110,6 +2109,10 @@ static int resize_chunks(struct r5conf *conf, int new_disks, int new_sectors) } put_online_cpus(); mddev_resume(conf->mddev); + if (!err) { + conf->scribble_disks = new_disks; + conf->scribble_sectors = new_sectors; + } return err; } @@ -2190,7 +2193,7 @@ static int resize_stripes(struct r5conf *conf, int newsize) cnt = 0; list_for_each_entry(nsh, &newstripes, lru) { lock_device_hash_lock(conf, hash); - wait_event_exclusive_cmd(conf->wait_for_stripe[hash], + wait_event_cmd(conf->wait_for_stripe, !list_empty(conf->inactive_list + hash), unlock_device_hash_lock(conf, hash), lock_device_hash_lock(conf, hash)); @@ -4236,7 +4239,6 @@ static void break_stripe_batch_list(struct stripe_head *head_sh, WARN_ON_ONCE(sh->state & ((1 << STRIPE_ACTIVE) | (1 << STRIPE_SYNCING) | (1 << STRIPE_REPLACED) | - (1 << STRIPE_PREREAD_ACTIVE) | (1 << STRIPE_DELAYED) | (1 << STRIPE_BIT_DELAY) | (1 << STRIPE_FULL_WRITE) | @@ -4251,6 +4253,7 @@ static void break_stripe_batch_list(struct stripe_head *head_sh, (1 << STRIPE_REPLACED))); set_mask_bits(&sh->state, ~(STRIPE_EXPAND_SYNC_FLAGS | + (1 << STRIPE_PREREAD_ACTIVE) | (1 << STRIPE_DEGRADED)), head_sh->state & (1 << STRIPE_INSYNC)); @@ -6413,6 +6416,12 @@ static int raid5_alloc_percpu(struct r5conf *conf) } put_online_cpus(); + if (!err) { + conf->scribble_disks = max(conf->raid_disks, + conf->previous_raid_disks); + conf->scribble_sectors = max(conf->chunk_sectors, + conf->prev_chunk_sectors); + } return err; } @@ -6503,9 +6512,7 @@ static struct r5conf *setup_conf(struct mddev *mddev) seqcount_init(&conf->gen_lock); mutex_init(&conf->cache_size_mutex); init_waitqueue_head(&conf->wait_for_quiescent); - for (i = 0; i < NR_STRIPE_HASH_LOCKS; i++) { - init_waitqueue_head(&conf->wait_for_stripe[i]); - } + init_waitqueue_head(&conf->wait_for_stripe); init_waitqueue_head(&conf->wait_for_overlap); INIT_LIST_HEAD(&conf->handle_list); INIT_LIST_HEAD(&conf->hold_list); @@ -7014,8 +7021,8 @@ static int raid5_run(struct mddev *mddev) } if (discard_supported && - mddev->queue->limits.max_discard_sectors >= stripe && - mddev->queue->limits.discard_granularity >= stripe) + mddev->queue->limits.max_discard_sectors >= (stripe >> 9) && + mddev->queue->limits.discard_granularity >= stripe) queue_flag_set_unlocked(QUEUE_FLAG_DISCARD, mddev->queue); else diff --git a/drivers/md/raid5.h b/drivers/md/raid5.h index a415e1cd3..517d4b68a 100644 --- a/drivers/md/raid5.h +++ b/drivers/md/raid5.h @@ -510,6 +510,8 @@ struct r5conf { * conversions */ } __percpu *percpu; + int scribble_disks; + int scribble_sectors; #ifdef CONFIG_HOTPLUG_CPU struct notifier_block cpu_notify; #endif @@ -522,7 +524,7 @@ struct r5conf { atomic_t empty_inactive_list_nr; struct llist_head released_stripes; wait_queue_head_t wait_for_quiescent; - wait_queue_head_t wait_for_stripe[NR_STRIPE_HASH_LOCKS]; + wait_queue_head_t wait_for_stripe; wait_queue_head_t wait_for_overlap; unsigned long cache_state; #define R5_INACTIVE_BLOCKED 1 /* release of inactive stripes blocked, diff --git a/drivers/media/i2c/adv7511.c b/drivers/media/i2c/adv7511.c index 471fd23b5..08d2c6bf7 100644 --- a/drivers/media/i2c/adv7511.c +++ b/drivers/media/i2c/adv7511.c @@ -1161,12 +1161,23 @@ static void adv7511_dbg_dump_edid(int lvl, int debug, struct v4l2_subdev *sd, in } } +static void adv7511_notify_no_edid(struct v4l2_subdev *sd) +{ + struct adv7511_state *state = get_adv7511_state(sd); + struct adv7511_edid_detect ed; + + /* We failed to read the EDID, so send an event for this. */ + ed.present = false; + ed.segment = adv7511_rd(sd, 0xc4); + v4l2_subdev_notify(sd, ADV7511_EDID_DETECT, (void *)&ed); + v4l2_ctrl_s_ctrl(state->have_edid0_ctrl, 0x0); +} + static void adv7511_edid_handler(struct work_struct *work) { struct delayed_work *dwork = to_delayed_work(work); struct adv7511_state *state = container_of(dwork, struct adv7511_state, edid_handler); struct v4l2_subdev *sd = &state->sd; - struct adv7511_edid_detect ed; v4l2_dbg(1, debug, sd, "%s:\n", __func__); @@ -1191,9 +1202,7 @@ static void adv7511_edid_handler(struct work_struct *work) } /* We failed to read the EDID, so send an event for this. */ - ed.present = false; - ed.segment = adv7511_rd(sd, 0xc4); - v4l2_subdev_notify(sd, ADV7511_EDID_DETECT, (void *)&ed); + adv7511_notify_no_edid(sd); v4l2_dbg(1, debug, sd, "%s: no edid found\n", __func__); } @@ -1264,7 +1273,6 @@ static void adv7511_check_monitor_present_status(struct v4l2_subdev *sd) /* update read only ctrls */ v4l2_ctrl_s_ctrl(state->hotplug_ctrl, adv7511_have_hotplug(sd) ? 0x1 : 0x0); v4l2_ctrl_s_ctrl(state->rx_sense_ctrl, adv7511_have_rx_sense(sd) ? 0x1 : 0x0); - v4l2_ctrl_s_ctrl(state->have_edid0_ctrl, state->edid.segments ? 0x1 : 0x0); if ((status & MASK_ADV7511_HPD_DETECT) && ((status & MASK_ADV7511_MSEN_DETECT) || state->edid.segments)) { v4l2_dbg(1, debug, sd, "%s: hotplug and (rx-sense or edid)\n", __func__); @@ -1294,6 +1302,7 @@ static void adv7511_check_monitor_present_status(struct v4l2_subdev *sd) } adv7511_s_power(sd, false); memset(&state->edid, 0, sizeof(struct adv7511_state_edid)); + adv7511_notify_no_edid(sd); } } @@ -1370,6 +1379,7 @@ static bool adv7511_check_edid_status(struct v4l2_subdev *sd) } /* one more segment read ok */ state->edid.segments = segment + 1; + v4l2_ctrl_s_ctrl(state->have_edid0_ctrl, 0x1); if (((state->edid.data[0x7e] >> 1) + 1) > state->edid.segments) { /* Request next EDID segment */ v4l2_dbg(1, debug, sd, "%s: request segment %d\n", __func__, state->edid.segments); @@ -1389,7 +1399,6 @@ static bool adv7511_check_edid_status(struct v4l2_subdev *sd) ed.present = true; ed.segment = 0; state->edid_detect_counter++; - v4l2_ctrl_s_ctrl(state->have_edid0_ctrl, state->edid.segments ? 0x1 : 0x0); v4l2_subdev_notify(sd, ADV7511_EDID_DETECT, (void *)&ed); return ed.present; } diff --git a/drivers/media/pci/bt8xx/bttv-driver.c b/drivers/media/pci/bt8xx/bttv-driver.c index 9400e9960..bedbd51fb 100644 --- a/drivers/media/pci/bt8xx/bttv-driver.c +++ b/drivers/media/pci/bt8xx/bttv-driver.c @@ -2334,6 +2334,19 @@ static int bttv_g_fmt_vid_overlay(struct file *file, void *priv, return 0; } +static void bttv_get_width_mask_vid_cap(const struct bttv_format *fmt, + unsigned int *width_mask, + unsigned int *width_bias) +{ + if (fmt->flags & FORMAT_FLAGS_PLANAR) { + *width_mask = ~15; /* width must be a multiple of 16 pixels */ + *width_bias = 8; /* nearest */ + } else { + *width_mask = ~3; /* width must be a multiple of 4 pixels */ + *width_bias = 2; /* nearest */ + } +} + static int bttv_try_fmt_vid_cap(struct file *file, void *priv, struct v4l2_format *f) { @@ -2343,6 +2356,7 @@ static int bttv_try_fmt_vid_cap(struct file *file, void *priv, enum v4l2_field field; __s32 width, height; __s32 height2; + unsigned int width_mask, width_bias; int rc; fmt = format_by_fourcc(f->fmt.pix.pixelformat); @@ -2375,9 +2389,9 @@ static int bttv_try_fmt_vid_cap(struct file *file, void *priv, width = f->fmt.pix.width; height = f->fmt.pix.height; + bttv_get_width_mask_vid_cap(fmt, &width_mask, &width_bias); rc = limit_scaled_size_lock(fh, &width, &height, field, - /* width_mask: 4 pixels */ ~3, - /* width_bias: nearest */ 2, + width_mask, width_bias, /* adjust_size */ 1, /* adjust_crop */ 0); if (0 != rc) @@ -2410,6 +2424,7 @@ static int bttv_s_fmt_vid_cap(struct file *file, void *priv, struct bttv_fh *fh = priv; struct bttv *btv = fh->btv; __s32 width, height; + unsigned int width_mask, width_bias; enum v4l2_field field; retval = bttv_switch_type(fh, f->type); @@ -2424,9 +2439,10 @@ static int bttv_s_fmt_vid_cap(struct file *file, void *priv, height = f->fmt.pix.height; field = f->fmt.pix.field; + fmt = format_by_fourcc(f->fmt.pix.pixelformat); + bttv_get_width_mask_vid_cap(fmt, &width_mask, &width_bias); retval = limit_scaled_size_lock(fh, &width, &height, f->fmt.pix.field, - /* width_mask: 4 pixels */ ~3, - /* width_bias: nearest */ 2, + width_mask, width_bias, /* adjust_size */ 1, /* adjust_crop */ 1); if (0 != retval) @@ -2434,8 +2450,6 @@ static int bttv_s_fmt_vid_cap(struct file *file, void *priv, f->fmt.pix.field = field; - fmt = format_by_fourcc(f->fmt.pix.pixelformat); - /* update our state informations */ fh->fmt = fmt; fh->cap.field = f->fmt.pix.field; diff --git a/drivers/media/pci/saa7134/saa7134-video.c b/drivers/media/pci/saa7134/saa7134-video.c index a63c1366a..1293563b7 100644 --- a/drivers/media/pci/saa7134/saa7134-video.c +++ b/drivers/media/pci/saa7134/saa7134-video.c @@ -1219,10 +1219,13 @@ static int saa7134_g_fmt_vid_cap(struct file *file, void *priv, f->fmt.pix.height = dev->height; f->fmt.pix.field = dev->field; f->fmt.pix.pixelformat = dev->fmt->fourcc; - f->fmt.pix.bytesperline = - (f->fmt.pix.width * dev->fmt->depth) >> 3; + if (dev->fmt->planar) + f->fmt.pix.bytesperline = f->fmt.pix.width; + else + f->fmt.pix.bytesperline = + (f->fmt.pix.width * dev->fmt->depth) / 8; f->fmt.pix.sizeimage = - f->fmt.pix.height * f->fmt.pix.bytesperline; + (f->fmt.pix.height * f->fmt.pix.width * dev->fmt->depth) / 8; f->fmt.pix.colorspace = V4L2_COLORSPACE_SMPTE170M; return 0; } @@ -1298,10 +1301,13 @@ static int saa7134_try_fmt_vid_cap(struct file *file, void *priv, if (f->fmt.pix.height > maxh) f->fmt.pix.height = maxh; f->fmt.pix.width &= ~0x03; - f->fmt.pix.bytesperline = - (f->fmt.pix.width * fmt->depth) >> 3; + if (fmt->planar) + f->fmt.pix.bytesperline = f->fmt.pix.width; + else + f->fmt.pix.bytesperline = + (f->fmt.pix.width * fmt->depth) / 8; f->fmt.pix.sizeimage = - f->fmt.pix.height * f->fmt.pix.bytesperline; + (f->fmt.pix.height * f->fmt.pix.width * fmt->depth) / 8; f->fmt.pix.colorspace = V4L2_COLORSPACE_SMPTE170M; return 0; diff --git a/drivers/media/platform/coda/coda-bit.c b/drivers/media/platform/coda/coda-bit.c index 7d28899f8..6efe9d002 100644 --- a/drivers/media/platform/coda/coda-bit.c +++ b/drivers/media/platform/coda/coda-bit.c @@ -1342,7 +1342,7 @@ static void coda_finish_encode(struct coda_ctx *ctx) /* Calculate bytesused field */ if (dst_buf->sequence == 0) { - vb2_set_plane_payload(&dst_buf->vb2_buf, 0, + vb2_set_plane_payload(&dst_buf->vb2_buf, 0, wr_ptr - start_ptr + ctx->vpu_header_size[0] + ctx->vpu_header_size[1] + ctx->vpu_header_size[2]); diff --git a/drivers/media/usb/pwc/pwc-if.c b/drivers/media/usb/pwc/pwc-if.c index 086cf1c7b..18aed5dd3 100644 --- a/drivers/media/usb/pwc/pwc-if.c +++ b/drivers/media/usb/pwc/pwc-if.c @@ -91,6 +91,7 @@ static const struct usb_device_id pwc_device_table [] = { { USB_DEVICE(0x0471, 0x0312) }, { USB_DEVICE(0x0471, 0x0313) }, /* the 'new' 720K */ { USB_DEVICE(0x0471, 0x0329) }, /* Philips SPC 900NC PC Camera */ + { USB_DEVICE(0x0471, 0x032C) }, /* Philips SPC 880NC PC Camera */ { USB_DEVICE(0x069A, 0x0001) }, /* Askey */ { USB_DEVICE(0x046D, 0x08B0) }, /* Logitech QuickCam Pro 3000 */ { USB_DEVICE(0x046D, 0x08B1) }, /* Logitech QuickCam Notebook Pro */ @@ -810,6 +811,11 @@ static int usb_pwc_probe(struct usb_interface *intf, const struct usb_device_id name = "Philips SPC 900NC webcam"; type_id = 740; break; + case 0x032C: + PWC_INFO("Philips SPC 880NC USB webcam detected.\n"); + name = "Philips SPC 880NC webcam"; + type_id = 740; + break; default: return -ENODEV; break; diff --git a/drivers/media/v4l2-core/v4l2-compat-ioctl32.c b/drivers/media/v4l2-core/v4l2-compat-ioctl32.c index 8fd84a674..019644ff6 100644 --- a/drivers/media/v4l2-core/v4l2-compat-ioctl32.c +++ b/drivers/media/v4l2-core/v4l2-compat-ioctl32.c @@ -415,7 +415,8 @@ static int get_v4l2_buffer32(struct v4l2_buffer *kp, struct v4l2_buffer32 __user get_user(kp->index, &up->index) || get_user(kp->type, &up->type) || get_user(kp->flags, &up->flags) || - get_user(kp->memory, &up->memory)) + get_user(kp->memory, &up->memory) || + get_user(kp->length, &up->length)) return -EFAULT; if (V4L2_TYPE_IS_OUTPUT(kp->type)) @@ -427,9 +428,6 @@ static int get_v4l2_buffer32(struct v4l2_buffer *kp, struct v4l2_buffer32 __user return -EFAULT; if (V4L2_TYPE_IS_MULTIPLANAR(kp->type)) { - if (get_user(kp->length, &up->length)) - return -EFAULT; - num_planes = kp->length; if (num_planes == 0) { kp->m.planes = NULL; @@ -462,16 +460,14 @@ static int get_v4l2_buffer32(struct v4l2_buffer *kp, struct v4l2_buffer32 __user } else { switch (kp->memory) { case V4L2_MEMORY_MMAP: - if (get_user(kp->length, &up->length) || - get_user(kp->m.offset, &up->m.offset)) + if (get_user(kp->m.offset, &up->m.offset)) return -EFAULT; break; case V4L2_MEMORY_USERPTR: { compat_long_t tmp; - if (get_user(kp->length, &up->length) || - get_user(tmp, &up->m.userptr)) + if (get_user(tmp, &up->m.userptr)) return -EFAULT; kp->m.userptr = (unsigned long)compat_ptr(tmp); @@ -513,7 +509,8 @@ static int put_v4l2_buffer32(struct v4l2_buffer *kp, struct v4l2_buffer32 __user copy_to_user(&up->timecode, &kp->timecode, sizeof(struct v4l2_timecode)) || put_user(kp->sequence, &up->sequence) || put_user(kp->reserved2, &up->reserved2) || - put_user(kp->reserved, &up->reserved)) + put_user(kp->reserved, &up->reserved) || + put_user(kp->length, &up->length)) return -EFAULT; if (V4L2_TYPE_IS_MULTIPLANAR(kp->type)) { @@ -536,13 +533,11 @@ static int put_v4l2_buffer32(struct v4l2_buffer *kp, struct v4l2_buffer32 __user } else { switch (kp->memory) { case V4L2_MEMORY_MMAP: - if (put_user(kp->length, &up->length) || - put_user(kp->m.offset, &up->m.offset)) + if (put_user(kp->m.offset, &up->m.offset)) return -EFAULT; break; case V4L2_MEMORY_USERPTR: - if (put_user(kp->length, &up->length) || - put_user(kp->m.userptr, &up->m.userptr)) + if (put_user(kp->m.userptr, &up->m.userptr)) return -EFAULT; break; case V4L2_MEMORY_OVERLAY: diff --git a/drivers/misc/mei/bus.c b/drivers/misc/mei/bus.c index 0b05aa938..1a173d0af 100644 --- a/drivers/misc/mei/bus.c +++ b/drivers/misc/mei/bus.c @@ -53,6 +53,11 @@ ssize_t __mei_cl_send(struct mei_cl *cl, u8 *buf, size_t length, bus = cl->dev; mutex_lock(&bus->device_lock); + if (bus->dev_state != MEI_DEV_ENABLED) { + rets = -ENODEV; + goto out; + } + if (!mei_cl_is_connected(cl)) { rets = -ENODEV; goto out; @@ -109,6 +114,10 @@ ssize_t __mei_cl_recv(struct mei_cl *cl, u8 *buf, size_t length) bus = cl->dev; mutex_lock(&bus->device_lock); + if (bus->dev_state != MEI_DEV_ENABLED) { + rets = -ENODEV; + goto out; + } cb = mei_cl_read_cb(cl, NULL); if (cb) diff --git a/drivers/mmc/card/block.c b/drivers/mmc/card/block.c index fe207e542..5fbffdb6b 100644 --- a/drivers/mmc/card/block.c +++ b/drivers/mmc/card/block.c @@ -589,6 +589,14 @@ static int mmc_blk_ioctl_cmd(struct block_device *bdev, struct mmc_card *card; int err = 0, ioc_err = 0; + /* + * The caller must have CAP_SYS_RAWIO, and must be calling this on the + * whole block device, not on a partition. This prevents overspray + * between sibling partitions. + */ + if ((!capable(CAP_SYS_RAWIO)) || (bdev != bdev->bd_contains)) + return -EPERM; + idata = mmc_blk_ioctl_copy_from_user(ic_ptr); if (IS_ERR(idata)) return PTR_ERR(idata); @@ -631,6 +639,14 @@ static int mmc_blk_ioctl_multi_cmd(struct block_device *bdev, int i, err = 0, ioc_err = 0; __u64 num_of_cmds; + /* + * The caller must have CAP_SYS_RAWIO, and must be calling this on the + * whole block device, not on a partition. This prevents overspray + * between sibling partitions. + */ + if ((!capable(CAP_SYS_RAWIO)) || (bdev != bdev->bd_contains)) + return -EPERM; + if (copy_from_user(&num_of_cmds, &user->num_of_cmds, sizeof(num_of_cmds))) return -EFAULT; @@ -688,14 +704,6 @@ cmd_err: static int mmc_blk_ioctl(struct block_device *bdev, fmode_t mode, unsigned int cmd, unsigned long arg) { - /* - * The caller must have CAP_SYS_RAWIO, and must be calling this on the - * whole block device, not on a partition. This prevents overspray - * between sibling partitions. - */ - if ((!capable(CAP_SYS_RAWIO)) || (bdev != bdev->bd_contains)) - return -EPERM; - switch (cmd) { case MMC_IOC_CMD: return mmc_blk_ioctl_cmd(bdev, diff --git a/drivers/mmc/host/atmel-mci.c b/drivers/mmc/host/atmel-mci.c index 851ccd9ac..25c179592 100644 --- a/drivers/mmc/host/atmel-mci.c +++ b/drivers/mmc/host/atmel-mci.c @@ -2443,7 +2443,7 @@ static int atmci_configure_dma(struct atmel_mci *host) struct mci_platform_data *pdata = host->pdev->dev.platform_data; dma_cap_mask_t mask; - if (!pdata->dma_filter) + if (!pdata || !pdata->dma_filter) return -ENODEV; dma_cap_zero(mask); diff --git a/drivers/mmc/host/mmc_spi.c b/drivers/mmc/host/mmc_spi.c index 3446097a4..e77d79c8c 100644 --- a/drivers/mmc/host/mmc_spi.c +++ b/drivers/mmc/host/mmc_spi.c @@ -1442,6 +1442,12 @@ static int mmc_spi_probe(struct spi_device *spi) host->pdata->cd_debounce); if (status != 0) goto fail_add_host; + + /* The platform has a CD GPIO signal that may support + * interrupts, so let mmc_gpiod_request_cd_irq() decide + * if polling is needed or not. + */ + mmc->caps &= ~MMC_CAP_NEEDS_POLL; mmc_gpiod_request_cd_irq(mmc); } diff --git a/drivers/mmc/host/sdhci-pxav3.c b/drivers/mmc/host/sdhci-pxav3.c index f5edf9d3a..c7f27fe48 100644 --- a/drivers/mmc/host/sdhci-pxav3.c +++ b/drivers/mmc/host/sdhci-pxav3.c @@ -137,6 +137,10 @@ static int armada_38x_quirks(struct platform_device *pdev, host->quirks &= ~SDHCI_QUIRK_CAP_CLOCK_BASE_BROKEN; host->quirks |= SDHCI_QUIRK_MISSING_CAPS; + + host->caps = sdhci_readl(host, SDHCI_CAPABILITIES); + host->caps1 = sdhci_readl(host, SDHCI_CAPABILITIES_1); + res = platform_get_resource_byname(pdev, IORESOURCE_MEM, "conf-sdio3"); if (res) { @@ -150,7 +154,6 @@ static int armada_38x_quirks(struct platform_device *pdev, * Configuration register, if the adjustment is not done, * remove them from the capabilities. */ - host->caps1 = sdhci_readl(host, SDHCI_CAPABILITIES_1); host->caps1 &= ~(SDHCI_SUPPORT_SDR50 | SDHCI_SUPPORT_DDR50); dev_warn(&pdev->dev, "conf-sdio3 register not found: disabling SDR50 and DDR50 modes.\nConsider updating your dtb\n"); @@ -161,7 +164,6 @@ static int armada_38x_quirks(struct platform_device *pdev, * controller has different capabilities than the ones shown * in its registers */ - host->caps = sdhci_readl(host, SDHCI_CAPABILITIES); if (of_property_read_bool(np, "no-1-8-v")) { host->caps &= ~SDHCI_CAN_VDD_180; host->mmc->caps &= ~MMC_CAP_1_8V_DDR; diff --git a/drivers/mmc/host/sdhci-tegra.c b/drivers/mmc/host/sdhci-tegra.c index 83c4bf7bc..0004721cd 100644 --- a/drivers/mmc/host/sdhci-tegra.c +++ b/drivers/mmc/host/sdhci-tegra.c @@ -147,10 +147,16 @@ static void tegra_sdhci_reset(struct sdhci_host *host, u8 mask) /* Advertise UHS modes as supported by host */ if (soc_data->nvquirks & NVQUIRK_ENABLE_SDR50) misc_ctrl |= SDHCI_MISC_CTRL_ENABLE_SDR50; + else + misc_ctrl &= ~SDHCI_MISC_CTRL_ENABLE_SDR50; if (soc_data->nvquirks & NVQUIRK_ENABLE_DDR50) misc_ctrl |= SDHCI_MISC_CTRL_ENABLE_DDR50; + else + misc_ctrl &= ~SDHCI_MISC_CTRL_ENABLE_DDR50; if (soc_data->nvquirks & NVQUIRK_ENABLE_SDR104) misc_ctrl |= SDHCI_MISC_CTRL_ENABLE_SDR104; + else + misc_ctrl &= ~SDHCI_MISC_CTRL_ENABLE_SDR104; sdhci_writel(host, misc_ctrl, SDHCI_TEGRA_VENDOR_MISC_CTRL); clk_ctrl = sdhci_readl(host, SDHCI_TEGRA_VENDOR_CLOCK_CTRL); @@ -188,7 +194,7 @@ static void tegra_sdhci_set_clock(struct sdhci_host *host, unsigned int clock) unsigned long host_clk; if (!clock) - return; + return sdhci_set_clock(host, clock); host_clk = tegra_host->ddr_signaling ? clock * 2 : clock; clk_set_rate(pltfm_host->clk, host_clk); @@ -335,6 +341,10 @@ static const struct sdhci_pltfm_data sdhci_tegra114_pdata = { static const struct sdhci_tegra_soc_data soc_data_tegra114 = { .pdata = &sdhci_tegra114_pdata, +}; + +static const struct sdhci_tegra_soc_data soc_data_tegra124 = { + .pdata = &sdhci_tegra114_pdata, .nvquirks = NVQUIRK_ENABLE_SDR50 | NVQUIRK_ENABLE_DDR50 | NVQUIRK_ENABLE_SDR104, @@ -357,7 +367,7 @@ static const struct sdhci_tegra_soc_data soc_data_tegra210 = { static const struct of_device_id sdhci_tegra_dt_match[] = { { .compatible = "nvidia,tegra210-sdhci", .data = &soc_data_tegra210 }, - { .compatible = "nvidia,tegra124-sdhci", .data = &soc_data_tegra114 }, + { .compatible = "nvidia,tegra124-sdhci", .data = &soc_data_tegra124 }, { .compatible = "nvidia,tegra114-sdhci", .data = &soc_data_tegra114 }, { .compatible = "nvidia,tegra30-sdhci", .data = &soc_data_tegra30 }, { .compatible = "nvidia,tegra20-sdhci", .data = &soc_data_tegra20 }, diff --git a/drivers/mmc/host/sdhci.c b/drivers/mmc/host/sdhci.c index add9fdfd1..8059d7248 100644 --- a/drivers/mmc/host/sdhci.c +++ b/drivers/mmc/host/sdhci.c @@ -465,8 +465,6 @@ static void sdhci_adma_mark_end(void *desc) static int sdhci_adma_table_pre(struct sdhci_host *host, struct mmc_data *data) { - int direction; - void *desc; void *align; dma_addr_t addr; @@ -483,20 +481,9 @@ static int sdhci_adma_table_pre(struct sdhci_host *host, * We currently guess that it is LE. */ - if (data->flags & MMC_DATA_READ) - direction = DMA_FROM_DEVICE; - else - direction = DMA_TO_DEVICE; - - host->align_addr = dma_map_single(mmc_dev(host->mmc), - host->align_buffer, host->align_buffer_sz, direction); - if (dma_mapping_error(mmc_dev(host->mmc), host->align_addr)) - goto fail; - BUG_ON(host->align_addr & SDHCI_ADMA2_MASK); - host->sg_count = sdhci_pre_dma_transfer(host, data); if (host->sg_count < 0) - goto unmap_align; + return -EINVAL; desc = host->adma_table; align = host->align_buffer; @@ -570,22 +557,7 @@ static int sdhci_adma_table_pre(struct sdhci_host *host, /* nop, end, valid */ sdhci_adma_write_desc(host, desc, 0, 0, ADMA2_NOP_END_VALID); } - - /* - * Resync align buffer as we might have changed it. - */ - if (data->flags & MMC_DATA_WRITE) { - dma_sync_single_for_device(mmc_dev(host->mmc), - host->align_addr, host->align_buffer_sz, direction); - } - return 0; - -unmap_align: - dma_unmap_single(mmc_dev(host->mmc), host->align_addr, - host->align_buffer_sz, direction); -fail: - return -EINVAL; } static void sdhci_adma_table_post(struct sdhci_host *host, @@ -605,9 +577,6 @@ static void sdhci_adma_table_post(struct sdhci_host *host, else direction = DMA_TO_DEVICE; - dma_unmap_single(mmc_dev(host->mmc), host->align_addr, - host->align_buffer_sz, direction); - /* Do a quick scan of the SG list for any unaligned mappings */ has_unaligned = false; for_each_sg(data->sg, sg, host->sg_count, i) @@ -666,9 +635,20 @@ static u8 sdhci_calc_timeout(struct sdhci_host *host, struct mmc_command *cmd) if (!data) target_timeout = cmd->busy_timeout * 1000; else { - target_timeout = data->timeout_ns / 1000; - if (host->clock) - target_timeout += data->timeout_clks / host->clock; + target_timeout = DIV_ROUND_UP(data->timeout_ns, 1000); + if (host->clock && data->timeout_clks) { + unsigned long long val; + + /* + * data->timeout_clks is in units of clock cycles. + * host->clock is in Hz. target_timeout is in us. + * Hence, us = 1000000 * cycles / Hz. Round up. + */ + val = 1000000 * data->timeout_clks; + if (do_div(val, host->clock)) + target_timeout++; + target_timeout += val; + } } /* @@ -1003,6 +983,9 @@ void sdhci_send_command(struct sdhci_host *host, struct mmc_command *cmd) WARN_ON(host->cmd); + /* Initially, a command has no error */ + cmd->error = 0; + /* Wait max 10 ms */ timeout = 10; @@ -1097,8 +1080,6 @@ static void sdhci_finish_command(struct sdhci_host *host) } } - host->cmd->error = 0; - /* Finished CMD23, now send actual command. */ if (host->cmd == host->mrq->sbc) { host->cmd = NULL; @@ -2114,14 +2095,13 @@ static void sdhci_post_req(struct mmc_host *mmc, struct mmc_request *mrq, struct sdhci_host *host = mmc_priv(mmc); struct mmc_data *data = mrq->data; - if (host->flags & SDHCI_REQ_USE_DMA) { - if (data->host_cookie == COOKIE_GIVEN || - data->host_cookie == COOKIE_MAPPED) - dma_unmap_sg(mmc_dev(host->mmc), data->sg, data->sg_len, - data->flags & MMC_DATA_WRITE ? - DMA_TO_DEVICE : DMA_FROM_DEVICE); - data->host_cookie = COOKIE_UNMAPPED; - } + if (data->host_cookie == COOKIE_GIVEN || + data->host_cookie == COOKIE_MAPPED) + dma_unmap_sg(mmc_dev(host->mmc), data->sg, data->sg_len, + data->flags & MMC_DATA_WRITE ? + DMA_TO_DEVICE : DMA_FROM_DEVICE); + + data->host_cookie = COOKIE_UNMAPPED; } static int sdhci_pre_dma_transfer(struct sdhci_host *host, @@ -2238,6 +2218,22 @@ static void sdhci_tasklet_finish(unsigned long param) mrq = host->mrq; /* + * Always unmap the data buffers if they were mapped by + * sdhci_prepare_data() whenever we finish with a request. + * This avoids leaking DMA mappings on error. + */ + if (host->flags & SDHCI_REQ_USE_DMA) { + struct mmc_data *data = mrq->data; + + if (data && data->host_cookie == COOKIE_MAPPED) { + dma_unmap_sg(mmc_dev(host->mmc), data->sg, data->sg_len, + (data->flags & MMC_DATA_READ) ? + DMA_FROM_DEVICE : DMA_TO_DEVICE); + data->host_cookie = COOKIE_UNMAPPED; + } + } + + /* * The controller needs a reset of internal state machines * upon error conditions. */ @@ -2322,13 +2318,30 @@ static void sdhci_cmd_irq(struct sdhci_host *host, u32 intmask, u32 *mask) return; } - if (intmask & SDHCI_INT_TIMEOUT) - host->cmd->error = -ETIMEDOUT; - else if (intmask & (SDHCI_INT_CRC | SDHCI_INT_END_BIT | - SDHCI_INT_INDEX)) - host->cmd->error = -EILSEQ; + if (intmask & (SDHCI_INT_TIMEOUT | SDHCI_INT_CRC | + SDHCI_INT_END_BIT | SDHCI_INT_INDEX)) { + if (intmask & SDHCI_INT_TIMEOUT) + host->cmd->error = -ETIMEDOUT; + else + host->cmd->error = -EILSEQ; + + /* + * If this command initiates a data phase and a response + * CRC error is signalled, the card can start transferring + * data - the card may have received the command without + * error. We must not terminate the mmc_request early. + * + * If the card did not receive the command or returned an + * error which prevented it sending data, the data phase + * will time out. + */ + if (host->cmd->data && + (intmask & (SDHCI_INT_CRC | SDHCI_INT_TIMEOUT)) == + SDHCI_INT_CRC) { + host->cmd = NULL; + return; + } - if (host->cmd->error) { tasklet_schedule(&host->finish_tasklet); return; } @@ -2967,14 +2980,21 @@ int sdhci_add_host(struct sdhci_host *host) &host->adma_addr, GFP_KERNEL); host->align_buffer_sz = SDHCI_MAX_SEGS * SDHCI_ADMA2_ALIGN; - host->align_buffer = kmalloc(host->align_buffer_sz, GFP_KERNEL); + host->align_buffer = dma_alloc_coherent(mmc_dev(mmc), + host->align_buffer_sz, + &host->align_addr, + GFP_KERNEL); if (!host->adma_table || !host->align_buffer) { if (host->adma_table) dma_free_coherent(mmc_dev(mmc), host->adma_table_sz, host->adma_table, host->adma_addr); - kfree(host->align_buffer); + if (host->align_buffer) + dma_free_coherent(mmc_dev(mmc), + host->align_buffer_sz, + host->align_buffer, + host->align_addr); pr_warn("%s: Unable to allocate ADMA buffers - falling back to standard DMA\n", mmc_hostname(mmc)); host->flags &= ~SDHCI_USE_ADMA; @@ -2986,10 +3006,14 @@ int sdhci_add_host(struct sdhci_host *host) host->flags &= ~SDHCI_USE_ADMA; dma_free_coherent(mmc_dev(mmc), host->adma_table_sz, host->adma_table, host->adma_addr); - kfree(host->align_buffer); + dma_free_coherent(mmc_dev(mmc), host->align_buffer_sz, + host->align_buffer, host->align_addr); host->adma_table = NULL; host->align_buffer = NULL; } + + /* dma_alloc_coherent returns page aligned and sized buffers */ + BUG_ON(host->align_addr & SDHCI_ADMA2_MASK); } /* @@ -3072,14 +3096,14 @@ int sdhci_add_host(struct sdhci_host *host) if (caps[0] & SDHCI_TIMEOUT_CLK_UNIT) host->timeout_clk *= 1000; + if (override_timeout_clk) + host->timeout_clk = override_timeout_clk; + mmc->max_busy_timeout = host->ops->get_max_timeout_count ? host->ops->get_max_timeout_count(host) : 1 << 27; mmc->max_busy_timeout /= host->timeout_clk; } - if (override_timeout_clk) - host->timeout_clk = override_timeout_clk; - mmc->caps |= MMC_CAP_SDIO_IRQ | MMC_CAP_ERASE | MMC_CAP_CMD23; mmc->caps2 |= MMC_CAP2_SDIO_IRQ_NOTHREAD; @@ -3452,7 +3476,9 @@ void sdhci_remove_host(struct sdhci_host *host, int dead) if (host->adma_table) dma_free_coherent(mmc_dev(mmc), host->adma_table_sz, host->adma_table, host->adma_addr); - kfree(host->align_buffer); + if (host->align_buffer) + dma_free_coherent(mmc_dev(mmc), host->align_buffer_sz, + host->align_buffer, host->align_addr); host->adma_table = NULL; host->align_buffer = NULL; diff --git a/drivers/mtd/onenand/onenand_base.c b/drivers/mtd/onenand/onenand_base.c index 43b3392ff..652d01832 100644 --- a/drivers/mtd/onenand/onenand_base.c +++ b/drivers/mtd/onenand/onenand_base.c @@ -2599,6 +2599,7 @@ static int onenand_default_block_markbad(struct mtd_info *mtd, loff_t ofs) */ static int onenand_block_markbad(struct mtd_info *mtd, loff_t ofs) { + struct onenand_chip *this = mtd->priv; int ret; ret = onenand_block_isbad(mtd, ofs); @@ -2610,7 +2611,7 @@ static int onenand_block_markbad(struct mtd_info *mtd, loff_t ofs) } onenand_get_device(mtd, FL_WRITING); - ret = mtd_block_markbad(mtd, ofs); + ret = this->block_markbad(mtd, ofs); onenand_release_device(mtd); return ret; } diff --git a/drivers/net/ethernet/marvell/mvneta.c b/drivers/net/ethernet/marvell/mvneta.c index b0ae69f84..acb1c5b2b 100644 --- a/drivers/net/ethernet/marvell/mvneta.c +++ b/drivers/net/ethernet/marvell/mvneta.c @@ -3720,7 +3720,7 @@ static int mvneta_probe(struct platform_device *pdev) dev->features = NETIF_F_SG | NETIF_F_IP_CSUM | NETIF_F_TSO; dev->hw_features |= dev->features; dev->vlan_features |= dev->features; - dev->priv_flags |= IFF_UNICAST_FLT; + dev->priv_flags |= IFF_UNICAST_FLT | IFF_LIVE_ADDR_CHANGE; dev->gso_max_segs = MVNETA_MAX_TSO_SEGS; err = register_netdev(dev); diff --git a/drivers/net/irda/irtty-sir.c b/drivers/net/irda/irtty-sir.c index 696852eb2..7a3f990c1 100644 --- a/drivers/net/irda/irtty-sir.c +++ b/drivers/net/irda/irtty-sir.c @@ -430,16 +430,6 @@ static int irtty_open(struct tty_struct *tty) /* Module stuff handled via irda_ldisc.owner - Jean II */ - /* First make sure we're not already connected. */ - if (tty->disc_data != NULL) { - priv = tty->disc_data; - if (priv && priv->magic == IRTTY_MAGIC) { - ret = -EEXIST; - goto out; - } - tty->disc_data = NULL; /* ### */ - } - /* stop the underlying driver */ irtty_stop_receiver(tty, TRUE); if (tty->ops->stop) diff --git a/drivers/net/rionet.c b/drivers/net/rionet.c index 01f08a775..e7034c55e 100644 --- a/drivers/net/rionet.c +++ b/drivers/net/rionet.c @@ -280,7 +280,7 @@ static void rionet_outb_msg_event(struct rio_mport *mport, void *dev_id, int mbo struct net_device *ndev = dev_id; struct rionet_private *rnet = netdev_priv(ndev); - spin_lock(&rnet->lock); + spin_lock(&rnet->tx_lock); if (netif_msg_intr(rnet)) printk(KERN_INFO @@ -299,7 +299,7 @@ static void rionet_outb_msg_event(struct rio_mport *mport, void *dev_id, int mbo if (rnet->tx_cnt < RIONET_TX_RING_SIZE) netif_wake_queue(ndev); - spin_unlock(&rnet->lock); + spin_unlock(&rnet->tx_lock); } static int rionet_open(struct net_device *ndev) diff --git a/drivers/net/wireless/broadcom/brcm80211/brcmfmac/flowring.c b/drivers/net/wireless/broadcom/brcm80211/brcmfmac/flowring.c index 2ca783fa5..7e269f9aa 100644 --- a/drivers/net/wireless/broadcom/brcm80211/brcmfmac/flowring.c +++ b/drivers/net/wireless/broadcom/brcm80211/brcmfmac/flowring.c @@ -32,7 +32,7 @@ #define BRCMF_FLOWRING_LOW (BRCMF_FLOWRING_HIGH - 256) #define BRCMF_FLOWRING_INVALID_IFIDX 0xff -#define BRCMF_FLOWRING_HASH_AP(da, fifo, ifidx) (da[5] + fifo + ifidx * 16) +#define BRCMF_FLOWRING_HASH_AP(da, fifo, ifidx) (da[5] * 2 + fifo + ifidx * 16) #define BRCMF_FLOWRING_HASH_STA(fifo, ifidx) (fifo + ifidx * 16) static const u8 brcmf_flowring_prio2fifo[] = { @@ -68,7 +68,7 @@ u32 brcmf_flowring_lookup(struct brcmf_flowring *flow, u8 da[ETH_ALEN], u8 prio, u8 ifidx) { struct brcmf_flowring_hash *hash; - u8 hash_idx; + u16 hash_idx; u32 i; bool found; bool sta; @@ -88,6 +88,7 @@ u32 brcmf_flowring_lookup(struct brcmf_flowring *flow, u8 da[ETH_ALEN], } hash_idx = sta ? BRCMF_FLOWRING_HASH_STA(fifo, ifidx) : BRCMF_FLOWRING_HASH_AP(mac, fifo, ifidx); + hash_idx &= (BRCMF_FLOWRING_HASHSIZE - 1); found = false; hash = flow->hash; for (i = 0; i < BRCMF_FLOWRING_HASHSIZE; i++) { @@ -98,6 +99,7 @@ u32 brcmf_flowring_lookup(struct brcmf_flowring *flow, u8 da[ETH_ALEN], break; } hash_idx++; + hash_idx &= (BRCMF_FLOWRING_HASHSIZE - 1); } if (found) return hash[hash_idx].flowid; @@ -111,7 +113,7 @@ u32 brcmf_flowring_create(struct brcmf_flowring *flow, u8 da[ETH_ALEN], { struct brcmf_flowring_ring *ring; struct brcmf_flowring_hash *hash; - u8 hash_idx; + u16 hash_idx; u32 i; bool found; u8 fifo; @@ -131,6 +133,7 @@ u32 brcmf_flowring_create(struct brcmf_flowring *flow, u8 da[ETH_ALEN], } hash_idx = sta ? BRCMF_FLOWRING_HASH_STA(fifo, ifidx) : BRCMF_FLOWRING_HASH_AP(mac, fifo, ifidx); + hash_idx &= (BRCMF_FLOWRING_HASHSIZE - 1); found = false; hash = flow->hash; for (i = 0; i < BRCMF_FLOWRING_HASHSIZE; i++) { @@ -140,6 +143,7 @@ u32 brcmf_flowring_create(struct brcmf_flowring *flow, u8 da[ETH_ALEN], break; } hash_idx++; + hash_idx &= (BRCMF_FLOWRING_HASHSIZE - 1); } if (found) { for (i = 0; i < flow->nrofrings; i++) { @@ -169,7 +173,7 @@ u32 brcmf_flowring_create(struct brcmf_flowring *flow, u8 da[ETH_ALEN], } -u8 brcmf_flowring_tid(struct brcmf_flowring *flow, u8 flowid) +u8 brcmf_flowring_tid(struct brcmf_flowring *flow, u16 flowid) { struct brcmf_flowring_ring *ring; @@ -179,7 +183,7 @@ u8 brcmf_flowring_tid(struct brcmf_flowring *flow, u8 flowid) } -static void brcmf_flowring_block(struct brcmf_flowring *flow, u8 flowid, +static void brcmf_flowring_block(struct brcmf_flowring *flow, u16 flowid, bool blocked) { struct brcmf_flowring_ring *ring; @@ -228,10 +232,10 @@ static void brcmf_flowring_block(struct brcmf_flowring *flow, u8 flowid, } -void brcmf_flowring_delete(struct brcmf_flowring *flow, u8 flowid) +void brcmf_flowring_delete(struct brcmf_flowring *flow, u16 flowid) { struct brcmf_flowring_ring *ring; - u8 hash_idx; + u16 hash_idx; struct sk_buff *skb; ring = flow->rings[flowid]; @@ -253,7 +257,7 @@ void brcmf_flowring_delete(struct brcmf_flowring *flow, u8 flowid) } -u32 brcmf_flowring_enqueue(struct brcmf_flowring *flow, u8 flowid, +u32 brcmf_flowring_enqueue(struct brcmf_flowring *flow, u16 flowid, struct sk_buff *skb) { struct brcmf_flowring_ring *ring; @@ -279,7 +283,7 @@ u32 brcmf_flowring_enqueue(struct brcmf_flowring *flow, u8 flowid, } -struct sk_buff *brcmf_flowring_dequeue(struct brcmf_flowring *flow, u8 flowid) +struct sk_buff *brcmf_flowring_dequeue(struct brcmf_flowring *flow, u16 flowid) { struct brcmf_flowring_ring *ring; struct sk_buff *skb; @@ -300,7 +304,7 @@ struct sk_buff *brcmf_flowring_dequeue(struct brcmf_flowring *flow, u8 flowid) } -void brcmf_flowring_reinsert(struct brcmf_flowring *flow, u8 flowid, +void brcmf_flowring_reinsert(struct brcmf_flowring *flow, u16 flowid, struct sk_buff *skb) { struct brcmf_flowring_ring *ring; @@ -311,7 +315,7 @@ void brcmf_flowring_reinsert(struct brcmf_flowring *flow, u8 flowid, } -u32 brcmf_flowring_qlen(struct brcmf_flowring *flow, u8 flowid) +u32 brcmf_flowring_qlen(struct brcmf_flowring *flow, u16 flowid) { struct brcmf_flowring_ring *ring; @@ -326,7 +330,7 @@ u32 brcmf_flowring_qlen(struct brcmf_flowring *flow, u8 flowid) } -void brcmf_flowring_open(struct brcmf_flowring *flow, u8 flowid) +void brcmf_flowring_open(struct brcmf_flowring *flow, u16 flowid) { struct brcmf_flowring_ring *ring; @@ -340,10 +344,10 @@ void brcmf_flowring_open(struct brcmf_flowring *flow, u8 flowid) } -u8 brcmf_flowring_ifidx_get(struct brcmf_flowring *flow, u8 flowid) +u8 brcmf_flowring_ifidx_get(struct brcmf_flowring *flow, u16 flowid) { struct brcmf_flowring_ring *ring; - u8 hash_idx; + u16 hash_idx; ring = flow->rings[flowid]; hash_idx = ring->hash_id; @@ -384,7 +388,7 @@ void brcmf_flowring_detach(struct brcmf_flowring *flow) struct brcmf_pub *drvr = bus_if->drvr; struct brcmf_flowring_tdls_entry *search; struct brcmf_flowring_tdls_entry *remove; - u8 flowid; + u16 flowid; for (flowid = 0; flowid < flow->nrofrings; flowid++) { if (flow->rings[flowid]) @@ -408,7 +412,7 @@ void brcmf_flowring_configure_addr_mode(struct brcmf_flowring *flow, int ifidx, struct brcmf_bus *bus_if = dev_get_drvdata(flow->dev); struct brcmf_pub *drvr = bus_if->drvr; u32 i; - u8 flowid; + u16 flowid; if (flow->addr_mode[ifidx] != addr_mode) { for (i = 0; i < ARRAY_SIZE(flow->hash); i++) { @@ -434,7 +438,7 @@ void brcmf_flowring_delete_peer(struct brcmf_flowring *flow, int ifidx, struct brcmf_flowring_tdls_entry *prev; struct brcmf_flowring_tdls_entry *search; u32 i; - u8 flowid; + u16 flowid; bool sta; sta = (flow->addr_mode[ifidx] == ADDR_INDIRECT); diff --git a/drivers/net/wireless/broadcom/brcm80211/brcmfmac/flowring.h b/drivers/net/wireless/broadcom/brcm80211/brcmfmac/flowring.h index 95fd1c967..068e68d94 100644 --- a/drivers/net/wireless/broadcom/brcm80211/brcmfmac/flowring.h +++ b/drivers/net/wireless/broadcom/brcm80211/brcmfmac/flowring.h @@ -16,7 +16,7 @@ #define BRCMFMAC_FLOWRING_H -#define BRCMF_FLOWRING_HASHSIZE 256 +#define BRCMF_FLOWRING_HASHSIZE 512 /* has to be 2^x */ #define BRCMF_FLOWRING_INVALID_ID 0xFFFFFFFF @@ -24,7 +24,7 @@ struct brcmf_flowring_hash { u8 mac[ETH_ALEN]; u8 fifo; u8 ifidx; - u8 flowid; + u16 flowid; }; enum ring_status { @@ -61,16 +61,16 @@ u32 brcmf_flowring_lookup(struct brcmf_flowring *flow, u8 da[ETH_ALEN], u8 prio, u8 ifidx); u32 brcmf_flowring_create(struct brcmf_flowring *flow, u8 da[ETH_ALEN], u8 prio, u8 ifidx); -void brcmf_flowring_delete(struct brcmf_flowring *flow, u8 flowid); -void brcmf_flowring_open(struct brcmf_flowring *flow, u8 flowid); -u8 brcmf_flowring_tid(struct brcmf_flowring *flow, u8 flowid); -u32 brcmf_flowring_enqueue(struct brcmf_flowring *flow, u8 flowid, +void brcmf_flowring_delete(struct brcmf_flowring *flow, u16 flowid); +void brcmf_flowring_open(struct brcmf_flowring *flow, u16 flowid); +u8 brcmf_flowring_tid(struct brcmf_flowring *flow, u16 flowid); +u32 brcmf_flowring_enqueue(struct brcmf_flowring *flow, u16 flowid, struct sk_buff *skb); -struct sk_buff *brcmf_flowring_dequeue(struct brcmf_flowring *flow, u8 flowid); -void brcmf_flowring_reinsert(struct brcmf_flowring *flow, u8 flowid, +struct sk_buff *brcmf_flowring_dequeue(struct brcmf_flowring *flow, u16 flowid); +void brcmf_flowring_reinsert(struct brcmf_flowring *flow, u16 flowid, struct sk_buff *skb); -u32 brcmf_flowring_qlen(struct brcmf_flowring *flow, u8 flowid); -u8 brcmf_flowring_ifidx_get(struct brcmf_flowring *flow, u8 flowid); +u32 brcmf_flowring_qlen(struct brcmf_flowring *flow, u16 flowid); +u8 brcmf_flowring_ifidx_get(struct brcmf_flowring *flow, u16 flowid); struct brcmf_flowring *brcmf_flowring_attach(struct device *dev, u16 nrofrings); void brcmf_flowring_detach(struct brcmf_flowring *flow); void brcmf_flowring_configure_addr_mode(struct brcmf_flowring *flow, int ifidx, diff --git a/drivers/net/wireless/broadcom/brcm80211/brcmfmac/msgbuf.c b/drivers/net/wireless/broadcom/brcm80211/brcmfmac/msgbuf.c index c2bdb9174..922966734 100644 --- a/drivers/net/wireless/broadcom/brcm80211/brcmfmac/msgbuf.c +++ b/drivers/net/wireless/broadcom/brcm80211/brcmfmac/msgbuf.c @@ -677,7 +677,7 @@ static u32 brcmf_msgbuf_flowring_create(struct brcmf_msgbuf *msgbuf, int ifidx, } -static void brcmf_msgbuf_txflow(struct brcmf_msgbuf *msgbuf, u8 flowid) +static void brcmf_msgbuf_txflow(struct brcmf_msgbuf *msgbuf, u16 flowid) { struct brcmf_flowring *flow = msgbuf->flow; struct brcmf_commonring *commonring; @@ -1310,7 +1310,7 @@ int brcmf_proto_msgbuf_rx_trigger(struct device *dev) } -void brcmf_msgbuf_delete_flowring(struct brcmf_pub *drvr, u8 flowid) +void brcmf_msgbuf_delete_flowring(struct brcmf_pub *drvr, u16 flowid) { struct brcmf_msgbuf *msgbuf = (struct brcmf_msgbuf *)drvr->proto->pd; struct msgbuf_tx_flowring_delete_req *delete; @@ -1415,6 +1415,13 @@ int brcmf_proto_msgbuf_attach(struct brcmf_pub *drvr) u32 count; if_msgbuf = drvr->bus_if->msgbuf; + + if (if_msgbuf->nrof_flowrings >= BRCMF_FLOWRING_HASHSIZE) { + brcmf_err("driver not configured for this many flowrings %d\n", + if_msgbuf->nrof_flowrings); + if_msgbuf->nrof_flowrings = BRCMF_FLOWRING_HASHSIZE - 1; + } + msgbuf = kzalloc(sizeof(*msgbuf), GFP_KERNEL); if (!msgbuf) goto fail; diff --git a/drivers/net/wireless/broadcom/brcm80211/brcmfmac/msgbuf.h b/drivers/net/wireless/broadcom/brcm80211/brcmfmac/msgbuf.h index 3d513e407..ee6906a3c 100644 --- a/drivers/net/wireless/broadcom/brcm80211/brcmfmac/msgbuf.h +++ b/drivers/net/wireless/broadcom/brcm80211/brcmfmac/msgbuf.h @@ -33,7 +33,7 @@ int brcmf_proto_msgbuf_rx_trigger(struct device *dev); -void brcmf_msgbuf_delete_flowring(struct brcmf_pub *drvr, u8 flowid); +void brcmf_msgbuf_delete_flowring(struct brcmf_pub *drvr, u16 flowid); int brcmf_proto_msgbuf_attach(struct brcmf_pub *drvr); void brcmf_proto_msgbuf_detach(struct brcmf_pub *drvr); #else diff --git a/drivers/net/wireless/ralink/rt2x00/rt2800usb.c b/drivers/net/wireless/ralink/rt2x00/rt2800usb.c index ee7d97482..e6a14a45f 100644 --- a/drivers/net/wireless/ralink/rt2x00/rt2800usb.c +++ b/drivers/net/wireless/ralink/rt2x00/rt2800usb.c @@ -1026,6 +1026,7 @@ static struct usb_device_id rt2800usb_device_table[] = { { USB_DEVICE(0x0411, 0x01a2) }, { USB_DEVICE(0x0411, 0x01ee) }, { USB_DEVICE(0x0411, 0x01a8) }, + { USB_DEVICE(0x0411, 0x01fd) }, /* Corega */ { USB_DEVICE(0x07aa, 0x002f) }, { USB_DEVICE(0x07aa, 0x003c) }, diff --git a/drivers/nvdimm/bus.c b/drivers/nvdimm/bus.c index 5d28e9405..576eb7013 100644 --- a/drivers/nvdimm/bus.c +++ b/drivers/nvdimm/bus.c @@ -513,10 +513,10 @@ static int __nd_ioctl(struct nvdimm_bus *nvdimm_bus, struct nvdimm *nvdimm, /* fail write commands (when read-only) */ if (read_only) - switch (ioctl_cmd) { - case ND_IOCTL_VENDOR: - case ND_IOCTL_SET_CONFIG_DATA: - case ND_IOCTL_ARS_START: + switch (cmd) { + case ND_CMD_VENDOR: + case ND_CMD_SET_CONFIG_DATA: + case ND_CMD_ARS_START: dev_dbg(&nvdimm_bus->dev, "'%s' command while read-only.\n", nvdimm ? nvdimm_cmd_name(cmd) : nvdimm_bus_cmd_name(cmd)); diff --git a/drivers/nvdimm/pmem.c b/drivers/nvdimm/pmem.c index 8d0b54670..544b802a5 100644 --- a/drivers/nvdimm/pmem.c +++ b/drivers/nvdimm/pmem.c @@ -66,22 +66,25 @@ static int pmem_do_bvec(struct pmem_device *pmem, struct page *page, unsigned int len, unsigned int off, int rw, sector_t sector) { + int rc = 0; void *mem = kmap_atomic(page); phys_addr_t pmem_off = sector * 512 + pmem->data_offset; void __pmem *pmem_addr = pmem->virt_addr + pmem_off; if (rw == READ) { if (unlikely(is_bad_pmem(&pmem->bb, sector, len))) - return -EIO; - memcpy_from_pmem(mem + off, pmem_addr, len); - flush_dcache_page(page); + rc = -EIO; + else { + memcpy_from_pmem(mem + off, pmem_addr, len); + flush_dcache_page(page); + } } else { flush_dcache_page(page); memcpy_to_pmem(pmem_addr, mem + off, len); } kunmap_atomic(mem); - return 0; + return rc; } static blk_qc_t pmem_make_request(struct request_queue *q, struct bio *bio) diff --git a/drivers/of/of_reserved_mem.c b/drivers/of/of_reserved_mem.c index 1a3556a9e..ed01c0172 100644 --- a/drivers/of/of_reserved_mem.c +++ b/drivers/of/of_reserved_mem.c @@ -32,11 +32,13 @@ int __init __weak early_init_dt_alloc_reserved_memory_arch(phys_addr_t size, phys_addr_t align, phys_addr_t start, phys_addr_t end, bool nomap, phys_addr_t *res_base) { + phys_addr_t base; /* * We use __memblock_alloc_base() because memblock_alloc_base() * panic()s on allocation failure. */ - phys_addr_t base = __memblock_alloc_base(size, align, end); + end = !end ? MEMBLOCK_ALLOC_ANYWHERE : end; + base = __memblock_alloc_base(size, align, end); if (!base) return -ENOMEM; diff --git a/drivers/pci/probe.c b/drivers/pci/probe.c index 6d7ab9bb0..6b0056e9c 100644 --- a/drivers/pci/probe.c +++ b/drivers/pci/probe.c @@ -179,6 +179,9 @@ int __pci_read_base(struct pci_dev *dev, enum pci_bar_type type, u16 orig_cmd; struct pci_bus_region region, inverted_region; + if (dev->non_compliant_bars) + return 0; + mask = type ? PCI_ROM_ADDRESS_MASK : ~0; /* No printks while decoding is disabled! */ @@ -1171,6 +1174,7 @@ static void pci_msi_setup_pci_dev(struct pci_dev *dev) int pci_setup_device(struct pci_dev *dev) { u32 class; + u16 cmd; u8 hdr_type; int pos = 0; struct pci_bus_region region; @@ -1214,6 +1218,16 @@ int pci_setup_device(struct pci_dev *dev) /* device class may be changed after fixup */ class = dev->class >> 8; + if (dev->non_compliant_bars) { + pci_read_config_word(dev, PCI_COMMAND, &cmd); + if (cmd & (PCI_COMMAND_IO | PCI_COMMAND_MEMORY)) { + dev_info(&dev->dev, "device has non-compliant BARs; disabling IO/MEM decoding\n"); + cmd &= ~PCI_COMMAND_IO; + cmd &= ~PCI_COMMAND_MEMORY; + pci_write_config_word(dev, PCI_COMMAND, cmd); + } + } + switch (dev->hdr_type) { /* header type */ case PCI_HEADER_TYPE_NORMAL: /* standard header */ if (class == PCI_CLASS_BRIDGE_PCI) diff --git a/drivers/pinctrl/bcm/pinctrl-bcm2835.c b/drivers/pinctrl/bcm/pinctrl-bcm2835.c index 0f5997ceb..08b1d93da 100644 --- a/drivers/pinctrl/bcm/pinctrl-bcm2835.c +++ b/drivers/pinctrl/bcm/pinctrl-bcm2835.c @@ -779,7 +779,7 @@ static int bcm2835_pctl_dt_node_to_map(struct pinctrl_dev *pctldev, } if (num_pulls) { err = of_property_read_u32_index(np, "brcm,pull", - (num_funcs > 1) ? i : 0, &pull); + (num_pulls > 1) ? i : 0, &pull); if (err) goto out; err = bcm2835_pctl_dt_node_to_map_pull(pc, np, pin, diff --git a/drivers/platform/x86/ideapad-laptop.c b/drivers/platform/x86/ideapad-laptop.c index d78ee151c..be3bc2f4e 100644 --- a/drivers/platform/x86/ideapad-laptop.c +++ b/drivers/platform/x86/ideapad-laptop.c @@ -865,6 +865,20 @@ static const struct dmi_system_id no_hw_rfkill_list[] = { }, }, { + .ident = "Lenovo ideapad Y700-15ISK", + .matches = { + DMI_MATCH(DMI_SYS_VENDOR, "LENOVO"), + DMI_MATCH(DMI_PRODUCT_VERSION, "Lenovo ideapad Y700-15ISK"), + }, + }, + { + .ident = "Lenovo ideapad Y700 Touch-15ISK", + .matches = { + DMI_MATCH(DMI_SYS_VENDOR, "LENOVO"), + DMI_MATCH(DMI_PRODUCT_VERSION, "Lenovo ideapad Y700 Touch-15ISK"), + }, + }, + { .ident = "Lenovo ideapad Y700-17ISK", .matches = { DMI_MATCH(DMI_SYS_VENDOR, "LENOVO"), diff --git a/drivers/scsi/NCR5380.c b/drivers/scsi/NCR5380.c index d72867257..3eff2a69f 100644 --- a/drivers/scsi/NCR5380.c +++ b/drivers/scsi/NCR5380.c @@ -760,7 +760,7 @@ static struct scsi_cmnd *dequeue_next_cmd(struct Scsi_Host *instance) struct NCR5380_cmd *ncmd; struct scsi_cmnd *cmd; - if (list_empty(&hostdata->autosense)) { + if (hostdata->sensing || list_empty(&hostdata->autosense)) { list_for_each_entry(ncmd, &hostdata->unissued, list) { cmd = NCR5380_to_scmd(ncmd); dsprintk(NDEBUG_QUEUES, instance, "dequeue: cmd=%p target=%d busy=0x%02x lun=%llu\n", @@ -793,7 +793,7 @@ static void requeue_cmd(struct Scsi_Host *instance, struct scsi_cmnd *cmd) struct NCR5380_hostdata *hostdata = shost_priv(instance); struct NCR5380_cmd *ncmd = scsi_cmd_priv(cmd); - if (hostdata->sensing) { + if (hostdata->sensing == cmd) { scsi_eh_restore_cmnd(cmd, &hostdata->ses); list_add(&ncmd->list, &hostdata->autosense); hostdata->sensing = NULL; @@ -815,15 +815,17 @@ static void NCR5380_main(struct work_struct *work) struct NCR5380_hostdata *hostdata = container_of(work, struct NCR5380_hostdata, main_task); struct Scsi_Host *instance = hostdata->host; - struct scsi_cmnd *cmd; int done; do { done = 1; spin_lock_irq(&hostdata->lock); - while (!hostdata->connected && - (cmd = dequeue_next_cmd(instance))) { + while (!hostdata->connected && !hostdata->selecting) { + struct scsi_cmnd *cmd = dequeue_next_cmd(instance); + + if (!cmd) + break; dsprintk(NDEBUG_MAIN, instance, "main: dequeued %p\n", cmd); @@ -840,8 +842,7 @@ static void NCR5380_main(struct work_struct *work) * entire unit. */ - cmd = NCR5380_select(instance, cmd); - if (!cmd) { + if (!NCR5380_select(instance, cmd)) { dsprintk(NDEBUG_MAIN, instance, "main: select complete\n"); } else { dsprintk(NDEBUG_MAIN | NDEBUG_QUEUES, instance, @@ -1056,6 +1057,11 @@ static struct scsi_cmnd *NCR5380_select(struct Scsi_Host *instance, /* Reselection interrupt */ goto out; } + if (!hostdata->selecting) { + /* Command was aborted */ + NCR5380_write(MODE_REG, MR_BASE); + goto out; + } if (err < 0) { NCR5380_write(MODE_REG, MR_BASE); shost_printk(KERN_ERR, instance, @@ -1759,9 +1765,7 @@ static void NCR5380_information_transfer(struct Scsi_Host *instance) unsigned char msgout = NOP; int sink = 0; int len; -#if defined(PSEUDO_DMA) || defined(REAL_DMA_POLL) int transfersize; -#endif unsigned char *data; unsigned char phase, tmp, extended_msg[10], old_phase = 0xff; struct scsi_cmnd *cmd; @@ -1798,6 +1802,7 @@ static void NCR5380_information_transfer(struct Scsi_Host *instance) do_abort(instance); cmd->result = DID_ERROR << 16; complete_cmd(instance, cmd); + hostdata->connected = NULL; return; #endif case PHASE_DATAIN: @@ -1847,20 +1852,23 @@ static void NCR5380_information_transfer(struct Scsi_Host *instance) sink = 1; do_abort(instance); cmd->result = DID_ERROR << 16; - complete_cmd(instance, cmd); /* XXX - need to source or sink data here, as appropriate */ } else cmd->SCp.this_residual -= transfersize - len; } else #endif /* defined(PSEUDO_DMA) || defined(REAL_DMA_POLL) */ { - spin_unlock_irq(&hostdata->lock); - NCR5380_transfer_pio(instance, &phase, - (int *)&cmd->SCp.this_residual, + /* Break up transfer into 3 ms chunks, + * presuming 6 accesses per handshake. + */ + transfersize = min((unsigned long)cmd->SCp.this_residual, + hostdata->accesses_per_ms / 2); + len = transfersize; + NCR5380_transfer_pio(instance, &phase, &len, (unsigned char **)&cmd->SCp.ptr); - spin_lock_irq(&hostdata->lock); + cmd->SCp.this_residual -= transfersize - len; } - break; + return; case PHASE_MSGIN: len = 1; data = &tmp; @@ -2292,14 +2300,17 @@ static bool list_del_cmd(struct list_head *haystack, * [disconnected -> connected ->]... * [autosense -> connected ->] done * - * If cmd is unissued then just remove it. - * If cmd is disconnected, try to select the target. - * If cmd is connected, try to send an abort message. - * If cmd is waiting for autosense, give it a chance to complete but check - * that it isn't left connected. * If cmd was not found at all then presumably it has already been completed, * in which case return SUCCESS to try to avoid further EH measures. + * * If the command has not completed yet, we must not fail to find it. + * We have no option but to forget the aborted command (even if it still + * lacks sense data). The mid-layer may re-issue a command that is in error + * recovery (see scsi_send_eh_cmnd), but the logic and data structures in + * this driver are such that a command can appear on one queue only. + * + * The lock protects driver data structures, but EH handlers also use it + * to serialize their own execution and prevent their own re-entry. */ static int NCR5380_abort(struct scsi_cmnd *cmd) @@ -2322,6 +2333,7 @@ static int NCR5380_abort(struct scsi_cmnd *cmd) "abort: removed %p from issue queue\n", cmd); cmd->result = DID_ABORT << 16; cmd->scsi_done(cmd); /* No tag or busy flag to worry about */ + goto out; } if (hostdata->selecting == cmd) { @@ -2336,58 +2348,21 @@ static int NCR5380_abort(struct scsi_cmnd *cmd) if (list_del_cmd(&hostdata->disconnected, cmd)) { dsprintk(NDEBUG_ABORT, instance, "abort: removed %p from disconnected list\n", cmd); - cmd->result = DID_ERROR << 16; - if (!hostdata->connected) - NCR5380_select(instance, cmd); - if (hostdata->connected != cmd) { - complete_cmd(instance, cmd); - result = FAILED; - goto out; - } + /* Can't call NCR5380_select() and send ABORT because that + * means releasing the lock. Need a bus reset. + */ + set_host_byte(cmd, DID_ERROR); + complete_cmd(instance, cmd); + result = FAILED; + goto out; } if (hostdata->connected == cmd) { dsprintk(NDEBUG_ABORT, instance, "abort: cmd %p is connected\n", cmd); hostdata->connected = NULL; - if (do_abort(instance)) { - set_host_byte(cmd, DID_ERROR); - complete_cmd(instance, cmd); - result = FAILED; - goto out; - } - set_host_byte(cmd, DID_ABORT); #ifdef REAL_DMA hostdata->dma_len = 0; #endif - if (cmd->cmnd[0] == REQUEST_SENSE) - complete_cmd(instance, cmd); - else { - struct NCR5380_cmd *ncmd = scsi_cmd_priv(cmd); - - /* Perform autosense for this command */ - list_add(&ncmd->list, &hostdata->autosense); - } - } - - if (list_find_cmd(&hostdata->autosense, cmd)) { - dsprintk(NDEBUG_ABORT, instance, - "abort: found %p on sense queue\n", cmd); - spin_unlock_irqrestore(&hostdata->lock, flags); - queue_work(hostdata->work_q, &hostdata->main_task); - msleep(1000); - spin_lock_irqsave(&hostdata->lock, flags); - if (list_del_cmd(&hostdata->autosense, cmd)) { - dsprintk(NDEBUG_ABORT, instance, - "abort: removed %p from sense queue\n", cmd); - set_host_byte(cmd, DID_ABORT); - complete_cmd(instance, cmd); - goto out; - } - } - - if (hostdata->connected == cmd) { - dsprintk(NDEBUG_ABORT, instance, "abort: cmd %p is connected\n", cmd); - hostdata->connected = NULL; if (do_abort(instance)) { set_host_byte(cmd, DID_ERROR); complete_cmd(instance, cmd); @@ -2395,9 +2370,14 @@ static int NCR5380_abort(struct scsi_cmnd *cmd) goto out; } set_host_byte(cmd, DID_ABORT); -#ifdef REAL_DMA - hostdata->dma_len = 0; -#endif + complete_cmd(instance, cmd); + goto out; + } + + if (list_del_cmd(&hostdata->autosense, cmd)) { + dsprintk(NDEBUG_ABORT, instance, + "abort: removed %p from sense queue\n", cmd); + set_host_byte(cmd, DID_ERROR); complete_cmd(instance, cmd); } @@ -2450,7 +2430,16 @@ static int NCR5380_bus_reset(struct scsi_cmnd *cmd) * commands! */ - hostdata->selecting = NULL; + if (list_del_cmd(&hostdata->unissued, cmd)) { + cmd->result = DID_RESET << 16; + cmd->scsi_done(cmd); + } + + if (hostdata->selecting) { + hostdata->selecting->result = DID_RESET << 16; + complete_cmd(instance, hostdata->selecting); + hostdata->selecting = NULL; + } list_for_each_entry(ncmd, &hostdata->disconnected, list) { struct scsi_cmnd *cmd = NCR5380_to_scmd(ncmd); @@ -2458,6 +2447,7 @@ static int NCR5380_bus_reset(struct scsi_cmnd *cmd) set_host_byte(cmd, DID_RESET); cmd->scsi_done(cmd); } + INIT_LIST_HEAD(&hostdata->disconnected); list_for_each_entry(ncmd, &hostdata->autosense, list) { struct scsi_cmnd *cmd = NCR5380_to_scmd(ncmd); @@ -2465,6 +2455,7 @@ static int NCR5380_bus_reset(struct scsi_cmnd *cmd) set_host_byte(cmd, DID_RESET); cmd->scsi_done(cmd); } + INIT_LIST_HEAD(&hostdata->autosense); if (hostdata->connected) { set_host_byte(hostdata->connected, DID_RESET); @@ -2472,12 +2463,6 @@ static int NCR5380_bus_reset(struct scsi_cmnd *cmd) hostdata->connected = NULL; } - if (hostdata->sensing) { - set_host_byte(hostdata->connected, DID_RESET); - complete_cmd(instance, hostdata->sensing); - hostdata->sensing = NULL; - } - for (i = 0; i < 8; ++i) hostdata->busy[i] = 0; #ifdef REAL_DMA diff --git a/drivers/scsi/aacraid/aacraid.h b/drivers/scsi/aacraid/aacraid.h index 074878b55..d044f3f27 100644 --- a/drivers/scsi/aacraid/aacraid.h +++ b/drivers/scsi/aacraid/aacraid.h @@ -944,6 +944,7 @@ struct fib { */ struct list_head fiblink; void *data; + u32 vector_no; struct hw_fib *hw_fib_va; /* Actual shared object */ dma_addr_t hw_fib_pa; /* physical address of hw_fib*/ }; @@ -2113,6 +2114,7 @@ static inline unsigned int cap_to_cyls(sector_t capacity, unsigned divisor) int aac_acquire_irq(struct aac_dev *dev); void aac_free_irq(struct aac_dev *dev); const char *aac_driverinfo(struct Scsi_Host *); +void aac_fib_vector_assign(struct aac_dev *dev); struct fib *aac_fib_alloc(struct aac_dev *dev); int aac_fib_setup(struct aac_dev *dev); void aac_fib_map_free(struct aac_dev *dev); diff --git a/drivers/scsi/aacraid/commsup.c b/drivers/scsi/aacraid/commsup.c index a1f90fe84..4cbf54928 100644 --- a/drivers/scsi/aacraid/commsup.c +++ b/drivers/scsi/aacraid/commsup.c @@ -83,13 +83,38 @@ static int fib_map_alloc(struct aac_dev *dev) void aac_fib_map_free(struct aac_dev *dev) { - pci_free_consistent(dev->pdev, - dev->max_fib_size * (dev->scsi_host_ptr->can_queue + AAC_NUM_MGT_FIB), - dev->hw_fib_va, dev->hw_fib_pa); + if (dev->hw_fib_va && dev->max_fib_size) { + pci_free_consistent(dev->pdev, + (dev->max_fib_size * + (dev->scsi_host_ptr->can_queue + AAC_NUM_MGT_FIB)), + dev->hw_fib_va, dev->hw_fib_pa); + } dev->hw_fib_va = NULL; dev->hw_fib_pa = 0; } +void aac_fib_vector_assign(struct aac_dev *dev) +{ + u32 i = 0; + u32 vector = 1; + struct fib *fibptr = NULL; + + for (i = 0, fibptr = &dev->fibs[i]; + i < (dev->scsi_host_ptr->can_queue + AAC_NUM_MGT_FIB); + i++, fibptr++) { + if ((dev->max_msix == 1) || + (i > ((dev->scsi_host_ptr->can_queue + AAC_NUM_MGT_FIB - 1) + - dev->vector_cap))) { + fibptr->vector_no = 0; + } else { + fibptr->vector_no = vector; + vector++; + if (vector == dev->max_msix) + vector = 1; + } + } +} + /** * aac_fib_setup - setup the fibs * @dev: Adapter to set up @@ -151,6 +176,12 @@ int aac_fib_setup(struct aac_dev * dev) hw_fib_pa = hw_fib_pa + dev->max_fib_size + sizeof(struct aac_fib_xporthdr); } + + /* + *Assign vector numbers to fibs + */ + aac_fib_vector_assign(dev); + /* * Add the fib chain to the free list */ diff --git a/drivers/scsi/aacraid/linit.c b/drivers/scsi/aacraid/linit.c index 76eaa38ff..8a8e84548 100644 --- a/drivers/scsi/aacraid/linit.c +++ b/drivers/scsi/aacraid/linit.c @@ -1404,8 +1404,18 @@ static int aac_acquire_resources(struct aac_dev *dev) aac_adapter_enable_int(dev); - if (!dev->sync_mode) + /*max msix may change after EEH + * Re-assign vectors to fibs + */ + aac_fib_vector_assign(dev); + + if (!dev->sync_mode) { + /* After EEH recovery or suspend resume, max_msix count + * may change, therfore updating in init as well. + */ aac_adapter_start(dev); + dev->init->Sa_MSIXVectors = cpu_to_le32(dev->max_msix); + } return 0; error_iounmap: diff --git a/drivers/scsi/aacraid/src.c b/drivers/scsi/aacraid/src.c index 2aa34ea8c..bc0203f3d 100644 --- a/drivers/scsi/aacraid/src.c +++ b/drivers/scsi/aacraid/src.c @@ -156,8 +156,8 @@ static irqreturn_t aac_src_intr_message(int irq, void *dev_id) break; if (dev->msi_enabled && dev->max_msix > 1) atomic_dec(&dev->rrq_outstanding[vector_no]); - aac_intr_normal(dev, handle-1, 0, isFastResponse, NULL); dev->host_rrq[index++] = 0; + aac_intr_normal(dev, handle-1, 0, isFastResponse, NULL); if (index == (vector_no + 1) * dev->vector_cap) index = vector_no * dev->vector_cap; dev->host_rrq_idx[vector_no] = index; @@ -452,36 +452,20 @@ static int aac_src_deliver_message(struct fib *fib) #endif u16 hdr_size = le16_to_cpu(fib->hw_fib_va->header.Size); + u16 vector_no; atomic_inc(&q->numpending); if (dev->msi_enabled && fib->hw_fib_va->header.Command != AifRequest && dev->max_msix > 1) { - u_int16_t vector_no, first_choice = 0xffff; - - vector_no = dev->fibs_pushed_no % dev->max_msix; - do { - vector_no += 1; - if (vector_no == dev->max_msix) - vector_no = 1; - if (atomic_read(&dev->rrq_outstanding[vector_no]) < - dev->vector_cap) - break; - if (0xffff == first_choice) - first_choice = vector_no; - else if (vector_no == first_choice) - break; - } while (1); - if (vector_no == first_choice) - vector_no = 0; - atomic_inc(&dev->rrq_outstanding[vector_no]); - if (dev->fibs_pushed_no == 0xffffffff) - dev->fibs_pushed_no = 0; - else - dev->fibs_pushed_no++; + vector_no = fib->vector_no; fib->hw_fib_va->header.Handle += (vector_no << 16); + } else { + vector_no = 0; } + atomic_inc(&dev->rrq_outstanding[vector_no]); + if (dev->comm_interface == AAC_COMM_MESSAGE_TYPE2) { /* Calculate the amount to the fibsize bits */ fibsize = (hdr_size + 127) / 128 - 1; diff --git a/drivers/scsi/aic7xxx/aic7xxx_osm.c b/drivers/scsi/aic7xxx/aic7xxx_osm.c index b846a4683..fc6a83188 100644 --- a/drivers/scsi/aic7xxx/aic7xxx_osm.c +++ b/drivers/scsi/aic7xxx/aic7xxx_osm.c @@ -1336,6 +1336,7 @@ ahc_platform_set_tags(struct ahc_softc *ahc, struct scsi_device *sdev, case AHC_DEV_Q_TAGGED: scsi_change_queue_depth(sdev, dev->openings + dev->active); + break; default: /* * We allow the OS to queue 2 untagged transactions to diff --git a/drivers/scsi/atari_NCR5380.c b/drivers/scsi/atari_NCR5380.c index e65478651..389825ba5 100644 --- a/drivers/scsi/atari_NCR5380.c +++ b/drivers/scsi/atari_NCR5380.c @@ -862,7 +862,7 @@ static struct scsi_cmnd *dequeue_next_cmd(struct Scsi_Host *instance) struct NCR5380_cmd *ncmd; struct scsi_cmnd *cmd; - if (list_empty(&hostdata->autosense)) { + if (hostdata->sensing || list_empty(&hostdata->autosense)) { list_for_each_entry(ncmd, &hostdata->unissued, list) { cmd = NCR5380_to_scmd(ncmd); dsprintk(NDEBUG_QUEUES, instance, "dequeue: cmd=%p target=%d busy=0x%02x lun=%llu\n", @@ -901,7 +901,7 @@ static void requeue_cmd(struct Scsi_Host *instance, struct scsi_cmnd *cmd) struct NCR5380_hostdata *hostdata = shost_priv(instance); struct NCR5380_cmd *ncmd = scsi_cmd_priv(cmd); - if (hostdata->sensing) { + if (hostdata->sensing == cmd) { scsi_eh_restore_cmnd(cmd, &hostdata->ses); list_add(&ncmd->list, &hostdata->autosense); hostdata->sensing = NULL; @@ -923,7 +923,6 @@ static void NCR5380_main(struct work_struct *work) struct NCR5380_hostdata *hostdata = container_of(work, struct NCR5380_hostdata, main_task); struct Scsi_Host *instance = hostdata->host; - struct scsi_cmnd *cmd; int done; /* @@ -936,8 +935,11 @@ static void NCR5380_main(struct work_struct *work) done = 1; spin_lock_irq(&hostdata->lock); - while (!hostdata->connected && - (cmd = dequeue_next_cmd(instance))) { + while (!hostdata->connected && !hostdata->selecting) { + struct scsi_cmnd *cmd = dequeue_next_cmd(instance); + + if (!cmd) + break; dsprintk(NDEBUG_MAIN, instance, "main: dequeued %p\n", cmd); @@ -960,8 +962,7 @@ static void NCR5380_main(struct work_struct *work) #ifdef SUPPORT_TAGS cmd_get_tag(cmd, cmd->cmnd[0] != REQUEST_SENSE); #endif - cmd = NCR5380_select(instance, cmd); - if (!cmd) { + if (!NCR5380_select(instance, cmd)) { dsprintk(NDEBUG_MAIN, instance, "main: select complete\n"); maybe_release_dma_irq(instance); } else { @@ -1257,6 +1258,11 @@ static struct scsi_cmnd *NCR5380_select(struct Scsi_Host *instance, /* Reselection interrupt */ goto out; } + if (!hostdata->selecting) { + /* Command was aborted */ + NCR5380_write(MODE_REG, MR_BASE); + goto out; + } if (err < 0) { NCR5380_write(MODE_REG, MR_BASE); shost_printk(KERN_ERR, instance, @@ -1838,9 +1844,7 @@ static void NCR5380_information_transfer(struct Scsi_Host *instance) unsigned char msgout = NOP; int sink = 0; int len; -#if defined(REAL_DMA) int transfersize; -#endif unsigned char *data; unsigned char phase, tmp, extended_msg[10], old_phase = 0xff; struct scsi_cmnd *cmd; @@ -1909,6 +1913,7 @@ static void NCR5380_information_transfer(struct Scsi_Host *instance) do_abort(instance); cmd->result = DID_ERROR << 16; complete_cmd(instance, cmd); + hostdata->connected = NULL; return; #endif case PHASE_DATAIN: @@ -1966,7 +1971,6 @@ static void NCR5380_information_transfer(struct Scsi_Host *instance) sink = 1; do_abort(instance); cmd->result = DID_ERROR << 16; - complete_cmd(instance, cmd); /* XXX - need to source or sink data here, as appropriate */ } else { #ifdef REAL_DMA @@ -1983,18 +1987,22 @@ static void NCR5380_information_transfer(struct Scsi_Host *instance) } else #endif /* defined(REAL_DMA) */ { - spin_unlock_irq(&hostdata->lock); - NCR5380_transfer_pio(instance, &phase, - (int *)&cmd->SCp.this_residual, + /* Break up transfer into 3 ms chunks, + * presuming 6 accesses per handshake. + */ + transfersize = min((unsigned long)cmd->SCp.this_residual, + hostdata->accesses_per_ms / 2); + len = transfersize; + NCR5380_transfer_pio(instance, &phase, &len, (unsigned char **)&cmd->SCp.ptr); - spin_lock_irq(&hostdata->lock); + cmd->SCp.this_residual -= transfersize - len; } #if defined(CONFIG_SUN3) && defined(REAL_DMA) /* if we had intended to dma that command clear it */ if (sun3_dma_setup_done == cmd) sun3_dma_setup_done = NULL; #endif - break; + return; case PHASE_MSGIN: len = 1; data = &tmp; @@ -2487,14 +2495,17 @@ static bool list_del_cmd(struct list_head *haystack, * [disconnected -> connected ->]... * [autosense -> connected ->] done * - * If cmd is unissued then just remove it. - * If cmd is disconnected, try to select the target. - * If cmd is connected, try to send an abort message. - * If cmd is waiting for autosense, give it a chance to complete but check - * that it isn't left connected. * If cmd was not found at all then presumably it has already been completed, * in which case return SUCCESS to try to avoid further EH measures. + * * If the command has not completed yet, we must not fail to find it. + * We have no option but to forget the aborted command (even if it still + * lacks sense data). The mid-layer may re-issue a command that is in error + * recovery (see scsi_send_eh_cmnd), but the logic and data structures in + * this driver are such that a command can appear on one queue only. + * + * The lock protects driver data structures, but EH handlers also use it + * to serialize their own execution and prevent their own re-entry. */ static int NCR5380_abort(struct scsi_cmnd *cmd) @@ -2517,6 +2528,7 @@ static int NCR5380_abort(struct scsi_cmnd *cmd) "abort: removed %p from issue queue\n", cmd); cmd->result = DID_ABORT << 16; cmd->scsi_done(cmd); /* No tag or busy flag to worry about */ + goto out; } if (hostdata->selecting == cmd) { @@ -2531,58 +2543,21 @@ static int NCR5380_abort(struct scsi_cmnd *cmd) if (list_del_cmd(&hostdata->disconnected, cmd)) { dsprintk(NDEBUG_ABORT, instance, "abort: removed %p from disconnected list\n", cmd); - cmd->result = DID_ERROR << 16; - if (!hostdata->connected) - NCR5380_select(instance, cmd); - if (hostdata->connected != cmd) { - complete_cmd(instance, cmd); - result = FAILED; - goto out; - } + /* Can't call NCR5380_select() and send ABORT because that + * means releasing the lock. Need a bus reset. + */ + set_host_byte(cmd, DID_ERROR); + complete_cmd(instance, cmd); + result = FAILED; + goto out; } if (hostdata->connected == cmd) { dsprintk(NDEBUG_ABORT, instance, "abort: cmd %p is connected\n", cmd); hostdata->connected = NULL; - if (do_abort(instance)) { - set_host_byte(cmd, DID_ERROR); - complete_cmd(instance, cmd); - result = FAILED; - goto out; - } - set_host_byte(cmd, DID_ABORT); #ifdef REAL_DMA hostdata->dma_len = 0; #endif - if (cmd->cmnd[0] == REQUEST_SENSE) - complete_cmd(instance, cmd); - else { - struct NCR5380_cmd *ncmd = scsi_cmd_priv(cmd); - - /* Perform autosense for this command */ - list_add(&ncmd->list, &hostdata->autosense); - } - } - - if (list_find_cmd(&hostdata->autosense, cmd)) { - dsprintk(NDEBUG_ABORT, instance, - "abort: found %p on sense queue\n", cmd); - spin_unlock_irqrestore(&hostdata->lock, flags); - queue_work(hostdata->work_q, &hostdata->main_task); - msleep(1000); - spin_lock_irqsave(&hostdata->lock, flags); - if (list_del_cmd(&hostdata->autosense, cmd)) { - dsprintk(NDEBUG_ABORT, instance, - "abort: removed %p from sense queue\n", cmd); - set_host_byte(cmd, DID_ABORT); - complete_cmd(instance, cmd); - goto out; - } - } - - if (hostdata->connected == cmd) { - dsprintk(NDEBUG_ABORT, instance, "abort: cmd %p is connected\n", cmd); - hostdata->connected = NULL; if (do_abort(instance)) { set_host_byte(cmd, DID_ERROR); complete_cmd(instance, cmd); @@ -2590,9 +2565,14 @@ static int NCR5380_abort(struct scsi_cmnd *cmd) goto out; } set_host_byte(cmd, DID_ABORT); -#ifdef REAL_DMA - hostdata->dma_len = 0; -#endif + complete_cmd(instance, cmd); + goto out; + } + + if (list_del_cmd(&hostdata->autosense, cmd)) { + dsprintk(NDEBUG_ABORT, instance, + "abort: removed %p from sense queue\n", cmd); + set_host_byte(cmd, DID_ERROR); complete_cmd(instance, cmd); } @@ -2646,7 +2626,16 @@ static int NCR5380_bus_reset(struct scsi_cmnd *cmd) * commands! */ - hostdata->selecting = NULL; + if (list_del_cmd(&hostdata->unissued, cmd)) { + cmd->result = DID_RESET << 16; + cmd->scsi_done(cmd); + } + + if (hostdata->selecting) { + hostdata->selecting->result = DID_RESET << 16; + complete_cmd(instance, hostdata->selecting); + hostdata->selecting = NULL; + } list_for_each_entry(ncmd, &hostdata->disconnected, list) { struct scsi_cmnd *cmd = NCR5380_to_scmd(ncmd); @@ -2654,6 +2643,7 @@ static int NCR5380_bus_reset(struct scsi_cmnd *cmd) set_host_byte(cmd, DID_RESET); cmd->scsi_done(cmd); } + INIT_LIST_HEAD(&hostdata->disconnected); list_for_each_entry(ncmd, &hostdata->autosense, list) { struct scsi_cmnd *cmd = NCR5380_to_scmd(ncmd); @@ -2661,6 +2651,7 @@ static int NCR5380_bus_reset(struct scsi_cmnd *cmd) set_host_byte(cmd, DID_RESET); cmd->scsi_done(cmd); } + INIT_LIST_HEAD(&hostdata->autosense); if (hostdata->connected) { set_host_byte(hostdata->connected, DID_RESET); @@ -2668,12 +2659,6 @@ static int NCR5380_bus_reset(struct scsi_cmnd *cmd) hostdata->connected = NULL; } - if (hostdata->sensing) { - set_host_byte(hostdata->connected, DID_RESET); - complete_cmd(instance, hostdata->sensing); - hostdata->sensing = NULL; - } - #ifdef SUPPORT_TAGS free_all_tags(hostdata); #endif diff --git a/drivers/scsi/be2iscsi/be_main.c b/drivers/scsi/be2iscsi/be_main.c index cb9072a84..069e5c50a 100644 --- a/drivers/scsi/be2iscsi/be_main.c +++ b/drivers/scsi/be2iscsi/be_main.c @@ -4468,6 +4468,7 @@ put_shost: scsi_host_put(phba->shost); free_kset: iscsi_boot_destroy_kset(phba->boot_kset); + phba->boot_kset = NULL; return -ENOMEM; } diff --git a/drivers/scsi/scsi_common.c b/drivers/scsi/scsi_common.c index c12696613..ce79de822 100644 --- a/drivers/scsi/scsi_common.c +++ b/drivers/scsi/scsi_common.c @@ -278,8 +278,16 @@ int scsi_set_sense_information(u8 *buf, int buf_len, u64 info) ucp[3] = 0; put_unaligned_be64(info, &ucp[4]); } else if ((buf[0] & 0x7f) == 0x70) { - buf[0] |= 0x80; - put_unaligned_be64(info, &buf[3]); + /* + * Only set the 'VALID' bit if we can represent the value + * correctly; otherwise just fill out the lower bytes and + * clear the 'VALID' flag. + */ + if (info <= 0xffffffffUL) + buf[0] |= 0x80; + else + buf[0] &= 0x7f; + put_unaligned_be32((u32)info, &buf[3]); } return 0; diff --git a/drivers/scsi/sd.c b/drivers/scsi/sd.c index d749da765..5a5457ac9 100644 --- a/drivers/scsi/sd.c +++ b/drivers/scsi/sd.c @@ -648,7 +648,7 @@ static void sd_config_discard(struct scsi_disk *sdkp, unsigned int mode) */ if (sdkp->lbprz) { q->limits.discard_alignment = 0; - q->limits.discard_granularity = 1; + q->limits.discard_granularity = logical_block_size; } else { q->limits.discard_alignment = sdkp->unmap_alignment * logical_block_size; diff --git a/drivers/scsi/sg.c b/drivers/scsi/sg.c index 5e8206744..ae7d9bdf4 100644 --- a/drivers/scsi/sg.c +++ b/drivers/scsi/sg.c @@ -652,7 +652,8 @@ sg_write(struct file *filp, const char __user *buf, size_t count, loff_t * ppos) else hp->dxfer_direction = (mxsize > 0) ? SG_DXFER_FROM_DEV : SG_DXFER_NONE; hp->dxfer_len = mxsize; - if (hp->dxfer_direction == SG_DXFER_TO_DEV) + if ((hp->dxfer_direction == SG_DXFER_TO_DEV) || + (hp->dxfer_direction == SG_DXFER_TO_FROM_DEV)) hp->dxferp = (char __user *)buf + cmd_size; else hp->dxferp = NULL; diff --git a/drivers/scsi/storvsc_drv.c b/drivers/scsi/storvsc_drv.c index 292c04eec..3ddcabb79 100644 --- a/drivers/scsi/storvsc_drv.c +++ b/drivers/scsi/storvsc_drv.c @@ -914,8 +914,9 @@ static void storvsc_handle_error(struct vmscsi_request *vm_srb, do_work = true; process_err_fn = storvsc_remove_lun; break; - case (SRB_STATUS_ABORTED | SRB_STATUS_AUTOSENSE_VALID): - if ((asc == 0x2a) && (ascq == 0x9)) { + case SRB_STATUS_ABORTED: + if (vm_srb->srb_status & SRB_STATUS_AUTOSENSE_VALID && + (asc == 0x2a) && (ascq == 0x9)) { do_work = true; process_err_fn = storvsc_device_scan; /* diff --git a/drivers/staging/android/ion/ion_test.c b/drivers/staging/android/ion/ion_test.c index b8dcf5a26..58d46893e 100644 --- a/drivers/staging/android/ion/ion_test.c +++ b/drivers/staging/android/ion/ion_test.c @@ -285,8 +285,8 @@ static int __init ion_test_init(void) { ion_test_pdev = platform_device_register_simple("ion-test", -1, NULL, 0); - if (!ion_test_pdev) - return -ENODEV; + if (IS_ERR(ion_test_pdev)) + return PTR_ERR(ion_test_pdev); return platform_driver_probe(&ion_test_platform_driver, ion_test_probe); } diff --git a/drivers/staging/comedi/drivers/ni_mio_common.c b/drivers/staging/comedi/drivers/ni_mio_common.c index 5e8130a7d..0e9f77924 100644 --- a/drivers/staging/comedi/drivers/ni_mio_common.c +++ b/drivers/staging/comedi/drivers/ni_mio_common.c @@ -246,24 +246,24 @@ static void ni_writel(struct comedi_device *dev, uint32_t data, int reg) { if (dev->mmio) writel(data, dev->mmio + reg); - - outl(data, dev->iobase + reg); + else + outl(data, dev->iobase + reg); } static void ni_writew(struct comedi_device *dev, uint16_t data, int reg) { if (dev->mmio) writew(data, dev->mmio + reg); - - outw(data, dev->iobase + reg); + else + outw(data, dev->iobase + reg); } static void ni_writeb(struct comedi_device *dev, uint8_t data, int reg) { if (dev->mmio) writeb(data, dev->mmio + reg); - - outb(data, dev->iobase + reg); + else + outb(data, dev->iobase + reg); } static uint32_t ni_readl(struct comedi_device *dev, int reg) diff --git a/drivers/staging/comedi/drivers/ni_tiocmd.c b/drivers/staging/comedi/drivers/ni_tiocmd.c index 437f723bb..823e47910 100644 --- a/drivers/staging/comedi/drivers/ni_tiocmd.c +++ b/drivers/staging/comedi/drivers/ni_tiocmd.c @@ -92,7 +92,7 @@ static int ni_tio_input_inttrig(struct comedi_device *dev, unsigned long flags; int ret = 0; - if (trig_num != cmd->start_src) + if (trig_num != cmd->start_arg) return -EINVAL; spin_lock_irqsave(&counter->lock, flags); diff --git a/drivers/target/target_core_transport.c b/drivers/target/target_core_transport.c index 867bc6d0a..43d8b42c0 100644 --- a/drivers/target/target_core_transport.c +++ b/drivers/target/target_core_transport.c @@ -2596,8 +2596,6 @@ void target_wait_for_sess_cmds(struct se_session *se_sess) list_for_each_entry_safe(se_cmd, tmp_cmd, &se_sess->sess_wait_list, se_cmd_list) { - list_del_init(&se_cmd->se_cmd_list); - pr_debug("Waiting for se_cmd: %p t_state: %d, fabric state:" " %d\n", se_cmd, se_cmd->t_state, se_cmd->se_tfo->get_cmd_state(se_cmd)); diff --git a/drivers/thermal/thermal_core.c b/drivers/thermal/thermal_core.c index a0a8fd123..d4b54653e 100644 --- a/drivers/thermal/thermal_core.c +++ b/drivers/thermal/thermal_core.c @@ -454,6 +454,10 @@ static void handle_thermal_trip(struct thermal_zone_device *tz, int trip) { enum thermal_trip_type type; + /* Ignore disabled trip points */ + if (test_bit(trip, &tz->trips_disabled)) + return; + tz->ops->get_trip_type(tz, trip, &type); if (type == THERMAL_TRIP_CRITICAL || type == THERMAL_TRIP_HOT) @@ -1800,6 +1804,7 @@ struct thermal_zone_device *thermal_zone_device_register(const char *type, { struct thermal_zone_device *tz; enum thermal_trip_type trip_type; + int trip_temp; int result; int count; int passive = 0; @@ -1871,9 +1876,15 @@ struct thermal_zone_device *thermal_zone_device_register(const char *type, goto unregister; for (count = 0; count < trips; count++) { - tz->ops->get_trip_type(tz, count, &trip_type); + if (tz->ops->get_trip_type(tz, count, &trip_type)) + set_bit(count, &tz->trips_disabled); if (trip_type == THERMAL_TRIP_PASSIVE) passive = 1; + if (tz->ops->get_trip_temp(tz, count, &trip_temp)) + set_bit(count, &tz->trips_disabled); + /* Check for bogus trip points */ + if (trip_temp == 0) + set_bit(count, &tz->trips_disabled); } if (!passive) { diff --git a/drivers/tty/serial/8250/8250_port.c b/drivers/tty/serial/8250/8250_port.c index 83aed6193..b1a6ba3c3 100644 --- a/drivers/tty/serial/8250/8250_port.c +++ b/drivers/tty/serial/8250/8250_port.c @@ -731,22 +731,16 @@ static int size_fifo(struct uart_8250_port *up) */ static unsigned int autoconfig_read_divisor_id(struct uart_8250_port *p) { - unsigned char old_dll, old_dlm, old_lcr; - unsigned int id; + unsigned char old_lcr; + unsigned int id, old_dl; old_lcr = serial_in(p, UART_LCR); serial_out(p, UART_LCR, UART_LCR_CONF_MODE_A); + old_dl = serial_dl_read(p); + serial_dl_write(p, 0); + id = serial_dl_read(p); + serial_dl_write(p, old_dl); - old_dll = serial_in(p, UART_DLL); - old_dlm = serial_in(p, UART_DLM); - - serial_out(p, UART_DLL, 0); - serial_out(p, UART_DLM, 0); - - id = serial_in(p, UART_DLL) | serial_in(p, UART_DLM) << 8; - - serial_out(p, UART_DLL, old_dll); - serial_out(p, UART_DLM, old_dlm); serial_out(p, UART_LCR, old_lcr); return id; diff --git a/drivers/usb/class/cdc-acm.c b/drivers/usb/class/cdc-acm.c index fa4e23930..d37fdcc31 100644 --- a/drivers/usb/class/cdc-acm.c +++ b/drivers/usb/class/cdc-acm.c @@ -1114,6 +1114,9 @@ static int acm_probe(struct usb_interface *intf, if (quirks == NO_UNION_NORMAL) { data_interface = usb_ifnum_to_if(usb_dev, 1); control_interface = usb_ifnum_to_if(usb_dev, 0); + /* we would crash */ + if (!data_interface || !control_interface) + return -ENODEV; goto skip_normal_probe; } diff --git a/drivers/usb/core/driver.c b/drivers/usb/core/driver.c index 56593a9a8..2057d91d8 100644 --- a/drivers/usb/core/driver.c +++ b/drivers/usb/core/driver.c @@ -502,11 +502,15 @@ static int usb_unbind_interface(struct device *dev) int usb_driver_claim_interface(struct usb_driver *driver, struct usb_interface *iface, void *priv) { - struct device *dev = &iface->dev; + struct device *dev; struct usb_device *udev; int retval = 0; int lpm_disable_error; + if (!iface) + return -ENODEV; + + dev = &iface->dev; if (dev->driver) return -EBUSY; diff --git a/drivers/usb/core/hub.c b/drivers/usb/core/hub.c index 51b436918..84f65743f 100644 --- a/drivers/usb/core/hub.c +++ b/drivers/usb/core/hub.c @@ -4292,7 +4292,7 @@ hub_port_init(struct usb_hub *hub, struct usb_device *udev, int port1, { struct usb_device *hdev = hub->hdev; struct usb_hcd *hcd = bus_to_hcd(hdev->bus); - int i, j, retval; + int retries, operations, retval, i; unsigned delay = HUB_SHORT_RESET_TIME; enum usb_device_speed oldspeed = udev->speed; const char *speed; @@ -4394,7 +4394,7 @@ hub_port_init(struct usb_hub *hub, struct usb_device *udev, int port1, * first 8 bytes of the device descriptor to get the ep0 maxpacket * value. */ - for (i = 0; i < GET_DESCRIPTOR_TRIES; (++i, msleep(100))) { + for (retries = 0; retries < GET_DESCRIPTOR_TRIES; (++retries, msleep(100))) { bool did_new_scheme = false; if (use_new_scheme(udev, retry_counter)) { @@ -4421,7 +4421,7 @@ hub_port_init(struct usb_hub *hub, struct usb_device *udev, int port1, * 255 is for WUSB devices, we actually need to use * 512 (WUSB1.0[4.8.1]). */ - for (j = 0; j < 3; ++j) { + for (operations = 0; operations < 3; ++operations) { buf->bMaxPacketSize0 = 0; r = usb_control_msg(udev, usb_rcvaddr0pipe(), USB_REQ_GET_DESCRIPTOR, USB_DIR_IN, @@ -4441,7 +4441,13 @@ hub_port_init(struct usb_hub *hub, struct usb_device *udev, int port1, r = -EPROTO; break; } - if (r == 0) + /* + * Some devices time out if they are powered on + * when already connected. They need a second + * reset. But only on the first attempt, + * lest we get into a time out/reset loop + */ + if (r == 0 || (r == -ETIMEDOUT && retries == 0)) break; } udev->descriptor.bMaxPacketSize0 = @@ -4473,7 +4479,7 @@ hub_port_init(struct usb_hub *hub, struct usb_device *udev, int port1, * authorization will assign the final address. */ if (udev->wusb == 0) { - for (j = 0; j < SET_ADDRESS_TRIES; ++j) { + for (operations = 0; operations < SET_ADDRESS_TRIES; ++operations) { retval = hub_set_address(udev, devnum); if (retval >= 0) break; diff --git a/drivers/usb/misc/iowarrior.c b/drivers/usb/misc/iowarrior.c index c6bfd13f6..1950e87b4 100644 --- a/drivers/usb/misc/iowarrior.c +++ b/drivers/usb/misc/iowarrior.c @@ -787,6 +787,12 @@ static int iowarrior_probe(struct usb_interface *interface, iface_desc = interface->cur_altsetting; dev->product_id = le16_to_cpu(udev->descriptor.idProduct); + if (iface_desc->desc.bNumEndpoints < 1) { + dev_err(&interface->dev, "Invalid number of endpoints\n"); + retval = -EINVAL; + goto error; + } + /* set up the endpoint information */ for (i = 0; i < iface_desc->desc.bNumEndpoints; ++i) { endpoint = &iface_desc->endpoint[i].desc; diff --git a/drivers/usb/serial/cp210x.c b/drivers/usb/serial/cp210x.c index 73a366de5..9bc0e090b 100644 --- a/drivers/usb/serial/cp210x.c +++ b/drivers/usb/serial/cp210x.c @@ -165,6 +165,7 @@ static const struct usb_device_id id_table[] = { { USB_DEVICE(0x18EF, 0xE025) }, /* ELV Marble Sound Board 1 */ { USB_DEVICE(0x1901, 0x0190) }, /* GE B850 CP2105 Recorder interface */ { USB_DEVICE(0x1901, 0x0193) }, /* GE B650 CP2104 PMC interface */ + { USB_DEVICE(0x1901, 0x0194) }, /* GE Healthcare Remote Alarm Box */ { USB_DEVICE(0x19CF, 0x3000) }, /* Parrot NMEA GPS Flight Recorder */ { USB_DEVICE(0x1ADB, 0x0001) }, /* Schweitzer Engineering C662 Cable */ { USB_DEVICE(0x1B1C, 0x1C00) }, /* Corsair USB Dongle */ diff --git a/drivers/usb/serial/cypress_m8.c b/drivers/usb/serial/cypress_m8.c index 01bf53392..244acb129 100644 --- a/drivers/usb/serial/cypress_m8.c +++ b/drivers/usb/serial/cypress_m8.c @@ -447,6 +447,11 @@ static int cypress_generic_port_probe(struct usb_serial_port *port) struct usb_serial *serial = port->serial; struct cypress_private *priv; + if (!port->interrupt_out_urb || !port->interrupt_in_urb) { + dev_err(&port->dev, "required endpoint is missing\n"); + return -ENODEV; + } + priv = kzalloc(sizeof(struct cypress_private), GFP_KERNEL); if (!priv) return -ENOMEM; @@ -606,12 +611,6 @@ static int cypress_open(struct tty_struct *tty, struct usb_serial_port *port) cypress_set_termios(tty, port, &priv->tmp_termios); /* setup the port and start reading from the device */ - if (!port->interrupt_in_urb) { - dev_err(&port->dev, "%s - interrupt_in_urb is empty!\n", - __func__); - return -1; - } - usb_fill_int_urb(port->interrupt_in_urb, serial->dev, usb_rcvintpipe(serial->dev, port->interrupt_in_endpointAddress), port->interrupt_in_urb->transfer_buffer, diff --git a/drivers/usb/serial/digi_acceleport.c b/drivers/usb/serial/digi_acceleport.c index 12b0e6747..3df7b7ec1 100644 --- a/drivers/usb/serial/digi_acceleport.c +++ b/drivers/usb/serial/digi_acceleport.c @@ -1251,8 +1251,27 @@ static int digi_port_init(struct usb_serial_port *port, unsigned port_num) static int digi_startup(struct usb_serial *serial) { + struct device *dev = &serial->interface->dev; struct digi_serial *serial_priv; int ret; + int i; + + /* check whether the device has the expected number of endpoints */ + if (serial->num_port_pointers < serial->type->num_ports + 1) { + dev_err(dev, "OOB endpoints missing\n"); + return -ENODEV; + } + + for (i = 0; i < serial->type->num_ports + 1 ; i++) { + if (!serial->port[i]->read_urb) { + dev_err(dev, "bulk-in endpoint missing\n"); + return -ENODEV; + } + if (!serial->port[i]->write_urb) { + dev_err(dev, "bulk-out endpoint missing\n"); + return -ENODEV; + } + } serial_priv = kzalloc(sizeof(*serial_priv), GFP_KERNEL); if (!serial_priv) diff --git a/drivers/usb/serial/ftdi_sio.c b/drivers/usb/serial/ftdi_sio.c index 8c660ae40..b61f12160 100644 --- a/drivers/usb/serial/ftdi_sio.c +++ b/drivers/usb/serial/ftdi_sio.c @@ -1004,6 +1004,10 @@ static const struct usb_device_id id_table_combined[] = { { USB_DEVICE(FTDI_VID, CHETCO_SEASMART_DISPLAY_PID) }, { USB_DEVICE(FTDI_VID, CHETCO_SEASMART_LITE_PID) }, { USB_DEVICE(FTDI_VID, CHETCO_SEASMART_ANALOG_PID) }, + /* ICP DAS I-756xU devices */ + { USB_DEVICE(ICPDAS_VID, ICPDAS_I7560U_PID) }, + { USB_DEVICE(ICPDAS_VID, ICPDAS_I7561U_PID) }, + { USB_DEVICE(ICPDAS_VID, ICPDAS_I7563U_PID) }, { } /* Terminating entry */ }; diff --git a/drivers/usb/serial/ftdi_sio_ids.h b/drivers/usb/serial/ftdi_sio_ids.h index a84df2513..c5d6c1e73 100644 --- a/drivers/usb/serial/ftdi_sio_ids.h +++ b/drivers/usb/serial/ftdi_sio_ids.h @@ -872,6 +872,14 @@ #define NOVITUS_BONO_E_PID 0x6010 /* + * ICPDAS I-756*U devices + */ +#define ICPDAS_VID 0x1b5c +#define ICPDAS_I7560U_PID 0x0103 +#define ICPDAS_I7561U_PID 0x0104 +#define ICPDAS_I7563U_PID 0x0105 + +/* * RT Systems programming cables for various ham radios */ #define RTSYSTEMS_VID 0x2100 /* Vendor ID */ diff --git a/drivers/usb/serial/mct_u232.c b/drivers/usb/serial/mct_u232.c index fd707d6a1..89726f702 100644 --- a/drivers/usb/serial/mct_u232.c +++ b/drivers/usb/serial/mct_u232.c @@ -376,14 +376,21 @@ static void mct_u232_msr_to_state(struct usb_serial_port *port, static int mct_u232_port_probe(struct usb_serial_port *port) { + struct usb_serial *serial = port->serial; struct mct_u232_private *priv; + /* check first to simplify error handling */ + if (!serial->port[1] || !serial->port[1]->interrupt_in_urb) { + dev_err(&port->dev, "expected endpoint missing\n"); + return -ENODEV; + } + priv = kzalloc(sizeof(*priv), GFP_KERNEL); if (!priv) return -ENOMEM; /* Use second interrupt-in endpoint for reading. */ - priv->read_urb = port->serial->port[1]->interrupt_in_urb; + priv->read_urb = serial->port[1]->interrupt_in_urb; priv->read_urb->context = port; spin_lock_init(&priv->lock); diff --git a/drivers/usb/serial/option.c b/drivers/usb/serial/option.c index 348e19834..c6f497f16 100644 --- a/drivers/usb/serial/option.c +++ b/drivers/usb/serial/option.c @@ -1818,6 +1818,8 @@ static const struct usb_device_id option_ids[] = { { USB_DEVICE_AND_INTERFACE_INFO(0x2001, 0x7d02, 0xff, 0x00, 0x00) }, { USB_DEVICE_AND_INTERFACE_INFO(0x2001, 0x7d03, 0xff, 0x02, 0x01) }, { USB_DEVICE_AND_INTERFACE_INFO(0x2001, 0x7d03, 0xff, 0x00, 0x00) }, + { USB_DEVICE_INTERFACE_CLASS(0x2001, 0x7e19, 0xff), /* D-Link DWM-221 B1 */ + .driver_info = (kernel_ulong_t)&net_intf4_blacklist }, { USB_DEVICE_AND_INTERFACE_INFO(0x07d1, 0x3e01, 0xff, 0xff, 0xff) }, /* D-Link DWM-152/C1 */ { USB_DEVICE_AND_INTERFACE_INFO(0x07d1, 0x3e02, 0xff, 0xff, 0xff) }, /* D-Link DWM-156/C1 */ { USB_DEVICE_INTERFACE_CLASS(0x2020, 0x4000, 0xff) }, /* OLICARD300 - MT6225 */ diff --git a/drivers/usb/storage/uas.c b/drivers/usb/storage/uas.c index 9ff9404f9..c90a7e46c 100644 --- a/drivers/usb/storage/uas.c +++ b/drivers/usb/storage/uas.c @@ -812,7 +812,7 @@ static struct scsi_host_template uas_host_template = { .slave_configure = uas_slave_configure, .eh_abort_handler = uas_eh_abort_handler, .eh_bus_reset_handler = uas_eh_bus_reset_handler, - .can_queue = 65536, /* Is there a limit on the _host_ ? */ + .can_queue = MAX_CMNDS, .this_id = -1, .sg_tablesize = SG_NONE, .skip_settle_delay = 1, diff --git a/drivers/watchdog/rc32434_wdt.c b/drivers/watchdog/rc32434_wdt.c index 71e78ef4b..3a75f3b53 100644 --- a/drivers/watchdog/rc32434_wdt.c +++ b/drivers/watchdog/rc32434_wdt.c @@ -237,7 +237,7 @@ static long rc32434_wdt_ioctl(struct file *file, unsigned int cmd, return -EINVAL; /* Fall through */ case WDIOC_GETTIMEOUT: - return copy_to_user(argp, &timeout, sizeof(int)); + return copy_to_user(argp, &timeout, sizeof(int)) ? -EFAULT : 0; default: return -ENOTTY; } diff --git a/fs/Kconfig b/fs/Kconfig index dc04844aa..5c1e6f459 100644 --- a/fs/Kconfig +++ b/fs/Kconfig @@ -233,6 +233,7 @@ source "fs/pstore/Kconfig" source "fs/sysv/Kconfig" source "fs/ufs/Kconfig" source "fs/exofs/Kconfig" +source "fs/aufs/Kconfig" endif # MISC_FILESYSTEMS diff --git a/fs/Makefile b/fs/Makefile index b8da1c2c3..0c61756a2 100644 --- a/fs/Makefile +++ b/fs/Makefile @@ -127,3 +127,4 @@ obj-y += exofs/ # Multiple modules obj-$(CONFIG_CEPH_FS) += ceph/ obj-$(CONFIG_PSTORE) += pstore/ obj-$(CONFIG_EFIVAR_FS) += efivarfs/ +obj-$(CONFIG_AUFS_FS) += aufs/ diff --git a/fs/aufs/Kconfig b/fs/aufs/Kconfig new file mode 100644 index 000000000..63560ceda --- /dev/null +++ b/fs/aufs/Kconfig @@ -0,0 +1,185 @@ +config AUFS_FS + tristate "Aufs (Advanced multi layered unification filesystem) support" + help + Aufs is a stackable unification filesystem such as Unionfs, + which unifies several directories and provides a merged single + directory. + In the early days, aufs was entirely re-designed and + re-implemented Unionfs Version 1.x series. Introducing many + original ideas, approaches and improvements, it becomes totally + different from Unionfs while keeping the basic features. + +if AUFS_FS +choice + prompt "Maximum number of branches" + default AUFS_BRANCH_MAX_127 + help + Specifies the maximum number of branches (or member directories) + in a single aufs. The larger value consumes more system + resources and has a minor impact to performance. +config AUFS_BRANCH_MAX_127 + bool "127" + help + Specifies the maximum number of branches (or member directories) + in a single aufs. The larger value consumes more system + resources and has a minor impact to performance. +config AUFS_BRANCH_MAX_511 + bool "511" + help + Specifies the maximum number of branches (or member directories) + in a single aufs. The larger value consumes more system + resources and has a minor impact to performance. +config AUFS_BRANCH_MAX_1023 + bool "1023" + help + Specifies the maximum number of branches (or member directories) + in a single aufs. The larger value consumes more system + resources and has a minor impact to performance. +config AUFS_BRANCH_MAX_32767 + bool "32767" + help + Specifies the maximum number of branches (or member directories) + in a single aufs. The larger value consumes more system + resources and has a minor impact to performance. +endchoice + +config AUFS_SBILIST + bool + depends on AUFS_MAGIC_SYSRQ || PROC_FS + default y + help + Automatic configuration for internal use. + When aufs supports Magic SysRq or /proc, enabled automatically. + +config AUFS_HNOTIFY + bool "Detect direct branch access (bypassing aufs)" + help + If you want to modify files on branches directly, eg. bypassing aufs, + and want aufs to detect the changes of them fully, then enable this + option and use 'udba=notify' mount option. + Currently there is only one available configuration, "fsnotify". + It will have a negative impact to the performance. + See detail in aufs.5. + +choice + prompt "method" if AUFS_HNOTIFY + default AUFS_HFSNOTIFY +config AUFS_HFSNOTIFY + bool "fsnotify" + select FSNOTIFY +endchoice + +config AUFS_EXPORT + bool "NFS-exportable aufs" + depends on EXPORTFS + help + If you want to export your mounted aufs via NFS, then enable this + option. There are several requirements for this configuration. + See detail in aufs.5. + +config AUFS_INO_T_64 + bool + depends on AUFS_EXPORT + depends on 64BIT && !(ALPHA || S390) + default y + help + Automatic configuration for internal use. + /* typedef unsigned long/int __kernel_ino_t */ + /* alpha and s390x are int */ + +config AUFS_XATTR + bool "support for XATTR/EA (including Security Labels)" + help + If your branch fs supports XATTR/EA and you want to make them + available in aufs too, then enable this opsion and specify the + branch attributes for EA. + See detail in aufs.5. + +config AUFS_FHSM + bool "File-based Hierarchical Storage Management" + help + Hierarchical Storage Management (or HSM) is a well-known feature + in the storage world. Aufs provides this feature as file-based. + with multiple branches. + These multiple branches are prioritized, ie. the topmost one + should be the fastest drive and be used heavily. + +config AUFS_RDU + bool "Readdir in userspace" + help + Aufs has two methods to provide a merged view for a directory, + by a user-space library and by kernel-space natively. The latter + is always enabled but sometimes large and slow. + If you enable this option, install the library in aufs2-util + package, and set some environment variables for your readdir(3), + then the work will be handled in user-space which generally + shows better performance in most cases. + See detail in aufs.5. + +config AUFS_SHWH + bool "Show whiteouts" + help + If you want to make the whiteouts in aufs visible, then enable + this option and specify 'shwh' mount option. Although it may + sounds like philosophy or something, but in technically it + simply shows the name of whiteout with keeping its behaviour. + +config AUFS_BR_RAMFS + bool "Ramfs (initramfs/rootfs) as an aufs branch" + help + If you want to use ramfs as an aufs branch fs, then enable this + option. Generally tmpfs is recommended. + Aufs prohibited them to be a branch fs by default, because + initramfs becomes unusable after switch_root or something + generally. If you sets initramfs as an aufs branch and boot your + system by switch_root, you will meet a problem easily since the + files in initramfs may be inaccessible. + Unless you are going to use ramfs as an aufs branch fs without + switch_root or something, leave it N. + +config AUFS_BR_FUSE + bool "Fuse fs as an aufs branch" + depends on FUSE_FS + select AUFS_POLL + help + If you want to use fuse-based userspace filesystem as an aufs + branch fs, then enable this option. + It implements the internal poll(2) operation which is + implemented by fuse only (curretnly). + +config AUFS_POLL + bool + help + Automatic configuration for internal use. + +config AUFS_BR_HFSPLUS + bool "Hfsplus as an aufs branch" + depends on HFSPLUS_FS + default y + help + If you want to use hfsplus fs as an aufs branch fs, then enable + this option. This option introduces a small overhead at + copying-up a file on hfsplus. + +config AUFS_BDEV_LOOP + bool + depends on BLK_DEV_LOOP + default y + help + Automatic configuration for internal use. + Convert =[ym] into =y. + +config AUFS_DEBUG + bool "Debug aufs" + help + Enable this to compile aufs internal debug code. + It will have a negative impact to the performance. + +config AUFS_MAGIC_SYSRQ + bool + depends on AUFS_DEBUG && MAGIC_SYSRQ + default y + help + Automatic configuration for internal use. + When aufs supports Magic SysRq, enabled automatically. +endif diff --git a/fs/aufs/Makefile b/fs/aufs/Makefile new file mode 100644 index 000000000..c7efb62b5 --- /dev/null +++ b/fs/aufs/Makefile @@ -0,0 +1,36 @@ + +include ${srctree}/${src}/magic.mk + +# cf. include/linux/kernel.h +# enable pr_debug +ccflags-y += -DDEBUG +# sparse requires the full pathname +ccflags-y += -include ${srctree}/include/uapi/linux/aufs_type.h + +obj-$(CONFIG_AUFS_FS) += aufs.o +aufs-y := module.o sbinfo.o super.o branch.o xino.o sysaufs.o opts.o \ + wkq.o vfsub.o dcsub.o \ + cpup.o whout.o wbr_policy.o \ + dinfo.o dentry.o \ + dynop.o \ + finfo.o file.o f_op.o \ + dir.o vdir.o \ + iinfo.o inode.o i_op.o i_op_add.o i_op_del.o i_op_ren.o \ + mvdown.o ioctl.o + +# all are boolean +aufs-$(CONFIG_PROC_FS) += procfs.o plink.o +aufs-$(CONFIG_SYSFS) += sysfs.o +aufs-$(CONFIG_DEBUG_FS) += dbgaufs.o +aufs-$(CONFIG_AUFS_BDEV_LOOP) += loop.o +aufs-$(CONFIG_AUFS_HNOTIFY) += hnotify.o +aufs-$(CONFIG_AUFS_HFSNOTIFY) += hfsnotify.o +aufs-$(CONFIG_AUFS_EXPORT) += export.o +aufs-$(CONFIG_AUFS_XATTR) += xattr.o +aufs-$(CONFIG_FS_POSIX_ACL) += posix_acl.o +aufs-$(CONFIG_AUFS_FHSM) += fhsm.o +aufs-$(CONFIG_AUFS_POLL) += poll.o +aufs-$(CONFIG_AUFS_RDU) += rdu.o +aufs-$(CONFIG_AUFS_BR_HFSPLUS) += hfsplus.o +aufs-$(CONFIG_AUFS_DEBUG) += debug.o +aufs-$(CONFIG_AUFS_MAGIC_SYSRQ) += sysrq.o diff --git a/fs/aufs/aufs.h b/fs/aufs/aufs.h new file mode 100644 index 000000000..49f43b433 --- /dev/null +++ b/fs/aufs/aufs.h @@ -0,0 +1,46 @@ +/* + * Copyright (C) 2005-2016 Junjiro R. Okajima + */ + +/* + * all header files + */ + +#ifndef __AUFS_H__ +#define __AUFS_H__ + +#ifdef __KERNEL__ + +#define AuStub(type, name, body, ...) \ + static inline type name(__VA_ARGS__) { body; } + +#define AuStubVoid(name, ...) \ + AuStub(void, name, , __VA_ARGS__) +#define AuStubInt0(name, ...) \ + AuStub(int, name, return 0, __VA_ARGS__) + +#include "debug.h" + +#include "branch.h" +#include "cpup.h" +#include "dcsub.h" +#include "dbgaufs.h" +#include "dentry.h" +#include "dir.h" +#include "dynop.h" +#include "file.h" +#include "fstype.h" +#include "inode.h" +#include "loop.h" +#include "module.h" +#include "opts.h" +#include "rwsem.h" +#include "spl.h" +#include "super.h" +#include "sysaufs.h" +#include "vfsub.h" +#include "whout.h" +#include "wkq.h" + +#endif /* __KERNEL__ */ +#endif /* __AUFS_H__ */ diff --git a/fs/aufs/branch.c b/fs/aufs/branch.c new file mode 100644 index 000000000..496bdaa99 --- /dev/null +++ b/fs/aufs/branch.c @@ -0,0 +1,1394 @@ +/* + * Copyright (C) 2005-2016 Junjiro R. Okajima + */ + +/* + * branch management + */ + +#include <linux/compat.h> +#include <linux/statfs.h> +#include "aufs.h" + +/* + * free a single branch + */ +static void au_br_do_free(struct au_branch *br) +{ + int i; + struct au_wbr *wbr; + struct au_dykey **key; + + au_hnotify_fin_br(br); + + if (br->br_xino.xi_file) + fput(br->br_xino.xi_file); + mutex_destroy(&br->br_xino.xi_nondir_mtx); + + AuDebugOn(atomic_read(&br->br_count)); + + wbr = br->br_wbr; + if (wbr) { + for (i = 0; i < AuBrWh_Last; i++) + dput(wbr->wbr_wh[i]); + AuDebugOn(atomic_read(&wbr->wbr_wh_running)); + AuRwDestroy(&wbr->wbr_wh_rwsem); + } + + if (br->br_fhsm) { + au_br_fhsm_fin(br->br_fhsm); + kfree(br->br_fhsm); + } + + key = br->br_dykey; + for (i = 0; i < AuBrDynOp; i++, key++) + if (*key) + au_dy_put(*key); + else + break; + + /* recursive lock, s_umount of branch's */ + lockdep_off(); + path_put(&br->br_path); + lockdep_on(); + kfree(wbr); + kfree(br); +} + +/* + * frees all branches + */ +void au_br_free(struct au_sbinfo *sbinfo) +{ + aufs_bindex_t bmax; + struct au_branch **br; + + AuRwMustWriteLock(&sbinfo->si_rwsem); + + bmax = sbinfo->si_bend + 1; + br = sbinfo->si_branch; + while (bmax--) + au_br_do_free(*br++); +} + +/* + * find the index of a branch which is specified by @br_id. + */ +int au_br_index(struct super_block *sb, aufs_bindex_t br_id) +{ + aufs_bindex_t bindex, bend; + + bend = au_sbend(sb); + for (bindex = 0; bindex <= bend; bindex++) + if (au_sbr_id(sb, bindex) == br_id) + return bindex; + return -1; +} + +/* ---------------------------------------------------------------------- */ + +/* + * add a branch + */ + +static int test_overlap(struct super_block *sb, struct dentry *h_adding, + struct dentry *h_root) +{ + if (unlikely(h_adding == h_root + || au_test_loopback_overlap(sb, h_adding))) + return 1; + if (h_adding->d_sb != h_root->d_sb) + return 0; + return au_test_subdir(h_adding, h_root) + || au_test_subdir(h_root, h_adding); +} + +/* + * returns a newly allocated branch. @new_nbranch is a number of branches + * after adding a branch. + */ +static struct au_branch *au_br_alloc(struct super_block *sb, int new_nbranch, + int perm) +{ + struct au_branch *add_branch; + struct dentry *root; + struct inode *inode; + int err; + + err = -ENOMEM; + root = sb->s_root; + add_branch = kzalloc(sizeof(*add_branch), GFP_NOFS); + if (unlikely(!add_branch)) + goto out; + + err = au_hnotify_init_br(add_branch, perm); + if (unlikely(err)) + goto out_br; + + if (au_br_writable(perm)) { + /* may be freed separately at changing the branch permission */ + add_branch->br_wbr = kzalloc(sizeof(*add_branch->br_wbr), + GFP_NOFS); + if (unlikely(!add_branch->br_wbr)) + goto out_hnotify; + } + + if (au_br_fhsm(perm)) { + err = au_fhsm_br_alloc(add_branch); + if (unlikely(err)) + goto out_wbr; + } + + err = au_sbr_realloc(au_sbi(sb), new_nbranch); + if (!err) + err = au_di_realloc(au_di(root), new_nbranch); + if (!err) { + inode = d_inode(root); + err = au_ii_realloc(au_ii(inode), new_nbranch); + } + if (!err) + return add_branch; /* success */ + +out_wbr: + kfree(add_branch->br_wbr); +out_hnotify: + au_hnotify_fin_br(add_branch); +out_br: + kfree(add_branch); +out: + return ERR_PTR(err); +} + +/* + * test if the branch permission is legal or not. + */ +static int test_br(struct inode *inode, int brperm, char *path) +{ + int err; + + err = (au_br_writable(brperm) && IS_RDONLY(inode)); + if (!err) + goto out; + + err = -EINVAL; + pr_err("write permission for readonly mount or inode, %s\n", path); + +out: + return err; +} + +/* + * returns: + * 0: success, the caller will add it + * plus: success, it is already unified, the caller should ignore it + * minus: error + */ +static int test_add(struct super_block *sb, struct au_opt_add *add, int remount) +{ + int err; + aufs_bindex_t bend, bindex; + struct dentry *root, *h_dentry; + struct inode *inode, *h_inode; + + root = sb->s_root; + bend = au_sbend(sb); + if (unlikely(bend >= 0 + && au_find_dbindex(root, add->path.dentry) >= 0)) { + err = 1; + if (!remount) { + err = -EINVAL; + pr_err("%s duplicated\n", add->pathname); + } + goto out; + } + + err = -ENOSPC; /* -E2BIG; */ + if (unlikely(AUFS_BRANCH_MAX <= add->bindex + || AUFS_BRANCH_MAX - 1 <= bend)) { + pr_err("number of branches exceeded %s\n", add->pathname); + goto out; + } + + err = -EDOM; + if (unlikely(add->bindex < 0 || bend + 1 < add->bindex)) { + pr_err("bad index %d\n", add->bindex); + goto out; + } + + inode = d_inode(add->path.dentry); + err = -ENOENT; + if (unlikely(!inode->i_nlink)) { + pr_err("no existence %s\n", add->pathname); + goto out; + } + + err = -EINVAL; + if (unlikely(inode->i_sb == sb)) { + pr_err("%s must be outside\n", add->pathname); + goto out; + } + + if (unlikely(au_test_fs_unsuppoted(inode->i_sb))) { + pr_err("unsupported filesystem, %s (%s)\n", + add->pathname, au_sbtype(inode->i_sb)); + goto out; + } + + if (unlikely(inode->i_sb->s_stack_depth)) { + pr_err("already stacked, %s (%s)\n", + add->pathname, au_sbtype(inode->i_sb)); + goto out; + } + + err = test_br(d_inode(add->path.dentry), add->perm, add->pathname); + if (unlikely(err)) + goto out; + + if (bend < 0) + return 0; /* success */ + + err = -EINVAL; + for (bindex = 0; bindex <= bend; bindex++) + if (unlikely(test_overlap(sb, add->path.dentry, + au_h_dptr(root, bindex)))) { + pr_err("%s is overlapped\n", add->pathname); + goto out; + } + + err = 0; + if (au_opt_test(au_mntflags(sb), WARN_PERM)) { + h_dentry = au_h_dptr(root, 0); + h_inode = d_inode(h_dentry); + if ((h_inode->i_mode & S_IALLUGO) != (inode->i_mode & S_IALLUGO) + || !uid_eq(h_inode->i_uid, inode->i_uid) + || !gid_eq(h_inode->i_gid, inode->i_gid)) + pr_warn("uid/gid/perm %s %u/%u/0%o, %u/%u/0%o\n", + add->pathname, + i_uid_read(inode), i_gid_read(inode), + (inode->i_mode & S_IALLUGO), + i_uid_read(h_inode), i_gid_read(h_inode), + (h_inode->i_mode & S_IALLUGO)); + } + +out: + return err; +} + +/* + * initialize or clean the whiteouts for an adding branch + */ +static int au_br_init_wh(struct super_block *sb, struct au_branch *br, + int new_perm) +{ + int err, old_perm; + aufs_bindex_t bindex; + struct inode *h_inode; + struct au_wbr *wbr; + struct au_hinode *hdir; + struct dentry *h_dentry; + + err = vfsub_mnt_want_write(au_br_mnt(br)); + if (unlikely(err)) + goto out; + + wbr = br->br_wbr; + old_perm = br->br_perm; + br->br_perm = new_perm; + hdir = NULL; + h_inode = NULL; + bindex = au_br_index(sb, br->br_id); + if (0 <= bindex) { + hdir = au_hi(d_inode(sb->s_root), bindex); + au_hn_imtx_lock_nested(hdir, AuLsc_I_PARENT); + } else { + h_dentry = au_br_dentry(br); + h_inode = d_inode(h_dentry); + inode_lock_nested(h_inode, AuLsc_I_PARENT); + } + if (!wbr) + err = au_wh_init(br, sb); + else { + wbr_wh_write_lock(wbr); + err = au_wh_init(br, sb); + wbr_wh_write_unlock(wbr); + } + if (hdir) + au_hn_imtx_unlock(hdir); + else + inode_unlock(h_inode); + vfsub_mnt_drop_write(au_br_mnt(br)); + br->br_perm = old_perm; + + if (!err && wbr && !au_br_writable(new_perm)) { + kfree(wbr); + br->br_wbr = NULL; + } + +out: + return err; +} + +static int au_wbr_init(struct au_branch *br, struct super_block *sb, + int perm) +{ + int err; + struct kstatfs kst; + struct au_wbr *wbr; + + wbr = br->br_wbr; + au_rw_init(&wbr->wbr_wh_rwsem); + atomic_set(&wbr->wbr_wh_running, 0); + + /* + * a limit for rmdir/rename a dir + * cf. AUFS_MAX_NAMELEN in include/uapi/linux/aufs_type.h + */ + err = vfs_statfs(&br->br_path, &kst); + if (unlikely(err)) + goto out; + err = -EINVAL; + if (kst.f_namelen >= NAME_MAX) + err = au_br_init_wh(sb, br, perm); + else + pr_err("%pd(%s), unsupported namelen %ld\n", + au_br_dentry(br), + au_sbtype(au_br_dentry(br)->d_sb), kst.f_namelen); + +out: + return err; +} + +/* initialize a new branch */ +static int au_br_init(struct au_branch *br, struct super_block *sb, + struct au_opt_add *add) +{ + int err; + struct inode *h_inode; + + err = 0; + mutex_init(&br->br_xino.xi_nondir_mtx); + br->br_perm = add->perm; + br->br_path = add->path; /* set first, path_get() later */ + spin_lock_init(&br->br_dykey_lock); + atomic_set(&br->br_count, 0); + atomic_set(&br->br_xino_running, 0); + br->br_id = au_new_br_id(sb); + AuDebugOn(br->br_id < 0); + + if (au_br_writable(add->perm)) { + err = au_wbr_init(br, sb, add->perm); + if (unlikely(err)) + goto out_err; + } + + if (au_opt_test(au_mntflags(sb), XINO)) { + h_inode = d_inode(add->path.dentry); + err = au_xino_br(sb, br, h_inode->i_ino, + au_sbr(sb, 0)->br_xino.xi_file, /*do_test*/1); + if (unlikely(err)) { + AuDebugOn(br->br_xino.xi_file); + goto out_err; + } + } + + sysaufs_br_init(br); + path_get(&br->br_path); + goto out; /* success */ + +out_err: + memset(&br->br_path, 0, sizeof(br->br_path)); +out: + return err; +} + +static void au_br_do_add_brp(struct au_sbinfo *sbinfo, aufs_bindex_t bindex, + struct au_branch *br, aufs_bindex_t bend, + aufs_bindex_t amount) +{ + struct au_branch **brp; + + AuRwMustWriteLock(&sbinfo->si_rwsem); + + brp = sbinfo->si_branch + bindex; + memmove(brp + 1, brp, sizeof(*brp) * amount); + *brp = br; + sbinfo->si_bend++; + if (unlikely(bend < 0)) + sbinfo->si_bend = 0; +} + +static void au_br_do_add_hdp(struct au_dinfo *dinfo, aufs_bindex_t bindex, + aufs_bindex_t bend, aufs_bindex_t amount) +{ + struct au_hdentry *hdp; + + AuRwMustWriteLock(&dinfo->di_rwsem); + + hdp = dinfo->di_hdentry + bindex; + memmove(hdp + 1, hdp, sizeof(*hdp) * amount); + au_h_dentry_init(hdp); + dinfo->di_bend++; + if (unlikely(bend < 0)) + dinfo->di_bstart = 0; +} + +static void au_br_do_add_hip(struct au_iinfo *iinfo, aufs_bindex_t bindex, + aufs_bindex_t bend, aufs_bindex_t amount) +{ + struct au_hinode *hip; + + AuRwMustWriteLock(&iinfo->ii_rwsem); + + hip = iinfo->ii_hinode + bindex; + memmove(hip + 1, hip, sizeof(*hip) * amount); + hip->hi_inode = NULL; + au_hn_init(hip); + iinfo->ii_bend++; + if (unlikely(bend < 0)) + iinfo->ii_bstart = 0; +} + +static void au_br_do_add(struct super_block *sb, struct au_branch *br, + aufs_bindex_t bindex) +{ + struct dentry *root, *h_dentry; + struct inode *root_inode, *h_inode; + aufs_bindex_t bend, amount; + + root = sb->s_root; + root_inode = d_inode(root); + bend = au_sbend(sb); + amount = bend + 1 - bindex; + h_dentry = au_br_dentry(br); + au_sbilist_lock(); + au_br_do_add_brp(au_sbi(sb), bindex, br, bend, amount); + au_br_do_add_hdp(au_di(root), bindex, bend, amount); + au_br_do_add_hip(au_ii(root_inode), bindex, bend, amount); + au_set_h_dptr(root, bindex, dget(h_dentry)); + h_inode = d_inode(h_dentry); + au_set_h_iptr(root_inode, bindex, au_igrab(h_inode), /*flags*/0); + au_sbilist_unlock(); +} + +int au_br_add(struct super_block *sb, struct au_opt_add *add, int remount) +{ + int err; + aufs_bindex_t bend, add_bindex; + struct dentry *root, *h_dentry; + struct inode *root_inode; + struct au_branch *add_branch; + + root = sb->s_root; + root_inode = d_inode(root); + IMustLock(root_inode); + err = test_add(sb, add, remount); + if (unlikely(err < 0)) + goto out; + if (err) { + err = 0; + goto out; /* success */ + } + + bend = au_sbend(sb); + add_branch = au_br_alloc(sb, bend + 2, add->perm); + err = PTR_ERR(add_branch); + if (IS_ERR(add_branch)) + goto out; + + err = au_br_init(add_branch, sb, add); + if (unlikely(err)) { + au_br_do_free(add_branch); + goto out; + } + + add_bindex = add->bindex; + if (!remount) + au_br_do_add(sb, add_branch, add_bindex); + else { + sysaufs_brs_del(sb, add_bindex); + au_br_do_add(sb, add_branch, add_bindex); + sysaufs_brs_add(sb, add_bindex); + } + + h_dentry = add->path.dentry; + if (!add_bindex) { + au_cpup_attr_all(root_inode, /*force*/1); + sb->s_maxbytes = h_dentry->d_sb->s_maxbytes; + } else + au_add_nlink(root_inode, d_inode(h_dentry)); + + /* + * this test/set prevents aufs from handling unnecesary notify events + * of xino files, in case of re-adding a writable branch which was + * once detached from aufs. + */ + if (au_xino_brid(sb) < 0 + && au_br_writable(add_branch->br_perm) + && !au_test_fs_bad_xino(h_dentry->d_sb) + && add_branch->br_xino.xi_file + && add_branch->br_xino.xi_file->f_path.dentry->d_parent == h_dentry) + au_xino_brid_set(sb, add_branch->br_id); + +out: + return err; +} + +/* ---------------------------------------------------------------------- */ + +static unsigned long long au_farray_cb(struct super_block *sb, void *a, + unsigned long long max __maybe_unused, + void *arg) +{ + unsigned long long n; + struct file **p, *f; + struct au_sphlhead *files; + struct au_finfo *finfo; + + n = 0; + p = a; + files = &au_sbi(sb)->si_files; + spin_lock(&files->spin); + hlist_for_each_entry(finfo, &files->head, fi_hlist) { + f = finfo->fi_file; + if (file_count(f) + && !special_file(file_inode(f)->i_mode)) { + get_file(f); + *p++ = f; + n++; + AuDebugOn(n > max); + } + } + spin_unlock(&files->spin); + + return n; +} + +static struct file **au_farray_alloc(struct super_block *sb, + unsigned long long *max) +{ + *max = atomic_long_read(&au_sbi(sb)->si_nfiles); + return au_array_alloc(max, au_farray_cb, sb, /*arg*/NULL); +} + +static void au_farray_free(struct file **a, unsigned long long max) +{ + unsigned long long ull; + + for (ull = 0; ull < max; ull++) + if (a[ull]) + fput(a[ull]); + kvfree(a); +} + +/* ---------------------------------------------------------------------- */ + +/* + * delete a branch + */ + +/* to show the line number, do not make it inlined function */ +#define AuVerbose(do_info, fmt, ...) do { \ + if (do_info) \ + pr_info(fmt, ##__VA_ARGS__); \ +} while (0) + +static int au_test_ibusy(struct inode *inode, aufs_bindex_t bstart, + aufs_bindex_t bend) +{ + return (inode && !S_ISDIR(inode->i_mode)) || bstart == bend; +} + +static int au_test_dbusy(struct dentry *dentry, aufs_bindex_t bstart, + aufs_bindex_t bend) +{ + return au_test_ibusy(d_inode(dentry), bstart, bend); +} + +/* + * test if the branch is deletable or not. + */ +static int test_dentry_busy(struct dentry *root, aufs_bindex_t bindex, + unsigned int sigen, const unsigned int verbose) +{ + int err, i, j, ndentry; + aufs_bindex_t bstart, bend; + struct au_dcsub_pages dpages; + struct au_dpage *dpage; + struct dentry *d; + + err = au_dpages_init(&dpages, GFP_NOFS); + if (unlikely(err)) + goto out; + err = au_dcsub_pages(&dpages, root, NULL, NULL); + if (unlikely(err)) + goto out_dpages; + + for (i = 0; !err && i < dpages.ndpage; i++) { + dpage = dpages.dpages + i; + ndentry = dpage->ndentry; + for (j = 0; !err && j < ndentry; j++) { + d = dpage->dentries[j]; + AuDebugOn(au_dcount(d) <= 0); + if (!au_digen_test(d, sigen)) { + di_read_lock_child(d, AuLock_IR); + if (unlikely(au_dbrange_test(d))) { + di_read_unlock(d, AuLock_IR); + continue; + } + } else { + di_write_lock_child(d); + if (unlikely(au_dbrange_test(d))) { + di_write_unlock(d); + continue; + } + err = au_reval_dpath(d, sigen); + if (!err) + di_downgrade_lock(d, AuLock_IR); + else { + di_write_unlock(d); + break; + } + } + + /* AuDbgDentry(d); */ + bstart = au_dbstart(d); + bend = au_dbend(d); + if (bstart <= bindex + && bindex <= bend + && au_h_dptr(d, bindex) + && au_test_dbusy(d, bstart, bend)) { + err = -EBUSY; + AuVerbose(verbose, "busy %pd\n", d); + AuDbgDentry(d); + } + di_read_unlock(d, AuLock_IR); + } + } + +out_dpages: + au_dpages_free(&dpages); +out: + return err; +} + +static int test_inode_busy(struct super_block *sb, aufs_bindex_t bindex, + unsigned int sigen, const unsigned int verbose) +{ + int err; + unsigned long long max, ull; + struct inode *i, **array; + aufs_bindex_t bstart, bend; + + array = au_iarray_alloc(sb, &max); + err = PTR_ERR(array); + if (IS_ERR(array)) + goto out; + + err = 0; + AuDbg("b%d\n", bindex); + for (ull = 0; !err && ull < max; ull++) { + i = array[ull]; + if (unlikely(!i)) + break; + if (i->i_ino == AUFS_ROOT_INO) + continue; + + /* AuDbgInode(i); */ + if (au_iigen(i, NULL) == sigen) + ii_read_lock_child(i); + else { + ii_write_lock_child(i); + err = au_refresh_hinode_self(i); + au_iigen_dec(i); + if (!err) + ii_downgrade_lock(i); + else { + ii_write_unlock(i); + break; + } + } + + bstart = au_ibstart(i); + bend = au_ibend(i); + if (bstart <= bindex + && bindex <= bend + && au_h_iptr(i, bindex) + && au_test_ibusy(i, bstart, bend)) { + err = -EBUSY; + AuVerbose(verbose, "busy i%lu\n", i->i_ino); + AuDbgInode(i); + } + ii_read_unlock(i); + } + au_iarray_free(array, max); + +out: + return err; +} + +static int test_children_busy(struct dentry *root, aufs_bindex_t bindex, + const unsigned int verbose) +{ + int err; + unsigned int sigen; + + sigen = au_sigen(root->d_sb); + DiMustNoWaiters(root); + IiMustNoWaiters(d_inode(root)); + di_write_unlock(root); + err = test_dentry_busy(root, bindex, sigen, verbose); + if (!err) + err = test_inode_busy(root->d_sb, bindex, sigen, verbose); + di_write_lock_child(root); /* aufs_write_lock() calls ..._child() */ + + return err; +} + +static int test_dir_busy(struct file *file, aufs_bindex_t br_id, + struct file **to_free, int *idx) +{ + int err; + unsigned char matched, root; + aufs_bindex_t bindex, bend; + struct au_fidir *fidir; + struct au_hfile *hfile; + + err = 0; + root = IS_ROOT(file->f_path.dentry); + if (root) { + get_file(file); + to_free[*idx] = file; + (*idx)++; + goto out; + } + + matched = 0; + fidir = au_fi(file)->fi_hdir; + AuDebugOn(!fidir); + bend = au_fbend_dir(file); + for (bindex = au_fbstart(file); bindex <= bend; bindex++) { + hfile = fidir->fd_hfile + bindex; + if (!hfile->hf_file) + continue; + + if (hfile->hf_br->br_id == br_id) { + matched = 1; + break; + } + } + if (matched) + err = -EBUSY; + +out: + return err; +} + +static int test_file_busy(struct super_block *sb, aufs_bindex_t br_id, + struct file **to_free, int opened) +{ + int err, idx; + unsigned long long ull, max; + aufs_bindex_t bstart; + struct file *file, **array; + struct dentry *root; + struct au_hfile *hfile; + + array = au_farray_alloc(sb, &max); + err = PTR_ERR(array); + if (IS_ERR(array)) + goto out; + + err = 0; + idx = 0; + root = sb->s_root; + di_write_unlock(root); + for (ull = 0; ull < max; ull++) { + file = array[ull]; + if (unlikely(!file)) + break; + + /* AuDbg("%pD\n", file); */ + fi_read_lock(file); + bstart = au_fbstart(file); + if (!d_is_dir(file->f_path.dentry)) { + hfile = &au_fi(file)->fi_htop; + if (hfile->hf_br->br_id == br_id) + err = -EBUSY; + } else + err = test_dir_busy(file, br_id, to_free, &idx); + fi_read_unlock(file); + if (unlikely(err)) + break; + } + di_write_lock_child(root); + au_farray_free(array, max); + AuDebugOn(idx > opened); + +out: + return err; +} + +static void br_del_file(struct file **to_free, unsigned long long opened, + aufs_bindex_t br_id) +{ + unsigned long long ull; + aufs_bindex_t bindex, bstart, bend, bfound; + struct file *file; + struct au_fidir *fidir; + struct au_hfile *hfile; + + for (ull = 0; ull < opened; ull++) { + file = to_free[ull]; + if (unlikely(!file)) + break; + + /* AuDbg("%pD\n", file); */ + AuDebugOn(!d_is_dir(file->f_path.dentry)); + bfound = -1; + fidir = au_fi(file)->fi_hdir; + AuDebugOn(!fidir); + fi_write_lock(file); + bstart = au_fbstart(file); + bend = au_fbend_dir(file); + for (bindex = bstart; bindex <= bend; bindex++) { + hfile = fidir->fd_hfile + bindex; + if (!hfile->hf_file) + continue; + + if (hfile->hf_br->br_id == br_id) { + bfound = bindex; + break; + } + } + AuDebugOn(bfound < 0); + au_set_h_fptr(file, bfound, NULL); + if (bfound == bstart) { + for (bstart++; bstart <= bend; bstart++) + if (au_hf_dir(file, bstart)) { + au_set_fbstart(file, bstart); + break; + } + } + fi_write_unlock(file); + } +} + +static void au_br_do_del_brp(struct au_sbinfo *sbinfo, + const aufs_bindex_t bindex, + const aufs_bindex_t bend) +{ + struct au_branch **brp, **p; + + AuRwMustWriteLock(&sbinfo->si_rwsem); + + brp = sbinfo->si_branch + bindex; + if (bindex < bend) + memmove(brp, brp + 1, sizeof(*brp) * (bend - bindex)); + sbinfo->si_branch[0 + bend] = NULL; + sbinfo->si_bend--; + + p = krealloc(sbinfo->si_branch, sizeof(*p) * bend, AuGFP_SBILIST); + if (p) + sbinfo->si_branch = p; + /* harmless error */ +} + +static void au_br_do_del_hdp(struct au_dinfo *dinfo, const aufs_bindex_t bindex, + const aufs_bindex_t bend) +{ + struct au_hdentry *hdp, *p; + + AuRwMustWriteLock(&dinfo->di_rwsem); + + hdp = dinfo->di_hdentry; + if (bindex < bend) + memmove(hdp + bindex, hdp + bindex + 1, + sizeof(*hdp) * (bend - bindex)); + hdp[0 + bend].hd_dentry = NULL; + dinfo->di_bend--; + + p = krealloc(hdp, sizeof(*p) * bend, AuGFP_SBILIST); + if (p) + dinfo->di_hdentry = p; + /* harmless error */ +} + +static void au_br_do_del_hip(struct au_iinfo *iinfo, const aufs_bindex_t bindex, + const aufs_bindex_t bend) +{ + struct au_hinode *hip, *p; + + AuRwMustWriteLock(&iinfo->ii_rwsem); + + hip = iinfo->ii_hinode + bindex; + if (bindex < bend) + memmove(hip, hip + 1, sizeof(*hip) * (bend - bindex)); + iinfo->ii_hinode[0 + bend].hi_inode = NULL; + au_hn_init(iinfo->ii_hinode + bend); + iinfo->ii_bend--; + + p = krealloc(iinfo->ii_hinode, sizeof(*p) * bend, AuGFP_SBILIST); + if (p) + iinfo->ii_hinode = p; + /* harmless error */ +} + +static void au_br_do_del(struct super_block *sb, aufs_bindex_t bindex, + struct au_branch *br) +{ + aufs_bindex_t bend; + struct au_sbinfo *sbinfo; + struct dentry *root, *h_root; + struct inode *inode, *h_inode; + struct au_hinode *hinode; + + SiMustWriteLock(sb); + + root = sb->s_root; + inode = d_inode(root); + sbinfo = au_sbi(sb); + bend = sbinfo->si_bend; + + h_root = au_h_dptr(root, bindex); + hinode = au_hi(inode, bindex); + h_inode = au_igrab(hinode->hi_inode); + au_hiput(hinode); + + au_sbilist_lock(); + au_br_do_del_brp(sbinfo, bindex, bend); + au_br_do_del_hdp(au_di(root), bindex, bend); + au_br_do_del_hip(au_ii(inode), bindex, bend); + au_sbilist_unlock(); + + dput(h_root); + iput(h_inode); + au_br_do_free(br); +} + +static unsigned long long empty_cb(struct super_block *sb, void *array, + unsigned long long max, void *arg) +{ + return max; +} + +int au_br_del(struct super_block *sb, struct au_opt_del *del, int remount) +{ + int err, rerr, i; + unsigned long long opened; + unsigned int mnt_flags; + aufs_bindex_t bindex, bend, br_id; + unsigned char do_wh, verbose; + struct au_branch *br; + struct au_wbr *wbr; + struct dentry *root; + struct file **to_free; + + err = 0; + opened = 0; + to_free = NULL; + root = sb->s_root; + bindex = au_find_dbindex(root, del->h_path.dentry); + if (bindex < 0) { + if (remount) + goto out; /* success */ + err = -ENOENT; + pr_err("%s no such branch\n", del->pathname); + goto out; + } + AuDbg("bindex b%d\n", bindex); + + err = -EBUSY; + mnt_flags = au_mntflags(sb); + verbose = !!au_opt_test(mnt_flags, VERBOSE); + bend = au_sbend(sb); + if (unlikely(!bend)) { + AuVerbose(verbose, "no more branches left\n"); + goto out; + } + br = au_sbr(sb, bindex); + AuDebugOn(!path_equal(&br->br_path, &del->h_path)); + + br_id = br->br_id; + opened = atomic_read(&br->br_count); + if (unlikely(opened)) { + to_free = au_array_alloc(&opened, empty_cb, sb, NULL); + err = PTR_ERR(to_free); + if (IS_ERR(to_free)) + goto out; + + err = test_file_busy(sb, br_id, to_free, opened); + if (unlikely(err)) { + AuVerbose(verbose, "%llu file(s) opened\n", opened); + goto out; + } + } + + wbr = br->br_wbr; + do_wh = wbr && (wbr->wbr_whbase || wbr->wbr_plink || wbr->wbr_orph); + if (do_wh) { + /* instead of WbrWhMustWriteLock(wbr) */ + SiMustWriteLock(sb); + for (i = 0; i < AuBrWh_Last; i++) { + dput(wbr->wbr_wh[i]); + wbr->wbr_wh[i] = NULL; + } + } + + err = test_children_busy(root, bindex, verbose); + if (unlikely(err)) { + if (do_wh) + goto out_wh; + goto out; + } + + err = 0; + if (to_free) { + /* + * now we confirmed the branch is deletable. + * let's free the remaining opened dirs on the branch. + */ + di_write_unlock(root); + br_del_file(to_free, opened, br_id); + di_write_lock_child(root); + } + + if (!remount) + au_br_do_del(sb, bindex, br); + else { + sysaufs_brs_del(sb, bindex); + au_br_do_del(sb, bindex, br); + sysaufs_brs_add(sb, bindex); + } + + if (!bindex) { + au_cpup_attr_all(d_inode(root), /*force*/1); + sb->s_maxbytes = au_sbr_sb(sb, 0)->s_maxbytes; + } else + au_sub_nlink(d_inode(root), d_inode(del->h_path.dentry)); + if (au_opt_test(mnt_flags, PLINK)) + au_plink_half_refresh(sb, br_id); + + if (au_xino_brid(sb) == br_id) + au_xino_brid_set(sb, -1); + goto out; /* success */ + +out_wh: + /* revert */ + rerr = au_br_init_wh(sb, br, br->br_perm); + if (rerr) + pr_warn("failed re-creating base whiteout, %s. (%d)\n", + del->pathname, rerr); +out: + if (to_free) + au_farray_free(to_free, opened); + return err; +} + +/* ---------------------------------------------------------------------- */ + +static int au_ibusy(struct super_block *sb, struct aufs_ibusy __user *arg) +{ + int err; + aufs_bindex_t bstart, bend; + struct aufs_ibusy ibusy; + struct inode *inode, *h_inode; + + err = -EPERM; + if (unlikely(!capable(CAP_SYS_ADMIN))) + goto out; + + err = copy_from_user(&ibusy, arg, sizeof(ibusy)); + if (!err) + err = !access_ok(VERIFY_WRITE, &arg->h_ino, sizeof(arg->h_ino)); + if (unlikely(err)) { + err = -EFAULT; + AuTraceErr(err); + goto out; + } + + err = -EINVAL; + si_read_lock(sb, AuLock_FLUSH); + if (unlikely(ibusy.bindex < 0 || ibusy.bindex > au_sbend(sb))) + goto out_unlock; + + err = 0; + ibusy.h_ino = 0; /* invalid */ + inode = ilookup(sb, ibusy.ino); + if (!inode + || inode->i_ino == AUFS_ROOT_INO + || is_bad_inode(inode)) + goto out_unlock; + + ii_read_lock_child(inode); + bstart = au_ibstart(inode); + bend = au_ibend(inode); + if (bstart <= ibusy.bindex && ibusy.bindex <= bend) { + h_inode = au_h_iptr(inode, ibusy.bindex); + if (h_inode && au_test_ibusy(inode, bstart, bend)) + ibusy.h_ino = h_inode->i_ino; + } + ii_read_unlock(inode); + iput(inode); + +out_unlock: + si_read_unlock(sb); + if (!err) { + err = __put_user(ibusy.h_ino, &arg->h_ino); + if (unlikely(err)) { + err = -EFAULT; + AuTraceErr(err); + } + } +out: + return err; +} + +long au_ibusy_ioctl(struct file *file, unsigned long arg) +{ + return au_ibusy(file->f_path.dentry->d_sb, (void __user *)arg); +} + +#ifdef CONFIG_COMPAT +long au_ibusy_compat_ioctl(struct file *file, unsigned long arg) +{ + return au_ibusy(file->f_path.dentry->d_sb, compat_ptr(arg)); +} +#endif + +/* ---------------------------------------------------------------------- */ + +/* + * change a branch permission + */ + +static void au_warn_ima(void) +{ +#ifdef CONFIG_IMA + /* since it doesn't support mark_files_ro() */ + AuWarn1("RW -> RO makes IMA to produce wrong message\n"); +#endif +} + +static int do_need_sigen_inc(int a, int b) +{ + return au_br_whable(a) && !au_br_whable(b); +} + +static int need_sigen_inc(int old, int new) +{ + return do_need_sigen_inc(old, new) + || do_need_sigen_inc(new, old); +} + +static int au_br_mod_files_ro(struct super_block *sb, aufs_bindex_t bindex) +{ + int err, do_warn; + unsigned int mnt_flags; + unsigned long long ull, max; + aufs_bindex_t br_id; + unsigned char verbose, writer; + struct file *file, *hf, **array; + struct au_hfile *hfile; + + mnt_flags = au_mntflags(sb); + verbose = !!au_opt_test(mnt_flags, VERBOSE); + + array = au_farray_alloc(sb, &max); + err = PTR_ERR(array); + if (IS_ERR(array)) + goto out; + + do_warn = 0; + br_id = au_sbr_id(sb, bindex); + for (ull = 0; ull < max; ull++) { + file = array[ull]; + if (unlikely(!file)) + break; + + /* AuDbg("%pD\n", file); */ + fi_read_lock(file); + if (unlikely(au_test_mmapped(file))) { + err = -EBUSY; + AuVerbose(verbose, "mmapped %pD\n", file); + AuDbgFile(file); + FiMustNoWaiters(file); + fi_read_unlock(file); + goto out_array; + } + + hfile = &au_fi(file)->fi_htop; + hf = hfile->hf_file; + if (!d_is_reg(file->f_path.dentry) + || !(file->f_mode & FMODE_WRITE) + || hfile->hf_br->br_id != br_id + || !(hf->f_mode & FMODE_WRITE)) + array[ull] = NULL; + else { + do_warn = 1; + get_file(file); + } + + FiMustNoWaiters(file); + fi_read_unlock(file); + fput(file); + } + + err = 0; + if (do_warn) + au_warn_ima(); + + for (ull = 0; ull < max; ull++) { + file = array[ull]; + if (!file) + continue; + + /* todo: already flushed? */ + /* + * fs/super.c:mark_files_ro() is gone, but aufs keeps its + * approach which resets f_mode and calls mnt_drop_write() and + * file_release_write() for each file, because the branch + * attribute in aufs world is totally different from the native + * fs rw/ro mode. + */ + /* fi_read_lock(file); */ + hfile = &au_fi(file)->fi_htop; + hf = hfile->hf_file; + /* fi_read_unlock(file); */ + spin_lock(&hf->f_lock); + writer = !!(hf->f_mode & FMODE_WRITER); + hf->f_mode &= ~(FMODE_WRITE | FMODE_WRITER); + spin_unlock(&hf->f_lock); + if (writer) { + put_write_access(file_inode(hf)); + __mnt_drop_write(hf->f_path.mnt); + } + } + +out_array: + au_farray_free(array, max); +out: + AuTraceErr(err); + return err; +} + +int au_br_mod(struct super_block *sb, struct au_opt_mod *mod, int remount, + int *do_refresh) +{ + int err, rerr; + aufs_bindex_t bindex; + struct dentry *root; + struct au_branch *br; + struct au_br_fhsm *bf; + + root = sb->s_root; + bindex = au_find_dbindex(root, mod->h_root); + if (bindex < 0) { + if (remount) + return 0; /* success */ + err = -ENOENT; + pr_err("%s no such branch\n", mod->path); + goto out; + } + AuDbg("bindex b%d\n", bindex); + + err = test_br(d_inode(mod->h_root), mod->perm, mod->path); + if (unlikely(err)) + goto out; + + br = au_sbr(sb, bindex); + AuDebugOn(mod->h_root != au_br_dentry(br)); + if (br->br_perm == mod->perm) + return 0; /* success */ + + /* pre-allocate for non-fhsm --> fhsm */ + bf = NULL; + if (!au_br_fhsm(br->br_perm) && au_br_fhsm(mod->perm)) { + err = au_fhsm_br_alloc(br); + if (unlikely(err)) + goto out; + bf = br->br_fhsm; + br->br_fhsm = NULL; + } + + if (au_br_writable(br->br_perm)) { + /* remove whiteout base */ + err = au_br_init_wh(sb, br, mod->perm); + if (unlikely(err)) + goto out_bf; + + if (!au_br_writable(mod->perm)) { + /* rw --> ro, file might be mmapped */ + DiMustNoWaiters(root); + IiMustNoWaiters(d_inode(root)); + di_write_unlock(root); + err = au_br_mod_files_ro(sb, bindex); + /* aufs_write_lock() calls ..._child() */ + di_write_lock_child(root); + + if (unlikely(err)) { + rerr = -ENOMEM; + br->br_wbr = kzalloc(sizeof(*br->br_wbr), + GFP_NOFS); + if (br->br_wbr) + rerr = au_wbr_init(br, sb, br->br_perm); + if (unlikely(rerr)) { + AuIOErr("nested error %d (%d)\n", + rerr, err); + br->br_perm = mod->perm; + } + } + } + } else if (au_br_writable(mod->perm)) { + /* ro --> rw */ + err = -ENOMEM; + br->br_wbr = kzalloc(sizeof(*br->br_wbr), GFP_NOFS); + if (br->br_wbr) { + err = au_wbr_init(br, sb, mod->perm); + if (unlikely(err)) { + kfree(br->br_wbr); + br->br_wbr = NULL; + } + } + } + if (unlikely(err)) + goto out_bf; + + if (au_br_fhsm(br->br_perm)) { + if (!au_br_fhsm(mod->perm)) { + /* fhsm --> non-fhsm */ + au_br_fhsm_fin(br->br_fhsm); + kfree(br->br_fhsm); + br->br_fhsm = NULL; + } + } else if (au_br_fhsm(mod->perm)) + /* non-fhsm --> fhsm */ + br->br_fhsm = bf; + + *do_refresh |= need_sigen_inc(br->br_perm, mod->perm); + br->br_perm = mod->perm; + goto out; /* success */ + +out_bf: + kfree(bf); +out: + AuTraceErr(err); + return err; +} + +/* ---------------------------------------------------------------------- */ + +int au_br_stfs(struct au_branch *br, struct aufs_stfs *stfs) +{ + int err; + struct kstatfs kstfs; + + err = vfs_statfs(&br->br_path, &kstfs); + if (!err) { + stfs->f_blocks = kstfs.f_blocks; + stfs->f_bavail = kstfs.f_bavail; + stfs->f_files = kstfs.f_files; + stfs->f_ffree = kstfs.f_ffree; + } + + return err; +} diff --git a/fs/aufs/branch.h b/fs/aufs/branch.h new file mode 100644 index 000000000..4c52ae166 --- /dev/null +++ b/fs/aufs/branch.h @@ -0,0 +1,266 @@ +/* + * Copyright (C) 2005-2016 Junjiro R. Okajima + */ + +/* + * branch filesystems and xino for them + */ + +#ifndef __AUFS_BRANCH_H__ +#define __AUFS_BRANCH_H__ + +#ifdef __KERNEL__ + +#include <linux/mount.h> +#include "dynop.h" +#include "rwsem.h" +#include "super.h" + +/* ---------------------------------------------------------------------- */ + +/* a xino file */ +struct au_xino_file { + struct file *xi_file; + struct mutex xi_nondir_mtx; + + /* todo: make xino files an array to support huge inode number */ + +#ifdef CONFIG_DEBUG_FS + struct dentry *xi_dbgaufs; +#endif +}; + +/* File-based Hierarchical Storage Management */ +struct au_br_fhsm { +#ifdef CONFIG_AUFS_FHSM + struct mutex bf_lock; + unsigned long bf_jiffy; + struct aufs_stfs bf_stfs; + int bf_readable; +#endif +}; + +/* members for writable branch only */ +enum {AuBrWh_BASE, AuBrWh_PLINK, AuBrWh_ORPH, AuBrWh_Last}; +struct au_wbr { + struct au_rwsem wbr_wh_rwsem; + struct dentry *wbr_wh[AuBrWh_Last]; + atomic_t wbr_wh_running; +#define wbr_whbase wbr_wh[AuBrWh_BASE] /* whiteout base */ +#define wbr_plink wbr_wh[AuBrWh_PLINK] /* pseudo-link dir */ +#define wbr_orph wbr_wh[AuBrWh_ORPH] /* dir for orphans */ + + /* mfs mode */ + unsigned long long wbr_bytes; +}; + +/* ext2 has 3 types of operations at least, ext3 has 4 */ +#define AuBrDynOp (AuDyLast * 4) + +#ifdef CONFIG_AUFS_HFSNOTIFY +/* support for asynchronous destruction */ +struct au_br_hfsnotify { + struct fsnotify_group *hfsn_group; +}; +#endif + +/* sysfs entries */ +struct au_brsysfs { + char name[16]; + struct attribute attr; +}; + +enum { + AuBrSysfs_BR, + AuBrSysfs_BRID, + AuBrSysfs_Last +}; + +/* protected by superblock rwsem */ +struct au_branch { + struct au_xino_file br_xino; + + aufs_bindex_t br_id; + + int br_perm; + struct path br_path; + spinlock_t br_dykey_lock; + struct au_dykey *br_dykey[AuBrDynOp]; + atomic_t br_count; + + struct au_wbr *br_wbr; + struct au_br_fhsm *br_fhsm; + + /* xino truncation */ + atomic_t br_xino_running; + +#ifdef CONFIG_AUFS_HFSNOTIFY + struct au_br_hfsnotify *br_hfsn; +#endif + +#ifdef CONFIG_SYSFS + /* entries under sysfs per mount-point */ + struct au_brsysfs br_sysfs[AuBrSysfs_Last]; +#endif +}; + +/* ---------------------------------------------------------------------- */ + +static inline struct vfsmount *au_br_mnt(struct au_branch *br) +{ + return br->br_path.mnt; +} + +static inline struct dentry *au_br_dentry(struct au_branch *br) +{ + return br->br_path.dentry; +} + +static inline struct super_block *au_br_sb(struct au_branch *br) +{ + return au_br_mnt(br)->mnt_sb; +} + +static inline int au_br_rdonly(struct au_branch *br) +{ + return ((au_br_sb(br)->s_flags & MS_RDONLY) + || !au_br_writable(br->br_perm)) + ? -EROFS : 0; +} + +static inline int au_br_hnotifyable(int brperm __maybe_unused) +{ +#ifdef CONFIG_AUFS_HNOTIFY + return !(brperm & AuBrPerm_RR); +#else + return 0; +#endif +} + +static inline int au_br_test_oflag(int oflag, struct au_branch *br) +{ + int err, exec_flag; + + err = 0; + exec_flag = oflag & __FMODE_EXEC; + if (unlikely(exec_flag && (au_br_mnt(br)->mnt_flags & MNT_NOEXEC))) + err = -EACCES; + + return err; +} + +/* ---------------------------------------------------------------------- */ + +/* branch.c */ +struct au_sbinfo; +void au_br_free(struct au_sbinfo *sinfo); +int au_br_index(struct super_block *sb, aufs_bindex_t br_id); +struct au_opt_add; +int au_br_add(struct super_block *sb, struct au_opt_add *add, int remount); +struct au_opt_del; +int au_br_del(struct super_block *sb, struct au_opt_del *del, int remount); +long au_ibusy_ioctl(struct file *file, unsigned long arg); +#ifdef CONFIG_COMPAT +long au_ibusy_compat_ioctl(struct file *file, unsigned long arg); +#endif +struct au_opt_mod; +int au_br_mod(struct super_block *sb, struct au_opt_mod *mod, int remount, + int *do_refresh); +struct aufs_stfs; +int au_br_stfs(struct au_branch *br, struct aufs_stfs *stfs); + +/* xino.c */ +static const loff_t au_loff_max = LLONG_MAX; + +int au_xib_trunc(struct super_block *sb); +ssize_t xino_fread(vfs_readf_t func, struct file *file, void *buf, size_t size, + loff_t *pos); +ssize_t xino_fwrite(vfs_writef_t func, struct file *file, void *buf, + size_t size, loff_t *pos); +struct file *au_xino_create2(struct file *base_file, struct file *copy_src); +struct file *au_xino_create(struct super_block *sb, char *fname, int silent); +ino_t au_xino_new_ino(struct super_block *sb); +void au_xino_delete_inode(struct inode *inode, const int unlinked); +int au_xino_write(struct super_block *sb, aufs_bindex_t bindex, ino_t h_ino, + ino_t ino); +int au_xino_read(struct super_block *sb, aufs_bindex_t bindex, ino_t h_ino, + ino_t *ino); +int au_xino_br(struct super_block *sb, struct au_branch *br, ino_t hino, + struct file *base_file, int do_test); +int au_xino_trunc(struct super_block *sb, aufs_bindex_t bindex); + +struct au_opt_xino; +int au_xino_set(struct super_block *sb, struct au_opt_xino *xino, int remount); +void au_xino_clr(struct super_block *sb); +struct file *au_xino_def(struct super_block *sb); +int au_xino_path(struct seq_file *seq, struct file *file); + +/* ---------------------------------------------------------------------- */ + +/* Superblock to branch */ +static inline +aufs_bindex_t au_sbr_id(struct super_block *sb, aufs_bindex_t bindex) +{ + return au_sbr(sb, bindex)->br_id; +} + +static inline +struct vfsmount *au_sbr_mnt(struct super_block *sb, aufs_bindex_t bindex) +{ + return au_br_mnt(au_sbr(sb, bindex)); +} + +static inline +struct super_block *au_sbr_sb(struct super_block *sb, aufs_bindex_t bindex) +{ + return au_br_sb(au_sbr(sb, bindex)); +} + +static inline void au_sbr_put(struct super_block *sb, aufs_bindex_t bindex) +{ + atomic_dec(&au_sbr(sb, bindex)->br_count); +} + +static inline int au_sbr_perm(struct super_block *sb, aufs_bindex_t bindex) +{ + return au_sbr(sb, bindex)->br_perm; +} + +static inline int au_sbr_whable(struct super_block *sb, aufs_bindex_t bindex) +{ + return au_br_whable(au_sbr_perm(sb, bindex)); +} + +/* ---------------------------------------------------------------------- */ + +/* + * wbr_wh_read_lock, wbr_wh_write_lock + * wbr_wh_read_unlock, wbr_wh_write_unlock, wbr_wh_downgrade_lock + */ +AuSimpleRwsemFuncs(wbr_wh, struct au_wbr *wbr, &wbr->wbr_wh_rwsem); + +#define WbrWhMustNoWaiters(wbr) AuRwMustNoWaiters(&wbr->wbr_wh_rwsem) +#define WbrWhMustAnyLock(wbr) AuRwMustAnyLock(&wbr->wbr_wh_rwsem) +#define WbrWhMustWriteLock(wbr) AuRwMustWriteLock(&wbr->wbr_wh_rwsem) + +/* ---------------------------------------------------------------------- */ + +#ifdef CONFIG_AUFS_FHSM +static inline void au_br_fhsm_init(struct au_br_fhsm *brfhsm) +{ + mutex_init(&brfhsm->bf_lock); + brfhsm->bf_jiffy = 0; + brfhsm->bf_readable = 0; +} + +static inline void au_br_fhsm_fin(struct au_br_fhsm *brfhsm) +{ + mutex_destroy(&brfhsm->bf_lock); +} +#else +AuStubVoid(au_br_fhsm_init, struct au_br_fhsm *brfhsm) +AuStubVoid(au_br_fhsm_fin, struct au_br_fhsm *brfhsm) +#endif + +#endif /* __KERNEL__ */ +#endif /* __AUFS_BRANCH_H__ */ diff --git a/fs/aufs/cpup.c b/fs/aufs/cpup.c new file mode 100644 index 000000000..9ff0963ad --- /dev/null +++ b/fs/aufs/cpup.c @@ -0,0 +1,1366 @@ +/* + * Copyright (C) 2005-2016 Junjiro R. Okajima + */ + +/* + * copy-up functions, see wbr_policy.c for copy-down + */ + +#include <linux/fs_stack.h> +#include <linux/mm.h> +#include <linux/task_work.h> +#include "aufs.h" + +void au_cpup_attr_flags(struct inode *dst, unsigned int iflags) +{ + const unsigned int mask = S_DEAD | S_SWAPFILE | S_PRIVATE + | S_NOATIME | S_NOCMTIME | S_AUTOMOUNT; + + BUILD_BUG_ON(sizeof(iflags) != sizeof(dst->i_flags)); + + dst->i_flags |= iflags & ~mask; + if (au_test_fs_notime(dst->i_sb)) + dst->i_flags |= S_NOATIME | S_NOCMTIME; +} + +void au_cpup_attr_timesizes(struct inode *inode) +{ + struct inode *h_inode; + + h_inode = au_h_iptr(inode, au_ibstart(inode)); + fsstack_copy_attr_times(inode, h_inode); + fsstack_copy_inode_size(inode, h_inode); +} + +void au_cpup_attr_nlink(struct inode *inode, int force) +{ + struct inode *h_inode; + struct super_block *sb; + aufs_bindex_t bindex, bend; + + sb = inode->i_sb; + bindex = au_ibstart(inode); + h_inode = au_h_iptr(inode, bindex); + if (!force + && !S_ISDIR(h_inode->i_mode) + && au_opt_test(au_mntflags(sb), PLINK) + && au_plink_test(inode)) + return; + + /* + * 0 can happen in revalidating. + * h_inode->i_mutex may not be held here, but it is harmless since once + * i_nlink reaches 0, it will never become positive except O_TMPFILE + * case. + * todo: O_TMPFILE+linkat(AT_SYMLINK_FOLLOW) bypassing aufs may cause + * the incorrect link count. + */ + set_nlink(inode, h_inode->i_nlink); + + /* + * fewer nlink makes find(1) noisy, but larger nlink doesn't. + * it may includes whplink directory. + */ + if (S_ISDIR(h_inode->i_mode)) { + bend = au_ibend(inode); + for (bindex++; bindex <= bend; bindex++) { + h_inode = au_h_iptr(inode, bindex); + if (h_inode) + au_add_nlink(inode, h_inode); + } + } +} + +void au_cpup_attr_changeable(struct inode *inode) +{ + struct inode *h_inode; + + h_inode = au_h_iptr(inode, au_ibstart(inode)); + inode->i_mode = h_inode->i_mode; + inode->i_uid = h_inode->i_uid; + inode->i_gid = h_inode->i_gid; + au_cpup_attr_timesizes(inode); + au_cpup_attr_flags(inode, h_inode->i_flags); +} + +void au_cpup_igen(struct inode *inode, struct inode *h_inode) +{ + struct au_iinfo *iinfo = au_ii(inode); + + IiMustWriteLock(inode); + + iinfo->ii_higen = h_inode->i_generation; + iinfo->ii_hsb1 = h_inode->i_sb; +} + +void au_cpup_attr_all(struct inode *inode, int force) +{ + struct inode *h_inode; + + h_inode = au_h_iptr(inode, au_ibstart(inode)); + au_cpup_attr_changeable(inode); + if (inode->i_nlink > 0) + au_cpup_attr_nlink(inode, force); + inode->i_rdev = h_inode->i_rdev; + inode->i_blkbits = h_inode->i_blkbits; + au_cpup_igen(inode, h_inode); +} + +/* ---------------------------------------------------------------------- */ + +/* Note: dt_dentry and dt_h_dentry are not dget/dput-ed */ + +/* keep the timestamps of the parent dir when cpup */ +void au_dtime_store(struct au_dtime *dt, struct dentry *dentry, + struct path *h_path) +{ + struct inode *h_inode; + + dt->dt_dentry = dentry; + dt->dt_h_path = *h_path; + h_inode = d_inode(h_path->dentry); + dt->dt_atime = h_inode->i_atime; + dt->dt_mtime = h_inode->i_mtime; + /* smp_mb(); */ +} + +void au_dtime_revert(struct au_dtime *dt) +{ + struct iattr attr; + int err; + + attr.ia_atime = dt->dt_atime; + attr.ia_mtime = dt->dt_mtime; + attr.ia_valid = ATTR_FORCE | ATTR_MTIME | ATTR_MTIME_SET + | ATTR_ATIME | ATTR_ATIME_SET; + + /* no delegation since this is a directory */ + err = vfsub_notify_change(&dt->dt_h_path, &attr, /*delegated*/NULL); + if (unlikely(err)) + pr_warn("restoring timestamps failed(%d). ignored\n", err); +} + +/* ---------------------------------------------------------------------- */ + +/* internal use only */ +struct au_cpup_reg_attr { + int valid; + struct kstat st; + unsigned int iflags; /* inode->i_flags */ +}; + +static noinline_for_stack +int cpup_iattr(struct dentry *dst, aufs_bindex_t bindex, struct dentry *h_src, + struct au_cpup_reg_attr *h_src_attr) +{ + int err, sbits, icex; + unsigned int mnt_flags; + unsigned char verbose; + struct iattr ia; + struct path h_path; + struct inode *h_isrc, *h_idst; + struct kstat *h_st; + struct au_branch *br; + + h_path.dentry = au_h_dptr(dst, bindex); + h_idst = d_inode(h_path.dentry); + br = au_sbr(dst->d_sb, bindex); + h_path.mnt = au_br_mnt(br); + h_isrc = d_inode(h_src); + ia.ia_valid = ATTR_FORCE | ATTR_UID | ATTR_GID + | ATTR_ATIME | ATTR_MTIME + | ATTR_ATIME_SET | ATTR_MTIME_SET; + if (h_src_attr && h_src_attr->valid) { + h_st = &h_src_attr->st; + ia.ia_uid = h_st->uid; + ia.ia_gid = h_st->gid; + ia.ia_atime = h_st->atime; + ia.ia_mtime = h_st->mtime; + if (h_idst->i_mode != h_st->mode + && !S_ISLNK(h_idst->i_mode)) { + ia.ia_valid |= ATTR_MODE; + ia.ia_mode = h_st->mode; + } + sbits = !!(h_st->mode & (S_ISUID | S_ISGID)); + au_cpup_attr_flags(h_idst, h_src_attr->iflags); + } else { + ia.ia_uid = h_isrc->i_uid; + ia.ia_gid = h_isrc->i_gid; + ia.ia_atime = h_isrc->i_atime; + ia.ia_mtime = h_isrc->i_mtime; + if (h_idst->i_mode != h_isrc->i_mode + && !S_ISLNK(h_idst->i_mode)) { + ia.ia_valid |= ATTR_MODE; + ia.ia_mode = h_isrc->i_mode; + } + sbits = !!(h_isrc->i_mode & (S_ISUID | S_ISGID)); + au_cpup_attr_flags(h_idst, h_isrc->i_flags); + } + /* no delegation since it is just created */ + err = vfsub_notify_change(&h_path, &ia, /*delegated*/NULL); + + /* is this nfs only? */ + if (!err && sbits && au_test_nfs(h_path.dentry->d_sb)) { + ia.ia_valid = ATTR_FORCE | ATTR_MODE; + ia.ia_mode = h_isrc->i_mode; + err = vfsub_notify_change(&h_path, &ia, /*delegated*/NULL); + } + + icex = br->br_perm & AuBrAttr_ICEX; + if (!err) { + mnt_flags = au_mntflags(dst->d_sb); + verbose = !!au_opt_test(mnt_flags, VERBOSE); + err = au_cpup_xattr(h_path.dentry, h_src, icex, verbose); + } + + return err; +} + +/* ---------------------------------------------------------------------- */ + +static int au_do_copy_file(struct file *dst, struct file *src, loff_t len, + char *buf, unsigned long blksize) +{ + int err; + size_t sz, rbytes, wbytes; + unsigned char all_zero; + char *p, *zp; + struct inode *h_inode; + /* reduce stack usage */ + struct iattr *ia; + + zp = page_address(ZERO_PAGE(0)); + if (unlikely(!zp)) + return -ENOMEM; /* possible? */ + + err = 0; + all_zero = 0; + while (len) { + AuDbg("len %lld\n", len); + sz = blksize; + if (len < blksize) + sz = len; + + rbytes = 0; + /* todo: signal_pending? */ + while (!rbytes || err == -EAGAIN || err == -EINTR) { + rbytes = vfsub_read_k(src, buf, sz, &src->f_pos); + err = rbytes; + } + if (unlikely(err < 0)) + break; + + all_zero = 0; + if (len >= rbytes && rbytes == blksize) + all_zero = !memcmp(buf, zp, rbytes); + if (!all_zero) { + wbytes = rbytes; + p = buf; + while (wbytes) { + size_t b; + + b = vfsub_write_k(dst, p, wbytes, &dst->f_pos); + err = b; + /* todo: signal_pending? */ + if (unlikely(err == -EAGAIN || err == -EINTR)) + continue; + if (unlikely(err < 0)) + break; + wbytes -= b; + p += b; + } + if (unlikely(err < 0)) + break; + } else { + loff_t res; + + AuLabel(hole); + res = vfsub_llseek(dst, rbytes, SEEK_CUR); + err = res; + if (unlikely(res < 0)) + break; + } + len -= rbytes; + err = 0; + } + + /* the last block may be a hole */ + if (!err && all_zero) { + AuLabel(last hole); + + err = 1; + if (au_test_nfs(dst->f_path.dentry->d_sb)) { + /* nfs requires this step to make last hole */ + /* is this only nfs? */ + do { + /* todo: signal_pending? */ + err = vfsub_write_k(dst, "\0", 1, &dst->f_pos); + } while (err == -EAGAIN || err == -EINTR); + if (err == 1) + dst->f_pos--; + } + + if (err == 1) { + ia = (void *)buf; + ia->ia_size = dst->f_pos; + ia->ia_valid = ATTR_SIZE | ATTR_FILE; + ia->ia_file = dst; + h_inode = file_inode(dst); + inode_lock_nested(h_inode, AuLsc_I_CHILD2); + /* no delegation since it is just created */ + err = vfsub_notify_change(&dst->f_path, ia, + /*delegated*/NULL); + inode_unlock(h_inode); + } + } + + return err; +} + +int au_copy_file(struct file *dst, struct file *src, loff_t len) +{ + int err; + unsigned long blksize; + unsigned char do_kfree; + char *buf; + + err = -ENOMEM; + blksize = dst->f_path.dentry->d_sb->s_blocksize; + if (!blksize || PAGE_SIZE < blksize) + blksize = PAGE_SIZE; + AuDbg("blksize %lu\n", blksize); + do_kfree = (blksize != PAGE_SIZE && blksize >= sizeof(struct iattr *)); + if (do_kfree) + buf = kmalloc(blksize, GFP_NOFS); + else + buf = (void *)__get_free_page(GFP_NOFS); + if (unlikely(!buf)) + goto out; + + if (len > (1 << 22)) + AuDbg("copying a large file %lld\n", (long long)len); + + src->f_pos = 0; + dst->f_pos = 0; + err = au_do_copy_file(dst, src, len, buf, blksize); + if (do_kfree) + kfree(buf); + else + free_page((unsigned long)buf); + +out: + return err; +} + +/* + * to support a sparse file which is opened with O_APPEND, + * we need to close the file. + */ +static int au_cp_regular(struct au_cp_generic *cpg) +{ + int err, i; + enum { SRC, DST }; + struct { + aufs_bindex_t bindex; + unsigned int flags; + struct dentry *dentry; + int force_wr; + struct file *file; + void *label; + } *f, file[] = { + { + .bindex = cpg->bsrc, + .flags = O_RDONLY | O_NOATIME | O_LARGEFILE, + .label = &&out + }, + { + .bindex = cpg->bdst, + .flags = O_WRONLY | O_NOATIME | O_LARGEFILE, + .force_wr = !!au_ftest_cpup(cpg->flags, RWDST), + .label = &&out_src + } + }; + struct super_block *sb; + struct task_struct *tsk = current; + + /* bsrc branch can be ro/rw. */ + sb = cpg->dentry->d_sb; + f = file; + for (i = 0; i < 2; i++, f++) { + f->dentry = au_h_dptr(cpg->dentry, f->bindex); + f->file = au_h_open(cpg->dentry, f->bindex, f->flags, + /*file*/NULL, f->force_wr); + err = PTR_ERR(f->file); + if (IS_ERR(f->file)) + goto *f->label; + } + + /* try stopping to update while we copyup */ + IMustLock(d_inode(file[SRC].dentry)); + err = au_copy_file(file[DST].file, file[SRC].file, cpg->len); + + /* i wonder if we had O_NO_DELAY_FPUT flag */ + if (tsk->flags & PF_KTHREAD) + __fput_sync(file[DST].file); + else { + WARN(1, "%pD\nPlease report this warning to aufs-users ML", + file[DST].file); + fput(file[DST].file); + /* + * too bad. + * we have to call both since we don't know which place the file + * was added to. + */ + task_work_run(); + flush_delayed_fput(); + } + au_sbr_put(sb, file[DST].bindex); + +out_src: + fput(file[SRC].file); + au_sbr_put(sb, file[SRC].bindex); +out: + return err; +} + +static int au_do_cpup_regular(struct au_cp_generic *cpg, + struct au_cpup_reg_attr *h_src_attr) +{ + int err, rerr; + loff_t l; + struct path h_path; + struct inode *h_src_inode, *h_dst_inode; + + err = 0; + h_src_inode = au_h_iptr(d_inode(cpg->dentry), cpg->bsrc); + l = i_size_read(h_src_inode); + if (cpg->len == -1 || l < cpg->len) + cpg->len = l; + if (cpg->len) { + /* try stopping to update while we are referencing */ + inode_lock_nested(h_src_inode, AuLsc_I_CHILD); + au_pin_hdir_unlock(cpg->pin); + + h_path.dentry = au_h_dptr(cpg->dentry, cpg->bsrc); + h_path.mnt = au_sbr_mnt(cpg->dentry->d_sb, cpg->bsrc); + h_src_attr->iflags = h_src_inode->i_flags; + if (!au_test_nfs(h_src_inode->i_sb)) + err = vfs_getattr(&h_path, &h_src_attr->st); + else { + inode_unlock(h_src_inode); + err = vfs_getattr(&h_path, &h_src_attr->st); + inode_lock_nested(h_src_inode, AuLsc_I_CHILD); + } + if (unlikely(err)) { + inode_unlock(h_src_inode); + goto out; + } + h_src_attr->valid = 1; + err = au_cp_regular(cpg); + inode_unlock(h_src_inode); + rerr = au_pin_hdir_relock(cpg->pin); + if (!err && rerr) + err = rerr; + } + if (!err && (h_src_inode->i_state & I_LINKABLE)) { + h_path.dentry = au_h_dptr(cpg->dentry, cpg->bdst); + h_dst_inode = d_inode(h_path.dentry); + spin_lock(&h_dst_inode->i_lock); + h_dst_inode->i_state |= I_LINKABLE; + spin_unlock(&h_dst_inode->i_lock); + } + +out: + return err; +} + +static int au_do_cpup_symlink(struct path *h_path, struct dentry *h_src, + struct inode *h_dir) +{ + int err, symlen; + mm_segment_t old_fs; + union { + char *k; + char __user *u; + } sym; + struct inode *h_inode = d_inode(h_src); + const struct inode_operations *h_iop = h_inode->i_op; + + err = -ENOSYS; + if (unlikely(!h_iop->readlink)) + goto out; + + err = -ENOMEM; + sym.k = (void *)__get_free_page(GFP_NOFS); + if (unlikely(!sym.k)) + goto out; + + /* unnecessary to support mmap_sem since symlink is not mmap-able */ + old_fs = get_fs(); + set_fs(KERNEL_DS); + symlen = h_iop->readlink(h_src, sym.u, PATH_MAX); + err = symlen; + set_fs(old_fs); + + if (symlen > 0) { + sym.k[symlen] = 0; + err = vfsub_symlink(h_dir, h_path, sym.k); + } + free_page((unsigned long)sym.k); + +out: + return err; +} + +/* + * regardless 'acl' option, reset all ACL. + * All ACL will be copied up later from the original entry on the lower branch. + */ +static int au_reset_acl(struct inode *h_dir, struct path *h_path, umode_t mode) +{ + int err; + struct dentry *h_dentry; + struct inode *h_inode; + + h_dentry = h_path->dentry; + h_inode = d_inode(h_dentry); + /* forget_all_cached_acls(h_inode)); */ + err = vfsub_removexattr(h_dentry, XATTR_NAME_POSIX_ACL_ACCESS); + AuTraceErr(err); + if (err == -EOPNOTSUPP) + err = 0; + if (!err) + err = vfsub_acl_chmod(h_inode, mode); + + AuTraceErr(err); + return err; +} + +static int au_do_cpup_dir(struct au_cp_generic *cpg, struct dentry *dst_parent, + struct inode *h_dir, struct path *h_path) +{ + int err; + struct inode *dir, *inode; + + err = vfsub_removexattr(h_path->dentry, XATTR_NAME_POSIX_ACL_DEFAULT); + AuTraceErr(err); + if (err == -EOPNOTSUPP) + err = 0; + if (unlikely(err)) + goto out; + + /* + * strange behaviour from the users view, + * particularry setattr case + */ + dir = d_inode(dst_parent); + if (au_ibstart(dir) == cpg->bdst) + au_cpup_attr_nlink(dir, /*force*/1); + inode = d_inode(cpg->dentry); + au_cpup_attr_nlink(inode, /*force*/1); + +out: + return err; +} + +static noinline_for_stack +int cpup_entry(struct au_cp_generic *cpg, struct dentry *dst_parent, + struct au_cpup_reg_attr *h_src_attr) +{ + int err; + umode_t mode; + unsigned int mnt_flags; + unsigned char isdir, isreg, force; + const unsigned char do_dt = !!au_ftest_cpup(cpg->flags, DTIME); + struct au_dtime dt; + struct path h_path; + struct dentry *h_src, *h_dst, *h_parent; + struct inode *h_inode, *h_dir; + struct super_block *sb; + + /* bsrc branch can be ro/rw. */ + h_src = au_h_dptr(cpg->dentry, cpg->bsrc); + h_inode = d_inode(h_src); + AuDebugOn(h_inode != au_h_iptr(d_inode(cpg->dentry), cpg->bsrc)); + + /* try stopping to be referenced while we are creating */ + h_dst = au_h_dptr(cpg->dentry, cpg->bdst); + if (au_ftest_cpup(cpg->flags, RENAME)) + AuDebugOn(strncmp(h_dst->d_name.name, AUFS_WH_PFX, + AUFS_WH_PFX_LEN)); + h_parent = h_dst->d_parent; /* dir inode is locked */ + h_dir = d_inode(h_parent); + IMustLock(h_dir); + AuDebugOn(h_parent != h_dst->d_parent); + + sb = cpg->dentry->d_sb; + h_path.mnt = au_sbr_mnt(sb, cpg->bdst); + if (do_dt) { + h_path.dentry = h_parent; + au_dtime_store(&dt, dst_parent, &h_path); + } + h_path.dentry = h_dst; + + isreg = 0; + isdir = 0; + mode = h_inode->i_mode; + switch (mode & S_IFMT) { + case S_IFREG: + isreg = 1; + err = vfsub_create(h_dir, &h_path, S_IRUSR | S_IWUSR, + /*want_excl*/true); + if (!err) + err = au_do_cpup_regular(cpg, h_src_attr); + break; + case S_IFDIR: + isdir = 1; + err = vfsub_mkdir(h_dir, &h_path, mode); + if (!err) + err = au_do_cpup_dir(cpg, dst_parent, h_dir, &h_path); + break; + case S_IFLNK: + err = au_do_cpup_symlink(&h_path, h_src, h_dir); + break; + case S_IFCHR: + case S_IFBLK: + AuDebugOn(!capable(CAP_MKNOD)); + /*FALLTHROUGH*/ + case S_IFIFO: + case S_IFSOCK: + err = vfsub_mknod(h_dir, &h_path, mode, h_inode->i_rdev); + break; + default: + AuIOErr("Unknown inode type 0%o\n", mode); + err = -EIO; + } + if (!err) + err = au_reset_acl(h_dir, &h_path, mode); + + mnt_flags = au_mntflags(sb); + if (!au_opt_test(mnt_flags, UDBA_NONE) + && !isdir + && au_opt_test(mnt_flags, XINO) + && (h_inode->i_nlink == 1 + || (h_inode->i_state & I_LINKABLE)) + /* todo: unnecessary? */ + /* && d_inode(cpg->dentry)->i_nlink == 1 */ + && cpg->bdst < cpg->bsrc + && !au_ftest_cpup(cpg->flags, KEEPLINO)) + au_xino_write(sb, cpg->bsrc, h_inode->i_ino, /*ino*/0); + /* ignore this error */ + + if (!err) { + force = 0; + if (isreg) { + force = !!cpg->len; + if (cpg->len == -1) + force = !!i_size_read(h_inode); + } + au_fhsm_wrote(sb, cpg->bdst, force); + } + + if (do_dt) + au_dtime_revert(&dt); + return err; +} + +static int au_do_ren_after_cpup(struct au_cp_generic *cpg, struct path *h_path) +{ + int err; + struct dentry *dentry, *h_dentry, *h_parent, *parent; + struct inode *h_dir; + aufs_bindex_t bdst; + + dentry = cpg->dentry; + bdst = cpg->bdst; + h_dentry = au_h_dptr(dentry, bdst); + if (!au_ftest_cpup(cpg->flags, OVERWRITE)) { + dget(h_dentry); + au_set_h_dptr(dentry, bdst, NULL); + err = au_lkup_neg(dentry, bdst, /*wh*/0); + if (!err) + h_path->dentry = dget(au_h_dptr(dentry, bdst)); + au_set_h_dptr(dentry, bdst, h_dentry); + } else { + err = 0; + parent = dget_parent(dentry); + h_parent = au_h_dptr(parent, bdst); + dput(parent); + h_path->dentry = vfsub_lkup_one(&dentry->d_name, h_parent); + if (IS_ERR(h_path->dentry)) + err = PTR_ERR(h_path->dentry); + } + if (unlikely(err)) + goto out; + + h_parent = h_dentry->d_parent; /* dir inode is locked */ + h_dir = d_inode(h_parent); + IMustLock(h_dir); + AuDbg("%pd %pd\n", h_dentry, h_path->dentry); + /* no delegation since it is just created */ + err = vfsub_rename(h_dir, h_dentry, h_dir, h_path, /*delegated*/NULL); + dput(h_path->dentry); + +out: + return err; +} + +/* + * copyup the @dentry from @bsrc to @bdst. + * the caller must set the both of lower dentries. + * @len is for truncating when it is -1 copyup the entire file. + * in link/rename cases, @dst_parent may be different from the real one. + * basic->bsrc can be larger than basic->bdst. + */ +static int au_cpup_single(struct au_cp_generic *cpg, struct dentry *dst_parent) +{ + int err, rerr; + aufs_bindex_t old_ibstart; + unsigned char isdir, plink; + struct dentry *h_src, *h_dst, *h_parent; + struct inode *dst_inode, *h_dir, *inode, *delegated, *src_inode; + struct super_block *sb; + struct au_branch *br; + /* to reuduce stack size */ + struct { + struct au_dtime dt; + struct path h_path; + struct au_cpup_reg_attr h_src_attr; + } *a; + + err = -ENOMEM; + a = kmalloc(sizeof(*a), GFP_NOFS); + if (unlikely(!a)) + goto out; + a->h_src_attr.valid = 0; + + sb = cpg->dentry->d_sb; + br = au_sbr(sb, cpg->bdst); + a->h_path.mnt = au_br_mnt(br); + h_dst = au_h_dptr(cpg->dentry, cpg->bdst); + h_parent = h_dst->d_parent; /* dir inode is locked */ + h_dir = d_inode(h_parent); + IMustLock(h_dir); + + h_src = au_h_dptr(cpg->dentry, cpg->bsrc); + inode = d_inode(cpg->dentry); + + if (!dst_parent) + dst_parent = dget_parent(cpg->dentry); + else + dget(dst_parent); + + plink = !!au_opt_test(au_mntflags(sb), PLINK); + dst_inode = au_h_iptr(inode, cpg->bdst); + if (dst_inode) { + if (unlikely(!plink)) { + err = -EIO; + AuIOErr("hi%lu(i%lu) exists on b%d " + "but plink is disabled\n", + dst_inode->i_ino, inode->i_ino, cpg->bdst); + goto out_parent; + } + + if (dst_inode->i_nlink) { + const int do_dt = au_ftest_cpup(cpg->flags, DTIME); + + h_src = au_plink_lkup(inode, cpg->bdst); + err = PTR_ERR(h_src); + if (IS_ERR(h_src)) + goto out_parent; + if (unlikely(d_is_negative(h_src))) { + err = -EIO; + AuIOErr("i%lu exists on b%d " + "but not pseudo-linked\n", + inode->i_ino, cpg->bdst); + dput(h_src); + goto out_parent; + } + + if (do_dt) { + a->h_path.dentry = h_parent; + au_dtime_store(&a->dt, dst_parent, &a->h_path); + } + + a->h_path.dentry = h_dst; + delegated = NULL; + err = vfsub_link(h_src, h_dir, &a->h_path, &delegated); + if (!err && au_ftest_cpup(cpg->flags, RENAME)) + err = au_do_ren_after_cpup(cpg, &a->h_path); + if (do_dt) + au_dtime_revert(&a->dt); + if (unlikely(err == -EWOULDBLOCK)) { + pr_warn("cannot retry for NFSv4 delegation" + " for an internal link\n"); + iput(delegated); + } + dput(h_src); + goto out_parent; + } else + /* todo: cpup_wh_file? */ + /* udba work */ + au_update_ibrange(inode, /*do_put_zero*/1); + } + + isdir = S_ISDIR(inode->i_mode); + old_ibstart = au_ibstart(inode); + err = cpup_entry(cpg, dst_parent, &a->h_src_attr); + if (unlikely(err)) + goto out_rev; + dst_inode = d_inode(h_dst); + inode_lock_nested(dst_inode, AuLsc_I_CHILD2); + /* todo: necessary? */ + /* au_pin_hdir_unlock(cpg->pin); */ + + err = cpup_iattr(cpg->dentry, cpg->bdst, h_src, &a->h_src_attr); + if (unlikely(err)) { + /* todo: necessary? */ + /* au_pin_hdir_relock(cpg->pin); */ /* ignore an error */ + inode_unlock(dst_inode); + goto out_rev; + } + + if (cpg->bdst < old_ibstart) { + if (S_ISREG(inode->i_mode)) { + err = au_dy_iaop(inode, cpg->bdst, dst_inode); + if (unlikely(err)) { + /* ignore an error */ + /* au_pin_hdir_relock(cpg->pin); */ + inode_unlock(dst_inode); + goto out_rev; + } + } + au_set_ibstart(inode, cpg->bdst); + } else + au_set_ibend(inode, cpg->bdst); + au_set_h_iptr(inode, cpg->bdst, au_igrab(dst_inode), + au_hi_flags(inode, isdir)); + + /* todo: necessary? */ + /* err = au_pin_hdir_relock(cpg->pin); */ + inode_unlock(dst_inode); + if (unlikely(err)) + goto out_rev; + + src_inode = d_inode(h_src); + if (!isdir + && (src_inode->i_nlink > 1 + || src_inode->i_state & I_LINKABLE) + && plink) + au_plink_append(inode, cpg->bdst, h_dst); + + if (au_ftest_cpup(cpg->flags, RENAME)) { + a->h_path.dentry = h_dst; + err = au_do_ren_after_cpup(cpg, &a->h_path); + } + if (!err) + goto out_parent; /* success */ + + /* revert */ +out_rev: + a->h_path.dentry = h_parent; + au_dtime_store(&a->dt, dst_parent, &a->h_path); + a->h_path.dentry = h_dst; + rerr = 0; + if (d_is_positive(h_dst)) { + if (!isdir) { + /* no delegation since it is just created */ + rerr = vfsub_unlink(h_dir, &a->h_path, + /*delegated*/NULL, /*force*/0); + } else + rerr = vfsub_rmdir(h_dir, &a->h_path); + } + au_dtime_revert(&a->dt); + if (rerr) { + AuIOErr("failed removing broken entry(%d, %d)\n", err, rerr); + err = -EIO; + } +out_parent: + dput(dst_parent); + kfree(a); +out: + return err; +} + +#if 0 /* reserved */ +struct au_cpup_single_args { + int *errp; + struct au_cp_generic *cpg; + struct dentry *dst_parent; +}; + +static void au_call_cpup_single(void *args) +{ + struct au_cpup_single_args *a = args; + + au_pin_hdir_acquire_nest(a->cpg->pin); + *a->errp = au_cpup_single(a->cpg, a->dst_parent); + au_pin_hdir_release(a->cpg->pin); +} +#endif + +/* + * prevent SIGXFSZ in copy-up. + * testing CAP_MKNOD is for generic fs, + * but CAP_FSETID is for xfs only, currently. + */ +static int au_cpup_sio_test(struct au_pin *pin, umode_t mode) +{ + int do_sio; + struct super_block *sb; + struct inode *h_dir; + + do_sio = 0; + sb = au_pinned_parent(pin)->d_sb; + if (!au_wkq_test() + && (!au_sbi(sb)->si_plink_maint_pid + || au_plink_maint(sb, AuLock_NOPLM))) { + switch (mode & S_IFMT) { + case S_IFREG: + /* no condition about RLIMIT_FSIZE and the file size */ + do_sio = 1; + break; + case S_IFCHR: + case S_IFBLK: + do_sio = !capable(CAP_MKNOD); + break; + } + if (!do_sio) + do_sio = ((mode & (S_ISUID | S_ISGID)) + && !capable(CAP_FSETID)); + /* this workaround may be removed in the future */ + if (!do_sio) { + h_dir = au_pinned_h_dir(pin); + do_sio = h_dir->i_mode & S_ISVTX; + } + } + + return do_sio; +} + +#if 0 /* reserved */ +int au_sio_cpup_single(struct au_cp_generic *cpg, struct dentry *dst_parent) +{ + int err, wkq_err; + struct dentry *h_dentry; + + h_dentry = au_h_dptr(cpg->dentry, cpg->bsrc); + if (!au_cpup_sio_test(pin, d_inode(h_dentry)->i_mode)) + err = au_cpup_single(cpg, dst_parent); + else { + struct au_cpup_single_args args = { + .errp = &err, + .cpg = cpg, + .dst_parent = dst_parent + }; + wkq_err = au_wkq_wait(au_call_cpup_single, &args); + if (unlikely(wkq_err)) + err = wkq_err; + } + + return err; +} +#endif + +/* + * copyup the @dentry from the first active lower branch to @bdst, + * using au_cpup_single(). + */ +static int au_cpup_simple(struct au_cp_generic *cpg) +{ + int err; + unsigned int flags_orig; + struct dentry *dentry; + + AuDebugOn(cpg->bsrc < 0); + + dentry = cpg->dentry; + DiMustWriteLock(dentry); + + err = au_lkup_neg(dentry, cpg->bdst, /*wh*/1); + if (!err) { + flags_orig = cpg->flags; + au_fset_cpup(cpg->flags, RENAME); + err = au_cpup_single(cpg, NULL); + cpg->flags = flags_orig; + if (!err) + return 0; /* success */ + + /* revert */ + au_set_h_dptr(dentry, cpg->bdst, NULL); + au_set_dbstart(dentry, cpg->bsrc); + } + + return err; +} + +struct au_cpup_simple_args { + int *errp; + struct au_cp_generic *cpg; +}; + +static void au_call_cpup_simple(void *args) +{ + struct au_cpup_simple_args *a = args; + + au_pin_hdir_acquire_nest(a->cpg->pin); + *a->errp = au_cpup_simple(a->cpg); + au_pin_hdir_release(a->cpg->pin); +} + +static int au_do_sio_cpup_simple(struct au_cp_generic *cpg) +{ + int err, wkq_err; + struct dentry *dentry, *parent; + struct file *h_file; + struct inode *h_dir; + + dentry = cpg->dentry; + h_file = NULL; + if (au_ftest_cpup(cpg->flags, HOPEN)) { + AuDebugOn(cpg->bsrc < 0); + h_file = au_h_open_pre(dentry, cpg->bsrc, /*force_wr*/0); + err = PTR_ERR(h_file); + if (IS_ERR(h_file)) + goto out; + } + + parent = dget_parent(dentry); + h_dir = au_h_iptr(d_inode(parent), cpg->bdst); + if (!au_test_h_perm_sio(h_dir, MAY_EXEC | MAY_WRITE) + && !au_cpup_sio_test(cpg->pin, d_inode(dentry)->i_mode)) + err = au_cpup_simple(cpg); + else { + struct au_cpup_simple_args args = { + .errp = &err, + .cpg = cpg + }; + wkq_err = au_wkq_wait(au_call_cpup_simple, &args); + if (unlikely(wkq_err)) + err = wkq_err; + } + + dput(parent); + if (h_file) + au_h_open_post(dentry, cpg->bsrc, h_file); + +out: + return err; +} + +int au_sio_cpup_simple(struct au_cp_generic *cpg) +{ + aufs_bindex_t bsrc, bend; + struct dentry *dentry, *h_dentry; + + if (cpg->bsrc < 0) { + dentry = cpg->dentry; + bend = au_dbend(dentry); + for (bsrc = cpg->bdst + 1; bsrc <= bend; bsrc++) { + h_dentry = au_h_dptr(dentry, bsrc); + if (h_dentry) { + AuDebugOn(d_is_negative(h_dentry)); + break; + } + } + AuDebugOn(bsrc > bend); + cpg->bsrc = bsrc; + } + AuDebugOn(cpg->bsrc <= cpg->bdst); + return au_do_sio_cpup_simple(cpg); +} + +int au_sio_cpdown_simple(struct au_cp_generic *cpg) +{ + AuDebugOn(cpg->bdst <= cpg->bsrc); + return au_do_sio_cpup_simple(cpg); +} + +/* ---------------------------------------------------------------------- */ + +/* + * copyup the deleted file for writing. + */ +static int au_do_cpup_wh(struct au_cp_generic *cpg, struct dentry *wh_dentry, + struct file *file) +{ + int err; + unsigned int flags_orig; + aufs_bindex_t bsrc_orig; + struct dentry *h_d_dst, *h_d_start; + struct au_dinfo *dinfo; + struct au_hdentry *hdp; + + dinfo = au_di(cpg->dentry); + AuRwMustWriteLock(&dinfo->di_rwsem); + + bsrc_orig = cpg->bsrc; + cpg->bsrc = dinfo->di_bstart; + hdp = dinfo->di_hdentry; + h_d_dst = hdp[0 + cpg->bdst].hd_dentry; + dinfo->di_bstart = cpg->bdst; + hdp[0 + cpg->bdst].hd_dentry = wh_dentry; + h_d_start = NULL; + if (file) { + h_d_start = hdp[0 + cpg->bsrc].hd_dentry; + hdp[0 + cpg->bsrc].hd_dentry = au_hf_top(file)->f_path.dentry; + } + flags_orig = cpg->flags; + cpg->flags = !AuCpup_DTIME; + err = au_cpup_single(cpg, /*h_parent*/NULL); + cpg->flags = flags_orig; + if (file) { + if (!err) + err = au_reopen_nondir(file); + hdp[0 + cpg->bsrc].hd_dentry = h_d_start; + } + hdp[0 + cpg->bdst].hd_dentry = h_d_dst; + dinfo->di_bstart = cpg->bsrc; + cpg->bsrc = bsrc_orig; + + return err; +} + +static int au_cpup_wh(struct au_cp_generic *cpg, struct file *file) +{ + int err; + aufs_bindex_t bdst; + struct au_dtime dt; + struct dentry *dentry, *parent, *h_parent, *wh_dentry; + struct au_branch *br; + struct path h_path; + + dentry = cpg->dentry; + bdst = cpg->bdst; + br = au_sbr(dentry->d_sb, bdst); + parent = dget_parent(dentry); + h_parent = au_h_dptr(parent, bdst); + wh_dentry = au_whtmp_lkup(h_parent, br, &dentry->d_name); + err = PTR_ERR(wh_dentry); + if (IS_ERR(wh_dentry)) + goto out; + + h_path.dentry = h_parent; + h_path.mnt = au_br_mnt(br); + au_dtime_store(&dt, parent, &h_path); + err = au_do_cpup_wh(cpg, wh_dentry, file); + if (unlikely(err)) + goto out_wh; + + dget(wh_dentry); + h_path.dentry = wh_dentry; + if (!d_is_dir(wh_dentry)) { + /* no delegation since it is just created */ + err = vfsub_unlink(d_inode(h_parent), &h_path, + /*delegated*/NULL, /*force*/0); + } else + err = vfsub_rmdir(d_inode(h_parent), &h_path); + if (unlikely(err)) { + AuIOErr("failed remove copied-up tmp file %pd(%d)\n", + wh_dentry, err); + err = -EIO; + } + au_dtime_revert(&dt); + au_set_hi_wh(d_inode(dentry), bdst, wh_dentry); + +out_wh: + dput(wh_dentry); +out: + dput(parent); + return err; +} + +struct au_cpup_wh_args { + int *errp; + struct au_cp_generic *cpg; + struct file *file; +}; + +static void au_call_cpup_wh(void *args) +{ + struct au_cpup_wh_args *a = args; + + au_pin_hdir_acquire_nest(a->cpg->pin); + *a->errp = au_cpup_wh(a->cpg, a->file); + au_pin_hdir_release(a->cpg->pin); +} + +int au_sio_cpup_wh(struct au_cp_generic *cpg, struct file *file) +{ + int err, wkq_err; + aufs_bindex_t bdst; + struct dentry *dentry, *parent, *h_orph, *h_parent; + struct inode *dir, *h_dir, *h_tmpdir; + struct au_wbr *wbr; + struct au_pin wh_pin, *pin_orig; + + dentry = cpg->dentry; + bdst = cpg->bdst; + parent = dget_parent(dentry); + dir = d_inode(parent); + h_orph = NULL; + h_parent = NULL; + h_dir = au_igrab(au_h_iptr(dir, bdst)); + h_tmpdir = h_dir; + pin_orig = NULL; + if (!h_dir->i_nlink) { + wbr = au_sbr(dentry->d_sb, bdst)->br_wbr; + h_orph = wbr->wbr_orph; + + h_parent = dget(au_h_dptr(parent, bdst)); + au_set_h_dptr(parent, bdst, dget(h_orph)); + h_tmpdir = d_inode(h_orph); + au_set_h_iptr(dir, bdst, au_igrab(h_tmpdir), /*flags*/0); + + inode_lock_nested(h_tmpdir, AuLsc_I_PARENT3); + /* todo: au_h_open_pre()? */ + + pin_orig = cpg->pin; + au_pin_init(&wh_pin, dentry, bdst, AuLsc_DI_PARENT, + AuLsc_I_PARENT3, cpg->pin->udba, AuPin_DI_LOCKED); + cpg->pin = &wh_pin; + } + + if (!au_test_h_perm_sio(h_tmpdir, MAY_EXEC | MAY_WRITE) + && !au_cpup_sio_test(cpg->pin, d_inode(dentry)->i_mode)) + err = au_cpup_wh(cpg, file); + else { + struct au_cpup_wh_args args = { + .errp = &err, + .cpg = cpg, + .file = file + }; + wkq_err = au_wkq_wait(au_call_cpup_wh, &args); + if (unlikely(wkq_err)) + err = wkq_err; + } + + if (h_orph) { + inode_unlock(h_tmpdir); + /* todo: au_h_open_post()? */ + au_set_h_iptr(dir, bdst, au_igrab(h_dir), /*flags*/0); + au_set_h_dptr(parent, bdst, h_parent); + AuDebugOn(!pin_orig); + cpg->pin = pin_orig; + } + iput(h_dir); + dput(parent); + + return err; +} + +/* ---------------------------------------------------------------------- */ + +/* + * generic routine for both of copy-up and copy-down. + */ +/* cf. revalidate function in file.c */ +int au_cp_dirs(struct dentry *dentry, aufs_bindex_t bdst, + int (*cp)(struct dentry *dentry, aufs_bindex_t bdst, + struct au_pin *pin, + struct dentry *h_parent, void *arg), + void *arg) +{ + int err; + struct au_pin pin; + struct dentry *d, *parent, *h_parent, *real_parent, *h_dentry; + + err = 0; + parent = dget_parent(dentry); + if (IS_ROOT(parent)) + goto out; + + au_pin_init(&pin, dentry, bdst, AuLsc_DI_PARENT2, AuLsc_I_PARENT2, + au_opt_udba(dentry->d_sb), AuPin_MNT_WRITE); + + /* do not use au_dpage */ + real_parent = parent; + while (1) { + dput(parent); + parent = dget_parent(dentry); + h_parent = au_h_dptr(parent, bdst); + if (h_parent) + goto out; /* success */ + + /* find top dir which is necessary to cpup */ + do { + d = parent; + dput(parent); + parent = dget_parent(d); + di_read_lock_parent3(parent, !AuLock_IR); + h_parent = au_h_dptr(parent, bdst); + di_read_unlock(parent, !AuLock_IR); + } while (!h_parent); + + if (d != real_parent) + di_write_lock_child3(d); + + /* somebody else might create while we were sleeping */ + h_dentry = au_h_dptr(d, bdst); + if (!h_dentry || d_is_negative(h_dentry)) { + if (h_dentry) + au_update_dbstart(d); + + au_pin_set_dentry(&pin, d); + err = au_do_pin(&pin); + if (!err) { + err = cp(d, bdst, &pin, h_parent, arg); + au_unpin(&pin); + } + } + + if (d != real_parent) + di_write_unlock(d); + if (unlikely(err)) + break; + } + +out: + dput(parent); + return err; +} + +static int au_cpup_dir(struct dentry *dentry, aufs_bindex_t bdst, + struct au_pin *pin, + struct dentry *h_parent __maybe_unused, + void *arg __maybe_unused) +{ + struct au_cp_generic cpg = { + .dentry = dentry, + .bdst = bdst, + .bsrc = -1, + .len = 0, + .pin = pin, + .flags = AuCpup_DTIME + }; + return au_sio_cpup_simple(&cpg); +} + +int au_cpup_dirs(struct dentry *dentry, aufs_bindex_t bdst) +{ + return au_cp_dirs(dentry, bdst, au_cpup_dir, NULL); +} + +int au_test_and_cpup_dirs(struct dentry *dentry, aufs_bindex_t bdst) +{ + int err; + struct dentry *parent; + struct inode *dir; + + parent = dget_parent(dentry); + dir = d_inode(parent); + err = 0; + if (au_h_iptr(dir, bdst)) + goto out; + + di_read_unlock(parent, AuLock_IR); + di_write_lock_parent(parent); + /* someone else might change our inode while we were sleeping */ + if (!au_h_iptr(dir, bdst)) + err = au_cpup_dirs(dentry, bdst); + di_downgrade_lock(parent, AuLock_IR); + +out: + dput(parent); + return err; +} diff --git a/fs/aufs/cpup.h b/fs/aufs/cpup.h new file mode 100644 index 000000000..ccba2c427 --- /dev/null +++ b/fs/aufs/cpup.h @@ -0,0 +1,81 @@ +/* + * Copyright (C) 2005-2016 Junjiro R. Okajima + */ + +/* + * copy-up/down functions + */ + +#ifndef __AUFS_CPUP_H__ +#define __AUFS_CPUP_H__ + +#ifdef __KERNEL__ + +#include <linux/path.h> + +struct inode; +struct file; +struct au_pin; + +void au_cpup_attr_flags(struct inode *dst, unsigned int iflags); +void au_cpup_attr_timesizes(struct inode *inode); +void au_cpup_attr_nlink(struct inode *inode, int force); +void au_cpup_attr_changeable(struct inode *inode); +void au_cpup_igen(struct inode *inode, struct inode *h_inode); +void au_cpup_attr_all(struct inode *inode, int force); + +/* ---------------------------------------------------------------------- */ + +struct au_cp_generic { + struct dentry *dentry; + aufs_bindex_t bdst, bsrc; + loff_t len; + struct au_pin *pin; + unsigned int flags; +}; + +/* cpup flags */ +#define AuCpup_DTIME 1 /* do dtime_store/revert */ +#define AuCpup_KEEPLINO (1 << 1) /* do not clear the lower xino, + for link(2) */ +#define AuCpup_RENAME (1 << 2) /* rename after cpup */ +#define AuCpup_HOPEN (1 << 3) /* call h_open_pre/post() in + cpup */ +#define AuCpup_OVERWRITE (1 << 4) /* allow overwriting the + existing entry */ +#define AuCpup_RWDST (1 << 5) /* force write target even if + the branch is marked as RO */ + +#define au_ftest_cpup(flags, name) ((flags) & AuCpup_##name) +#define au_fset_cpup(flags, name) \ + do { (flags) |= AuCpup_##name; } while (0) +#define au_fclr_cpup(flags, name) \ + do { (flags) &= ~AuCpup_##name; } while (0) + +int au_copy_file(struct file *dst, struct file *src, loff_t len); +int au_sio_cpup_simple(struct au_cp_generic *cpg); +int au_sio_cpdown_simple(struct au_cp_generic *cpg); +int au_sio_cpup_wh(struct au_cp_generic *cpg, struct file *file); + +int au_cp_dirs(struct dentry *dentry, aufs_bindex_t bdst, + int (*cp)(struct dentry *dentry, aufs_bindex_t bdst, + struct au_pin *pin, + struct dentry *h_parent, void *arg), + void *arg); +int au_cpup_dirs(struct dentry *dentry, aufs_bindex_t bdst); +int au_test_and_cpup_dirs(struct dentry *dentry, aufs_bindex_t bdst); + +/* ---------------------------------------------------------------------- */ + +/* keep timestamps when copyup */ +struct au_dtime { + struct dentry *dt_dentry; + struct path dt_h_path; + struct timespec dt_atime, dt_mtime; +}; +void au_dtime_store(struct au_dtime *dt, struct dentry *dentry, + struct path *h_path); +void au_dtime_revert(struct au_dtime *dt); + +#endif /* __KERNEL__ */ +#endif /* __AUFS_CPUP_H__ */ diff --git a/fs/aufs/dbgaufs.c b/fs/aufs/dbgaufs.c new file mode 100644 index 000000000..0aefb5ed8 --- /dev/null +++ b/fs/aufs/dbgaufs.c @@ -0,0 +1,419 @@ +/* + * Copyright (C) 2005-2016 Junjiro R. Okajima + */ + +/* + * debugfs interface + */ + +#include <linux/debugfs.h> +#include "aufs.h" + +#ifndef CONFIG_SYSFS +#error DEBUG_FS depends upon SYSFS +#endif + +static struct dentry *dbgaufs; +static const mode_t dbgaufs_mode = S_IRUSR | S_IRGRP | S_IROTH; + +/* 20 is max digits length of ulong 64 */ +struct dbgaufs_arg { + int n; + char a[20 * 4]; +}; + +/* + * common function for all XINO files + */ +static int dbgaufs_xi_release(struct inode *inode __maybe_unused, + struct file *file) +{ + kfree(file->private_data); + return 0; +} + +static int dbgaufs_xi_open(struct file *xf, struct file *file, int do_fcnt) +{ + int err; + struct kstat st; + struct dbgaufs_arg *p; + + err = -ENOMEM; + p = kmalloc(sizeof(*p), GFP_NOFS); + if (unlikely(!p)) + goto out; + + err = 0; + p->n = 0; + file->private_data = p; + if (!xf) + goto out; + + err = vfs_getattr(&xf->f_path, &st); + if (!err) { + if (do_fcnt) + p->n = snprintf + (p->a, sizeof(p->a), "%ld, %llux%lu %lld\n", + (long)file_count(xf), st.blocks, st.blksize, + (long long)st.size); + else + p->n = snprintf(p->a, sizeof(p->a), "%llux%lu %lld\n", + st.blocks, st.blksize, + (long long)st.size); + AuDebugOn(p->n >= sizeof(p->a)); + } else { + p->n = snprintf(p->a, sizeof(p->a), "err %d\n", err); + err = 0; + } + +out: + return err; + +} + +static ssize_t dbgaufs_xi_read(struct file *file, char __user *buf, + size_t count, loff_t *ppos) +{ + struct dbgaufs_arg *p; + + p = file->private_data; + return simple_read_from_buffer(buf, count, ppos, p->a, p->n); +} + +/* ---------------------------------------------------------------------- */ + +struct dbgaufs_plink_arg { + int n; + char a[]; +}; + +static int dbgaufs_plink_release(struct inode *inode __maybe_unused, + struct file *file) +{ + free_page((unsigned long)file->private_data); + return 0; +} + +static int dbgaufs_plink_open(struct inode *inode, struct file *file) +{ + int err, i, limit; + unsigned long n, sum; + struct dbgaufs_plink_arg *p; + struct au_sbinfo *sbinfo; + struct super_block *sb; + struct au_sphlhead *sphl; + + err = -ENOMEM; + p = (void *)get_zeroed_page(GFP_NOFS); + if (unlikely(!p)) + goto out; + + err = -EFBIG; + sbinfo = inode->i_private; + sb = sbinfo->si_sb; + si_noflush_read_lock(sb); + if (au_opt_test(au_mntflags(sb), PLINK)) { + limit = PAGE_SIZE - sizeof(p->n); + + /* the number of buckets */ + n = snprintf(p->a + p->n, limit, "%d\n", AuPlink_NHASH); + p->n += n; + limit -= n; + + sum = 0; + for (i = 0, sphl = sbinfo->si_plink; + i < AuPlink_NHASH; + i++, sphl++) { + n = au_sphl_count(sphl); + sum += n; + + n = snprintf(p->a + p->n, limit, "%lu ", n); + p->n += n; + limit -= n; + if (unlikely(limit <= 0)) + goto out_free; + } + p->a[p->n - 1] = '\n'; + + /* the sum of plinks */ + n = snprintf(p->a + p->n, limit, "%lu\n", sum); + p->n += n; + limit -= n; + if (unlikely(limit <= 0)) + goto out_free; + } else { +#define str "1\n0\n0\n" + p->n = sizeof(str) - 1; + strcpy(p->a, str); +#undef str + } + si_read_unlock(sb); + + err = 0; + file->private_data = p; + goto out; /* success */ + +out_free: + free_page((unsigned long)p); +out: + return err; +} + +static ssize_t dbgaufs_plink_read(struct file *file, char __user *buf, + size_t count, loff_t *ppos) +{ + struct dbgaufs_plink_arg *p; + + p = file->private_data; + return simple_read_from_buffer(buf, count, ppos, p->a, p->n); +} + +static const struct file_operations dbgaufs_plink_fop = { + .owner = THIS_MODULE, + .open = dbgaufs_plink_open, + .release = dbgaufs_plink_release, + .read = dbgaufs_plink_read +}; + +/* ---------------------------------------------------------------------- */ + +static int dbgaufs_xib_open(struct inode *inode, struct file *file) +{ + int err; + struct au_sbinfo *sbinfo; + struct super_block *sb; + + sbinfo = inode->i_private; + sb = sbinfo->si_sb; + si_noflush_read_lock(sb); + err = dbgaufs_xi_open(sbinfo->si_xib, file, /*do_fcnt*/0); + si_read_unlock(sb); + return err; +} + +static const struct file_operations dbgaufs_xib_fop = { + .owner = THIS_MODULE, + .open = dbgaufs_xib_open, + .release = dbgaufs_xi_release, + .read = dbgaufs_xi_read +}; + +/* ---------------------------------------------------------------------- */ + +#define DbgaufsXi_PREFIX "xi" + +static int dbgaufs_xino_open(struct inode *inode, struct file *file) +{ + int err; + long l; + struct au_sbinfo *sbinfo; + struct super_block *sb; + struct file *xf; + struct qstr *name; + + err = -ENOENT; + xf = NULL; + name = &file->f_path.dentry->d_name; + if (unlikely(name->len < sizeof(DbgaufsXi_PREFIX) + || memcmp(name->name, DbgaufsXi_PREFIX, + sizeof(DbgaufsXi_PREFIX) - 1))) + goto out; + err = kstrtol(name->name + sizeof(DbgaufsXi_PREFIX) - 1, 10, &l); + if (unlikely(err)) + goto out; + + sbinfo = inode->i_private; + sb = sbinfo->si_sb; + si_noflush_read_lock(sb); + if (l <= au_sbend(sb)) { + xf = au_sbr(sb, (aufs_bindex_t)l)->br_xino.xi_file; + err = dbgaufs_xi_open(xf, file, /*do_fcnt*/1); + } else + err = -ENOENT; + si_read_unlock(sb); + +out: + return err; +} + +static const struct file_operations dbgaufs_xino_fop = { + .owner = THIS_MODULE, + .open = dbgaufs_xino_open, + .release = dbgaufs_xi_release, + .read = dbgaufs_xi_read +}; + +void dbgaufs_brs_del(struct super_block *sb, aufs_bindex_t bindex) +{ + aufs_bindex_t bend; + struct au_branch *br; + struct au_xino_file *xi; + + if (!au_sbi(sb)->si_dbgaufs) + return; + + bend = au_sbend(sb); + for (; bindex <= bend; bindex++) { + br = au_sbr(sb, bindex); + xi = &br->br_xino; + debugfs_remove(xi->xi_dbgaufs); + xi->xi_dbgaufs = NULL; + } +} + +void dbgaufs_brs_add(struct super_block *sb, aufs_bindex_t bindex) +{ + struct au_sbinfo *sbinfo; + struct dentry *parent; + struct au_branch *br; + struct au_xino_file *xi; + aufs_bindex_t bend; + char name[sizeof(DbgaufsXi_PREFIX) + 5]; /* "xi" bindex NULL */ + + sbinfo = au_sbi(sb); + parent = sbinfo->si_dbgaufs; + if (!parent) + return; + + bend = au_sbend(sb); + for (; bindex <= bend; bindex++) { + snprintf(name, sizeof(name), DbgaufsXi_PREFIX "%d", bindex); + br = au_sbr(sb, bindex); + xi = &br->br_xino; + AuDebugOn(xi->xi_dbgaufs); + xi->xi_dbgaufs = debugfs_create_file(name, dbgaufs_mode, parent, + sbinfo, &dbgaufs_xino_fop); + /* ignore an error */ + if (unlikely(!xi->xi_dbgaufs)) + AuWarn1("failed %s under debugfs\n", name); + } +} + +/* ---------------------------------------------------------------------- */ + +#ifdef CONFIG_AUFS_EXPORT +static int dbgaufs_xigen_open(struct inode *inode, struct file *file) +{ + int err; + struct au_sbinfo *sbinfo; + struct super_block *sb; + + sbinfo = inode->i_private; + sb = sbinfo->si_sb; + si_noflush_read_lock(sb); + err = dbgaufs_xi_open(sbinfo->si_xigen, file, /*do_fcnt*/0); + si_read_unlock(sb); + return err; +} + +static const struct file_operations dbgaufs_xigen_fop = { + .owner = THIS_MODULE, + .open = dbgaufs_xigen_open, + .release = dbgaufs_xi_release, + .read = dbgaufs_xi_read +}; + +static int dbgaufs_xigen_init(struct au_sbinfo *sbinfo) +{ + int err; + + /* + * This function is a dynamic '__init' function actually, + * so the tiny check for si_rwsem is unnecessary. + */ + /* AuRwMustWriteLock(&sbinfo->si_rwsem); */ + + err = -EIO; + sbinfo->si_dbgaufs_xigen = debugfs_create_file + ("xigen", dbgaufs_mode, sbinfo->si_dbgaufs, sbinfo, + &dbgaufs_xigen_fop); + if (sbinfo->si_dbgaufs_xigen) + err = 0; + + return err; +} +#else +static int dbgaufs_xigen_init(struct au_sbinfo *sbinfo) +{ + return 0; +} +#endif /* CONFIG_AUFS_EXPORT */ + +/* ---------------------------------------------------------------------- */ + +void dbgaufs_si_fin(struct au_sbinfo *sbinfo) +{ + /* + * This function is a dynamic '__fin' function actually, + * so the tiny check for si_rwsem is unnecessary. + */ + /* AuRwMustWriteLock(&sbinfo->si_rwsem); */ + + debugfs_remove_recursive(sbinfo->si_dbgaufs); + sbinfo->si_dbgaufs = NULL; + kobject_put(&sbinfo->si_kobj); +} + +int dbgaufs_si_init(struct au_sbinfo *sbinfo) +{ + int err; + char name[SysaufsSiNameLen]; + + /* + * This function is a dynamic '__init' function actually, + * so the tiny check for si_rwsem is unnecessary. + */ + /* AuRwMustWriteLock(&sbinfo->si_rwsem); */ + + err = -ENOENT; + if (!dbgaufs) { + AuErr1("/debug/aufs is uninitialized\n"); + goto out; + } + + err = -EIO; + sysaufs_name(sbinfo, name); + sbinfo->si_dbgaufs = debugfs_create_dir(name, dbgaufs); + if (unlikely(!sbinfo->si_dbgaufs)) + goto out; + kobject_get(&sbinfo->si_kobj); + + sbinfo->si_dbgaufs_xib = debugfs_create_file + ("xib", dbgaufs_mode, sbinfo->si_dbgaufs, sbinfo, + &dbgaufs_xib_fop); + if (unlikely(!sbinfo->si_dbgaufs_xib)) + goto out_dir; + + sbinfo->si_dbgaufs_plink = debugfs_create_file + ("plink", dbgaufs_mode, sbinfo->si_dbgaufs, sbinfo, + &dbgaufs_plink_fop); + if (unlikely(!sbinfo->si_dbgaufs_plink)) + goto out_dir; + + err = dbgaufs_xigen_init(sbinfo); + if (!err) + goto out; /* success */ + +out_dir: + dbgaufs_si_fin(sbinfo); +out: + return err; +} + +/* ---------------------------------------------------------------------- */ + +void dbgaufs_fin(void) +{ + debugfs_remove(dbgaufs); +} + +int __init dbgaufs_init(void) +{ + int err; + + err = -EIO; + dbgaufs = debugfs_create_dir(AUFS_NAME, NULL); + if (dbgaufs) + err = 0; + return err; +} diff --git a/fs/aufs/dbgaufs.h b/fs/aufs/dbgaufs.h new file mode 100644 index 000000000..81f272e42 --- /dev/null +++ b/fs/aufs/dbgaufs.h @@ -0,0 +1,35 @@ +/* + * Copyright (C) 2005-2016 Junjiro R. Okajima + */ + +/* + * debugfs interface + */ + +#ifndef __DBGAUFS_H__ +#define __DBGAUFS_H__ + +#ifdef __KERNEL__ + +struct super_block; +struct au_sbinfo; + +#ifdef CONFIG_DEBUG_FS +/* dbgaufs.c */ +void dbgaufs_brs_del(struct super_block *sb, aufs_bindex_t bindex); +void dbgaufs_brs_add(struct super_block *sb, aufs_bindex_t bindex); +void dbgaufs_si_fin(struct au_sbinfo *sbinfo); +int dbgaufs_si_init(struct au_sbinfo *sbinfo); +void dbgaufs_fin(void); +int __init dbgaufs_init(void); +#else +AuStubVoid(dbgaufs_brs_del, struct super_block *sb, aufs_bindex_t bindex) +AuStubVoid(dbgaufs_brs_add, struct super_block *sb, aufs_bindex_t bindex) +AuStubVoid(dbgaufs_si_fin, struct au_sbinfo *sbinfo) +AuStubInt0(dbgaufs_si_init, struct au_sbinfo *sbinfo) +AuStubVoid(dbgaufs_fin, void) +AuStubInt0(__init dbgaufs_init, void) +#endif /* CONFIG_DEBUG_FS */ + +#endif /* __KERNEL__ */ +#endif /* __DBGAUFS_H__ */ diff --git a/fs/aufs/dcsub.c b/fs/aufs/dcsub.c new file mode 100644 index 000000000..e72accebb --- /dev/null +++ b/fs/aufs/dcsub.c @@ -0,0 +1,211 @@ +/* + * Copyright (C) 2005-2016 Junjiro R. Okajima + */ + +/* + * sub-routines for dentry cache + */ + +#include "aufs.h" + +static void au_dpage_free(struct au_dpage *dpage) +{ + int i; + struct dentry **p; + + p = dpage->dentries; + for (i = 0; i < dpage->ndentry; i++) + dput(*p++); + free_page((unsigned long)dpage->dentries); +} + +int au_dpages_init(struct au_dcsub_pages *dpages, gfp_t gfp) +{ + int err; + void *p; + + err = -ENOMEM; + dpages->dpages = kmalloc(sizeof(*dpages->dpages), gfp); + if (unlikely(!dpages->dpages)) + goto out; + + p = (void *)__get_free_page(gfp); + if (unlikely(!p)) + goto out_dpages; + + dpages->dpages[0].ndentry = 0; + dpages->dpages[0].dentries = p; + dpages->ndpage = 1; + return 0; /* success */ + +out_dpages: + kfree(dpages->dpages); +out: + return err; +} + +void au_dpages_free(struct au_dcsub_pages *dpages) +{ + int i; + struct au_dpage *p; + + p = dpages->dpages; + for (i = 0; i < dpages->ndpage; i++) + au_dpage_free(p++); + kfree(dpages->dpages); +} + +static int au_dpages_append(struct au_dcsub_pages *dpages, + struct dentry *dentry, gfp_t gfp) +{ + int err, sz; + struct au_dpage *dpage; + void *p; + + dpage = dpages->dpages + dpages->ndpage - 1; + sz = PAGE_SIZE / sizeof(dentry); + if (unlikely(dpage->ndentry >= sz)) { + AuLabel(new dpage); + err = -ENOMEM; + sz = dpages->ndpage * sizeof(*dpages->dpages); + p = au_kzrealloc(dpages->dpages, sz, + sz + sizeof(*dpages->dpages), gfp); + if (unlikely(!p)) + goto out; + + dpages->dpages = p; + dpage = dpages->dpages + dpages->ndpage; + p = (void *)__get_free_page(gfp); + if (unlikely(!p)) + goto out; + + dpage->ndentry = 0; + dpage->dentries = p; + dpages->ndpage++; + } + + AuDebugOn(au_dcount(dentry) <= 0); + dpage->dentries[dpage->ndentry++] = dget_dlock(dentry); + return 0; /* success */ + +out: + return err; +} + +/* todo: BAD approach */ +/* copied from linux/fs/dcache.c */ +enum d_walk_ret { + D_WALK_CONTINUE, + D_WALK_QUIT, + D_WALK_NORETRY, + D_WALK_SKIP, +}; + +extern void d_walk(struct dentry *parent, void *data, + enum d_walk_ret (*enter)(void *, struct dentry *), + void (*finish)(void *)); + +struct ac_dpages_arg { + int err; + struct au_dcsub_pages *dpages; + struct super_block *sb; + au_dpages_test test; + void *arg; +}; + +static enum d_walk_ret au_call_dpages_append(void *_arg, struct dentry *dentry) +{ + enum d_walk_ret ret; + struct ac_dpages_arg *arg = _arg; + + ret = D_WALK_CONTINUE; + if (dentry->d_sb == arg->sb + && !IS_ROOT(dentry) + && au_dcount(dentry) > 0 + && au_di(dentry) + && (!arg->test || arg->test(dentry, arg->arg))) { + arg->err = au_dpages_append(arg->dpages, dentry, GFP_ATOMIC); + if (unlikely(arg->err)) + ret = D_WALK_QUIT; + } + + return ret; +} + +int au_dcsub_pages(struct au_dcsub_pages *dpages, struct dentry *root, + au_dpages_test test, void *arg) +{ + struct ac_dpages_arg args = { + .err = 0, + .dpages = dpages, + .sb = root->d_sb, + .test = test, + .arg = arg + }; + + d_walk(root, &args, au_call_dpages_append, NULL); + + return args.err; +} + +int au_dcsub_pages_rev(struct au_dcsub_pages *dpages, struct dentry *dentry, + int do_include, au_dpages_test test, void *arg) +{ + int err; + + err = 0; + write_seqlock(&rename_lock); + spin_lock(&dentry->d_lock); + if (do_include + && au_dcount(dentry) > 0 + && (!test || test(dentry, arg))) + err = au_dpages_append(dpages, dentry, GFP_ATOMIC); + spin_unlock(&dentry->d_lock); + if (unlikely(err)) + goto out; + + /* + * RCU for vfsmount is unnecessary since this is a traverse in a single + * mount + */ + while (!IS_ROOT(dentry)) { + dentry = dentry->d_parent; /* rename_lock is locked */ + spin_lock(&dentry->d_lock); + if (au_dcount(dentry) > 0 + && (!test || test(dentry, arg))) + err = au_dpages_append(dpages, dentry, GFP_ATOMIC); + spin_unlock(&dentry->d_lock); + if (unlikely(err)) + break; + } + +out: + write_sequnlock(&rename_lock); + return err; +} + +static inline int au_dcsub_dpages_aufs(struct dentry *dentry, void *arg) +{ + return au_di(dentry) && dentry->d_sb == arg; +} + +int au_dcsub_pages_rev_aufs(struct au_dcsub_pages *dpages, + struct dentry *dentry, int do_include) +{ + return au_dcsub_pages_rev(dpages, dentry, do_include, + au_dcsub_dpages_aufs, dentry->d_sb); +} + +int au_test_subdir(struct dentry *d1, struct dentry *d2) +{ + struct path path[2] = { + { + .dentry = d1 + }, + { + .dentry = d2 + } + }; + + return path_is_under(path + 0, path + 1); +} diff --git a/fs/aufs/dcsub.h b/fs/aufs/dcsub.h new file mode 100644 index 000000000..5d2cf661d --- /dev/null +++ b/fs/aufs/dcsub.h @@ -0,0 +1,123 @@ +/* + * Copyright (C) 2005-2016 Junjiro R. Okajima + */ + +/* + * sub-routines for dentry cache + */ + +#ifndef __AUFS_DCSUB_H__ +#define __AUFS_DCSUB_H__ + +#ifdef __KERNEL__ + +#include <linux/dcache.h> +#include <linux/fs.h> + +struct au_dpage { + int ndentry; + struct dentry **dentries; +}; + +struct au_dcsub_pages { + int ndpage; + struct au_dpage *dpages; +}; + +/* ---------------------------------------------------------------------- */ + +/* dcsub.c */ +int au_dpages_init(struct au_dcsub_pages *dpages, gfp_t gfp); +void au_dpages_free(struct au_dcsub_pages *dpages); +typedef int (*au_dpages_test)(struct dentry *dentry, void *arg); +int au_dcsub_pages(struct au_dcsub_pages *dpages, struct dentry *root, + au_dpages_test test, void *arg); +int au_dcsub_pages_rev(struct au_dcsub_pages *dpages, struct dentry *dentry, + int do_include, au_dpages_test test, void *arg); +int au_dcsub_pages_rev_aufs(struct au_dcsub_pages *dpages, + struct dentry *dentry, int do_include); +int au_test_subdir(struct dentry *d1, struct dentry *d2); + +/* ---------------------------------------------------------------------- */ + +/* + * todo: in linux-3.13, several similar (but faster) helpers are added to + * include/linux/dcache.h. Try them (in the future). + */ + +static inline int au_d_hashed_positive(struct dentry *d) +{ + int err; + struct inode *inode = d_inode(d); + + err = 0; + if (unlikely(d_unhashed(d) + || d_is_negative(d) + || !inode->i_nlink)) + err = -ENOENT; + return err; +} + +static inline int au_d_linkable(struct dentry *d) +{ + int err; + struct inode *inode = d_inode(d); + + err = au_d_hashed_positive(d); + if (err + && d_is_positive(d) + && (inode->i_state & I_LINKABLE)) + err = 0; + return err; +} + +static inline int au_d_alive(struct dentry *d) +{ + int err; + struct inode *inode; + + err = 0; + if (!IS_ROOT(d)) + err = au_d_hashed_positive(d); + else { + inode = d_inode(d); + if (unlikely(d_unlinked(d) + || d_is_negative(d) + || !inode->i_nlink)) + err = -ENOENT; + } + return err; +} + +static inline int au_alive_dir(struct dentry *d) +{ + int err; + + err = au_d_alive(d); + if (unlikely(err || IS_DEADDIR(d_inode(d)))) + err = -ENOENT; + return err; +} + +static inline int au_qstreq(struct qstr *a, struct qstr *b) +{ + return a->len == b->len + && !memcmp(a->name, b->name, a->len); +} + +/* + * by the commit + * 360f547 2015-01-25 dcache: let the dentry count go down to zero without + * taking d_lock + * the type of d_lockref.count became int, but the inlined function d_count() + * still returns unsigned int. + * I don't know why. Maybe it is for every d_count() users? + * Anyway au_dcount() lives on. + */ +static inline int au_dcount(struct dentry *d) +{ + return (int)d_count(d); +} + +#endif /* __KERNEL__ */ +#endif /* __AUFS_DCSUB_H__ */ diff --git a/fs/aufs/debug.c b/fs/aufs/debug.c new file mode 100644 index 000000000..4529831a9 --- /dev/null +++ b/fs/aufs/debug.c @@ -0,0 +1,425 @@ +/* + * Copyright (C) 2005-2016 Junjiro R. Okajima + */ + +/* + * debug print functions + */ + +#include "aufs.h" + +/* Returns 0, or -errno. arg is in kp->arg. */ +static int param_atomic_t_set(const char *val, const struct kernel_param *kp) +{ + int err, n; + + err = kstrtoint(val, 0, &n); + if (!err) { + if (n > 0) + au_debug_on(); + else + au_debug_off(); + } + return err; +} + +/* Returns length written or -errno. Buffer is 4k (ie. be short!) */ +static int param_atomic_t_get(char *buffer, const struct kernel_param *kp) +{ + atomic_t *a; + + a = kp->arg; + return sprintf(buffer, "%d", atomic_read(a)); +} + +static struct kernel_param_ops param_ops_atomic_t = { + .set = param_atomic_t_set, + .get = param_atomic_t_get + /* void (*free)(void *arg) */ +}; + +atomic_t aufs_debug = ATOMIC_INIT(0); +MODULE_PARM_DESC(debug, "debug print"); +module_param_named(debug, aufs_debug, atomic_t, S_IRUGO | S_IWUSR | S_IWGRP); + +DEFINE_MUTEX(au_dbg_mtx); /* just to serialize the dbg msgs */ +char *au_plevel = KERN_DEBUG; +#define dpri(fmt, ...) do { \ + if ((au_plevel \ + && strcmp(au_plevel, KERN_DEBUG)) \ + || au_debug_test()) \ + printk("%s" fmt, au_plevel, ##__VA_ARGS__); \ +} while (0) + +/* ---------------------------------------------------------------------- */ + +void au_dpri_whlist(struct au_nhash *whlist) +{ + unsigned long ul, n; + struct hlist_head *head; + struct au_vdir_wh *pos; + + n = whlist->nh_num; + head = whlist->nh_head; + for (ul = 0; ul < n; ul++) { + hlist_for_each_entry(pos, head, wh_hash) + dpri("b%d, %.*s, %d\n", + pos->wh_bindex, + pos->wh_str.len, pos->wh_str.name, + pos->wh_str.len); + head++; + } +} + +void au_dpri_vdir(struct au_vdir *vdir) +{ + unsigned long ul; + union au_vdir_deblk_p p; + unsigned char *o; + + if (!vdir || IS_ERR(vdir)) { + dpri("err %ld\n", PTR_ERR(vdir)); + return; + } + + dpri("deblk %u, nblk %lu, deblk %p, last{%lu, %p}, ver %lu\n", + vdir->vd_deblk_sz, vdir->vd_nblk, vdir->vd_deblk, + vdir->vd_last.ul, vdir->vd_last.p.deblk, vdir->vd_version); + for (ul = 0; ul < vdir->vd_nblk; ul++) { + p.deblk = vdir->vd_deblk[ul]; + o = p.deblk; + dpri("[%lu]: %p\n", ul, o); + } +} + +static int do_pri_inode(aufs_bindex_t bindex, struct inode *inode, int hn, + struct dentry *wh) +{ + char *n = NULL; + int l = 0; + + if (!inode || IS_ERR(inode)) { + dpri("i%d: err %ld\n", bindex, PTR_ERR(inode)); + return -1; + } + + /* the type of i_blocks depends upon CONFIG_LBDAF */ + BUILD_BUG_ON(sizeof(inode->i_blocks) != sizeof(unsigned long) + && sizeof(inode->i_blocks) != sizeof(u64)); + if (wh) { + n = (void *)wh->d_name.name; + l = wh->d_name.len; + } + + dpri("i%d: %p, i%lu, %s, cnt %d, nl %u, 0%o, sz %llu, blk %llu," + " hn %d, ct %lld, np %lu, st 0x%lx, f 0x%x, v %llu, g %x%s%.*s\n", + bindex, inode, + inode->i_ino, inode->i_sb ? au_sbtype(inode->i_sb) : "??", + atomic_read(&inode->i_count), inode->i_nlink, inode->i_mode, + i_size_read(inode), (unsigned long long)inode->i_blocks, + hn, (long long)timespec_to_ns(&inode->i_ctime) & 0x0ffff, + inode->i_mapping ? inode->i_mapping->nrpages : 0, + inode->i_state, inode->i_flags, inode->i_version, + inode->i_generation, + l ? ", wh " : "", l, n); + return 0; +} + +void au_dpri_inode(struct inode *inode) +{ + struct au_iinfo *iinfo; + aufs_bindex_t bindex; + int err, hn; + + err = do_pri_inode(-1, inode, -1, NULL); + if (err || !au_test_aufs(inode->i_sb)) + return; + + iinfo = au_ii(inode); + if (!iinfo) + return; + dpri("i-1: bstart %d, bend %d, gen %d\n", + iinfo->ii_bstart, iinfo->ii_bend, au_iigen(inode, NULL)); + if (iinfo->ii_bstart < 0) + return; + hn = 0; + for (bindex = iinfo->ii_bstart; bindex <= iinfo->ii_bend; bindex++) { + hn = !!au_hn(iinfo->ii_hinode + bindex); + do_pri_inode(bindex, iinfo->ii_hinode[0 + bindex].hi_inode, hn, + iinfo->ii_hinode[0 + bindex].hi_whdentry); + } +} + +void au_dpri_dalias(struct inode *inode) +{ + struct dentry *d; + + spin_lock(&inode->i_lock); + hlist_for_each_entry(d, &inode->i_dentry, d_u.d_alias) + au_dpri_dentry(d); + spin_unlock(&inode->i_lock); +} + +static int do_pri_dentry(aufs_bindex_t bindex, struct dentry *dentry) +{ + struct dentry *wh = NULL; + int hn; + struct au_iinfo *iinfo; + + if (!dentry || IS_ERR(dentry)) { + dpri("d%d: err %ld\n", bindex, PTR_ERR(dentry)); + return -1; + } + /* do not call dget_parent() here */ + /* note: access d_xxx without d_lock */ + dpri("d%d: %p, %pd2?, %s, cnt %d, flags 0x%x, %shashed\n", + bindex, dentry, dentry, + dentry->d_sb ? au_sbtype(dentry->d_sb) : "??", + au_dcount(dentry), dentry->d_flags, + d_unhashed(dentry) ? "un" : ""); + hn = -1; + if (bindex >= 0 + && d_is_positive(dentry) + && au_test_aufs(dentry->d_sb)) { + iinfo = au_ii(d_inode(dentry)); + if (iinfo) { + hn = !!au_hn(iinfo->ii_hinode + bindex); + wh = iinfo->ii_hinode[0 + bindex].hi_whdentry; + } + } + do_pri_inode(bindex, d_inode(dentry), hn, wh); + return 0; +} + +void au_dpri_dentry(struct dentry *dentry) +{ + struct au_dinfo *dinfo; + aufs_bindex_t bindex; + int err; + struct au_hdentry *hdp; + + err = do_pri_dentry(-1, dentry); + if (err || !au_test_aufs(dentry->d_sb)) + return; + + dinfo = au_di(dentry); + if (!dinfo) + return; + dpri("d-1: bstart %d, bend %d, bwh %d, bdiropq %d, gen %d, tmp %d\n", + dinfo->di_bstart, dinfo->di_bend, + dinfo->di_bwh, dinfo->di_bdiropq, au_digen(dentry), + dinfo->di_tmpfile); + if (dinfo->di_bstart < 0) + return; + hdp = dinfo->di_hdentry; + for (bindex = dinfo->di_bstart; bindex <= dinfo->di_bend; bindex++) + do_pri_dentry(bindex, hdp[0 + bindex].hd_dentry); +} + +static int do_pri_file(aufs_bindex_t bindex, struct file *file) +{ + char a[32]; + + if (!file || IS_ERR(file)) { + dpri("f%d: err %ld\n", bindex, PTR_ERR(file)); + return -1; + } + a[0] = 0; + if (bindex < 0 + && !IS_ERR_OR_NULL(file->f_path.dentry) + && au_test_aufs(file->f_path.dentry->d_sb) + && au_fi(file)) + snprintf(a, sizeof(a), ", gen %d, mmapped %d", + au_figen(file), atomic_read(&au_fi(file)->fi_mmapped)); + dpri("f%d: mode 0x%x, flags 0%o, cnt %ld, v %llu, pos %llu%s\n", + bindex, file->f_mode, file->f_flags, (long)file_count(file), + file->f_version, file->f_pos, a); + if (!IS_ERR_OR_NULL(file->f_path.dentry)) + do_pri_dentry(bindex, file->f_path.dentry); + return 0; +} + +void au_dpri_file(struct file *file) +{ + struct au_finfo *finfo; + struct au_fidir *fidir; + struct au_hfile *hfile; + aufs_bindex_t bindex; + int err; + + err = do_pri_file(-1, file); + if (err + || IS_ERR_OR_NULL(file->f_path.dentry) + || !au_test_aufs(file->f_path.dentry->d_sb)) + return; + + finfo = au_fi(file); + if (!finfo) + return; + if (finfo->fi_btop < 0) + return; + fidir = finfo->fi_hdir; + if (!fidir) + do_pri_file(finfo->fi_btop, finfo->fi_htop.hf_file); + else + for (bindex = finfo->fi_btop; + bindex >= 0 && bindex <= fidir->fd_bbot; + bindex++) { + hfile = fidir->fd_hfile + bindex; + do_pri_file(bindex, hfile ? hfile->hf_file : NULL); + } +} + +static int do_pri_br(aufs_bindex_t bindex, struct au_branch *br) +{ + struct vfsmount *mnt; + struct super_block *sb; + + if (!br || IS_ERR(br)) + goto out; + mnt = au_br_mnt(br); + if (!mnt || IS_ERR(mnt)) + goto out; + sb = mnt->mnt_sb; + if (!sb || IS_ERR(sb)) + goto out; + + dpri("s%d: {perm 0x%x, id %d, cnt %d, wbr %p}, " + "%s, dev 0x%02x%02x, flags 0x%lx, cnt %d, active %d, " + "xino %d\n", + bindex, br->br_perm, br->br_id, atomic_read(&br->br_count), + br->br_wbr, au_sbtype(sb), MAJOR(sb->s_dev), MINOR(sb->s_dev), + sb->s_flags, sb->s_count, + atomic_read(&sb->s_active), !!br->br_xino.xi_file); + return 0; + +out: + dpri("s%d: err %ld\n", bindex, PTR_ERR(br)); + return -1; +} + +void au_dpri_sb(struct super_block *sb) +{ + struct au_sbinfo *sbinfo; + aufs_bindex_t bindex; + int err; + /* to reuduce stack size */ + struct { + struct vfsmount mnt; + struct au_branch fake; + } *a; + + /* this function can be called from magic sysrq */ + a = kzalloc(sizeof(*a), GFP_ATOMIC); + if (unlikely(!a)) { + dpri("no memory\n"); + return; + } + + a->mnt.mnt_sb = sb; + a->fake.br_path.mnt = &a->mnt; + atomic_set(&a->fake.br_count, 0); + smp_mb(); /* atomic_set */ + err = do_pri_br(-1, &a->fake); + kfree(a); + dpri("dev 0x%x\n", sb->s_dev); + if (err || !au_test_aufs(sb)) + return; + + sbinfo = au_sbi(sb); + if (!sbinfo) + return; + dpri("nw %d, gen %u, kobj %d\n", + atomic_read(&sbinfo->si_nowait.nw_len), sbinfo->si_generation, + atomic_read(&sbinfo->si_kobj.kref.refcount)); + for (bindex = 0; bindex <= sbinfo->si_bend; bindex++) + do_pri_br(bindex, sbinfo->si_branch[0 + bindex]); +} + +/* ---------------------------------------------------------------------- */ + +void __au_dbg_verify_dinode(struct dentry *dentry, const char *func, int line) +{ + struct inode *h_inode, *inode = d_inode(dentry); + struct dentry *h_dentry; + aufs_bindex_t bindex, bend, bi; + + if (!inode /* || au_di(dentry)->di_lsc == AuLsc_DI_TMP */) + return; + + bend = au_dbend(dentry); + bi = au_ibend(inode); + if (bi < bend) + bend = bi; + bindex = au_dbstart(dentry); + bi = au_ibstart(inode); + if (bi > bindex) + bindex = bi; + + for (; bindex <= bend; bindex++) { + h_dentry = au_h_dptr(dentry, bindex); + if (!h_dentry) + continue; + h_inode = au_h_iptr(inode, bindex); + if (unlikely(h_inode != d_inode(h_dentry))) { + au_debug_on(); + AuDbg("b%d, %s:%d\n", bindex, func, line); + AuDbgDentry(dentry); + AuDbgInode(inode); + au_debug_off(); + BUG(); + } + } +} + +void au_dbg_verify_gen(struct dentry *parent, unsigned int sigen) +{ + int err, i, j; + struct au_dcsub_pages dpages; + struct au_dpage *dpage; + struct dentry **dentries; + + err = au_dpages_init(&dpages, GFP_NOFS); + AuDebugOn(err); + err = au_dcsub_pages_rev_aufs(&dpages, parent, /*do_include*/1); + AuDebugOn(err); + for (i = dpages.ndpage - 1; !err && i >= 0; i--) { + dpage = dpages.dpages + i; + dentries = dpage->dentries; + for (j = dpage->ndentry - 1; !err && j >= 0; j--) + AuDebugOn(au_digen_test(dentries[j], sigen)); + } + au_dpages_free(&dpages); +} + +void au_dbg_verify_kthread(void) +{ + if (au_wkq_test()) { + au_dbg_blocked(); + /* + * It may be recursive, but udba=notify between two aufs mounts, + * where a single ro branch is shared, is not a problem. + */ + /* WARN_ON(1); */ + } +} + +/* ---------------------------------------------------------------------- */ + +int __init au_debug_init(void) +{ + aufs_bindex_t bindex; + struct au_vdir_destr destr; + + bindex = -1; + AuDebugOn(bindex >= 0); + + destr.len = -1; + AuDebugOn(destr.len < NAME_MAX); + +#ifdef CONFIG_4KSTACKS + pr_warn("CONFIG_4KSTACKS is defined.\n"); +#endif + + return 0; +} diff --git a/fs/aufs/debug.h b/fs/aufs/debug.h new file mode 100644 index 000000000..0567f31d0 --- /dev/null +++ b/fs/aufs/debug.h @@ -0,0 +1,212 @@ +/* + * Copyright (C) 2005-2016 Junjiro R. Okajima + */ + +/* + * debug print functions + */ + +#ifndef __AUFS_DEBUG_H__ +#define __AUFS_DEBUG_H__ + +#ifdef __KERNEL__ + +#include <linux/atomic.h> +#include <linux/module.h> +#include <linux/kallsyms.h> +#include <linux/sysrq.h> + +#ifdef CONFIG_AUFS_DEBUG +#define AuDebugOn(a) BUG_ON(a) + +/* module parameter */ +extern atomic_t aufs_debug; +static inline void au_debug_on(void) +{ + atomic_inc(&aufs_debug); +} +static inline void au_debug_off(void) +{ + atomic_dec_if_positive(&aufs_debug); +} + +static inline int au_debug_test(void) +{ + return atomic_read(&aufs_debug) > 0; +} +#else +#define AuDebugOn(a) do {} while (0) +AuStubVoid(au_debug_on, void) +AuStubVoid(au_debug_off, void) +AuStubInt0(au_debug_test, void) +#endif /* CONFIG_AUFS_DEBUG */ + +#define param_check_atomic_t(name, p) __param_check(name, p, atomic_t) + +/* ---------------------------------------------------------------------- */ + +/* debug print */ + +#define AuDbg(fmt, ...) do { \ + if (au_debug_test()) \ + pr_debug("DEBUG: " fmt, ##__VA_ARGS__); \ +} while (0) +#define AuLabel(l) AuDbg(#l "\n") +#define AuIOErr(fmt, ...) pr_err("I/O Error, " fmt, ##__VA_ARGS__) +#define AuWarn1(fmt, ...) do { \ + static unsigned char _c; \ + if (!_c++) \ + pr_warn(fmt, ##__VA_ARGS__); \ +} while (0) + +#define AuErr1(fmt, ...) do { \ + static unsigned char _c; \ + if (!_c++) \ + pr_err(fmt, ##__VA_ARGS__); \ +} while (0) + +#define AuIOErr1(fmt, ...) do { \ + static unsigned char _c; \ + if (!_c++) \ + AuIOErr(fmt, ##__VA_ARGS__); \ +} while (0) + +#define AuUnsupportMsg "This operation is not supported." \ + " Please report this application to aufs-users ML." +#define AuUnsupport(fmt, ...) do { \ + pr_err(AuUnsupportMsg "\n" fmt, ##__VA_ARGS__); \ + dump_stack(); \ +} while (0) + +#define AuTraceErr(e) do { \ + if (unlikely((e) < 0)) \ + AuDbg("err %d\n", (int)(e)); \ +} while (0) + +#define AuTraceErrPtr(p) do { \ + if (IS_ERR(p)) \ + AuDbg("err %ld\n", PTR_ERR(p)); \ +} while (0) + +/* dirty macros for debug print, use with "%.*s" and caution */ +#define AuLNPair(qstr) (qstr)->len, (qstr)->name + +/* ---------------------------------------------------------------------- */ + +struct dentry; +#ifdef CONFIG_AUFS_DEBUG +extern struct mutex au_dbg_mtx; +extern char *au_plevel; +struct au_nhash; +void au_dpri_whlist(struct au_nhash *whlist); +struct au_vdir; +void au_dpri_vdir(struct au_vdir *vdir); +struct inode; +void au_dpri_inode(struct inode *inode); +void au_dpri_dalias(struct inode *inode); +void au_dpri_dentry(struct dentry *dentry); +struct file; +void au_dpri_file(struct file *filp); +struct super_block; +void au_dpri_sb(struct super_block *sb); + +#define au_dbg_verify_dinode(d) __au_dbg_verify_dinode(d, __func__, __LINE__) +void __au_dbg_verify_dinode(struct dentry *dentry, const char *func, int line); +void au_dbg_verify_gen(struct dentry *parent, unsigned int sigen); +void au_dbg_verify_kthread(void); + +int __init au_debug_init(void); + +#define AuDbgWhlist(w) do { \ + mutex_lock(&au_dbg_mtx); \ + AuDbg(#w "\n"); \ + au_dpri_whlist(w); \ + mutex_unlock(&au_dbg_mtx); \ +} while (0) + +#define AuDbgVdir(v) do { \ + mutex_lock(&au_dbg_mtx); \ + AuDbg(#v "\n"); \ + au_dpri_vdir(v); \ + mutex_unlock(&au_dbg_mtx); \ +} while (0) + +#define AuDbgInode(i) do { \ + mutex_lock(&au_dbg_mtx); \ + AuDbg(#i "\n"); \ + au_dpri_inode(i); \ + mutex_unlock(&au_dbg_mtx); \ +} while (0) + +#define AuDbgDAlias(i) do { \ + mutex_lock(&au_dbg_mtx); \ + AuDbg(#i "\n"); \ + au_dpri_dalias(i); \ + mutex_unlock(&au_dbg_mtx); \ +} while (0) + +#define AuDbgDentry(d) do { \ + mutex_lock(&au_dbg_mtx); \ + AuDbg(#d "\n"); \ + au_dpri_dentry(d); \ + mutex_unlock(&au_dbg_mtx); \ +} while (0) + +#define AuDbgFile(f) do { \ + mutex_lock(&au_dbg_mtx); \ + AuDbg(#f "\n"); \ + au_dpri_file(f); \ + mutex_unlock(&au_dbg_mtx); \ +} while (0) + +#define AuDbgSb(sb) do { \ + mutex_lock(&au_dbg_mtx); \ + AuDbg(#sb "\n"); \ + au_dpri_sb(sb); \ + mutex_unlock(&au_dbg_mtx); \ +} while (0) + +#define AuDbgSym(addr) do { \ + char sym[KSYM_SYMBOL_LEN]; \ + sprint_symbol(sym, (unsigned long)addr); \ + AuDbg("%s\n", sym); \ +} while (0) +#else +AuStubVoid(au_dbg_verify_dinode, struct dentry *dentry) +AuStubVoid(au_dbg_verify_gen, struct dentry *parent, unsigned int sigen) +AuStubVoid(au_dbg_verify_kthread, void) +AuStubInt0(__init au_debug_init, void) + +#define AuDbgWhlist(w) do {} while (0) +#define AuDbgVdir(v) do {} while (0) +#define AuDbgInode(i) do {} while (0) +#define AuDbgDAlias(i) do {} while (0) +#define AuDbgDentry(d) do {} while (0) +#define AuDbgFile(f) do {} while (0) +#define AuDbgSb(sb) do {} while (0) +#define AuDbgSym(addr) do {} while (0) +#endif /* CONFIG_AUFS_DEBUG */ + +/* ---------------------------------------------------------------------- */ + +#ifdef CONFIG_AUFS_MAGIC_SYSRQ +int __init au_sysrq_init(void); +void au_sysrq_fin(void); + +#ifdef CONFIG_HW_CONSOLE +#define au_dbg_blocked() do { \ + WARN_ON(1); \ + handle_sysrq('w'); \ +} while (0) +#else +AuStubVoid(au_dbg_blocked, void) +#endif + +#else +AuStubInt0(__init au_sysrq_init, void) +AuStubVoid(au_sysrq_fin, void) +AuStubVoid(au_dbg_blocked, void) +#endif /* CONFIG_AUFS_MAGIC_SYSRQ */ + +#endif /* __KERNEL__ */ +#endif /* __AUFS_DEBUG_H__ */ diff --git a/fs/aufs/dentry.c b/fs/aufs/dentry.c new file mode 100644 index 000000000..e65fadd05 --- /dev/null +++ b/fs/aufs/dentry.c @@ -0,0 +1,1123 @@ +/* + * Copyright (C) 2005-2016 Junjiro R. Okajima + */ + +/* + * lookup and dentry operations + */ + +#include <linux/namei.h> +#include "aufs.h" + +#define AuLkup_ALLOW_NEG 1 +#define AuLkup_IGNORE_PERM (1 << 1) +#define au_ftest_lkup(flags, name) ((flags) & AuLkup_##name) +#define au_fset_lkup(flags, name) \ + do { (flags) |= AuLkup_##name; } while (0) +#define au_fclr_lkup(flags, name) \ + do { (flags) &= ~AuLkup_##name; } while (0) + +struct au_do_lookup_args { + unsigned int flags; + mode_t type; +}; + +/* + * returns positive/negative dentry, NULL or an error. + * NULL means whiteout-ed or not-found. + */ +static struct dentry* +au_do_lookup(struct dentry *h_parent, struct dentry *dentry, + aufs_bindex_t bindex, struct qstr *wh_name, + struct au_do_lookup_args *args) +{ + struct dentry *h_dentry; + struct inode *h_inode; + struct au_branch *br; + int wh_found, opq; + unsigned char wh_able; + const unsigned char allow_neg = !!au_ftest_lkup(args->flags, ALLOW_NEG); + const unsigned char ignore_perm = !!au_ftest_lkup(args->flags, + IGNORE_PERM); + + wh_found = 0; + br = au_sbr(dentry->d_sb, bindex); + wh_able = !!au_br_whable(br->br_perm); + if (wh_able) + wh_found = au_wh_test(h_parent, wh_name, /*try_sio*/0); + h_dentry = ERR_PTR(wh_found); + if (!wh_found) + goto real_lookup; + if (unlikely(wh_found < 0)) + goto out; + + /* We found a whiteout */ + /* au_set_dbend(dentry, bindex); */ + au_set_dbwh(dentry, bindex); + if (!allow_neg) + return NULL; /* success */ + +real_lookup: + if (!ignore_perm) + h_dentry = vfsub_lkup_one(&dentry->d_name, h_parent); + else + h_dentry = au_sio_lkup_one(&dentry->d_name, h_parent); + if (IS_ERR(h_dentry)) { + if (PTR_ERR(h_dentry) == -ENAMETOOLONG + && !allow_neg) + h_dentry = NULL; + goto out; + } + + h_inode = d_inode(h_dentry); + if (d_is_negative(h_dentry)) { + if (!allow_neg) + goto out_neg; + } else if (wh_found + || (args->type && args->type != (h_inode->i_mode & S_IFMT))) + goto out_neg; + + if (au_dbend(dentry) <= bindex) + au_set_dbend(dentry, bindex); + if (au_dbstart(dentry) < 0 || bindex < au_dbstart(dentry)) + au_set_dbstart(dentry, bindex); + au_set_h_dptr(dentry, bindex, h_dentry); + + if (!d_is_dir(h_dentry) + || !wh_able + || (d_really_is_positive(dentry) && !d_is_dir(dentry))) + goto out; /* success */ + + inode_lock_nested(h_inode, AuLsc_I_CHILD); + opq = au_diropq_test(h_dentry); + inode_unlock(h_inode); + if (opq > 0) + au_set_dbdiropq(dentry, bindex); + else if (unlikely(opq < 0)) { + au_set_h_dptr(dentry, bindex, NULL); + h_dentry = ERR_PTR(opq); + } + goto out; + +out_neg: + dput(h_dentry); + h_dentry = NULL; +out: + return h_dentry; +} + +static int au_test_shwh(struct super_block *sb, const struct qstr *name) +{ + if (unlikely(!au_opt_test(au_mntflags(sb), SHWH) + && !strncmp(name->name, AUFS_WH_PFX, AUFS_WH_PFX_LEN))) + return -EPERM; + return 0; +} + +/* + * returns the number of lower positive dentries, + * otherwise an error. + * can be called at unlinking with @type is zero. + */ +int au_lkup_dentry(struct dentry *dentry, aufs_bindex_t bstart, mode_t type) +{ + int npositive, err; + aufs_bindex_t bindex, btail, bdiropq; + unsigned char isdir, dirperm1; + struct qstr whname; + struct au_do_lookup_args args = { + .flags = 0, + .type = type + }; + const struct qstr *name = &dentry->d_name; + struct dentry *parent; + struct super_block *sb; + + sb = dentry->d_sb; + err = au_test_shwh(sb, name); + if (unlikely(err)) + goto out; + + err = au_wh_name_alloc(&whname, name); + if (unlikely(err)) + goto out; + + isdir = !!d_is_dir(dentry); + if (!type) + au_fset_lkup(args.flags, ALLOW_NEG); + dirperm1 = !!au_opt_test(au_mntflags(sb), DIRPERM1); + + npositive = 0; + parent = dget_parent(dentry); + btail = au_dbtaildir(parent); + for (bindex = bstart; bindex <= btail; bindex++) { + struct dentry *h_parent, *h_dentry; + struct inode *h_inode, *h_dir; + + h_dentry = au_h_dptr(dentry, bindex); + if (h_dentry) { + if (d_is_positive(h_dentry)) + npositive++; + if (type != S_IFDIR) + break; + continue; + } + h_parent = au_h_dptr(parent, bindex); + if (!h_parent || !d_is_dir(h_parent)) + continue; + + h_dir = d_inode(h_parent); + inode_lock_nested(h_dir, AuLsc_I_PARENT); + h_dentry = au_do_lookup(h_parent, dentry, bindex, &whname, + &args); + inode_unlock(h_dir); + err = PTR_ERR(h_dentry); + if (IS_ERR(h_dentry)) + goto out_parent; + if (h_dentry) + au_fclr_lkup(args.flags, ALLOW_NEG); + if (dirperm1) + au_fset_lkup(args.flags, IGNORE_PERM); + + if (au_dbwh(dentry) == bindex) + break; + if (!h_dentry) + continue; + if (d_is_negative(h_dentry)) + continue; + h_inode = d_inode(h_dentry); + npositive++; + if (!args.type) + args.type = h_inode->i_mode & S_IFMT; + if (args.type != S_IFDIR) + break; + else if (isdir) { + /* the type of lower may be different */ + bdiropq = au_dbdiropq(dentry); + if (bdiropq >= 0 && bdiropq <= bindex) + break; + } + } + + if (npositive) { + AuLabel(positive); + au_update_dbstart(dentry); + } + err = npositive; + if (unlikely(!au_opt_test(au_mntflags(sb), UDBA_NONE) + && au_dbstart(dentry) < 0)) { + err = -EIO; + AuIOErr("both of real entry and whiteout found, %pd, err %d\n", + dentry, err); + } + +out_parent: + dput(parent); + kfree(whname.name); +out: + return err; +} + +struct dentry *au_sio_lkup_one(struct qstr *name, struct dentry *parent) +{ + struct dentry *dentry; + int wkq_err; + + if (!au_test_h_perm_sio(d_inode(parent), MAY_EXEC)) + dentry = vfsub_lkup_one(name, parent); + else { + struct vfsub_lkup_one_args args = { + .errp = &dentry, + .name = name, + .parent = parent + }; + + wkq_err = au_wkq_wait(vfsub_call_lkup_one, &args); + if (unlikely(wkq_err)) + dentry = ERR_PTR(wkq_err); + } + + return dentry; +} + +/* + * lookup @dentry on @bindex which should be negative. + */ +int au_lkup_neg(struct dentry *dentry, aufs_bindex_t bindex, int wh) +{ + int err; + struct dentry *parent, *h_parent, *h_dentry; + struct au_branch *br; + + parent = dget_parent(dentry); + h_parent = au_h_dptr(parent, bindex); + br = au_sbr(dentry->d_sb, bindex); + if (wh) + h_dentry = au_whtmp_lkup(h_parent, br, &dentry->d_name); + else + h_dentry = au_sio_lkup_one(&dentry->d_name, h_parent); + err = PTR_ERR(h_dentry); + if (IS_ERR(h_dentry)) + goto out; + if (unlikely(d_is_positive(h_dentry))) { + err = -EIO; + AuIOErr("%pd should be negative on b%d.\n", h_dentry, bindex); + dput(h_dentry); + goto out; + } + + err = 0; + if (bindex < au_dbstart(dentry)) + au_set_dbstart(dentry, bindex); + if (au_dbend(dentry) < bindex) + au_set_dbend(dentry, bindex); + au_set_h_dptr(dentry, bindex, h_dentry); + +out: + dput(parent); + return err; +} + +/* ---------------------------------------------------------------------- */ + +/* subset of struct inode */ +struct au_iattr { + unsigned long i_ino; + /* unsigned int i_nlink; */ + kuid_t i_uid; + kgid_t i_gid; + u64 i_version; +/* + loff_t i_size; + blkcnt_t i_blocks; +*/ + umode_t i_mode; +}; + +static void au_iattr_save(struct au_iattr *ia, struct inode *h_inode) +{ + ia->i_ino = h_inode->i_ino; + /* ia->i_nlink = h_inode->i_nlink; */ + ia->i_uid = h_inode->i_uid; + ia->i_gid = h_inode->i_gid; + ia->i_version = h_inode->i_version; +/* + ia->i_size = h_inode->i_size; + ia->i_blocks = h_inode->i_blocks; +*/ + ia->i_mode = (h_inode->i_mode & S_IFMT); +} + +static int au_iattr_test(struct au_iattr *ia, struct inode *h_inode) +{ + return ia->i_ino != h_inode->i_ino + /* || ia->i_nlink != h_inode->i_nlink */ + || !uid_eq(ia->i_uid, h_inode->i_uid) + || !gid_eq(ia->i_gid, h_inode->i_gid) + || ia->i_version != h_inode->i_version +/* + || ia->i_size != h_inode->i_size + || ia->i_blocks != h_inode->i_blocks +*/ + || ia->i_mode != (h_inode->i_mode & S_IFMT); +} + +static int au_h_verify_dentry(struct dentry *h_dentry, struct dentry *h_parent, + struct au_branch *br) +{ + int err; + struct au_iattr ia; + struct inode *h_inode; + struct dentry *h_d; + struct super_block *h_sb; + + err = 0; + memset(&ia, -1, sizeof(ia)); + h_sb = h_dentry->d_sb; + h_inode = NULL; + if (d_is_positive(h_dentry)) { + h_inode = d_inode(h_dentry); + au_iattr_save(&ia, h_inode); + } else if (au_test_nfs(h_sb) || au_test_fuse(h_sb)) + /* nfs d_revalidate may return 0 for negative dentry */ + /* fuse d_revalidate always return 0 for negative dentry */ + goto out; + + /* main purpose is namei.c:cached_lookup() and d_revalidate */ + h_d = vfsub_lkup_one(&h_dentry->d_name, h_parent); + err = PTR_ERR(h_d); + if (IS_ERR(h_d)) + goto out; + + err = 0; + if (unlikely(h_d != h_dentry + || d_inode(h_d) != h_inode + || (h_inode && au_iattr_test(&ia, h_inode)))) + err = au_busy_or_stale(); + dput(h_d); + +out: + AuTraceErr(err); + return err; +} + +int au_h_verify(struct dentry *h_dentry, unsigned int udba, struct inode *h_dir, + struct dentry *h_parent, struct au_branch *br) +{ + int err; + + err = 0; + if (udba == AuOpt_UDBA_REVAL + && !au_test_fs_remote(h_dentry->d_sb)) { + IMustLock(h_dir); + err = (d_inode(h_dentry->d_parent) != h_dir); + } else if (udba != AuOpt_UDBA_NONE) + err = au_h_verify_dentry(h_dentry, h_parent, br); + + return err; +} + +/* ---------------------------------------------------------------------- */ + +static int au_do_refresh_hdentry(struct dentry *dentry, struct dentry *parent) +{ + int err; + aufs_bindex_t new_bindex, bindex, bend, bwh, bdiropq; + struct au_hdentry tmp, *p, *q; + struct au_dinfo *dinfo; + struct super_block *sb; + + DiMustWriteLock(dentry); + + sb = dentry->d_sb; + dinfo = au_di(dentry); + bend = dinfo->di_bend; + bwh = dinfo->di_bwh; + bdiropq = dinfo->di_bdiropq; + p = dinfo->di_hdentry + dinfo->di_bstart; + for (bindex = dinfo->di_bstart; bindex <= bend; bindex++, p++) { + if (!p->hd_dentry) + continue; + + new_bindex = au_br_index(sb, p->hd_id); + if (new_bindex == bindex) + continue; + + if (dinfo->di_bwh == bindex) + bwh = new_bindex; + if (dinfo->di_bdiropq == bindex) + bdiropq = new_bindex; + if (new_bindex < 0) { + au_hdput(p); + p->hd_dentry = NULL; + continue; + } + + /* swap two lower dentries, and loop again */ + q = dinfo->di_hdentry + new_bindex; + tmp = *q; + *q = *p; + *p = tmp; + if (tmp.hd_dentry) { + bindex--; + p--; + } + } + + dinfo->di_bwh = -1; + if (bwh >= 0 && bwh <= au_sbend(sb) && au_sbr_whable(sb, bwh)) + dinfo->di_bwh = bwh; + + dinfo->di_bdiropq = -1; + if (bdiropq >= 0 + && bdiropq <= au_sbend(sb) + && au_sbr_whable(sb, bdiropq)) + dinfo->di_bdiropq = bdiropq; + + err = -EIO; + dinfo->di_bstart = -1; + dinfo->di_bend = -1; + bend = au_dbend(parent); + p = dinfo->di_hdentry; + for (bindex = 0; bindex <= bend; bindex++, p++) + if (p->hd_dentry) { + dinfo->di_bstart = bindex; + break; + } + + if (dinfo->di_bstart >= 0) { + p = dinfo->di_hdentry + bend; + for (bindex = bend; bindex >= 0; bindex--, p--) + if (p->hd_dentry) { + dinfo->di_bend = bindex; + err = 0; + break; + } + } + + return err; +} + +static void au_do_hide(struct dentry *dentry) +{ + struct inode *inode; + + if (d_really_is_positive(dentry)) { + inode = d_inode(dentry); + if (!d_is_dir(dentry)) { + if (inode->i_nlink && !d_unhashed(dentry)) + drop_nlink(inode); + } else { + clear_nlink(inode); + /* stop next lookup */ + inode->i_flags |= S_DEAD; + } + smp_mb(); /* necessary? */ + } + d_drop(dentry); +} + +static int au_hide_children(struct dentry *parent) +{ + int err, i, j, ndentry; + struct au_dcsub_pages dpages; + struct au_dpage *dpage; + struct dentry *dentry; + + err = au_dpages_init(&dpages, GFP_NOFS); + if (unlikely(err)) + goto out; + err = au_dcsub_pages(&dpages, parent, NULL, NULL); + if (unlikely(err)) + goto out_dpages; + + /* in reverse order */ + for (i = dpages.ndpage - 1; i >= 0; i--) { + dpage = dpages.dpages + i; + ndentry = dpage->ndentry; + for (j = ndentry - 1; j >= 0; j--) { + dentry = dpage->dentries[j]; + if (dentry != parent) + au_do_hide(dentry); + } + } + +out_dpages: + au_dpages_free(&dpages); +out: + return err; +} + +static void au_hide(struct dentry *dentry) +{ + int err; + + AuDbgDentry(dentry); + if (d_is_dir(dentry)) { + /* shrink_dcache_parent(dentry); */ + err = au_hide_children(dentry); + if (unlikely(err)) + AuIOErr("%pd, failed hiding children, ignored %d\n", + dentry, err); + } + au_do_hide(dentry); +} + +/* + * By adding a dirty branch, a cached dentry may be affected in various ways. + * + * a dirty branch is added + * - on the top of layers + * - in the middle of layers + * - to the bottom of layers + * + * on the added branch there exists + * - a whiteout + * - a diropq + * - a same named entry + * + exist + * * negative --> positive + * * positive --> positive + * - type is unchanged + * - type is changed + * + doesn't exist + * * negative --> negative + * * positive --> negative (rejected by au_br_del() for non-dir case) + * - none + */ +static int au_refresh_by_dinfo(struct dentry *dentry, struct au_dinfo *dinfo, + struct au_dinfo *tmp) +{ + int err; + aufs_bindex_t bindex, bend; + struct { + struct dentry *dentry; + struct inode *inode; + mode_t mode; + } orig_h, tmp_h = { + .dentry = NULL + }; + struct au_hdentry *hd; + struct inode *inode, *h_inode; + struct dentry *h_dentry; + + err = 0; + AuDebugOn(dinfo->di_bstart < 0); + orig_h.mode = 0; + orig_h.dentry = dinfo->di_hdentry[dinfo->di_bstart].hd_dentry; + orig_h.inode = NULL; + if (d_is_positive(orig_h.dentry)) { + orig_h.inode = d_inode(orig_h.dentry); + orig_h.mode = orig_h.inode->i_mode & S_IFMT; + } + if (tmp->di_bstart >= 0) { + tmp_h.dentry = tmp->di_hdentry[tmp->di_bstart].hd_dentry; + if (d_is_positive(tmp_h.dentry)) { + tmp_h.inode = d_inode(tmp_h.dentry); + tmp_h.mode = tmp_h.inode->i_mode & S_IFMT; + } + } + + inode = NULL; + if (d_really_is_positive(dentry)) + inode = d_inode(dentry); + if (!orig_h.inode) { + AuDbg("nagative originally\n"); + if (inode) { + au_hide(dentry); + goto out; + } + AuDebugOn(inode); + AuDebugOn(dinfo->di_bstart != dinfo->di_bend); + AuDebugOn(dinfo->di_bdiropq != -1); + + if (!tmp_h.inode) { + AuDbg("negative --> negative\n"); + /* should have only one negative lower */ + if (tmp->di_bstart >= 0 + && tmp->di_bstart < dinfo->di_bstart) { + AuDebugOn(tmp->di_bstart != tmp->di_bend); + AuDebugOn(dinfo->di_bstart != dinfo->di_bend); + au_set_h_dptr(dentry, dinfo->di_bstart, NULL); + au_di_cp(dinfo, tmp); + hd = tmp->di_hdentry + tmp->di_bstart; + au_set_h_dptr(dentry, tmp->di_bstart, + dget(hd->hd_dentry)); + } + au_dbg_verify_dinode(dentry); + } else { + AuDbg("negative --> positive\n"); + /* + * similar to the behaviour of creating with bypassing + * aufs. + * unhash it in order to force an error in the + * succeeding create operation. + * we should not set S_DEAD here. + */ + d_drop(dentry); + /* au_di_swap(tmp, dinfo); */ + au_dbg_verify_dinode(dentry); + } + } else { + AuDbg("positive originally\n"); + /* inode may be NULL */ + AuDebugOn(inode && (inode->i_mode & S_IFMT) != orig_h.mode); + if (!tmp_h.inode) { + AuDbg("positive --> negative\n"); + /* or bypassing aufs */ + au_hide(dentry); + if (tmp->di_bwh >= 0 && tmp->di_bwh <= dinfo->di_bstart) + dinfo->di_bwh = tmp->di_bwh; + if (inode) + err = au_refresh_hinode_self(inode); + au_dbg_verify_dinode(dentry); + } else if (orig_h.mode == tmp_h.mode) { + AuDbg("positive --> positive, same type\n"); + if (!S_ISDIR(orig_h.mode) + && dinfo->di_bstart > tmp->di_bstart) { + /* + * similar to the behaviour of removing and + * creating. + */ + au_hide(dentry); + if (inode) + err = au_refresh_hinode_self(inode); + au_dbg_verify_dinode(dentry); + } else { + /* fill empty slots */ + if (dinfo->di_bstart > tmp->di_bstart) + dinfo->di_bstart = tmp->di_bstart; + if (dinfo->di_bend < tmp->di_bend) + dinfo->di_bend = tmp->di_bend; + dinfo->di_bwh = tmp->di_bwh; + dinfo->di_bdiropq = tmp->di_bdiropq; + hd = tmp->di_hdentry; + bend = dinfo->di_bend; + for (bindex = tmp->di_bstart; bindex <= bend; + bindex++) { + if (au_h_dptr(dentry, bindex)) + continue; + h_dentry = hd[bindex].hd_dentry; + if (!h_dentry) + continue; + AuDebugOn(d_is_negative(h_dentry)); + h_inode = d_inode(h_dentry); + AuDebugOn(orig_h.mode + != (h_inode->i_mode + & S_IFMT)); + au_set_h_dptr(dentry, bindex, + dget(h_dentry)); + } + err = au_refresh_hinode(inode, dentry); + au_dbg_verify_dinode(dentry); + } + } else { + AuDbg("positive --> positive, different type\n"); + /* similar to the behaviour of removing and creating */ + au_hide(dentry); + if (inode) + err = au_refresh_hinode_self(inode); + au_dbg_verify_dinode(dentry); + } + } + +out: + return err; +} + +void au_refresh_dop(struct dentry *dentry, int force_reval) +{ + const struct dentry_operations *dop + = force_reval ? &aufs_dop : dentry->d_sb->s_d_op; + static const unsigned int mask + = DCACHE_OP_REVALIDATE | DCACHE_OP_WEAK_REVALIDATE; + + BUILD_BUG_ON(sizeof(mask) != sizeof(dentry->d_flags)); + + if (dentry->d_op == dop) + return; + + AuDbg("%pd\n", dentry); + spin_lock(&dentry->d_lock); + if (dop == &aufs_dop) + dentry->d_flags |= mask; + else + dentry->d_flags &= ~mask; + dentry->d_op = dop; + spin_unlock(&dentry->d_lock); +} + +int au_refresh_dentry(struct dentry *dentry, struct dentry *parent) +{ + int err, ebrange; + unsigned int sigen; + struct au_dinfo *dinfo, *tmp; + struct super_block *sb; + struct inode *inode; + + DiMustWriteLock(dentry); + AuDebugOn(IS_ROOT(dentry)); + AuDebugOn(d_really_is_negative(parent)); + + sb = dentry->d_sb; + sigen = au_sigen(sb); + err = au_digen_test(parent, sigen); + if (unlikely(err)) + goto out; + + dinfo = au_di(dentry); + err = au_di_realloc(dinfo, au_sbend(sb) + 1); + if (unlikely(err)) + goto out; + ebrange = au_dbrange_test(dentry); + if (!ebrange) + ebrange = au_do_refresh_hdentry(dentry, parent); + + if (d_unhashed(dentry) || ebrange /* || dinfo->di_tmpfile */) { + AuDebugOn(au_dbstart(dentry) < 0 && au_dbend(dentry) >= 0); + if (d_really_is_positive(dentry)) { + inode = d_inode(dentry); + err = au_refresh_hinode_self(inode); + } + au_dbg_verify_dinode(dentry); + if (!err) + goto out_dgen; /* success */ + goto out; + } + + /* temporary dinfo */ + AuDbgDentry(dentry); + err = -ENOMEM; + tmp = au_di_alloc(sb, AuLsc_DI_TMP); + if (unlikely(!tmp)) + goto out; + au_di_swap(tmp, dinfo); + /* returns the number of positive dentries */ + /* + * if current working dir is removed, it returns an error. + * but the dentry is legal. + */ + err = au_lkup_dentry(dentry, /*bstart*/0, /*type*/0); + AuDbgDentry(dentry); + au_di_swap(tmp, dinfo); + if (err == -ENOENT) + err = 0; + if (err >= 0) { + /* compare/refresh by dinfo */ + AuDbgDentry(dentry); + err = au_refresh_by_dinfo(dentry, dinfo, tmp); + au_dbg_verify_dinode(dentry); + AuTraceErr(err); + } + au_rw_write_unlock(&tmp->di_rwsem); + au_di_free(tmp); + if (unlikely(err)) + goto out; + +out_dgen: + au_update_digen(dentry); +out: + if (unlikely(err && !(dentry->d_flags & DCACHE_NFSFS_RENAMED))) { + AuIOErr("failed refreshing %pd, %d\n", dentry, err); + AuDbgDentry(dentry); + } + AuTraceErr(err); + return err; +} + +static int au_do_h_d_reval(struct dentry *h_dentry, unsigned int flags, + struct dentry *dentry, aufs_bindex_t bindex) +{ + int err, valid; + + err = 0; + if (!(h_dentry->d_flags & DCACHE_OP_REVALIDATE)) + goto out; + + AuDbg("b%d\n", bindex); + /* + * gave up supporting LOOKUP_CREATE/OPEN for lower fs, + * due to whiteout and branch permission. + */ + flags &= ~(/*LOOKUP_PARENT |*/ LOOKUP_OPEN | LOOKUP_CREATE + | LOOKUP_FOLLOW | LOOKUP_EXCL); + /* it may return tri-state */ + valid = h_dentry->d_op->d_revalidate(h_dentry, flags); + + if (unlikely(valid < 0)) + err = valid; + else if (!valid) + err = -EINVAL; + +out: + AuTraceErr(err); + return err; +} + +/* todo: remove this */ +static int h_d_revalidate(struct dentry *dentry, struct inode *inode, + unsigned int flags, int do_udba) +{ + int err; + umode_t mode, h_mode; + aufs_bindex_t bindex, btail, bstart, ibs, ibe; + unsigned char plus, unhashed, is_root, h_plus, h_nfs, tmpfile; + struct inode *h_inode, *h_cached_inode; + struct dentry *h_dentry; + struct qstr *name, *h_name; + + err = 0; + plus = 0; + mode = 0; + ibs = -1; + ibe = -1; + unhashed = !!d_unhashed(dentry); + is_root = !!IS_ROOT(dentry); + name = &dentry->d_name; + tmpfile = au_di(dentry)->di_tmpfile; + + /* + * Theoretically, REVAL test should be unnecessary in case of + * {FS,I}NOTIFY. + * But {fs,i}notify doesn't fire some necessary events, + * IN_ATTRIB for atime/nlink/pageio + * Let's do REVAL test too. + */ + if (do_udba && inode) { + mode = (inode->i_mode & S_IFMT); + plus = (inode->i_nlink > 0); + ibs = au_ibstart(inode); + ibe = au_ibend(inode); + } + + bstart = au_dbstart(dentry); + btail = bstart; + if (inode && S_ISDIR(inode->i_mode)) + btail = au_dbtaildir(dentry); + for (bindex = bstart; bindex <= btail; bindex++) { + h_dentry = au_h_dptr(dentry, bindex); + if (!h_dentry) + continue; + + AuDbg("b%d, %pd\n", bindex, h_dentry); + h_nfs = !!au_test_nfs(h_dentry->d_sb); + spin_lock(&h_dentry->d_lock); + h_name = &h_dentry->d_name; + if (unlikely(do_udba + && !is_root + && ((!h_nfs + && (unhashed != !!d_unhashed(h_dentry) + || (!tmpfile + && !au_qstreq(name, h_name)) + )) + || (h_nfs + && !(flags & LOOKUP_OPEN) + && (h_dentry->d_flags + & DCACHE_NFSFS_RENAMED))) + )) { + int h_unhashed; + + h_unhashed = d_unhashed(h_dentry); + spin_unlock(&h_dentry->d_lock); + AuDbg("unhash 0x%x 0x%x, %pd %pd\n", + unhashed, h_unhashed, dentry, h_dentry); + goto err; + } + spin_unlock(&h_dentry->d_lock); + + err = au_do_h_d_reval(h_dentry, flags, dentry, bindex); + if (unlikely(err)) + /* do not goto err, to keep the errno */ + break; + + /* todo: plink too? */ + if (!do_udba) + continue; + + /* UDBA tests */ + if (unlikely(!!inode != d_is_positive(h_dentry))) + goto err; + + h_inode = NULL; + if (d_is_positive(h_dentry)) + h_inode = d_inode(h_dentry); + h_plus = plus; + h_mode = mode; + h_cached_inode = h_inode; + if (h_inode) { + h_mode = (h_inode->i_mode & S_IFMT); + h_plus = (h_inode->i_nlink > 0); + } + if (inode && ibs <= bindex && bindex <= ibe) + h_cached_inode = au_h_iptr(inode, bindex); + + if (!h_nfs) { + if (unlikely(plus != h_plus && !tmpfile)) + goto err; + } else { + if (unlikely(!(h_dentry->d_flags & DCACHE_NFSFS_RENAMED) + && !is_root + && !IS_ROOT(h_dentry) + && unhashed != d_unhashed(h_dentry))) + goto err; + } + if (unlikely(mode != h_mode + || h_cached_inode != h_inode)) + goto err; + continue; + +err: + err = -EINVAL; + break; + } + + AuTraceErr(err); + return err; +} + +/* todo: consolidate with do_refresh() and au_reval_for_attr() */ +static int simple_reval_dpath(struct dentry *dentry, unsigned int sigen) +{ + int err; + struct dentry *parent; + + if (!au_digen_test(dentry, sigen)) + return 0; + + parent = dget_parent(dentry); + di_read_lock_parent(parent, AuLock_IR); + AuDebugOn(au_digen_test(parent, sigen)); + au_dbg_verify_gen(parent, sigen); + err = au_refresh_dentry(dentry, parent); + di_read_unlock(parent, AuLock_IR); + dput(parent); + AuTraceErr(err); + return err; +} + +int au_reval_dpath(struct dentry *dentry, unsigned int sigen) +{ + int err; + struct dentry *d, *parent; + + if (!au_ftest_si(au_sbi(dentry->d_sb), FAILED_REFRESH_DIR)) + return simple_reval_dpath(dentry, sigen); + + /* slow loop, keep it simple and stupid */ + /* cf: au_cpup_dirs() */ + err = 0; + parent = NULL; + while (au_digen_test(dentry, sigen)) { + d = dentry; + while (1) { + dput(parent); + parent = dget_parent(d); + if (!au_digen_test(parent, sigen)) + break; + d = parent; + } + + if (d != dentry) + di_write_lock_child2(d); + + /* someone might update our dentry while we were sleeping */ + if (au_digen_test(d, sigen)) { + /* + * todo: consolidate with simple_reval_dpath(), + * do_refresh() and au_reval_for_attr(). + */ + di_read_lock_parent(parent, AuLock_IR); + err = au_refresh_dentry(d, parent); + di_read_unlock(parent, AuLock_IR); + } + + if (d != dentry) + di_write_unlock(d); + dput(parent); + if (unlikely(err)) + break; + } + + return err; +} + +/* + * if valid returns 1, otherwise 0. + */ +static int aufs_d_revalidate(struct dentry *dentry, unsigned int flags) +{ + int valid, err; + unsigned int sigen; + unsigned char do_udba; + struct super_block *sb; + struct inode *inode; + + /* todo: support rcu-walk? */ + if (flags & LOOKUP_RCU) + return -ECHILD; + + valid = 0; + if (unlikely(!au_di(dentry))) + goto out; + + valid = 1; + sb = dentry->d_sb; + /* + * todo: very ugly + * i_mutex of parent dir may be held, + * but we should not return 'invalid' due to busy. + */ + err = aufs_read_lock(dentry, AuLock_FLUSH | AuLock_DW | AuLock_NOPLM); + if (unlikely(err)) { + valid = err; + AuTraceErr(err); + goto out; + } + inode = NULL; + if (d_really_is_positive(dentry)) + inode = d_inode(dentry); + if (unlikely(inode && is_bad_inode(inode))) { + err = -EINVAL; + AuTraceErr(err); + goto out_dgrade; + } + if (unlikely(au_dbrange_test(dentry))) { + err = -EINVAL; + AuTraceErr(err); + goto out_dgrade; + } + + sigen = au_sigen(sb); + if (au_digen_test(dentry, sigen)) { + AuDebugOn(IS_ROOT(dentry)); + err = au_reval_dpath(dentry, sigen); + if (unlikely(err)) { + AuTraceErr(err); + goto out_dgrade; + } + } + di_downgrade_lock(dentry, AuLock_IR); + + err = -EINVAL; + if (!(flags & (LOOKUP_OPEN | LOOKUP_EMPTY)) + && inode + && !(inode->i_state && I_LINKABLE) + && (IS_DEADDIR(inode) || !inode->i_nlink)) { + AuTraceErr(err); + goto out_inval; + } + + do_udba = !au_opt_test(au_mntflags(sb), UDBA_NONE); + if (do_udba && inode) { + aufs_bindex_t bstart = au_ibstart(inode); + struct inode *h_inode; + + if (bstart >= 0) { + h_inode = au_h_iptr(inode, bstart); + if (h_inode && au_test_higen(inode, h_inode)) { + AuTraceErr(err); + goto out_inval; + } + } + } + + err = h_d_revalidate(dentry, inode, flags, do_udba); + if (unlikely(!err && do_udba && au_dbstart(dentry) < 0)) { + err = -EIO; + AuDbg("both of real entry and whiteout found, %p, err %d\n", + dentry, err); + } + goto out_inval; + +out_dgrade: + di_downgrade_lock(dentry, AuLock_IR); +out_inval: + aufs_read_unlock(dentry, AuLock_IR); + AuTraceErr(err); + valid = !err; +out: + if (!valid) { + AuDbg("%pd invalid, %d\n", dentry, valid); + d_drop(dentry); + } + return valid; +} + +static void aufs_d_release(struct dentry *dentry) +{ + if (au_di(dentry)) { + au_di_fin(dentry); + au_hn_di_reinit(dentry); + } +} + +const struct dentry_operations aufs_dop = { + .d_revalidate = aufs_d_revalidate, + .d_weak_revalidate = aufs_d_revalidate, + .d_release = aufs_d_release +}; + +/* aufs_dop without d_revalidate */ +const struct dentry_operations aufs_dop_noreval = { + .d_release = aufs_d_release +}; diff --git a/fs/aufs/dentry.h b/fs/aufs/dentry.h new file mode 100644 index 000000000..c794adf59 --- /dev/null +++ b/fs/aufs/dentry.h @@ -0,0 +1,221 @@ +/* + * Copyright (C) 2005-2016 Junjiro R. Okajima + */ + +/* + * lookup and dentry operations + */ + +#ifndef __AUFS_DENTRY_H__ +#define __AUFS_DENTRY_H__ + +#ifdef __KERNEL__ + +#include <linux/dcache.h> +#include "rwsem.h" + +struct au_hdentry { + struct dentry *hd_dentry; + aufs_bindex_t hd_id; +}; + +struct au_dinfo { + atomic_t di_generation; + + struct au_rwsem di_rwsem; + aufs_bindex_t di_bstart, di_bend, di_bwh, di_bdiropq; + unsigned char di_tmpfile; /* to allow the different name */ + struct au_hdentry *di_hdentry; +} ____cacheline_aligned_in_smp; + +/* ---------------------------------------------------------------------- */ + +/* dentry.c */ +extern const struct dentry_operations aufs_dop, aufs_dop_noreval; +struct au_branch; +struct dentry *au_sio_lkup_one(struct qstr *name, struct dentry *parent); +int au_h_verify(struct dentry *h_dentry, unsigned int udba, struct inode *h_dir, + struct dentry *h_parent, struct au_branch *br); + +int au_lkup_dentry(struct dentry *dentry, aufs_bindex_t bstart, mode_t type); +int au_lkup_neg(struct dentry *dentry, aufs_bindex_t bindex, int wh); +int au_refresh_dentry(struct dentry *dentry, struct dentry *parent); +int au_reval_dpath(struct dentry *dentry, unsigned int sigen); +void au_refresh_dop(struct dentry *dentry, int force_reval); + +/* dinfo.c */ +void au_di_init_once(void *_di); +struct au_dinfo *au_di_alloc(struct super_block *sb, unsigned int lsc); +void au_di_free(struct au_dinfo *dinfo); +void au_di_swap(struct au_dinfo *a, struct au_dinfo *b); +void au_di_cp(struct au_dinfo *dst, struct au_dinfo *src); +int au_di_init(struct dentry *dentry); +void au_di_fin(struct dentry *dentry); +int au_di_realloc(struct au_dinfo *dinfo, int nbr); + +void di_read_lock(struct dentry *d, int flags, unsigned int lsc); +void di_read_unlock(struct dentry *d, int flags); +void di_downgrade_lock(struct dentry *d, int flags); +void di_write_lock(struct dentry *d, unsigned int lsc); +void di_write_unlock(struct dentry *d); +void di_write_lock2_child(struct dentry *d1, struct dentry *d2, int isdir); +void di_write_lock2_parent(struct dentry *d1, struct dentry *d2, int isdir); +void di_write_unlock2(struct dentry *d1, struct dentry *d2); + +struct dentry *au_h_dptr(struct dentry *dentry, aufs_bindex_t bindex); +struct dentry *au_h_d_alias(struct dentry *dentry, aufs_bindex_t bindex); +aufs_bindex_t au_dbtail(struct dentry *dentry); +aufs_bindex_t au_dbtaildir(struct dentry *dentry); + +void au_set_h_dptr(struct dentry *dentry, aufs_bindex_t bindex, + struct dentry *h_dentry); +int au_digen_test(struct dentry *dentry, unsigned int sigen); +int au_dbrange_test(struct dentry *dentry); +void au_update_digen(struct dentry *dentry); +void au_update_dbrange(struct dentry *dentry, int do_put_zero); +void au_update_dbstart(struct dentry *dentry); +void au_update_dbend(struct dentry *dentry); +int au_find_dbindex(struct dentry *dentry, struct dentry *h_dentry); + +/* ---------------------------------------------------------------------- */ + +static inline struct au_dinfo *au_di(struct dentry *dentry) +{ + return dentry->d_fsdata; +} + +/* ---------------------------------------------------------------------- */ + +/* lock subclass for dinfo */ +enum { + AuLsc_DI_CHILD, /* child first */ + AuLsc_DI_CHILD2, /* rename(2), link(2), and cpup at hnotify */ + AuLsc_DI_CHILD3, /* copyup dirs */ + AuLsc_DI_PARENT, + AuLsc_DI_PARENT2, + AuLsc_DI_PARENT3, + AuLsc_DI_TMP /* temp for replacing dinfo */ +}; + +/* + * di_read_lock_child, di_write_lock_child, + * di_read_lock_child2, di_write_lock_child2, + * di_read_lock_child3, di_write_lock_child3, + * di_read_lock_parent, di_write_lock_parent, + * di_read_lock_parent2, di_write_lock_parent2, + * di_read_lock_parent3, di_write_lock_parent3, + */ +#define AuReadLockFunc(name, lsc) \ +static inline void di_read_lock_##name(struct dentry *d, int flags) \ +{ di_read_lock(d, flags, AuLsc_DI_##lsc); } + +#define AuWriteLockFunc(name, lsc) \ +static inline void di_write_lock_##name(struct dentry *d) \ +{ di_write_lock(d, AuLsc_DI_##lsc); } + +#define AuRWLockFuncs(name, lsc) \ + AuReadLockFunc(name, lsc) \ + AuWriteLockFunc(name, lsc) + +AuRWLockFuncs(child, CHILD); +AuRWLockFuncs(child2, CHILD2); +AuRWLockFuncs(child3, CHILD3); +AuRWLockFuncs(parent, PARENT); +AuRWLockFuncs(parent2, PARENT2); +AuRWLockFuncs(parent3, PARENT3); + +#undef AuReadLockFunc +#undef AuWriteLockFunc +#undef AuRWLockFuncs + +#define DiMustNoWaiters(d) AuRwMustNoWaiters(&au_di(d)->di_rwsem) +#define DiMustAnyLock(d) AuRwMustAnyLock(&au_di(d)->di_rwsem) +#define DiMustWriteLock(d) AuRwMustWriteLock(&au_di(d)->di_rwsem) + +/* ---------------------------------------------------------------------- */ + +/* todo: memory barrier? */ +static inline unsigned int au_digen(struct dentry *d) +{ + return atomic_read(&au_di(d)->di_generation); +} + +static inline void au_h_dentry_init(struct au_hdentry *hdentry) +{ + hdentry->hd_dentry = NULL; +} + +static inline void au_hdput(struct au_hdentry *hd) +{ + if (hd) + dput(hd->hd_dentry); +} + +static inline aufs_bindex_t au_dbstart(struct dentry *dentry) +{ + DiMustAnyLock(dentry); + return au_di(dentry)->di_bstart; +} + +static inline aufs_bindex_t au_dbend(struct dentry *dentry) +{ + DiMustAnyLock(dentry); + return au_di(dentry)->di_bend; +} + +static inline aufs_bindex_t au_dbwh(struct dentry *dentry) +{ + DiMustAnyLock(dentry); + return au_di(dentry)->di_bwh; +} + +static inline aufs_bindex_t au_dbdiropq(struct dentry *dentry) +{ + DiMustAnyLock(dentry); + return au_di(dentry)->di_bdiropq; +} + +/* todo: hard/soft set? */ +static inline void au_set_dbstart(struct dentry *dentry, aufs_bindex_t bindex) +{ + DiMustWriteLock(dentry); + au_di(dentry)->di_bstart = bindex; +} + +static inline void au_set_dbend(struct dentry *dentry, aufs_bindex_t bindex) +{ + DiMustWriteLock(dentry); + au_di(dentry)->di_bend = bindex; +} + +static inline void au_set_dbwh(struct dentry *dentry, aufs_bindex_t bindex) +{ + DiMustWriteLock(dentry); + /* dbwh can be outside of bstart - bend range */ + au_di(dentry)->di_bwh = bindex; +} + +static inline void au_set_dbdiropq(struct dentry *dentry, aufs_bindex_t bindex) +{ + DiMustWriteLock(dentry); + au_di(dentry)->di_bdiropq = bindex; +} + +/* ---------------------------------------------------------------------- */ + +#ifdef CONFIG_AUFS_HNOTIFY +static inline void au_digen_dec(struct dentry *d) +{ + atomic_dec(&au_di(d)->di_generation); +} + +static inline void au_hn_di_reinit(struct dentry *dentry) +{ + dentry->d_fsdata = NULL; +} +#else +AuStubVoid(au_hn_di_reinit, struct dentry *dentry __maybe_unused) +#endif /* CONFIG_AUFS_HNOTIFY */ + +#endif /* __KERNEL__ */ +#endif /* __AUFS_DENTRY_H__ */ diff --git a/fs/aufs/dinfo.c b/fs/aufs/dinfo.c new file mode 100644 index 000000000..ad6d045c4 --- /dev/null +++ b/fs/aufs/dinfo.c @@ -0,0 +1,537 @@ +/* + * Copyright (C) 2005-2016 Junjiro R. Okajima + */ + +/* + * dentry private data + */ + +#include "aufs.h" + +void au_di_init_once(void *_dinfo) +{ + struct au_dinfo *dinfo = _dinfo; + static struct lock_class_key aufs_di; + + au_rw_init(&dinfo->di_rwsem); + au_rw_class(&dinfo->di_rwsem, &aufs_di); +} + +struct au_dinfo *au_di_alloc(struct super_block *sb, unsigned int lsc) +{ + struct au_dinfo *dinfo; + int nbr, i; + + dinfo = au_cache_alloc_dinfo(); + if (unlikely(!dinfo)) + goto out; + + nbr = au_sbend(sb) + 1; + if (nbr <= 0) + nbr = 1; + dinfo->di_hdentry = kcalloc(nbr, sizeof(*dinfo->di_hdentry), GFP_NOFS); + if (dinfo->di_hdentry) { + au_rw_write_lock_nested(&dinfo->di_rwsem, lsc); + dinfo->di_bstart = -1; + dinfo->di_bend = -1; + dinfo->di_bwh = -1; + dinfo->di_bdiropq = -1; + dinfo->di_tmpfile = 0; + for (i = 0; i < nbr; i++) + dinfo->di_hdentry[i].hd_id = -1; + goto out; + } + + au_cache_free_dinfo(dinfo); + dinfo = NULL; + +out: + return dinfo; +} + +void au_di_free(struct au_dinfo *dinfo) +{ + struct au_hdentry *p; + aufs_bindex_t bend, bindex; + + /* dentry may not be revalidated */ + bindex = dinfo->di_bstart; + if (bindex >= 0) { + bend = dinfo->di_bend; + p = dinfo->di_hdentry + bindex; + while (bindex++ <= bend) + au_hdput(p++); + } + kfree(dinfo->di_hdentry); + au_cache_free_dinfo(dinfo); +} + +void au_di_swap(struct au_dinfo *a, struct au_dinfo *b) +{ + struct au_hdentry *p; + aufs_bindex_t bi; + + AuRwMustWriteLock(&a->di_rwsem); + AuRwMustWriteLock(&b->di_rwsem); + +#define DiSwap(v, name) \ + do { \ + v = a->di_##name; \ + a->di_##name = b->di_##name; \ + b->di_##name = v; \ + } while (0) + + DiSwap(p, hdentry); + DiSwap(bi, bstart); + DiSwap(bi, bend); + DiSwap(bi, bwh); + DiSwap(bi, bdiropq); + /* smp_mb(); */ + +#undef DiSwap +} + +void au_di_cp(struct au_dinfo *dst, struct au_dinfo *src) +{ + AuRwMustWriteLock(&dst->di_rwsem); + AuRwMustWriteLock(&src->di_rwsem); + + dst->di_bstart = src->di_bstart; + dst->di_bend = src->di_bend; + dst->di_bwh = src->di_bwh; + dst->di_bdiropq = src->di_bdiropq; + /* smp_mb(); */ +} + +int au_di_init(struct dentry *dentry) +{ + int err; + struct super_block *sb; + struct au_dinfo *dinfo; + + err = 0; + sb = dentry->d_sb; + dinfo = au_di_alloc(sb, AuLsc_DI_CHILD); + if (dinfo) { + atomic_set(&dinfo->di_generation, au_sigen(sb)); + /* smp_mb(); */ /* atomic_set */ + dentry->d_fsdata = dinfo; + } else + err = -ENOMEM; + + return err; +} + +void au_di_fin(struct dentry *dentry) +{ + struct au_dinfo *dinfo; + + dinfo = au_di(dentry); + AuRwDestroy(&dinfo->di_rwsem); + au_di_free(dinfo); +} + +int au_di_realloc(struct au_dinfo *dinfo, int nbr) +{ + int err, sz; + struct au_hdentry *hdp; + + AuRwMustWriteLock(&dinfo->di_rwsem); + + err = -ENOMEM; + sz = sizeof(*hdp) * (dinfo->di_bend + 1); + if (!sz) + sz = sizeof(*hdp); + hdp = au_kzrealloc(dinfo->di_hdentry, sz, sizeof(*hdp) * nbr, GFP_NOFS); + if (hdp) { + dinfo->di_hdentry = hdp; + err = 0; + } + + return err; +} + +/* ---------------------------------------------------------------------- */ + +static void do_ii_write_lock(struct inode *inode, unsigned int lsc) +{ + switch (lsc) { + case AuLsc_DI_CHILD: + ii_write_lock_child(inode); + break; + case AuLsc_DI_CHILD2: + ii_write_lock_child2(inode); + break; + case AuLsc_DI_CHILD3: + ii_write_lock_child3(inode); + break; + case AuLsc_DI_PARENT: + ii_write_lock_parent(inode); + break; + case AuLsc_DI_PARENT2: + ii_write_lock_parent2(inode); + break; + case AuLsc_DI_PARENT3: + ii_write_lock_parent3(inode); + break; + default: + BUG(); + } +} + +static void do_ii_read_lock(struct inode *inode, unsigned int lsc) +{ + switch (lsc) { + case AuLsc_DI_CHILD: + ii_read_lock_child(inode); + break; + case AuLsc_DI_CHILD2: + ii_read_lock_child2(inode); + break; + case AuLsc_DI_CHILD3: + ii_read_lock_child3(inode); + break; + case AuLsc_DI_PARENT: + ii_read_lock_parent(inode); + break; + case AuLsc_DI_PARENT2: + ii_read_lock_parent2(inode); + break; + case AuLsc_DI_PARENT3: + ii_read_lock_parent3(inode); + break; + default: + BUG(); + } +} + +void di_read_lock(struct dentry *d, int flags, unsigned int lsc) +{ + struct inode *inode; + + au_rw_read_lock_nested(&au_di(d)->di_rwsem, lsc); + if (d_really_is_positive(d)) { + inode = d_inode(d); + if (au_ftest_lock(flags, IW)) + do_ii_write_lock(inode, lsc); + else if (au_ftest_lock(flags, IR)) + do_ii_read_lock(inode, lsc); + } +} + +void di_read_unlock(struct dentry *d, int flags) +{ + struct inode *inode; + + if (d_really_is_positive(d)) { + inode = d_inode(d); + if (au_ftest_lock(flags, IW)) { + au_dbg_verify_dinode(d); + ii_write_unlock(inode); + } else if (au_ftest_lock(flags, IR)) { + au_dbg_verify_dinode(d); + ii_read_unlock(inode); + } + } + au_rw_read_unlock(&au_di(d)->di_rwsem); +} + +void di_downgrade_lock(struct dentry *d, int flags) +{ + if (d_really_is_positive(d) && au_ftest_lock(flags, IR)) + ii_downgrade_lock(d_inode(d)); + au_rw_dgrade_lock(&au_di(d)->di_rwsem); +} + +void di_write_lock(struct dentry *d, unsigned int lsc) +{ + au_rw_write_lock_nested(&au_di(d)->di_rwsem, lsc); + if (d_really_is_positive(d)) + do_ii_write_lock(d_inode(d), lsc); +} + +void di_write_unlock(struct dentry *d) +{ + au_dbg_verify_dinode(d); + if (d_really_is_positive(d)) + ii_write_unlock(d_inode(d)); + au_rw_write_unlock(&au_di(d)->di_rwsem); +} + +void di_write_lock2_child(struct dentry *d1, struct dentry *d2, int isdir) +{ + AuDebugOn(d1 == d2 + || d_inode(d1) == d_inode(d2) + || d1->d_sb != d2->d_sb); + + if (isdir && au_test_subdir(d1, d2)) { + di_write_lock_child(d1); + di_write_lock_child2(d2); + } else { + /* there should be no races */ + di_write_lock_child(d2); + di_write_lock_child2(d1); + } +} + +void di_write_lock2_parent(struct dentry *d1, struct dentry *d2, int isdir) +{ + AuDebugOn(d1 == d2 + || d_inode(d1) == d_inode(d2) + || d1->d_sb != d2->d_sb); + + if (isdir && au_test_subdir(d1, d2)) { + di_write_lock_parent(d1); + di_write_lock_parent2(d2); + } else { + /* there should be no races */ + di_write_lock_parent(d2); + di_write_lock_parent2(d1); + } +} + +void di_write_unlock2(struct dentry *d1, struct dentry *d2) +{ + di_write_unlock(d1); + if (d_inode(d1) == d_inode(d2)) + au_rw_write_unlock(&au_di(d2)->di_rwsem); + else + di_write_unlock(d2); +} + +/* ---------------------------------------------------------------------- */ + +struct dentry *au_h_dptr(struct dentry *dentry, aufs_bindex_t bindex) +{ + struct dentry *d; + + DiMustAnyLock(dentry); + + if (au_dbstart(dentry) < 0 || bindex < au_dbstart(dentry)) + return NULL; + AuDebugOn(bindex < 0); + d = au_di(dentry)->di_hdentry[0 + bindex].hd_dentry; + AuDebugOn(d && au_dcount(d) <= 0); + return d; +} + +/* + * extended version of au_h_dptr(). + * returns a hashed and positive (or linkable) h_dentry in bindex, NULL, or + * error. + */ +struct dentry *au_h_d_alias(struct dentry *dentry, aufs_bindex_t bindex) +{ + struct dentry *h_dentry; + struct inode *inode, *h_inode; + + AuDebugOn(d_really_is_negative(dentry)); + + h_dentry = NULL; + if (au_dbstart(dentry) <= bindex + && bindex <= au_dbend(dentry)) + h_dentry = au_h_dptr(dentry, bindex); + if (h_dentry && !au_d_linkable(h_dentry)) { + dget(h_dentry); + goto out; /* success */ + } + + inode = d_inode(dentry); + AuDebugOn(bindex < au_ibstart(inode)); + AuDebugOn(au_ibend(inode) < bindex); + h_inode = au_h_iptr(inode, bindex); + h_dentry = d_find_alias(h_inode); + if (h_dentry) { + if (!IS_ERR(h_dentry)) { + if (!au_d_linkable(h_dentry)) + goto out; /* success */ + dput(h_dentry); + } else + goto out; + } + + if (au_opt_test(au_mntflags(dentry->d_sb), PLINK)) { + h_dentry = au_plink_lkup(inode, bindex); + AuDebugOn(!h_dentry); + if (!IS_ERR(h_dentry)) { + if (!au_d_hashed_positive(h_dentry)) + goto out; /* success */ + dput(h_dentry); + h_dentry = NULL; + } + } + +out: + AuDbgDentry(h_dentry); + return h_dentry; +} + +aufs_bindex_t au_dbtail(struct dentry *dentry) +{ + aufs_bindex_t bend, bwh; + + bend = au_dbend(dentry); + if (0 <= bend) { + bwh = au_dbwh(dentry); + if (!bwh) + return bwh; + if (0 < bwh && bwh < bend) + return bwh - 1; + } + return bend; +} + +aufs_bindex_t au_dbtaildir(struct dentry *dentry) +{ + aufs_bindex_t bend, bopq; + + bend = au_dbtail(dentry); + if (0 <= bend) { + bopq = au_dbdiropq(dentry); + if (0 <= bopq && bopq < bend) + bend = bopq; + } + return bend; +} + +/* ---------------------------------------------------------------------- */ + +void au_set_h_dptr(struct dentry *dentry, aufs_bindex_t bindex, + struct dentry *h_dentry) +{ + struct au_hdentry *hd = au_di(dentry)->di_hdentry + bindex; + struct au_branch *br; + + DiMustWriteLock(dentry); + + au_hdput(hd); + hd->hd_dentry = h_dentry; + if (h_dentry) { + br = au_sbr(dentry->d_sb, bindex); + hd->hd_id = br->br_id; + } +} + +int au_dbrange_test(struct dentry *dentry) +{ + int err; + aufs_bindex_t bstart, bend; + + err = 0; + bstart = au_dbstart(dentry); + bend = au_dbend(dentry); + if (bstart >= 0) + AuDebugOn(bend < 0 && bstart > bend); + else { + err = -EIO; + AuDebugOn(bend >= 0); + } + + return err; +} + +int au_digen_test(struct dentry *dentry, unsigned int sigen) +{ + int err; + + err = 0; + if (unlikely(au_digen(dentry) != sigen + || au_iigen_test(d_inode(dentry), sigen))) + err = -EIO; + + return err; +} + +void au_update_digen(struct dentry *dentry) +{ + atomic_set(&au_di(dentry)->di_generation, au_sigen(dentry->d_sb)); + /* smp_mb(); */ /* atomic_set */ +} + +void au_update_dbrange(struct dentry *dentry, int do_put_zero) +{ + struct au_dinfo *dinfo; + struct dentry *h_d; + struct au_hdentry *hdp; + + DiMustWriteLock(dentry); + + dinfo = au_di(dentry); + if (!dinfo || dinfo->di_bstart < 0) + return; + + hdp = dinfo->di_hdentry; + if (do_put_zero) { + aufs_bindex_t bindex, bend; + + bend = dinfo->di_bend; + for (bindex = dinfo->di_bstart; bindex <= bend; bindex++) { + h_d = hdp[0 + bindex].hd_dentry; + if (h_d && d_is_negative(h_d)) + au_set_h_dptr(dentry, bindex, NULL); + } + } + + dinfo->di_bstart = -1; + while (++dinfo->di_bstart <= dinfo->di_bend) + if (hdp[0 + dinfo->di_bstart].hd_dentry) + break; + if (dinfo->di_bstart > dinfo->di_bend) { + dinfo->di_bstart = -1; + dinfo->di_bend = -1; + return; + } + + dinfo->di_bend++; + while (0 <= --dinfo->di_bend) + if (hdp[0 + dinfo->di_bend].hd_dentry) + break; + AuDebugOn(dinfo->di_bstart > dinfo->di_bend || dinfo->di_bend < 0); +} + +void au_update_dbstart(struct dentry *dentry) +{ + aufs_bindex_t bindex, bend; + struct dentry *h_dentry; + + bend = au_dbend(dentry); + for (bindex = au_dbstart(dentry); bindex <= bend; bindex++) { + h_dentry = au_h_dptr(dentry, bindex); + if (!h_dentry) + continue; + if (d_is_positive(h_dentry)) { + au_set_dbstart(dentry, bindex); + return; + } + au_set_h_dptr(dentry, bindex, NULL); + } +} + +void au_update_dbend(struct dentry *dentry) +{ + aufs_bindex_t bindex, bstart; + struct dentry *h_dentry; + + bstart = au_dbstart(dentry); + for (bindex = au_dbend(dentry); bindex >= bstart; bindex--) { + h_dentry = au_h_dptr(dentry, bindex); + if (!h_dentry) + continue; + if (d_is_positive(h_dentry)) { + au_set_dbend(dentry, bindex); + return; + } + au_set_h_dptr(dentry, bindex, NULL); + } +} + +int au_find_dbindex(struct dentry *dentry, struct dentry *h_dentry) +{ + aufs_bindex_t bindex, bend; + + bend = au_dbend(dentry); + for (bindex = au_dbstart(dentry); bindex <= bend; bindex++) + if (au_h_dptr(dentry, bindex) == h_dentry) + return bindex; + return -1; +} diff --git a/fs/aufs/dir.c b/fs/aufs/dir.c new file mode 100644 index 000000000..4644de4a9 --- /dev/null +++ b/fs/aufs/dir.c @@ -0,0 +1,743 @@ +/* + * Copyright (C) 2005-2016 Junjiro R. Okajima + */ + +/* + * directory operations + */ + +#include <linux/fs_stack.h> +#include "aufs.h" + +void au_add_nlink(struct inode *dir, struct inode *h_dir) +{ + unsigned int nlink; + + AuDebugOn(!S_ISDIR(dir->i_mode) || !S_ISDIR(h_dir->i_mode)); + + nlink = dir->i_nlink; + nlink += h_dir->i_nlink - 2; + if (h_dir->i_nlink < 2) + nlink += 2; + smp_mb(); /* for i_nlink */ + /* 0 can happen in revaliding */ + set_nlink(dir, nlink); +} + +void au_sub_nlink(struct inode *dir, struct inode *h_dir) +{ + unsigned int nlink; + + AuDebugOn(!S_ISDIR(dir->i_mode) || !S_ISDIR(h_dir->i_mode)); + + nlink = dir->i_nlink; + nlink -= h_dir->i_nlink - 2; + if (h_dir->i_nlink < 2) + nlink -= 2; + smp_mb(); /* for i_nlink */ + /* nlink == 0 means the branch-fs is broken */ + set_nlink(dir, nlink); +} + +loff_t au_dir_size(struct file *file, struct dentry *dentry) +{ + loff_t sz; + aufs_bindex_t bindex, bend; + struct file *h_file; + struct dentry *h_dentry; + + sz = 0; + if (file) { + AuDebugOn(!d_is_dir(file->f_path.dentry)); + + bend = au_fbend_dir(file); + for (bindex = au_fbstart(file); + bindex <= bend && sz < KMALLOC_MAX_SIZE; + bindex++) { + h_file = au_hf_dir(file, bindex); + if (h_file && file_inode(h_file)) + sz += vfsub_f_size_read(h_file); + } + } else { + AuDebugOn(!dentry); + AuDebugOn(!d_is_dir(dentry)); + + bend = au_dbtaildir(dentry); + for (bindex = au_dbstart(dentry); + bindex <= bend && sz < KMALLOC_MAX_SIZE; + bindex++) { + h_dentry = au_h_dptr(dentry, bindex); + if (h_dentry && d_is_positive(h_dentry)) + sz += i_size_read(d_inode(h_dentry)); + } + } + if (sz < KMALLOC_MAX_SIZE) + sz = roundup_pow_of_two(sz); + if (sz > KMALLOC_MAX_SIZE) + sz = KMALLOC_MAX_SIZE; + else if (sz < NAME_MAX) { + BUILD_BUG_ON(AUFS_RDBLK_DEF < NAME_MAX); + sz = AUFS_RDBLK_DEF; + } + return sz; +} + +struct au_dir_ts_arg { + struct dentry *dentry; + aufs_bindex_t brid; +}; + +static void au_do_dir_ts(void *arg) +{ + struct au_dir_ts_arg *a = arg; + struct au_dtime dt; + struct path h_path; + struct inode *dir, *h_dir; + struct super_block *sb; + struct au_branch *br; + struct au_hinode *hdir; + int err; + aufs_bindex_t bstart, bindex; + + sb = a->dentry->d_sb; + if (d_really_is_negative(a->dentry)) + goto out; + /* no dir->i_mutex lock */ + aufs_read_lock(a->dentry, AuLock_DW); /* noflush */ + + dir = d_inode(a->dentry); + bstart = au_ibstart(dir); + bindex = au_br_index(sb, a->brid); + if (bindex < bstart) + goto out_unlock; + + br = au_sbr(sb, bindex); + h_path.dentry = au_h_dptr(a->dentry, bindex); + if (!h_path.dentry) + goto out_unlock; + h_path.mnt = au_br_mnt(br); + au_dtime_store(&dt, a->dentry, &h_path); + + br = au_sbr(sb, bstart); + if (!au_br_writable(br->br_perm)) + goto out_unlock; + h_path.dentry = au_h_dptr(a->dentry, bstart); + h_path.mnt = au_br_mnt(br); + err = vfsub_mnt_want_write(h_path.mnt); + if (err) + goto out_unlock; + hdir = au_hi(dir, bstart); + au_hn_imtx_lock_nested(hdir, AuLsc_I_PARENT); + h_dir = au_h_iptr(dir, bstart); + if (h_dir->i_nlink + && timespec_compare(&h_dir->i_mtime, &dt.dt_mtime) < 0) { + dt.dt_h_path = h_path; + au_dtime_revert(&dt); + } + au_hn_imtx_unlock(hdir); + vfsub_mnt_drop_write(h_path.mnt); + au_cpup_attr_timesizes(dir); + +out_unlock: + aufs_read_unlock(a->dentry, AuLock_DW); +out: + dput(a->dentry); + au_nwt_done(&au_sbi(sb)->si_nowait); + kfree(arg); +} + +void au_dir_ts(struct inode *dir, aufs_bindex_t bindex) +{ + int perm, wkq_err; + aufs_bindex_t bstart; + struct au_dir_ts_arg *arg; + struct dentry *dentry; + struct super_block *sb; + + IMustLock(dir); + + dentry = d_find_any_alias(dir); + AuDebugOn(!dentry); + sb = dentry->d_sb; + bstart = au_ibstart(dir); + if (bstart == bindex) { + au_cpup_attr_timesizes(dir); + goto out; + } + + perm = au_sbr_perm(sb, bstart); + if (!au_br_writable(perm)) + goto out; + + arg = kmalloc(sizeof(*arg), GFP_NOFS); + if (!arg) + goto out; + + arg->dentry = dget(dentry); /* will be dput-ted by au_do_dir_ts() */ + arg->brid = au_sbr_id(sb, bindex); + wkq_err = au_wkq_nowait(au_do_dir_ts, arg, sb, /*flags*/0); + if (unlikely(wkq_err)) { + pr_err("wkq %d\n", wkq_err); + dput(dentry); + kfree(arg); + } + +out: + dput(dentry); +} + +/* ---------------------------------------------------------------------- */ + +static int reopen_dir(struct file *file) +{ + int err; + unsigned int flags; + aufs_bindex_t bindex, btail, bstart; + struct dentry *dentry, *h_dentry; + struct file *h_file; + + /* open all lower dirs */ + dentry = file->f_path.dentry; + bstart = au_dbstart(dentry); + for (bindex = au_fbstart(file); bindex < bstart; bindex++) + au_set_h_fptr(file, bindex, NULL); + au_set_fbstart(file, bstart); + + btail = au_dbtaildir(dentry); + for (bindex = au_fbend_dir(file); btail < bindex; bindex--) + au_set_h_fptr(file, bindex, NULL); + au_set_fbend_dir(file, btail); + + flags = vfsub_file_flags(file); + for (bindex = bstart; bindex <= btail; bindex++) { + h_dentry = au_h_dptr(dentry, bindex); + if (!h_dentry) + continue; + h_file = au_hf_dir(file, bindex); + if (h_file) + continue; + + h_file = au_h_open(dentry, bindex, flags, file, /*force_wr*/0); + err = PTR_ERR(h_file); + if (IS_ERR(h_file)) + goto out; /* close all? */ + au_set_h_fptr(file, bindex, h_file); + } + au_update_figen(file); + /* todo: necessary? */ + /* file->f_ra = h_file->f_ra; */ + err = 0; + +out: + return err; +} + +static int do_open_dir(struct file *file, int flags, struct file *h_file) +{ + int err; + aufs_bindex_t bindex, btail; + struct dentry *dentry, *h_dentry; + struct vfsmount *mnt; + + FiMustWriteLock(file); + AuDebugOn(h_file); + + err = 0; + mnt = file->f_path.mnt; + dentry = file->f_path.dentry; + file->f_version = d_inode(dentry)->i_version; + bindex = au_dbstart(dentry); + au_set_fbstart(file, bindex); + btail = au_dbtaildir(dentry); + au_set_fbend_dir(file, btail); + for (; !err && bindex <= btail; bindex++) { + h_dentry = au_h_dptr(dentry, bindex); + if (!h_dentry) + continue; + + err = vfsub_test_mntns(mnt, h_dentry->d_sb); + if (unlikely(err)) + break; + h_file = au_h_open(dentry, bindex, flags, file, /*force_wr*/0); + if (IS_ERR(h_file)) { + err = PTR_ERR(h_file); + break; + } + au_set_h_fptr(file, bindex, h_file); + } + au_update_figen(file); + /* todo: necessary? */ + /* file->f_ra = h_file->f_ra; */ + if (!err) + return 0; /* success */ + + /* close all */ + for (bindex = au_fbstart(file); bindex <= btail; bindex++) + au_set_h_fptr(file, bindex, NULL); + au_set_fbstart(file, -1); + au_set_fbend_dir(file, -1); + + return err; +} + +static int aufs_open_dir(struct inode *inode __maybe_unused, + struct file *file) +{ + int err; + struct super_block *sb; + struct au_fidir *fidir; + + err = -ENOMEM; + sb = file->f_path.dentry->d_sb; + si_read_lock(sb, AuLock_FLUSH); + fidir = au_fidir_alloc(sb); + if (fidir) { + struct au_do_open_args args = { + .open = do_open_dir, + .fidir = fidir + }; + err = au_do_open(file, &args); + if (unlikely(err)) + kfree(fidir); + } + si_read_unlock(sb); + return err; +} + +static int aufs_release_dir(struct inode *inode __maybe_unused, + struct file *file) +{ + struct au_vdir *vdir_cache; + struct au_finfo *finfo; + struct au_fidir *fidir; + aufs_bindex_t bindex, bend; + + finfo = au_fi(file); + fidir = finfo->fi_hdir; + if (fidir) { + au_sphl_del(&finfo->fi_hlist, + &au_sbi(file->f_path.dentry->d_sb)->si_files); + vdir_cache = fidir->fd_vdir_cache; /* lock-free */ + if (vdir_cache) + au_vdir_free(vdir_cache); + + bindex = finfo->fi_btop; + if (bindex >= 0) { + /* + * calls fput() instead of filp_close(), + * since no dnotify or lock for the lower file. + */ + bend = fidir->fd_bbot; + for (; bindex <= bend; bindex++) + au_set_h_fptr(file, bindex, NULL); + } + kfree(fidir); + finfo->fi_hdir = NULL; + } + au_finfo_fin(file); + return 0; +} + +/* ---------------------------------------------------------------------- */ + +static int au_do_flush_dir(struct file *file, fl_owner_t id) +{ + int err; + aufs_bindex_t bindex, bend; + struct file *h_file; + + err = 0; + bend = au_fbend_dir(file); + for (bindex = au_fbstart(file); !err && bindex <= bend; bindex++) { + h_file = au_hf_dir(file, bindex); + if (h_file) + err = vfsub_flush(h_file, id); + } + return err; +} + +static int aufs_flush_dir(struct file *file, fl_owner_t id) +{ + return au_do_flush(file, id, au_do_flush_dir); +} + +/* ---------------------------------------------------------------------- */ + +static int au_do_fsync_dir_no_file(struct dentry *dentry, int datasync) +{ + int err; + aufs_bindex_t bend, bindex; + struct inode *inode; + struct super_block *sb; + + err = 0; + sb = dentry->d_sb; + inode = d_inode(dentry); + IMustLock(inode); + bend = au_dbend(dentry); + for (bindex = au_dbstart(dentry); !err && bindex <= bend; bindex++) { + struct path h_path; + + if (au_test_ro(sb, bindex, inode)) + continue; + h_path.dentry = au_h_dptr(dentry, bindex); + if (!h_path.dentry) + continue; + + h_path.mnt = au_sbr_mnt(sb, bindex); + err = vfsub_fsync(NULL, &h_path, datasync); + } + + return err; +} + +static int au_do_fsync_dir(struct file *file, int datasync) +{ + int err; + aufs_bindex_t bend, bindex; + struct file *h_file; + struct super_block *sb; + struct inode *inode; + + err = au_reval_and_lock_fdi(file, reopen_dir, /*wlock*/1); + if (unlikely(err)) + goto out; + + inode = file_inode(file); + sb = inode->i_sb; + bend = au_fbend_dir(file); + for (bindex = au_fbstart(file); !err && bindex <= bend; bindex++) { + h_file = au_hf_dir(file, bindex); + if (!h_file || au_test_ro(sb, bindex, inode)) + continue; + + err = vfsub_fsync(h_file, &h_file->f_path, datasync); + } + +out: + return err; +} + +/* + * @file may be NULL + */ +static int aufs_fsync_dir(struct file *file, loff_t start, loff_t end, + int datasync) +{ + int err; + struct dentry *dentry; + struct inode *inode; + struct super_block *sb; + + err = 0; + dentry = file->f_path.dentry; + inode = d_inode(dentry); + inode_lock(inode); + sb = dentry->d_sb; + si_noflush_read_lock(sb); + if (file) + err = au_do_fsync_dir(file, datasync); + else { + di_write_lock_child(dentry); + err = au_do_fsync_dir_no_file(dentry, datasync); + } + au_cpup_attr_timesizes(inode); + di_write_unlock(dentry); + if (file) + fi_write_unlock(file); + + si_read_unlock(sb); + inode_unlock(inode); + return err; +} + +/* ---------------------------------------------------------------------- */ + +static int aufs_iterate(struct file *file, struct dir_context *ctx) +{ + int err; + struct dentry *dentry; + struct inode *inode, *h_inode; + struct super_block *sb; + + AuDbg("%pD, ctx{%pf, %llu}\n", file, ctx->actor, ctx->pos); + + dentry = file->f_path.dentry; + inode = d_inode(dentry); + IMustLock(inode); + + sb = dentry->d_sb; + si_read_lock(sb, AuLock_FLUSH); + err = au_reval_and_lock_fdi(file, reopen_dir, /*wlock*/1); + if (unlikely(err)) + goto out; + err = au_alive_dir(dentry); + if (!err) + err = au_vdir_init(file); + di_downgrade_lock(dentry, AuLock_IR); + if (unlikely(err)) + goto out_unlock; + + h_inode = au_h_iptr(inode, au_ibstart(inode)); + if (!au_test_nfsd()) { + err = au_vdir_fill_de(file, ctx); + fsstack_copy_attr_atime(inode, h_inode); + } else { + /* + * nfsd filldir may call lookup_one_len(), vfs_getattr(), + * encode_fh() and others. + */ + atomic_inc(&h_inode->i_count); + di_read_unlock(dentry, AuLock_IR); + si_read_unlock(sb); + err = au_vdir_fill_de(file, ctx); + fsstack_copy_attr_atime(inode, h_inode); + fi_write_unlock(file); + iput(h_inode); + + AuTraceErr(err); + return err; + } + +out_unlock: + di_read_unlock(dentry, AuLock_IR); + fi_write_unlock(file); +out: + si_read_unlock(sb); + return err; +} + +/* ---------------------------------------------------------------------- */ + +#define AuTestEmpty_WHONLY 1 +#define AuTestEmpty_CALLED (1 << 1) +#define AuTestEmpty_SHWH (1 << 2) +#define au_ftest_testempty(flags, name) ((flags) & AuTestEmpty_##name) +#define au_fset_testempty(flags, name) \ + do { (flags) |= AuTestEmpty_##name; } while (0) +#define au_fclr_testempty(flags, name) \ + do { (flags) &= ~AuTestEmpty_##name; } while (0) + +#ifndef CONFIG_AUFS_SHWH +#undef AuTestEmpty_SHWH +#define AuTestEmpty_SHWH 0 +#endif + +struct test_empty_arg { + struct dir_context ctx; + struct au_nhash *whlist; + unsigned int flags; + int err; + aufs_bindex_t bindex; +}; + +static int test_empty_cb(struct dir_context *ctx, const char *__name, + int namelen, loff_t offset __maybe_unused, u64 ino, + unsigned int d_type) +{ + struct test_empty_arg *arg = container_of(ctx, struct test_empty_arg, + ctx); + char *name = (void *)__name; + + arg->err = 0; + au_fset_testempty(arg->flags, CALLED); + /* smp_mb(); */ + if (name[0] == '.' + && (namelen == 1 || (name[1] == '.' && namelen == 2))) + goto out; /* success */ + + if (namelen <= AUFS_WH_PFX_LEN + || memcmp(name, AUFS_WH_PFX, AUFS_WH_PFX_LEN)) { + if (au_ftest_testempty(arg->flags, WHONLY) + && !au_nhash_test_known_wh(arg->whlist, name, namelen)) + arg->err = -ENOTEMPTY; + goto out; + } + + name += AUFS_WH_PFX_LEN; + namelen -= AUFS_WH_PFX_LEN; + if (!au_nhash_test_known_wh(arg->whlist, name, namelen)) + arg->err = au_nhash_append_wh + (arg->whlist, name, namelen, ino, d_type, arg->bindex, + au_ftest_testempty(arg->flags, SHWH)); + +out: + /* smp_mb(); */ + AuTraceErr(arg->err); + return arg->err; +} + +static int do_test_empty(struct dentry *dentry, struct test_empty_arg *arg) +{ + int err; + struct file *h_file; + + h_file = au_h_open(dentry, arg->bindex, + O_RDONLY | O_NONBLOCK | O_DIRECTORY | O_LARGEFILE, + /*file*/NULL, /*force_wr*/0); + err = PTR_ERR(h_file); + if (IS_ERR(h_file)) + goto out; + + err = 0; + if (!au_opt_test(au_mntflags(dentry->d_sb), UDBA_NONE) + && !file_inode(h_file)->i_nlink) + goto out_put; + + do { + arg->err = 0; + au_fclr_testempty(arg->flags, CALLED); + /* smp_mb(); */ + err = vfsub_iterate_dir(h_file, &arg->ctx); + if (err >= 0) + err = arg->err; + } while (!err && au_ftest_testempty(arg->flags, CALLED)); + +out_put: + fput(h_file); + au_sbr_put(dentry->d_sb, arg->bindex); +out: + return err; +} + +struct do_test_empty_args { + int *errp; + struct dentry *dentry; + struct test_empty_arg *arg; +}; + +static void call_do_test_empty(void *args) +{ + struct do_test_empty_args *a = args; + *a->errp = do_test_empty(a->dentry, a->arg); +} + +static int sio_test_empty(struct dentry *dentry, struct test_empty_arg *arg) +{ + int err, wkq_err; + struct dentry *h_dentry; + struct inode *h_inode; + + h_dentry = au_h_dptr(dentry, arg->bindex); + h_inode = d_inode(h_dentry); + /* todo: i_mode changes anytime? */ + inode_lock_nested(h_inode, AuLsc_I_CHILD); + err = au_test_h_perm_sio(h_inode, MAY_EXEC | MAY_READ); + inode_unlock(h_inode); + if (!err) + err = do_test_empty(dentry, arg); + else { + struct do_test_empty_args args = { + .errp = &err, + .dentry = dentry, + .arg = arg + }; + unsigned int flags = arg->flags; + + wkq_err = au_wkq_wait(call_do_test_empty, &args); + if (unlikely(wkq_err)) + err = wkq_err; + arg->flags = flags; + } + + return err; +} + +int au_test_empty_lower(struct dentry *dentry) +{ + int err; + unsigned int rdhash; + aufs_bindex_t bindex, bstart, btail; + struct au_nhash whlist; + struct test_empty_arg arg = { + .ctx = { + .actor = test_empty_cb + } + }; + int (*test_empty)(struct dentry *dentry, struct test_empty_arg *arg); + + SiMustAnyLock(dentry->d_sb); + + rdhash = au_sbi(dentry->d_sb)->si_rdhash; + if (!rdhash) + rdhash = au_rdhash_est(au_dir_size(/*file*/NULL, dentry)); + err = au_nhash_alloc(&whlist, rdhash, GFP_NOFS); + if (unlikely(err)) + goto out; + + arg.flags = 0; + arg.whlist = &whlist; + bstart = au_dbstart(dentry); + if (au_opt_test(au_mntflags(dentry->d_sb), SHWH)) + au_fset_testempty(arg.flags, SHWH); + test_empty = do_test_empty; + if (au_opt_test(au_mntflags(dentry->d_sb), DIRPERM1)) + test_empty = sio_test_empty; + arg.bindex = bstart; + err = test_empty(dentry, &arg); + if (unlikely(err)) + goto out_whlist; + + au_fset_testempty(arg.flags, WHONLY); + btail = au_dbtaildir(dentry); + for (bindex = bstart + 1; !err && bindex <= btail; bindex++) { + struct dentry *h_dentry; + + h_dentry = au_h_dptr(dentry, bindex); + if (h_dentry && d_is_positive(h_dentry)) { + arg.bindex = bindex; + err = test_empty(dentry, &arg); + } + } + +out_whlist: + au_nhash_wh_free(&whlist); +out: + return err; +} + +int au_test_empty(struct dentry *dentry, struct au_nhash *whlist) +{ + int err; + struct test_empty_arg arg = { + .ctx = { + .actor = test_empty_cb + } + }; + aufs_bindex_t bindex, btail; + + err = 0; + arg.whlist = whlist; + arg.flags = AuTestEmpty_WHONLY; + if (au_opt_test(au_mntflags(dentry->d_sb), SHWH)) + au_fset_testempty(arg.flags, SHWH); + btail = au_dbtaildir(dentry); + for (bindex = au_dbstart(dentry); !err && bindex <= btail; bindex++) { + struct dentry *h_dentry; + + h_dentry = au_h_dptr(dentry, bindex); + if (h_dentry && d_is_positive(h_dentry)) { + arg.bindex = bindex; + err = sio_test_empty(dentry, &arg); + } + } + + return err; +} + +/* ---------------------------------------------------------------------- */ + +const struct file_operations aufs_dir_fop = { + .owner = THIS_MODULE, + .llseek = default_llseek, + .read = generic_read_dir, + .iterate = aufs_iterate, + .unlocked_ioctl = aufs_ioctl_dir, +#ifdef CONFIG_COMPAT + .compat_ioctl = aufs_compat_ioctl_dir, +#endif + .open = aufs_open_dir, + .release = aufs_release_dir, + .flush = aufs_flush_dir, + .fsync = aufs_fsync_dir +}; diff --git a/fs/aufs/dir.h b/fs/aufs/dir.h new file mode 100644 index 000000000..b0a79d722 --- /dev/null +++ b/fs/aufs/dir.h @@ -0,0 +1,118 @@ +/* + * Copyright (C) 2005-2016 Junjiro R. Okajima + */ + +/* + * directory operations + */ + +#ifndef __AUFS_DIR_H__ +#define __AUFS_DIR_H__ + +#ifdef __KERNEL__ + +#include <linux/fs.h> + +/* ---------------------------------------------------------------------- */ + +/* need to be faster and smaller */ + +struct au_nhash { + unsigned int nh_num; + struct hlist_head *nh_head; +}; + +struct au_vdir_destr { + unsigned char len; + unsigned char name[0]; +} __packed; + +struct au_vdir_dehstr { + struct hlist_node hash; + struct au_vdir_destr *str; +} ____cacheline_aligned_in_smp; + +struct au_vdir_de { + ino_t de_ino; + unsigned char de_type; + /* caution: packed */ + struct au_vdir_destr de_str; +} __packed; + +struct au_vdir_wh { + struct hlist_node wh_hash; +#ifdef CONFIG_AUFS_SHWH + ino_t wh_ino; + aufs_bindex_t wh_bindex; + unsigned char wh_type; +#else + aufs_bindex_t wh_bindex; +#endif + /* caution: packed */ + struct au_vdir_destr wh_str; +} __packed; + +union au_vdir_deblk_p { + unsigned char *deblk; + struct au_vdir_de *de; +}; + +struct au_vdir { + unsigned char **vd_deblk; + unsigned long vd_nblk; + struct { + unsigned long ul; + union au_vdir_deblk_p p; + } vd_last; + + unsigned long vd_version; + unsigned int vd_deblk_sz; + unsigned long vd_jiffy; +} ____cacheline_aligned_in_smp; + +/* ---------------------------------------------------------------------- */ + +/* dir.c */ +extern const struct file_operations aufs_dir_fop; +void au_add_nlink(struct inode *dir, struct inode *h_dir); +void au_sub_nlink(struct inode *dir, struct inode *h_dir); +loff_t au_dir_size(struct file *file, struct dentry *dentry); +void au_dir_ts(struct inode *dir, aufs_bindex_t bsrc); +int au_test_empty_lower(struct dentry *dentry); +int au_test_empty(struct dentry *dentry, struct au_nhash *whlist); + +/* vdir.c */ +unsigned int au_rdhash_est(loff_t sz); +int au_nhash_alloc(struct au_nhash *nhash, unsigned int num_hash, gfp_t gfp); +void au_nhash_wh_free(struct au_nhash *whlist); +int au_nhash_test_longer_wh(struct au_nhash *whlist, aufs_bindex_t btgt, + int limit); +int au_nhash_test_known_wh(struct au_nhash *whlist, char *name, int nlen); +int au_nhash_append_wh(struct au_nhash *whlist, char *name, int nlen, ino_t ino, + unsigned int d_type, aufs_bindex_t bindex, + unsigned char shwh); +void au_vdir_free(struct au_vdir *vdir); +int au_vdir_init(struct file *file); +int au_vdir_fill_de(struct file *file, struct dir_context *ctx); + +/* ioctl.c */ +long aufs_ioctl_dir(struct file *file, unsigned int cmd, unsigned long arg); + +#ifdef CONFIG_AUFS_RDU +/* rdu.c */ +long au_rdu_ioctl(struct file *file, unsigned int cmd, unsigned long arg); +#ifdef CONFIG_COMPAT +long au_rdu_compat_ioctl(struct file *file, unsigned int cmd, + unsigned long arg); +#endif +#else +AuStub(long, au_rdu_ioctl, return -EINVAL, struct file *file, + unsigned int cmd, unsigned long arg) +#ifdef CONFIG_COMPAT +AuStub(long, au_rdu_compat_ioctl, return -EINVAL, struct file *file, + unsigned int cmd, unsigned long arg) +#endif +#endif + +#endif /* __KERNEL__ */ +#endif /* __AUFS_DIR_H__ */ diff --git a/fs/aufs/dynop.c b/fs/aufs/dynop.c new file mode 100644 index 000000000..53a8b55d8 --- /dev/null +++ b/fs/aufs/dynop.c @@ -0,0 +1,356 @@ +/* + * Copyright (C) 2010-2016 Junjiro R. Okajima + */ + +/* + * dynamically customizable operations for regular files + */ + +#include "aufs.h" + +#define DyPrSym(key) AuDbgSym(key->dk_op.dy_hop) + +/* + * How large will these lists be? + * Usually just a few elements, 20-30 at most for each, I guess. + */ +static struct au_splhead dynop[AuDyLast]; + +static struct au_dykey *dy_gfind_get(struct au_splhead *spl, const void *h_op) +{ + struct au_dykey *key, *tmp; + struct list_head *head; + + key = NULL; + head = &spl->head; + rcu_read_lock(); + list_for_each_entry_rcu(tmp, head, dk_list) + if (tmp->dk_op.dy_hop == h_op) { + key = tmp; + kref_get(&key->dk_kref); + break; + } + rcu_read_unlock(); + + return key; +} + +static struct au_dykey *dy_bradd(struct au_branch *br, struct au_dykey *key) +{ + struct au_dykey **k, *found; + const void *h_op = key->dk_op.dy_hop; + int i; + + found = NULL; + k = br->br_dykey; + for (i = 0; i < AuBrDynOp; i++) + if (k[i]) { + if (k[i]->dk_op.dy_hop == h_op) { + found = k[i]; + break; + } + } else + break; + if (!found) { + spin_lock(&br->br_dykey_lock); + for (; i < AuBrDynOp; i++) + if (k[i]) { + if (k[i]->dk_op.dy_hop == h_op) { + found = k[i]; + break; + } + } else { + k[i] = key; + break; + } + spin_unlock(&br->br_dykey_lock); + BUG_ON(i == AuBrDynOp); /* expand the array */ + } + + return found; +} + +/* kref_get() if @key is already added */ +static struct au_dykey *dy_gadd(struct au_splhead *spl, struct au_dykey *key) +{ + struct au_dykey *tmp, *found; + struct list_head *head; + const void *h_op = key->dk_op.dy_hop; + + found = NULL; + head = &spl->head; + spin_lock(&spl->spin); + list_for_each_entry(tmp, head, dk_list) + if (tmp->dk_op.dy_hop == h_op) { + kref_get(&tmp->dk_kref); + found = tmp; + break; + } + if (!found) + list_add_rcu(&key->dk_list, head); + spin_unlock(&spl->spin); + + if (!found) + DyPrSym(key); + return found; +} + +static void dy_free_rcu(struct rcu_head *rcu) +{ + struct au_dykey *key; + + key = container_of(rcu, struct au_dykey, dk_rcu); + DyPrSym(key); + kfree(key); +} + +static void dy_free(struct kref *kref) +{ + struct au_dykey *key; + struct au_splhead *spl; + + key = container_of(kref, struct au_dykey, dk_kref); + spl = dynop + key->dk_op.dy_type; + au_spl_del_rcu(&key->dk_list, spl); + call_rcu(&key->dk_rcu, dy_free_rcu); +} + +void au_dy_put(struct au_dykey *key) +{ + kref_put(&key->dk_kref, dy_free); +} + +/* ---------------------------------------------------------------------- */ + +#define DyDbgSize(cnt, op) AuDebugOn(cnt != sizeof(op)/sizeof(void *)) + +#ifdef CONFIG_AUFS_DEBUG +#define DyDbgDeclare(cnt) unsigned int cnt = 0 +#define DyDbgInc(cnt) do { cnt++; } while (0) +#else +#define DyDbgDeclare(cnt) do {} while (0) +#define DyDbgInc(cnt) do {} while (0) +#endif + +#define DySet(func, dst, src, h_op, h_sb) do { \ + DyDbgInc(cnt); \ + if (h_op->func) { \ + if (src.func) \ + dst.func = src.func; \ + else \ + AuDbg("%s %s\n", au_sbtype(h_sb), #func); \ + } \ +} while (0) + +#define DySetForce(func, dst, src) do { \ + AuDebugOn(!src.func); \ + DyDbgInc(cnt); \ + dst.func = src.func; \ +} while (0) + +#define DySetAop(func) \ + DySet(func, dyaop->da_op, aufs_aop, h_aop, h_sb) +#define DySetAopForce(func) \ + DySetForce(func, dyaop->da_op, aufs_aop) + +static void dy_aop(struct au_dykey *key, const void *h_op, + struct super_block *h_sb __maybe_unused) +{ + struct au_dyaop *dyaop = (void *)key; + const struct address_space_operations *h_aop = h_op; + DyDbgDeclare(cnt); + + AuDbg("%s\n", au_sbtype(h_sb)); + + DySetAop(writepage); + DySetAopForce(readpage); /* force */ + DySetAop(writepages); + DySetAop(set_page_dirty); + DySetAop(readpages); + DySetAop(write_begin); + DySetAop(write_end); + DySetAop(bmap); + DySetAop(invalidatepage); + DySetAop(releasepage); + DySetAop(freepage); + /* this one will be changed according to an aufs mount option */ + DySetAop(direct_IO); + DySetAop(migratepage); + DySetAop(launder_page); + DySetAop(is_partially_uptodate); + DySetAop(is_dirty_writeback); + DySetAop(error_remove_page); + DySetAop(swap_activate); + DySetAop(swap_deactivate); + + DyDbgSize(cnt, *h_aop); +} + +/* ---------------------------------------------------------------------- */ + +static void dy_bug(struct kref *kref) +{ + BUG(); +} + +static struct au_dykey *dy_get(struct au_dynop *op, struct au_branch *br) +{ + struct au_dykey *key, *old; + struct au_splhead *spl; + struct op { + unsigned int sz; + void (*set)(struct au_dykey *key, const void *h_op, + struct super_block *h_sb __maybe_unused); + }; + static const struct op a[] = { + [AuDy_AOP] = { + .sz = sizeof(struct au_dyaop), + .set = dy_aop + } + }; + const struct op *p; + + spl = dynop + op->dy_type; + key = dy_gfind_get(spl, op->dy_hop); + if (key) + goto out_add; /* success */ + + p = a + op->dy_type; + key = kzalloc(p->sz, GFP_NOFS); + if (unlikely(!key)) { + key = ERR_PTR(-ENOMEM); + goto out; + } + + key->dk_op.dy_hop = op->dy_hop; + kref_init(&key->dk_kref); + p->set(key, op->dy_hop, au_br_sb(br)); + old = dy_gadd(spl, key); + if (old) { + kfree(key); + key = old; + } + +out_add: + old = dy_bradd(br, key); + if (old) + /* its ref-count should never be zero here */ + kref_put(&key->dk_kref, dy_bug); +out: + return key; +} + +/* ---------------------------------------------------------------------- */ +/* + * Aufs prohibits O_DIRECT by defaut even if the branch supports it. + * This behaviour is necessary to return an error from open(O_DIRECT) instead + * of the succeeding I/O. The dio mount option enables O_DIRECT and makes + * open(O_DIRECT) always succeed, but the succeeding I/O may return an error. + * See the aufs manual in detail. + */ +static void dy_adx(struct au_dyaop *dyaop, int do_dx) +{ + if (!do_dx) + dyaop->da_op.direct_IO = NULL; + else + dyaop->da_op.direct_IO = aufs_aop.direct_IO; +} + +static struct au_dyaop *dy_aget(struct au_branch *br, + const struct address_space_operations *h_aop, + int do_dx) +{ + struct au_dyaop *dyaop; + struct au_dynop op; + + op.dy_type = AuDy_AOP; + op.dy_haop = h_aop; + dyaop = (void *)dy_get(&op, br); + if (IS_ERR(dyaop)) + goto out; + dy_adx(dyaop, do_dx); + +out: + return dyaop; +} + +int au_dy_iaop(struct inode *inode, aufs_bindex_t bindex, + struct inode *h_inode) +{ + int err, do_dx; + struct super_block *sb; + struct au_branch *br; + struct au_dyaop *dyaop; + + AuDebugOn(!S_ISREG(h_inode->i_mode)); + IiMustWriteLock(inode); + + sb = inode->i_sb; + br = au_sbr(sb, bindex); + do_dx = !!au_opt_test(au_mntflags(sb), DIO); + dyaop = dy_aget(br, h_inode->i_mapping->a_ops, do_dx); + err = PTR_ERR(dyaop); + if (IS_ERR(dyaop)) + /* unnecessary to call dy_fput() */ + goto out; + + err = 0; + inode->i_mapping->a_ops = &dyaop->da_op; + +out: + return err; +} + +/* + * Is it safe to replace a_ops during the inode/file is in operation? + * Yes, I hope so. + */ +int au_dy_irefresh(struct inode *inode) +{ + int err; + aufs_bindex_t bstart; + struct inode *h_inode; + + err = 0; + if (S_ISREG(inode->i_mode)) { + bstart = au_ibstart(inode); + h_inode = au_h_iptr(inode, bstart); + err = au_dy_iaop(inode, bstart, h_inode); + } + return err; +} + +void au_dy_arefresh(int do_dx) +{ + struct au_splhead *spl; + struct list_head *head; + struct au_dykey *key; + + spl = dynop + AuDy_AOP; + head = &spl->head; + spin_lock(&spl->spin); + list_for_each_entry(key, head, dk_list) + dy_adx((void *)key, do_dx); + spin_unlock(&spl->spin); +} + +/* ---------------------------------------------------------------------- */ + +void __init au_dy_init(void) +{ + int i; + + /* make sure that 'struct au_dykey *' can be any type */ + BUILD_BUG_ON(offsetof(struct au_dyaop, da_key)); + + for (i = 0; i < AuDyLast; i++) + au_spl_init(dynop + i); +} + +void au_dy_fin(void) +{ + int i; + + for (i = 0; i < AuDyLast; i++) + WARN_ON(!list_empty(&dynop[i].head)); +} diff --git a/fs/aufs/dynop.h b/fs/aufs/dynop.h new file mode 100644 index 000000000..8680bfc53 --- /dev/null +++ b/fs/aufs/dynop.h @@ -0,0 +1,61 @@ +/* + * Copyright (C) 2010-2016 Junjiro R. Okajima + */ + +/* + * dynamically customizable operations (for regular files only) + */ + +#ifndef __AUFS_DYNOP_H__ +#define __AUFS_DYNOP_H__ + +#ifdef __KERNEL__ + +#include <linux/fs.h> +#include <linux/kref.h> + +enum {AuDy_AOP, AuDyLast}; + +struct au_dynop { + int dy_type; + union { + const void *dy_hop; + const struct address_space_operations *dy_haop; + }; +}; + +struct au_dykey { + union { + struct list_head dk_list; + struct rcu_head dk_rcu; + }; + struct au_dynop dk_op; + + /* + * during I am in the branch local array, kref is gotten. when the + * branch is removed, kref is put. + */ + struct kref dk_kref; +}; + +/* stop unioning since their sizes are very different from each other */ +struct au_dyaop { + struct au_dykey da_key; + struct address_space_operations da_op; /* not const */ +}; + +/* ---------------------------------------------------------------------- */ + +/* dynop.c */ +struct au_branch; +void au_dy_put(struct au_dykey *key); +int au_dy_iaop(struct inode *inode, aufs_bindex_t bindex, + struct inode *h_inode); +int au_dy_irefresh(struct inode *inode); +void au_dy_arefresh(int do_dio); + +void __init au_dy_init(void); +void au_dy_fin(void); + +#endif /* __KERNEL__ */ +#endif /* __AUFS_DYNOP_H__ */ diff --git a/fs/aufs/export.c b/fs/aufs/export.c new file mode 100644 index 000000000..69b70f2a1 --- /dev/null +++ b/fs/aufs/export.c @@ -0,0 +1,817 @@ +/* + * Copyright (C) 2005-2016 Junjiro R. Okajima + */ + +/* + * export via nfs + */ + +#include <linux/exportfs.h> +#include <linux/fs_struct.h> +#include <linux/namei.h> +#include <linux/nsproxy.h> +#include <linux/random.h> +#include <linux/writeback.h> +#include "../fs/mount.h" +#include "aufs.h" + +union conv { +#ifdef CONFIG_AUFS_INO_T_64 + __u32 a[2]; +#else + __u32 a[1]; +#endif + ino_t ino; +}; + +static ino_t decode_ino(__u32 *a) +{ + union conv u; + + BUILD_BUG_ON(sizeof(u.ino) != sizeof(u.a)); + u.a[0] = a[0]; +#ifdef CONFIG_AUFS_INO_T_64 + u.a[1] = a[1]; +#endif + return u.ino; +} + +static void encode_ino(__u32 *a, ino_t ino) +{ + union conv u; + + u.ino = ino; + a[0] = u.a[0]; +#ifdef CONFIG_AUFS_INO_T_64 + a[1] = u.a[1]; +#endif +} + +/* NFS file handle */ +enum { + Fh_br_id, + Fh_sigen, +#ifdef CONFIG_AUFS_INO_T_64 + /* support 64bit inode number */ + Fh_ino1, + Fh_ino2, + Fh_dir_ino1, + Fh_dir_ino2, +#else + Fh_ino1, + Fh_dir_ino1, +#endif + Fh_igen, + Fh_h_type, + Fh_tail, + + Fh_ino = Fh_ino1, + Fh_dir_ino = Fh_dir_ino1 +}; + +static int au_test_anon(struct dentry *dentry) +{ + /* note: read d_flags without d_lock */ + return !!(dentry->d_flags & DCACHE_DISCONNECTED); +} + +int au_test_nfsd(void) +{ + int ret; + struct task_struct *tsk = current; + char comm[sizeof(tsk->comm)]; + + ret = 0; + if (tsk->flags & PF_KTHREAD) { + get_task_comm(comm, tsk); + ret = !strcmp(comm, "nfsd"); + } + + return ret; +} + +/* ---------------------------------------------------------------------- */ +/* inode generation external table */ + +void au_xigen_inc(struct inode *inode) +{ + loff_t pos; + ssize_t sz; + __u32 igen; + struct super_block *sb; + struct au_sbinfo *sbinfo; + + sb = inode->i_sb; + AuDebugOn(!au_opt_test(au_mntflags(sb), XINO)); + + sbinfo = au_sbi(sb); + pos = inode->i_ino; + pos *= sizeof(igen); + igen = inode->i_generation + 1; + sz = xino_fwrite(sbinfo->si_xwrite, sbinfo->si_xigen, &igen, + sizeof(igen), &pos); + if (sz == sizeof(igen)) + return; /* success */ + + if (unlikely(sz >= 0)) + AuIOErr("xigen error (%zd)\n", sz); +} + +int au_xigen_new(struct inode *inode) +{ + int err; + loff_t pos; + ssize_t sz; + struct super_block *sb; + struct au_sbinfo *sbinfo; + struct file *file; + + err = 0; + /* todo: dirty, at mount time */ + if (inode->i_ino == AUFS_ROOT_INO) + goto out; + sb = inode->i_sb; + SiMustAnyLock(sb); + if (unlikely(!au_opt_test(au_mntflags(sb), XINO))) + goto out; + + err = -EFBIG; + pos = inode->i_ino; + if (unlikely(au_loff_max / sizeof(inode->i_generation) - 1 < pos)) { + AuIOErr1("too large i%lld\n", pos); + goto out; + } + pos *= sizeof(inode->i_generation); + + err = 0; + sbinfo = au_sbi(sb); + file = sbinfo->si_xigen; + BUG_ON(!file); + + if (vfsub_f_size_read(file) + < pos + sizeof(inode->i_generation)) { + inode->i_generation = atomic_inc_return(&sbinfo->si_xigen_next); + sz = xino_fwrite(sbinfo->si_xwrite, file, &inode->i_generation, + sizeof(inode->i_generation), &pos); + } else + sz = xino_fread(sbinfo->si_xread, file, &inode->i_generation, + sizeof(inode->i_generation), &pos); + if (sz == sizeof(inode->i_generation)) + goto out; /* success */ + + err = sz; + if (unlikely(sz >= 0)) { + err = -EIO; + AuIOErr("xigen error (%zd)\n", sz); + } + +out: + return err; +} + +int au_xigen_set(struct super_block *sb, struct file *base) +{ + int err; + struct au_sbinfo *sbinfo; + struct file *file; + + SiMustWriteLock(sb); + + sbinfo = au_sbi(sb); + file = au_xino_create2(base, sbinfo->si_xigen); + err = PTR_ERR(file); + if (IS_ERR(file)) + goto out; + err = 0; + if (sbinfo->si_xigen) + fput(sbinfo->si_xigen); + sbinfo->si_xigen = file; + +out: + return err; +} + +void au_xigen_clr(struct super_block *sb) +{ + struct au_sbinfo *sbinfo; + + SiMustWriteLock(sb); + + sbinfo = au_sbi(sb); + if (sbinfo->si_xigen) { + fput(sbinfo->si_xigen); + sbinfo->si_xigen = NULL; + } +} + +/* ---------------------------------------------------------------------- */ + +static struct dentry *decode_by_ino(struct super_block *sb, ino_t ino, + ino_t dir_ino) +{ + struct dentry *dentry, *d; + struct inode *inode; + unsigned int sigen; + + dentry = NULL; + inode = ilookup(sb, ino); + if (!inode) + goto out; + + dentry = ERR_PTR(-ESTALE); + sigen = au_sigen(sb); + if (unlikely(is_bad_inode(inode) + || IS_DEADDIR(inode) + || sigen != au_iigen(inode, NULL))) + goto out_iput; + + dentry = NULL; + if (!dir_ino || S_ISDIR(inode->i_mode)) + dentry = d_find_alias(inode); + else { + spin_lock(&inode->i_lock); + hlist_for_each_entry(d, &inode->i_dentry, d_u.d_alias) { + spin_lock(&d->d_lock); + if (!au_test_anon(d) + && d_inode(d->d_parent)->i_ino == dir_ino) { + dentry = dget_dlock(d); + spin_unlock(&d->d_lock); + break; + } + spin_unlock(&d->d_lock); + } + spin_unlock(&inode->i_lock); + } + if (unlikely(dentry && au_digen_test(dentry, sigen))) { + /* need to refresh */ + dput(dentry); + dentry = NULL; + } + +out_iput: + iput(inode); +out: + AuTraceErrPtr(dentry); + return dentry; +} + +/* ---------------------------------------------------------------------- */ + +/* todo: dirty? */ +/* if exportfs_decode_fh() passed vfsmount*, we could be happy */ + +struct au_compare_mnt_args { + /* input */ + struct super_block *sb; + + /* output */ + struct vfsmount *mnt; +}; + +static int au_compare_mnt(struct vfsmount *mnt, void *arg) +{ + struct au_compare_mnt_args *a = arg; + + if (mnt->mnt_sb != a->sb) + return 0; + a->mnt = mntget(mnt); + return 1; +} + +static struct vfsmount *au_mnt_get(struct super_block *sb) +{ + int err; + struct path root; + struct au_compare_mnt_args args = { + .sb = sb + }; + + get_fs_root(current->fs, &root); + rcu_read_lock(); + err = iterate_mounts(au_compare_mnt, &args, root.mnt); + rcu_read_unlock(); + path_put(&root); + AuDebugOn(!err); + AuDebugOn(!args.mnt); + return args.mnt; +} + +struct au_nfsd_si_lock { + unsigned int sigen; + aufs_bindex_t bindex, br_id; + unsigned char force_lock; +}; + +static int si_nfsd_read_lock(struct super_block *sb, + struct au_nfsd_si_lock *nsi_lock) +{ + int err; + aufs_bindex_t bindex; + + si_read_lock(sb, AuLock_FLUSH); + + /* branch id may be wrapped around */ + err = 0; + bindex = au_br_index(sb, nsi_lock->br_id); + if (bindex >= 0 && nsi_lock->sigen + AUFS_BRANCH_MAX > au_sigen(sb)) + goto out; /* success */ + + err = -ESTALE; + bindex = -1; + if (!nsi_lock->force_lock) + si_read_unlock(sb); + +out: + nsi_lock->bindex = bindex; + return err; +} + +struct find_name_by_ino { + struct dir_context ctx; + int called, found; + ino_t ino; + char *name; + int namelen; +}; + +static int +find_name_by_ino(struct dir_context *ctx, const char *name, int namelen, + loff_t offset, u64 ino, unsigned int d_type) +{ + struct find_name_by_ino *a = container_of(ctx, struct find_name_by_ino, + ctx); + + a->called++; + if (a->ino != ino) + return 0; + + memcpy(a->name, name, namelen); + a->namelen = namelen; + a->found = 1; + return 1; +} + +static struct dentry *au_lkup_by_ino(struct path *path, ino_t ino, + struct au_nfsd_si_lock *nsi_lock) +{ + struct dentry *dentry, *parent; + struct file *file; + struct inode *dir; + struct find_name_by_ino arg = { + .ctx = { + .actor = find_name_by_ino + } + }; + int err; + + parent = path->dentry; + if (nsi_lock) + si_read_unlock(parent->d_sb); + file = vfsub_dentry_open(path, au_dir_roflags); + dentry = (void *)file; + if (IS_ERR(file)) + goto out; + + dentry = ERR_PTR(-ENOMEM); + arg.name = (void *)__get_free_page(GFP_NOFS); + if (unlikely(!arg.name)) + goto out_file; + arg.ino = ino; + arg.found = 0; + do { + arg.called = 0; + /* smp_mb(); */ + err = vfsub_iterate_dir(file, &arg.ctx); + } while (!err && !arg.found && arg.called); + dentry = ERR_PTR(err); + if (unlikely(err)) + goto out_name; + /* instead of ENOENT */ + dentry = ERR_PTR(-ESTALE); + if (!arg.found) + goto out_name; + + /* do not call vfsub_lkup_one() */ + dir = d_inode(parent); + dentry = vfsub_lookup_one_len_unlocked(arg.name, parent, arg.namelen); + AuTraceErrPtr(dentry); + if (IS_ERR(dentry)) + goto out_name; + AuDebugOn(au_test_anon(dentry)); + if (unlikely(d_really_is_negative(dentry))) { + dput(dentry); + dentry = ERR_PTR(-ENOENT); + } + +out_name: + free_page((unsigned long)arg.name); +out_file: + fput(file); +out: + if (unlikely(nsi_lock + && si_nfsd_read_lock(parent->d_sb, nsi_lock) < 0)) + if (!IS_ERR(dentry)) { + dput(dentry); + dentry = ERR_PTR(-ESTALE); + } + AuTraceErrPtr(dentry); + return dentry; +} + +static struct dentry *decode_by_dir_ino(struct super_block *sb, ino_t ino, + ino_t dir_ino, + struct au_nfsd_si_lock *nsi_lock) +{ + struct dentry *dentry; + struct path path; + + if (dir_ino != AUFS_ROOT_INO) { + path.dentry = decode_by_ino(sb, dir_ino, 0); + dentry = path.dentry; + if (!path.dentry || IS_ERR(path.dentry)) + goto out; + AuDebugOn(au_test_anon(path.dentry)); + } else + path.dentry = dget(sb->s_root); + + path.mnt = au_mnt_get(sb); + dentry = au_lkup_by_ino(&path, ino, nsi_lock); + path_put(&path); + +out: + AuTraceErrPtr(dentry); + return dentry; +} + +/* ---------------------------------------------------------------------- */ + +static int h_acceptable(void *expv, struct dentry *dentry) +{ + return 1; +} + +static char *au_build_path(struct dentry *h_parent, struct path *h_rootpath, + char *buf, int len, struct super_block *sb) +{ + char *p; + int n; + struct path path; + + p = d_path(h_rootpath, buf, len); + if (IS_ERR(p)) + goto out; + n = strlen(p); + + path.mnt = h_rootpath->mnt; + path.dentry = h_parent; + p = d_path(&path, buf, len); + if (IS_ERR(p)) + goto out; + if (n != 1) + p += n; + + path.mnt = au_mnt_get(sb); + path.dentry = sb->s_root; + p = d_path(&path, buf, len - strlen(p)); + mntput(path.mnt); + if (IS_ERR(p)) + goto out; + if (n != 1) + p[strlen(p)] = '/'; + +out: + AuTraceErrPtr(p); + return p; +} + +static +struct dentry *decode_by_path(struct super_block *sb, ino_t ino, __u32 *fh, + int fh_len, struct au_nfsd_si_lock *nsi_lock) +{ + struct dentry *dentry, *h_parent, *root; + struct super_block *h_sb; + char *pathname, *p; + struct vfsmount *h_mnt; + struct au_branch *br; + int err; + struct path path; + + br = au_sbr(sb, nsi_lock->bindex); + h_mnt = au_br_mnt(br); + h_sb = h_mnt->mnt_sb; + /* todo: call lower fh_to_dentry()? fh_to_parent()? */ + h_parent = exportfs_decode_fh(h_mnt, (void *)(fh + Fh_tail), + fh_len - Fh_tail, fh[Fh_h_type], + h_acceptable, /*context*/NULL); + dentry = h_parent; + if (unlikely(!h_parent || IS_ERR(h_parent))) { + AuWarn1("%s decode_fh failed, %ld\n", + au_sbtype(h_sb), PTR_ERR(h_parent)); + goto out; + } + dentry = NULL; + if (unlikely(au_test_anon(h_parent))) { + AuWarn1("%s decode_fh returned a disconnected dentry\n", + au_sbtype(h_sb)); + goto out_h_parent; + } + + dentry = ERR_PTR(-ENOMEM); + pathname = (void *)__get_free_page(GFP_NOFS); + if (unlikely(!pathname)) + goto out_h_parent; + + root = sb->s_root; + path.mnt = h_mnt; + di_read_lock_parent(root, !AuLock_IR); + path.dentry = au_h_dptr(root, nsi_lock->bindex); + di_read_unlock(root, !AuLock_IR); + p = au_build_path(h_parent, &path, pathname, PAGE_SIZE, sb); + dentry = (void *)p; + if (IS_ERR(p)) + goto out_pathname; + + si_read_unlock(sb); + err = vfsub_kern_path(p, LOOKUP_FOLLOW | LOOKUP_DIRECTORY, &path); + dentry = ERR_PTR(err); + if (unlikely(err)) + goto out_relock; + + dentry = ERR_PTR(-ENOENT); + AuDebugOn(au_test_anon(path.dentry)); + if (unlikely(d_really_is_negative(path.dentry))) + goto out_path; + + if (ino != d_inode(path.dentry)->i_ino) + dentry = au_lkup_by_ino(&path, ino, /*nsi_lock*/NULL); + else + dentry = dget(path.dentry); + +out_path: + path_put(&path); +out_relock: + if (unlikely(si_nfsd_read_lock(sb, nsi_lock) < 0)) + if (!IS_ERR(dentry)) { + dput(dentry); + dentry = ERR_PTR(-ESTALE); + } +out_pathname: + free_page((unsigned long)pathname); +out_h_parent: + dput(h_parent); +out: + AuTraceErrPtr(dentry); + return dentry; +} + +/* ---------------------------------------------------------------------- */ + +static struct dentry * +aufs_fh_to_dentry(struct super_block *sb, struct fid *fid, int fh_len, + int fh_type) +{ + struct dentry *dentry; + __u32 *fh = fid->raw; + struct au_branch *br; + ino_t ino, dir_ino; + struct au_nfsd_si_lock nsi_lock = { + .force_lock = 0 + }; + + dentry = ERR_PTR(-ESTALE); + /* it should never happen, but the file handle is unreliable */ + if (unlikely(fh_len < Fh_tail)) + goto out; + nsi_lock.sigen = fh[Fh_sigen]; + nsi_lock.br_id = fh[Fh_br_id]; + + /* branch id may be wrapped around */ + br = NULL; + if (unlikely(si_nfsd_read_lock(sb, &nsi_lock))) + goto out; + nsi_lock.force_lock = 1; + + /* is this inode still cached? */ + ino = decode_ino(fh + Fh_ino); + /* it should never happen */ + if (unlikely(ino == AUFS_ROOT_INO)) + goto out_unlock; + + dir_ino = decode_ino(fh + Fh_dir_ino); + dentry = decode_by_ino(sb, ino, dir_ino); + if (IS_ERR(dentry)) + goto out_unlock; + if (dentry) + goto accept; + + /* is the parent dir cached? */ + br = au_sbr(sb, nsi_lock.bindex); + atomic_inc(&br->br_count); + dentry = decode_by_dir_ino(sb, ino, dir_ino, &nsi_lock); + if (IS_ERR(dentry)) + goto out_unlock; + if (dentry) + goto accept; + + /* lookup path */ + dentry = decode_by_path(sb, ino, fh, fh_len, &nsi_lock); + if (IS_ERR(dentry)) + goto out_unlock; + if (unlikely(!dentry)) + /* todo?: make it ESTALE */ + goto out_unlock; + +accept: + if (!au_digen_test(dentry, au_sigen(sb)) + && d_inode(dentry)->i_generation == fh[Fh_igen]) + goto out_unlock; /* success */ + + dput(dentry); + dentry = ERR_PTR(-ESTALE); +out_unlock: + if (br) + atomic_dec(&br->br_count); + si_read_unlock(sb); +out: + AuTraceErrPtr(dentry); + return dentry; +} + +#if 0 /* reserved for future use */ +/* support subtreecheck option */ +static struct dentry *aufs_fh_to_parent(struct super_block *sb, struct fid *fid, + int fh_len, int fh_type) +{ + struct dentry *parent; + __u32 *fh = fid->raw; + ino_t dir_ino; + + dir_ino = decode_ino(fh + Fh_dir_ino); + parent = decode_by_ino(sb, dir_ino, 0); + if (IS_ERR(parent)) + goto out; + if (!parent) + parent = decode_by_path(sb, au_br_index(sb, fh[Fh_br_id]), + dir_ino, fh, fh_len); + +out: + AuTraceErrPtr(parent); + return parent; +} +#endif + +/* ---------------------------------------------------------------------- */ + +static int aufs_encode_fh(struct inode *inode, __u32 *fh, int *max_len, + struct inode *dir) +{ + int err; + aufs_bindex_t bindex; + struct super_block *sb, *h_sb; + struct dentry *dentry, *parent, *h_parent; + struct inode *h_dir; + struct au_branch *br; + + err = -ENOSPC; + if (unlikely(*max_len <= Fh_tail)) { + AuWarn1("NFSv2 client (max_len %d)?\n", *max_len); + goto out; + } + + err = FILEID_ROOT; + if (inode->i_ino == AUFS_ROOT_INO) { + AuDebugOn(inode->i_ino != AUFS_ROOT_INO); + goto out; + } + + h_parent = NULL; + sb = inode->i_sb; + err = si_read_lock(sb, AuLock_FLUSH); + if (unlikely(err)) + goto out; + +#ifdef CONFIG_AUFS_DEBUG + if (unlikely(!au_opt_test(au_mntflags(sb), XINO))) + AuWarn1("NFS-exporting requires xino\n"); +#endif + err = -EIO; + parent = NULL; + ii_read_lock_child(inode); + bindex = au_ibstart(inode); + if (!dir) { + dentry = d_find_any_alias(inode); + if (unlikely(!dentry)) + goto out_unlock; + AuDebugOn(au_test_anon(dentry)); + parent = dget_parent(dentry); + dput(dentry); + if (unlikely(!parent)) + goto out_unlock; + if (d_really_is_positive(parent)) + dir = d_inode(parent); + } + + ii_read_lock_parent(dir); + h_dir = au_h_iptr(dir, bindex); + ii_read_unlock(dir); + if (unlikely(!h_dir)) + goto out_parent; + h_parent = d_find_any_alias(h_dir); + if (unlikely(!h_parent)) + goto out_hparent; + + err = -EPERM; + br = au_sbr(sb, bindex); + h_sb = au_br_sb(br); + if (unlikely(!h_sb->s_export_op)) { + AuErr1("%s branch is not exportable\n", au_sbtype(h_sb)); + goto out_hparent; + } + + fh[Fh_br_id] = br->br_id; + fh[Fh_sigen] = au_sigen(sb); + encode_ino(fh + Fh_ino, inode->i_ino); + encode_ino(fh + Fh_dir_ino, dir->i_ino); + fh[Fh_igen] = inode->i_generation; + + *max_len -= Fh_tail; + fh[Fh_h_type] = exportfs_encode_fh(h_parent, (void *)(fh + Fh_tail), + max_len, + /*connectable or subtreecheck*/0); + err = fh[Fh_h_type]; + *max_len += Fh_tail; + /* todo: macros? */ + if (err != FILEID_INVALID) + err = 99; + else + AuWarn1("%s encode_fh failed\n", au_sbtype(h_sb)); + +out_hparent: + dput(h_parent); +out_parent: + dput(parent); +out_unlock: + ii_read_unlock(inode); + si_read_unlock(sb); +out: + if (unlikely(err < 0)) + err = FILEID_INVALID; + return err; +} + +/* ---------------------------------------------------------------------- */ + +static int aufs_commit_metadata(struct inode *inode) +{ + int err; + aufs_bindex_t bindex; + struct super_block *sb; + struct inode *h_inode; + int (*f)(struct inode *inode); + + sb = inode->i_sb; + si_read_lock(sb, AuLock_FLUSH | AuLock_NOPLMW); + ii_write_lock_child(inode); + bindex = au_ibstart(inode); + AuDebugOn(bindex < 0); + h_inode = au_h_iptr(inode, bindex); + + f = h_inode->i_sb->s_export_op->commit_metadata; + if (f) + err = f(h_inode); + else { + struct writeback_control wbc = { + .sync_mode = WB_SYNC_ALL, + .nr_to_write = 0 /* metadata only */ + }; + + err = sync_inode(h_inode, &wbc); + } + + au_cpup_attr_timesizes(inode); + ii_write_unlock(inode); + si_read_unlock(sb); + return err; +} + +/* ---------------------------------------------------------------------- */ + +static struct export_operations aufs_export_op = { + .fh_to_dentry = aufs_fh_to_dentry, + /* .fh_to_parent = aufs_fh_to_parent, */ + .encode_fh = aufs_encode_fh, + .commit_metadata = aufs_commit_metadata +}; + +void au_export_init(struct super_block *sb) +{ + struct au_sbinfo *sbinfo; + __u32 u; + + sb->s_export_op = &aufs_export_op; + sbinfo = au_sbi(sb); + sbinfo->si_xigen = NULL; + get_random_bytes(&u, sizeof(u)); + BUILD_BUG_ON(sizeof(u) != sizeof(int)); + atomic_set(&sbinfo->si_xigen_next, u); +} diff --git a/fs/aufs/f_op.c b/fs/aufs/f_op.c new file mode 100644 index 000000000..ac1304a18 --- /dev/null +++ b/fs/aufs/f_op.c @@ -0,0 +1,757 @@ +/* + * Copyright (C) 2005-2016 Junjiro R. Okajima + */ + +/* + * file and vm operations + */ + +#include <linux/aio.h> +#include <linux/fs_stack.h> +#include <linux/mman.h> +#include <linux/security.h> +#include "aufs.h" + +int au_do_open_nondir(struct file *file, int flags, struct file *h_file) +{ + int err; + aufs_bindex_t bindex; + struct dentry *dentry, *h_dentry; + struct au_finfo *finfo; + struct inode *h_inode; + + FiMustWriteLock(file); + + err = 0; + dentry = file->f_path.dentry; + AuDebugOn(IS_ERR_OR_NULL(dentry)); + finfo = au_fi(file); + memset(&finfo->fi_htop, 0, sizeof(finfo->fi_htop)); + atomic_set(&finfo->fi_mmapped, 0); + bindex = au_dbstart(dentry); + if (!h_file) { + h_dentry = au_h_dptr(dentry, bindex); + err = vfsub_test_mntns(file->f_path.mnt, h_dentry->d_sb); + if (unlikely(err)) + goto out; + h_file = au_h_open(dentry, bindex, flags, file, /*force_wr*/0); + } else { + h_dentry = h_file->f_path.dentry; + err = vfsub_test_mntns(file->f_path.mnt, h_dentry->d_sb); + if (unlikely(err)) + goto out; + get_file(h_file); + } + if (IS_ERR(h_file)) + err = PTR_ERR(h_file); + else { + if ((flags & __O_TMPFILE) + && !(flags & O_EXCL)) { + h_inode = file_inode(h_file); + spin_lock(&h_inode->i_lock); + h_inode->i_state |= I_LINKABLE; + spin_unlock(&h_inode->i_lock); + } + au_set_fbstart(file, bindex); + au_set_h_fptr(file, bindex, h_file); + au_update_figen(file); + /* todo: necessary? */ + /* file->f_ra = h_file->f_ra; */ + } + +out: + return err; +} + +static int aufs_open_nondir(struct inode *inode __maybe_unused, + struct file *file) +{ + int err; + struct super_block *sb; + struct au_do_open_args args = { + .open = au_do_open_nondir + }; + + AuDbg("%pD, f_flags 0x%x, f_mode 0x%x\n", + file, vfsub_file_flags(file), file->f_mode); + + sb = file->f_path.dentry->d_sb; + si_read_lock(sb, AuLock_FLUSH); + err = au_do_open(file, &args); + si_read_unlock(sb); + return err; +} + +int aufs_release_nondir(struct inode *inode __maybe_unused, struct file *file) +{ + struct au_finfo *finfo; + aufs_bindex_t bindex; + + finfo = au_fi(file); + au_sphl_del(&finfo->fi_hlist, + &au_sbi(file->f_path.dentry->d_sb)->si_files); + bindex = finfo->fi_btop; + if (bindex >= 0) + au_set_h_fptr(file, bindex, NULL); + + au_finfo_fin(file); + return 0; +} + +/* ---------------------------------------------------------------------- */ + +static int au_do_flush_nondir(struct file *file, fl_owner_t id) +{ + int err; + struct file *h_file; + + err = 0; + h_file = au_hf_top(file); + if (h_file) + err = vfsub_flush(h_file, id); + return err; +} + +static int aufs_flush_nondir(struct file *file, fl_owner_t id) +{ + return au_do_flush(file, id, au_do_flush_nondir); +} + +/* ---------------------------------------------------------------------- */ +/* + * read and write functions acquire [fdi]_rwsem once, but release before + * mmap_sem. This is because to stop a race condition between mmap(2). + * Releasing these aufs-rwsem should be safe, no branch-mamagement (by keeping + * si_rwsem), no harmful copy-up should happen. Actually copy-up may happen in + * read functions after [fdi]_rwsem are released, but it should be harmless. + */ + +/* Callers should call au_read_post() or fput() in the end */ +struct file *au_read_pre(struct file *file, int keep_fi) +{ + struct file *h_file; + int err; + + err = au_reval_and_lock_fdi(file, au_reopen_nondir, /*wlock*/0); + if (!err) { + di_read_unlock(file->f_path.dentry, AuLock_IR); + h_file = au_hf_top(file); + get_file(h_file); + if (!keep_fi) + fi_read_unlock(file); + } else + h_file = ERR_PTR(err); + + return h_file; +} + +static void au_read_post(struct inode *inode, struct file *h_file) +{ + /* update without lock, I don't think it a problem */ + fsstack_copy_attr_atime(inode, file_inode(h_file)); + fput(h_file); +} + +struct au_write_pre { + blkcnt_t blks; + aufs_bindex_t bstart; +}; + +/* + * return with iinfo is write-locked + * callers should call au_write_post() or iinfo_write_unlock() + fput() in the + * end + */ +static struct file *au_write_pre(struct file *file, int do_ready, + struct au_write_pre *wpre) +{ + struct file *h_file; + struct dentry *dentry; + int err; + struct au_pin pin; + + err = au_reval_and_lock_fdi(file, au_reopen_nondir, /*wlock*/1); + h_file = ERR_PTR(err); + if (unlikely(err)) + goto out; + + dentry = file->f_path.dentry; + if (do_ready) { + err = au_ready_to_write(file, -1, &pin); + if (unlikely(err)) { + h_file = ERR_PTR(err); + di_write_unlock(dentry); + goto out_fi; + } + } + + di_downgrade_lock(dentry, /*flags*/0); + if (wpre) + wpre->bstart = au_fbstart(file); + h_file = au_hf_top(file); + get_file(h_file); + if (wpre) + wpre->blks = file_inode(h_file)->i_blocks; + if (do_ready) + au_unpin(&pin); + di_read_unlock(dentry, /*flags*/0); + +out_fi: + fi_write_unlock(file); +out: + return h_file; +} + +static void au_write_post(struct inode *inode, struct file *h_file, + struct au_write_pre *wpre, ssize_t written) +{ + struct inode *h_inode; + + au_cpup_attr_timesizes(inode); + AuDebugOn(au_ibstart(inode) != wpre->bstart); + h_inode = file_inode(h_file); + inode->i_mode = h_inode->i_mode; + ii_write_unlock(inode); + fput(h_file); + + /* AuDbg("blks %llu, %llu\n", (u64)blks, (u64)h_inode->i_blocks); */ + if (written > 0) + au_fhsm_wrote(inode->i_sb, wpre->bstart, + /*force*/h_inode->i_blocks > wpre->blks); +} + +static ssize_t aufs_read(struct file *file, char __user *buf, size_t count, + loff_t *ppos) +{ + ssize_t err; + struct inode *inode; + struct file *h_file; + struct super_block *sb; + + inode = file_inode(file); + sb = inode->i_sb; + si_read_lock(sb, AuLock_FLUSH | AuLock_NOPLMW); + + h_file = au_read_pre(file, /*keep_fi*/0); + err = PTR_ERR(h_file); + if (IS_ERR(h_file)) + goto out; + + /* filedata may be obsoleted by concurrent copyup, but no problem */ + err = vfsub_read_u(h_file, buf, count, ppos); + /* todo: necessary? */ + /* file->f_ra = h_file->f_ra; */ + au_read_post(inode, h_file); + +out: + si_read_unlock(sb); + return err; +} + +/* + * todo: very ugly + * it locks both of i_mutex and si_rwsem for read in safe. + * if the plink maintenance mode continues forever (that is the problem), + * may loop forever. + */ +static void au_mtx_and_read_lock(struct inode *inode) +{ + int err; + struct super_block *sb = inode->i_sb; + + while (1) { + inode_lock(inode); + err = si_read_lock(sb, AuLock_FLUSH | AuLock_NOPLM); + if (!err) + break; + inode_unlock(inode); + si_read_lock(sb, AuLock_NOPLMW); + si_read_unlock(sb); + } +} + +static ssize_t aufs_write(struct file *file, const char __user *ubuf, + size_t count, loff_t *ppos) +{ + ssize_t err; + struct au_write_pre wpre; + struct inode *inode; + struct file *h_file; + char __user *buf = (char __user *)ubuf; + + inode = file_inode(file); + au_mtx_and_read_lock(inode); + + h_file = au_write_pre(file, /*do_ready*/1, &wpre); + err = PTR_ERR(h_file); + if (IS_ERR(h_file)) + goto out; + + err = vfsub_write_u(h_file, buf, count, ppos); + au_write_post(inode, h_file, &wpre, err); + +out: + si_read_unlock(inode->i_sb); + inode_unlock(inode); + return err; +} + +static ssize_t au_do_iter(struct file *h_file, int rw, struct kiocb *kio, + struct iov_iter *iov_iter) +{ + ssize_t err; + struct file *file; + ssize_t (*iter)(struct kiocb *, struct iov_iter *); + + err = security_file_permission(h_file, rw); + if (unlikely(err)) + goto out; + + err = -ENOSYS; + iter = NULL; + if (rw == MAY_READ) + iter = h_file->f_op->read_iter; + else if (rw == MAY_WRITE) + iter = h_file->f_op->write_iter; + + file = kio->ki_filp; + kio->ki_filp = h_file; + if (iter) { + lockdep_off(); + err = iter(kio, iov_iter); + lockdep_on(); + } else + /* currently there is no such fs */ + WARN_ON_ONCE(1); + kio->ki_filp = file; + +out: + return err; +} + +static ssize_t aufs_read_iter(struct kiocb *kio, struct iov_iter *iov_iter) +{ + ssize_t err; + struct file *file, *h_file; + struct inode *inode; + struct super_block *sb; + + file = kio->ki_filp; + inode = file_inode(file); + sb = inode->i_sb; + si_read_lock(sb, AuLock_FLUSH | AuLock_NOPLMW); + + h_file = au_read_pre(file, /*keep_fi*/0); + err = PTR_ERR(h_file); + if (IS_ERR(h_file)) + goto out; + + err = au_do_iter(h_file, MAY_READ, kio, iov_iter); + /* todo: necessary? */ + /* file->f_ra = h_file->f_ra; */ + au_read_post(inode, h_file); + +out: + si_read_unlock(sb); + return err; +} + +static ssize_t aufs_write_iter(struct kiocb *kio, struct iov_iter *iov_iter) +{ + ssize_t err; + struct au_write_pre wpre; + struct inode *inode; + struct file *file, *h_file; + + file = kio->ki_filp; + inode = file_inode(file); + au_mtx_and_read_lock(inode); + + h_file = au_write_pre(file, /*do_ready*/1, &wpre); + err = PTR_ERR(h_file); + if (IS_ERR(h_file)) + goto out; + + err = au_do_iter(h_file, MAY_WRITE, kio, iov_iter); + au_write_post(inode, h_file, &wpre, err); + +out: + si_read_unlock(inode->i_sb); + inode_unlock(inode); + return err; +} + +static ssize_t aufs_splice_read(struct file *file, loff_t *ppos, + struct pipe_inode_info *pipe, size_t len, + unsigned int flags) +{ + ssize_t err; + struct file *h_file; + struct inode *inode; + struct super_block *sb; + + inode = file_inode(file); + sb = inode->i_sb; + si_read_lock(sb, AuLock_FLUSH | AuLock_NOPLMW); + + h_file = au_read_pre(file, /*keep_fi*/1); + err = PTR_ERR(h_file); + if (IS_ERR(h_file)) + goto out; + + if (au_test_loopback_kthread()) { + au_warn_loopback(h_file->f_path.dentry->d_sb); + if (file->f_mapping != h_file->f_mapping) { + file->f_mapping = h_file->f_mapping; + smp_mb(); /* unnecessary? */ + } + } + fi_read_unlock(file); + + err = vfsub_splice_to(h_file, ppos, pipe, len, flags); + /* todo: necessasry? */ + /* file->f_ra = h_file->f_ra; */ + au_read_post(inode, h_file); + +out: + si_read_unlock(sb); + return err; +} + +static ssize_t +aufs_splice_write(struct pipe_inode_info *pipe, struct file *file, loff_t *ppos, + size_t len, unsigned int flags) +{ + ssize_t err; + struct au_write_pre wpre; + struct inode *inode; + struct file *h_file; + + inode = file_inode(file); + au_mtx_and_read_lock(inode); + + h_file = au_write_pre(file, /*do_ready*/1, &wpre); + err = PTR_ERR(h_file); + if (IS_ERR(h_file)) + goto out; + + err = vfsub_splice_from(pipe, h_file, ppos, len, flags); + au_write_post(inode, h_file, &wpre, err); + +out: + si_read_unlock(inode->i_sb); + inode_unlock(inode); + return err; +} + +static long aufs_fallocate(struct file *file, int mode, loff_t offset, + loff_t len) +{ + long err; + struct au_write_pre wpre; + struct inode *inode; + struct file *h_file; + + inode = file_inode(file); + au_mtx_and_read_lock(inode); + + h_file = au_write_pre(file, /*do_ready*/1, &wpre); + err = PTR_ERR(h_file); + if (IS_ERR(h_file)) + goto out; + + lockdep_off(); + err = vfs_fallocate(h_file, mode, offset, len); + lockdep_on(); + au_write_post(inode, h_file, &wpre, /*written*/1); + +out: + si_read_unlock(inode->i_sb); + inode_unlock(inode); + return err; +} + +/* ---------------------------------------------------------------------- */ + +/* + * The locking order around current->mmap_sem. + * - in most and regular cases + * file I/O syscall -- aufs_read() or something + * -- si_rwsem for read -- mmap_sem + * (Note that [fdi]i_rwsem are released before mmap_sem). + * - in mmap case + * mmap(2) -- mmap_sem -- aufs_mmap() -- si_rwsem for read -- [fdi]i_rwsem + * This AB-BA order is definitly bad, but is not a problem since "si_rwsem for + * read" allows muliple processes to acquire it and [fdi]i_rwsem are not held in + * file I/O. Aufs needs to stop lockdep in aufs_mmap() though. + * It means that when aufs acquires si_rwsem for write, the process should never + * acquire mmap_sem. + * + * Actually aufs_iterate() holds [fdi]i_rwsem before mmap_sem, but this is not a + * problem either since any directory is not able to be mmap-ed. + * The similar scenario is applied to aufs_readlink() too. + */ + +#if 0 /* stop calling security_file_mmap() */ +/* cf. linux/include/linux/mman.h: calc_vm_prot_bits() */ +#define AuConv_VM_PROT(f, b) _calc_vm_trans(f, VM_##b, PROT_##b) + +static unsigned long au_arch_prot_conv(unsigned long flags) +{ + /* currently ppc64 only */ +#ifdef CONFIG_PPC64 + /* cf. linux/arch/powerpc/include/asm/mman.h */ + AuDebugOn(arch_calc_vm_prot_bits(-1) != VM_SAO); + return AuConv_VM_PROT(flags, SAO); +#else + AuDebugOn(arch_calc_vm_prot_bits(-1)); + return 0; +#endif +} + +static unsigned long au_prot_conv(unsigned long flags) +{ + return AuConv_VM_PROT(flags, READ) + | AuConv_VM_PROT(flags, WRITE) + | AuConv_VM_PROT(flags, EXEC) + | au_arch_prot_conv(flags); +} + +/* cf. linux/include/linux/mman.h: calc_vm_flag_bits() */ +#define AuConv_VM_MAP(f, b) _calc_vm_trans(f, VM_##b, MAP_##b) + +static unsigned long au_flag_conv(unsigned long flags) +{ + return AuConv_VM_MAP(flags, GROWSDOWN) + | AuConv_VM_MAP(flags, DENYWRITE) + | AuConv_VM_MAP(flags, LOCKED); +} +#endif + +static int aufs_mmap(struct file *file, struct vm_area_struct *vma) +{ + int err; + const unsigned char wlock + = (file->f_mode & FMODE_WRITE) && (vma->vm_flags & VM_SHARED); + struct super_block *sb; + struct file *h_file; + struct inode *inode; + + AuDbgVmRegion(file, vma); + + inode = file_inode(file); + sb = inode->i_sb; + lockdep_off(); + si_read_lock(sb, AuLock_NOPLMW); + + h_file = au_write_pre(file, wlock, /*wpre*/NULL); + lockdep_on(); + err = PTR_ERR(h_file); + if (IS_ERR(h_file)) + goto out; + + err = 0; + au_set_mmapped(file); + au_vm_file_reset(vma, h_file); + /* + * we cannot call security_mmap_file() here since it may acquire + * mmap_sem or i_mutex. + * + * err = security_mmap_file(h_file, au_prot_conv(vma->vm_flags), + * au_flag_conv(vma->vm_flags)); + */ + if (!err) + err = h_file->f_op->mmap(h_file, vma); + if (!err) { + au_vm_prfile_set(vma, file); + fsstack_copy_attr_atime(inode, file_inode(h_file)); + goto out_fput; /* success */ + } + au_unset_mmapped(file); + au_vm_file_reset(vma, file); + +out_fput: + lockdep_off(); + ii_write_unlock(inode); + lockdep_on(); + fput(h_file); +out: + lockdep_off(); + si_read_unlock(sb); + lockdep_on(); + AuTraceErr(err); + return err; +} + +/* ---------------------------------------------------------------------- */ + +static int aufs_fsync_nondir(struct file *file, loff_t start, loff_t end, + int datasync) +{ + int err; + struct au_write_pre wpre; + struct inode *inode; + struct file *h_file; + + err = 0; /* -EBADF; */ /* posix? */ + if (unlikely(!(file->f_mode & FMODE_WRITE))) + goto out; + + inode = file_inode(file); + au_mtx_and_read_lock(inode); + + h_file = au_write_pre(file, /*do_ready*/1, &wpre); + err = PTR_ERR(h_file); + if (IS_ERR(h_file)) + goto out_unlock; + + err = vfsub_fsync(h_file, &h_file->f_path, datasync); + au_write_post(inode, h_file, &wpre, /*written*/0); + +out_unlock: + si_read_unlock(inode->i_sb); + inode_unlock(inode); +out: + return err; +} + +/* no one supports this operation, currently */ +#if 0 +static int aufs_aio_fsync_nondir(struct kiocb *kio, int datasync) +{ + int err; + struct au_write_pre wpre; + struct inode *inode, *h_inode; + struct file *file, *h_file; + + err = 0; /* -EBADF; */ /* posix? */ + if (unlikely(!(file->f_mode & FMODE_WRITE))) + goto out; + + file = kio->ki_filp; + inode = file_inode(file); + au_mtx_and_read_lock(inode); + + h_file = au_write_pre(file, /*do_ready*/1, &wpre); + err = PTR_ERR(h_file); + if (IS_ERR(h_file)) + goto out_unlock; + + err = -ENOSYS; + h_file = au_hf_top(file); + if (h_file->f_op->aio_fsync) { + h_inode = file_inode(h_file); + if (!is_sync_kiocb(kio)) { + get_file(h_file); + fput(file); + } + kio->ki_filp = h_file; + err = h_file->f_op->aio_fsync(kio, datasync); + inode_lock_nested(h_inode, AuLsc_I_CHILD); + if (!err) + vfsub_update_h_iattr(&h_file->f_path, /*did*/NULL); + /*ignore*/ + inode_unlock(h_inode); + } + au_write_post(inode, h_file, &wpre, /*written*/0); + +out_unlock: + si_read_unlock(inode->sb); + inode_unlock(inode); +out: + return err; +} +#endif + +static int aufs_fasync(int fd, struct file *file, int flag) +{ + int err; + struct file *h_file; + struct super_block *sb; + + sb = file->f_path.dentry->d_sb; + si_read_lock(sb, AuLock_FLUSH | AuLock_NOPLMW); + + h_file = au_read_pre(file, /*keep_fi*/0); + err = PTR_ERR(h_file); + if (IS_ERR(h_file)) + goto out; + + if (h_file->f_op->fasync) + err = h_file->f_op->fasync(fd, h_file, flag); + fput(h_file); /* instead of au_read_post() */ + +out: + si_read_unlock(sb); + return err; +} + +static int aufs_setfl(struct file *file, unsigned long arg) +{ + int err; + struct file *h_file; + struct super_block *sb; + + sb = file->f_path.dentry->d_sb; + si_read_lock(sb, AuLock_FLUSH | AuLock_NOPLMW); + + h_file = au_read_pre(file, /*keep_fi*/0); + err = PTR_ERR(h_file); + if (IS_ERR(h_file)) + goto out; + + arg |= vfsub_file_flags(file) & FASYNC; /* stop calling h_file->fasync */ + err = setfl(/*unused fd*/-1, h_file, arg); + fput(h_file); /* instead of au_read_post() */ + +out: + si_read_unlock(sb); + return err; +} + +/* ---------------------------------------------------------------------- */ + +/* no one supports this operation, currently */ +#if 0 +static ssize_t aufs_sendpage(struct file *file, struct page *page, int offset, + size_t len, loff_t *pos, int more) +{ +} +#endif + +/* ---------------------------------------------------------------------- */ + +const struct file_operations aufs_file_fop = { + .owner = THIS_MODULE, + + .llseek = default_llseek, + + .read = aufs_read, + .write = aufs_write, + .read_iter = aufs_read_iter, + .write_iter = aufs_write_iter, + +#ifdef CONFIG_AUFS_POLL + .poll = aufs_poll, +#endif + .unlocked_ioctl = aufs_ioctl_nondir, +#ifdef CONFIG_COMPAT + .compat_ioctl = aufs_compat_ioctl_nondir, +#endif + .mmap = aufs_mmap, + .open = aufs_open_nondir, + .flush = aufs_flush_nondir, + .release = aufs_release_nondir, + .fsync = aufs_fsync_nondir, + /* .aio_fsync = aufs_aio_fsync_nondir, */ + .fasync = aufs_fasync, + /* .sendpage = aufs_sendpage, */ + .setfl = aufs_setfl, + .splice_write = aufs_splice_write, + .splice_read = aufs_splice_read, +#if 0 + .aio_splice_write = aufs_aio_splice_write, + .aio_splice_read = aufs_aio_splice_read, +#endif + .fallocate = aufs_fallocate +}; diff --git a/fs/aufs/fhsm.c b/fs/aufs/fhsm.c new file mode 100644 index 000000000..db079d6ee --- /dev/null +++ b/fs/aufs/fhsm.c @@ -0,0 +1,412 @@ +/* + * Copyright (C) 2011-2016 Junjiro R. Okajima + */ + +/* + * File-based Hierarchy Storage Management + */ + +#include <linux/anon_inodes.h> +#include <linux/poll.h> +#include <linux/seq_file.h> +#include <linux/statfs.h> +#include "aufs.h" + +static aufs_bindex_t au_fhsm_bottom(struct super_block *sb) +{ + struct au_sbinfo *sbinfo; + struct au_fhsm *fhsm; + + SiMustAnyLock(sb); + + sbinfo = au_sbi(sb); + fhsm = &sbinfo->si_fhsm; + AuDebugOn(!fhsm); + return fhsm->fhsm_bottom; +} + +void au_fhsm_set_bottom(struct super_block *sb, aufs_bindex_t bindex) +{ + struct au_sbinfo *sbinfo; + struct au_fhsm *fhsm; + + SiMustWriteLock(sb); + + sbinfo = au_sbi(sb); + fhsm = &sbinfo->si_fhsm; + AuDebugOn(!fhsm); + fhsm->fhsm_bottom = bindex; +} + +/* ---------------------------------------------------------------------- */ + +static int au_fhsm_test_jiffy(struct au_sbinfo *sbinfo, struct au_branch *br) +{ + struct au_br_fhsm *bf; + + bf = br->br_fhsm; + MtxMustLock(&bf->bf_lock); + + return !bf->bf_readable + || time_after(jiffies, + bf->bf_jiffy + sbinfo->si_fhsm.fhsm_expire); +} + +/* ---------------------------------------------------------------------- */ + +static void au_fhsm_notify(struct super_block *sb, int val) +{ + struct au_sbinfo *sbinfo; + struct au_fhsm *fhsm; + + SiMustAnyLock(sb); + + sbinfo = au_sbi(sb); + fhsm = &sbinfo->si_fhsm; + if (au_fhsm_pid(fhsm) + && atomic_read(&fhsm->fhsm_readable) != -1) { + atomic_set(&fhsm->fhsm_readable, val); + if (val) + wake_up(&fhsm->fhsm_wqh); + } +} + +static int au_fhsm_stfs(struct super_block *sb, aufs_bindex_t bindex, + struct aufs_stfs *rstfs, int do_lock, int do_notify) +{ + int err; + struct au_branch *br; + struct au_br_fhsm *bf; + + br = au_sbr(sb, bindex); + AuDebugOn(au_br_rdonly(br)); + bf = br->br_fhsm; + AuDebugOn(!bf); + + if (do_lock) + mutex_lock(&bf->bf_lock); + else + MtxMustLock(&bf->bf_lock); + + /* sb->s_root for NFS is unreliable */ + err = au_br_stfs(br, &bf->bf_stfs); + if (unlikely(err)) { + AuErr1("FHSM failed (%d), b%d, ignored.\n", bindex, err); + goto out; + } + + bf->bf_jiffy = jiffies; + bf->bf_readable = 1; + if (do_notify) + au_fhsm_notify(sb, /*val*/1); + if (rstfs) + *rstfs = bf->bf_stfs; + +out: + if (do_lock) + mutex_unlock(&bf->bf_lock); + au_fhsm_notify(sb, /*val*/1); + + return err; +} + +void au_fhsm_wrote(struct super_block *sb, aufs_bindex_t bindex, int force) +{ + int err; + struct au_sbinfo *sbinfo; + struct au_fhsm *fhsm; + struct au_branch *br; + struct au_br_fhsm *bf; + + AuDbg("b%d, force %d\n", bindex, force); + SiMustAnyLock(sb); + + sbinfo = au_sbi(sb); + fhsm = &sbinfo->si_fhsm; + if (!au_ftest_si(sbinfo, FHSM) + || fhsm->fhsm_bottom == bindex) + return; + + br = au_sbr(sb, bindex); + bf = br->br_fhsm; + AuDebugOn(!bf); + mutex_lock(&bf->bf_lock); + if (force + || au_fhsm_pid(fhsm) + || au_fhsm_test_jiffy(sbinfo, br)) + err = au_fhsm_stfs(sb, bindex, /*rstfs*/NULL, /*do_lock*/0, + /*do_notify*/1); + mutex_unlock(&bf->bf_lock); +} + +void au_fhsm_wrote_all(struct super_block *sb, int force) +{ + aufs_bindex_t bindex, bend; + struct au_branch *br; + + /* exclude the bottom */ + bend = au_fhsm_bottom(sb); + for (bindex = 0; bindex < bend; bindex++) { + br = au_sbr(sb, bindex); + if (au_br_fhsm(br->br_perm)) + au_fhsm_wrote(sb, bindex, force); + } +} + +/* ---------------------------------------------------------------------- */ + +static unsigned int au_fhsm_poll(struct file *file, + struct poll_table_struct *wait) +{ + unsigned int mask; + struct au_sbinfo *sbinfo; + struct au_fhsm *fhsm; + + mask = 0; + sbinfo = file->private_data; + fhsm = &sbinfo->si_fhsm; + poll_wait(file, &fhsm->fhsm_wqh, wait); + if (atomic_read(&fhsm->fhsm_readable)) + mask = POLLIN /* | POLLRDNORM */; + + AuTraceErr((int)mask); + return mask; +} + +static int au_fhsm_do_read_one(struct aufs_stbr __user *stbr, + struct aufs_stfs *stfs, __s16 brid) +{ + int err; + + err = copy_to_user(&stbr->stfs, stfs, sizeof(*stfs)); + if (!err) + err = __put_user(brid, &stbr->brid); + if (unlikely(err)) + err = -EFAULT; + + return err; +} + +static ssize_t au_fhsm_do_read(struct super_block *sb, + struct aufs_stbr __user *stbr, size_t count) +{ + ssize_t err; + int nstbr; + aufs_bindex_t bindex, bend; + struct au_branch *br; + struct au_br_fhsm *bf; + + /* except the bottom branch */ + err = 0; + nstbr = 0; + bend = au_fhsm_bottom(sb); + for (bindex = 0; !err && bindex < bend; bindex++) { + br = au_sbr(sb, bindex); + if (!au_br_fhsm(br->br_perm)) + continue; + + bf = br->br_fhsm; + mutex_lock(&bf->bf_lock); + if (bf->bf_readable) { + err = -EFAULT; + if (count >= sizeof(*stbr)) + err = au_fhsm_do_read_one(stbr++, &bf->bf_stfs, + br->br_id); + if (!err) { + bf->bf_readable = 0; + count -= sizeof(*stbr); + nstbr++; + } + } + mutex_unlock(&bf->bf_lock); + } + if (!err) + err = sizeof(*stbr) * nstbr; + + return err; +} + +static ssize_t au_fhsm_read(struct file *file, char __user *buf, size_t count, + loff_t *pos) +{ + ssize_t err; + int readable; + aufs_bindex_t nfhsm, bindex, bend; + struct au_sbinfo *sbinfo; + struct au_fhsm *fhsm; + struct au_branch *br; + struct super_block *sb; + + err = 0; + sbinfo = file->private_data; + fhsm = &sbinfo->si_fhsm; +need_data: + spin_lock_irq(&fhsm->fhsm_wqh.lock); + if (!atomic_read(&fhsm->fhsm_readable)) { + if (vfsub_file_flags(file) & O_NONBLOCK) + err = -EAGAIN; + else + err = wait_event_interruptible_locked_irq + (fhsm->fhsm_wqh, + atomic_read(&fhsm->fhsm_readable)); + } + spin_unlock_irq(&fhsm->fhsm_wqh.lock); + if (unlikely(err)) + goto out; + + /* sb may already be dead */ + au_rw_read_lock(&sbinfo->si_rwsem); + readable = atomic_read(&fhsm->fhsm_readable); + if (readable > 0) { + sb = sbinfo->si_sb; + AuDebugOn(!sb); + /* exclude the bottom branch */ + nfhsm = 0; + bend = au_fhsm_bottom(sb); + for (bindex = 0; bindex < bend; bindex++) { + br = au_sbr(sb, bindex); + if (au_br_fhsm(br->br_perm)) + nfhsm++; + } + err = -EMSGSIZE; + if (nfhsm * sizeof(struct aufs_stbr) <= count) { + atomic_set(&fhsm->fhsm_readable, 0); + err = au_fhsm_do_read(sbinfo->si_sb, (void __user *)buf, + count); + } + } + au_rw_read_unlock(&sbinfo->si_rwsem); + if (!readable) + goto need_data; + +out: + return err; +} + +static int au_fhsm_release(struct inode *inode, struct file *file) +{ + struct au_sbinfo *sbinfo; + struct au_fhsm *fhsm; + + /* sb may already be dead */ + sbinfo = file->private_data; + fhsm = &sbinfo->si_fhsm; + spin_lock(&fhsm->fhsm_spin); + fhsm->fhsm_pid = 0; + spin_unlock(&fhsm->fhsm_spin); + kobject_put(&sbinfo->si_kobj); + + return 0; +} + +static const struct file_operations au_fhsm_fops = { + .owner = THIS_MODULE, + .llseek = noop_llseek, + .read = au_fhsm_read, + .poll = au_fhsm_poll, + .release = au_fhsm_release +}; + +int au_fhsm_fd(struct super_block *sb, int oflags) +{ + int err, fd; + struct au_sbinfo *sbinfo; + struct au_fhsm *fhsm; + + err = -EPERM; + if (unlikely(!capable(CAP_SYS_ADMIN))) + goto out; + + err = -EINVAL; + if (unlikely(oflags & ~(O_CLOEXEC | O_NONBLOCK))) + goto out; + + err = 0; + sbinfo = au_sbi(sb); + fhsm = &sbinfo->si_fhsm; + spin_lock(&fhsm->fhsm_spin); + if (!fhsm->fhsm_pid) + fhsm->fhsm_pid = current->pid; + else + err = -EBUSY; + spin_unlock(&fhsm->fhsm_spin); + if (unlikely(err)) + goto out; + + oflags |= O_RDONLY; + /* oflags |= FMODE_NONOTIFY; */ + fd = anon_inode_getfd("[aufs_fhsm]", &au_fhsm_fops, sbinfo, oflags); + err = fd; + if (unlikely(fd < 0)) + goto out_pid; + + /* succeed reglardless 'fhsm' status */ + kobject_get(&sbinfo->si_kobj); + si_noflush_read_lock(sb); + if (au_ftest_si(sbinfo, FHSM)) + au_fhsm_wrote_all(sb, /*force*/0); + si_read_unlock(sb); + goto out; /* success */ + +out_pid: + spin_lock(&fhsm->fhsm_spin); + fhsm->fhsm_pid = 0; + spin_unlock(&fhsm->fhsm_spin); +out: + AuTraceErr(err); + return err; +} + +/* ---------------------------------------------------------------------- */ + +int au_fhsm_br_alloc(struct au_branch *br) +{ + int err; + + err = 0; + br->br_fhsm = kmalloc(sizeof(*br->br_fhsm), GFP_NOFS); + if (br->br_fhsm) + au_br_fhsm_init(br->br_fhsm); + else + err = -ENOMEM; + + return err; +} + +/* ---------------------------------------------------------------------- */ + +void au_fhsm_fin(struct super_block *sb) +{ + au_fhsm_notify(sb, /*val*/-1); +} + +void au_fhsm_init(struct au_sbinfo *sbinfo) +{ + struct au_fhsm *fhsm; + + fhsm = &sbinfo->si_fhsm; + spin_lock_init(&fhsm->fhsm_spin); + init_waitqueue_head(&fhsm->fhsm_wqh); + atomic_set(&fhsm->fhsm_readable, 0); + fhsm->fhsm_expire + = msecs_to_jiffies(AUFS_FHSM_CACHE_DEF_SEC * MSEC_PER_SEC); + fhsm->fhsm_bottom = -1; +} + +void au_fhsm_set(struct au_sbinfo *sbinfo, unsigned int sec) +{ + sbinfo->si_fhsm.fhsm_expire + = msecs_to_jiffies(sec * MSEC_PER_SEC); +} + +void au_fhsm_show(struct seq_file *seq, struct au_sbinfo *sbinfo) +{ + unsigned int u; + + if (!au_ftest_si(sbinfo, FHSM)) + return; + + u = jiffies_to_msecs(sbinfo->si_fhsm.fhsm_expire) / MSEC_PER_SEC; + if (u != AUFS_FHSM_CACHE_DEF_SEC) + seq_printf(seq, ",fhsm_sec=%u", u); +} diff --git a/fs/aufs/file.c b/fs/aufs/file.c new file mode 100644 index 000000000..6b8a66b4a --- /dev/null +++ b/fs/aufs/file.c @@ -0,0 +1,831 @@ +/* + * Copyright (C) 2005-2016 Junjiro R. Okajima + */ + +/* + * handling file/dir, and address_space operation + */ + +#ifdef CONFIG_AUFS_DEBUG +#include <linux/migrate.h> +#endif +#include <linux/pagemap.h> +#include "aufs.h" + +/* drop flags for writing */ +unsigned int au_file_roflags(unsigned int flags) +{ + flags &= ~(O_WRONLY | O_RDWR | O_APPEND | O_CREAT | O_TRUNC); + flags |= O_RDONLY | O_NOATIME; + return flags; +} + +/* common functions to regular file and dir */ +struct file *au_h_open(struct dentry *dentry, aufs_bindex_t bindex, int flags, + struct file *file, int force_wr) +{ + struct file *h_file; + struct dentry *h_dentry; + struct inode *h_inode; + struct super_block *sb; + struct au_branch *br; + struct path h_path; + int err; + + /* a race condition can happen between open and unlink/rmdir */ + h_file = ERR_PTR(-ENOENT); + h_dentry = au_h_dptr(dentry, bindex); + if (au_test_nfsd() && (!h_dentry || d_is_negative(h_dentry))) + goto out; + h_inode = d_inode(h_dentry); + spin_lock(&h_dentry->d_lock); + err = (!d_unhashed(dentry) && d_unlinked(h_dentry)) + /* || !d_inode(dentry)->i_nlink */ + ; + spin_unlock(&h_dentry->d_lock); + if (unlikely(err)) + goto out; + + sb = dentry->d_sb; + br = au_sbr(sb, bindex); + err = au_br_test_oflag(flags, br); + h_file = ERR_PTR(err); + if (unlikely(err)) + goto out; + + /* drop flags for writing */ + if (au_test_ro(sb, bindex, d_inode(dentry))) { + if (force_wr && !(flags & O_WRONLY)) + force_wr = 0; + flags = au_file_roflags(flags); + if (force_wr) { + h_file = ERR_PTR(-EROFS); + flags = au_file_roflags(flags); + if (unlikely(vfsub_native_ro(h_inode) + || IS_APPEND(h_inode))) + goto out; + flags &= ~O_ACCMODE; + flags |= O_WRONLY; + } + } + flags &= ~O_CREAT; + atomic_inc(&br->br_count); + h_path.dentry = h_dentry; + h_path.mnt = au_br_mnt(br); + h_file = vfsub_dentry_open(&h_path, flags); + if (IS_ERR(h_file)) + goto out_br; + + if (flags & __FMODE_EXEC) { + err = deny_write_access(h_file); + if (unlikely(err)) { + fput(h_file); + h_file = ERR_PTR(err); + goto out_br; + } + } + fsnotify_open(h_file); + goto out; /* success */ + +out_br: + atomic_dec(&br->br_count); +out: + return h_file; +} + +static int au_cmoo(struct dentry *dentry) +{ + int err, cmoo; + unsigned int udba; + struct path h_path; + struct au_pin pin; + struct au_cp_generic cpg = { + .dentry = dentry, + .bdst = -1, + .bsrc = -1, + .len = -1, + .pin = &pin, + .flags = AuCpup_DTIME | AuCpup_HOPEN + }; + struct inode *delegated; + struct super_block *sb; + struct au_sbinfo *sbinfo; + struct au_fhsm *fhsm; + pid_t pid; + struct au_branch *br; + struct dentry *parent; + struct au_hinode *hdir; + + DiMustWriteLock(dentry); + IiMustWriteLock(d_inode(dentry)); + + err = 0; + if (IS_ROOT(dentry)) + goto out; + cpg.bsrc = au_dbstart(dentry); + if (!cpg.bsrc) + goto out; + + sb = dentry->d_sb; + sbinfo = au_sbi(sb); + fhsm = &sbinfo->si_fhsm; + pid = au_fhsm_pid(fhsm); + if (pid + && (current->pid == pid + || current->real_parent->pid == pid)) + goto out; + + br = au_sbr(sb, cpg.bsrc); + cmoo = au_br_cmoo(br->br_perm); + if (!cmoo) + goto out; + if (!d_is_reg(dentry)) + cmoo &= AuBrAttr_COO_ALL; + if (!cmoo) + goto out; + + parent = dget_parent(dentry); + di_write_lock_parent(parent); + err = au_wbr_do_copyup_bu(dentry, cpg.bsrc - 1); + cpg.bdst = err; + if (unlikely(err < 0)) { + err = 0; /* there is no upper writable branch */ + goto out_dgrade; + } + AuDbg("bsrc %d, bdst %d\n", cpg.bsrc, cpg.bdst); + + /* do not respect the coo attrib for the target branch */ + err = au_cpup_dirs(dentry, cpg.bdst); + if (unlikely(err)) + goto out_dgrade; + + di_downgrade_lock(parent, AuLock_IR); + udba = au_opt_udba(sb); + err = au_pin(&pin, dentry, cpg.bdst, udba, + AuPin_DI_LOCKED | AuPin_MNT_WRITE); + if (unlikely(err)) + goto out_parent; + + err = au_sio_cpup_simple(&cpg); + au_unpin(&pin); + if (unlikely(err)) + goto out_parent; + if (!(cmoo & AuBrWAttr_MOO)) + goto out_parent; /* success */ + + err = au_pin(&pin, dentry, cpg.bsrc, udba, + AuPin_DI_LOCKED | AuPin_MNT_WRITE); + if (unlikely(err)) + goto out_parent; + + h_path.mnt = au_br_mnt(br); + h_path.dentry = au_h_dptr(dentry, cpg.bsrc); + hdir = au_hi(d_inode(parent), cpg.bsrc); + delegated = NULL; + err = vfsub_unlink(hdir->hi_inode, &h_path, &delegated, /*force*/1); + au_unpin(&pin); + /* todo: keep h_dentry or not? */ + if (unlikely(err == -EWOULDBLOCK)) { + pr_warn("cannot retry for NFSv4 delegation" + " for an internal unlink\n"); + iput(delegated); + } + if (unlikely(err)) { + pr_err("unlink %pd after coo failed (%d), ignored\n", + dentry, err); + err = 0; + } + goto out_parent; /* success */ + +out_dgrade: + di_downgrade_lock(parent, AuLock_IR); +out_parent: + di_read_unlock(parent, AuLock_IR); + dput(parent); +out: + AuTraceErr(err); + return err; +} + +int au_do_open(struct file *file, struct au_do_open_args *args) +{ + int err, no_lock = args->no_lock; + struct dentry *dentry; + struct au_finfo *finfo; + + if (!no_lock) + err = au_finfo_init(file, args->fidir); + else { + lockdep_off(); + err = au_finfo_init(file, args->fidir); + lockdep_on(); + } + if (unlikely(err)) + goto out; + + dentry = file->f_path.dentry; + AuDebugOn(IS_ERR_OR_NULL(dentry)); + if (!no_lock) { + di_write_lock_child(dentry); + err = au_cmoo(dentry); + di_downgrade_lock(dentry, AuLock_IR); + if (!err) + err = args->open(file, vfsub_file_flags(file), NULL); + di_read_unlock(dentry, AuLock_IR); + } else { + err = au_cmoo(dentry); + if (!err) + err = args->open(file, vfsub_file_flags(file), + args->h_file); + if (!err && au_fbstart(file) != au_dbstart(dentry)) + /* + * cmoo happens after h_file was opened. + * need to refresh file later. + */ + atomic_dec(&au_fi(file)->fi_generation); + } + + finfo = au_fi(file); + if (!err) { + finfo->fi_file = file; + au_sphl_add(&finfo->fi_hlist, + &au_sbi(file->f_path.dentry->d_sb)->si_files); + } + if (!no_lock) + fi_write_unlock(file); + else { + lockdep_off(); + fi_write_unlock(file); + lockdep_on(); + } + if (unlikely(err)) { + finfo->fi_hdir = NULL; + au_finfo_fin(file); + } + +out: + return err; +} + +int au_reopen_nondir(struct file *file) +{ + int err; + aufs_bindex_t bstart; + struct dentry *dentry; + struct file *h_file, *h_file_tmp; + + dentry = file->f_path.dentry; + bstart = au_dbstart(dentry); + h_file_tmp = NULL; + if (au_fbstart(file) == bstart) { + h_file = au_hf_top(file); + if (file->f_mode == h_file->f_mode) + return 0; /* success */ + h_file_tmp = h_file; + get_file(h_file_tmp); + au_set_h_fptr(file, bstart, NULL); + } + AuDebugOn(au_fi(file)->fi_hdir); + /* + * it can happen + * file exists on both of rw and ro + * open --> dbstart and fbstart are both 0 + * prepend a branch as rw, "rw" become ro + * remove rw/file + * delete the top branch, "rw" becomes rw again + * --> dbstart is 1, fbstart is still 0 + * write --> fbstart is 0 but dbstart is 1 + */ + /* AuDebugOn(au_fbstart(file) < bstart); */ + + h_file = au_h_open(dentry, bstart, vfsub_file_flags(file) & ~O_TRUNC, + file, /*force_wr*/0); + err = PTR_ERR(h_file); + if (IS_ERR(h_file)) { + if (h_file_tmp) { + atomic_inc(&au_sbr(dentry->d_sb, bstart)->br_count); + au_set_h_fptr(file, bstart, h_file_tmp); + h_file_tmp = NULL; + } + goto out; /* todo: close all? */ + } + + err = 0; + au_set_fbstart(file, bstart); + au_set_h_fptr(file, bstart, h_file); + au_update_figen(file); + /* todo: necessary? */ + /* file->f_ra = h_file->f_ra; */ + +out: + if (h_file_tmp) + fput(h_file_tmp); + return err; +} + +/* ---------------------------------------------------------------------- */ + +static int au_reopen_wh(struct file *file, aufs_bindex_t btgt, + struct dentry *hi_wh) +{ + int err; + aufs_bindex_t bstart; + struct au_dinfo *dinfo; + struct dentry *h_dentry; + struct au_hdentry *hdp; + + dinfo = au_di(file->f_path.dentry); + AuRwMustWriteLock(&dinfo->di_rwsem); + + bstart = dinfo->di_bstart; + dinfo->di_bstart = btgt; + hdp = dinfo->di_hdentry; + h_dentry = hdp[0 + btgt].hd_dentry; + hdp[0 + btgt].hd_dentry = hi_wh; + err = au_reopen_nondir(file); + hdp[0 + btgt].hd_dentry = h_dentry; + dinfo->di_bstart = bstart; + + return err; +} + +static int au_ready_to_write_wh(struct file *file, loff_t len, + aufs_bindex_t bcpup, struct au_pin *pin) +{ + int err; + struct inode *inode, *h_inode; + struct dentry *h_dentry, *hi_wh; + struct au_cp_generic cpg = { + .dentry = file->f_path.dentry, + .bdst = bcpup, + .bsrc = -1, + .len = len, + .pin = pin + }; + + au_update_dbstart(cpg.dentry); + inode = d_inode(cpg.dentry); + h_inode = NULL; + if (au_dbstart(cpg.dentry) <= bcpup + && au_dbend(cpg.dentry) >= bcpup) { + h_dentry = au_h_dptr(cpg.dentry, bcpup); + if (h_dentry && d_is_positive(h_dentry)) + h_inode = d_inode(h_dentry); + } + hi_wh = au_hi_wh(inode, bcpup); + if (!hi_wh && !h_inode) + err = au_sio_cpup_wh(&cpg, file); + else + /* already copied-up after unlink */ + err = au_reopen_wh(file, bcpup, hi_wh); + + if (!err + && (inode->i_nlink > 1 + || (inode->i_state & I_LINKABLE)) + && au_opt_test(au_mntflags(cpg.dentry->d_sb), PLINK)) + au_plink_append(inode, bcpup, au_h_dptr(cpg.dentry, bcpup)); + + return err; +} + +/* + * prepare the @file for writing. + */ +int au_ready_to_write(struct file *file, loff_t len, struct au_pin *pin) +{ + int err; + aufs_bindex_t dbstart; + struct dentry *parent; + struct inode *inode; + struct super_block *sb; + struct file *h_file; + struct au_cp_generic cpg = { + .dentry = file->f_path.dentry, + .bdst = -1, + .bsrc = -1, + .len = len, + .pin = pin, + .flags = AuCpup_DTIME + }; + + sb = cpg.dentry->d_sb; + inode = d_inode(cpg.dentry); + cpg.bsrc = au_fbstart(file); + err = au_test_ro(sb, cpg.bsrc, inode); + if (!err && (au_hf_top(file)->f_mode & FMODE_WRITE)) { + err = au_pin(pin, cpg.dentry, cpg.bsrc, AuOpt_UDBA_NONE, + /*flags*/0); + goto out; + } + + /* need to cpup or reopen */ + parent = dget_parent(cpg.dentry); + di_write_lock_parent(parent); + err = AuWbrCopyup(au_sbi(sb), cpg.dentry); + cpg.bdst = err; + if (unlikely(err < 0)) + goto out_dgrade; + err = 0; + + if (!d_unhashed(cpg.dentry) && !au_h_dptr(parent, cpg.bdst)) { + err = au_cpup_dirs(cpg.dentry, cpg.bdst); + if (unlikely(err)) + goto out_dgrade; + } + + err = au_pin(pin, cpg.dentry, cpg.bdst, AuOpt_UDBA_NONE, + AuPin_DI_LOCKED | AuPin_MNT_WRITE); + if (unlikely(err)) + goto out_dgrade; + + dbstart = au_dbstart(cpg.dentry); + if (dbstart <= cpg.bdst) + cpg.bsrc = cpg.bdst; + + if (dbstart <= cpg.bdst /* just reopen */ + || !d_unhashed(cpg.dentry) /* copyup and reopen */ + ) { + h_file = au_h_open_pre(cpg.dentry, cpg.bsrc, /*force_wr*/0); + if (IS_ERR(h_file)) + err = PTR_ERR(h_file); + else { + di_downgrade_lock(parent, AuLock_IR); + if (dbstart > cpg.bdst) + err = au_sio_cpup_simple(&cpg); + if (!err) + err = au_reopen_nondir(file); + au_h_open_post(cpg.dentry, cpg.bsrc, h_file); + } + } else { /* copyup as wh and reopen */ + /* + * since writable hfsplus branch is not supported, + * h_open_pre/post() are unnecessary. + */ + err = au_ready_to_write_wh(file, len, cpg.bdst, pin); + di_downgrade_lock(parent, AuLock_IR); + } + + if (!err) { + au_pin_set_parent_lflag(pin, /*lflag*/0); + goto out_dput; /* success */ + } + au_unpin(pin); + goto out_unlock; + +out_dgrade: + di_downgrade_lock(parent, AuLock_IR); +out_unlock: + di_read_unlock(parent, AuLock_IR); +out_dput: + dput(parent); +out: + return err; +} + +/* ---------------------------------------------------------------------- */ + +int au_do_flush(struct file *file, fl_owner_t id, + int (*flush)(struct file *file, fl_owner_t id)) +{ + int err; + struct super_block *sb; + struct inode *inode; + + inode = file_inode(file); + sb = inode->i_sb; + si_noflush_read_lock(sb); + fi_read_lock(file); + ii_read_lock_child(inode); + + err = flush(file, id); + au_cpup_attr_timesizes(inode); + + ii_read_unlock(inode); + fi_read_unlock(file); + si_read_unlock(sb); + return err; +} + +/* ---------------------------------------------------------------------- */ + +static int au_file_refresh_by_inode(struct file *file, int *need_reopen) +{ + int err; + struct au_pin pin; + struct au_finfo *finfo; + struct dentry *parent, *hi_wh; + struct inode *inode; + struct super_block *sb; + struct au_cp_generic cpg = { + .dentry = file->f_path.dentry, + .bdst = -1, + .bsrc = -1, + .len = -1, + .pin = &pin, + .flags = AuCpup_DTIME + }; + + FiMustWriteLock(file); + + err = 0; + finfo = au_fi(file); + sb = cpg.dentry->d_sb; + inode = d_inode(cpg.dentry); + cpg.bdst = au_ibstart(inode); + if (cpg.bdst == finfo->fi_btop || IS_ROOT(cpg.dentry)) + goto out; + + parent = dget_parent(cpg.dentry); + if (au_test_ro(sb, cpg.bdst, inode)) { + di_read_lock_parent(parent, !AuLock_IR); + err = AuWbrCopyup(au_sbi(sb), cpg.dentry); + cpg.bdst = err; + di_read_unlock(parent, !AuLock_IR); + if (unlikely(err < 0)) + goto out_parent; + err = 0; + } + + di_read_lock_parent(parent, AuLock_IR); + hi_wh = au_hi_wh(inode, cpg.bdst); + if (!S_ISDIR(inode->i_mode) + && au_opt_test(au_mntflags(sb), PLINK) + && au_plink_test(inode) + && !d_unhashed(cpg.dentry) + && cpg.bdst < au_dbstart(cpg.dentry)) { + err = au_test_and_cpup_dirs(cpg.dentry, cpg.bdst); + if (unlikely(err)) + goto out_unlock; + + /* always superio. */ + err = au_pin(&pin, cpg.dentry, cpg.bdst, AuOpt_UDBA_NONE, + AuPin_DI_LOCKED | AuPin_MNT_WRITE); + if (!err) { + err = au_sio_cpup_simple(&cpg); + au_unpin(&pin); + } + } else if (hi_wh) { + /* already copied-up after unlink */ + err = au_reopen_wh(file, cpg.bdst, hi_wh); + *need_reopen = 0; + } + +out_unlock: + di_read_unlock(parent, AuLock_IR); +out_parent: + dput(parent); +out: + return err; +} + +static void au_do_refresh_dir(struct file *file) +{ + aufs_bindex_t bindex, bend, new_bindex, brid; + struct au_hfile *p, tmp, *q; + struct au_finfo *finfo; + struct super_block *sb; + struct au_fidir *fidir; + + FiMustWriteLock(file); + + sb = file->f_path.dentry->d_sb; + finfo = au_fi(file); + fidir = finfo->fi_hdir; + AuDebugOn(!fidir); + p = fidir->fd_hfile + finfo->fi_btop; + brid = p->hf_br->br_id; + bend = fidir->fd_bbot; + for (bindex = finfo->fi_btop; bindex <= bend; bindex++, p++) { + if (!p->hf_file) + continue; + + new_bindex = au_br_index(sb, p->hf_br->br_id); + if (new_bindex == bindex) + continue; + if (new_bindex < 0) { + au_set_h_fptr(file, bindex, NULL); + continue; + } + + /* swap two lower inode, and loop again */ + q = fidir->fd_hfile + new_bindex; + tmp = *q; + *q = *p; + *p = tmp; + if (tmp.hf_file) { + bindex--; + p--; + } + } + + p = fidir->fd_hfile; + if (!au_test_mmapped(file) && !d_unlinked(file->f_path.dentry)) { + bend = au_sbend(sb); + for (finfo->fi_btop = 0; finfo->fi_btop <= bend; + finfo->fi_btop++, p++) + if (p->hf_file) { + if (file_inode(p->hf_file)) + break; + au_hfput(p, file); + } + } else { + bend = au_br_index(sb, brid); + for (finfo->fi_btop = 0; finfo->fi_btop < bend; + finfo->fi_btop++, p++) + if (p->hf_file) + au_hfput(p, file); + bend = au_sbend(sb); + } + + p = fidir->fd_hfile + bend; + for (fidir->fd_bbot = bend; fidir->fd_bbot >= finfo->fi_btop; + fidir->fd_bbot--, p--) + if (p->hf_file) { + if (file_inode(p->hf_file)) + break; + au_hfput(p, file); + } + AuDebugOn(fidir->fd_bbot < finfo->fi_btop); +} + +/* + * after branch manipulating, refresh the file. + */ +static int refresh_file(struct file *file, int (*reopen)(struct file *file)) +{ + int err, need_reopen; + aufs_bindex_t bend, bindex; + struct dentry *dentry; + struct au_finfo *finfo; + struct au_hfile *hfile; + + dentry = file->f_path.dentry; + finfo = au_fi(file); + if (!finfo->fi_hdir) { + hfile = &finfo->fi_htop; + AuDebugOn(!hfile->hf_file); + bindex = au_br_index(dentry->d_sb, hfile->hf_br->br_id); + AuDebugOn(bindex < 0); + if (bindex != finfo->fi_btop) + au_set_fbstart(file, bindex); + } else { + err = au_fidir_realloc(finfo, au_sbend(dentry->d_sb) + 1); + if (unlikely(err)) + goto out; + au_do_refresh_dir(file); + } + + err = 0; + need_reopen = 1; + if (!au_test_mmapped(file)) + err = au_file_refresh_by_inode(file, &need_reopen); + if (!err && need_reopen && !d_unlinked(dentry)) + err = reopen(file); + if (!err) { + au_update_figen(file); + goto out; /* success */ + } + + /* error, close all lower files */ + if (finfo->fi_hdir) { + bend = au_fbend_dir(file); + for (bindex = au_fbstart(file); bindex <= bend; bindex++) + au_set_h_fptr(file, bindex, NULL); + } + +out: + return err; +} + +/* common function to regular file and dir */ +int au_reval_and_lock_fdi(struct file *file, int (*reopen)(struct file *file), + int wlock) +{ + int err; + unsigned int sigen, figen; + aufs_bindex_t bstart; + unsigned char pseudo_link; + struct dentry *dentry; + struct inode *inode; + + err = 0; + dentry = file->f_path.dentry; + inode = d_inode(dentry); + sigen = au_sigen(dentry->d_sb); + fi_write_lock(file); + figen = au_figen(file); + di_write_lock_child(dentry); + bstart = au_dbstart(dentry); + pseudo_link = (bstart != au_ibstart(inode)); + if (sigen == figen && !pseudo_link && au_fbstart(file) == bstart) { + if (!wlock) { + di_downgrade_lock(dentry, AuLock_IR); + fi_downgrade_lock(file); + } + goto out; /* success */ + } + + AuDbg("sigen %d, figen %d\n", sigen, figen); + if (au_digen_test(dentry, sigen)) { + err = au_reval_dpath(dentry, sigen); + AuDebugOn(!err && au_digen_test(dentry, sigen)); + } + + if (!err) + err = refresh_file(file, reopen); + if (!err) { + if (!wlock) { + di_downgrade_lock(dentry, AuLock_IR); + fi_downgrade_lock(file); + } + } else { + di_write_unlock(dentry); + fi_write_unlock(file); + } + +out: + return err; +} + +/* ---------------------------------------------------------------------- */ + +/* cf. aufs_nopage() */ +/* for madvise(2) */ +static int aufs_readpage(struct file *file __maybe_unused, struct page *page) +{ + unlock_page(page); + return 0; +} + +/* it will never be called, but necessary to support O_DIRECT */ +static ssize_t aufs_direct_IO(struct kiocb *iocb, struct iov_iter *iter, + loff_t offset) +{ BUG(); return 0; } + +/* they will never be called. */ +#ifdef CONFIG_AUFS_DEBUG +static int aufs_write_begin(struct file *file, struct address_space *mapping, + loff_t pos, unsigned len, unsigned flags, + struct page **pagep, void **fsdata) +{ AuUnsupport(); return 0; } +static int aufs_write_end(struct file *file, struct address_space *mapping, + loff_t pos, unsigned len, unsigned copied, + struct page *page, void *fsdata) +{ AuUnsupport(); return 0; } +static int aufs_writepage(struct page *page, struct writeback_control *wbc) +{ AuUnsupport(); return 0; } + +static int aufs_set_page_dirty(struct page *page) +{ AuUnsupport(); return 0; } +static void aufs_invalidatepage(struct page *page, unsigned int offset, + unsigned int length) +{ AuUnsupport(); } +static int aufs_releasepage(struct page *page, gfp_t gfp) +{ AuUnsupport(); return 0; } +#if 0 /* called by memory compaction regardless file */ +static int aufs_migratepage(struct address_space *mapping, struct page *newpage, + struct page *page, enum migrate_mode mode) +{ AuUnsupport(); return 0; } +#endif +static int aufs_launder_page(struct page *page) +{ AuUnsupport(); return 0; } +static int aufs_is_partially_uptodate(struct page *page, + unsigned long from, + unsigned long count) +{ AuUnsupport(); return 0; } +static void aufs_is_dirty_writeback(struct page *page, bool *dirty, + bool *writeback) +{ AuUnsupport(); } +static int aufs_error_remove_page(struct address_space *mapping, + struct page *page) +{ AuUnsupport(); return 0; } +static int aufs_swap_activate(struct swap_info_struct *sis, struct file *file, + sector_t *span) +{ AuUnsupport(); return 0; } +static void aufs_swap_deactivate(struct file *file) +{ AuUnsupport(); } +#endif /* CONFIG_AUFS_DEBUG */ + +const struct address_space_operations aufs_aop = { + .readpage = aufs_readpage, + .direct_IO = aufs_direct_IO, +#ifdef CONFIG_AUFS_DEBUG + .writepage = aufs_writepage, + /* no writepages, because of writepage */ + .set_page_dirty = aufs_set_page_dirty, + /* no readpages, because of readpage */ + .write_begin = aufs_write_begin, + .write_end = aufs_write_end, + /* no bmap, no block device */ + .invalidatepage = aufs_invalidatepage, + .releasepage = aufs_releasepage, + /* is fallback_migrate_page ok? */ + /* .migratepage = aufs_migratepage, */ + .launder_page = aufs_launder_page, + .is_partially_uptodate = aufs_is_partially_uptodate, + .is_dirty_writeback = aufs_is_dirty_writeback, + .error_remove_page = aufs_error_remove_page, + .swap_activate = aufs_swap_activate, + .swap_deactivate = aufs_swap_deactivate +#endif /* CONFIG_AUFS_DEBUG */ +}; diff --git a/fs/aufs/file.h b/fs/aufs/file.h new file mode 100644 index 000000000..27d802487 --- /dev/null +++ b/fs/aufs/file.h @@ -0,0 +1,278 @@ +/* + * Copyright (C) 2005-2016 Junjiro R. Okajima + */ + +/* + * file operations + */ + +#ifndef __AUFS_FILE_H__ +#define __AUFS_FILE_H__ + +#ifdef __KERNEL__ + +#include <linux/file.h> +#include <linux/fs.h> +#include <linux/poll.h> +#include "rwsem.h" + +struct au_branch; +struct au_hfile { + struct file *hf_file; + struct au_branch *hf_br; +}; + +struct au_vdir; +struct au_fidir { + aufs_bindex_t fd_bbot; + aufs_bindex_t fd_nent; + struct au_vdir *fd_vdir_cache; + struct au_hfile fd_hfile[]; +}; + +static inline int au_fidir_sz(int nent) +{ + AuDebugOn(nent < 0); + return sizeof(struct au_fidir) + sizeof(struct au_hfile) * nent; +} + +struct au_finfo { + atomic_t fi_generation; + + struct au_rwsem fi_rwsem; + aufs_bindex_t fi_btop; + + /* do not union them */ + struct { /* for non-dir */ + struct au_hfile fi_htop; + atomic_t fi_mmapped; + }; + struct au_fidir *fi_hdir; /* for dir only */ + + struct hlist_node fi_hlist; + struct file *fi_file; /* very ugly */ +} ____cacheline_aligned_in_smp; + +/* ---------------------------------------------------------------------- */ + +/* file.c */ +extern const struct address_space_operations aufs_aop; +unsigned int au_file_roflags(unsigned int flags); +struct file *au_h_open(struct dentry *dentry, aufs_bindex_t bindex, int flags, + struct file *file, int force_wr); +struct au_do_open_args { + int no_lock; + int (*open)(struct file *file, int flags, + struct file *h_file); + struct au_fidir *fidir; + struct file *h_file; +}; +int au_do_open(struct file *file, struct au_do_open_args *args); +int au_reopen_nondir(struct file *file); +struct au_pin; +int au_ready_to_write(struct file *file, loff_t len, struct au_pin *pin); +int au_reval_and_lock_fdi(struct file *file, int (*reopen)(struct file *file), + int wlock); +int au_do_flush(struct file *file, fl_owner_t id, + int (*flush)(struct file *file, fl_owner_t id)); + +/* poll.c */ +#ifdef CONFIG_AUFS_POLL +unsigned int aufs_poll(struct file *file, poll_table *wait); +#endif + +#ifdef CONFIG_AUFS_BR_HFSPLUS +/* hfsplus.c */ +struct file *au_h_open_pre(struct dentry *dentry, aufs_bindex_t bindex, + int force_wr); +void au_h_open_post(struct dentry *dentry, aufs_bindex_t bindex, + struct file *h_file); +#else +AuStub(struct file *, au_h_open_pre, return NULL, struct dentry *dentry, + aufs_bindex_t bindex, int force_wr) +AuStubVoid(au_h_open_post, struct dentry *dentry, aufs_bindex_t bindex, + struct file *h_file); +#endif + +/* f_op.c */ +extern const struct file_operations aufs_file_fop; +int au_do_open_nondir(struct file *file, int flags, struct file *h_file); +int aufs_release_nondir(struct inode *inode __maybe_unused, struct file *file); +struct file *au_read_pre(struct file *file, int keep_fi); + +/* finfo.c */ +void au_hfput(struct au_hfile *hf, struct file *file); +void au_set_h_fptr(struct file *file, aufs_bindex_t bindex, + struct file *h_file); + +void au_update_figen(struct file *file); +struct au_fidir *au_fidir_alloc(struct super_block *sb); +int au_fidir_realloc(struct au_finfo *finfo, int nbr); + +void au_fi_init_once(void *_fi); +void au_finfo_fin(struct file *file); +int au_finfo_init(struct file *file, struct au_fidir *fidir); + +/* ioctl.c */ +long aufs_ioctl_nondir(struct file *file, unsigned int cmd, unsigned long arg); +#ifdef CONFIG_COMPAT +long aufs_compat_ioctl_dir(struct file *file, unsigned int cmd, + unsigned long arg); +long aufs_compat_ioctl_nondir(struct file *file, unsigned int cmd, + unsigned long arg); +#endif + +/* ---------------------------------------------------------------------- */ + +static inline struct au_finfo *au_fi(struct file *file) +{ + return file->private_data; +} + +/* ---------------------------------------------------------------------- */ + +/* + * fi_read_lock, fi_write_lock, + * fi_read_unlock, fi_write_unlock, fi_downgrade_lock + */ +AuSimpleRwsemFuncs(fi, struct file *f, &au_fi(f)->fi_rwsem); + +#define FiMustNoWaiters(f) AuRwMustNoWaiters(&au_fi(f)->fi_rwsem) +#define FiMustAnyLock(f) AuRwMustAnyLock(&au_fi(f)->fi_rwsem) +#define FiMustWriteLock(f) AuRwMustWriteLock(&au_fi(f)->fi_rwsem) + +/* ---------------------------------------------------------------------- */ + +/* todo: hard/soft set? */ +static inline aufs_bindex_t au_fbstart(struct file *file) +{ + FiMustAnyLock(file); + return au_fi(file)->fi_btop; +} + +static inline aufs_bindex_t au_fbend_dir(struct file *file) +{ + FiMustAnyLock(file); + AuDebugOn(!au_fi(file)->fi_hdir); + return au_fi(file)->fi_hdir->fd_bbot; +} + +static inline struct au_vdir *au_fvdir_cache(struct file *file) +{ + FiMustAnyLock(file); + AuDebugOn(!au_fi(file)->fi_hdir); + return au_fi(file)->fi_hdir->fd_vdir_cache; +} + +static inline void au_set_fbstart(struct file *file, aufs_bindex_t bindex) +{ + FiMustWriteLock(file); + au_fi(file)->fi_btop = bindex; +} + +static inline void au_set_fbend_dir(struct file *file, aufs_bindex_t bindex) +{ + FiMustWriteLock(file); + AuDebugOn(!au_fi(file)->fi_hdir); + au_fi(file)->fi_hdir->fd_bbot = bindex; +} + +static inline void au_set_fvdir_cache(struct file *file, + struct au_vdir *vdir_cache) +{ + FiMustWriteLock(file); + AuDebugOn(!au_fi(file)->fi_hdir); + au_fi(file)->fi_hdir->fd_vdir_cache = vdir_cache; +} + +static inline struct file *au_hf_top(struct file *file) +{ + FiMustAnyLock(file); + AuDebugOn(au_fi(file)->fi_hdir); + return au_fi(file)->fi_htop.hf_file; +} + +static inline struct file *au_hf_dir(struct file *file, aufs_bindex_t bindex) +{ + FiMustAnyLock(file); + AuDebugOn(!au_fi(file)->fi_hdir); + return au_fi(file)->fi_hdir->fd_hfile[0 + bindex].hf_file; +} + +/* todo: memory barrier? */ +static inline unsigned int au_figen(struct file *f) +{ + return atomic_read(&au_fi(f)->fi_generation); +} + +static inline void au_set_mmapped(struct file *f) +{ + if (atomic_inc_return(&au_fi(f)->fi_mmapped)) + return; + pr_warn("fi_mmapped wrapped around\n"); + while (!atomic_inc_return(&au_fi(f)->fi_mmapped)) + ; +} + +static inline void au_unset_mmapped(struct file *f) +{ + atomic_dec(&au_fi(f)->fi_mmapped); +} + +static inline int au_test_mmapped(struct file *f) +{ + return atomic_read(&au_fi(f)->fi_mmapped); +} + +/* customize vma->vm_file */ + +static inline void au_do_vm_file_reset(struct vm_area_struct *vma, + struct file *file) +{ + struct file *f; + + f = vma->vm_file; + get_file(file); + vma->vm_file = file; + fput(f); +} + +#ifdef CONFIG_MMU +#define AuDbgVmRegion(file, vma) do {} while (0) + +static inline void au_vm_file_reset(struct vm_area_struct *vma, + struct file *file) +{ + au_do_vm_file_reset(vma, file); +} +#else +#define AuDbgVmRegion(file, vma) \ + AuDebugOn((vma)->vm_region && (vma)->vm_region->vm_file != (file)) + +static inline void au_vm_file_reset(struct vm_area_struct *vma, + struct file *file) +{ + struct file *f; + + au_do_vm_file_reset(vma, file); + f = vma->vm_region->vm_file; + get_file(file); + vma->vm_region->vm_file = file; + fput(f); +} +#endif /* CONFIG_MMU */ + +/* handle vma->vm_prfile */ +static inline void au_vm_prfile_set(struct vm_area_struct *vma, + struct file *file) +{ + get_file(file); + vma->vm_prfile = file; +#ifndef CONFIG_MMU + get_file(file); + vma->vm_region->vm_prfile = file; +#endif +} + +#endif /* __KERNEL__ */ +#endif /* __AUFS_FILE_H__ */ diff --git a/fs/aufs/finfo.c b/fs/aufs/finfo.c new file mode 100644 index 000000000..b5eb55dfb --- /dev/null +++ b/fs/aufs/finfo.c @@ -0,0 +1,143 @@ +/* + * Copyright (C) 2005-2016 Junjiro R. Okajima + */ + +/* + * file private data + */ + +#include "aufs.h" + +void au_hfput(struct au_hfile *hf, struct file *file) +{ + /* todo: direct access f_flags */ + if (vfsub_file_flags(file) & __FMODE_EXEC) + allow_write_access(hf->hf_file); + fput(hf->hf_file); + hf->hf_file = NULL; + atomic_dec(&hf->hf_br->br_count); + hf->hf_br = NULL; +} + +void au_set_h_fptr(struct file *file, aufs_bindex_t bindex, struct file *val) +{ + struct au_finfo *finfo = au_fi(file); + struct au_hfile *hf; + struct au_fidir *fidir; + + fidir = finfo->fi_hdir; + if (!fidir) { + AuDebugOn(finfo->fi_btop != bindex); + hf = &finfo->fi_htop; + } else + hf = fidir->fd_hfile + bindex; + + if (hf && hf->hf_file) + au_hfput(hf, file); + if (val) { + FiMustWriteLock(file); + AuDebugOn(IS_ERR_OR_NULL(file->f_path.dentry)); + hf->hf_file = val; + hf->hf_br = au_sbr(file->f_path.dentry->d_sb, bindex); + } +} + +void au_update_figen(struct file *file) +{ + atomic_set(&au_fi(file)->fi_generation, au_digen(file->f_path.dentry)); + /* smp_mb(); */ /* atomic_set */ +} + +/* ---------------------------------------------------------------------- */ + +struct au_fidir *au_fidir_alloc(struct super_block *sb) +{ + struct au_fidir *fidir; + int nbr; + + nbr = au_sbend(sb) + 1; + if (nbr < 2) + nbr = 2; /* initial allocate for 2 branches */ + fidir = kzalloc(au_fidir_sz(nbr), GFP_NOFS); + if (fidir) { + fidir->fd_bbot = -1; + fidir->fd_nent = nbr; + } + + return fidir; +} + +int au_fidir_realloc(struct au_finfo *finfo, int nbr) +{ + int err; + struct au_fidir *fidir, *p; + + AuRwMustWriteLock(&finfo->fi_rwsem); + fidir = finfo->fi_hdir; + AuDebugOn(!fidir); + + err = -ENOMEM; + p = au_kzrealloc(fidir, au_fidir_sz(fidir->fd_nent), au_fidir_sz(nbr), + GFP_NOFS); + if (p) { + p->fd_nent = nbr; + finfo->fi_hdir = p; + err = 0; + } + + return err; +} + +/* ---------------------------------------------------------------------- */ + +void au_finfo_fin(struct file *file) +{ + struct au_finfo *finfo; + + au_nfiles_dec(file->f_path.dentry->d_sb); + + finfo = au_fi(file); + AuDebugOn(finfo->fi_hdir); + AuRwDestroy(&finfo->fi_rwsem); + au_cache_free_finfo(finfo); +} + +void au_fi_init_once(void *_finfo) +{ + struct au_finfo *finfo = _finfo; + static struct lock_class_key aufs_fi; + + au_rw_init(&finfo->fi_rwsem); + au_rw_class(&finfo->fi_rwsem, &aufs_fi); +} + +int au_finfo_init(struct file *file, struct au_fidir *fidir) +{ + int err; + struct au_finfo *finfo; + struct dentry *dentry; + + err = -ENOMEM; + dentry = file->f_path.dentry; + finfo = au_cache_alloc_finfo(); + if (unlikely(!finfo)) + goto out; + + err = 0; + au_nfiles_inc(dentry->d_sb); + /* verbose coding for lock class name */ + if (!fidir) + au_rw_class(&finfo->fi_rwsem, au_lc_key + AuLcNonDir_FIINFO); + else + au_rw_class(&finfo->fi_rwsem, au_lc_key + AuLcDir_FIINFO); + au_rw_write_lock(&finfo->fi_rwsem); + finfo->fi_btop = -1; + finfo->fi_hdir = fidir; + atomic_set(&finfo->fi_generation, au_digen(dentry)); + /* smp_mb(); */ /* atomic_set */ + + file->private_data = finfo; + +out: + return err; +} diff --git a/fs/aufs/fstype.h b/fs/aufs/fstype.h new file mode 100644 index 000000000..725b2ffff --- /dev/null +++ b/fs/aufs/fstype.h @@ -0,0 +1,387 @@ +/* + * Copyright (C) 2005-2016 Junjiro R. Okajima + */ + +/* + * judging filesystem type + */ + +#ifndef __AUFS_FSTYPE_H__ +#define __AUFS_FSTYPE_H__ + +#ifdef __KERNEL__ + +#include <linux/fs.h> +#include <linux/magic.h> +#include <linux/nfs_fs.h> +#include <linux/romfs_fs.h> + +static inline int au_test_aufs(struct super_block *sb) +{ + return sb->s_magic == AUFS_SUPER_MAGIC; +} + +static inline const char *au_sbtype(struct super_block *sb) +{ + return sb->s_type->name; +} + +static inline int au_test_iso9660(struct super_block *sb __maybe_unused) +{ +#if defined(CONFIG_ISO9660_FS) || defined(CONFIG_ISO9660_FS_MODULE) + return sb->s_magic == ISOFS_SUPER_MAGIC; +#else + return 0; +#endif +} + +static inline int au_test_romfs(struct super_block *sb __maybe_unused) +{ +#if defined(CONFIG_ROMFS_FS) || defined(CONFIG_ROMFS_FS_MODULE) + return sb->s_magic == ROMFS_MAGIC; +#else + return 0; +#endif +} + +static inline int au_test_cramfs(struct super_block *sb __maybe_unused) +{ +#if defined(CONFIG_CRAMFS) || defined(CONFIG_CRAMFS_MODULE) + return sb->s_magic == CRAMFS_MAGIC; +#endif + return 0; +} + +static inline int au_test_nfs(struct super_block *sb __maybe_unused) +{ +#if defined(CONFIG_NFS_FS) || defined(CONFIG_NFS_FS_MODULE) + return sb->s_magic == NFS_SUPER_MAGIC; +#else + return 0; +#endif +} + +static inline int au_test_fuse(struct super_block *sb __maybe_unused) +{ +#if defined(CONFIG_FUSE_FS) || defined(CONFIG_FUSE_FS_MODULE) + return sb->s_magic == FUSE_SUPER_MAGIC; +#else + return 0; +#endif +} + +static inline int au_test_xfs(struct super_block *sb __maybe_unused) +{ +#if defined(CONFIG_XFS_FS) || defined(CONFIG_XFS_FS_MODULE) + return sb->s_magic == XFS_SB_MAGIC; +#else + return 0; +#endif +} + +static inline int au_test_tmpfs(struct super_block *sb __maybe_unused) +{ +#ifdef CONFIG_TMPFS + return sb->s_magic == TMPFS_MAGIC; +#else + return 0; +#endif +} + +static inline int au_test_ecryptfs(struct super_block *sb __maybe_unused) +{ +#if defined(CONFIG_ECRYPT_FS) || defined(CONFIG_ECRYPT_FS_MODULE) + return !strcmp(au_sbtype(sb), "ecryptfs"); +#else + return 0; +#endif +} + +static inline int au_test_ramfs(struct super_block *sb) +{ + return sb->s_magic == RAMFS_MAGIC; +} + +static inline int au_test_ubifs(struct super_block *sb __maybe_unused) +{ +#if defined(CONFIG_UBIFS_FS) || defined(CONFIG_UBIFS_FS_MODULE) + return sb->s_magic == UBIFS_SUPER_MAGIC; +#else + return 0; +#endif +} + +static inline int au_test_procfs(struct super_block *sb __maybe_unused) +{ +#ifdef CONFIG_PROC_FS + return sb->s_magic == PROC_SUPER_MAGIC; +#else + return 0; +#endif +} + +static inline int au_test_sysfs(struct super_block *sb __maybe_unused) +{ +#ifdef CONFIG_SYSFS + return sb->s_magic == SYSFS_MAGIC; +#else + return 0; +#endif +} + +static inline int au_test_configfs(struct super_block *sb __maybe_unused) +{ +#if defined(CONFIG_CONFIGFS_FS) || defined(CONFIG_CONFIGFS_FS_MODULE) + return sb->s_magic == CONFIGFS_MAGIC; +#else + return 0; +#endif +} + +static inline int au_test_minix(struct super_block *sb __maybe_unused) +{ +#if defined(CONFIG_MINIX_FS) || defined(CONFIG_MINIX_FS_MODULE) + return sb->s_magic == MINIX3_SUPER_MAGIC + || sb->s_magic == MINIX2_SUPER_MAGIC + || sb->s_magic == MINIX2_SUPER_MAGIC2 + || sb->s_magic == MINIX_SUPER_MAGIC + || sb->s_magic == MINIX_SUPER_MAGIC2; +#else + return 0; +#endif +} + +static inline int au_test_fat(struct super_block *sb __maybe_unused) +{ +#if defined(CONFIG_FAT_FS) || defined(CONFIG_FAT_FS_MODULE) + return sb->s_magic == MSDOS_SUPER_MAGIC; +#else + return 0; +#endif +} + +static inline int au_test_msdos(struct super_block *sb) +{ + return au_test_fat(sb); +} + +static inline int au_test_vfat(struct super_block *sb) +{ + return au_test_fat(sb); +} + +static inline int au_test_securityfs(struct super_block *sb __maybe_unused) +{ +#ifdef CONFIG_SECURITYFS + return sb->s_magic == SECURITYFS_MAGIC; +#else + return 0; +#endif +} + +static inline int au_test_squashfs(struct super_block *sb __maybe_unused) +{ +#if defined(CONFIG_SQUASHFS) || defined(CONFIG_SQUASHFS_MODULE) + return sb->s_magic == SQUASHFS_MAGIC; +#else + return 0; +#endif +} + +static inline int au_test_btrfs(struct super_block *sb __maybe_unused) +{ +#if defined(CONFIG_BTRFS_FS) || defined(CONFIG_BTRFS_FS_MODULE) + return sb->s_magic == BTRFS_SUPER_MAGIC; +#else + return 0; +#endif +} + +static inline int au_test_xenfs(struct super_block *sb __maybe_unused) +{ +#if defined(CONFIG_XENFS) || defined(CONFIG_XENFS_MODULE) + return sb->s_magic == XENFS_SUPER_MAGIC; +#else + return 0; +#endif +} + +static inline int au_test_debugfs(struct super_block *sb __maybe_unused) +{ +#ifdef CONFIG_DEBUG_FS + return sb->s_magic == DEBUGFS_MAGIC; +#else + return 0; +#endif +} + +static inline int au_test_nilfs(struct super_block *sb __maybe_unused) +{ +#if defined(CONFIG_NILFS) || defined(CONFIG_NILFS_MODULE) + return sb->s_magic == NILFS_SUPER_MAGIC; +#else + return 0; +#endif +} + +static inline int au_test_hfsplus(struct super_block *sb __maybe_unused) +{ +#if defined(CONFIG_HFSPLUS_FS) || defined(CONFIG_HFSPLUS_FS_MODULE) + return sb->s_magic == HFSPLUS_SUPER_MAGIC; +#else + return 0; +#endif +} + +/* ---------------------------------------------------------------------- */ +/* + * they can't be an aufs branch. + */ +static inline int au_test_fs_unsuppoted(struct super_block *sb) +{ + return +#ifndef CONFIG_AUFS_BR_RAMFS + au_test_ramfs(sb) || +#endif + au_test_procfs(sb) + || au_test_sysfs(sb) + || au_test_configfs(sb) + || au_test_debugfs(sb) + || au_test_securityfs(sb) + || au_test_xenfs(sb) + || au_test_ecryptfs(sb) + /* || !strcmp(au_sbtype(sb), "unionfs") */ + || au_test_aufs(sb); /* will be supported in next version */ +} + +static inline int au_test_fs_remote(struct super_block *sb) +{ + return !au_test_tmpfs(sb) +#ifdef CONFIG_AUFS_BR_RAMFS + && !au_test_ramfs(sb) +#endif + && !(sb->s_type->fs_flags & FS_REQUIRES_DEV); +} + +/* ---------------------------------------------------------------------- */ + +/* + * Note: these functions (below) are created after reading ->getattr() in all + * filesystems under linux/fs. it means we have to do so in every update... + */ + +/* + * some filesystems require getattr to refresh the inode attributes before + * referencing. + * in most cases, we can rely on the inode attribute in NFS (or every remote fs) + * and leave the work for d_revalidate() + */ +static inline int au_test_fs_refresh_iattr(struct super_block *sb) +{ + return au_test_nfs(sb) + || au_test_fuse(sb) + /* || au_test_btrfs(sb) */ /* untested */ + ; +} + +/* + * filesystems which don't maintain i_size or i_blocks. + */ +static inline int au_test_fs_bad_iattr_size(struct super_block *sb) +{ + return au_test_xfs(sb) + || au_test_btrfs(sb) + || au_test_ubifs(sb) + || au_test_hfsplus(sb) /* maintained, but incorrect */ + /* || au_test_minix(sb) */ /* untested */ + ; +} + +/* + * filesystems which don't store the correct value in some of their inode + * attributes. + */ +static inline int au_test_fs_bad_iattr(struct super_block *sb) +{ + return au_test_fs_bad_iattr_size(sb) + || au_test_fat(sb) + || au_test_msdos(sb) + || au_test_vfat(sb); +} + +/* they don't check i_nlink in link(2) */ +static inline int au_test_fs_no_limit_nlink(struct super_block *sb) +{ + return au_test_tmpfs(sb) +#ifdef CONFIG_AUFS_BR_RAMFS + || au_test_ramfs(sb) +#endif + || au_test_ubifs(sb) + || au_test_hfsplus(sb); +} + +/* + * filesystems which sets S_NOATIME and S_NOCMTIME. + */ +static inline int au_test_fs_notime(struct super_block *sb) +{ + return au_test_nfs(sb) + || au_test_fuse(sb) + || au_test_ubifs(sb) + ; +} + +/* temporary support for i#1 in cramfs */ +static inline int au_test_fs_unique_ino(struct inode *inode) +{ + if (au_test_cramfs(inode->i_sb)) + return inode->i_ino != 1; + return 1; +} + +/* ---------------------------------------------------------------------- */ + +/* + * the filesystem where the xino files placed must support i/o after unlink and + * maintain i_size and i_blocks. + */ +static inline int au_test_fs_bad_xino(struct super_block *sb) +{ + return au_test_fs_remote(sb) + || au_test_fs_bad_iattr_size(sb) + /* don't want unnecessary work for xino */ + || au_test_aufs(sb) + || au_test_ecryptfs(sb) + || au_test_nilfs(sb); +} + +static inline int au_test_fs_trunc_xino(struct super_block *sb) +{ + return au_test_tmpfs(sb) + || au_test_ramfs(sb); +} + +/* + * test if the @sb is real-readonly. + */ +static inline int au_test_fs_rr(struct super_block *sb) +{ + return au_test_squashfs(sb) + || au_test_iso9660(sb) + || au_test_cramfs(sb) + || au_test_romfs(sb); +} + +/* + * test if the @inode is nfs with 'noacl' option + * NFS always sets MS_POSIXACL regardless its mount option 'noacl.' + */ +static inline int au_test_nfs_noacl(struct inode *inode) +{ + return au_test_nfs(inode->i_sb) + /* && IS_POSIXACL(inode) */ + && !nfs_server_capable(inode, NFS_CAP_ACLS); +} + +#endif /* __KERNEL__ */ +#endif /* __AUFS_FSTYPE_H__ */ diff --git a/fs/aufs/hfsnotify.c b/fs/aufs/hfsnotify.c new file mode 100644 index 000000000..c0a1a63a9 --- /dev/null +++ b/fs/aufs/hfsnotify.c @@ -0,0 +1,275 @@ +/* + * Copyright (C) 2005-2016 Junjiro R. Okajima + */ + +/* + * fsnotify for the lower directories + */ + +#include "aufs.h" + +/* FS_IN_IGNORED is unnecessary */ +static const __u32 AuHfsnMask = (FS_MOVED_TO | FS_MOVED_FROM | FS_DELETE + | FS_CREATE | FS_EVENT_ON_CHILD); +static DECLARE_WAIT_QUEUE_HEAD(au_hfsn_wq); +static __cacheline_aligned_in_smp atomic64_t au_hfsn_ifree = ATOMIC64_INIT(0); + +static void au_hfsn_free_mark(struct fsnotify_mark *mark) +{ + struct au_hnotify *hn = container_of(mark, struct au_hnotify, + hn_mark); + AuDbg("here\n"); + au_cache_free_hnotify(hn); + smp_mb__before_atomic(); + if (atomic64_dec_and_test(&au_hfsn_ifree)) + wake_up(&au_hfsn_wq); +} + +static int au_hfsn_alloc(struct au_hinode *hinode) +{ + int err; + struct au_hnotify *hn; + struct super_block *sb; + struct au_branch *br; + struct fsnotify_mark *mark; + aufs_bindex_t bindex; + + hn = hinode->hi_notify; + sb = hn->hn_aufs_inode->i_sb; + bindex = au_br_index(sb, hinode->hi_id); + br = au_sbr(sb, bindex); + AuDebugOn(!br->br_hfsn); + + mark = &hn->hn_mark; + fsnotify_init_mark(mark, au_hfsn_free_mark); + mark->mask = AuHfsnMask; + /* + * by udba rename or rmdir, aufs assign a new inode to the known + * h_inode, so specify 1 to allow dups. + */ + lockdep_off(); + err = fsnotify_add_mark(mark, br->br_hfsn->hfsn_group, hinode->hi_inode, + /*mnt*/NULL, /*allow_dups*/1); + /* even if err */ + fsnotify_put_mark(mark); + lockdep_on(); + + return err; +} + +static int au_hfsn_free(struct au_hinode *hinode, struct au_hnotify *hn) +{ + struct fsnotify_mark *mark; + unsigned long long ull; + struct fsnotify_group *group; + + ull = atomic64_inc_return(&au_hfsn_ifree); + BUG_ON(!ull); + + mark = &hn->hn_mark; + spin_lock(&mark->lock); + group = mark->group; + fsnotify_get_group(group); + spin_unlock(&mark->lock); + lockdep_off(); + fsnotify_destroy_mark(mark, group); + fsnotify_put_group(group); + lockdep_on(); + + /* free hn by myself */ + return 0; +} + +/* ---------------------------------------------------------------------- */ + +static void au_hfsn_ctl(struct au_hinode *hinode, int do_set) +{ + struct fsnotify_mark *mark; + + mark = &hinode->hi_notify->hn_mark; + spin_lock(&mark->lock); + if (do_set) { + AuDebugOn(mark->mask & AuHfsnMask); + mark->mask |= AuHfsnMask; + } else { + AuDebugOn(!(mark->mask & AuHfsnMask)); + mark->mask &= ~AuHfsnMask; + } + spin_unlock(&mark->lock); + /* fsnotify_recalc_inode_mask(hinode->hi_inode); */ +} + +/* ---------------------------------------------------------------------- */ + +/* #define AuDbgHnotify */ +#ifdef AuDbgHnotify +static char *au_hfsn_name(u32 mask) +{ +#ifdef CONFIG_AUFS_DEBUG +#define test_ret(flag) \ + do { \ + if (mask & flag) \ + return #flag; \ + } while (0) + test_ret(FS_ACCESS); + test_ret(FS_MODIFY); + test_ret(FS_ATTRIB); + test_ret(FS_CLOSE_WRITE); + test_ret(FS_CLOSE_NOWRITE); + test_ret(FS_OPEN); + test_ret(FS_MOVED_FROM); + test_ret(FS_MOVED_TO); + test_ret(FS_CREATE); + test_ret(FS_DELETE); + test_ret(FS_DELETE_SELF); + test_ret(FS_MOVE_SELF); + test_ret(FS_UNMOUNT); + test_ret(FS_Q_OVERFLOW); + test_ret(FS_IN_IGNORED); + test_ret(FS_ISDIR); + test_ret(FS_IN_ONESHOT); + test_ret(FS_EVENT_ON_CHILD); + return ""; +#undef test_ret +#else + return "??"; +#endif +} +#endif + +/* ---------------------------------------------------------------------- */ + +static void au_hfsn_free_group(struct fsnotify_group *group) +{ + struct au_br_hfsnotify *hfsn = group->private; + + AuDbg("here\n"); + kfree(hfsn); +} + +static int au_hfsn_handle_event(struct fsnotify_group *group, + struct inode *inode, + struct fsnotify_mark *inode_mark, + struct fsnotify_mark *vfsmount_mark, + u32 mask, void *data, int data_type, + const unsigned char *file_name, u32 cookie) +{ + int err; + struct au_hnotify *hnotify; + struct inode *h_dir, *h_inode; + struct qstr h_child_qstr = QSTR_INIT(file_name, strlen(file_name)); + + AuDebugOn(data_type != FSNOTIFY_EVENT_INODE); + + err = 0; + /* if FS_UNMOUNT happens, there must be another bug */ + AuDebugOn(mask & FS_UNMOUNT); + if (mask & (FS_IN_IGNORED | FS_UNMOUNT)) + goto out; + + h_dir = inode; + h_inode = NULL; +#ifdef AuDbgHnotify + au_debug_on(); + if (1 || h_child_qstr.len != sizeof(AUFS_XINO_FNAME) - 1 + || strncmp(h_child_qstr.name, AUFS_XINO_FNAME, h_child_qstr.len)) { + AuDbg("i%lu, mask 0x%x %s, hcname %.*s, hi%lu\n", + h_dir->i_ino, mask, au_hfsn_name(mask), + AuLNPair(&h_child_qstr), h_inode ? h_inode->i_ino : 0); + /* WARN_ON(1); */ + } + au_debug_off(); +#endif + + AuDebugOn(!inode_mark); + hnotify = container_of(inode_mark, struct au_hnotify, hn_mark); + err = au_hnotify(h_dir, hnotify, mask, &h_child_qstr, h_inode); + +out: + return err; +} + +static struct fsnotify_ops au_hfsn_ops = { + .handle_event = au_hfsn_handle_event, + .free_group_priv = au_hfsn_free_group +}; + +/* ---------------------------------------------------------------------- */ + +static void au_hfsn_fin_br(struct au_branch *br) +{ + struct au_br_hfsnotify *hfsn; + + hfsn = br->br_hfsn; + if (hfsn) { + lockdep_off(); + fsnotify_put_group(hfsn->hfsn_group); + lockdep_on(); + } +} + +static int au_hfsn_init_br(struct au_branch *br, int perm) +{ + int err; + struct fsnotify_group *group; + struct au_br_hfsnotify *hfsn; + + err = 0; + br->br_hfsn = NULL; + if (!au_br_hnotifyable(perm)) + goto out; + + err = -ENOMEM; + hfsn = kmalloc(sizeof(*hfsn), GFP_NOFS); + if (unlikely(!hfsn)) + goto out; + + err = 0; + group = fsnotify_alloc_group(&au_hfsn_ops); + if (IS_ERR(group)) { + err = PTR_ERR(group); + pr_err("fsnotify_alloc_group() failed, %d\n", err); + goto out_hfsn; + } + + group->private = hfsn; + hfsn->hfsn_group = group; + br->br_hfsn = hfsn; + goto out; /* success */ + +out_hfsn: + kfree(hfsn); +out: + return err; +} + +static int au_hfsn_reset_br(unsigned int udba, struct au_branch *br, int perm) +{ + int err; + + err = 0; + if (!br->br_hfsn) + err = au_hfsn_init_br(br, perm); + + return err; +} + +/* ---------------------------------------------------------------------- */ + +static void au_hfsn_fin(void) +{ + AuDbg("au_hfsn_ifree %lld\n", (long long)atomic64_read(&au_hfsn_ifree)); + wait_event(au_hfsn_wq, !atomic64_read(&au_hfsn_ifree)); +} + +const struct au_hnotify_op au_hnotify_op = { + .ctl = au_hfsn_ctl, + .alloc = au_hfsn_alloc, + .free = au_hfsn_free, + + .fin = au_hfsn_fin, + + .reset_br = au_hfsn_reset_br, + .fin_br = au_hfsn_fin_br, + .init_br = au_hfsn_init_br +}; diff --git a/fs/aufs/hfsplus.c b/fs/aufs/hfsplus.c new file mode 100644 index 000000000..145c6ac2f --- /dev/null +++ b/fs/aufs/hfsplus.c @@ -0,0 +1,43 @@ +/* + * Copyright (C) 2010-2016 Junjiro R. Okajima + */ + +/* + * special support for filesystems which aqucires an inode mutex + * at final closing a file, eg, hfsplus. + * + * This trick is very simple and stupid, just to open the file before really + * neceeary open to tell hfsplus that this is not the final closing. + * The caller should call au_h_open_pre() after acquiring the inode mutex, + * and au_h_open_post() after releasing it. + */ + +#include "aufs.h" + +struct file *au_h_open_pre(struct dentry *dentry, aufs_bindex_t bindex, + int force_wr) +{ + struct file *h_file; + struct dentry *h_dentry; + + h_dentry = au_h_dptr(dentry, bindex); + AuDebugOn(!h_dentry); + AuDebugOn(d_is_negative(h_dentry)); + + h_file = NULL; + if (au_test_hfsplus(h_dentry->d_sb) + && d_is_reg(h_dentry)) + h_file = au_h_open(dentry, bindex, + O_RDONLY | O_NOATIME | O_LARGEFILE, + /*file*/NULL, force_wr); + return h_file; +} + +void au_h_open_post(struct dentry *dentry, aufs_bindex_t bindex, + struct file *h_file) +{ + if (h_file) { + fput(h_file); + au_sbr_put(dentry->d_sb, bindex); + } +} diff --git a/fs/aufs/hnotify.c b/fs/aufs/hnotify.c new file mode 100644 index 000000000..92e432b6a --- /dev/null +++ b/fs/aufs/hnotify.c @@ -0,0 +1,697 @@ +/* + * Copyright (C) 2005-2016 Junjiro R. Okajima + */ + +/* + * abstraction to notify the direct changes on lower directories + */ + +#include "aufs.h" + +int au_hn_alloc(struct au_hinode *hinode, struct inode *inode) +{ + int err; + struct au_hnotify *hn; + + err = -ENOMEM; + hn = au_cache_alloc_hnotify(); + if (hn) { + hn->hn_aufs_inode = inode; + hinode->hi_notify = hn; + err = au_hnotify_op.alloc(hinode); + AuTraceErr(err); + if (unlikely(err)) { + hinode->hi_notify = NULL; + au_cache_free_hnotify(hn); + /* + * The upper dir was removed by udba, but the same named + * dir left. In this case, aufs assignes a new inode + * number and set the monitor again. + * For the lower dir, the old monitnor is still left. + */ + if (err == -EEXIST) + err = 0; + } + } + + AuTraceErr(err); + return err; +} + +void au_hn_free(struct au_hinode *hinode) +{ + struct au_hnotify *hn; + + hn = hinode->hi_notify; + if (hn) { + hinode->hi_notify = NULL; + if (au_hnotify_op.free(hinode, hn)) + au_cache_free_hnotify(hn); + } +} + +/* ---------------------------------------------------------------------- */ + +void au_hn_ctl(struct au_hinode *hinode, int do_set) +{ + if (hinode->hi_notify) + au_hnotify_op.ctl(hinode, do_set); +} + +void au_hn_reset(struct inode *inode, unsigned int flags) +{ + aufs_bindex_t bindex, bend; + struct inode *hi; + struct dentry *iwhdentry; + + bend = au_ibend(inode); + for (bindex = au_ibstart(inode); bindex <= bend; bindex++) { + hi = au_h_iptr(inode, bindex); + if (!hi) + continue; + + /* inode_lock_nested(hi, AuLsc_I_CHILD); */ + iwhdentry = au_hi_wh(inode, bindex); + if (iwhdentry) + dget(iwhdentry); + au_igrab(hi); + au_set_h_iptr(inode, bindex, NULL, 0); + au_set_h_iptr(inode, bindex, au_igrab(hi), + flags & ~AuHi_XINO); + iput(hi); + dput(iwhdentry); + /* inode_unlock(hi); */ + } +} + +/* ---------------------------------------------------------------------- */ + +static int hn_xino(struct inode *inode, struct inode *h_inode) +{ + int err; + aufs_bindex_t bindex, bend, bfound, bstart; + struct inode *h_i; + + err = 0; + if (unlikely(inode->i_ino == AUFS_ROOT_INO)) { + pr_warn("branch root dir was changed\n"); + goto out; + } + + bfound = -1; + bend = au_ibend(inode); + bstart = au_ibstart(inode); +#if 0 /* reserved for future use */ + if (bindex == bend) { + /* keep this ino in rename case */ + goto out; + } +#endif + for (bindex = bstart; bindex <= bend; bindex++) + if (au_h_iptr(inode, bindex) == h_inode) { + bfound = bindex; + break; + } + if (bfound < 0) + goto out; + + for (bindex = bstart; bindex <= bend; bindex++) { + h_i = au_h_iptr(inode, bindex); + if (!h_i) + continue; + + err = au_xino_write(inode->i_sb, bindex, h_i->i_ino, /*ino*/0); + /* ignore this error */ + /* bad action? */ + } + + /* children inode number will be broken */ + +out: + AuTraceErr(err); + return err; +} + +static int hn_gen_tree(struct dentry *dentry) +{ + int err, i, j, ndentry; + struct au_dcsub_pages dpages; + struct au_dpage *dpage; + struct dentry **dentries; + + err = au_dpages_init(&dpages, GFP_NOFS); + if (unlikely(err)) + goto out; + err = au_dcsub_pages(&dpages, dentry, NULL, NULL); + if (unlikely(err)) + goto out_dpages; + + for (i = 0; i < dpages.ndpage; i++) { + dpage = dpages.dpages + i; + dentries = dpage->dentries; + ndentry = dpage->ndentry; + for (j = 0; j < ndentry; j++) { + struct dentry *d; + + d = dentries[j]; + if (IS_ROOT(d)) + continue; + + au_digen_dec(d); + if (d_really_is_positive(d)) + /* todo: reset children xino? + cached children only? */ + au_iigen_dec(d_inode(d)); + } + } + +out_dpages: + au_dpages_free(&dpages); + +#if 0 + /* discard children */ + dentry_unhash(dentry); + dput(dentry); +#endif +out: + return err; +} + +/* + * return 0 if processed. + */ +static int hn_gen_by_inode(char *name, unsigned int nlen, struct inode *inode, + const unsigned int isdir) +{ + int err; + struct dentry *d; + struct qstr *dname; + + err = 1; + if (unlikely(inode->i_ino == AUFS_ROOT_INO)) { + pr_warn("branch root dir was changed\n"); + err = 0; + goto out; + } + + if (!isdir) { + AuDebugOn(!name); + au_iigen_dec(inode); + spin_lock(&inode->i_lock); + hlist_for_each_entry(d, &inode->i_dentry, d_u.d_alias) { + spin_lock(&d->d_lock); + dname = &d->d_name; + if (dname->len != nlen + && memcmp(dname->name, name, nlen)) { + spin_unlock(&d->d_lock); + continue; + } + err = 0; + au_digen_dec(d); + spin_unlock(&d->d_lock); + break; + } + spin_unlock(&inode->i_lock); + } else { + au_fset_si(au_sbi(inode->i_sb), FAILED_REFRESH_DIR); + d = d_find_any_alias(inode); + if (!d) { + au_iigen_dec(inode); + goto out; + } + + spin_lock(&d->d_lock); + dname = &d->d_name; + if (dname->len == nlen && !memcmp(dname->name, name, nlen)) { + spin_unlock(&d->d_lock); + err = hn_gen_tree(d); + spin_lock(&d->d_lock); + } + spin_unlock(&d->d_lock); + dput(d); + } + +out: + AuTraceErr(err); + return err; +} + +static int hn_gen_by_name(struct dentry *dentry, const unsigned int isdir) +{ + int err; + + if (IS_ROOT(dentry)) { + pr_warn("branch root dir was changed\n"); + return 0; + } + + err = 0; + if (!isdir) { + au_digen_dec(dentry); + if (d_really_is_positive(dentry)) + au_iigen_dec(d_inode(dentry)); + } else { + au_fset_si(au_sbi(dentry->d_sb), FAILED_REFRESH_DIR); + if (d_really_is_positive(dentry)) + err = hn_gen_tree(dentry); + } + + AuTraceErr(err); + return err; +} + +/* ---------------------------------------------------------------------- */ + +/* hnotify job flags */ +#define AuHnJob_XINO0 1 +#define AuHnJob_GEN (1 << 1) +#define AuHnJob_DIRENT (1 << 2) +#define AuHnJob_ISDIR (1 << 3) +#define AuHnJob_TRYXINO0 (1 << 4) +#define AuHnJob_MNTPNT (1 << 5) +#define au_ftest_hnjob(flags, name) ((flags) & AuHnJob_##name) +#define au_fset_hnjob(flags, name) \ + do { (flags) |= AuHnJob_##name; } while (0) +#define au_fclr_hnjob(flags, name) \ + do { (flags) &= ~AuHnJob_##name; } while (0) + +enum { + AuHn_CHILD, + AuHn_PARENT, + AuHnLast +}; + +struct au_hnotify_args { + struct inode *h_dir, *dir, *h_child_inode; + u32 mask; + unsigned int flags[AuHnLast]; + unsigned int h_child_nlen; + char h_child_name[]; +}; + +struct hn_job_args { + unsigned int flags; + struct inode *inode, *h_inode, *dir, *h_dir; + struct dentry *dentry; + char *h_name; + int h_nlen; +}; + +static int hn_job(struct hn_job_args *a) +{ + const unsigned int isdir = au_ftest_hnjob(a->flags, ISDIR); + int e; + + /* reset xino */ + if (au_ftest_hnjob(a->flags, XINO0) && a->inode) + hn_xino(a->inode, a->h_inode); /* ignore this error */ + + if (au_ftest_hnjob(a->flags, TRYXINO0) + && a->inode + && a->h_inode) { + inode_lock_nested(a->h_inode, AuLsc_I_CHILD); + if (!a->h_inode->i_nlink + && !(a->h_inode->i_state & I_LINKABLE)) + hn_xino(a->inode, a->h_inode); /* ignore this error */ + inode_unlock(a->h_inode); + } + + /* make the generation obsolete */ + if (au_ftest_hnjob(a->flags, GEN)) { + e = -1; + if (a->inode) + e = hn_gen_by_inode(a->h_name, a->h_nlen, a->inode, + isdir); + if (e && a->dentry) + hn_gen_by_name(a->dentry, isdir); + /* ignore this error */ + } + + /* make dir entries obsolete */ + if (au_ftest_hnjob(a->flags, DIRENT) && a->inode) { + struct au_vdir *vdir; + + vdir = au_ivdir(a->inode); + if (vdir) + vdir->vd_jiffy = 0; + /* IMustLock(a->inode); */ + /* a->inode->i_version++; */ + } + + /* can do nothing but warn */ + if (au_ftest_hnjob(a->flags, MNTPNT) + && a->dentry + && d_mountpoint(a->dentry)) + pr_warn("mount-point %pd is removed or renamed\n", a->dentry); + + return 0; +} + +/* ---------------------------------------------------------------------- */ + +static struct dentry *lookup_wlock_by_name(char *name, unsigned int nlen, + struct inode *dir) +{ + struct dentry *dentry, *d, *parent; + struct qstr *dname; + + parent = d_find_any_alias(dir); + if (!parent) + return NULL; + + dentry = NULL; + spin_lock(&parent->d_lock); + list_for_each_entry(d, &parent->d_subdirs, d_child) { + /* AuDbg("%pd\n", d); */ + spin_lock_nested(&d->d_lock, DENTRY_D_LOCK_NESTED); + dname = &d->d_name; + if (dname->len != nlen || memcmp(dname->name, name, nlen)) + goto cont_unlock; + if (au_di(d)) + au_digen_dec(d); + else + goto cont_unlock; + if (au_dcount(d) > 0) { + dentry = dget_dlock(d); + spin_unlock(&d->d_lock); + break; + } + +cont_unlock: + spin_unlock(&d->d_lock); + } + spin_unlock(&parent->d_lock); + dput(parent); + + if (dentry) + di_write_lock_child(dentry); + + return dentry; +} + +static struct inode *lookup_wlock_by_ino(struct super_block *sb, + aufs_bindex_t bindex, ino_t h_ino) +{ + struct inode *inode; + ino_t ino; + int err; + + inode = NULL; + err = au_xino_read(sb, bindex, h_ino, &ino); + if (!err && ino) + inode = ilookup(sb, ino); + if (!inode) + goto out; + + if (unlikely(inode->i_ino == AUFS_ROOT_INO)) { + pr_warn("wrong root branch\n"); + iput(inode); + inode = NULL; + goto out; + } + + ii_write_lock_child(inode); + +out: + return inode; +} + +static void au_hn_bh(void *_args) +{ + struct au_hnotify_args *a = _args; + struct super_block *sb; + aufs_bindex_t bindex, bend, bfound; + unsigned char xino, try_iput; + int err; + struct inode *inode; + ino_t h_ino; + struct hn_job_args args; + struct dentry *dentry; + struct au_sbinfo *sbinfo; + + AuDebugOn(!_args); + AuDebugOn(!a->h_dir); + AuDebugOn(!a->dir); + AuDebugOn(!a->mask); + AuDbg("mask 0x%x, i%lu, hi%lu, hci%lu\n", + a->mask, a->dir->i_ino, a->h_dir->i_ino, + a->h_child_inode ? a->h_child_inode->i_ino : 0); + + inode = NULL; + dentry = NULL; + /* + * do not lock a->dir->i_mutex here + * because of d_revalidate() may cause a deadlock. + */ + sb = a->dir->i_sb; + AuDebugOn(!sb); + sbinfo = au_sbi(sb); + AuDebugOn(!sbinfo); + si_write_lock(sb, AuLock_NOPLMW); + + ii_read_lock_parent(a->dir); + bfound = -1; + bend = au_ibend(a->dir); + for (bindex = au_ibstart(a->dir); bindex <= bend; bindex++) + if (au_h_iptr(a->dir, bindex) == a->h_dir) { + bfound = bindex; + break; + } + ii_read_unlock(a->dir); + if (unlikely(bfound < 0)) + goto out; + + xino = !!au_opt_test(au_mntflags(sb), XINO); + h_ino = 0; + if (a->h_child_inode) + h_ino = a->h_child_inode->i_ino; + + if (a->h_child_nlen + && (au_ftest_hnjob(a->flags[AuHn_CHILD], GEN) + || au_ftest_hnjob(a->flags[AuHn_CHILD], MNTPNT))) + dentry = lookup_wlock_by_name(a->h_child_name, a->h_child_nlen, + a->dir); + try_iput = 0; + if (dentry && d_really_is_positive(dentry)) + inode = d_inode(dentry); + if (xino && !inode && h_ino + && (au_ftest_hnjob(a->flags[AuHn_CHILD], XINO0) + || au_ftest_hnjob(a->flags[AuHn_CHILD], TRYXINO0) + || au_ftest_hnjob(a->flags[AuHn_CHILD], GEN))) { + inode = lookup_wlock_by_ino(sb, bfound, h_ino); + try_iput = 1; + } + + args.flags = a->flags[AuHn_CHILD]; + args.dentry = dentry; + args.inode = inode; + args.h_inode = a->h_child_inode; + args.dir = a->dir; + args.h_dir = a->h_dir; + args.h_name = a->h_child_name; + args.h_nlen = a->h_child_nlen; + err = hn_job(&args); + if (dentry) { + if (au_di(dentry)) + di_write_unlock(dentry); + dput(dentry); + } + if (inode && try_iput) { + ii_write_unlock(inode); + iput(inode); + } + + ii_write_lock_parent(a->dir); + args.flags = a->flags[AuHn_PARENT]; + args.dentry = NULL; + args.inode = a->dir; + args.h_inode = a->h_dir; + args.dir = NULL; + args.h_dir = NULL; + args.h_name = NULL; + args.h_nlen = 0; + err = hn_job(&args); + ii_write_unlock(a->dir); + +out: + iput(a->h_child_inode); + iput(a->h_dir); + iput(a->dir); + si_write_unlock(sb); + au_nwt_done(&sbinfo->si_nowait); + kfree(a); +} + +/* ---------------------------------------------------------------------- */ + +int au_hnotify(struct inode *h_dir, struct au_hnotify *hnotify, u32 mask, + struct qstr *h_child_qstr, struct inode *h_child_inode) +{ + int err, len; + unsigned int flags[AuHnLast], f; + unsigned char isdir, isroot, wh; + struct inode *dir; + struct au_hnotify_args *args; + char *p, *h_child_name; + + err = 0; + AuDebugOn(!hnotify || !hnotify->hn_aufs_inode); + dir = igrab(hnotify->hn_aufs_inode); + if (!dir) + goto out; + + isroot = (dir->i_ino == AUFS_ROOT_INO); + wh = 0; + h_child_name = (void *)h_child_qstr->name; + len = h_child_qstr->len; + if (h_child_name) { + if (len > AUFS_WH_PFX_LEN + && !memcmp(h_child_name, AUFS_WH_PFX, AUFS_WH_PFX_LEN)) { + h_child_name += AUFS_WH_PFX_LEN; + len -= AUFS_WH_PFX_LEN; + wh = 1; + } + } + + isdir = 0; + if (h_child_inode) + isdir = !!S_ISDIR(h_child_inode->i_mode); + flags[AuHn_PARENT] = AuHnJob_ISDIR; + flags[AuHn_CHILD] = 0; + if (isdir) + flags[AuHn_CHILD] = AuHnJob_ISDIR; + au_fset_hnjob(flags[AuHn_PARENT], DIRENT); + au_fset_hnjob(flags[AuHn_CHILD], GEN); + switch (mask & FS_EVENTS_POSS_ON_CHILD) { + case FS_MOVED_FROM: + case FS_MOVED_TO: + au_fset_hnjob(flags[AuHn_CHILD], XINO0); + au_fset_hnjob(flags[AuHn_CHILD], MNTPNT); + /*FALLTHROUGH*/ + case FS_CREATE: + AuDebugOn(!h_child_name); + break; + + case FS_DELETE: + /* + * aufs never be able to get this child inode. + * revalidation should be in d_revalidate() + * by checking i_nlink, i_generation or d_unhashed(). + */ + AuDebugOn(!h_child_name); + au_fset_hnjob(flags[AuHn_CHILD], TRYXINO0); + au_fset_hnjob(flags[AuHn_CHILD], MNTPNT); + break; + + default: + AuDebugOn(1); + } + + if (wh) + h_child_inode = NULL; + + err = -ENOMEM; + /* iput() and kfree() will be called in au_hnotify() */ + args = kmalloc(sizeof(*args) + len + 1, GFP_NOFS); + if (unlikely(!args)) { + AuErr1("no memory\n"); + iput(dir); + goto out; + } + args->flags[AuHn_PARENT] = flags[AuHn_PARENT]; + args->flags[AuHn_CHILD] = flags[AuHn_CHILD]; + args->mask = mask; + args->dir = dir; + args->h_dir = igrab(h_dir); + if (h_child_inode) + h_child_inode = igrab(h_child_inode); /* can be NULL */ + args->h_child_inode = h_child_inode; + args->h_child_nlen = len; + if (len) { + p = (void *)args; + p += sizeof(*args); + memcpy(p, h_child_name, len); + p[len] = 0; + } + + /* NFS fires the event for silly-renamed one from kworker */ + f = 0; + if (!dir->i_nlink + || (au_test_nfs(h_dir->i_sb) && (mask & FS_DELETE))) + f = AuWkq_NEST; + err = au_wkq_nowait(au_hn_bh, args, dir->i_sb, f); + if (unlikely(err)) { + pr_err("wkq %d\n", err); + iput(args->h_child_inode); + iput(args->h_dir); + iput(args->dir); + kfree(args); + } + +out: + return err; +} + +/* ---------------------------------------------------------------------- */ + +int au_hnotify_reset_br(unsigned int udba, struct au_branch *br, int perm) +{ + int err; + + AuDebugOn(!(udba & AuOptMask_UDBA)); + + err = 0; + if (au_hnotify_op.reset_br) + err = au_hnotify_op.reset_br(udba, br, perm); + + return err; +} + +int au_hnotify_init_br(struct au_branch *br, int perm) +{ + int err; + + err = 0; + if (au_hnotify_op.init_br) + err = au_hnotify_op.init_br(br, perm); + + return err; +} + +void au_hnotify_fin_br(struct au_branch *br) +{ + if (au_hnotify_op.fin_br) + au_hnotify_op.fin_br(br); +} + +static void au_hn_destroy_cache(void) +{ + kmem_cache_destroy(au_cachep[AuCache_HNOTIFY]); + au_cachep[AuCache_HNOTIFY] = NULL; +} + +int __init au_hnotify_init(void) +{ + int err; + + err = -ENOMEM; + au_cachep[AuCache_HNOTIFY] = AuCache(au_hnotify); + if (au_cachep[AuCache_HNOTIFY]) { + err = 0; + if (au_hnotify_op.init) + err = au_hnotify_op.init(); + if (unlikely(err)) + au_hn_destroy_cache(); + } + AuTraceErr(err); + return err; +} + +void au_hnotify_fin(void) +{ + if (au_hnotify_op.fin) + au_hnotify_op.fin(); + /* cf. au_cache_fin() */ + if (au_cachep[AuCache_HNOTIFY]) + au_hn_destroy_cache(); +} diff --git a/fs/aufs/i_op.c b/fs/aufs/i_op.c new file mode 100644 index 000000000..a70d5b050 --- /dev/null +++ b/fs/aufs/i_op.c @@ -0,0 +1,1406 @@ +/* + * Copyright (C) 2005-2016 Junjiro R. Okajima + */ + +/* + * inode operations (except add/del/rename) + */ + +#include <linux/device_cgroup.h> +#include <linux/fs_stack.h> +#include <linux/namei.h> +#include <linux/security.h> +#include "aufs.h" + +static int h_permission(struct inode *h_inode, int mask, + struct vfsmount *h_mnt, int brperm) +{ + int err; + const unsigned char write_mask = !!(mask & (MAY_WRITE | MAY_APPEND)); + + err = -EACCES; + if ((write_mask && IS_IMMUTABLE(h_inode)) + || ((mask & MAY_EXEC) + && S_ISREG(h_inode->i_mode) + && ((h_mnt->mnt_flags & MNT_NOEXEC) + || !(h_inode->i_mode & S_IXUGO)))) + goto out; + + /* + * - skip the lower fs test in the case of write to ro branch. + * - nfs dir permission write check is optimized, but a policy for + * link/rename requires a real check. + * - nfs always sets MS_POSIXACL regardless its mount option 'noacl.' + * in this case, generic_permission() returns -EOPNOTSUPP. + */ + if ((write_mask && !au_br_writable(brperm)) + || (au_test_nfs(h_inode->i_sb) && S_ISDIR(h_inode->i_mode) + && write_mask && !(mask & MAY_READ)) + || !h_inode->i_op->permission) { + /* AuLabel(generic_permission); */ + /* AuDbg("get_acl %pf\n", h_inode->i_op->get_acl); */ + err = generic_permission(h_inode, mask); + if (err == -EOPNOTSUPP && au_test_nfs_noacl(h_inode)) + err = h_inode->i_op->permission(h_inode, mask); + AuTraceErr(err); + } else { + /* AuLabel(h_inode->permission); */ + err = h_inode->i_op->permission(h_inode, mask); + AuTraceErr(err); + } + + if (!err) + err = devcgroup_inode_permission(h_inode, mask); + if (!err) + err = security_inode_permission(h_inode, mask); + +#if 0 + if (!err) { + /* todo: do we need to call ima_path_check()? */ + struct path h_path = { + .dentry = + .mnt = h_mnt + }; + err = ima_path_check(&h_path, + mask & (MAY_READ | MAY_WRITE | MAY_EXEC), + IMA_COUNT_LEAVE); + } +#endif + +out: + return err; +} + +static int aufs_permission(struct inode *inode, int mask) +{ + int err; + aufs_bindex_t bindex, bend; + const unsigned char isdir = !!S_ISDIR(inode->i_mode), + write_mask = !!(mask & (MAY_WRITE | MAY_APPEND)); + struct inode *h_inode; + struct super_block *sb; + struct au_branch *br; + + /* todo: support rcu-walk? */ + if (mask & MAY_NOT_BLOCK) + return -ECHILD; + + sb = inode->i_sb; + si_read_lock(sb, AuLock_FLUSH); + ii_read_lock_child(inode); +#if 0 + err = au_iigen_test(inode, au_sigen(sb)); + if (unlikely(err)) + goto out; +#endif + + if (!isdir + || write_mask + || au_opt_test(au_mntflags(sb), DIRPERM1)) { + err = au_busy_or_stale(); + h_inode = au_h_iptr(inode, au_ibstart(inode)); + if (unlikely(!h_inode + || (h_inode->i_mode & S_IFMT) + != (inode->i_mode & S_IFMT))) + goto out; + + err = 0; + bindex = au_ibstart(inode); + br = au_sbr(sb, bindex); + err = h_permission(h_inode, mask, au_br_mnt(br), br->br_perm); + if (write_mask + && !err + && !special_file(h_inode->i_mode)) { + /* test whether the upper writable branch exists */ + err = -EROFS; + for (; bindex >= 0; bindex--) + if (!au_br_rdonly(au_sbr(sb, bindex))) { + err = 0; + break; + } + } + goto out; + } + + /* non-write to dir */ + err = 0; + bend = au_ibend(inode); + for (bindex = au_ibstart(inode); !err && bindex <= bend; bindex++) { + h_inode = au_h_iptr(inode, bindex); + if (h_inode) { + err = au_busy_or_stale(); + if (unlikely(!S_ISDIR(h_inode->i_mode))) + break; + + br = au_sbr(sb, bindex); + err = h_permission(h_inode, mask, au_br_mnt(br), + br->br_perm); + } + } + +out: + ii_read_unlock(inode); + si_read_unlock(sb); + return err; +} + +/* ---------------------------------------------------------------------- */ + +static struct dentry *aufs_lookup(struct inode *dir, struct dentry *dentry, + unsigned int flags) +{ + struct dentry *ret, *parent; + struct inode *inode; + struct super_block *sb; + int err, npositive; + + IMustLock(dir); + + /* todo: support rcu-walk? */ + ret = ERR_PTR(-ECHILD); + if (flags & LOOKUP_RCU) + goto out; + + ret = ERR_PTR(-ENAMETOOLONG); + if (unlikely(dentry->d_name.len > AUFS_MAX_NAMELEN)) + goto out; + + sb = dir->i_sb; + err = si_read_lock(sb, AuLock_FLUSH | AuLock_NOPLM); + ret = ERR_PTR(err); + if (unlikely(err)) + goto out; + + err = au_di_init(dentry); + ret = ERR_PTR(err); + if (unlikely(err)) + goto out_si; + + inode = NULL; + npositive = 0; /* suppress a warning */ + parent = dentry->d_parent; /* dir inode is locked */ + di_read_lock_parent(parent, AuLock_IR); + err = au_alive_dir(parent); + if (!err) + err = au_digen_test(parent, au_sigen(sb)); + if (!err) { + npositive = au_lkup_dentry(dentry, au_dbstart(parent), + /*type*/0); + err = npositive; + } + di_read_unlock(parent, AuLock_IR); + ret = ERR_PTR(err); + if (unlikely(err < 0)) + goto out_unlock; + + if (npositive) { + inode = au_new_inode(dentry, /*must_new*/0); + if (IS_ERR(inode)) { + ret = (void *)inode; + inode = NULL; + goto out_unlock; + } + } + + if (inode) + atomic_inc(&inode->i_count); + ret = d_splice_alias(inode, dentry); +#if 0 + if (unlikely(d_need_lookup(dentry))) { + spin_lock(&dentry->d_lock); + dentry->d_flags &= ~DCACHE_NEED_LOOKUP; + spin_unlock(&dentry->d_lock); + } else +#endif + if (inode) { + if (!IS_ERR(ret)) { + iput(inode); + if (ret && ret != dentry) + ii_write_unlock(inode); + } else { + ii_write_unlock(inode); + iput(inode); + inode = NULL; + } + } + +out_unlock: + di_write_unlock(dentry); + if (inode) { + /* verbose coding for lock class name */ + if (unlikely(S_ISLNK(inode->i_mode))) + au_rw_class(&au_di(dentry)->di_rwsem, + au_lc_key + AuLcSymlink_DIINFO); + else if (unlikely(S_ISDIR(inode->i_mode))) + au_rw_class(&au_di(dentry)->di_rwsem, + au_lc_key + AuLcDir_DIINFO); + else /* likely */ + au_rw_class(&au_di(dentry)->di_rwsem, + au_lc_key + AuLcNonDir_DIINFO); + } +out_si: + si_read_unlock(sb); +out: + return ret; +} + +/* ---------------------------------------------------------------------- */ + +struct aopen_node { + struct hlist_node hlist; + struct file *file, *h_file; +}; + +static int au_do_aopen(struct inode *inode, struct file *file) +{ + struct au_sphlhead *aopen; + struct aopen_node *node; + struct au_do_open_args args = { + .no_lock = 1, + .open = au_do_open_nondir + }; + + aopen = &au_sbi(inode->i_sb)->si_aopen; + spin_lock(&aopen->spin); + hlist_for_each_entry(node, &aopen->head, hlist) + if (node->file == file) { + args.h_file = node->h_file; + break; + } + spin_unlock(&aopen->spin); + /* AuDebugOn(!args.h_file); */ + + return au_do_open(file, &args); +} + +static int aufs_atomic_open(struct inode *dir, struct dentry *dentry, + struct file *file, unsigned int open_flag, + umode_t create_mode, int *opened) +{ + int err, h_opened = *opened; + struct dentry *parent; + struct dentry *d; + struct au_sphlhead *aopen; + struct vfsub_aopen_args args = { + .open_flag = open_flag, + .create_mode = create_mode, + .opened = &h_opened + }; + struct aopen_node aopen_node = { + .file = file + }; + + IMustLock(dir); + AuDbg("open_flag 0x%x\n", open_flag); + AuDbgDentry(dentry); + + err = 0; + if (!au_di(dentry)) { + d = aufs_lookup(dir, dentry, /*flags*/0); + if (IS_ERR(d)) { + err = PTR_ERR(d); + goto out; + } else if (d) { + /* + * obsoleted dentry found. + * another error will be returned later. + */ + d_drop(d); + dput(d); + AuDbgDentry(d); + } + AuDbgDentry(dentry); + } + + if (d_is_positive(dentry) + || d_unhashed(dentry) + || d_unlinked(dentry) + || !(open_flag & O_CREAT)) + goto out_no_open; + + err = aufs_read_lock(dentry, AuLock_DW | AuLock_FLUSH | AuLock_GEN); + if (unlikely(err)) + goto out; + + parent = dentry->d_parent; /* dir is locked */ + di_write_lock_parent(parent); + err = au_lkup_dentry(dentry, /*bstart*/0, /*type*/0); + if (unlikely(err)) + goto out_unlock; + + AuDbgDentry(dentry); + if (d_is_positive(dentry)) + goto out_unlock; + + args.file = get_empty_filp(); + err = PTR_ERR(args.file); + if (IS_ERR(args.file)) + goto out_unlock; + + args.file->f_flags = file->f_flags; + err = au_aopen_or_create(dir, dentry, &args); + AuTraceErr(err); + AuDbgFile(args.file); + if (unlikely(err < 0)) { + if (h_opened & FILE_OPENED) + fput(args.file); + else + put_filp(args.file); + goto out_unlock; + } + + /* some filesystems don't set FILE_CREATED while succeeded? */ + *opened |= FILE_CREATED; + if (h_opened & FILE_OPENED) + aopen_node.h_file = args.file; + else { + put_filp(args.file); + args.file = NULL; + } + aopen = &au_sbi(dir->i_sb)->si_aopen; + au_sphl_add(&aopen_node.hlist, aopen); + err = finish_open(file, dentry, au_do_aopen, opened); + au_sphl_del(&aopen_node.hlist, aopen); + AuTraceErr(err); + AuDbgFile(file); + if (aopen_node.h_file) + fput(aopen_node.h_file); + +out_unlock: + di_write_unlock(parent); + aufs_read_unlock(dentry, AuLock_DW); + AuDbgDentry(dentry); + if (unlikely(err)) + goto out; +out_no_open: + if (!err && !(*opened & FILE_CREATED)) { + AuLabel(out_no_open); + dget(dentry); + err = finish_no_open(file, dentry); + } +out: + AuDbg("%pd%s%s\n", dentry, + (*opened & FILE_CREATED) ? " created" : "", + (*opened & FILE_OPENED) ? " opened" : ""); + AuTraceErr(err); + return err; +} + + +/* ---------------------------------------------------------------------- */ + +static int au_wr_dir_cpup(struct dentry *dentry, struct dentry *parent, + const unsigned char add_entry, aufs_bindex_t bcpup, + aufs_bindex_t bstart) +{ + int err; + struct dentry *h_parent; + struct inode *h_dir; + + if (add_entry) + IMustLock(d_inode(parent)); + else + di_write_lock_parent(parent); + + err = 0; + if (!au_h_dptr(parent, bcpup)) { + if (bstart > bcpup) + err = au_cpup_dirs(dentry, bcpup); + else if (bstart < bcpup) + err = au_cpdown_dirs(dentry, bcpup); + else + BUG(); + } + if (!err && add_entry && !au_ftest_wrdir(add_entry, TMPFILE)) { + h_parent = au_h_dptr(parent, bcpup); + h_dir = d_inode(h_parent); + inode_lock_nested(h_dir, AuLsc_I_PARENT); + err = au_lkup_neg(dentry, bcpup, /*wh*/0); + /* todo: no unlock here */ + inode_unlock(h_dir); + + AuDbg("bcpup %d\n", bcpup); + if (!err) { + if (d_really_is_negative(dentry)) + au_set_h_dptr(dentry, bstart, NULL); + au_update_dbrange(dentry, /*do_put_zero*/0); + } + } + + if (!add_entry) + di_write_unlock(parent); + if (!err) + err = bcpup; /* success */ + + AuTraceErr(err); + return err; +} + +/* + * decide the branch and the parent dir where we will create a new entry. + * returns new bindex or an error. + * copyup the parent dir if needed. + */ +int au_wr_dir(struct dentry *dentry, struct dentry *src_dentry, + struct au_wr_dir_args *args) +{ + int err; + unsigned int flags; + aufs_bindex_t bcpup, bstart, src_bstart; + const unsigned char add_entry + = au_ftest_wrdir(args->flags, ADD_ENTRY) + | au_ftest_wrdir(args->flags, TMPFILE); + struct super_block *sb; + struct dentry *parent; + struct au_sbinfo *sbinfo; + + sb = dentry->d_sb; + sbinfo = au_sbi(sb); + parent = dget_parent(dentry); + bstart = au_dbstart(dentry); + bcpup = bstart; + if (args->force_btgt < 0) { + if (src_dentry) { + src_bstart = au_dbstart(src_dentry); + if (src_bstart < bstart) + bcpup = src_bstart; + } else if (add_entry) { + flags = 0; + if (au_ftest_wrdir(args->flags, ISDIR)) + au_fset_wbr(flags, DIR); + err = AuWbrCreate(sbinfo, dentry, flags); + bcpup = err; + } + + if (bcpup < 0 || au_test_ro(sb, bcpup, d_inode(dentry))) { + if (add_entry) + err = AuWbrCopyup(sbinfo, dentry); + else { + if (!IS_ROOT(dentry)) { + di_read_lock_parent(parent, !AuLock_IR); + err = AuWbrCopyup(sbinfo, dentry); + di_read_unlock(parent, !AuLock_IR); + } else + err = AuWbrCopyup(sbinfo, dentry); + } + bcpup = err; + if (unlikely(err < 0)) + goto out; + } + } else { + bcpup = args->force_btgt; + AuDebugOn(au_test_ro(sb, bcpup, d_inode(dentry))); + } + + AuDbg("bstart %d, bcpup %d\n", bstart, bcpup); + err = bcpup; + if (bcpup == bstart) + goto out; /* success */ + + /* copyup the new parent into the branch we process */ + err = au_wr_dir_cpup(dentry, parent, add_entry, bcpup, bstart); + if (err >= 0) { + if (d_really_is_negative(dentry)) { + au_set_h_dptr(dentry, bstart, NULL); + au_set_dbstart(dentry, bcpup); + au_set_dbend(dentry, bcpup); + } + AuDebugOn(add_entry + && !au_ftest_wrdir(args->flags, TMPFILE) + && !au_h_dptr(dentry, bcpup)); + } + +out: + dput(parent); + return err; +} + +/* ---------------------------------------------------------------------- */ + +void au_pin_hdir_unlock(struct au_pin *p) +{ + if (p->hdir) + au_hn_imtx_unlock(p->hdir); +} + +int au_pin_hdir_lock(struct au_pin *p) +{ + int err; + + err = 0; + if (!p->hdir) + goto out; + + /* even if an error happens later, keep this lock */ + au_hn_imtx_lock_nested(p->hdir, p->lsc_hi); + + err = -EBUSY; + if (unlikely(p->hdir->hi_inode != d_inode(p->h_parent))) + goto out; + + err = 0; + if (p->h_dentry) + err = au_h_verify(p->h_dentry, p->udba, p->hdir->hi_inode, + p->h_parent, p->br); + +out: + return err; +} + +int au_pin_hdir_relock(struct au_pin *p) +{ + int err, i; + struct inode *h_i; + struct dentry *h_d[] = { + p->h_dentry, + p->h_parent + }; + + err = au_pin_hdir_lock(p); + if (unlikely(err)) + goto out; + + for (i = 0; !err && i < sizeof(h_d)/sizeof(*h_d); i++) { + if (!h_d[i]) + continue; + if (d_is_positive(h_d[i])) { + h_i = d_inode(h_d[i]); + err = !h_i->i_nlink; + } + } + +out: + return err; +} + +void au_pin_hdir_set_owner(struct au_pin *p, struct task_struct *task) +{ +#if defined(CONFIG_DEBUG_MUTEXES) || defined(CONFIG_SMP) + p->hdir->hi_inode->i_mutex.owner = task; +#endif +} + +void au_pin_hdir_acquire_nest(struct au_pin *p) +{ + if (p->hdir) { + mutex_acquire_nest(&p->hdir->hi_inode->i_mutex.dep_map, + p->lsc_hi, 0, NULL, _RET_IP_); + au_pin_hdir_set_owner(p, current); + } +} + +void au_pin_hdir_release(struct au_pin *p) +{ + if (p->hdir) { + au_pin_hdir_set_owner(p, p->task); + mutex_release(&p->hdir->hi_inode->i_mutex.dep_map, 1, _RET_IP_); + } +} + +struct dentry *au_pinned_h_parent(struct au_pin *pin) +{ + if (pin && pin->parent) + return au_h_dptr(pin->parent, pin->bindex); + return NULL; +} + +void au_unpin(struct au_pin *p) +{ + if (p->hdir) + au_pin_hdir_unlock(p); + if (p->h_mnt && au_ftest_pin(p->flags, MNT_WRITE)) + vfsub_mnt_drop_write(p->h_mnt); + if (!p->hdir) + return; + + if (!au_ftest_pin(p->flags, DI_LOCKED)) + di_read_unlock(p->parent, AuLock_IR); + iput(p->hdir->hi_inode); + dput(p->parent); + p->parent = NULL; + p->hdir = NULL; + p->h_mnt = NULL; + /* do not clear p->task */ +} + +int au_do_pin(struct au_pin *p) +{ + int err; + struct super_block *sb; + struct inode *h_dir; + + err = 0; + sb = p->dentry->d_sb; + p->br = au_sbr(sb, p->bindex); + if (IS_ROOT(p->dentry)) { + if (au_ftest_pin(p->flags, MNT_WRITE)) { + p->h_mnt = au_br_mnt(p->br); + err = vfsub_mnt_want_write(p->h_mnt); + if (unlikely(err)) { + au_fclr_pin(p->flags, MNT_WRITE); + goto out_err; + } + } + goto out; + } + + p->h_dentry = NULL; + if (p->bindex <= au_dbend(p->dentry)) + p->h_dentry = au_h_dptr(p->dentry, p->bindex); + + p->parent = dget_parent(p->dentry); + if (!au_ftest_pin(p->flags, DI_LOCKED)) + di_read_lock(p->parent, AuLock_IR, p->lsc_di); + + h_dir = NULL; + p->h_parent = au_h_dptr(p->parent, p->bindex); + p->hdir = au_hi(d_inode(p->parent), p->bindex); + if (p->hdir) + h_dir = p->hdir->hi_inode; + + /* + * udba case, or + * if DI_LOCKED is not set, then p->parent may be different + * and h_parent can be NULL. + */ + if (unlikely(!p->hdir || !h_dir || !p->h_parent)) { + err = -EBUSY; + if (!au_ftest_pin(p->flags, DI_LOCKED)) + di_read_unlock(p->parent, AuLock_IR); + dput(p->parent); + p->parent = NULL; + goto out_err; + } + + if (au_ftest_pin(p->flags, MNT_WRITE)) { + p->h_mnt = au_br_mnt(p->br); + err = vfsub_mnt_want_write(p->h_mnt); + if (unlikely(err)) { + au_fclr_pin(p->flags, MNT_WRITE); + if (!au_ftest_pin(p->flags, DI_LOCKED)) + di_read_unlock(p->parent, AuLock_IR); + dput(p->parent); + p->parent = NULL; + goto out_err; + } + } + + au_igrab(h_dir); + err = au_pin_hdir_lock(p); + if (!err) + goto out; /* success */ + + au_unpin(p); + +out_err: + pr_err("err %d\n", err); + err = au_busy_or_stale(); +out: + return err; +} + +void au_pin_init(struct au_pin *p, struct dentry *dentry, + aufs_bindex_t bindex, int lsc_di, int lsc_hi, + unsigned int udba, unsigned char flags) +{ + p->dentry = dentry; + p->udba = udba; + p->lsc_di = lsc_di; + p->lsc_hi = lsc_hi; + p->flags = flags; + p->bindex = bindex; + + p->parent = NULL; + p->hdir = NULL; + p->h_mnt = NULL; + + p->h_dentry = NULL; + p->h_parent = NULL; + p->br = NULL; + p->task = current; +} + +int au_pin(struct au_pin *pin, struct dentry *dentry, aufs_bindex_t bindex, + unsigned int udba, unsigned char flags) +{ + au_pin_init(pin, dentry, bindex, AuLsc_DI_PARENT, AuLsc_I_PARENT2, + udba, flags); + return au_do_pin(pin); +} + +/* ---------------------------------------------------------------------- */ + +/* + * ->setattr() and ->getattr() are called in various cases. + * chmod, stat: dentry is revalidated. + * fchmod, fstat: file and dentry are not revalidated, additionally they may be + * unhashed. + * for ->setattr(), ia->ia_file is passed from ftruncate only. + */ +/* todo: consolidate with do_refresh() and simple_reval_dpath() */ +int au_reval_for_attr(struct dentry *dentry, unsigned int sigen) +{ + int err; + struct dentry *parent; + + err = 0; + if (au_digen_test(dentry, sigen)) { + parent = dget_parent(dentry); + di_read_lock_parent(parent, AuLock_IR); + err = au_refresh_dentry(dentry, parent); + di_read_unlock(parent, AuLock_IR); + dput(parent); + } + + AuTraceErr(err); + return err; +} + +int au_pin_and_icpup(struct dentry *dentry, struct iattr *ia, + struct au_icpup_args *a) +{ + int err; + loff_t sz; + aufs_bindex_t bstart, ibstart; + struct dentry *hi_wh, *parent; + struct inode *inode; + struct au_wr_dir_args wr_dir_args = { + .force_btgt = -1, + .flags = 0 + }; + + if (d_is_dir(dentry)) + au_fset_wrdir(wr_dir_args.flags, ISDIR); + /* plink or hi_wh() case */ + bstart = au_dbstart(dentry); + inode = d_inode(dentry); + ibstart = au_ibstart(inode); + if (bstart != ibstart && !au_test_ro(inode->i_sb, ibstart, inode)) + wr_dir_args.force_btgt = ibstart; + err = au_wr_dir(dentry, /*src_dentry*/NULL, &wr_dir_args); + if (unlikely(err < 0)) + goto out; + a->btgt = err; + if (err != bstart) + au_fset_icpup(a->flags, DID_CPUP); + + err = 0; + a->pin_flags = AuPin_MNT_WRITE; + parent = NULL; + if (!IS_ROOT(dentry)) { + au_fset_pin(a->pin_flags, DI_LOCKED); + parent = dget_parent(dentry); + di_write_lock_parent(parent); + } + + err = au_pin(&a->pin, dentry, a->btgt, a->udba, a->pin_flags); + if (unlikely(err)) + goto out_parent; + + a->h_path.dentry = au_h_dptr(dentry, bstart); + sz = -1; + a->h_inode = d_inode(a->h_path.dentry); + if (ia && (ia->ia_valid & ATTR_SIZE)) { + inode_lock_nested(a->h_inode, AuLsc_I_CHILD); + if (ia->ia_size < i_size_read(a->h_inode)) + sz = ia->ia_size; + inode_unlock(a->h_inode); + } + + hi_wh = NULL; + if (au_ftest_icpup(a->flags, DID_CPUP) && d_unlinked(dentry)) { + hi_wh = au_hi_wh(inode, a->btgt); + if (!hi_wh) { + struct au_cp_generic cpg = { + .dentry = dentry, + .bdst = a->btgt, + .bsrc = -1, + .len = sz, + .pin = &a->pin + }; + err = au_sio_cpup_wh(&cpg, /*file*/NULL); + if (unlikely(err)) + goto out_unlock; + hi_wh = au_hi_wh(inode, a->btgt); + /* todo: revalidate hi_wh? */ + } + } + + if (parent) { + au_pin_set_parent_lflag(&a->pin, /*lflag*/0); + di_downgrade_lock(parent, AuLock_IR); + dput(parent); + parent = NULL; + } + if (!au_ftest_icpup(a->flags, DID_CPUP)) + goto out; /* success */ + + if (!d_unhashed(dentry)) { + struct au_cp_generic cpg = { + .dentry = dentry, + .bdst = a->btgt, + .bsrc = bstart, + .len = sz, + .pin = &a->pin, + .flags = AuCpup_DTIME | AuCpup_HOPEN + }; + err = au_sio_cpup_simple(&cpg); + if (!err) + a->h_path.dentry = au_h_dptr(dentry, a->btgt); + } else if (!hi_wh) + a->h_path.dentry = au_h_dptr(dentry, a->btgt); + else + a->h_path.dentry = hi_wh; /* do not dget here */ + +out_unlock: + a->h_inode = d_inode(a->h_path.dentry); + if (!err) + goto out; /* success */ + au_unpin(&a->pin); +out_parent: + if (parent) { + di_write_unlock(parent); + dput(parent); + } +out: + if (!err) + inode_lock_nested(a->h_inode, AuLsc_I_CHILD); + return err; +} + +static int aufs_setattr(struct dentry *dentry, struct iattr *ia) +{ + int err; + struct inode *inode, *delegated; + struct super_block *sb; + struct file *file; + struct au_icpup_args *a; + + inode = d_inode(dentry); + IMustLock(inode); + + err = -ENOMEM; + a = kzalloc(sizeof(*a), GFP_NOFS); + if (unlikely(!a)) + goto out; + + if (ia->ia_valid & (ATTR_KILL_SUID | ATTR_KILL_SGID)) + ia->ia_valid &= ~ATTR_MODE; + + file = NULL; + sb = dentry->d_sb; + err = si_read_lock(sb, AuLock_FLUSH | AuLock_NOPLM); + if (unlikely(err)) + goto out_kfree; + + if (ia->ia_valid & ATTR_FILE) { + /* currently ftruncate(2) only */ + AuDebugOn(!d_is_reg(dentry)); + file = ia->ia_file; + err = au_reval_and_lock_fdi(file, au_reopen_nondir, /*wlock*/1); + if (unlikely(err)) + goto out_si; + ia->ia_file = au_hf_top(file); + a->udba = AuOpt_UDBA_NONE; + } else { + /* fchmod() doesn't pass ia_file */ + a->udba = au_opt_udba(sb); + di_write_lock_child(dentry); + /* no d_unlinked(), to set UDBA_NONE for root */ + if (d_unhashed(dentry)) + a->udba = AuOpt_UDBA_NONE; + if (a->udba != AuOpt_UDBA_NONE) { + AuDebugOn(IS_ROOT(dentry)); + err = au_reval_for_attr(dentry, au_sigen(sb)); + if (unlikely(err)) + goto out_dentry; + } + } + + err = au_pin_and_icpup(dentry, ia, a); + if (unlikely(err < 0)) + goto out_dentry; + if (au_ftest_icpup(a->flags, DID_CPUP)) { + ia->ia_file = NULL; + ia->ia_valid &= ~ATTR_FILE; + } + + a->h_path.mnt = au_sbr_mnt(sb, a->btgt); + if ((ia->ia_valid & (ATTR_MODE | ATTR_CTIME)) + == (ATTR_MODE | ATTR_CTIME)) { + err = security_path_chmod(&a->h_path, ia->ia_mode); + if (unlikely(err)) + goto out_unlock; + } else if ((ia->ia_valid & (ATTR_UID | ATTR_GID)) + && (ia->ia_valid & ATTR_CTIME)) { + err = security_path_chown(&a->h_path, ia->ia_uid, ia->ia_gid); + if (unlikely(err)) + goto out_unlock; + } + + if (ia->ia_valid & ATTR_SIZE) { + struct file *f; + + if (ia->ia_size < i_size_read(inode)) + /* unmap only */ + truncate_setsize(inode, ia->ia_size); + + f = NULL; + if (ia->ia_valid & ATTR_FILE) + f = ia->ia_file; + inode_unlock(a->h_inode); + err = vfsub_trunc(&a->h_path, ia->ia_size, ia->ia_valid, f); + inode_lock_nested(a->h_inode, AuLsc_I_CHILD); + } else { + delegated = NULL; + while (1) { + err = vfsub_notify_change(&a->h_path, ia, &delegated); + if (delegated) { + err = break_deleg_wait(&delegated); + if (!err) + continue; + } + break; + } + } + /* + * regardless aufs 'acl' option setting. + * why don't all acl-aware fs call this func from their ->setattr()? + */ + if (!err && (ia->ia_valid & ATTR_MODE)) + err = vfsub_acl_chmod(a->h_inode, ia->ia_mode); + if (!err) + au_cpup_attr_changeable(inode); + +out_unlock: + inode_unlock(a->h_inode); + au_unpin(&a->pin); + if (unlikely(err)) + au_update_dbstart(dentry); +out_dentry: + di_write_unlock(dentry); + if (file) { + fi_write_unlock(file); + ia->ia_file = file; + ia->ia_valid |= ATTR_FILE; + } +out_si: + si_read_unlock(sb); +out_kfree: + kfree(a); +out: + AuTraceErr(err); + return err; +} + +#if IS_ENABLED(CONFIG_AUFS_XATTR) || IS_ENABLED(CONFIG_FS_POSIX_ACL) +static int au_h_path_to_set_attr(struct dentry *dentry, + struct au_icpup_args *a, struct path *h_path) +{ + int err; + struct super_block *sb; + + sb = dentry->d_sb; + a->udba = au_opt_udba(sb); + /* no d_unlinked(), to set UDBA_NONE for root */ + if (d_unhashed(dentry)) + a->udba = AuOpt_UDBA_NONE; + if (a->udba != AuOpt_UDBA_NONE) { + AuDebugOn(IS_ROOT(dentry)); + err = au_reval_for_attr(dentry, au_sigen(sb)); + if (unlikely(err)) + goto out; + } + err = au_pin_and_icpup(dentry, /*ia*/NULL, a); + if (unlikely(err < 0)) + goto out; + + h_path->dentry = a->h_path.dentry; + h_path->mnt = au_sbr_mnt(sb, a->btgt); + +out: + return err; +} + +ssize_t au_srxattr(struct dentry *dentry, struct au_srxattr *arg) +{ + int err; + struct path h_path; + struct super_block *sb; + struct au_icpup_args *a; + struct inode *inode, *h_inode; + + inode = d_inode(dentry); + IMustLock(inode); + + err = -ENOMEM; + a = kzalloc(sizeof(*a), GFP_NOFS); + if (unlikely(!a)) + goto out; + + sb = dentry->d_sb; + err = si_read_lock(sb, AuLock_FLUSH | AuLock_NOPLM); + if (unlikely(err)) + goto out_kfree; + + h_path.dentry = NULL; /* silence gcc */ + di_write_lock_child(dentry); + err = au_h_path_to_set_attr(dentry, a, &h_path); + if (unlikely(err)) + goto out_di; + + inode_unlock(a->h_inode); + switch (arg->type) { + case AU_XATTR_SET: + err = vfsub_setxattr(h_path.dentry, + arg->u.set.name, arg->u.set.value, + arg->u.set.size, arg->u.set.flags); + break; + case AU_XATTR_REMOVE: + err = vfsub_removexattr(h_path.dentry, arg->u.remove.name); + break; + case AU_ACL_SET: + err = -EOPNOTSUPP; + h_inode = d_inode(h_path.dentry); + if (h_inode->i_op->set_acl) + err = h_inode->i_op->set_acl(h_inode, + arg->u.acl_set.acl, + arg->u.acl_set.type); + break; + } + if (!err) + au_cpup_attr_timesizes(inode); + + au_unpin(&a->pin); + if (unlikely(err)) + au_update_dbstart(dentry); + +out_di: + di_write_unlock(dentry); + si_read_unlock(sb); +out_kfree: + kfree(a); +out: + AuTraceErr(err); + return err; +} +#endif + +static void au_refresh_iattr(struct inode *inode, struct kstat *st, + unsigned int nlink) +{ + unsigned int n; + + inode->i_mode = st->mode; + /* don't i_[ug]id_write() here */ + inode->i_uid = st->uid; + inode->i_gid = st->gid; + inode->i_atime = st->atime; + inode->i_mtime = st->mtime; + inode->i_ctime = st->ctime; + + au_cpup_attr_nlink(inode, /*force*/0); + if (S_ISDIR(inode->i_mode)) { + n = inode->i_nlink; + n -= nlink; + n += st->nlink; + smp_mb(); /* for i_nlink */ + /* 0 can happen */ + set_nlink(inode, n); + } + + spin_lock(&inode->i_lock); + inode->i_blocks = st->blocks; + i_size_write(inode, st->size); + spin_unlock(&inode->i_lock); +} + +/* + * common routine for aufs_getattr() and aufs_getxattr(). + * returns zero or negative (an error). + * @dentry will be read-locked in success. + */ +int au_h_path_getattr(struct dentry *dentry, int force, struct path *h_path) +{ + int err; + unsigned int mnt_flags, sigen; + unsigned char udba_none; + aufs_bindex_t bindex; + struct super_block *sb, *h_sb; + struct inode *inode; + + h_path->mnt = NULL; + h_path->dentry = NULL; + + err = 0; + sb = dentry->d_sb; + mnt_flags = au_mntflags(sb); + udba_none = !!au_opt_test(mnt_flags, UDBA_NONE); + + /* support fstat(2) */ + if (!d_unlinked(dentry) && !udba_none) { + sigen = au_sigen(sb); + err = au_digen_test(dentry, sigen); + if (!err) { + di_read_lock_child(dentry, AuLock_IR); + err = au_dbrange_test(dentry); + if (unlikely(err)) { + di_read_unlock(dentry, AuLock_IR); + goto out; + } + } else { + AuDebugOn(IS_ROOT(dentry)); + di_write_lock_child(dentry); + err = au_dbrange_test(dentry); + if (!err) + err = au_reval_for_attr(dentry, sigen); + if (!err) + di_downgrade_lock(dentry, AuLock_IR); + else { + di_write_unlock(dentry); + goto out; + } + } + } else + di_read_lock_child(dentry, AuLock_IR); + + inode = d_inode(dentry); + bindex = au_ibstart(inode); + h_path->mnt = au_sbr_mnt(sb, bindex); + h_sb = h_path->mnt->mnt_sb; + if (!force + && !au_test_fs_bad_iattr(h_sb) + && udba_none) + goto out; /* success */ + + if (au_dbstart(dentry) == bindex) + h_path->dentry = au_h_dptr(dentry, bindex); + else if (au_opt_test(mnt_flags, PLINK) && au_plink_test(inode)) { + h_path->dentry = au_plink_lkup(inode, bindex); + if (IS_ERR(h_path->dentry)) + /* pretending success */ + h_path->dentry = NULL; + else + dput(h_path->dentry); + } + +out: + return err; +} + +static int aufs_getattr(struct vfsmount *mnt __maybe_unused, + struct dentry *dentry, struct kstat *st) +{ + int err; + unsigned char positive; + struct path h_path; + struct inode *inode; + struct super_block *sb; + + inode = d_inode(dentry); + sb = dentry->d_sb; + err = si_read_lock(sb, AuLock_FLUSH | AuLock_NOPLM); + if (unlikely(err)) + goto out; + err = au_h_path_getattr(dentry, /*force*/0, &h_path); + if (unlikely(err)) + goto out_si; + if (unlikely(!h_path.dentry)) + /* illegally overlapped or something */ + goto out_fill; /* pretending success */ + + positive = d_is_positive(h_path.dentry); + if (positive) + err = vfs_getattr(&h_path, st); + if (!err) { + if (positive) + au_refresh_iattr(inode, st, + d_inode(h_path.dentry)->i_nlink); + goto out_fill; /* success */ + } + AuTraceErr(err); + goto out_di; + +out_fill: + generic_fillattr(inode, st); +out_di: + di_read_unlock(dentry, AuLock_IR); +out_si: + si_read_unlock(sb); +out: + AuTraceErr(err); + return err; +} + +/* ---------------------------------------------------------------------- */ + +static const char *aufs_get_link(struct dentry *dentry, struct inode *inode, + struct delayed_call *done) +{ + const char *ret; + struct dentry *h_dentry; + struct inode *h_inode; + int err; + aufs_bindex_t bindex; + + ret = NULL; /* suppress a warning */ + err = -ECHILD; + if (!dentry) + goto out; + + err = aufs_read_lock(dentry, AuLock_IR | AuLock_GEN); + if (unlikely(err)) + goto out; + + err = au_d_hashed_positive(dentry); + if (unlikely(err)) + goto out_unlock; + + err = -EINVAL; + inode = d_inode(dentry); + bindex = au_ibstart(inode); + h_inode = au_h_iptr(inode, bindex); + if (unlikely(!h_inode->i_op->get_link)) + goto out_unlock; + + err = -EBUSY; + h_dentry = NULL; + if (au_dbstart(dentry) <= bindex) { + h_dentry = au_h_dptr(dentry, bindex); + if (h_dentry) + dget(h_dentry); + } + if (!h_dentry) { + h_dentry = d_find_any_alias(h_inode); + if (IS_ERR(h_dentry)) { + err = PTR_ERR(h_dentry); + goto out_unlock; + } + } + if (unlikely(!h_dentry)) + goto out_unlock; + + err = 0; + AuDbg("%pf\n", h_inode->i_op->get_link); + AuDbgDentry(h_dentry); + ret = h_inode->i_op->get_link(h_dentry, h_inode, done); + dput(h_dentry); + if (IS_ERR(ret)) + err = PTR_ERR(ret); + +out_unlock: + aufs_read_unlock(dentry, AuLock_IR); +out: + if (unlikely(err)) + ret = ERR_PTR(err); + AuTraceErrPtr(ret); + return ret; +} + +/* ---------------------------------------------------------------------- */ + +static int aufs_update_time(struct inode *inode, struct timespec *ts, int flags) +{ + int err; + struct super_block *sb; + struct inode *h_inode; + + sb = inode->i_sb; + /* mmap_sem might be acquired already, cf. aufs_mmap() */ + lockdep_off(); + si_read_lock(sb, AuLock_FLUSH); + ii_write_lock_child(inode); + lockdep_on(); + h_inode = au_h_iptr(inode, au_ibstart(inode)); + err = vfsub_update_time(h_inode, ts, flags); + lockdep_off(); + if (!err) + au_cpup_attr_timesizes(inode); + ii_write_unlock(inode); + si_read_unlock(sb); + lockdep_on(); + + if (!err && (flags & S_VERSION)) + inode_inc_iversion(inode); + + return err; +} + +/* ---------------------------------------------------------------------- */ + +/* no getattr version will be set by module.c:aufs_init() */ +struct inode_operations aufs_iop_nogetattr[AuIop_Last], + aufs_iop[] = { + [AuIop_SYMLINK] = { + .permission = aufs_permission, +#ifdef CONFIG_FS_POSIX_ACL + .get_acl = aufs_get_acl, + .set_acl = aufs_set_acl, /* unsupport for symlink? */ +#endif + + .setattr = aufs_setattr, + .getattr = aufs_getattr, + +#ifdef CONFIG_AUFS_XATTR + .setxattr = aufs_setxattr, + .getxattr = aufs_getxattr, + .listxattr = aufs_listxattr, + .removexattr = aufs_removexattr, +#endif + + .readlink = generic_readlink, + .get_link = aufs_get_link, + + /* .update_time = aufs_update_time */ + }, + [AuIop_DIR] = { + .create = aufs_create, + .lookup = aufs_lookup, + .link = aufs_link, + .unlink = aufs_unlink, + .symlink = aufs_symlink, + .mkdir = aufs_mkdir, + .rmdir = aufs_rmdir, + .mknod = aufs_mknod, + .rename = aufs_rename, + + .permission = aufs_permission, +#ifdef CONFIG_FS_POSIX_ACL + .get_acl = aufs_get_acl, + .set_acl = aufs_set_acl, +#endif + + .setattr = aufs_setattr, + .getattr = aufs_getattr, + +#ifdef CONFIG_AUFS_XATTR + .setxattr = aufs_setxattr, + .getxattr = aufs_getxattr, + .listxattr = aufs_listxattr, + .removexattr = aufs_removexattr, +#endif + + .update_time = aufs_update_time, + .atomic_open = aufs_atomic_open, + .tmpfile = aufs_tmpfile + }, + [AuIop_OTHER] = { + .permission = aufs_permission, +#ifdef CONFIG_FS_POSIX_ACL + .get_acl = aufs_get_acl, + .set_acl = aufs_set_acl, +#endif + + .setattr = aufs_setattr, + .getattr = aufs_getattr, + +#ifdef CONFIG_AUFS_XATTR + .setxattr = aufs_setxattr, + .getxattr = aufs_getxattr, + .listxattr = aufs_listxattr, + .removexattr = aufs_removexattr, +#endif + + .update_time = aufs_update_time + } +}; diff --git a/fs/aufs/i_op_add.c b/fs/aufs/i_op_add.c new file mode 100644 index 000000000..533b36a2a --- /dev/null +++ b/fs/aufs/i_op_add.c @@ -0,0 +1,919 @@ +/* + * Copyright (C) 2005-2016 Junjiro R. Okajima + */ + +/* + * inode operations (add entry) + */ + +#include "aufs.h" + +/* + * final procedure of adding a new entry, except link(2). + * remove whiteout, instantiate, copyup the parent dir's times and size + * and update version. + * if it failed, re-create the removed whiteout. + */ +static int epilog(struct inode *dir, aufs_bindex_t bindex, + struct dentry *wh_dentry, struct dentry *dentry) +{ + int err, rerr; + aufs_bindex_t bwh; + struct path h_path; + struct super_block *sb; + struct inode *inode, *h_dir; + struct dentry *wh; + + bwh = -1; + sb = dir->i_sb; + if (wh_dentry) { + h_dir = d_inode(wh_dentry->d_parent); /* dir inode is locked */ + IMustLock(h_dir); + AuDebugOn(au_h_iptr(dir, bindex) != h_dir); + bwh = au_dbwh(dentry); + h_path.dentry = wh_dentry; + h_path.mnt = au_sbr_mnt(sb, bindex); + err = au_wh_unlink_dentry(au_h_iptr(dir, bindex), &h_path, + dentry); + if (unlikely(err)) + goto out; + } + + inode = au_new_inode(dentry, /*must_new*/1); + if (!IS_ERR(inode)) { + d_instantiate(dentry, inode); + dir = d_inode(dentry->d_parent); /* dir inode is locked */ + IMustLock(dir); + au_dir_ts(dir, bindex); + dir->i_version++; + au_fhsm_wrote(sb, bindex, /*force*/0); + return 0; /* success */ + } + + err = PTR_ERR(inode); + if (!wh_dentry) + goto out; + + /* revert */ + /* dir inode is locked */ + wh = au_wh_create(dentry, bwh, wh_dentry->d_parent); + rerr = PTR_ERR(wh); + if (IS_ERR(wh)) { + AuIOErr("%pd reverting whiteout failed(%d, %d)\n", + dentry, err, rerr); + err = -EIO; + } else + dput(wh); + +out: + return err; +} + +static int au_d_may_add(struct dentry *dentry) +{ + int err; + + err = 0; + if (unlikely(d_unhashed(dentry))) + err = -ENOENT; + if (unlikely(d_really_is_positive(dentry))) + err = -EEXIST; + return err; +} + +/* + * simple tests for the adding inode operations. + * following the checks in vfs, plus the parent-child relationship. + */ +int au_may_add(struct dentry *dentry, aufs_bindex_t bindex, + struct dentry *h_parent, int isdir) +{ + int err; + umode_t h_mode; + struct dentry *h_dentry; + struct inode *h_inode; + + err = -ENAMETOOLONG; + if (unlikely(dentry->d_name.len > AUFS_MAX_NAMELEN)) + goto out; + + h_dentry = au_h_dptr(dentry, bindex); + if (d_really_is_negative(dentry)) { + err = -EEXIST; + if (unlikely(d_is_positive(h_dentry))) + goto out; + } else { + /* rename(2) case */ + err = -EIO; + if (unlikely(d_is_negative(h_dentry))) + goto out; + h_inode = d_inode(h_dentry); + if (unlikely(!h_inode->i_nlink)) + goto out; + + h_mode = h_inode->i_mode; + if (!isdir) { + err = -EISDIR; + if (unlikely(S_ISDIR(h_mode))) + goto out; + } else if (unlikely(!S_ISDIR(h_mode))) { + err = -ENOTDIR; + goto out; + } + } + + err = 0; + /* expected parent dir is locked */ + if (unlikely(h_parent != h_dentry->d_parent)) + err = -EIO; + +out: + AuTraceErr(err); + return err; +} + +/* + * initial procedure of adding a new entry. + * prepare writable branch and the parent dir, lock it, + * and lookup whiteout for the new entry. + */ +static struct dentry* +lock_hdir_lkup_wh(struct dentry *dentry, struct au_dtime *dt, + struct dentry *src_dentry, struct au_pin *pin, + struct au_wr_dir_args *wr_dir_args) +{ + struct dentry *wh_dentry, *h_parent; + struct super_block *sb; + struct au_branch *br; + int err; + unsigned int udba; + aufs_bindex_t bcpup; + + AuDbg("%pd\n", dentry); + + err = au_wr_dir(dentry, src_dentry, wr_dir_args); + bcpup = err; + wh_dentry = ERR_PTR(err); + if (unlikely(err < 0)) + goto out; + + sb = dentry->d_sb; + udba = au_opt_udba(sb); + err = au_pin(pin, dentry, bcpup, udba, + AuPin_DI_LOCKED | AuPin_MNT_WRITE); + wh_dentry = ERR_PTR(err); + if (unlikely(err)) + goto out; + + h_parent = au_pinned_h_parent(pin); + if (udba != AuOpt_UDBA_NONE + && au_dbstart(dentry) == bcpup) + err = au_may_add(dentry, bcpup, h_parent, + au_ftest_wrdir(wr_dir_args->flags, ISDIR)); + else if (unlikely(dentry->d_name.len > AUFS_MAX_NAMELEN)) + err = -ENAMETOOLONG; + wh_dentry = ERR_PTR(err); + if (unlikely(err)) + goto out_unpin; + + br = au_sbr(sb, bcpup); + if (dt) { + struct path tmp = { + .dentry = h_parent, + .mnt = au_br_mnt(br) + }; + au_dtime_store(dt, au_pinned_parent(pin), &tmp); + } + + wh_dentry = NULL; + if (bcpup != au_dbwh(dentry)) + goto out; /* success */ + + /* + * ENAMETOOLONG here means that if we allowed create such name, then it + * would not be able to removed in the future. So we don't allow such + * name here and we don't handle ENAMETOOLONG differently here. + */ + wh_dentry = au_wh_lkup(h_parent, &dentry->d_name, br); + +out_unpin: + if (IS_ERR(wh_dentry)) + au_unpin(pin); +out: + return wh_dentry; +} + +/* ---------------------------------------------------------------------- */ + +enum { Mknod, Symlink, Creat }; +struct simple_arg { + int type; + union { + struct { + umode_t mode; + bool want_excl; + bool try_aopen; + struct vfsub_aopen_args *aopen; + } c; + struct { + const char *symname; + } s; + struct { + umode_t mode; + dev_t dev; + } m; + } u; +}; + +static int add_simple(struct inode *dir, struct dentry *dentry, + struct simple_arg *arg) +{ + int err, rerr; + aufs_bindex_t bstart; + unsigned char created; + const unsigned char try_aopen + = (arg->type == Creat && arg->u.c.try_aopen); + struct dentry *wh_dentry, *parent; + struct inode *h_dir; + struct super_block *sb; + struct au_branch *br; + /* to reuduce stack size */ + struct { + struct au_dtime dt; + struct au_pin pin; + struct path h_path; + struct au_wr_dir_args wr_dir_args; + } *a; + + AuDbg("%pd\n", dentry); + IMustLock(dir); + + err = -ENOMEM; + a = kmalloc(sizeof(*a), GFP_NOFS); + if (unlikely(!a)) + goto out; + a->wr_dir_args.force_btgt = -1; + a->wr_dir_args.flags = AuWrDir_ADD_ENTRY; + + parent = dentry->d_parent; /* dir inode is locked */ + if (!try_aopen) { + err = aufs_read_lock(dentry, AuLock_DW | AuLock_GEN); + if (unlikely(err)) + goto out_free; + } + err = au_d_may_add(dentry); + if (unlikely(err)) + goto out_unlock; + if (!try_aopen) + di_write_lock_parent(parent); + wh_dentry = lock_hdir_lkup_wh(dentry, &a->dt, /*src_dentry*/NULL, + &a->pin, &a->wr_dir_args); + err = PTR_ERR(wh_dentry); + if (IS_ERR(wh_dentry)) + goto out_parent; + + bstart = au_dbstart(dentry); + sb = dentry->d_sb; + br = au_sbr(sb, bstart); + a->h_path.dentry = au_h_dptr(dentry, bstart); + a->h_path.mnt = au_br_mnt(br); + h_dir = au_pinned_h_dir(&a->pin); + switch (arg->type) { + case Creat: + err = 0; + if (!try_aopen || !h_dir->i_op->atomic_open) + err = vfsub_create(h_dir, &a->h_path, arg->u.c.mode, + arg->u.c.want_excl); + else + err = vfsub_atomic_open(h_dir, a->h_path.dentry, + arg->u.c.aopen, br); + break; + case Symlink: + err = vfsub_symlink(h_dir, &a->h_path, arg->u.s.symname); + break; + case Mknod: + err = vfsub_mknod(h_dir, &a->h_path, arg->u.m.mode, + arg->u.m.dev); + break; + default: + BUG(); + } + created = !err; + if (!err) + err = epilog(dir, bstart, wh_dentry, dentry); + + /* revert */ + if (unlikely(created && err && d_is_positive(a->h_path.dentry))) { + /* no delegation since it is just created */ + rerr = vfsub_unlink(h_dir, &a->h_path, /*delegated*/NULL, + /*force*/0); + if (rerr) { + AuIOErr("%pd revert failure(%d, %d)\n", + dentry, err, rerr); + err = -EIO; + } + au_dtime_revert(&a->dt); + } + + if (!err && try_aopen && !h_dir->i_op->atomic_open) + *arg->u.c.aopen->opened |= FILE_CREATED; + + au_unpin(&a->pin); + dput(wh_dentry); + +out_parent: + if (!try_aopen) + di_write_unlock(parent); +out_unlock: + if (unlikely(err)) { + au_update_dbstart(dentry); + d_drop(dentry); + } + if (!try_aopen) + aufs_read_unlock(dentry, AuLock_DW); +out_free: + kfree(a); +out: + return err; +} + +int aufs_mknod(struct inode *dir, struct dentry *dentry, umode_t mode, + dev_t dev) +{ + struct simple_arg arg = { + .type = Mknod, + .u.m = { + .mode = mode, + .dev = dev + } + }; + return add_simple(dir, dentry, &arg); +} + +int aufs_symlink(struct inode *dir, struct dentry *dentry, const char *symname) +{ + struct simple_arg arg = { + .type = Symlink, + .u.s.symname = symname + }; + return add_simple(dir, dentry, &arg); +} + +int aufs_create(struct inode *dir, struct dentry *dentry, umode_t mode, + bool want_excl) +{ + struct simple_arg arg = { + .type = Creat, + .u.c = { + .mode = mode, + .want_excl = want_excl + } + }; + return add_simple(dir, dentry, &arg); +} + +int au_aopen_or_create(struct inode *dir, struct dentry *dentry, + struct vfsub_aopen_args *aopen_args) +{ + struct simple_arg arg = { + .type = Creat, + .u.c = { + .mode = aopen_args->create_mode, + .want_excl = aopen_args->open_flag & O_EXCL, + .try_aopen = true, + .aopen = aopen_args + } + }; + return add_simple(dir, dentry, &arg); +} + +int aufs_tmpfile(struct inode *dir, struct dentry *dentry, umode_t mode) +{ + int err; + aufs_bindex_t bindex; + struct super_block *sb; + struct dentry *parent, *h_parent, *h_dentry; + struct inode *h_dir, *inode; + struct vfsmount *h_mnt; + struct au_wr_dir_args wr_dir_args = { + .force_btgt = -1, + .flags = AuWrDir_TMPFILE + }; + + /* copy-up may happen */ + inode_lock(dir); + + sb = dir->i_sb; + err = si_read_lock(sb, AuLock_FLUSH | AuLock_NOPLM); + if (unlikely(err)) + goto out; + + err = au_di_init(dentry); + if (unlikely(err)) + goto out_si; + + err = -EBUSY; + parent = d_find_any_alias(dir); + AuDebugOn(!parent); + di_write_lock_parent(parent); + if (unlikely(d_inode(parent) != dir)) + goto out_parent; + + err = au_digen_test(parent, au_sigen(sb)); + if (unlikely(err)) + goto out_parent; + + bindex = au_dbstart(parent); + au_set_dbstart(dentry, bindex); + au_set_dbend(dentry, bindex); + err = au_wr_dir(dentry, /*src_dentry*/NULL, &wr_dir_args); + bindex = err; + if (unlikely(err < 0)) + goto out_parent; + + err = -EOPNOTSUPP; + h_dir = au_h_iptr(dir, bindex); + if (unlikely(!h_dir->i_op->tmpfile)) + goto out_parent; + + h_mnt = au_sbr_mnt(sb, bindex); + err = vfsub_mnt_want_write(h_mnt); + if (unlikely(err)) + goto out_parent; + + h_parent = au_h_dptr(parent, bindex); + err = inode_permission(d_inode(h_parent), MAY_WRITE | MAY_EXEC); + if (unlikely(err)) + goto out_mnt; + + err = -ENOMEM; + h_dentry = d_alloc(h_parent, &dentry->d_name); + if (unlikely(!h_dentry)) + goto out_mnt; + + err = h_dir->i_op->tmpfile(h_dir, h_dentry, mode); + if (unlikely(err)) + goto out_dentry; + + au_set_dbstart(dentry, bindex); + au_set_dbend(dentry, bindex); + au_set_h_dptr(dentry, bindex, dget(h_dentry)); + inode = au_new_inode(dentry, /*must_new*/1); + if (IS_ERR(inode)) { + err = PTR_ERR(inode); + au_set_h_dptr(dentry, bindex, NULL); + au_set_dbstart(dentry, -1); + au_set_dbend(dentry, -1); + } else { + if (!inode->i_nlink) + set_nlink(inode, 1); + d_tmpfile(dentry, inode); + au_di(dentry)->di_tmpfile = 1; + + /* update without i_mutex */ + if (au_ibstart(dir) == au_dbstart(dentry)) + au_cpup_attr_timesizes(dir); + } + +out_dentry: + dput(h_dentry); +out_mnt: + vfsub_mnt_drop_write(h_mnt); +out_parent: + di_write_unlock(parent); + dput(parent); + di_write_unlock(dentry); + if (!err) +#if 0 + /* verbose coding for lock class name */ + au_rw_class(&au_di(dentry)->di_rwsem, + au_lc_key + AuLcNonDir_DIINFO); +#else + ; +#endif + else { + au_di_fin(dentry); + dentry->d_fsdata = NULL; + } +out_si: + si_read_unlock(sb); +out: + inode_unlock(dir); + return err; +} + +/* ---------------------------------------------------------------------- */ + +struct au_link_args { + aufs_bindex_t bdst, bsrc; + struct au_pin pin; + struct path h_path; + struct dentry *src_parent, *parent; +}; + +static int au_cpup_before_link(struct dentry *src_dentry, + struct au_link_args *a) +{ + int err; + struct dentry *h_src_dentry; + struct au_cp_generic cpg = { + .dentry = src_dentry, + .bdst = a->bdst, + .bsrc = a->bsrc, + .len = -1, + .pin = &a->pin, + .flags = AuCpup_DTIME | AuCpup_HOPEN /* | AuCpup_KEEPLINO */ + }; + + di_read_lock_parent(a->src_parent, AuLock_IR); + err = au_test_and_cpup_dirs(src_dentry, a->bdst); + if (unlikely(err)) + goto out; + + h_src_dentry = au_h_dptr(src_dentry, a->bsrc); + err = au_pin(&a->pin, src_dentry, a->bdst, + au_opt_udba(src_dentry->d_sb), + AuPin_DI_LOCKED | AuPin_MNT_WRITE); + if (unlikely(err)) + goto out; + + err = au_sio_cpup_simple(&cpg); + au_unpin(&a->pin); + +out: + di_read_unlock(a->src_parent, AuLock_IR); + return err; +} + +static int au_cpup_or_link(struct dentry *src_dentry, struct dentry *dentry, + struct au_link_args *a) +{ + int err; + unsigned char plink; + aufs_bindex_t bend; + struct dentry *h_src_dentry; + struct inode *h_inode, *inode, *delegated; + struct super_block *sb; + struct file *h_file; + + plink = 0; + h_inode = NULL; + sb = src_dentry->d_sb; + inode = d_inode(src_dentry); + if (au_ibstart(inode) <= a->bdst) + h_inode = au_h_iptr(inode, a->bdst); + if (!h_inode || !h_inode->i_nlink) { + /* copyup src_dentry as the name of dentry. */ + bend = au_dbend(dentry); + if (bend < a->bsrc) + au_set_dbend(dentry, a->bsrc); + au_set_h_dptr(dentry, a->bsrc, + dget(au_h_dptr(src_dentry, a->bsrc))); + dget(a->h_path.dentry); + au_set_h_dptr(dentry, a->bdst, NULL); + AuDbg("temporary d_inode...\n"); + spin_lock(&dentry->d_lock); + dentry->d_inode = d_inode(src_dentry); /* tmp */ + spin_unlock(&dentry->d_lock); + h_file = au_h_open_pre(dentry, a->bsrc, /*force_wr*/0); + if (IS_ERR(h_file)) + err = PTR_ERR(h_file); + else { + struct au_cp_generic cpg = { + .dentry = dentry, + .bdst = a->bdst, + .bsrc = -1, + .len = -1, + .pin = &a->pin, + .flags = AuCpup_KEEPLINO + }; + err = au_sio_cpup_simple(&cpg); + au_h_open_post(dentry, a->bsrc, h_file); + if (!err) { + dput(a->h_path.dentry); + a->h_path.dentry = au_h_dptr(dentry, a->bdst); + } else + au_set_h_dptr(dentry, a->bdst, + a->h_path.dentry); + } + spin_lock(&dentry->d_lock); + dentry->d_inode = NULL; /* restore */ + spin_unlock(&dentry->d_lock); + AuDbg("temporary d_inode...done\n"); + au_set_h_dptr(dentry, a->bsrc, NULL); + au_set_dbend(dentry, bend); + } else { + /* the inode of src_dentry already exists on a.bdst branch */ + h_src_dentry = d_find_alias(h_inode); + if (!h_src_dentry && au_plink_test(inode)) { + plink = 1; + h_src_dentry = au_plink_lkup(inode, a->bdst); + err = PTR_ERR(h_src_dentry); + if (IS_ERR(h_src_dentry)) + goto out; + + if (unlikely(d_is_negative(h_src_dentry))) { + dput(h_src_dentry); + h_src_dentry = NULL; + } + + } + if (h_src_dentry) { + delegated = NULL; + err = vfsub_link(h_src_dentry, au_pinned_h_dir(&a->pin), + &a->h_path, &delegated); + if (unlikely(err == -EWOULDBLOCK)) { + pr_warn("cannot retry for NFSv4 delegation" + " for an internal link\n"); + iput(delegated); + } + dput(h_src_dentry); + } else { + AuIOErr("no dentry found for hi%lu on b%d\n", + h_inode->i_ino, a->bdst); + err = -EIO; + } + } + + if (!err && !plink) + au_plink_append(inode, a->bdst, a->h_path.dentry); + +out: + AuTraceErr(err); + return err; +} + +int aufs_link(struct dentry *src_dentry, struct inode *dir, + struct dentry *dentry) +{ + int err, rerr; + struct au_dtime dt; + struct au_link_args *a; + struct dentry *wh_dentry, *h_src_dentry; + struct inode *inode, *delegated; + struct super_block *sb; + struct au_wr_dir_args wr_dir_args = { + /* .force_btgt = -1, */ + .flags = AuWrDir_ADD_ENTRY + }; + + IMustLock(dir); + inode = d_inode(src_dentry); + IMustLock(inode); + + err = -ENOMEM; + a = kzalloc(sizeof(*a), GFP_NOFS); + if (unlikely(!a)) + goto out; + + a->parent = dentry->d_parent; /* dir inode is locked */ + err = aufs_read_and_write_lock2(dentry, src_dentry, + AuLock_NOPLM | AuLock_GEN); + if (unlikely(err)) + goto out_kfree; + err = au_d_linkable(src_dentry); + if (unlikely(err)) + goto out_unlock; + err = au_d_may_add(dentry); + if (unlikely(err)) + goto out_unlock; + + a->src_parent = dget_parent(src_dentry); + wr_dir_args.force_btgt = au_ibstart(inode); + + di_write_lock_parent(a->parent); + wr_dir_args.force_btgt = au_wbr(dentry, wr_dir_args.force_btgt); + wh_dentry = lock_hdir_lkup_wh(dentry, &dt, src_dentry, &a->pin, + &wr_dir_args); + err = PTR_ERR(wh_dentry); + if (IS_ERR(wh_dentry)) + goto out_parent; + + err = 0; + sb = dentry->d_sb; + a->bdst = au_dbstart(dentry); + a->h_path.dentry = au_h_dptr(dentry, a->bdst); + a->h_path.mnt = au_sbr_mnt(sb, a->bdst); + a->bsrc = au_ibstart(inode); + h_src_dentry = au_h_d_alias(src_dentry, a->bsrc); + if (!h_src_dentry && au_di(src_dentry)->di_tmpfile) + h_src_dentry = dget(au_hi_wh(inode, a->bsrc)); + if (!h_src_dentry) { + a->bsrc = au_dbstart(src_dentry); + h_src_dentry = au_h_d_alias(src_dentry, a->bsrc); + AuDebugOn(!h_src_dentry); + } else if (IS_ERR(h_src_dentry)) { + err = PTR_ERR(h_src_dentry); + goto out_parent; + } + + if (au_opt_test(au_mntflags(sb), PLINK)) { + if (a->bdst < a->bsrc + /* && h_src_dentry->d_sb != a->h_path.dentry->d_sb */) + err = au_cpup_or_link(src_dentry, dentry, a); + else { + delegated = NULL; + err = vfsub_link(h_src_dentry, au_pinned_h_dir(&a->pin), + &a->h_path, &delegated); + if (unlikely(err == -EWOULDBLOCK)) { + pr_warn("cannot retry for NFSv4 delegation" + " for an internal link\n"); + iput(delegated); + } + } + dput(h_src_dentry); + } else { + /* + * copyup src_dentry to the branch we process, + * and then link(2) to it. + */ + dput(h_src_dentry); + if (a->bdst < a->bsrc + /* && h_src_dentry->d_sb != a->h_path.dentry->d_sb */) { + au_unpin(&a->pin); + di_write_unlock(a->parent); + err = au_cpup_before_link(src_dentry, a); + di_write_lock_parent(a->parent); + if (!err) + err = au_pin(&a->pin, dentry, a->bdst, + au_opt_udba(sb), + AuPin_DI_LOCKED | AuPin_MNT_WRITE); + if (unlikely(err)) + goto out_wh; + } + if (!err) { + h_src_dentry = au_h_dptr(src_dentry, a->bdst); + err = -ENOENT; + if (h_src_dentry && d_is_positive(h_src_dentry)) { + delegated = NULL; + err = vfsub_link(h_src_dentry, + au_pinned_h_dir(&a->pin), + &a->h_path, &delegated); + if (unlikely(err == -EWOULDBLOCK)) { + pr_warn("cannot retry" + " for NFSv4 delegation" + " for an internal link\n"); + iput(delegated); + } + } + } + } + if (unlikely(err)) + goto out_unpin; + + if (wh_dentry) { + a->h_path.dentry = wh_dentry; + err = au_wh_unlink_dentry(au_pinned_h_dir(&a->pin), &a->h_path, + dentry); + if (unlikely(err)) + goto out_revert; + } + + au_dir_ts(dir, a->bdst); + dir->i_version++; + inc_nlink(inode); + inode->i_ctime = dir->i_ctime; + d_instantiate(dentry, au_igrab(inode)); + if (d_unhashed(a->h_path.dentry)) + /* some filesystem calls d_drop() */ + d_drop(dentry); + /* some filesystems consume an inode even hardlink */ + au_fhsm_wrote(sb, a->bdst, /*force*/0); + goto out_unpin; /* success */ + +out_revert: + /* no delegation since it is just created */ + rerr = vfsub_unlink(au_pinned_h_dir(&a->pin), &a->h_path, + /*delegated*/NULL, /*force*/0); + if (unlikely(rerr)) { + AuIOErr("%pd reverting failed(%d, %d)\n", dentry, err, rerr); + err = -EIO; + } + au_dtime_revert(&dt); +out_unpin: + au_unpin(&a->pin); +out_wh: + dput(wh_dentry); +out_parent: + di_write_unlock(a->parent); + dput(a->src_parent); +out_unlock: + if (unlikely(err)) { + au_update_dbstart(dentry); + d_drop(dentry); + } + aufs_read_and_write_unlock2(dentry, src_dentry); +out_kfree: + kfree(a); +out: + AuTraceErr(err); + return err; +} + +int aufs_mkdir(struct inode *dir, struct dentry *dentry, umode_t mode) +{ + int err, rerr; + aufs_bindex_t bindex; + unsigned char diropq; + struct path h_path; + struct dentry *wh_dentry, *parent, *opq_dentry; + struct inode *h_inode; + struct super_block *sb; + struct { + struct au_pin pin; + struct au_dtime dt; + } *a; /* reduce the stack usage */ + struct au_wr_dir_args wr_dir_args = { + .force_btgt = -1, + .flags = AuWrDir_ADD_ENTRY | AuWrDir_ISDIR + }; + + IMustLock(dir); + + err = -ENOMEM; + a = kmalloc(sizeof(*a), GFP_NOFS); + if (unlikely(!a)) + goto out; + + err = aufs_read_lock(dentry, AuLock_DW | AuLock_GEN); + if (unlikely(err)) + goto out_free; + err = au_d_may_add(dentry); + if (unlikely(err)) + goto out_unlock; + + parent = dentry->d_parent; /* dir inode is locked */ + di_write_lock_parent(parent); + wh_dentry = lock_hdir_lkup_wh(dentry, &a->dt, /*src_dentry*/NULL, + &a->pin, &wr_dir_args); + err = PTR_ERR(wh_dentry); + if (IS_ERR(wh_dentry)) + goto out_parent; + + sb = dentry->d_sb; + bindex = au_dbstart(dentry); + h_path.dentry = au_h_dptr(dentry, bindex); + h_path.mnt = au_sbr_mnt(sb, bindex); + err = vfsub_mkdir(au_pinned_h_dir(&a->pin), &h_path, mode); + if (unlikely(err)) + goto out_unpin; + + /* make the dir opaque */ + diropq = 0; + h_inode = d_inode(h_path.dentry); + if (wh_dentry + || au_opt_test(au_mntflags(sb), ALWAYS_DIROPQ)) { + inode_lock_nested(h_inode, AuLsc_I_CHILD); + opq_dentry = au_diropq_create(dentry, bindex); + inode_unlock(h_inode); + err = PTR_ERR(opq_dentry); + if (IS_ERR(opq_dentry)) + goto out_dir; + dput(opq_dentry); + diropq = 1; + } + + err = epilog(dir, bindex, wh_dentry, dentry); + if (!err) { + inc_nlink(dir); + goto out_unpin; /* success */ + } + + /* revert */ + if (diropq) { + AuLabel(revert opq); + inode_lock_nested(h_inode, AuLsc_I_CHILD); + rerr = au_diropq_remove(dentry, bindex); + inode_unlock(h_inode); + if (rerr) { + AuIOErr("%pd reverting diropq failed(%d, %d)\n", + dentry, err, rerr); + err = -EIO; + } + } + +out_dir: + AuLabel(revert dir); + rerr = vfsub_rmdir(au_pinned_h_dir(&a->pin), &h_path); + if (rerr) { + AuIOErr("%pd reverting dir failed(%d, %d)\n", + dentry, err, rerr); + err = -EIO; + } + au_dtime_revert(&a->dt); +out_unpin: + au_unpin(&a->pin); + dput(wh_dentry); +out_parent: + di_write_unlock(parent); +out_unlock: + if (unlikely(err)) { + au_update_dbstart(dentry); + d_drop(dentry); + } + aufs_read_unlock(dentry, AuLock_DW); +out_free: + kfree(a); +out: + return err; +} diff --git a/fs/aufs/i_op_del.c b/fs/aufs/i_op_del.c new file mode 100644 index 000000000..68741aadb --- /dev/null +++ b/fs/aufs/i_op_del.c @@ -0,0 +1,497 @@ +/* + * Copyright (C) 2005-2016 Junjiro R. Okajima + */ + +/* + * inode operations (del entry) + */ + +#include "aufs.h" + +/* + * decide if a new whiteout for @dentry is necessary or not. + * when it is necessary, prepare the parent dir for the upper branch whose + * branch index is @bcpup for creation. the actual creation of the whiteout will + * be done by caller. + * return value: + * 0: wh is unnecessary + * plus: wh is necessary + * minus: error + */ +int au_wr_dir_need_wh(struct dentry *dentry, int isdir, aufs_bindex_t *bcpup) +{ + int need_wh, err; + aufs_bindex_t bstart; + struct super_block *sb; + + sb = dentry->d_sb; + bstart = au_dbstart(dentry); + if (*bcpup < 0) { + *bcpup = bstart; + if (au_test_ro(sb, bstart, d_inode(dentry))) { + err = AuWbrCopyup(au_sbi(sb), dentry); + *bcpup = err; + if (unlikely(err < 0)) + goto out; + } + } else + AuDebugOn(bstart < *bcpup + || au_test_ro(sb, *bcpup, d_inode(dentry))); + AuDbg("bcpup %d, bstart %d\n", *bcpup, bstart); + + if (*bcpup != bstart) { + err = au_cpup_dirs(dentry, *bcpup); + if (unlikely(err)) + goto out; + need_wh = 1; + } else { + struct au_dinfo *dinfo, *tmp; + + need_wh = -ENOMEM; + dinfo = au_di(dentry); + tmp = au_di_alloc(sb, AuLsc_DI_TMP); + if (tmp) { + au_di_cp(tmp, dinfo); + au_di_swap(tmp, dinfo); + /* returns the number of positive dentries */ + need_wh = au_lkup_dentry(dentry, bstart + 1, /*type*/0); + au_di_swap(tmp, dinfo); + au_rw_write_unlock(&tmp->di_rwsem); + au_di_free(tmp); + } + } + AuDbg("need_wh %d\n", need_wh); + err = need_wh; + +out: + return err; +} + +/* + * simple tests for the del-entry operations. + * following the checks in vfs, plus the parent-child relationship. + */ +int au_may_del(struct dentry *dentry, aufs_bindex_t bindex, + struct dentry *h_parent, int isdir) +{ + int err; + umode_t h_mode; + struct dentry *h_dentry, *h_latest; + struct inode *h_inode; + + h_dentry = au_h_dptr(dentry, bindex); + if (d_really_is_positive(dentry)) { + err = -ENOENT; + if (unlikely(d_is_negative(h_dentry))) + goto out; + h_inode = d_inode(h_dentry); + if (unlikely(!h_inode->i_nlink)) + goto out; + + h_mode = h_inode->i_mode; + if (!isdir) { + err = -EISDIR; + if (unlikely(S_ISDIR(h_mode))) + goto out; + } else if (unlikely(!S_ISDIR(h_mode))) { + err = -ENOTDIR; + goto out; + } + } else { + /* rename(2) case */ + err = -EIO; + if (unlikely(d_is_positive(h_dentry))) + goto out; + } + + err = -ENOENT; + /* expected parent dir is locked */ + if (unlikely(h_parent != h_dentry->d_parent)) + goto out; + err = 0; + + /* + * rmdir a dir may break the consistency on some filesystem. + * let's try heavy test. + */ + err = -EACCES; + if (unlikely(!au_opt_test(au_mntflags(dentry->d_sb), DIRPERM1) + && au_test_h_perm(d_inode(h_parent), + MAY_EXEC | MAY_WRITE))) + goto out; + + h_latest = au_sio_lkup_one(&dentry->d_name, h_parent); + err = -EIO; + if (IS_ERR(h_latest)) + goto out; + if (h_latest == h_dentry) + err = 0; + dput(h_latest); + +out: + return err; +} + +/* + * decide the branch where we operate for @dentry. the branch index will be set + * @rbcpup. after diciding it, 'pin' it and store the timestamps of the parent + * dir for reverting. + * when a new whiteout is necessary, create it. + */ +static struct dentry* +lock_hdir_create_wh(struct dentry *dentry, int isdir, aufs_bindex_t *rbcpup, + struct au_dtime *dt, struct au_pin *pin) +{ + struct dentry *wh_dentry; + struct super_block *sb; + struct path h_path; + int err, need_wh; + unsigned int udba; + aufs_bindex_t bcpup; + + need_wh = au_wr_dir_need_wh(dentry, isdir, rbcpup); + wh_dentry = ERR_PTR(need_wh); + if (unlikely(need_wh < 0)) + goto out; + + sb = dentry->d_sb; + udba = au_opt_udba(sb); + bcpup = *rbcpup; + err = au_pin(pin, dentry, bcpup, udba, + AuPin_DI_LOCKED | AuPin_MNT_WRITE); + wh_dentry = ERR_PTR(err); + if (unlikely(err)) + goto out; + + h_path.dentry = au_pinned_h_parent(pin); + if (udba != AuOpt_UDBA_NONE + && au_dbstart(dentry) == bcpup) { + err = au_may_del(dentry, bcpup, h_path.dentry, isdir); + wh_dentry = ERR_PTR(err); + if (unlikely(err)) + goto out_unpin; + } + + h_path.mnt = au_sbr_mnt(sb, bcpup); + au_dtime_store(dt, au_pinned_parent(pin), &h_path); + wh_dentry = NULL; + if (!need_wh) + goto out; /* success, no need to create whiteout */ + + wh_dentry = au_wh_create(dentry, bcpup, h_path.dentry); + if (IS_ERR(wh_dentry)) + goto out_unpin; + + /* returns with the parent is locked and wh_dentry is dget-ed */ + goto out; /* success */ + +out_unpin: + au_unpin(pin); +out: + return wh_dentry; +} + +/* + * when removing a dir, rename it to a unique temporary whiteout-ed name first + * in order to be revertible and save time for removing many child whiteouts + * under the dir. + * returns 1 when there are too many child whiteout and caller should remove + * them asynchronously. returns 0 when the number of children is enough small to + * remove now or the branch fs is a remote fs. + * otherwise return an error. + */ +static int renwh_and_rmdir(struct dentry *dentry, aufs_bindex_t bindex, + struct au_nhash *whlist, struct inode *dir) +{ + int rmdir_later, err, dirwh; + struct dentry *h_dentry; + struct super_block *sb; + struct inode *inode; + + sb = dentry->d_sb; + SiMustAnyLock(sb); + h_dentry = au_h_dptr(dentry, bindex); + err = au_whtmp_ren(h_dentry, au_sbr(sb, bindex)); + if (unlikely(err)) + goto out; + + /* stop monitoring */ + inode = d_inode(dentry); + au_hn_free(au_hi(inode, bindex)); + + if (!au_test_fs_remote(h_dentry->d_sb)) { + dirwh = au_sbi(sb)->si_dirwh; + rmdir_later = (dirwh <= 1); + if (!rmdir_later) + rmdir_later = au_nhash_test_longer_wh(whlist, bindex, + dirwh); + if (rmdir_later) + return rmdir_later; + } + + err = au_whtmp_rmdir(dir, bindex, h_dentry, whlist); + if (unlikely(err)) { + AuIOErr("rmdir %pd, b%d failed, %d. ignored\n", + h_dentry, bindex, err); + err = 0; + } + +out: + AuTraceErr(err); + return err; +} + +/* + * final procedure for deleting a entry. + * maintain dentry and iattr. + */ +static void epilog(struct inode *dir, struct dentry *dentry, + aufs_bindex_t bindex) +{ + struct inode *inode; + + inode = d_inode(dentry); + d_drop(dentry); + inode->i_ctime = dir->i_ctime; + + au_dir_ts(dir, bindex); + dir->i_version++; +} + +/* + * when an error happened, remove the created whiteout and revert everything. + */ +static int do_revert(int err, struct inode *dir, aufs_bindex_t bindex, + aufs_bindex_t bwh, struct dentry *wh_dentry, + struct dentry *dentry, struct au_dtime *dt) +{ + int rerr; + struct path h_path = { + .dentry = wh_dentry, + .mnt = au_sbr_mnt(dir->i_sb, bindex) + }; + + rerr = au_wh_unlink_dentry(au_h_iptr(dir, bindex), &h_path, dentry); + if (!rerr) { + au_set_dbwh(dentry, bwh); + au_dtime_revert(dt); + return 0; + } + + AuIOErr("%pd reverting whiteout failed(%d, %d)\n", dentry, err, rerr); + return -EIO; +} + +/* ---------------------------------------------------------------------- */ + +int aufs_unlink(struct inode *dir, struct dentry *dentry) +{ + int err; + aufs_bindex_t bwh, bindex, bstart; + struct inode *inode, *h_dir, *delegated; + struct dentry *parent, *wh_dentry; + /* to reuduce stack size */ + struct { + struct au_dtime dt; + struct au_pin pin; + struct path h_path; + } *a; + + IMustLock(dir); + + err = -ENOMEM; + a = kmalloc(sizeof(*a), GFP_NOFS); + if (unlikely(!a)) + goto out; + + err = aufs_read_lock(dentry, AuLock_DW | AuLock_GEN); + if (unlikely(err)) + goto out_free; + err = au_d_hashed_positive(dentry); + if (unlikely(err)) + goto out_unlock; + inode = d_inode(dentry); + IMustLock(inode); + err = -EISDIR; + if (unlikely(d_is_dir(dentry))) + goto out_unlock; /* possible? */ + + bstart = au_dbstart(dentry); + bwh = au_dbwh(dentry); + bindex = -1; + parent = dentry->d_parent; /* dir inode is locked */ + di_write_lock_parent(parent); + wh_dentry = lock_hdir_create_wh(dentry, /*isdir*/0, &bindex, &a->dt, + &a->pin); + err = PTR_ERR(wh_dentry); + if (IS_ERR(wh_dentry)) + goto out_parent; + + a->h_path.mnt = au_sbr_mnt(dentry->d_sb, bstart); + a->h_path.dentry = au_h_dptr(dentry, bstart); + dget(a->h_path.dentry); + if (bindex == bstart) { + h_dir = au_pinned_h_dir(&a->pin); + delegated = NULL; + err = vfsub_unlink(h_dir, &a->h_path, &delegated, /*force*/0); + if (unlikely(err == -EWOULDBLOCK)) { + pr_warn("cannot retry for NFSv4 delegation" + " for an internal unlink\n"); + iput(delegated); + } + } else { + /* dir inode is locked */ + h_dir = d_inode(wh_dentry->d_parent); + IMustLock(h_dir); + err = 0; + } + + if (!err) { + vfsub_drop_nlink(inode); + epilog(dir, dentry, bindex); + + /* update target timestamps */ + if (bindex == bstart) { + vfsub_update_h_iattr(&a->h_path, /*did*/NULL); + /*ignore*/ + inode->i_ctime = d_inode(a->h_path.dentry)->i_ctime; + } else + /* todo: this timestamp may be reverted later */ + inode->i_ctime = h_dir->i_ctime; + goto out_unpin; /* success */ + } + + /* revert */ + if (wh_dentry) { + int rerr; + + rerr = do_revert(err, dir, bindex, bwh, wh_dentry, dentry, + &a->dt); + if (rerr) + err = rerr; + } + +out_unpin: + au_unpin(&a->pin); + dput(wh_dentry); + dput(a->h_path.dentry); +out_parent: + di_write_unlock(parent); +out_unlock: + aufs_read_unlock(dentry, AuLock_DW); +out_free: + kfree(a); +out: + return err; +} + +int aufs_rmdir(struct inode *dir, struct dentry *dentry) +{ + int err, rmdir_later; + aufs_bindex_t bwh, bindex, bstart; + struct inode *inode; + struct dentry *parent, *wh_dentry, *h_dentry; + struct au_whtmp_rmdir *args; + /* to reuduce stack size */ + struct { + struct au_dtime dt; + struct au_pin pin; + } *a; + + IMustLock(dir); + + err = -ENOMEM; + a = kmalloc(sizeof(*a), GFP_NOFS); + if (unlikely(!a)) + goto out; + + err = aufs_read_lock(dentry, AuLock_DW | AuLock_FLUSH | AuLock_GEN); + if (unlikely(err)) + goto out_free; + err = au_alive_dir(dentry); + if (unlikely(err)) + goto out_unlock; + inode = d_inode(dentry); + IMustLock(inode); + err = -ENOTDIR; + if (unlikely(!d_is_dir(dentry))) + goto out_unlock; /* possible? */ + + err = -ENOMEM; + args = au_whtmp_rmdir_alloc(dir->i_sb, GFP_NOFS); + if (unlikely(!args)) + goto out_unlock; + + parent = dentry->d_parent; /* dir inode is locked */ + di_write_lock_parent(parent); + err = au_test_empty(dentry, &args->whlist); + if (unlikely(err)) + goto out_parent; + + bstart = au_dbstart(dentry); + bwh = au_dbwh(dentry); + bindex = -1; + wh_dentry = lock_hdir_create_wh(dentry, /*isdir*/1, &bindex, &a->dt, + &a->pin); + err = PTR_ERR(wh_dentry); + if (IS_ERR(wh_dentry)) + goto out_parent; + + h_dentry = au_h_dptr(dentry, bstart); + dget(h_dentry); + rmdir_later = 0; + if (bindex == bstart) { + err = renwh_and_rmdir(dentry, bstart, &args->whlist, dir); + if (err > 0) { + rmdir_later = err; + err = 0; + } + } else { + /* stop monitoring */ + au_hn_free(au_hi(inode, bstart)); + + /* dir inode is locked */ + IMustLock(d_inode(wh_dentry->d_parent)); + err = 0; + } + + if (!err) { + vfsub_dead_dir(inode); + au_set_dbdiropq(dentry, -1); + epilog(dir, dentry, bindex); + + if (rmdir_later) { + au_whtmp_kick_rmdir(dir, bstart, h_dentry, args); + args = NULL; + } + + goto out_unpin; /* success */ + } + + /* revert */ + AuLabel(revert); + if (wh_dentry) { + int rerr; + + rerr = do_revert(err, dir, bindex, bwh, wh_dentry, dentry, + &a->dt); + if (rerr) + err = rerr; + } + +out_unpin: + au_unpin(&a->pin); + dput(wh_dentry); + dput(h_dentry); +out_parent: + di_write_unlock(parent); + if (args) + au_whtmp_rmdir_free(args); +out_unlock: + aufs_read_unlock(dentry, AuLock_DW); +out_free: + kfree(a); +out: + AuTraceErr(err); + return err; +} diff --git a/fs/aufs/i_op_ren.c b/fs/aufs/i_op_ren.c new file mode 100644 index 000000000..de4e03bf0 --- /dev/null +++ b/fs/aufs/i_op_ren.c @@ -0,0 +1,1002 @@ +/* + * Copyright (C) 2005-2016 Junjiro R. Okajima + */ + +/* + * inode operation (rename entry) + * todo: this is crazy monster + */ + +#include "aufs.h" + +enum { AuSRC, AuDST, AuSrcDst }; +enum { AuPARENT, AuCHILD, AuParentChild }; + +#define AuRen_ISDIR 1 +#define AuRen_ISSAMEDIR (1 << 1) +#define AuRen_WHSRC (1 << 2) +#define AuRen_WHDST (1 << 3) +#define AuRen_MNT_WRITE (1 << 4) +#define AuRen_DT_DSTDIR (1 << 5) +#define AuRen_DIROPQ (1 << 6) +#define au_ftest_ren(flags, name) ((flags) & AuRen_##name) +#define au_fset_ren(flags, name) \ + do { (flags) |= AuRen_##name; } while (0) +#define au_fclr_ren(flags, name) \ + do { (flags) &= ~AuRen_##name; } while (0) + +struct au_ren_args { + struct { + struct dentry *dentry, *h_dentry, *parent, *h_parent, + *wh_dentry; + struct inode *dir, *inode; + struct au_hinode *hdir; + struct au_dtime dt[AuParentChild]; + aufs_bindex_t bstart; + } sd[AuSrcDst]; + +#define src_dentry sd[AuSRC].dentry +#define src_dir sd[AuSRC].dir +#define src_inode sd[AuSRC].inode +#define src_h_dentry sd[AuSRC].h_dentry +#define src_parent sd[AuSRC].parent +#define src_h_parent sd[AuSRC].h_parent +#define src_wh_dentry sd[AuSRC].wh_dentry +#define src_hdir sd[AuSRC].hdir +#define src_h_dir sd[AuSRC].hdir->hi_inode +#define src_dt sd[AuSRC].dt +#define src_bstart sd[AuSRC].bstart + +#define dst_dentry sd[AuDST].dentry +#define dst_dir sd[AuDST].dir +#define dst_inode sd[AuDST].inode +#define dst_h_dentry sd[AuDST].h_dentry +#define dst_parent sd[AuDST].parent +#define dst_h_parent sd[AuDST].h_parent +#define dst_wh_dentry sd[AuDST].wh_dentry +#define dst_hdir sd[AuDST].hdir +#define dst_h_dir sd[AuDST].hdir->hi_inode +#define dst_dt sd[AuDST].dt +#define dst_bstart sd[AuDST].bstart + + struct dentry *h_trap; + struct au_branch *br; + struct au_hinode *src_hinode; + struct path h_path; + struct au_nhash whlist; + aufs_bindex_t btgt, src_bwh, src_bdiropq; + + unsigned int flags; + + struct au_whtmp_rmdir *thargs; + struct dentry *h_dst; +}; + +/* ---------------------------------------------------------------------- */ + +/* + * functions for reverting. + * when an error happened in a single rename systemcall, we should revert + * everything as if nothing happened. + * we don't need to revert the copied-up/down the parent dir since they are + * harmless. + */ + +#define RevertFailure(fmt, ...) do { \ + AuIOErr("revert failure: " fmt " (%d, %d)\n", \ + ##__VA_ARGS__, err, rerr); \ + err = -EIO; \ +} while (0) + +static void au_ren_rev_diropq(int err, struct au_ren_args *a) +{ + int rerr; + + au_hn_imtx_lock_nested(a->src_hinode, AuLsc_I_CHILD); + rerr = au_diropq_remove(a->src_dentry, a->btgt); + au_hn_imtx_unlock(a->src_hinode); + au_set_dbdiropq(a->src_dentry, a->src_bdiropq); + if (rerr) + RevertFailure("remove diropq %pd", a->src_dentry); +} + +static void au_ren_rev_rename(int err, struct au_ren_args *a) +{ + int rerr; + struct inode *delegated; + + a->h_path.dentry = vfsub_lkup_one(&a->src_dentry->d_name, + a->src_h_parent); + rerr = PTR_ERR(a->h_path.dentry); + if (IS_ERR(a->h_path.dentry)) { + RevertFailure("lkup one %pd", a->src_dentry); + return; + } + + delegated = NULL; + rerr = vfsub_rename(a->dst_h_dir, + au_h_dptr(a->src_dentry, a->btgt), + a->src_h_dir, &a->h_path, &delegated); + if (unlikely(rerr == -EWOULDBLOCK)) { + pr_warn("cannot retry for NFSv4 delegation" + " for an internal rename\n"); + iput(delegated); + } + d_drop(a->h_path.dentry); + dput(a->h_path.dentry); + /* au_set_h_dptr(a->src_dentry, a->btgt, NULL); */ + if (rerr) + RevertFailure("rename %pd", a->src_dentry); +} + +static void au_ren_rev_whtmp(int err, struct au_ren_args *a) +{ + int rerr; + struct inode *delegated; + + a->h_path.dentry = vfsub_lkup_one(&a->dst_dentry->d_name, + a->dst_h_parent); + rerr = PTR_ERR(a->h_path.dentry); + if (IS_ERR(a->h_path.dentry)) { + RevertFailure("lkup one %pd", a->dst_dentry); + return; + } + if (d_is_positive(a->h_path.dentry)) { + d_drop(a->h_path.dentry); + dput(a->h_path.dentry); + return; + } + + delegated = NULL; + rerr = vfsub_rename(a->dst_h_dir, a->h_dst, a->dst_h_dir, &a->h_path, + &delegated); + if (unlikely(rerr == -EWOULDBLOCK)) { + pr_warn("cannot retry for NFSv4 delegation" + " for an internal rename\n"); + iput(delegated); + } + d_drop(a->h_path.dentry); + dput(a->h_path.dentry); + if (!rerr) + au_set_h_dptr(a->dst_dentry, a->btgt, dget(a->h_dst)); + else + RevertFailure("rename %pd", a->h_dst); +} + +static void au_ren_rev_whsrc(int err, struct au_ren_args *a) +{ + int rerr; + + a->h_path.dentry = a->src_wh_dentry; + rerr = au_wh_unlink_dentry(a->src_h_dir, &a->h_path, a->src_dentry); + au_set_dbwh(a->src_dentry, a->src_bwh); + if (rerr) + RevertFailure("unlink %pd", a->src_wh_dentry); +} +#undef RevertFailure + +/* ---------------------------------------------------------------------- */ + +/* + * when we have to copyup the renaming entry, do it with the rename-target name + * in order to minimize the cost (the later actual rename is unnecessary). + * otherwise rename it on the target branch. + */ +static int au_ren_or_cpup(struct au_ren_args *a) +{ + int err; + struct dentry *d; + struct inode *delegated; + + d = a->src_dentry; + if (au_dbstart(d) == a->btgt) { + a->h_path.dentry = a->dst_h_dentry; + if (au_ftest_ren(a->flags, DIROPQ) + && au_dbdiropq(d) == a->btgt) + au_fclr_ren(a->flags, DIROPQ); + AuDebugOn(au_dbstart(d) != a->btgt); + delegated = NULL; + err = vfsub_rename(a->src_h_dir, au_h_dptr(d, a->btgt), + a->dst_h_dir, &a->h_path, &delegated); + if (unlikely(err == -EWOULDBLOCK)) { + pr_warn("cannot retry for NFSv4 delegation" + " for an internal rename\n"); + iput(delegated); + } + } else + BUG(); + + if (!err && a->h_dst) + /* it will be set to dinfo later */ + dget(a->h_dst); + + return err; +} + +/* cf. aufs_rmdir() */ +static int au_ren_del_whtmp(struct au_ren_args *a) +{ + int err; + struct inode *dir; + + dir = a->dst_dir; + SiMustAnyLock(dir->i_sb); + if (!au_nhash_test_longer_wh(&a->whlist, a->btgt, + au_sbi(dir->i_sb)->si_dirwh) + || au_test_fs_remote(a->h_dst->d_sb)) { + err = au_whtmp_rmdir(dir, a->btgt, a->h_dst, &a->whlist); + if (unlikely(err)) + pr_warn("failed removing whtmp dir %pd (%d), " + "ignored.\n", a->h_dst, err); + } else { + au_nhash_wh_free(&a->thargs->whlist); + a->thargs->whlist = a->whlist; + a->whlist.nh_num = 0; + au_whtmp_kick_rmdir(dir, a->btgt, a->h_dst, a->thargs); + dput(a->h_dst); + a->thargs = NULL; + } + + return 0; +} + +/* make it 'opaque' dir. */ +static int au_ren_diropq(struct au_ren_args *a) +{ + int err; + struct dentry *diropq; + + err = 0; + a->src_bdiropq = au_dbdiropq(a->src_dentry); + a->src_hinode = au_hi(a->src_inode, a->btgt); + au_hn_imtx_lock_nested(a->src_hinode, AuLsc_I_CHILD); + diropq = au_diropq_create(a->src_dentry, a->btgt); + au_hn_imtx_unlock(a->src_hinode); + if (IS_ERR(diropq)) + err = PTR_ERR(diropq); + else + dput(diropq); + + return err; +} + +static int do_rename(struct au_ren_args *a) +{ + int err; + struct dentry *d, *h_d; + + /* prepare workqueue args for asynchronous rmdir */ + h_d = a->dst_h_dentry; + if (au_ftest_ren(a->flags, ISDIR) && d_is_positive(h_d)) { + err = -ENOMEM; + a->thargs = au_whtmp_rmdir_alloc(a->src_dentry->d_sb, GFP_NOFS); + if (unlikely(!a->thargs)) + goto out; + a->h_dst = dget(h_d); + } + + /* create whiteout for src_dentry */ + if (au_ftest_ren(a->flags, WHSRC)) { + a->src_bwh = au_dbwh(a->src_dentry); + AuDebugOn(a->src_bwh >= 0); + a->src_wh_dentry + = au_wh_create(a->src_dentry, a->btgt, a->src_h_parent); + err = PTR_ERR(a->src_wh_dentry); + if (IS_ERR(a->src_wh_dentry)) + goto out_thargs; + } + + /* lookup whiteout for dentry */ + if (au_ftest_ren(a->flags, WHDST)) { + h_d = au_wh_lkup(a->dst_h_parent, &a->dst_dentry->d_name, + a->br); + err = PTR_ERR(h_d); + if (IS_ERR(h_d)) + goto out_whsrc; + if (d_is_negative(h_d)) + dput(h_d); + else + a->dst_wh_dentry = h_d; + } + + /* rename dentry to tmpwh */ + if (a->thargs) { + err = au_whtmp_ren(a->dst_h_dentry, a->br); + if (unlikely(err)) + goto out_whdst; + + d = a->dst_dentry; + au_set_h_dptr(d, a->btgt, NULL); + err = au_lkup_neg(d, a->btgt, /*wh*/0); + if (unlikely(err)) + goto out_whtmp; + a->dst_h_dentry = au_h_dptr(d, a->btgt); + } + + BUG_ON(d_is_positive(a->dst_h_dentry) && a->src_bstart != a->btgt); + + /* rename by vfs_rename or cpup */ + d = a->dst_dentry; + if (au_ftest_ren(a->flags, ISDIR) + && (a->dst_wh_dentry + || au_dbdiropq(d) == a->btgt + /* hide the lower to keep xino */ + || a->btgt < au_dbend(d) + || au_opt_test(au_mntflags(d->d_sb), ALWAYS_DIROPQ))) + au_fset_ren(a->flags, DIROPQ); + err = au_ren_or_cpup(a); + if (unlikely(err)) + /* leave the copied-up one */ + goto out_whtmp; + + /* make dir opaque */ + if (au_ftest_ren(a->flags, DIROPQ)) { + err = au_ren_diropq(a); + if (unlikely(err)) + goto out_rename; + } + + /* update target timestamps */ + AuDebugOn(au_dbstart(a->src_dentry) != a->btgt); + a->h_path.dentry = au_h_dptr(a->src_dentry, a->btgt); + vfsub_update_h_iattr(&a->h_path, /*did*/NULL); /*ignore*/ + a->src_inode->i_ctime = d_inode(a->h_path.dentry)->i_ctime; + + /* remove whiteout for dentry */ + if (a->dst_wh_dentry) { + a->h_path.dentry = a->dst_wh_dentry; + err = au_wh_unlink_dentry(a->dst_h_dir, &a->h_path, + a->dst_dentry); + if (unlikely(err)) + goto out_diropq; + } + + /* remove whtmp */ + if (a->thargs) + au_ren_del_whtmp(a); /* ignore this error */ + + au_fhsm_wrote(a->src_dentry->d_sb, a->btgt, /*force*/0); + err = 0; + goto out_success; + +out_diropq: + if (au_ftest_ren(a->flags, DIROPQ)) + au_ren_rev_diropq(err, a); +out_rename: + au_ren_rev_rename(err, a); + dput(a->h_dst); +out_whtmp: + if (a->thargs) + au_ren_rev_whtmp(err, a); +out_whdst: + dput(a->dst_wh_dentry); + a->dst_wh_dentry = NULL; +out_whsrc: + if (a->src_wh_dentry) + au_ren_rev_whsrc(err, a); +out_success: + dput(a->src_wh_dentry); + dput(a->dst_wh_dentry); +out_thargs: + if (a->thargs) { + dput(a->h_dst); + au_whtmp_rmdir_free(a->thargs); + a->thargs = NULL; + } +out: + return err; +} + +/* ---------------------------------------------------------------------- */ + +/* + * test if @dentry dir can be rename destination or not. + * success means, it is a logically empty dir. + */ +static int may_rename_dstdir(struct dentry *dentry, struct au_nhash *whlist) +{ + return au_test_empty(dentry, whlist); +} + +/* + * test if @dentry dir can be rename source or not. + * if it can, return 0 and @children is filled. + * success means, + * - it is a logically empty dir. + * - or, it exists on writable branch and has no children including whiteouts + * on the lower branch. + */ +static int may_rename_srcdir(struct dentry *dentry, aufs_bindex_t btgt) +{ + int err; + unsigned int rdhash; + aufs_bindex_t bstart; + + bstart = au_dbstart(dentry); + if (bstart != btgt) { + struct au_nhash whlist; + + SiMustAnyLock(dentry->d_sb); + rdhash = au_sbi(dentry->d_sb)->si_rdhash; + if (!rdhash) + rdhash = au_rdhash_est(au_dir_size(/*file*/NULL, + dentry)); + err = au_nhash_alloc(&whlist, rdhash, GFP_NOFS); + if (unlikely(err)) + goto out; + err = au_test_empty(dentry, &whlist); + au_nhash_wh_free(&whlist); + goto out; + } + + if (bstart == au_dbtaildir(dentry)) + return 0; /* success */ + + err = au_test_empty_lower(dentry); + +out: + if (err == -ENOTEMPTY) { + AuWarn1("renaming dir who has child(ren) on multiple branches," + " is not supported\n"); + err = -EXDEV; + } + return err; +} + +/* side effect: sets whlist and h_dentry */ +static int au_ren_may_dir(struct au_ren_args *a) +{ + int err; + unsigned int rdhash; + struct dentry *d; + + d = a->dst_dentry; + SiMustAnyLock(d->d_sb); + + err = 0; + if (au_ftest_ren(a->flags, ISDIR) && a->dst_inode) { + rdhash = au_sbi(d->d_sb)->si_rdhash; + if (!rdhash) + rdhash = au_rdhash_est(au_dir_size(/*file*/NULL, d)); + err = au_nhash_alloc(&a->whlist, rdhash, GFP_NOFS); + if (unlikely(err)) + goto out; + + au_set_dbstart(d, a->dst_bstart); + err = may_rename_dstdir(d, &a->whlist); + au_set_dbstart(d, a->btgt); + } + a->dst_h_dentry = au_h_dptr(d, au_dbstart(d)); + if (unlikely(err)) + goto out; + + d = a->src_dentry; + a->src_h_dentry = au_h_dptr(d, au_dbstart(d)); + if (au_ftest_ren(a->flags, ISDIR)) { + err = may_rename_srcdir(d, a->btgt); + if (unlikely(err)) { + au_nhash_wh_free(&a->whlist); + a->whlist.nh_num = 0; + } + } +out: + return err; +} + +/* ---------------------------------------------------------------------- */ + +/* + * simple tests for rename. + * following the checks in vfs, plus the parent-child relationship. + */ +static int au_may_ren(struct au_ren_args *a) +{ + int err, isdir; + struct inode *h_inode; + + if (a->src_bstart == a->btgt) { + err = au_may_del(a->src_dentry, a->btgt, a->src_h_parent, + au_ftest_ren(a->flags, ISDIR)); + if (unlikely(err)) + goto out; + err = -EINVAL; + if (unlikely(a->src_h_dentry == a->h_trap)) + goto out; + } + + err = 0; + if (a->dst_bstart != a->btgt) + goto out; + + err = -ENOTEMPTY; + if (unlikely(a->dst_h_dentry == a->h_trap)) + goto out; + + err = -EIO; + isdir = !!au_ftest_ren(a->flags, ISDIR); + if (d_really_is_negative(a->dst_dentry)) { + if (d_is_negative(a->dst_h_dentry)) + err = au_may_add(a->dst_dentry, a->btgt, + a->dst_h_parent, isdir); + } else { + if (unlikely(d_is_negative(a->dst_h_dentry))) + goto out; + h_inode = d_inode(a->dst_h_dentry); + if (h_inode->i_nlink) + err = au_may_del(a->dst_dentry, a->btgt, + a->dst_h_parent, isdir); + } + +out: + if (unlikely(err == -ENOENT || err == -EEXIST)) + err = -EIO; + AuTraceErr(err); + return err; +} + +/* ---------------------------------------------------------------------- */ + +/* + * locking order + * (VFS) + * - src_dir and dir by lock_rename() + * - inode if exitsts + * (aufs) + * - lock all + * + src_dentry and dentry by aufs_read_and_write_lock2() which calls, + * + si_read_lock + * + di_write_lock2_child() + * + di_write_lock_child() + * + ii_write_lock_child() + * + di_write_lock_child2() + * + ii_write_lock_child2() + * + src_parent and parent + * + di_write_lock_parent() + * + ii_write_lock_parent() + * + di_write_lock_parent2() + * + ii_write_lock_parent2() + * + lower src_dir and dir by vfsub_lock_rename() + * + verify the every relationships between child and parent. if any + * of them failed, unlock all and return -EBUSY. + */ +static void au_ren_unlock(struct au_ren_args *a) +{ + vfsub_unlock_rename(a->src_h_parent, a->src_hdir, + a->dst_h_parent, a->dst_hdir); + if (au_ftest_ren(a->flags, MNT_WRITE)) + vfsub_mnt_drop_write(au_br_mnt(a->br)); +} + +static int au_ren_lock(struct au_ren_args *a) +{ + int err; + unsigned int udba; + + err = 0; + a->src_h_parent = au_h_dptr(a->src_parent, a->btgt); + a->src_hdir = au_hi(a->src_dir, a->btgt); + a->dst_h_parent = au_h_dptr(a->dst_parent, a->btgt); + a->dst_hdir = au_hi(a->dst_dir, a->btgt); + + err = vfsub_mnt_want_write(au_br_mnt(a->br)); + if (unlikely(err)) + goto out; + au_fset_ren(a->flags, MNT_WRITE); + a->h_trap = vfsub_lock_rename(a->src_h_parent, a->src_hdir, + a->dst_h_parent, a->dst_hdir); + udba = au_opt_udba(a->src_dentry->d_sb); + if (unlikely(a->src_hdir->hi_inode != d_inode(a->src_h_parent) + || a->dst_hdir->hi_inode != d_inode(a->dst_h_parent))) + err = au_busy_or_stale(); + if (!err && au_dbstart(a->src_dentry) == a->btgt) + err = au_h_verify(a->src_h_dentry, udba, + d_inode(a->src_h_parent), a->src_h_parent, + a->br); + if (!err && au_dbstart(a->dst_dentry) == a->btgt) + err = au_h_verify(a->dst_h_dentry, udba, + d_inode(a->dst_h_parent), a->dst_h_parent, + a->br); + if (!err) + goto out; /* success */ + + err = au_busy_or_stale(); + au_ren_unlock(a); + +out: + return err; +} + +/* ---------------------------------------------------------------------- */ + +static void au_ren_refresh_dir(struct au_ren_args *a) +{ + struct inode *dir; + + dir = a->dst_dir; + dir->i_version++; + if (au_ftest_ren(a->flags, ISDIR)) { + /* is this updating defined in POSIX? */ + au_cpup_attr_timesizes(a->src_inode); + au_cpup_attr_nlink(dir, /*force*/1); + } + + au_dir_ts(dir, a->btgt); + + if (au_ftest_ren(a->flags, ISSAMEDIR)) + return; + + dir = a->src_dir; + dir->i_version++; + if (au_ftest_ren(a->flags, ISDIR)) + au_cpup_attr_nlink(dir, /*force*/1); + au_dir_ts(dir, a->btgt); +} + +static void au_ren_refresh(struct au_ren_args *a) +{ + aufs_bindex_t bend, bindex; + struct dentry *d, *h_d; + struct inode *i, *h_i; + struct super_block *sb; + + d = a->dst_dentry; + d_drop(d); + if (a->h_dst) + /* already dget-ed by au_ren_or_cpup() */ + au_set_h_dptr(d, a->btgt, a->h_dst); + + i = a->dst_inode; + if (i) { + if (!au_ftest_ren(a->flags, ISDIR)) + vfsub_drop_nlink(i); + else { + vfsub_dead_dir(i); + au_cpup_attr_timesizes(i); + } + au_update_dbrange(d, /*do_put_zero*/1); + } else { + bend = a->btgt; + for (bindex = au_dbstart(d); bindex < bend; bindex++) + au_set_h_dptr(d, bindex, NULL); + bend = au_dbend(d); + for (bindex = a->btgt + 1; bindex <= bend; bindex++) + au_set_h_dptr(d, bindex, NULL); + au_update_dbrange(d, /*do_put_zero*/0); + } + + d = a->src_dentry; + au_set_dbwh(d, -1); + bend = au_dbend(d); + for (bindex = a->btgt + 1; bindex <= bend; bindex++) { + h_d = au_h_dptr(d, bindex); + if (h_d) + au_set_h_dptr(d, bindex, NULL); + } + au_set_dbend(d, a->btgt); + + sb = d->d_sb; + i = a->src_inode; + if (au_opt_test(au_mntflags(sb), PLINK) && au_plink_test(i)) + return; /* success */ + + bend = au_ibend(i); + for (bindex = a->btgt + 1; bindex <= bend; bindex++) { + h_i = au_h_iptr(i, bindex); + if (h_i) { + au_xino_write(sb, bindex, h_i->i_ino, /*ino*/0); + /* ignore this error */ + au_set_h_iptr(i, bindex, NULL, 0); + } + } + au_set_ibend(i, a->btgt); +} + +/* ---------------------------------------------------------------------- */ + +/* mainly for link(2) and rename(2) */ +int au_wbr(struct dentry *dentry, aufs_bindex_t btgt) +{ + aufs_bindex_t bdiropq, bwh; + struct dentry *parent; + struct au_branch *br; + + parent = dentry->d_parent; + IMustLock(d_inode(parent)); /* dir is locked */ + + bdiropq = au_dbdiropq(parent); + bwh = au_dbwh(dentry); + br = au_sbr(dentry->d_sb, btgt); + if (au_br_rdonly(br) + || (0 <= bdiropq && bdiropq < btgt) + || (0 <= bwh && bwh < btgt)) + btgt = -1; + + AuDbg("btgt %d\n", btgt); + return btgt; +} + +/* sets src_bstart, dst_bstart and btgt */ +static int au_ren_wbr(struct au_ren_args *a) +{ + int err; + struct au_wr_dir_args wr_dir_args = { + /* .force_btgt = -1, */ + .flags = AuWrDir_ADD_ENTRY + }; + + a->src_bstart = au_dbstart(a->src_dentry); + a->dst_bstart = au_dbstart(a->dst_dentry); + if (au_ftest_ren(a->flags, ISDIR)) + au_fset_wrdir(wr_dir_args.flags, ISDIR); + wr_dir_args.force_btgt = a->src_bstart; + if (a->dst_inode && a->dst_bstart < a->src_bstart) + wr_dir_args.force_btgt = a->dst_bstart; + wr_dir_args.force_btgt = au_wbr(a->dst_dentry, wr_dir_args.force_btgt); + err = au_wr_dir(a->dst_dentry, a->src_dentry, &wr_dir_args); + a->btgt = err; + + return err; +} + +static void au_ren_dt(struct au_ren_args *a) +{ + a->h_path.dentry = a->src_h_parent; + au_dtime_store(a->src_dt + AuPARENT, a->src_parent, &a->h_path); + if (!au_ftest_ren(a->flags, ISSAMEDIR)) { + a->h_path.dentry = a->dst_h_parent; + au_dtime_store(a->dst_dt + AuPARENT, a->dst_parent, &a->h_path); + } + + au_fclr_ren(a->flags, DT_DSTDIR); + if (!au_ftest_ren(a->flags, ISDIR)) + return; + + a->h_path.dentry = a->src_h_dentry; + au_dtime_store(a->src_dt + AuCHILD, a->src_dentry, &a->h_path); + if (d_is_positive(a->dst_h_dentry)) { + au_fset_ren(a->flags, DT_DSTDIR); + a->h_path.dentry = a->dst_h_dentry; + au_dtime_store(a->dst_dt + AuCHILD, a->dst_dentry, &a->h_path); + } +} + +static void au_ren_rev_dt(int err, struct au_ren_args *a) +{ + struct dentry *h_d; + struct inode *h_inode; + + au_dtime_revert(a->src_dt + AuPARENT); + if (!au_ftest_ren(a->flags, ISSAMEDIR)) + au_dtime_revert(a->dst_dt + AuPARENT); + + if (au_ftest_ren(a->flags, ISDIR) && err != -EIO) { + h_d = a->src_dt[AuCHILD].dt_h_path.dentry; + h_inode = d_inode(h_d); + inode_lock_nested(h_inode, AuLsc_I_CHILD); + au_dtime_revert(a->src_dt + AuCHILD); + inode_unlock(h_inode); + + if (au_ftest_ren(a->flags, DT_DSTDIR)) { + h_d = a->dst_dt[AuCHILD].dt_h_path.dentry; + h_inode = d_inode(h_d); + inode_lock_nested(h_inode, AuLsc_I_CHILD); + au_dtime_revert(a->dst_dt + AuCHILD); + inode_unlock(h_inode); + } + } +} + +/* ---------------------------------------------------------------------- */ + +int aufs_rename(struct inode *_src_dir, struct dentry *_src_dentry, + struct inode *_dst_dir, struct dentry *_dst_dentry) +{ + int err, flags; + /* reduce stack space */ + struct au_ren_args *a; + + AuDbg("%pd, %pd\n", _src_dentry, _dst_dentry); + IMustLock(_src_dir); + IMustLock(_dst_dir); + + err = -ENOMEM; + BUILD_BUG_ON(sizeof(*a) > PAGE_SIZE); + a = kzalloc(sizeof(*a), GFP_NOFS); + if (unlikely(!a)) + goto out; + + a->src_dir = _src_dir; + a->src_dentry = _src_dentry; + a->src_inode = NULL; + if (d_really_is_positive(a->src_dentry)) + a->src_inode = d_inode(a->src_dentry); + a->src_parent = a->src_dentry->d_parent; /* dir inode is locked */ + a->dst_dir = _dst_dir; + a->dst_dentry = _dst_dentry; + a->dst_inode = NULL; + if (d_really_is_positive(a->dst_dentry)) + a->dst_inode = d_inode(a->dst_dentry); + a->dst_parent = a->dst_dentry->d_parent; /* dir inode is locked */ + if (a->dst_inode) { + IMustLock(a->dst_inode); + au_igrab(a->dst_inode); + } + + err = -ENOTDIR; + flags = AuLock_FLUSH | AuLock_NOPLM | AuLock_GEN; + if (d_is_dir(a->src_dentry)) { + au_fset_ren(a->flags, ISDIR); + if (unlikely(d_really_is_positive(a->dst_dentry) + && !d_is_dir(a->dst_dentry))) + goto out_free; + flags |= AuLock_DIRS; + } + err = aufs_read_and_write_lock2(a->dst_dentry, a->src_dentry, flags); + if (unlikely(err)) + goto out_free; + + err = au_d_hashed_positive(a->src_dentry); + if (unlikely(err)) + goto out_unlock; + err = -ENOENT; + if (a->dst_inode) { + /* + * If it is a dir, VFS unhash dst_dentry before this + * function. It means we cannot rely upon d_unhashed(). + */ + if (unlikely(!a->dst_inode->i_nlink)) + goto out_unlock; + if (!S_ISDIR(a->dst_inode->i_mode)) { + err = au_d_hashed_positive(a->dst_dentry); + if (unlikely(err)) + goto out_unlock; + } else if (unlikely(IS_DEADDIR(a->dst_inode))) + goto out_unlock; + } else if (unlikely(d_unhashed(a->dst_dentry))) + goto out_unlock; + + /* + * is it possible? + * yes, it happened (in linux-3.3-rcN) but I don't know why. + * there may exist a problem somewhere else. + */ + err = -EINVAL; + if (unlikely(d_inode(a->dst_parent) == d_inode(a->src_dentry))) + goto out_unlock; + + au_fset_ren(a->flags, ISSAMEDIR); /* temporary */ + di_write_lock_parent(a->dst_parent); + + /* which branch we process */ + err = au_ren_wbr(a); + if (unlikely(err < 0)) + goto out_parent; + a->br = au_sbr(a->dst_dentry->d_sb, a->btgt); + a->h_path.mnt = au_br_mnt(a->br); + + /* are they available to be renamed */ + err = au_ren_may_dir(a); + if (unlikely(err)) + goto out_children; + + /* prepare the writable parent dir on the same branch */ + if (a->dst_bstart == a->btgt) { + au_fset_ren(a->flags, WHDST); + } else { + err = au_cpup_dirs(a->dst_dentry, a->btgt); + if (unlikely(err)) + goto out_children; + } + + if (a->src_dir != a->dst_dir) { + /* + * this temporary unlock is safe, + * because both dir->i_mutex are locked. + */ + di_write_unlock(a->dst_parent); + di_write_lock_parent(a->src_parent); + err = au_wr_dir_need_wh(a->src_dentry, + au_ftest_ren(a->flags, ISDIR), + &a->btgt); + di_write_unlock(a->src_parent); + di_write_lock2_parent(a->src_parent, a->dst_parent, /*isdir*/1); + au_fclr_ren(a->flags, ISSAMEDIR); + } else + err = au_wr_dir_need_wh(a->src_dentry, + au_ftest_ren(a->flags, ISDIR), + &a->btgt); + if (unlikely(err < 0)) + goto out_children; + if (err) + au_fset_ren(a->flags, WHSRC); + + /* cpup src */ + if (a->src_bstart != a->btgt) { + struct au_pin pin; + + err = au_pin(&pin, a->src_dentry, a->btgt, + au_opt_udba(a->src_dentry->d_sb), + AuPin_DI_LOCKED | AuPin_MNT_WRITE); + if (!err) { + struct au_cp_generic cpg = { + .dentry = a->src_dentry, + .bdst = a->btgt, + .bsrc = a->src_bstart, + .len = -1, + .pin = &pin, + .flags = AuCpup_DTIME | AuCpup_HOPEN + }; + AuDebugOn(au_dbstart(a->src_dentry) != a->src_bstart); + err = au_sio_cpup_simple(&cpg); + au_unpin(&pin); + } + if (unlikely(err)) + goto out_children; + a->src_bstart = a->btgt; + a->src_h_dentry = au_h_dptr(a->src_dentry, a->btgt); + au_fset_ren(a->flags, WHSRC); + } + + /* lock them all */ + err = au_ren_lock(a); + if (unlikely(err)) + /* leave the copied-up one */ + goto out_children; + + if (!au_opt_test(au_mntflags(a->dst_dir->i_sb), UDBA_NONE)) + err = au_may_ren(a); + else if (unlikely(a->dst_dentry->d_name.len > AUFS_MAX_NAMELEN)) + err = -ENAMETOOLONG; + if (unlikely(err)) + goto out_hdir; + + /* store timestamps to be revertible */ + au_ren_dt(a); + + /* here we go */ + err = do_rename(a); + if (unlikely(err)) + goto out_dt; + + /* update dir attributes */ + au_ren_refresh_dir(a); + + /* dput/iput all lower dentries */ + au_ren_refresh(a); + + goto out_hdir; /* success */ + +out_dt: + au_ren_rev_dt(err, a); +out_hdir: + au_ren_unlock(a); +out_children: + au_nhash_wh_free(&a->whlist); + if (err && a->dst_inode && a->dst_bstart != a->btgt) { + AuDbg("bstart %d, btgt %d\n", a->dst_bstart, a->btgt); + au_set_h_dptr(a->dst_dentry, a->btgt, NULL); + au_set_dbstart(a->dst_dentry, a->dst_bstart); + } +out_parent: + if (!err) + d_move(a->src_dentry, a->dst_dentry); + else { + au_update_dbstart(a->dst_dentry); + if (!a->dst_inode) + d_drop(a->dst_dentry); + } + if (au_ftest_ren(a->flags, ISSAMEDIR)) + di_write_unlock(a->dst_parent); + else + di_write_unlock2(a->src_parent, a->dst_parent); +out_unlock: + aufs_read_and_write_unlock2(a->dst_dentry, a->src_dentry); +out_free: + iput(a->dst_inode); + if (a->thargs) + au_whtmp_rmdir_free(a->thargs); + kfree(a); +out: + AuTraceErr(err); + return err; +} diff --git a/fs/aufs/iinfo.c b/fs/aufs/iinfo.c new file mode 100644 index 000000000..67ef672a0 --- /dev/null +++ b/fs/aufs/iinfo.c @@ -0,0 +1,264 @@ +/* + * Copyright (C) 2005-2016 Junjiro R. Okajima + */ + +/* + * inode private data + */ + +#include "aufs.h" + +struct inode *au_h_iptr(struct inode *inode, aufs_bindex_t bindex) +{ + struct inode *h_inode; + + IiMustAnyLock(inode); + + h_inode = au_ii(inode)->ii_hinode[0 + bindex].hi_inode; + AuDebugOn(h_inode && atomic_read(&h_inode->i_count) <= 0); + return h_inode; +} + +/* todo: hard/soft set? */ +void au_hiput(struct au_hinode *hinode) +{ + au_hn_free(hinode); + dput(hinode->hi_whdentry); + iput(hinode->hi_inode); +} + +unsigned int au_hi_flags(struct inode *inode, int isdir) +{ + unsigned int flags; + const unsigned int mnt_flags = au_mntflags(inode->i_sb); + + flags = 0; + if (au_opt_test(mnt_flags, XINO)) + au_fset_hi(flags, XINO); + if (isdir && au_opt_test(mnt_flags, UDBA_HNOTIFY)) + au_fset_hi(flags, HNOTIFY); + return flags; +} + +void au_set_h_iptr(struct inode *inode, aufs_bindex_t bindex, + struct inode *h_inode, unsigned int flags) +{ + struct au_hinode *hinode; + struct inode *hi; + struct au_iinfo *iinfo = au_ii(inode); + + IiMustWriteLock(inode); + + hinode = iinfo->ii_hinode + bindex; + hi = hinode->hi_inode; + AuDebugOn(h_inode && atomic_read(&h_inode->i_count) <= 0); + + if (hi) + au_hiput(hinode); + hinode->hi_inode = h_inode; + if (h_inode) { + int err; + struct super_block *sb = inode->i_sb; + struct au_branch *br; + + AuDebugOn(inode->i_mode + && (h_inode->i_mode & S_IFMT) + != (inode->i_mode & S_IFMT)); + if (bindex == iinfo->ii_bstart) + au_cpup_igen(inode, h_inode); + br = au_sbr(sb, bindex); + hinode->hi_id = br->br_id; + if (au_ftest_hi(flags, XINO)) { + err = au_xino_write(sb, bindex, h_inode->i_ino, + inode->i_ino); + if (unlikely(err)) + AuIOErr1("failed au_xino_write() %d\n", err); + } + + if (au_ftest_hi(flags, HNOTIFY) + && au_br_hnotifyable(br->br_perm)) { + err = au_hn_alloc(hinode, inode); + if (unlikely(err)) + AuIOErr1("au_hn_alloc() %d\n", err); + } + } +} + +void au_set_hi_wh(struct inode *inode, aufs_bindex_t bindex, + struct dentry *h_wh) +{ + struct au_hinode *hinode; + + IiMustWriteLock(inode); + + hinode = au_ii(inode)->ii_hinode + bindex; + AuDebugOn(hinode->hi_whdentry); + hinode->hi_whdentry = h_wh; +} + +void au_update_iigen(struct inode *inode, int half) +{ + struct au_iinfo *iinfo; + struct au_iigen *iigen; + unsigned int sigen; + + sigen = au_sigen(inode->i_sb); + iinfo = au_ii(inode); + iigen = &iinfo->ii_generation; + spin_lock(&iigen->ig_spin); + iigen->ig_generation = sigen; + if (half) + au_ig_fset(iigen->ig_flags, HALF_REFRESHED); + else + au_ig_fclr(iigen->ig_flags, HALF_REFRESHED); + spin_unlock(&iigen->ig_spin); +} + +/* it may be called at remount time, too */ +void au_update_ibrange(struct inode *inode, int do_put_zero) +{ + struct au_iinfo *iinfo; + aufs_bindex_t bindex, bend; + + iinfo = au_ii(inode); + if (!iinfo) + return; + + IiMustWriteLock(inode); + + if (do_put_zero && iinfo->ii_bstart >= 0) { + for (bindex = iinfo->ii_bstart; bindex <= iinfo->ii_bend; + bindex++) { + struct inode *h_i; + + h_i = iinfo->ii_hinode[0 + bindex].hi_inode; + if (h_i + && !h_i->i_nlink + && !(h_i->i_state & I_LINKABLE)) + au_set_h_iptr(inode, bindex, NULL, 0); + } + } + + iinfo->ii_bstart = -1; + iinfo->ii_bend = -1; + bend = au_sbend(inode->i_sb); + for (bindex = 0; bindex <= bend; bindex++) + if (iinfo->ii_hinode[0 + bindex].hi_inode) { + iinfo->ii_bstart = bindex; + break; + } + if (iinfo->ii_bstart >= 0) + for (bindex = bend; bindex >= iinfo->ii_bstart; bindex--) + if (iinfo->ii_hinode[0 + bindex].hi_inode) { + iinfo->ii_bend = bindex; + break; + } + AuDebugOn(iinfo->ii_bstart > iinfo->ii_bend); +} + +/* ---------------------------------------------------------------------- */ + +void au_icntnr_init_once(void *_c) +{ + struct au_icntnr *c = _c; + struct au_iinfo *iinfo = &c->iinfo; + static struct lock_class_key aufs_ii; + + spin_lock_init(&iinfo->ii_generation.ig_spin); + au_rw_init(&iinfo->ii_rwsem); + au_rw_class(&iinfo->ii_rwsem, &aufs_ii); + inode_init_once(&c->vfs_inode); +} + +int au_iinfo_init(struct inode *inode) +{ + struct au_iinfo *iinfo; + struct super_block *sb; + int nbr, i; + + sb = inode->i_sb; + iinfo = &(container_of(inode, struct au_icntnr, vfs_inode)->iinfo); + nbr = au_sbend(sb) + 1; + if (unlikely(nbr <= 0)) + nbr = 1; + iinfo->ii_hinode = kcalloc(nbr, sizeof(*iinfo->ii_hinode), GFP_NOFS); + if (iinfo->ii_hinode) { + au_ninodes_inc(sb); + for (i = 0; i < nbr; i++) + iinfo->ii_hinode[i].hi_id = -1; + + iinfo->ii_generation.ig_generation = au_sigen(sb); + iinfo->ii_bstart = -1; + iinfo->ii_bend = -1; + iinfo->ii_vdir = NULL; + return 0; + } + return -ENOMEM; +} + +int au_ii_realloc(struct au_iinfo *iinfo, int nbr) +{ + int err, sz; + struct au_hinode *hip; + + AuRwMustWriteLock(&iinfo->ii_rwsem); + + err = -ENOMEM; + sz = sizeof(*hip) * (iinfo->ii_bend + 1); + if (!sz) + sz = sizeof(*hip); + hip = au_kzrealloc(iinfo->ii_hinode, sz, sizeof(*hip) * nbr, GFP_NOFS); + if (hip) { + iinfo->ii_hinode = hip; + err = 0; + } + + return err; +} + +void au_iinfo_fin(struct inode *inode) +{ + struct au_iinfo *iinfo; + struct au_hinode *hi; + struct super_block *sb; + aufs_bindex_t bindex, bend; + const unsigned char unlinked = !inode->i_nlink; + + iinfo = au_ii(inode); + /* bad_inode case */ + if (!iinfo) + return; + + sb = inode->i_sb; + au_ninodes_dec(sb); + if (si_pid_test(sb)) + au_xino_delete_inode(inode, unlinked); + else { + /* + * it is safe to hide the dependency between sbinfo and + * sb->s_umount. + */ + lockdep_off(); + si_noflush_read_lock(sb); + au_xino_delete_inode(inode, unlinked); + si_read_unlock(sb); + lockdep_on(); + } + + if (iinfo->ii_vdir) + au_vdir_free(iinfo->ii_vdir); + + bindex = iinfo->ii_bstart; + if (bindex >= 0) { + hi = iinfo->ii_hinode + bindex; + bend = iinfo->ii_bend; + while (bindex++ <= bend) { + if (hi->hi_inode) + au_hiput(hi); + hi++; + } + } + kfree(iinfo->ii_hinode); + iinfo->ii_hinode = NULL; + AuRwDestroy(&iinfo->ii_rwsem); +} diff --git a/fs/aufs/inode.c b/fs/aufs/inode.c new file mode 100644 index 000000000..5a87727ba --- /dev/null +++ b/fs/aufs/inode.c @@ -0,0 +1,514 @@ +/* + * Copyright (C) 2005-2016 Junjiro R. Okajima + */ + +/* + * inode functions + */ + +#include "aufs.h" + +struct inode *au_igrab(struct inode *inode) +{ + if (inode) { + AuDebugOn(!atomic_read(&inode->i_count)); + ihold(inode); + } + return inode; +} + +static void au_refresh_hinode_attr(struct inode *inode, int do_version) +{ + au_cpup_attr_all(inode, /*force*/0); + au_update_iigen(inode, /*half*/1); + if (do_version) + inode->i_version++; +} + +static int au_ii_refresh(struct inode *inode, int *update) +{ + int err, e; + umode_t type; + aufs_bindex_t bindex, new_bindex; + struct super_block *sb; + struct au_iinfo *iinfo; + struct au_hinode *p, *q, tmp; + + IiMustWriteLock(inode); + + *update = 0; + sb = inode->i_sb; + type = inode->i_mode & S_IFMT; + iinfo = au_ii(inode); + err = au_ii_realloc(iinfo, au_sbend(sb) + 1); + if (unlikely(err)) + goto out; + + AuDebugOn(iinfo->ii_bstart < 0); + p = iinfo->ii_hinode + iinfo->ii_bstart; + for (bindex = iinfo->ii_bstart; bindex <= iinfo->ii_bend; + bindex++, p++) { + if (!p->hi_inode) + continue; + + AuDebugOn(type != (p->hi_inode->i_mode & S_IFMT)); + new_bindex = au_br_index(sb, p->hi_id); + if (new_bindex == bindex) + continue; + + if (new_bindex < 0) { + *update = 1; + au_hiput(p); + p->hi_inode = NULL; + continue; + } + + if (new_bindex < iinfo->ii_bstart) + iinfo->ii_bstart = new_bindex; + if (iinfo->ii_bend < new_bindex) + iinfo->ii_bend = new_bindex; + /* swap two lower inode, and loop again */ + q = iinfo->ii_hinode + new_bindex; + tmp = *q; + *q = *p; + *p = tmp; + if (tmp.hi_inode) { + bindex--; + p--; + } + } + au_update_ibrange(inode, /*do_put_zero*/0); + e = au_dy_irefresh(inode); + if (unlikely(e && !err)) + err = e; + +out: + AuTraceErr(err); + return err; +} + +void au_refresh_iop(struct inode *inode, int force_getattr) +{ + int type; + struct au_sbinfo *sbi = au_sbi(inode->i_sb); + const struct inode_operations *iop + = force_getattr ? aufs_iop : sbi->si_iop_array; + + if (inode->i_op == iop) + return; + + switch (inode->i_mode & S_IFMT) { + case S_IFDIR: + type = AuIop_DIR; + break; + case S_IFLNK: + type = AuIop_SYMLINK; + break; + default: + type = AuIop_OTHER; + break; + } + + inode->i_op = iop + type; + /* unnecessary smp_wmb() */ +} + +int au_refresh_hinode_self(struct inode *inode) +{ + int err, update; + + err = au_ii_refresh(inode, &update); + if (!err) + au_refresh_hinode_attr(inode, update && S_ISDIR(inode->i_mode)); + + AuTraceErr(err); + return err; +} + +int au_refresh_hinode(struct inode *inode, struct dentry *dentry) +{ + int err, e, update; + unsigned int flags; + umode_t mode; + aufs_bindex_t bindex, bend; + unsigned char isdir; + struct au_hinode *p; + struct au_iinfo *iinfo; + + err = au_ii_refresh(inode, &update); + if (unlikely(err)) + goto out; + + update = 0; + iinfo = au_ii(inode); + p = iinfo->ii_hinode + iinfo->ii_bstart; + mode = (inode->i_mode & S_IFMT); + isdir = S_ISDIR(mode); + flags = au_hi_flags(inode, isdir); + bend = au_dbend(dentry); + for (bindex = au_dbstart(dentry); bindex <= bend; bindex++) { + struct inode *h_i, *h_inode; + struct dentry *h_d; + + h_d = au_h_dptr(dentry, bindex); + if (!h_d || d_is_negative(h_d)) + continue; + + h_inode = d_inode(h_d); + AuDebugOn(mode != (h_inode->i_mode & S_IFMT)); + if (iinfo->ii_bstart <= bindex && bindex <= iinfo->ii_bend) { + h_i = au_h_iptr(inode, bindex); + if (h_i) { + if (h_i == h_inode) + continue; + err = -EIO; + break; + } + } + if (bindex < iinfo->ii_bstart) + iinfo->ii_bstart = bindex; + if (iinfo->ii_bend < bindex) + iinfo->ii_bend = bindex; + au_set_h_iptr(inode, bindex, au_igrab(h_inode), flags); + update = 1; + } + au_update_ibrange(inode, /*do_put_zero*/0); + e = au_dy_irefresh(inode); + if (unlikely(e && !err)) + err = e; + if (!err) + au_refresh_hinode_attr(inode, update && isdir); + +out: + AuTraceErr(err); + return err; +} + +static int set_inode(struct inode *inode, struct dentry *dentry) +{ + int err; + unsigned int flags; + umode_t mode; + aufs_bindex_t bindex, bstart, btail; + unsigned char isdir; + struct dentry *h_dentry; + struct inode *h_inode; + struct au_iinfo *iinfo; + struct inode_operations *iop; + + IiMustWriteLock(inode); + + err = 0; + isdir = 0; + iop = au_sbi(inode->i_sb)->si_iop_array; + bstart = au_dbstart(dentry); + h_dentry = au_h_dptr(dentry, bstart); + h_inode = d_inode(h_dentry); + mode = h_inode->i_mode; + switch (mode & S_IFMT) { + case S_IFREG: + btail = au_dbtail(dentry); + inode->i_op = iop + AuIop_OTHER; + inode->i_fop = &aufs_file_fop; + err = au_dy_iaop(inode, bstart, h_inode); + if (unlikely(err)) + goto out; + break; + case S_IFDIR: + isdir = 1; + btail = au_dbtaildir(dentry); + inode->i_op = iop + AuIop_DIR; + inode->i_fop = &aufs_dir_fop; + break; + case S_IFLNK: + btail = au_dbtail(dentry); + inode->i_op = iop + AuIop_SYMLINK; + break; + case S_IFBLK: + case S_IFCHR: + case S_IFIFO: + case S_IFSOCK: + btail = au_dbtail(dentry); + inode->i_op = iop + AuIop_OTHER; + init_special_inode(inode, mode, h_inode->i_rdev); + break; + default: + AuIOErr("Unknown file type 0%o\n", mode); + err = -EIO; + goto out; + } + + /* do not set hnotify for whiteouted dirs (SHWH mode) */ + flags = au_hi_flags(inode, isdir); + if (au_opt_test(au_mntflags(dentry->d_sb), SHWH) + && au_ftest_hi(flags, HNOTIFY) + && dentry->d_name.len > AUFS_WH_PFX_LEN + && !memcmp(dentry->d_name.name, AUFS_WH_PFX, AUFS_WH_PFX_LEN)) + au_fclr_hi(flags, HNOTIFY); + iinfo = au_ii(inode); + iinfo->ii_bstart = bstart; + iinfo->ii_bend = btail; + for (bindex = bstart; bindex <= btail; bindex++) { + h_dentry = au_h_dptr(dentry, bindex); + if (h_dentry) + au_set_h_iptr(inode, bindex, + au_igrab(d_inode(h_dentry)), flags); + } + au_cpup_attr_all(inode, /*force*/1); + /* + * to force calling aufs_get_acl() every time, + * do not call cache_no_acl() for aufs inode. + */ + +out: + return err; +} + +/* + * successful returns with iinfo write_locked + * minus: errno + * zero: success, matched + * plus: no error, but unmatched + */ +static int reval_inode(struct inode *inode, struct dentry *dentry) +{ + int err; + unsigned int gen, igflags; + aufs_bindex_t bindex, bend; + struct inode *h_inode, *h_dinode; + struct dentry *h_dentry; + + /* + * before this function, if aufs got any iinfo lock, it must be only + * one, the parent dir. + * it can happen by UDBA and the obsoleted inode number. + */ + err = -EIO; + if (unlikely(inode->i_ino == parent_ino(dentry))) + goto out; + + err = 1; + ii_write_lock_new_child(inode); + h_dentry = au_h_dptr(dentry, au_dbstart(dentry)); + h_dinode = d_inode(h_dentry); + bend = au_ibend(inode); + for (bindex = au_ibstart(inode); bindex <= bend; bindex++) { + h_inode = au_h_iptr(inode, bindex); + if (!h_inode || h_inode != h_dinode) + continue; + + err = 0; + gen = au_iigen(inode, &igflags); + if (gen == au_digen(dentry) + && !au_ig_ftest(igflags, HALF_REFRESHED)) + break; + + /* fully refresh inode using dentry */ + err = au_refresh_hinode(inode, dentry); + if (!err) + au_update_iigen(inode, /*half*/0); + break; + } + + if (unlikely(err)) + ii_write_unlock(inode); +out: + return err; +} + +int au_ino(struct super_block *sb, aufs_bindex_t bindex, ino_t h_ino, + unsigned int d_type, ino_t *ino) +{ + int err; + struct mutex *mtx; + + /* prevent hardlinked inode number from race condition */ + mtx = NULL; + if (d_type != DT_DIR) { + mtx = &au_sbr(sb, bindex)->br_xino.xi_nondir_mtx; + mutex_lock(mtx); + } + err = au_xino_read(sb, bindex, h_ino, ino); + if (unlikely(err)) + goto out; + + if (!*ino) { + err = -EIO; + *ino = au_xino_new_ino(sb); + if (unlikely(!*ino)) + goto out; + err = au_xino_write(sb, bindex, h_ino, *ino); + if (unlikely(err)) + goto out; + } + +out: + if (mtx) + mutex_unlock(mtx); + return err; +} + +/* successful returns with iinfo write_locked */ +/* todo: return with unlocked? */ +struct inode *au_new_inode(struct dentry *dentry, int must_new) +{ + struct inode *inode, *h_inode; + struct dentry *h_dentry; + struct super_block *sb; + struct mutex *mtx; + ino_t h_ino, ino; + int err; + aufs_bindex_t bstart; + + sb = dentry->d_sb; + bstart = au_dbstart(dentry); + h_dentry = au_h_dptr(dentry, bstart); + h_inode = d_inode(h_dentry); + h_ino = h_inode->i_ino; + + /* + * stop 'race'-ing between hardlinks under different + * parents. + */ + mtx = NULL; + if (!d_is_dir(h_dentry)) + mtx = &au_sbr(sb, bstart)->br_xino.xi_nondir_mtx; + +new_ino: + if (mtx) + mutex_lock(mtx); + err = au_xino_read(sb, bstart, h_ino, &ino); + inode = ERR_PTR(err); + if (unlikely(err)) + goto out; + + if (!ino) { + ino = au_xino_new_ino(sb); + if (unlikely(!ino)) { + inode = ERR_PTR(-EIO); + goto out; + } + } + + AuDbg("i%lu\n", (unsigned long)ino); + inode = au_iget_locked(sb, ino); + err = PTR_ERR(inode); + if (IS_ERR(inode)) + goto out; + + AuDbg("%lx, new %d\n", inode->i_state, !!(inode->i_state & I_NEW)); + if (inode->i_state & I_NEW) { + /* verbose coding for lock class name */ + if (unlikely(d_is_symlink(h_dentry))) + au_rw_class(&au_ii(inode)->ii_rwsem, + au_lc_key + AuLcSymlink_IIINFO); + else if (unlikely(d_is_dir(h_dentry))) + au_rw_class(&au_ii(inode)->ii_rwsem, + au_lc_key + AuLcDir_IIINFO); + else /* likely */ + au_rw_class(&au_ii(inode)->ii_rwsem, + au_lc_key + AuLcNonDir_IIINFO); + + ii_write_lock_new_child(inode); + err = set_inode(inode, dentry); + if (!err) { + unlock_new_inode(inode); + goto out; /* success */ + } + + /* + * iget_failed() calls iput(), but we need to call + * ii_write_unlock() after iget_failed(). so dirty hack for + * i_count. + */ + atomic_inc(&inode->i_count); + iget_failed(inode); + ii_write_unlock(inode); + au_xino_write(sb, bstart, h_ino, /*ino*/0); + /* ignore this error */ + goto out_iput; + } else if (!must_new && !IS_DEADDIR(inode) && inode->i_nlink) { + /* + * horrible race condition between lookup, readdir and copyup + * (or something). + */ + if (mtx) + mutex_unlock(mtx); + err = reval_inode(inode, dentry); + if (unlikely(err < 0)) { + mtx = NULL; + goto out_iput; + } + + if (!err) { + mtx = NULL; + goto out; /* success */ + } else if (mtx) + mutex_lock(mtx); + } + + if (unlikely(au_test_fs_unique_ino(h_inode))) + AuWarn1("Warning: Un-notified UDBA or repeatedly renamed dir," + " b%d, %s, %pd, hi%lu, i%lu.\n", + bstart, au_sbtype(h_dentry->d_sb), dentry, + (unsigned long)h_ino, (unsigned long)ino); + ino = 0; + err = au_xino_write(sb, bstart, h_ino, /*ino*/0); + if (!err) { + iput(inode); + if (mtx) + mutex_unlock(mtx); + goto new_ino; + } + +out_iput: + iput(inode); + inode = ERR_PTR(err); +out: + if (mtx) + mutex_unlock(mtx); + return inode; +} + +/* ---------------------------------------------------------------------- */ + +int au_test_ro(struct super_block *sb, aufs_bindex_t bindex, + struct inode *inode) +{ + int err; + struct inode *hi; + + err = au_br_rdonly(au_sbr(sb, bindex)); + + /* pseudo-link after flushed may happen out of bounds */ + if (!err + && inode + && au_ibstart(inode) <= bindex + && bindex <= au_ibend(inode)) { + /* + * permission check is unnecessary since vfsub routine + * will be called later + */ + hi = au_h_iptr(inode, bindex); + if (hi) + err = IS_IMMUTABLE(hi) ? -EROFS : 0; + } + + return err; +} + +int au_test_h_perm(struct inode *h_inode, int mask) +{ + if (uid_eq(current_fsuid(), GLOBAL_ROOT_UID)) + return 0; + return inode_permission(h_inode, mask); +} + +int au_test_h_perm_sio(struct inode *h_inode, int mask) +{ + if (au_test_nfs(h_inode->i_sb) + && (mask & MAY_WRITE) + && S_ISDIR(h_inode->i_mode)) + mask |= MAY_READ; /* force permission check */ + return au_test_h_perm(h_inode, mask); +} diff --git a/fs/aufs/inode.h b/fs/aufs/inode.h new file mode 100644 index 000000000..eeb43d220 --- /dev/null +++ b/fs/aufs/inode.h @@ -0,0 +1,672 @@ +/* + * Copyright (C) 2005-2016 Junjiro R. Okajima + */ + +/* + * inode operations + */ + +#ifndef __AUFS_INODE_H__ +#define __AUFS_INODE_H__ + +#ifdef __KERNEL__ + +#include <linux/fsnotify.h> +#include "rwsem.h" + +struct vfsmount; + +struct au_hnotify { +#ifdef CONFIG_AUFS_HNOTIFY +#ifdef CONFIG_AUFS_HFSNOTIFY + /* never use fsnotify_add_vfsmount_mark() */ + struct fsnotify_mark hn_mark; +#endif + struct inode *hn_aufs_inode; /* no get/put */ +#endif +} ____cacheline_aligned_in_smp; + +struct au_hinode { + struct inode *hi_inode; + aufs_bindex_t hi_id; +#ifdef CONFIG_AUFS_HNOTIFY + struct au_hnotify *hi_notify; +#endif + + /* reference to the copied-up whiteout with get/put */ + struct dentry *hi_whdentry; +}; + +/* ig_flags */ +#define AuIG_HALF_REFRESHED 1 +#define au_ig_ftest(flags, name) ((flags) & AuIG_##name) +#define au_ig_fset(flags, name) \ + do { (flags) |= AuIG_##name; } while (0) +#define au_ig_fclr(flags, name) \ + do { (flags) &= ~AuIG_##name; } while (0) + +struct au_iigen { + spinlock_t ig_spin; + __u32 ig_generation, ig_flags; +}; + +struct au_vdir; +struct au_iinfo { + struct au_iigen ii_generation; + struct super_block *ii_hsb1; /* no get/put */ + + struct au_rwsem ii_rwsem; + aufs_bindex_t ii_bstart, ii_bend; + __u32 ii_higen; + struct au_hinode *ii_hinode; + struct au_vdir *ii_vdir; +}; + +struct au_icntnr { + struct au_iinfo iinfo; + struct inode vfs_inode; +} ____cacheline_aligned_in_smp; + +/* au_pin flags */ +#define AuPin_DI_LOCKED 1 +#define AuPin_MNT_WRITE (1 << 1) +#define au_ftest_pin(flags, name) ((flags) & AuPin_##name) +#define au_fset_pin(flags, name) \ + do { (flags) |= AuPin_##name; } while (0) +#define au_fclr_pin(flags, name) \ + do { (flags) &= ~AuPin_##name; } while (0) + +struct au_pin { + /* input */ + struct dentry *dentry; + unsigned int udba; + unsigned char lsc_di, lsc_hi, flags; + aufs_bindex_t bindex; + + /* output */ + struct dentry *parent; + struct au_hinode *hdir; + struct vfsmount *h_mnt; + + /* temporary unlock/relock for copyup */ + struct dentry *h_dentry, *h_parent; + struct au_branch *br; + struct task_struct *task; +}; + +void au_pin_hdir_unlock(struct au_pin *p); +int au_pin_hdir_lock(struct au_pin *p); +int au_pin_hdir_relock(struct au_pin *p); +void au_pin_hdir_set_owner(struct au_pin *p, struct task_struct *task); +void au_pin_hdir_acquire_nest(struct au_pin *p); +void au_pin_hdir_release(struct au_pin *p); + +/* ---------------------------------------------------------------------- */ + +static inline struct au_iinfo *au_ii(struct inode *inode) +{ + struct au_iinfo *iinfo; + + iinfo = &(container_of(inode, struct au_icntnr, vfs_inode)->iinfo); + if (iinfo->ii_hinode) + return iinfo; + return NULL; /* debugging bad_inode case */ +} + +/* ---------------------------------------------------------------------- */ + +/* inode.c */ +struct inode *au_igrab(struct inode *inode); +void au_refresh_iop(struct inode *inode, int force_getattr); +int au_refresh_hinode_self(struct inode *inode); +int au_refresh_hinode(struct inode *inode, struct dentry *dentry); +int au_ino(struct super_block *sb, aufs_bindex_t bindex, ino_t h_ino, + unsigned int d_type, ino_t *ino); +struct inode *au_new_inode(struct dentry *dentry, int must_new); +int au_test_ro(struct super_block *sb, aufs_bindex_t bindex, + struct inode *inode); +int au_test_h_perm(struct inode *h_inode, int mask); +int au_test_h_perm_sio(struct inode *h_inode, int mask); + +static inline int au_wh_ino(struct super_block *sb, aufs_bindex_t bindex, + ino_t h_ino, unsigned int d_type, ino_t *ino) +{ +#ifdef CONFIG_AUFS_SHWH + return au_ino(sb, bindex, h_ino, d_type, ino); +#else + return 0; +#endif +} + +/* i_op.c */ +enum { + AuIop_SYMLINK, + AuIop_DIR, + AuIop_OTHER, + AuIop_Last +}; +extern struct inode_operations aufs_iop[AuIop_Last], + aufs_iop_nogetattr[AuIop_Last]; + +/* au_wr_dir flags */ +#define AuWrDir_ADD_ENTRY 1 +#define AuWrDir_ISDIR (1 << 1) +#define AuWrDir_TMPFILE (1 << 2) +#define au_ftest_wrdir(flags, name) ((flags) & AuWrDir_##name) +#define au_fset_wrdir(flags, name) \ + do { (flags) |= AuWrDir_##name; } while (0) +#define au_fclr_wrdir(flags, name) \ + do { (flags) &= ~AuWrDir_##name; } while (0) + +struct au_wr_dir_args { + aufs_bindex_t force_btgt; + unsigned char flags; +}; +int au_wr_dir(struct dentry *dentry, struct dentry *src_dentry, + struct au_wr_dir_args *args); + +struct dentry *au_pinned_h_parent(struct au_pin *pin); +void au_pin_init(struct au_pin *pin, struct dentry *dentry, + aufs_bindex_t bindex, int lsc_di, int lsc_hi, + unsigned int udba, unsigned char flags); +int au_pin(struct au_pin *pin, struct dentry *dentry, aufs_bindex_t bindex, + unsigned int udba, unsigned char flags) __must_check; +int au_do_pin(struct au_pin *pin) __must_check; +void au_unpin(struct au_pin *pin); +int au_reval_for_attr(struct dentry *dentry, unsigned int sigen); + +#define AuIcpup_DID_CPUP 1 +#define au_ftest_icpup(flags, name) ((flags) & AuIcpup_##name) +#define au_fset_icpup(flags, name) \ + do { (flags) |= AuIcpup_##name; } while (0) +#define au_fclr_icpup(flags, name) \ + do { (flags) &= ~AuIcpup_##name; } while (0) + +struct au_icpup_args { + unsigned char flags; + unsigned char pin_flags; + aufs_bindex_t btgt; + unsigned int udba; + struct au_pin pin; + struct path h_path; + struct inode *h_inode; +}; + +int au_pin_and_icpup(struct dentry *dentry, struct iattr *ia, + struct au_icpup_args *a); + +int au_h_path_getattr(struct dentry *dentry, int force, struct path *h_path); + +/* i_op_add.c */ +int au_may_add(struct dentry *dentry, aufs_bindex_t bindex, + struct dentry *h_parent, int isdir); +int aufs_mknod(struct inode *dir, struct dentry *dentry, umode_t mode, + dev_t dev); +int aufs_symlink(struct inode *dir, struct dentry *dentry, const char *symname); +int aufs_create(struct inode *dir, struct dentry *dentry, umode_t mode, + bool want_excl); +struct vfsub_aopen_args; +int au_aopen_or_create(struct inode *dir, struct dentry *dentry, + struct vfsub_aopen_args *args); +int aufs_tmpfile(struct inode *dir, struct dentry *dentry, umode_t mode); +int aufs_link(struct dentry *src_dentry, struct inode *dir, + struct dentry *dentry); +int aufs_mkdir(struct inode *dir, struct dentry *dentry, umode_t mode); + +/* i_op_del.c */ +int au_wr_dir_need_wh(struct dentry *dentry, int isdir, aufs_bindex_t *bcpup); +int au_may_del(struct dentry *dentry, aufs_bindex_t bindex, + struct dentry *h_parent, int isdir); +int aufs_unlink(struct inode *dir, struct dentry *dentry); +int aufs_rmdir(struct inode *dir, struct dentry *dentry); + +/* i_op_ren.c */ +int au_wbr(struct dentry *dentry, aufs_bindex_t btgt); +int aufs_rename(struct inode *src_dir, struct dentry *src_dentry, + struct inode *dir, struct dentry *dentry); + +/* iinfo.c */ +struct inode *au_h_iptr(struct inode *inode, aufs_bindex_t bindex); +void au_hiput(struct au_hinode *hinode); +void au_set_hi_wh(struct inode *inode, aufs_bindex_t bindex, + struct dentry *h_wh); +unsigned int au_hi_flags(struct inode *inode, int isdir); + +/* hinode flags */ +#define AuHi_XINO 1 +#define AuHi_HNOTIFY (1 << 1) +#define au_ftest_hi(flags, name) ((flags) & AuHi_##name) +#define au_fset_hi(flags, name) \ + do { (flags) |= AuHi_##name; } while (0) +#define au_fclr_hi(flags, name) \ + do { (flags) &= ~AuHi_##name; } while (0) + +#ifndef CONFIG_AUFS_HNOTIFY +#undef AuHi_HNOTIFY +#define AuHi_HNOTIFY 0 +#endif + +void au_set_h_iptr(struct inode *inode, aufs_bindex_t bindex, + struct inode *h_inode, unsigned int flags); + +void au_update_iigen(struct inode *inode, int half); +void au_update_ibrange(struct inode *inode, int do_put_zero); + +void au_icntnr_init_once(void *_c); +int au_iinfo_init(struct inode *inode); +void au_iinfo_fin(struct inode *inode); +int au_ii_realloc(struct au_iinfo *iinfo, int nbr); + +#ifdef CONFIG_PROC_FS +/* plink.c */ +int au_plink_maint(struct super_block *sb, int flags); +struct au_sbinfo; +void au_plink_maint_leave(struct au_sbinfo *sbinfo); +int au_plink_maint_enter(struct super_block *sb); +#ifdef CONFIG_AUFS_DEBUG +void au_plink_list(struct super_block *sb); +#else +AuStubVoid(au_plink_list, struct super_block *sb) +#endif +int au_plink_test(struct inode *inode); +struct dentry *au_plink_lkup(struct inode *inode, aufs_bindex_t bindex); +void au_plink_append(struct inode *inode, aufs_bindex_t bindex, + struct dentry *h_dentry); +void au_plink_put(struct super_block *sb, int verbose); +void au_plink_clean(struct super_block *sb, int verbose); +void au_plink_half_refresh(struct super_block *sb, aufs_bindex_t br_id); +#else +AuStubInt0(au_plink_maint, struct super_block *sb, int flags); +AuStubVoid(au_plink_maint_leave, struct au_sbinfo *sbinfo); +AuStubInt0(au_plink_maint_enter, struct super_block *sb); +AuStubVoid(au_plink_list, struct super_block *sb); +AuStubInt0(au_plink_test, struct inode *inode); +AuStub(struct dentry *, au_plink_lkup, return NULL, + struct inode *inode, aufs_bindex_t bindex); +AuStubVoid(au_plink_append, struct inode *inode, aufs_bindex_t bindex, + struct dentry *h_dentry); +AuStubVoid(au_plink_put, struct super_block *sb, int verbose); +AuStubVoid(au_plink_clean, struct super_block *sb, int verbose); +AuStubVoid(au_plink_half_refresh, struct super_block *sb, aufs_bindex_t br_id); +#endif /* CONFIG_PROC_FS */ + +#ifdef CONFIG_AUFS_XATTR +/* xattr.c */ +int au_cpup_xattr(struct dentry *h_dst, struct dentry *h_src, int ignore_flags, + unsigned int verbose); +ssize_t aufs_listxattr(struct dentry *dentry, char *list, size_t size); +ssize_t aufs_getxattr(struct dentry *dentry, const char *name, void *value, + size_t size); +int aufs_setxattr(struct dentry *dentry, const char *name, const void *value, + size_t size, int flags); +int aufs_removexattr(struct dentry *dentry, const char *name); + +/* void au_xattr_init(struct super_block *sb); */ +#else +AuStubInt0(au_cpup_xattr, struct dentry *h_dst, struct dentry *h_src, + int ignore_flags, unsigned int verbose); +/* AuStubVoid(au_xattr_init, struct super_block *sb); */ +#endif + +#ifdef CONFIG_FS_POSIX_ACL +struct posix_acl *aufs_get_acl(struct inode *inode, int type); +int aufs_set_acl(struct inode *inode, struct posix_acl *acl, int type); +#endif + +#if IS_ENABLED(CONFIG_AUFS_XATTR) || IS_ENABLED(CONFIG_FS_POSIX_ACL) +enum { + AU_XATTR_SET, + AU_XATTR_REMOVE, + AU_ACL_SET +}; + +struct au_srxattr { + int type; + union { + struct { + const char *name; + const void *value; + size_t size; + int flags; + } set; + struct { + const char *name; + } remove; + struct { + struct posix_acl *acl; + int type; + } acl_set; + } u; +}; +ssize_t au_srxattr(struct dentry *dentry, struct au_srxattr *arg); +#endif + +/* ---------------------------------------------------------------------- */ + +/* lock subclass for iinfo */ +enum { + AuLsc_II_CHILD, /* child first */ + AuLsc_II_CHILD2, /* rename(2), link(2), and cpup at hnotify */ + AuLsc_II_CHILD3, /* copyup dirs */ + AuLsc_II_PARENT, /* see AuLsc_I_PARENT in vfsub.h */ + AuLsc_II_PARENT2, + AuLsc_II_PARENT3, /* copyup dirs */ + AuLsc_II_NEW_CHILD +}; + +/* + * ii_read_lock_child, ii_write_lock_child, + * ii_read_lock_child2, ii_write_lock_child2, + * ii_read_lock_child3, ii_write_lock_child3, + * ii_read_lock_parent, ii_write_lock_parent, + * ii_read_lock_parent2, ii_write_lock_parent2, + * ii_read_lock_parent3, ii_write_lock_parent3, + * ii_read_lock_new_child, ii_write_lock_new_child, + */ +#define AuReadLockFunc(name, lsc) \ +static inline void ii_read_lock_##name(struct inode *i) \ +{ \ + au_rw_read_lock_nested(&au_ii(i)->ii_rwsem, AuLsc_II_##lsc); \ +} + +#define AuWriteLockFunc(name, lsc) \ +static inline void ii_write_lock_##name(struct inode *i) \ +{ \ + au_rw_write_lock_nested(&au_ii(i)->ii_rwsem, AuLsc_II_##lsc); \ +} + +#define AuRWLockFuncs(name, lsc) \ + AuReadLockFunc(name, lsc) \ + AuWriteLockFunc(name, lsc) + +AuRWLockFuncs(child, CHILD); +AuRWLockFuncs(child2, CHILD2); +AuRWLockFuncs(child3, CHILD3); +AuRWLockFuncs(parent, PARENT); +AuRWLockFuncs(parent2, PARENT2); +AuRWLockFuncs(parent3, PARENT3); +AuRWLockFuncs(new_child, NEW_CHILD); + +#undef AuReadLockFunc +#undef AuWriteLockFunc +#undef AuRWLockFuncs + +/* + * ii_read_unlock, ii_write_unlock, ii_downgrade_lock + */ +AuSimpleUnlockRwsemFuncs(ii, struct inode *i, &au_ii(i)->ii_rwsem); + +#define IiMustNoWaiters(i) AuRwMustNoWaiters(&au_ii(i)->ii_rwsem) +#define IiMustAnyLock(i) AuRwMustAnyLock(&au_ii(i)->ii_rwsem) +#define IiMustWriteLock(i) AuRwMustWriteLock(&au_ii(i)->ii_rwsem) + +/* ---------------------------------------------------------------------- */ + +static inline void au_icntnr_init(struct au_icntnr *c) +{ +#ifdef CONFIG_AUFS_DEBUG + c->vfs_inode.i_mode = 0; +#endif +} + +static inline unsigned int au_iigen(struct inode *inode, unsigned int *igflags) +{ + unsigned int gen; + struct au_iinfo *iinfo; + struct au_iigen *iigen; + + iinfo = au_ii(inode); + iigen = &iinfo->ii_generation; + spin_lock(&iigen->ig_spin); + if (igflags) + *igflags = iigen->ig_flags; + gen = iigen->ig_generation; + spin_unlock(&iigen->ig_spin); + + return gen; +} + +/* tiny test for inode number */ +/* tmpfs generation is too rough */ +static inline int au_test_higen(struct inode *inode, struct inode *h_inode) +{ + struct au_iinfo *iinfo; + + iinfo = au_ii(inode); + AuRwMustAnyLock(&iinfo->ii_rwsem); + return !(iinfo->ii_hsb1 == h_inode->i_sb + && iinfo->ii_higen == h_inode->i_generation); +} + +static inline void au_iigen_dec(struct inode *inode) +{ + struct au_iinfo *iinfo; + struct au_iigen *iigen; + + iinfo = au_ii(inode); + iigen = &iinfo->ii_generation; + spin_lock(&iigen->ig_spin); + iigen->ig_generation--; + spin_unlock(&iigen->ig_spin); +} + +static inline int au_iigen_test(struct inode *inode, unsigned int sigen) +{ + int err; + + err = 0; + if (unlikely(inode && au_iigen(inode, NULL) != sigen)) + err = -EIO; + + return err; +} + +/* ---------------------------------------------------------------------- */ + +static inline aufs_bindex_t au_ii_br_id(struct inode *inode, + aufs_bindex_t bindex) +{ + IiMustAnyLock(inode); + return au_ii(inode)->ii_hinode[0 + bindex].hi_id; +} + +static inline aufs_bindex_t au_ibstart(struct inode *inode) +{ + IiMustAnyLock(inode); + return au_ii(inode)->ii_bstart; +} + +static inline aufs_bindex_t au_ibend(struct inode *inode) +{ + IiMustAnyLock(inode); + return au_ii(inode)->ii_bend; +} + +static inline struct au_vdir *au_ivdir(struct inode *inode) +{ + IiMustAnyLock(inode); + return au_ii(inode)->ii_vdir; +} + +static inline struct dentry *au_hi_wh(struct inode *inode, aufs_bindex_t bindex) +{ + IiMustAnyLock(inode); + return au_ii(inode)->ii_hinode[0 + bindex].hi_whdentry; +} + +static inline void au_set_ibstart(struct inode *inode, aufs_bindex_t bindex) +{ + IiMustWriteLock(inode); + au_ii(inode)->ii_bstart = bindex; +} + +static inline void au_set_ibend(struct inode *inode, aufs_bindex_t bindex) +{ + IiMustWriteLock(inode); + au_ii(inode)->ii_bend = bindex; +} + +static inline void au_set_ivdir(struct inode *inode, struct au_vdir *vdir) +{ + IiMustWriteLock(inode); + au_ii(inode)->ii_vdir = vdir; +} + +static inline struct au_hinode *au_hi(struct inode *inode, aufs_bindex_t bindex) +{ + IiMustAnyLock(inode); + return au_ii(inode)->ii_hinode + bindex; +} + +/* ---------------------------------------------------------------------- */ + +static inline struct dentry *au_pinned_parent(struct au_pin *pin) +{ + if (pin) + return pin->parent; + return NULL; +} + +static inline struct inode *au_pinned_h_dir(struct au_pin *pin) +{ + if (pin && pin->hdir) + return pin->hdir->hi_inode; + return NULL; +} + +static inline struct au_hinode *au_pinned_hdir(struct au_pin *pin) +{ + if (pin) + return pin->hdir; + return NULL; +} + +static inline void au_pin_set_dentry(struct au_pin *pin, struct dentry *dentry) +{ + if (pin) + pin->dentry = dentry; +} + +static inline void au_pin_set_parent_lflag(struct au_pin *pin, + unsigned char lflag) +{ + if (pin) { + if (lflag) + au_fset_pin(pin->flags, DI_LOCKED); + else + au_fclr_pin(pin->flags, DI_LOCKED); + } +} + +#if 0 /* reserved */ +static inline void au_pin_set_parent(struct au_pin *pin, struct dentry *parent) +{ + if (pin) { + dput(pin->parent); + pin->parent = dget(parent); + } +} +#endif + +/* ---------------------------------------------------------------------- */ + +struct au_branch; +#ifdef CONFIG_AUFS_HNOTIFY +struct au_hnotify_op { + void (*ctl)(struct au_hinode *hinode, int do_set); + int (*alloc)(struct au_hinode *hinode); + + /* + * if it returns true, the the caller should free hinode->hi_notify, + * otherwise ->free() frees it. + */ + int (*free)(struct au_hinode *hinode, + struct au_hnotify *hn) __must_check; + + void (*fin)(void); + int (*init)(void); + + int (*reset_br)(unsigned int udba, struct au_branch *br, int perm); + void (*fin_br)(struct au_branch *br); + int (*init_br)(struct au_branch *br, int perm); +}; + +/* hnotify.c */ +int au_hn_alloc(struct au_hinode *hinode, struct inode *inode); +void au_hn_free(struct au_hinode *hinode); +void au_hn_ctl(struct au_hinode *hinode, int do_set); +void au_hn_reset(struct inode *inode, unsigned int flags); +int au_hnotify(struct inode *h_dir, struct au_hnotify *hnotify, u32 mask, + struct qstr *h_child_qstr, struct inode *h_child_inode); +int au_hnotify_reset_br(unsigned int udba, struct au_branch *br, int perm); +int au_hnotify_init_br(struct au_branch *br, int perm); +void au_hnotify_fin_br(struct au_branch *br); +int __init au_hnotify_init(void); +void au_hnotify_fin(void); + +/* hfsnotify.c */ +extern const struct au_hnotify_op au_hnotify_op; + +static inline +void au_hn_init(struct au_hinode *hinode) +{ + hinode->hi_notify = NULL; +} + +static inline struct au_hnotify *au_hn(struct au_hinode *hinode) +{ + return hinode->hi_notify; +} + +#else +AuStub(int, au_hn_alloc, return -EOPNOTSUPP, + struct au_hinode *hinode __maybe_unused, + struct inode *inode __maybe_unused) +AuStub(struct au_hnotify *, au_hn, return NULL, struct au_hinode *hinode) +AuStubVoid(au_hn_free, struct au_hinode *hinode __maybe_unused) +AuStubVoid(au_hn_ctl, struct au_hinode *hinode __maybe_unused, + int do_set __maybe_unused) +AuStubVoid(au_hn_reset, struct inode *inode __maybe_unused, + unsigned int flags __maybe_unused) +AuStubInt0(au_hnotify_reset_br, unsigned int udba __maybe_unused, + struct au_branch *br __maybe_unused, + int perm __maybe_unused) +AuStubInt0(au_hnotify_init_br, struct au_branch *br __maybe_unused, + int perm __maybe_unused) +AuStubVoid(au_hnotify_fin_br, struct au_branch *br __maybe_unused) +AuStubInt0(__init au_hnotify_init, void) +AuStubVoid(au_hnotify_fin, void) +AuStubVoid(au_hn_init, struct au_hinode *hinode __maybe_unused) +#endif /* CONFIG_AUFS_HNOTIFY */ + +static inline void au_hn_suspend(struct au_hinode *hdir) +{ + au_hn_ctl(hdir, /*do_set*/0); +} + +static inline void au_hn_resume(struct au_hinode *hdir) +{ + au_hn_ctl(hdir, /*do_set*/1); +} + +static inline void au_hn_imtx_lock(struct au_hinode *hdir) +{ + inode_lock(hdir->hi_inode); + au_hn_suspend(hdir); +} + +static inline void au_hn_imtx_lock_nested(struct au_hinode *hdir, + unsigned int sc __maybe_unused) +{ + inode_lock_nested(hdir->hi_inode, sc); + au_hn_suspend(hdir); +} + +static inline void au_hn_imtx_unlock(struct au_hinode *hdir) +{ + au_hn_resume(hdir); + inode_unlock(hdir->hi_inode); +} + +#endif /* __KERNEL__ */ +#endif /* __AUFS_INODE_H__ */ diff --git a/fs/aufs/ioctl.c b/fs/aufs/ioctl.c new file mode 100644 index 000000000..6528fb911 --- /dev/null +++ b/fs/aufs/ioctl.c @@ -0,0 +1,206 @@ +/* + * Copyright (C) 2005-2016 Junjiro R. Okajima + */ + +/* + * ioctl + * plink-management and readdir in userspace. + * assist the pathconf(3) wrapper library. + * move-down + * File-based Hierarchical Storage Management. + */ + +#include <linux/compat.h> +#include <linux/file.h> +#include "aufs.h" + +static int au_wbr_fd(struct path *path, struct aufs_wbr_fd __user *arg) +{ + int err, fd; + aufs_bindex_t wbi, bindex, bend; + struct file *h_file; + struct super_block *sb; + struct dentry *root; + struct au_branch *br; + struct aufs_wbr_fd wbrfd = { + .oflags = au_dir_roflags, + .brid = -1 + }; + const int valid = O_RDONLY | O_NONBLOCK | O_LARGEFILE | O_DIRECTORY + | O_NOATIME | O_CLOEXEC; + + AuDebugOn(wbrfd.oflags & ~valid); + + if (arg) { + err = copy_from_user(&wbrfd, arg, sizeof(wbrfd)); + if (unlikely(err)) { + err = -EFAULT; + goto out; + } + + err = -EINVAL; + AuDbg("wbrfd{0%o, %d}\n", wbrfd.oflags, wbrfd.brid); + wbrfd.oflags |= au_dir_roflags; + AuDbg("0%o\n", wbrfd.oflags); + if (unlikely(wbrfd.oflags & ~valid)) + goto out; + } + + fd = get_unused_fd_flags(0); + err = fd; + if (unlikely(fd < 0)) + goto out; + + h_file = ERR_PTR(-EINVAL); + wbi = 0; + br = NULL; + sb = path->dentry->d_sb; + root = sb->s_root; + aufs_read_lock(root, AuLock_IR); + bend = au_sbend(sb); + if (wbrfd.brid >= 0) { + wbi = au_br_index(sb, wbrfd.brid); + if (unlikely(wbi < 0 || wbi > bend)) + goto out_unlock; + } + + h_file = ERR_PTR(-ENOENT); + br = au_sbr(sb, wbi); + if (!au_br_writable(br->br_perm)) { + if (arg) + goto out_unlock; + + bindex = wbi + 1; + wbi = -1; + for (; bindex <= bend; bindex++) { + br = au_sbr(sb, bindex); + if (au_br_writable(br->br_perm)) { + wbi = bindex; + br = au_sbr(sb, wbi); + break; + } + } + } + AuDbg("wbi %d\n", wbi); + if (wbi >= 0) + h_file = au_h_open(root, wbi, wbrfd.oflags, NULL, + /*force_wr*/0); + +out_unlock: + aufs_read_unlock(root, AuLock_IR); + err = PTR_ERR(h_file); + if (IS_ERR(h_file)) + goto out_fd; + + atomic_dec(&br->br_count); /* cf. au_h_open() */ + fd_install(fd, h_file); + err = fd; + goto out; /* success */ + +out_fd: + put_unused_fd(fd); +out: + AuTraceErr(err); + return err; +} + +/* ---------------------------------------------------------------------- */ + +long aufs_ioctl_dir(struct file *file, unsigned int cmd, unsigned long arg) +{ + long err; + struct dentry *dentry; + + switch (cmd) { + case AUFS_CTL_RDU: + case AUFS_CTL_RDU_INO: + err = au_rdu_ioctl(file, cmd, arg); + break; + + case AUFS_CTL_WBR_FD: + err = au_wbr_fd(&file->f_path, (void __user *)arg); + break; + + case AUFS_CTL_IBUSY: + err = au_ibusy_ioctl(file, arg); + break; + + case AUFS_CTL_BRINFO: + err = au_brinfo_ioctl(file, arg); + break; + + case AUFS_CTL_FHSM_FD: + dentry = file->f_path.dentry; + if (IS_ROOT(dentry)) + err = au_fhsm_fd(dentry->d_sb, arg); + else + err = -ENOTTY; + break; + + default: + /* do not call the lower */ + AuDbg("0x%x\n", cmd); + err = -ENOTTY; + } + + AuTraceErr(err); + return err; +} + +long aufs_ioctl_nondir(struct file *file, unsigned int cmd, unsigned long arg) +{ + long err; + + switch (cmd) { + case AUFS_CTL_MVDOWN: + err = au_mvdown(file->f_path.dentry, (void __user *)arg); + break; + + case AUFS_CTL_WBR_FD: + err = au_wbr_fd(&file->f_path, (void __user *)arg); + break; + + default: + /* do not call the lower */ + AuDbg("0x%x\n", cmd); + err = -ENOTTY; + } + + AuTraceErr(err); + return err; +} + +#ifdef CONFIG_COMPAT +long aufs_compat_ioctl_dir(struct file *file, unsigned int cmd, + unsigned long arg) +{ + long err; + + switch (cmd) { + case AUFS_CTL_RDU: + case AUFS_CTL_RDU_INO: + err = au_rdu_compat_ioctl(file, cmd, arg); + break; + + case AUFS_CTL_IBUSY: + err = au_ibusy_compat_ioctl(file, arg); + break; + + case AUFS_CTL_BRINFO: + err = au_brinfo_compat_ioctl(file, arg); + break; + + default: + err = aufs_ioctl_dir(file, cmd, arg); + } + + AuTraceErr(err); + return err; +} + +long aufs_compat_ioctl_nondir(struct file *file, unsigned int cmd, + unsigned long arg) +{ + return aufs_ioctl_nondir(file, cmd, (unsigned long)compat_ptr(arg)); +} +#endif diff --git a/fs/aufs/loop.c b/fs/aufs/loop.c new file mode 100644 index 000000000..5711e7a2f --- /dev/null +++ b/fs/aufs/loop.c @@ -0,0 +1,133 @@ +/* + * Copyright (C) 2005-2016 Junjiro R. Okajima + */ + +/* + * support for loopback block device as a branch + */ + +#include "aufs.h" + +/* added into drivers/block/loop.c */ +static struct file *(*backing_file_func)(struct super_block *sb); + +/* + * test if two lower dentries have overlapping branches. + */ +int au_test_loopback_overlap(struct super_block *sb, struct dentry *h_adding) +{ + struct super_block *h_sb; + struct file *backing_file; + + if (unlikely(!backing_file_func)) { + /* don't load "loop" module here */ + backing_file_func = symbol_get(loop_backing_file); + if (unlikely(!backing_file_func)) + /* "loop" module is not loaded */ + return 0; + } + + h_sb = h_adding->d_sb; + backing_file = backing_file_func(h_sb); + if (!backing_file) + return 0; + + h_adding = backing_file->f_path.dentry; + /* + * h_adding can be local NFS. + * in this case aufs cannot detect the loop. + */ + if (unlikely(h_adding->d_sb == sb)) + return 1; + return !!au_test_subdir(h_adding, sb->s_root); +} + +/* true if a kernel thread named 'loop[0-9].*' accesses a file */ +int au_test_loopback_kthread(void) +{ + int ret; + struct task_struct *tsk = current; + char c, comm[sizeof(tsk->comm)]; + + ret = 0; + if (tsk->flags & PF_KTHREAD) { + get_task_comm(comm, tsk); + c = comm[4]; + ret = ('0' <= c && c <= '9' + && !strncmp(comm, "loop", 4)); + } + + return ret; +} + +/* ---------------------------------------------------------------------- */ + +#define au_warn_loopback_step 16 +static int au_warn_loopback_nelem = au_warn_loopback_step; +static unsigned long *au_warn_loopback_array; + +void au_warn_loopback(struct super_block *h_sb) +{ + int i, new_nelem; + unsigned long *a, magic; + static DEFINE_SPINLOCK(spin); + + magic = h_sb->s_magic; + spin_lock(&spin); + a = au_warn_loopback_array; + for (i = 0; i < au_warn_loopback_nelem && *a; i++) + if (a[i] == magic) { + spin_unlock(&spin); + return; + } + + /* h_sb is new to us, print it */ + if (i < au_warn_loopback_nelem) { + a[i] = magic; + goto pr; + } + + /* expand the array */ + new_nelem = au_warn_loopback_nelem + au_warn_loopback_step; + a = au_kzrealloc(au_warn_loopback_array, + au_warn_loopback_nelem * sizeof(unsigned long), + new_nelem * sizeof(unsigned long), GFP_ATOMIC); + if (a) { + au_warn_loopback_nelem = new_nelem; + au_warn_loopback_array = a; + a[i] = magic; + goto pr; + } + + spin_unlock(&spin); + AuWarn1("realloc failed, ignored\n"); + return; + +pr: + spin_unlock(&spin); + pr_warn("you may want to try another patch for loopback file " + "on %s(0x%lx) branch\n", au_sbtype(h_sb), magic); +} + +int au_loopback_init(void) +{ + int err; + struct super_block *sb __maybe_unused; + + BUILD_BUG_ON(sizeof(sb->s_magic) != sizeof(unsigned long)); + + err = 0; + au_warn_loopback_array = kcalloc(au_warn_loopback_step, + sizeof(unsigned long), GFP_NOFS); + if (unlikely(!au_warn_loopback_array)) + err = -ENOMEM; + + return err; +} + +void au_loopback_fin(void) +{ + if (backing_file_func) + symbol_put(loop_backing_file); + kfree(au_warn_loopback_array); +} diff --git a/fs/aufs/loop.h b/fs/aufs/loop.h new file mode 100644 index 000000000..48bf070e8 --- /dev/null +++ b/fs/aufs/loop.h @@ -0,0 +1,39 @@ +/* + * Copyright (C) 2005-2016 Junjiro R. Okajima + */ + +/* + * support for loopback mount as a branch + */ + +#ifndef __AUFS_LOOP_H__ +#define __AUFS_LOOP_H__ + +#ifdef __KERNEL__ + +struct dentry; +struct super_block; + +#ifdef CONFIG_AUFS_BDEV_LOOP +/* drivers/block/loop.c */ +struct file *loop_backing_file(struct super_block *sb); + +/* loop.c */ +int au_test_loopback_overlap(struct super_block *sb, struct dentry *h_adding); +int au_test_loopback_kthread(void); +void au_warn_loopback(struct super_block *h_sb); + +int au_loopback_init(void); +void au_loopback_fin(void); +#else +AuStubInt0(au_test_loopback_overlap, struct super_block *sb, + struct dentry *h_adding) +AuStubInt0(au_test_loopback_kthread, void) +AuStubVoid(au_warn_loopback, struct super_block *h_sb) + +AuStubInt0(au_loopback_init, void) +AuStubVoid(au_loopback_fin, void) +#endif /* BLK_DEV_LOOP */ + +#endif /* __KERNEL__ */ +#endif /* __AUFS_LOOP_H__ */ diff --git a/fs/aufs/magic.mk b/fs/aufs/magic.mk new file mode 100644 index 000000000..4f83bdf1d --- /dev/null +++ b/fs/aufs/magic.mk @@ -0,0 +1,30 @@ + +# defined in ${srctree}/fs/fuse/inode.c +# tristate +ifdef CONFIG_FUSE_FS +ccflags-y += -DFUSE_SUPER_MAGIC=0x65735546 +endif + +# defined in ${srctree}/fs/xfs/xfs_sb.h +# tristate +ifdef CONFIG_XFS_FS +ccflags-y += -DXFS_SB_MAGIC=0x58465342 +endif + +# defined in ${srctree}/fs/configfs/mount.c +# tristate +ifdef CONFIG_CONFIGFS_FS +ccflags-y += -DCONFIGFS_MAGIC=0x62656570 +endif + +# defined in ${srctree}/fs/ubifs/ubifs.h +# tristate +ifdef CONFIG_UBIFS_FS +ccflags-y += -DUBIFS_SUPER_MAGIC=0x24051905 +endif + +# defined in ${srctree}/fs/hfsplus/hfsplus_raw.h +# tristate +ifdef CONFIG_HFSPLUS_FS +ccflags-y += -DHFSPLUS_SUPER_MAGIC=0x482b +endif diff --git a/fs/aufs/module.c b/fs/aufs/module.c new file mode 100644 index 000000000..8a28377c5 --- /dev/null +++ b/fs/aufs/module.c @@ -0,0 +1,207 @@ +/* + * Copyright (C) 2005-2016 Junjiro R. Okajima + */ + +/* + * module global variables and operations + */ + +#include <linux/module.h> +#include <linux/seq_file.h> +#include "aufs.h" + +void *au_kzrealloc(void *p, unsigned int nused, unsigned int new_sz, gfp_t gfp) +{ + if (new_sz <= nused) + return p; + + p = krealloc(p, new_sz, gfp); + if (p) + memset(p + nused, 0, new_sz - nused); + return p; +} + +/* ---------------------------------------------------------------------- */ + +/* + * aufs caches + */ +struct kmem_cache *au_cachep[AuCache_Last]; +static int __init au_cache_init(void) +{ + au_cachep[AuCache_DINFO] = AuCacheCtor(au_dinfo, au_di_init_once); + if (au_cachep[AuCache_DINFO]) + /* SLAB_DESTROY_BY_RCU */ + au_cachep[AuCache_ICNTNR] = AuCacheCtor(au_icntnr, + au_icntnr_init_once); + if (au_cachep[AuCache_ICNTNR]) + au_cachep[AuCache_FINFO] = AuCacheCtor(au_finfo, + au_fi_init_once); + if (au_cachep[AuCache_FINFO]) + au_cachep[AuCache_VDIR] = AuCache(au_vdir); + if (au_cachep[AuCache_VDIR]) + au_cachep[AuCache_DEHSTR] = AuCache(au_vdir_dehstr); + if (au_cachep[AuCache_DEHSTR]) + return 0; + + return -ENOMEM; +} + +static void au_cache_fin(void) +{ + int i; + + /* + * Make sure all delayed rcu free inodes are flushed before we + * destroy cache. + */ + rcu_barrier(); + + /* excluding AuCache_HNOTIFY */ + BUILD_BUG_ON(AuCache_HNOTIFY + 1 != AuCache_Last); + for (i = 0; i < AuCache_HNOTIFY; i++) { + kmem_cache_destroy(au_cachep[i]); + au_cachep[i] = NULL; + } +} + +/* ---------------------------------------------------------------------- */ + +int au_dir_roflags; + +#ifdef CONFIG_AUFS_SBILIST +/* + * iterate_supers_type() doesn't protect us from + * remounting (branch management) + */ +struct au_splhead au_sbilist; +#endif + +struct lock_class_key au_lc_key[AuLcKey_Last]; + +/* + * functions for module interface. + */ +MODULE_LICENSE("GPL"); +/* MODULE_LICENSE("GPL v2"); */ +MODULE_AUTHOR("Junjiro R. Okajima <aufs-users@lists.sourceforge.net>"); +MODULE_DESCRIPTION(AUFS_NAME + " -- Advanced multi layered unification filesystem"); +MODULE_VERSION(AUFS_VERSION); + +/* this module parameter has no meaning when SYSFS is disabled */ +int sysaufs_brs = 1; +MODULE_PARM_DESC(brs, "use <sysfs>/fs/aufs/si_*/brN"); +module_param_named(brs, sysaufs_brs, int, S_IRUGO); + +/* this module parameter has no meaning when USER_NS is disabled */ +bool au_userns; +MODULE_PARM_DESC(allow_userns, "allow unprivileged to mount under userns"); +module_param_named(allow_userns, au_userns, bool, S_IRUGO); + +/* ---------------------------------------------------------------------- */ + +static char au_esc_chars[0x20 + 3]; /* 0x01-0x20, backslash, del, and NULL */ + +int au_seq_path(struct seq_file *seq, struct path *path) +{ + int err; + + err = seq_path(seq, path, au_esc_chars); + if (err > 0) + err = 0; + else if (err < 0) + err = -ENOMEM; + + return err; +} + +/* ---------------------------------------------------------------------- */ + +static int __init aufs_init(void) +{ + int err, i; + char *p; + + p = au_esc_chars; + for (i = 1; i <= ' '; i++) + *p++ = i; + *p++ = '\\'; + *p++ = '\x7f'; + *p = 0; + + au_dir_roflags = au_file_roflags(O_DIRECTORY | O_LARGEFILE); + + memcpy(aufs_iop_nogetattr, aufs_iop, sizeof(aufs_iop)); + for (i = 0; i < AuIop_Last; i++) + aufs_iop_nogetattr[i].getattr = NULL; + + au_sbilist_init(); + sysaufs_brs_init(); + au_debug_init(); + au_dy_init(); + err = sysaufs_init(); + if (unlikely(err)) + goto out; + err = au_procfs_init(); + if (unlikely(err)) + goto out_sysaufs; + err = au_wkq_init(); + if (unlikely(err)) + goto out_procfs; + err = au_loopback_init(); + if (unlikely(err)) + goto out_wkq; + err = au_hnotify_init(); + if (unlikely(err)) + goto out_loopback; + err = au_sysrq_init(); + if (unlikely(err)) + goto out_hin; + err = au_cache_init(); + if (unlikely(err)) + goto out_sysrq; + + aufs_fs_type.fs_flags |= au_userns ? FS_USERNS_MOUNT : 0; + err = register_filesystem(&aufs_fs_type); + if (unlikely(err)) + goto out_cache; + + /* since we define pr_fmt, call printk directly */ + printk(KERN_INFO AUFS_NAME " " AUFS_VERSION "\n"); + goto out; /* success */ + +out_cache: + au_cache_fin(); +out_sysrq: + au_sysrq_fin(); +out_hin: + au_hnotify_fin(); +out_loopback: + au_loopback_fin(); +out_wkq: + au_wkq_fin(); +out_procfs: + au_procfs_fin(); +out_sysaufs: + sysaufs_fin(); + au_dy_fin(); +out: + return err; +} + +static void __exit aufs_exit(void) +{ + unregister_filesystem(&aufs_fs_type); + au_cache_fin(); + au_sysrq_fin(); + au_hnotify_fin(); + au_loopback_fin(); + au_wkq_fin(); + au_procfs_fin(); + sysaufs_fin(); + au_dy_fin(); +} + +module_init(aufs_init); +module_exit(aufs_exit); diff --git a/fs/aufs/module.h b/fs/aufs/module.h new file mode 100644 index 000000000..bb8644730 --- /dev/null +++ b/fs/aufs/module.h @@ -0,0 +1,92 @@ +/* + * Copyright (C) 2005-2016 Junjiro R. Okajima + */ + +/* + * module initialization and module-global + */ + +#ifndef __AUFS_MODULE_H__ +#define __AUFS_MODULE_H__ + +#ifdef __KERNEL__ + +#include <linux/slab.h> + +struct path; +struct seq_file; + +/* module parameters */ +extern int sysaufs_brs; +extern bool au_userns; + +/* ---------------------------------------------------------------------- */ + +extern int au_dir_roflags; + +enum { + AuLcNonDir_FIINFO, + AuLcNonDir_DIINFO, + AuLcNonDir_IIINFO, + + AuLcDir_FIINFO, + AuLcDir_DIINFO, + AuLcDir_IIINFO, + + AuLcSymlink_DIINFO, + AuLcSymlink_IIINFO, + + AuLcKey_Last +}; +extern struct lock_class_key au_lc_key[AuLcKey_Last]; + +void *au_kzrealloc(void *p, unsigned int nused, unsigned int new_sz, gfp_t gfp); +int au_seq_path(struct seq_file *seq, struct path *path); + +#ifdef CONFIG_PROC_FS +/* procfs.c */ +int __init au_procfs_init(void); +void au_procfs_fin(void); +#else +AuStubInt0(au_procfs_init, void); +AuStubVoid(au_procfs_fin, void); +#endif + +/* ---------------------------------------------------------------------- */ + +/* kmem cache */ +enum { + AuCache_DINFO, + AuCache_ICNTNR, + AuCache_FINFO, + AuCache_VDIR, + AuCache_DEHSTR, + AuCache_HNOTIFY, /* must be last */ + AuCache_Last +}; + +#define AuCacheFlags (SLAB_RECLAIM_ACCOUNT | SLAB_MEM_SPREAD) +#define AuCache(type) KMEM_CACHE(type, AuCacheFlags) +#define AuCacheCtor(type, ctor) \ + kmem_cache_create(#type, sizeof(struct type), \ + __alignof__(struct type), AuCacheFlags, ctor) + +extern struct kmem_cache *au_cachep[]; + +#define AuCacheFuncs(name, index) \ +static inline struct au_##name *au_cache_alloc_##name(void) \ +{ return kmem_cache_alloc(au_cachep[AuCache_##index], GFP_NOFS); } \ +static inline void au_cache_free_##name(struct au_##name *p) \ +{ kmem_cache_free(au_cachep[AuCache_##index], p); } + +AuCacheFuncs(dinfo, DINFO); +AuCacheFuncs(icntnr, ICNTNR); +AuCacheFuncs(finfo, FINFO); +AuCacheFuncs(vdir, VDIR); +AuCacheFuncs(vdir_dehstr, DEHSTR); +#ifdef CONFIG_AUFS_HNOTIFY +AuCacheFuncs(hnotify, HNOTIFY); +#endif + +#endif /* __KERNEL__ */ +#endif /* __AUFS_MODULE_H__ */ diff --git a/fs/aufs/mvdown.c b/fs/aufs/mvdown.c new file mode 100644 index 000000000..d8415422b --- /dev/null +++ b/fs/aufs/mvdown.c @@ -0,0 +1,690 @@ +/* + * Copyright (C) 2011-2016 Junjiro R. Okajima + */ + +/* + * move-down, opposite of copy-up + */ + +#include "aufs.h" + +struct au_mvd_args { + struct { + struct super_block *h_sb; + struct dentry *h_parent; + struct au_hinode *hdir; + struct inode *h_dir, *h_inode; + struct au_pin pin; + } info[AUFS_MVDOWN_NARRAY]; + + struct aufs_mvdown mvdown; + struct dentry *dentry, *parent; + struct inode *inode, *dir; + struct super_block *sb; + aufs_bindex_t bopq, bwh, bfound; + unsigned char rename_lock; +}; + +#define mvd_errno mvdown.au_errno +#define mvd_bsrc mvdown.stbr[AUFS_MVDOWN_UPPER].bindex +#define mvd_src_brid mvdown.stbr[AUFS_MVDOWN_UPPER].brid +#define mvd_bdst mvdown.stbr[AUFS_MVDOWN_LOWER].bindex +#define mvd_dst_brid mvdown.stbr[AUFS_MVDOWN_LOWER].brid + +#define mvd_h_src_sb info[AUFS_MVDOWN_UPPER].h_sb +#define mvd_h_src_parent info[AUFS_MVDOWN_UPPER].h_parent +#define mvd_hdir_src info[AUFS_MVDOWN_UPPER].hdir +#define mvd_h_src_dir info[AUFS_MVDOWN_UPPER].h_dir +#define mvd_h_src_inode info[AUFS_MVDOWN_UPPER].h_inode +#define mvd_pin_src info[AUFS_MVDOWN_UPPER].pin + +#define mvd_h_dst_sb info[AUFS_MVDOWN_LOWER].h_sb +#define mvd_h_dst_parent info[AUFS_MVDOWN_LOWER].h_parent +#define mvd_hdir_dst info[AUFS_MVDOWN_LOWER].hdir +#define mvd_h_dst_dir info[AUFS_MVDOWN_LOWER].h_dir +#define mvd_h_dst_inode info[AUFS_MVDOWN_LOWER].h_inode +#define mvd_pin_dst info[AUFS_MVDOWN_LOWER].pin + +#define AU_MVD_PR(flag, ...) do { \ + if (flag) \ + pr_err(__VA_ARGS__); \ + } while (0) + +static int find_lower_writable(struct au_mvd_args *a) +{ + struct super_block *sb; + aufs_bindex_t bindex, bend; + struct au_branch *br; + + sb = a->sb; + bindex = a->mvd_bsrc; + bend = au_sbend(sb); + if (a->mvdown.flags & AUFS_MVDOWN_FHSM_LOWER) + for (bindex++; bindex <= bend; bindex++) { + br = au_sbr(sb, bindex); + if (au_br_fhsm(br->br_perm) + && (!(au_br_sb(br)->s_flags & MS_RDONLY))) + return bindex; + } + else if (!(a->mvdown.flags & AUFS_MVDOWN_ROLOWER)) + for (bindex++; bindex <= bend; bindex++) { + br = au_sbr(sb, bindex); + if (!au_br_rdonly(br)) + return bindex; + } + else + for (bindex++; bindex <= bend; bindex++) { + br = au_sbr(sb, bindex); + if (!(au_br_sb(br)->s_flags & MS_RDONLY)) { + if (au_br_rdonly(br)) + a->mvdown.flags + |= AUFS_MVDOWN_ROLOWER_R; + return bindex; + } + } + + return -1; +} + +/* make the parent dir on bdst */ +static int au_do_mkdir(const unsigned char dmsg, struct au_mvd_args *a) +{ + int err; + + err = 0; + a->mvd_hdir_src = au_hi(a->dir, a->mvd_bsrc); + a->mvd_hdir_dst = au_hi(a->dir, a->mvd_bdst); + a->mvd_h_src_parent = au_h_dptr(a->parent, a->mvd_bsrc); + a->mvd_h_dst_parent = NULL; + if (au_dbend(a->parent) >= a->mvd_bdst) + a->mvd_h_dst_parent = au_h_dptr(a->parent, a->mvd_bdst); + if (!a->mvd_h_dst_parent) { + err = au_cpdown_dirs(a->dentry, a->mvd_bdst); + if (unlikely(err)) { + AU_MVD_PR(dmsg, "cpdown_dirs failed\n"); + goto out; + } + a->mvd_h_dst_parent = au_h_dptr(a->parent, a->mvd_bdst); + } + +out: + AuTraceErr(err); + return err; +} + +/* lock them all */ +static int au_do_lock(const unsigned char dmsg, struct au_mvd_args *a) +{ + int err; + struct dentry *h_trap; + + a->mvd_h_src_sb = au_sbr_sb(a->sb, a->mvd_bsrc); + a->mvd_h_dst_sb = au_sbr_sb(a->sb, a->mvd_bdst); + err = au_pin(&a->mvd_pin_dst, a->dentry, a->mvd_bdst, + au_opt_udba(a->sb), + AuPin_MNT_WRITE | AuPin_DI_LOCKED); + AuTraceErr(err); + if (unlikely(err)) { + AU_MVD_PR(dmsg, "pin_dst failed\n"); + goto out; + } + + if (a->mvd_h_src_sb != a->mvd_h_dst_sb) { + a->rename_lock = 0; + au_pin_init(&a->mvd_pin_src, a->dentry, a->mvd_bsrc, + AuLsc_DI_PARENT, AuLsc_I_PARENT3, + au_opt_udba(a->sb), + AuPin_MNT_WRITE | AuPin_DI_LOCKED); + err = au_do_pin(&a->mvd_pin_src); + AuTraceErr(err); + a->mvd_h_src_dir = d_inode(a->mvd_h_src_parent); + if (unlikely(err)) { + AU_MVD_PR(dmsg, "pin_src failed\n"); + goto out_dst; + } + goto out; /* success */ + } + + a->rename_lock = 1; + au_pin_hdir_unlock(&a->mvd_pin_dst); + err = au_pin(&a->mvd_pin_src, a->dentry, a->mvd_bsrc, + au_opt_udba(a->sb), + AuPin_MNT_WRITE | AuPin_DI_LOCKED); + AuTraceErr(err); + a->mvd_h_src_dir = d_inode(a->mvd_h_src_parent); + if (unlikely(err)) { + AU_MVD_PR(dmsg, "pin_src failed\n"); + au_pin_hdir_lock(&a->mvd_pin_dst); + goto out_dst; + } + au_pin_hdir_unlock(&a->mvd_pin_src); + h_trap = vfsub_lock_rename(a->mvd_h_src_parent, a->mvd_hdir_src, + a->mvd_h_dst_parent, a->mvd_hdir_dst); + if (h_trap) { + err = (h_trap != a->mvd_h_src_parent); + if (err) + err = (h_trap != a->mvd_h_dst_parent); + } + BUG_ON(err); /* it should never happen */ + if (unlikely(a->mvd_h_src_dir != au_pinned_h_dir(&a->mvd_pin_src))) { + err = -EBUSY; + AuTraceErr(err); + vfsub_unlock_rename(a->mvd_h_src_parent, a->mvd_hdir_src, + a->mvd_h_dst_parent, a->mvd_hdir_dst); + au_pin_hdir_lock(&a->mvd_pin_src); + au_unpin(&a->mvd_pin_src); + au_pin_hdir_lock(&a->mvd_pin_dst); + goto out_dst; + } + goto out; /* success */ + +out_dst: + au_unpin(&a->mvd_pin_dst); +out: + AuTraceErr(err); + return err; +} + +static void au_do_unlock(const unsigned char dmsg, struct au_mvd_args *a) +{ + if (!a->rename_lock) + au_unpin(&a->mvd_pin_src); + else { + vfsub_unlock_rename(a->mvd_h_src_parent, a->mvd_hdir_src, + a->mvd_h_dst_parent, a->mvd_hdir_dst); + au_pin_hdir_lock(&a->mvd_pin_src); + au_unpin(&a->mvd_pin_src); + au_pin_hdir_lock(&a->mvd_pin_dst); + } + au_unpin(&a->mvd_pin_dst); +} + +/* copy-down the file */ +static int au_do_cpdown(const unsigned char dmsg, struct au_mvd_args *a) +{ + int err; + struct au_cp_generic cpg = { + .dentry = a->dentry, + .bdst = a->mvd_bdst, + .bsrc = a->mvd_bsrc, + .len = -1, + .pin = &a->mvd_pin_dst, + .flags = AuCpup_DTIME | AuCpup_HOPEN + }; + + AuDbg("b%d, b%d\n", cpg.bsrc, cpg.bdst); + if (a->mvdown.flags & AUFS_MVDOWN_OWLOWER) + au_fset_cpup(cpg.flags, OVERWRITE); + if (a->mvdown.flags & AUFS_MVDOWN_ROLOWER) + au_fset_cpup(cpg.flags, RWDST); + err = au_sio_cpdown_simple(&cpg); + if (unlikely(err)) + AU_MVD_PR(dmsg, "cpdown failed\n"); + + AuTraceErr(err); + return err; +} + +/* + * unlink the whiteout on bdst if exist which may be created by UDBA while we + * were sleeping + */ +static int au_do_unlink_wh(const unsigned char dmsg, struct au_mvd_args *a) +{ + int err; + struct path h_path; + struct au_branch *br; + struct inode *delegated; + + br = au_sbr(a->sb, a->mvd_bdst); + h_path.dentry = au_wh_lkup(a->mvd_h_dst_parent, &a->dentry->d_name, br); + err = PTR_ERR(h_path.dentry); + if (IS_ERR(h_path.dentry)) { + AU_MVD_PR(dmsg, "wh_lkup failed\n"); + goto out; + } + + err = 0; + if (d_is_positive(h_path.dentry)) { + h_path.mnt = au_br_mnt(br); + delegated = NULL; + err = vfsub_unlink(d_inode(a->mvd_h_dst_parent), &h_path, + &delegated, /*force*/0); + if (unlikely(err == -EWOULDBLOCK)) { + pr_warn("cannot retry for NFSv4 delegation" + " for an internal unlink\n"); + iput(delegated); + } + if (unlikely(err)) + AU_MVD_PR(dmsg, "wh_unlink failed\n"); + } + dput(h_path.dentry); + +out: + AuTraceErr(err); + return err; +} + +/* + * unlink the topmost h_dentry + */ +static int au_do_unlink(const unsigned char dmsg, struct au_mvd_args *a) +{ + int err; + struct path h_path; + struct inode *delegated; + + h_path.mnt = au_sbr_mnt(a->sb, a->mvd_bsrc); + h_path.dentry = au_h_dptr(a->dentry, a->mvd_bsrc); + delegated = NULL; + err = vfsub_unlink(a->mvd_h_src_dir, &h_path, &delegated, /*force*/0); + if (unlikely(err == -EWOULDBLOCK)) { + pr_warn("cannot retry for NFSv4 delegation" + " for an internal unlink\n"); + iput(delegated); + } + if (unlikely(err)) + AU_MVD_PR(dmsg, "unlink failed\n"); + + AuTraceErr(err); + return err; +} + +/* Since mvdown succeeded, we ignore an error of this function */ +static void au_do_stfs(const unsigned char dmsg, struct au_mvd_args *a) +{ + int err; + struct au_branch *br; + + a->mvdown.flags |= AUFS_MVDOWN_STFS_FAILED; + br = au_sbr(a->sb, a->mvd_bsrc); + err = au_br_stfs(br, &a->mvdown.stbr[AUFS_MVDOWN_UPPER].stfs); + if (!err) { + br = au_sbr(a->sb, a->mvd_bdst); + a->mvdown.stbr[AUFS_MVDOWN_LOWER].brid = br->br_id; + err = au_br_stfs(br, &a->mvdown.stbr[AUFS_MVDOWN_LOWER].stfs); + } + if (!err) + a->mvdown.flags &= ~AUFS_MVDOWN_STFS_FAILED; + else + AU_MVD_PR(dmsg, "statfs failed (%d), ignored\n", err); +} + +/* + * copy-down the file and unlink the bsrc file. + * - unlink the bdst whout if exist + * - copy-down the file (with whtmp name and rename) + * - unlink the bsrc file + */ +static int au_do_mvdown(const unsigned char dmsg, struct au_mvd_args *a) +{ + int err; + + err = au_do_mkdir(dmsg, a); + if (!err) + err = au_do_lock(dmsg, a); + if (unlikely(err)) + goto out; + + /* + * do not revert the activities we made on bdst since they should be + * harmless in aufs. + */ + + err = au_do_cpdown(dmsg, a); + if (!err) + err = au_do_unlink_wh(dmsg, a); + if (!err && !(a->mvdown.flags & AUFS_MVDOWN_KUPPER)) + err = au_do_unlink(dmsg, a); + if (unlikely(err)) + goto out_unlock; + + AuDbg("%pd2, 0x%x, %d --> %d\n", + a->dentry, a->mvdown.flags, a->mvd_bsrc, a->mvd_bdst); + if (find_lower_writable(a) < 0) + a->mvdown.flags |= AUFS_MVDOWN_BOTTOM; + + if (a->mvdown.flags & AUFS_MVDOWN_STFS) + au_do_stfs(dmsg, a); + + /* maintain internal array */ + if (!(a->mvdown.flags & AUFS_MVDOWN_KUPPER)) { + au_set_h_dptr(a->dentry, a->mvd_bsrc, NULL); + au_set_dbstart(a->dentry, a->mvd_bdst); + au_set_h_iptr(a->inode, a->mvd_bsrc, NULL, /*flags*/0); + au_set_ibstart(a->inode, a->mvd_bdst); + } else { + /* hide the lower */ + au_set_h_dptr(a->dentry, a->mvd_bdst, NULL); + au_set_dbend(a->dentry, a->mvd_bsrc); + au_set_h_iptr(a->inode, a->mvd_bdst, NULL, /*flags*/0); + au_set_ibend(a->inode, a->mvd_bsrc); + } + if (au_dbend(a->dentry) < a->mvd_bdst) + au_set_dbend(a->dentry, a->mvd_bdst); + if (au_ibend(a->inode) < a->mvd_bdst) + au_set_ibend(a->inode, a->mvd_bdst); + +out_unlock: + au_do_unlock(dmsg, a); +out: + AuTraceErr(err); + return err; +} + +/* ---------------------------------------------------------------------- */ + +/* make sure the file is idle */ +static int au_mvd_args_busy(const unsigned char dmsg, struct au_mvd_args *a) +{ + int err, plinked; + + err = 0; + plinked = !!au_opt_test(au_mntflags(a->sb), PLINK); + if (au_dbstart(a->dentry) == a->mvd_bsrc + && au_dcount(a->dentry) == 1 + && atomic_read(&a->inode->i_count) == 1 + /* && a->mvd_h_src_inode->i_nlink == 1 */ + && (!plinked || !au_plink_test(a->inode)) + && a->inode->i_nlink == 1) + goto out; + + err = -EBUSY; + AU_MVD_PR(dmsg, + "b%d, d{b%d, c%d?}, i{c%d?, l%u}, hi{l%u}, p{%d, %d}\n", + a->mvd_bsrc, au_dbstart(a->dentry), au_dcount(a->dentry), + atomic_read(&a->inode->i_count), a->inode->i_nlink, + a->mvd_h_src_inode->i_nlink, + plinked, plinked ? au_plink_test(a->inode) : 0); + +out: + AuTraceErr(err); + return err; +} + +/* make sure the parent dir is fine */ +static int au_mvd_args_parent(const unsigned char dmsg, + struct au_mvd_args *a) +{ + int err; + aufs_bindex_t bindex; + + err = 0; + if (unlikely(au_alive_dir(a->parent))) { + err = -ENOENT; + AU_MVD_PR(dmsg, "parent dir is dead\n"); + goto out; + } + + a->bopq = au_dbdiropq(a->parent); + bindex = au_wbr_nonopq(a->dentry, a->mvd_bdst); + AuDbg("b%d\n", bindex); + if (unlikely((bindex >= 0 && bindex < a->mvd_bdst) + || (a->bopq != -1 && a->bopq < a->mvd_bdst))) { + err = -EINVAL; + a->mvd_errno = EAU_MVDOWN_OPAQUE; + AU_MVD_PR(dmsg, "ancestor is opaque b%d, b%d\n", + a->bopq, a->mvd_bdst); + } + +out: + AuTraceErr(err); + return err; +} + +static int au_mvd_args_intermediate(const unsigned char dmsg, + struct au_mvd_args *a) +{ + int err; + struct au_dinfo *dinfo, *tmp; + + /* lookup the next lower positive entry */ + err = -ENOMEM; + tmp = au_di_alloc(a->sb, AuLsc_DI_TMP); + if (unlikely(!tmp)) + goto out; + + a->bfound = -1; + a->bwh = -1; + dinfo = au_di(a->dentry); + au_di_cp(tmp, dinfo); + au_di_swap(tmp, dinfo); + + /* returns the number of positive dentries */ + err = au_lkup_dentry(a->dentry, a->mvd_bsrc + 1, /*type*/0); + if (!err) + a->bwh = au_dbwh(a->dentry); + else if (err > 0) + a->bfound = au_dbstart(a->dentry); + + au_di_swap(tmp, dinfo); + au_rw_write_unlock(&tmp->di_rwsem); + au_di_free(tmp); + if (unlikely(err < 0)) + AU_MVD_PR(dmsg, "failed look-up lower\n"); + + /* + * here, we have these cases. + * bfound == -1 + * no positive dentry under bsrc. there are more sub-cases. + * bwh < 0 + * there no whiteout, we can safely move-down. + * bwh <= bsrc + * impossible + * bsrc < bwh && bwh < bdst + * there is a whiteout on RO branch. cannot proceed. + * bwh == bdst + * there is a whiteout on the RW target branch. it should + * be removed. + * bdst < bwh + * there is a whiteout somewhere unrelated branch. + * -1 < bfound && bfound <= bsrc + * impossible. + * bfound < bdst + * found, but it is on RO branch between bsrc and bdst. cannot + * proceed. + * bfound == bdst + * found, replace it if AUFS_MVDOWN_FORCE is set. otherwise return + * error. + * bdst < bfound + * found, after we create the file on bdst, it will be hidden. + */ + + AuDebugOn(a->bfound == -1 + && a->bwh != -1 + && a->bwh <= a->mvd_bsrc); + AuDebugOn(-1 < a->bfound + && a->bfound <= a->mvd_bsrc); + + err = -EINVAL; + if (a->bfound == -1 + && a->mvd_bsrc < a->bwh + && a->bwh != -1 + && a->bwh < a->mvd_bdst) { + a->mvd_errno = EAU_MVDOWN_WHITEOUT; + AU_MVD_PR(dmsg, "bsrc %d, bdst %d, bfound %d, bwh %d\n", + a->mvd_bsrc, a->mvd_bdst, a->bfound, a->bwh); + goto out; + } else if (a->bfound != -1 && a->bfound < a->mvd_bdst) { + a->mvd_errno = EAU_MVDOWN_UPPER; + AU_MVD_PR(dmsg, "bdst %d, bfound %d\n", + a->mvd_bdst, a->bfound); + goto out; + } + + err = 0; /* success */ + +out: + AuTraceErr(err); + return err; +} + +static int au_mvd_args_exist(const unsigned char dmsg, struct au_mvd_args *a) +{ + int err; + + err = 0; + if (!(a->mvdown.flags & AUFS_MVDOWN_OWLOWER) + && a->bfound == a->mvd_bdst) + err = -EEXIST; + AuTraceErr(err); + return err; +} + +static int au_mvd_args(const unsigned char dmsg, struct au_mvd_args *a) +{ + int err; + struct au_branch *br; + + err = -EISDIR; + if (unlikely(S_ISDIR(a->inode->i_mode))) + goto out; + + err = -EINVAL; + if (!(a->mvdown.flags & AUFS_MVDOWN_BRID_UPPER)) + a->mvd_bsrc = au_ibstart(a->inode); + else { + a->mvd_bsrc = au_br_index(a->sb, a->mvd_src_brid); + if (unlikely(a->mvd_bsrc < 0 + || (a->mvd_bsrc < au_dbstart(a->dentry) + || au_dbend(a->dentry) < a->mvd_bsrc + || !au_h_dptr(a->dentry, a->mvd_bsrc)) + || (a->mvd_bsrc < au_ibstart(a->inode) + || au_ibend(a->inode) < a->mvd_bsrc + || !au_h_iptr(a->inode, a->mvd_bsrc)))) { + a->mvd_errno = EAU_MVDOWN_NOUPPER; + AU_MVD_PR(dmsg, "no upper\n"); + goto out; + } + } + if (unlikely(a->mvd_bsrc == au_sbend(a->sb))) { + a->mvd_errno = EAU_MVDOWN_BOTTOM; + AU_MVD_PR(dmsg, "on the bottom\n"); + goto out; + } + a->mvd_h_src_inode = au_h_iptr(a->inode, a->mvd_bsrc); + br = au_sbr(a->sb, a->mvd_bsrc); + err = au_br_rdonly(br); + if (!(a->mvdown.flags & AUFS_MVDOWN_ROUPPER)) { + if (unlikely(err)) + goto out; + } else if (!(vfsub_native_ro(a->mvd_h_src_inode) + || IS_APPEND(a->mvd_h_src_inode))) { + if (err) + a->mvdown.flags |= AUFS_MVDOWN_ROUPPER_R; + /* go on */ + } else + goto out; + + err = -EINVAL; + if (!(a->mvdown.flags & AUFS_MVDOWN_BRID_LOWER)) { + a->mvd_bdst = find_lower_writable(a); + if (unlikely(a->mvd_bdst < 0)) { + a->mvd_errno = EAU_MVDOWN_BOTTOM; + AU_MVD_PR(dmsg, "no writable lower branch\n"); + goto out; + } + } else { + a->mvd_bdst = au_br_index(a->sb, a->mvd_dst_brid); + if (unlikely(a->mvd_bdst < 0 + || au_sbend(a->sb) < a->mvd_bdst)) { + a->mvd_errno = EAU_MVDOWN_NOLOWERBR; + AU_MVD_PR(dmsg, "no lower brid\n"); + goto out; + } + } + + err = au_mvd_args_busy(dmsg, a); + if (!err) + err = au_mvd_args_parent(dmsg, a); + if (!err) + err = au_mvd_args_intermediate(dmsg, a); + if (!err) + err = au_mvd_args_exist(dmsg, a); + if (!err) + AuDbg("b%d, b%d\n", a->mvd_bsrc, a->mvd_bdst); + +out: + AuTraceErr(err); + return err; +} + +int au_mvdown(struct dentry *dentry, struct aufs_mvdown __user *uarg) +{ + int err, e; + unsigned char dmsg; + struct au_mvd_args *args; + struct inode *inode; + + inode = d_inode(dentry); + err = -EPERM; + if (unlikely(!capable(CAP_SYS_ADMIN))) + goto out; + + err = -ENOMEM; + args = kmalloc(sizeof(*args), GFP_NOFS); + if (unlikely(!args)) + goto out; + + err = copy_from_user(&args->mvdown, uarg, sizeof(args->mvdown)); + if (!err) + err = !access_ok(VERIFY_WRITE, uarg, sizeof(*uarg)); + if (unlikely(err)) { + err = -EFAULT; + AuTraceErr(err); + goto out_free; + } + AuDbg("flags 0x%x\n", args->mvdown.flags); + args->mvdown.flags &= ~(AUFS_MVDOWN_ROLOWER_R | AUFS_MVDOWN_ROUPPER_R); + args->mvdown.au_errno = 0; + args->dentry = dentry; + args->inode = inode; + args->sb = dentry->d_sb; + + err = -ENOENT; + dmsg = !!(args->mvdown.flags & AUFS_MVDOWN_DMSG); + args->parent = dget_parent(dentry); + args->dir = d_inode(args->parent); + inode_lock_nested(args->dir, I_MUTEX_PARENT); + dput(args->parent); + if (unlikely(args->parent != dentry->d_parent)) { + AU_MVD_PR(dmsg, "parent dir is moved\n"); + goto out_dir; + } + + inode_lock_nested(inode, I_MUTEX_CHILD); + err = aufs_read_lock(dentry, AuLock_DW | AuLock_FLUSH | AuLock_NOPLMW); + if (unlikely(err)) + goto out_inode; + + di_write_lock_parent(args->parent); + err = au_mvd_args(dmsg, args); + if (unlikely(err)) + goto out_parent; + + err = au_do_mvdown(dmsg, args); + if (unlikely(err)) + goto out_parent; + + au_cpup_attr_timesizes(args->dir); + au_cpup_attr_timesizes(inode); + if (!(args->mvdown.flags & AUFS_MVDOWN_KUPPER)) + au_cpup_igen(inode, au_h_iptr(inode, args->mvd_bdst)); + /* au_digen_dec(dentry); */ + +out_parent: + di_write_unlock(args->parent); + aufs_read_unlock(dentry, AuLock_DW); +out_inode: + inode_unlock(inode); +out_dir: + inode_unlock(args->dir); +out_free: + e = copy_to_user(uarg, &args->mvdown, sizeof(args->mvdown)); + if (unlikely(e)) + err = -EFAULT; + kfree(args); +out: + AuTraceErr(err); + return err; +} diff --git a/fs/aufs/opts.c b/fs/aufs/opts.c new file mode 100644 index 000000000..fc90b657a --- /dev/null +++ b/fs/aufs/opts.c @@ -0,0 +1,1846 @@ +/* + * Copyright (C) 2005-2016 Junjiro R. Okajima + */ + +/* + * mount options/flags + */ + +#include <linux/namei.h> +#include <linux/types.h> /* a distribution requires */ +#include <linux/parser.h> +#include "aufs.h" + +/* ---------------------------------------------------------------------- */ + +enum { + Opt_br, + Opt_add, Opt_del, Opt_mod, Opt_append, Opt_prepend, + Opt_idel, Opt_imod, + Opt_dirwh, Opt_rdcache, Opt_rdblk, Opt_rdhash, + Opt_rdblk_def, Opt_rdhash_def, + Opt_xino, Opt_noxino, + Opt_trunc_xino, Opt_trunc_xino_v, Opt_notrunc_xino, + Opt_trunc_xino_path, Opt_itrunc_xino, + Opt_trunc_xib, Opt_notrunc_xib, + Opt_shwh, Opt_noshwh, + Opt_plink, Opt_noplink, Opt_list_plink, + Opt_udba, + Opt_dio, Opt_nodio, + Opt_diropq_a, Opt_diropq_w, + Opt_warn_perm, Opt_nowarn_perm, + Opt_wbr_copyup, Opt_wbr_create, + Opt_fhsm_sec, + Opt_verbose, Opt_noverbose, + Opt_sum, Opt_nosum, Opt_wsum, + Opt_dirperm1, Opt_nodirperm1, + Opt_acl, Opt_noacl, + Opt_tail, Opt_ignore, Opt_ignore_silent, Opt_err +}; + +static match_table_t options = { + {Opt_br, "br=%s"}, + {Opt_br, "br:%s"}, + + {Opt_add, "add=%d:%s"}, + {Opt_add, "add:%d:%s"}, + {Opt_add, "ins=%d:%s"}, + {Opt_add, "ins:%d:%s"}, + {Opt_append, "append=%s"}, + {Opt_append, "append:%s"}, + {Opt_prepend, "prepend=%s"}, + {Opt_prepend, "prepend:%s"}, + + {Opt_del, "del=%s"}, + {Opt_del, "del:%s"}, + /* {Opt_idel, "idel:%d"}, */ + {Opt_mod, "mod=%s"}, + {Opt_mod, "mod:%s"}, + /* {Opt_imod, "imod:%d:%s"}, */ + + {Opt_dirwh, "dirwh=%d"}, + + {Opt_xino, "xino=%s"}, + {Opt_noxino, "noxino"}, + {Opt_trunc_xino, "trunc_xino"}, + {Opt_trunc_xino_v, "trunc_xino_v=%d:%d"}, + {Opt_notrunc_xino, "notrunc_xino"}, + {Opt_trunc_xino_path, "trunc_xino=%s"}, + {Opt_itrunc_xino, "itrunc_xino=%d"}, + /* {Opt_zxino, "zxino=%s"}, */ + {Opt_trunc_xib, "trunc_xib"}, + {Opt_notrunc_xib, "notrunc_xib"}, + +#ifdef CONFIG_PROC_FS + {Opt_plink, "plink"}, +#else + {Opt_ignore_silent, "plink"}, +#endif + + {Opt_noplink, "noplink"}, + +#ifdef CONFIG_AUFS_DEBUG + {Opt_list_plink, "list_plink"}, +#endif + + {Opt_udba, "udba=%s"}, + + {Opt_dio, "dio"}, + {Opt_nodio, "nodio"}, + +#ifdef CONFIG_AUFS_FHSM + {Opt_fhsm_sec, "fhsm_sec=%d"}, +#else + {Opt_ignore_silent, "fhsm_sec=%d"}, +#endif + + {Opt_diropq_a, "diropq=always"}, + {Opt_diropq_a, "diropq=a"}, + {Opt_diropq_w, "diropq=whiteouted"}, + {Opt_diropq_w, "diropq=w"}, + + {Opt_warn_perm, "warn_perm"}, + {Opt_nowarn_perm, "nowarn_perm"}, + + /* keep them temporary */ + {Opt_ignore_silent, "nodlgt"}, + {Opt_ignore_silent, "clean_plink"}, + +#ifdef CONFIG_AUFS_SHWH + {Opt_shwh, "shwh"}, +#endif + {Opt_noshwh, "noshwh"}, + + {Opt_dirperm1, "dirperm1"}, + {Opt_nodirperm1, "nodirperm1"}, + + {Opt_verbose, "verbose"}, + {Opt_verbose, "v"}, + {Opt_noverbose, "noverbose"}, + {Opt_noverbose, "quiet"}, + {Opt_noverbose, "q"}, + {Opt_noverbose, "silent"}, + + {Opt_sum, "sum"}, + {Opt_nosum, "nosum"}, + {Opt_wsum, "wsum"}, + + {Opt_rdcache, "rdcache=%d"}, + {Opt_rdblk, "rdblk=%d"}, + {Opt_rdblk_def, "rdblk=def"}, + {Opt_rdhash, "rdhash=%d"}, + {Opt_rdhash_def, "rdhash=def"}, + + {Opt_wbr_create, "create=%s"}, + {Opt_wbr_create, "create_policy=%s"}, + {Opt_wbr_copyup, "cpup=%s"}, + {Opt_wbr_copyup, "copyup=%s"}, + {Opt_wbr_copyup, "copyup_policy=%s"}, + + /* generic VFS flag */ +#ifdef CONFIG_FS_POSIX_ACL + {Opt_acl, "acl"}, + {Opt_noacl, "noacl"}, +#else + {Opt_ignore_silent, "acl"}, + {Opt_ignore_silent, "noacl"}, +#endif + + /* internal use for the scripts */ + {Opt_ignore_silent, "si=%s"}, + + {Opt_br, "dirs=%s"}, + {Opt_ignore, "debug=%d"}, + {Opt_ignore, "delete=whiteout"}, + {Opt_ignore, "delete=all"}, + {Opt_ignore, "imap=%s"}, + + /* temporary workaround, due to old mount(8)? */ + {Opt_ignore_silent, "relatime"}, + + {Opt_err, NULL} +}; + +/* ---------------------------------------------------------------------- */ + +static const char *au_parser_pattern(int val, match_table_t tbl) +{ + struct match_token *p; + + p = tbl; + while (p->pattern) { + if (p->token == val) + return p->pattern; + p++; + } + BUG(); + return "??"; +} + +static const char *au_optstr(int *val, match_table_t tbl) +{ + struct match_token *p; + int v; + + v = *val; + if (!v) + goto out; + p = tbl; + while (p->pattern) { + if (p->token + && (v & p->token) == p->token) { + *val &= ~p->token; + return p->pattern; + } + p++; + } + +out: + return NULL; +} + +/* ---------------------------------------------------------------------- */ + +static match_table_t brperm = { + {AuBrPerm_RO, AUFS_BRPERM_RO}, + {AuBrPerm_RR, AUFS_BRPERM_RR}, + {AuBrPerm_RW, AUFS_BRPERM_RW}, + {0, NULL} +}; + +static match_table_t brattr = { + /* general */ + {AuBrAttr_COO_REG, AUFS_BRATTR_COO_REG}, + {AuBrAttr_COO_ALL, AUFS_BRATTR_COO_ALL}, + /* 'unpin' attrib is meaningless since linux-3.18-rc1 */ + {AuBrAttr_UNPIN, AUFS_BRATTR_UNPIN}, +#ifdef CONFIG_AUFS_FHSM + {AuBrAttr_FHSM, AUFS_BRATTR_FHSM}, +#endif +#ifdef CONFIG_AUFS_XATTR + {AuBrAttr_ICEX, AUFS_BRATTR_ICEX}, + {AuBrAttr_ICEX_SEC, AUFS_BRATTR_ICEX_SEC}, + {AuBrAttr_ICEX_SYS, AUFS_BRATTR_ICEX_SYS}, + {AuBrAttr_ICEX_TR, AUFS_BRATTR_ICEX_TR}, + {AuBrAttr_ICEX_USR, AUFS_BRATTR_ICEX_USR}, + {AuBrAttr_ICEX_OTH, AUFS_BRATTR_ICEX_OTH}, +#endif + + /* ro/rr branch */ + {AuBrRAttr_WH, AUFS_BRRATTR_WH}, + + /* rw branch */ + {AuBrWAttr_MOO, AUFS_BRWATTR_MOO}, + {AuBrWAttr_NoLinkWH, AUFS_BRWATTR_NLWH}, + + {0, NULL} +}; + +static int br_attr_val(char *str, match_table_t table, substring_t args[]) +{ + int attr, v; + char *p; + + attr = 0; + do { + p = strchr(str, '+'); + if (p) + *p = 0; + v = match_token(str, table, args); + if (v) { + if (v & AuBrAttr_CMOO_Mask) + attr &= ~AuBrAttr_CMOO_Mask; + attr |= v; + } else { + if (p) + *p = '+'; + pr_warn("ignored branch attribute %s\n", str); + break; + } + if (p) + str = p + 1; + } while (p); + + return attr; +} + +static int au_do_optstr_br_attr(au_br_perm_str_t *str, int perm) +{ + int sz; + const char *p; + char *q; + + q = str->a; + *q = 0; + p = au_optstr(&perm, brattr); + if (p) { + sz = strlen(p); + memcpy(q, p, sz + 1); + q += sz; + } else + goto out; + + do { + p = au_optstr(&perm, brattr); + if (p) { + *q++ = '+'; + sz = strlen(p); + memcpy(q, p, sz + 1); + q += sz; + } + } while (p); + +out: + return q - str->a; +} + +static int noinline_for_stack br_perm_val(char *perm) +{ + int val, bad, sz; + char *p; + substring_t args[MAX_OPT_ARGS]; + au_br_perm_str_t attr; + + p = strchr(perm, '+'); + if (p) + *p = 0; + val = match_token(perm, brperm, args); + if (!val) { + if (p) + *p = '+'; + pr_warn("ignored branch permission %s\n", perm); + val = AuBrPerm_RO; + goto out; + } + if (!p) + goto out; + + val |= br_attr_val(p + 1, brattr, args); + + bad = 0; + switch (val & AuBrPerm_Mask) { + case AuBrPerm_RO: + case AuBrPerm_RR: + bad = val & AuBrWAttr_Mask; + val &= ~AuBrWAttr_Mask; + break; + case AuBrPerm_RW: + bad = val & AuBrRAttr_Mask; + val &= ~AuBrRAttr_Mask; + break; + } + + /* + * 'unpin' attrib becomes meaningless since linux-3.18-rc1, but aufs + * does not treat it as an error, just warning. + * this is a tiny guard for the user operation. + */ + if (val & AuBrAttr_UNPIN) { + bad |= AuBrAttr_UNPIN; + val &= ~AuBrAttr_UNPIN; + } + + if (unlikely(bad)) { + sz = au_do_optstr_br_attr(&attr, bad); + AuDebugOn(!sz); + pr_warn("ignored branch attribute %s\n", attr.a); + } + +out: + return val; +} + +void au_optstr_br_perm(au_br_perm_str_t *str, int perm) +{ + au_br_perm_str_t attr; + const char *p; + char *q; + int sz; + + q = str->a; + p = au_optstr(&perm, brperm); + AuDebugOn(!p || !*p); + sz = strlen(p); + memcpy(q, p, sz + 1); + q += sz; + + sz = au_do_optstr_br_attr(&attr, perm); + if (sz) { + *q++ = '+'; + memcpy(q, attr.a, sz + 1); + } + + AuDebugOn(strlen(str->a) >= sizeof(str->a)); +} + +/* ---------------------------------------------------------------------- */ + +static match_table_t udbalevel = { + {AuOpt_UDBA_REVAL, "reval"}, + {AuOpt_UDBA_NONE, "none"}, +#ifdef CONFIG_AUFS_HNOTIFY + {AuOpt_UDBA_HNOTIFY, "notify"}, /* abstraction */ +#ifdef CONFIG_AUFS_HFSNOTIFY + {AuOpt_UDBA_HNOTIFY, "fsnotify"}, +#endif +#endif + {-1, NULL} +}; + +static int noinline_for_stack udba_val(char *str) +{ + substring_t args[MAX_OPT_ARGS]; + + return match_token(str, udbalevel, args); +} + +const char *au_optstr_udba(int udba) +{ + return au_parser_pattern(udba, udbalevel); +} + +/* ---------------------------------------------------------------------- */ + +static match_table_t au_wbr_create_policy = { + {AuWbrCreate_TDP, "tdp"}, + {AuWbrCreate_TDP, "top-down-parent"}, + {AuWbrCreate_RR, "rr"}, + {AuWbrCreate_RR, "round-robin"}, + {AuWbrCreate_MFS, "mfs"}, + {AuWbrCreate_MFS, "most-free-space"}, + {AuWbrCreate_MFSV, "mfs:%d"}, + {AuWbrCreate_MFSV, "most-free-space:%d"}, + + {AuWbrCreate_MFSRR, "mfsrr:%d"}, + {AuWbrCreate_MFSRRV, "mfsrr:%d:%d"}, + {AuWbrCreate_PMFS, "pmfs"}, + {AuWbrCreate_PMFSV, "pmfs:%d"}, + {AuWbrCreate_PMFSRR, "pmfsrr:%d"}, + {AuWbrCreate_PMFSRRV, "pmfsrr:%d:%d"}, + + {-1, NULL} +}; + +/* + * cf. linux/lib/parser.c and cmdline.c + * gave up calling memparse() since it uses simple_strtoull() instead of + * kstrto...(). + */ +static int noinline_for_stack +au_match_ull(substring_t *s, unsigned long long *result) +{ + int err; + unsigned int len; + char a[32]; + + err = -ERANGE; + len = s->to - s->from; + if (len + 1 <= sizeof(a)) { + memcpy(a, s->from, len); + a[len] = '\0'; + err = kstrtoull(a, 0, result); + } + return err; +} + +static int au_wbr_mfs_wmark(substring_t *arg, char *str, + struct au_opt_wbr_create *create) +{ + int err; + unsigned long long ull; + + err = 0; + if (!au_match_ull(arg, &ull)) + create->mfsrr_watermark = ull; + else { + pr_err("bad integer in %s\n", str); + err = -EINVAL; + } + + return err; +} + +static int au_wbr_mfs_sec(substring_t *arg, char *str, + struct au_opt_wbr_create *create) +{ + int n, err; + + err = 0; + if (!match_int(arg, &n) && 0 <= n && n <= AUFS_MFS_MAX_SEC) + create->mfs_second = n; + else { + pr_err("bad integer in %s\n", str); + err = -EINVAL; + } + + return err; +} + +static int noinline_for_stack +au_wbr_create_val(char *str, struct au_opt_wbr_create *create) +{ + int err, e; + substring_t args[MAX_OPT_ARGS]; + + err = match_token(str, au_wbr_create_policy, args); + create->wbr_create = err; + switch (err) { + case AuWbrCreate_MFSRRV: + case AuWbrCreate_PMFSRRV: + e = au_wbr_mfs_wmark(&args[0], str, create); + if (!e) + e = au_wbr_mfs_sec(&args[1], str, create); + if (unlikely(e)) + err = e; + break; + case AuWbrCreate_MFSRR: + case AuWbrCreate_PMFSRR: + e = au_wbr_mfs_wmark(&args[0], str, create); + if (unlikely(e)) { + err = e; + break; + } + /*FALLTHROUGH*/ + case AuWbrCreate_MFS: + case AuWbrCreate_PMFS: + create->mfs_second = AUFS_MFS_DEF_SEC; + break; + case AuWbrCreate_MFSV: + case AuWbrCreate_PMFSV: + e = au_wbr_mfs_sec(&args[0], str, create); + if (unlikely(e)) + err = e; + break; + } + + return err; +} + +const char *au_optstr_wbr_create(int wbr_create) +{ + return au_parser_pattern(wbr_create, au_wbr_create_policy); +} + +static match_table_t au_wbr_copyup_policy = { + {AuWbrCopyup_TDP, "tdp"}, + {AuWbrCopyup_TDP, "top-down-parent"}, + {AuWbrCopyup_BUP, "bup"}, + {AuWbrCopyup_BUP, "bottom-up-parent"}, + {AuWbrCopyup_BU, "bu"}, + {AuWbrCopyup_BU, "bottom-up"}, + {-1, NULL} +}; + +static int noinline_for_stack au_wbr_copyup_val(char *str) +{ + substring_t args[MAX_OPT_ARGS]; + + return match_token(str, au_wbr_copyup_policy, args); +} + +const char *au_optstr_wbr_copyup(int wbr_copyup) +{ + return au_parser_pattern(wbr_copyup, au_wbr_copyup_policy); +} + +/* ---------------------------------------------------------------------- */ + +static const int lkup_dirflags = LOOKUP_FOLLOW | LOOKUP_DIRECTORY; + +static void dump_opts(struct au_opts *opts) +{ +#ifdef CONFIG_AUFS_DEBUG + /* reduce stack space */ + union { + struct au_opt_add *add; + struct au_opt_del *del; + struct au_opt_mod *mod; + struct au_opt_xino *xino; + struct au_opt_xino_itrunc *xino_itrunc; + struct au_opt_wbr_create *create; + } u; + struct au_opt *opt; + + opt = opts->opt; + while (opt->type != Opt_tail) { + switch (opt->type) { + case Opt_add: + u.add = &opt->add; + AuDbg("add {b%d, %s, 0x%x, %p}\n", + u.add->bindex, u.add->pathname, u.add->perm, + u.add->path.dentry); + break; + case Opt_del: + case Opt_idel: + u.del = &opt->del; + AuDbg("del {%s, %p}\n", + u.del->pathname, u.del->h_path.dentry); + break; + case Opt_mod: + case Opt_imod: + u.mod = &opt->mod; + AuDbg("mod {%s, 0x%x, %p}\n", + u.mod->path, u.mod->perm, u.mod->h_root); + break; + case Opt_append: + u.add = &opt->add; + AuDbg("append {b%d, %s, 0x%x, %p}\n", + u.add->bindex, u.add->pathname, u.add->perm, + u.add->path.dentry); + break; + case Opt_prepend: + u.add = &opt->add; + AuDbg("prepend {b%d, %s, 0x%x, %p}\n", + u.add->bindex, u.add->pathname, u.add->perm, + u.add->path.dentry); + break; + case Opt_dirwh: + AuDbg("dirwh %d\n", opt->dirwh); + break; + case Opt_rdcache: + AuDbg("rdcache %d\n", opt->rdcache); + break; + case Opt_rdblk: + AuDbg("rdblk %u\n", opt->rdblk); + break; + case Opt_rdblk_def: + AuDbg("rdblk_def\n"); + break; + case Opt_rdhash: + AuDbg("rdhash %u\n", opt->rdhash); + break; + case Opt_rdhash_def: + AuDbg("rdhash_def\n"); + break; + case Opt_xino: + u.xino = &opt->xino; + AuDbg("xino {%s %pD}\n", u.xino->path, u.xino->file); + break; + case Opt_trunc_xino: + AuLabel(trunc_xino); + break; + case Opt_notrunc_xino: + AuLabel(notrunc_xino); + break; + case Opt_trunc_xino_path: + case Opt_itrunc_xino: + u.xino_itrunc = &opt->xino_itrunc; + AuDbg("trunc_xino %d\n", u.xino_itrunc->bindex); + break; + case Opt_noxino: + AuLabel(noxino); + break; + case Opt_trunc_xib: + AuLabel(trunc_xib); + break; + case Opt_notrunc_xib: + AuLabel(notrunc_xib); + break; + case Opt_shwh: + AuLabel(shwh); + break; + case Opt_noshwh: + AuLabel(noshwh); + break; + case Opt_dirperm1: + AuLabel(dirperm1); + break; + case Opt_nodirperm1: + AuLabel(nodirperm1); + break; + case Opt_plink: + AuLabel(plink); + break; + case Opt_noplink: + AuLabel(noplink); + break; + case Opt_list_plink: + AuLabel(list_plink); + break; + case Opt_udba: + AuDbg("udba %d, %s\n", + opt->udba, au_optstr_udba(opt->udba)); + break; + case Opt_dio: + AuLabel(dio); + break; + case Opt_nodio: + AuLabel(nodio); + break; + case Opt_diropq_a: + AuLabel(diropq_a); + break; + case Opt_diropq_w: + AuLabel(diropq_w); + break; + case Opt_warn_perm: + AuLabel(warn_perm); + break; + case Opt_nowarn_perm: + AuLabel(nowarn_perm); + break; + case Opt_verbose: + AuLabel(verbose); + break; + case Opt_noverbose: + AuLabel(noverbose); + break; + case Opt_sum: + AuLabel(sum); + break; + case Opt_nosum: + AuLabel(nosum); + break; + case Opt_wsum: + AuLabel(wsum); + break; + case Opt_wbr_create: + u.create = &opt->wbr_create; + AuDbg("create %d, %s\n", u.create->wbr_create, + au_optstr_wbr_create(u.create->wbr_create)); + switch (u.create->wbr_create) { + case AuWbrCreate_MFSV: + case AuWbrCreate_PMFSV: + AuDbg("%d sec\n", u.create->mfs_second); + break; + case AuWbrCreate_MFSRR: + AuDbg("%llu watermark\n", + u.create->mfsrr_watermark); + break; + case AuWbrCreate_MFSRRV: + case AuWbrCreate_PMFSRRV: + AuDbg("%llu watermark, %d sec\n", + u.create->mfsrr_watermark, + u.create->mfs_second); + break; + } + break; + case Opt_wbr_copyup: + AuDbg("copyup %d, %s\n", opt->wbr_copyup, + au_optstr_wbr_copyup(opt->wbr_copyup)); + break; + case Opt_fhsm_sec: + AuDbg("fhsm_sec %u\n", opt->fhsm_second); + break; + case Opt_acl: + AuLabel(acl); + break; + case Opt_noacl: + AuLabel(noacl); + break; + default: + BUG(); + } + opt++; + } +#endif +} + +void au_opts_free(struct au_opts *opts) +{ + struct au_opt *opt; + + opt = opts->opt; + while (opt->type != Opt_tail) { + switch (opt->type) { + case Opt_add: + case Opt_append: + case Opt_prepend: + path_put(&opt->add.path); + break; + case Opt_del: + case Opt_idel: + path_put(&opt->del.h_path); + break; + case Opt_mod: + case Opt_imod: + dput(opt->mod.h_root); + break; + case Opt_xino: + fput(opt->xino.file); + break; + } + opt++; + } +} + +static int opt_add(struct au_opt *opt, char *opt_str, unsigned long sb_flags, + aufs_bindex_t bindex) +{ + int err; + struct au_opt_add *add = &opt->add; + char *p; + + add->bindex = bindex; + add->perm = AuBrPerm_RO; + add->pathname = opt_str; + p = strchr(opt_str, '='); + if (p) { + *p++ = 0; + if (*p) + add->perm = br_perm_val(p); + } + + err = vfsub_kern_path(add->pathname, lkup_dirflags, &add->path); + if (!err) { + if (!p) { + add->perm = AuBrPerm_RO; + if (au_test_fs_rr(add->path.dentry->d_sb)) + add->perm = AuBrPerm_RR; + else if (!bindex && !(sb_flags & MS_RDONLY)) + add->perm = AuBrPerm_RW; + } + opt->type = Opt_add; + goto out; + } + pr_err("lookup failed %s (%d)\n", add->pathname, err); + err = -EINVAL; + +out: + return err; +} + +static int au_opts_parse_del(struct au_opt_del *del, substring_t args[]) +{ + int err; + + del->pathname = args[0].from; + AuDbg("del path %s\n", del->pathname); + + err = vfsub_kern_path(del->pathname, lkup_dirflags, &del->h_path); + if (unlikely(err)) + pr_err("lookup failed %s (%d)\n", del->pathname, err); + + return err; +} + +#if 0 /* reserved for future use */ +static int au_opts_parse_idel(struct super_block *sb, aufs_bindex_t bindex, + struct au_opt_del *del, substring_t args[]) +{ + int err; + struct dentry *root; + + err = -EINVAL; + root = sb->s_root; + aufs_read_lock(root, AuLock_FLUSH); + if (bindex < 0 || au_sbend(sb) < bindex) { + pr_err("out of bounds, %d\n", bindex); + goto out; + } + + err = 0; + del->h_path.dentry = dget(au_h_dptr(root, bindex)); + del->h_path.mnt = mntget(au_sbr_mnt(sb, bindex)); + +out: + aufs_read_unlock(root, !AuLock_IR); + return err; +} +#endif + +static int noinline_for_stack +au_opts_parse_mod(struct au_opt_mod *mod, substring_t args[]) +{ + int err; + struct path path; + char *p; + + err = -EINVAL; + mod->path = args[0].from; + p = strchr(mod->path, '='); + if (unlikely(!p)) { + pr_err("no permssion %s\n", args[0].from); + goto out; + } + + *p++ = 0; + err = vfsub_kern_path(mod->path, lkup_dirflags, &path); + if (unlikely(err)) { + pr_err("lookup failed %s (%d)\n", mod->path, err); + goto out; + } + + mod->perm = br_perm_val(p); + AuDbg("mod path %s, perm 0x%x, %s\n", mod->path, mod->perm, p); + mod->h_root = dget(path.dentry); + path_put(&path); + +out: + return err; +} + +#if 0 /* reserved for future use */ +static int au_opts_parse_imod(struct super_block *sb, aufs_bindex_t bindex, + struct au_opt_mod *mod, substring_t args[]) +{ + int err; + struct dentry *root; + + err = -EINVAL; + root = sb->s_root; + aufs_read_lock(root, AuLock_FLUSH); + if (bindex < 0 || au_sbend(sb) < bindex) { + pr_err("out of bounds, %d\n", bindex); + goto out; + } + + err = 0; + mod->perm = br_perm_val(args[1].from); + AuDbg("mod path %s, perm 0x%x, %s\n", + mod->path, mod->perm, args[1].from); + mod->h_root = dget(au_h_dptr(root, bindex)); + +out: + aufs_read_unlock(root, !AuLock_IR); + return err; +} +#endif + +static int au_opts_parse_xino(struct super_block *sb, struct au_opt_xino *xino, + substring_t args[]) +{ + int err; + struct file *file; + + file = au_xino_create(sb, args[0].from, /*silent*/0); + err = PTR_ERR(file); + if (IS_ERR(file)) + goto out; + + err = -EINVAL; + if (unlikely(file->f_path.dentry->d_sb == sb)) { + fput(file); + pr_err("%s must be outside\n", args[0].from); + goto out; + } + + err = 0; + xino->file = file; + xino->path = args[0].from; + +out: + return err; +} + +static int noinline_for_stack +au_opts_parse_xino_itrunc_path(struct super_block *sb, + struct au_opt_xino_itrunc *xino_itrunc, + substring_t args[]) +{ + int err; + aufs_bindex_t bend, bindex; + struct path path; + struct dentry *root; + + err = vfsub_kern_path(args[0].from, lkup_dirflags, &path); + if (unlikely(err)) { + pr_err("lookup failed %s (%d)\n", args[0].from, err); + goto out; + } + + xino_itrunc->bindex = -1; + root = sb->s_root; + aufs_read_lock(root, AuLock_FLUSH); + bend = au_sbend(sb); + for (bindex = 0; bindex <= bend; bindex++) { + if (au_h_dptr(root, bindex) == path.dentry) { + xino_itrunc->bindex = bindex; + break; + } + } + aufs_read_unlock(root, !AuLock_IR); + path_put(&path); + + if (unlikely(xino_itrunc->bindex < 0)) { + pr_err("no such branch %s\n", args[0].from); + err = -EINVAL; + } + +out: + return err; +} + +/* called without aufs lock */ +int au_opts_parse(struct super_block *sb, char *str, struct au_opts *opts) +{ + int err, n, token; + aufs_bindex_t bindex; + unsigned char skipped; + struct dentry *root; + struct au_opt *opt, *opt_tail; + char *opt_str; + /* reduce the stack space */ + union { + struct au_opt_xino_itrunc *xino_itrunc; + struct au_opt_wbr_create *create; + } u; + struct { + substring_t args[MAX_OPT_ARGS]; + } *a; + + err = -ENOMEM; + a = kmalloc(sizeof(*a), GFP_NOFS); + if (unlikely(!a)) + goto out; + + root = sb->s_root; + err = 0; + bindex = 0; + opt = opts->opt; + opt_tail = opt + opts->max_opt - 1; + opt->type = Opt_tail; + while (!err && (opt_str = strsep(&str, ",")) && *opt_str) { + err = -EINVAL; + skipped = 0; + token = match_token(opt_str, options, a->args); + switch (token) { + case Opt_br: + err = 0; + while (!err && (opt_str = strsep(&a->args[0].from, ":")) + && *opt_str) { + err = opt_add(opt, opt_str, opts->sb_flags, + bindex++); + if (unlikely(!err && ++opt > opt_tail)) { + err = -E2BIG; + break; + } + opt->type = Opt_tail; + skipped = 1; + } + break; + case Opt_add: + if (unlikely(match_int(&a->args[0], &n))) { + pr_err("bad integer in %s\n", opt_str); + break; + } + bindex = n; + err = opt_add(opt, a->args[1].from, opts->sb_flags, + bindex); + if (!err) + opt->type = token; + break; + case Opt_append: + err = opt_add(opt, a->args[0].from, opts->sb_flags, + /*dummy bindex*/1); + if (!err) + opt->type = token; + break; + case Opt_prepend: + err = opt_add(opt, a->args[0].from, opts->sb_flags, + /*bindex*/0); + if (!err) + opt->type = token; + break; + case Opt_del: + err = au_opts_parse_del(&opt->del, a->args); + if (!err) + opt->type = token; + break; +#if 0 /* reserved for future use */ + case Opt_idel: + del->pathname = "(indexed)"; + if (unlikely(match_int(&args[0], &n))) { + pr_err("bad integer in %s\n", opt_str); + break; + } + err = au_opts_parse_idel(sb, n, &opt->del, a->args); + if (!err) + opt->type = token; + break; +#endif + case Opt_mod: + err = au_opts_parse_mod(&opt->mod, a->args); + if (!err) + opt->type = token; + break; +#ifdef IMOD /* reserved for future use */ + case Opt_imod: + u.mod->path = "(indexed)"; + if (unlikely(match_int(&a->args[0], &n))) { + pr_err("bad integer in %s\n", opt_str); + break; + } + err = au_opts_parse_imod(sb, n, &opt->mod, a->args); + if (!err) + opt->type = token; + break; +#endif + case Opt_xino: + err = au_opts_parse_xino(sb, &opt->xino, a->args); + if (!err) + opt->type = token; + break; + + case Opt_trunc_xino_path: + err = au_opts_parse_xino_itrunc_path + (sb, &opt->xino_itrunc, a->args); + if (!err) + opt->type = token; + break; + + case Opt_itrunc_xino: + u.xino_itrunc = &opt->xino_itrunc; + if (unlikely(match_int(&a->args[0], &n))) { + pr_err("bad integer in %s\n", opt_str); + break; + } + u.xino_itrunc->bindex = n; + aufs_read_lock(root, AuLock_FLUSH); + if (n < 0 || au_sbend(sb) < n) { + pr_err("out of bounds, %d\n", n); + aufs_read_unlock(root, !AuLock_IR); + break; + } + aufs_read_unlock(root, !AuLock_IR); + err = 0; + opt->type = token; + break; + + case Opt_dirwh: + if (unlikely(match_int(&a->args[0], &opt->dirwh))) + break; + err = 0; + opt->type = token; + break; + + case Opt_rdcache: + if (unlikely(match_int(&a->args[0], &n))) { + pr_err("bad integer in %s\n", opt_str); + break; + } + if (unlikely(n > AUFS_RDCACHE_MAX)) { + pr_err("rdcache must be smaller than %d\n", + AUFS_RDCACHE_MAX); + break; + } + opt->rdcache = n; + err = 0; + opt->type = token; + break; + case Opt_rdblk: + if (unlikely(match_int(&a->args[0], &n) + || n < 0 + || n > KMALLOC_MAX_SIZE)) { + pr_err("bad integer in %s\n", opt_str); + break; + } + if (unlikely(n && n < NAME_MAX)) { + pr_err("rdblk must be larger than %d\n", + NAME_MAX); + break; + } + opt->rdblk = n; + err = 0; + opt->type = token; + break; + case Opt_rdhash: + if (unlikely(match_int(&a->args[0], &n) + || n < 0 + || n * sizeof(struct hlist_head) + > KMALLOC_MAX_SIZE)) { + pr_err("bad integer in %s\n", opt_str); + break; + } + opt->rdhash = n; + err = 0; + opt->type = token; + break; + + case Opt_trunc_xino: + case Opt_notrunc_xino: + case Opt_noxino: + case Opt_trunc_xib: + case Opt_notrunc_xib: + case Opt_shwh: + case Opt_noshwh: + case Opt_dirperm1: + case Opt_nodirperm1: + case Opt_plink: + case Opt_noplink: + case Opt_list_plink: + case Opt_dio: + case Opt_nodio: + case Opt_diropq_a: + case Opt_diropq_w: + case Opt_warn_perm: + case Opt_nowarn_perm: + case Opt_verbose: + case Opt_noverbose: + case Opt_sum: + case Opt_nosum: + case Opt_wsum: + case Opt_rdblk_def: + case Opt_rdhash_def: + case Opt_acl: + case Opt_noacl: + err = 0; + opt->type = token; + break; + + case Opt_udba: + opt->udba = udba_val(a->args[0].from); + if (opt->udba >= 0) { + err = 0; + opt->type = token; + } else + pr_err("wrong value, %s\n", opt_str); + break; + + case Opt_wbr_create: + u.create = &opt->wbr_create; + u.create->wbr_create + = au_wbr_create_val(a->args[0].from, u.create); + if (u.create->wbr_create >= 0) { + err = 0; + opt->type = token; + } else + pr_err("wrong value, %s\n", opt_str); + break; + case Opt_wbr_copyup: + opt->wbr_copyup = au_wbr_copyup_val(a->args[0].from); + if (opt->wbr_copyup >= 0) { + err = 0; + opt->type = token; + } else + pr_err("wrong value, %s\n", opt_str); + break; + + case Opt_fhsm_sec: + if (unlikely(match_int(&a->args[0], &n) + || n < 0)) { + pr_err("bad integer in %s\n", opt_str); + break; + } + if (sysaufs_brs) { + opt->fhsm_second = n; + opt->type = token; + } else + pr_warn("ignored %s\n", opt_str); + err = 0; + break; + + case Opt_ignore: + pr_warn("ignored %s\n", opt_str); + /*FALLTHROUGH*/ + case Opt_ignore_silent: + skipped = 1; + err = 0; + break; + case Opt_err: + pr_err("unknown option %s\n", opt_str); + break; + } + + if (!err && !skipped) { + if (unlikely(++opt > opt_tail)) { + err = -E2BIG; + opt--; + opt->type = Opt_tail; + break; + } + opt->type = Opt_tail; + } + } + + kfree(a); + dump_opts(opts); + if (unlikely(err)) + au_opts_free(opts); + +out: + return err; +} + +static int au_opt_wbr_create(struct super_block *sb, + struct au_opt_wbr_create *create) +{ + int err; + struct au_sbinfo *sbinfo; + + SiMustWriteLock(sb); + + err = 1; /* handled */ + sbinfo = au_sbi(sb); + if (sbinfo->si_wbr_create_ops->fin) { + err = sbinfo->si_wbr_create_ops->fin(sb); + if (!err) + err = 1; + } + + sbinfo->si_wbr_create = create->wbr_create; + sbinfo->si_wbr_create_ops = au_wbr_create_ops + create->wbr_create; + switch (create->wbr_create) { + case AuWbrCreate_MFSRRV: + case AuWbrCreate_MFSRR: + case AuWbrCreate_PMFSRR: + case AuWbrCreate_PMFSRRV: + sbinfo->si_wbr_mfs.mfsrr_watermark = create->mfsrr_watermark; + /*FALLTHROUGH*/ + case AuWbrCreate_MFS: + case AuWbrCreate_MFSV: + case AuWbrCreate_PMFS: + case AuWbrCreate_PMFSV: + sbinfo->si_wbr_mfs.mfs_expire + = msecs_to_jiffies(create->mfs_second * MSEC_PER_SEC); + break; + } + + if (sbinfo->si_wbr_create_ops->init) + sbinfo->si_wbr_create_ops->init(sb); /* ignore */ + + return err; +} + +/* + * returns, + * plus: processed without an error + * zero: unprocessed + */ +static int au_opt_simple(struct super_block *sb, struct au_opt *opt, + struct au_opts *opts) +{ + int err; + struct au_sbinfo *sbinfo; + + SiMustWriteLock(sb); + + err = 1; /* handled */ + sbinfo = au_sbi(sb); + switch (opt->type) { + case Opt_udba: + sbinfo->si_mntflags &= ~AuOptMask_UDBA; + sbinfo->si_mntflags |= opt->udba; + opts->given_udba |= opt->udba; + break; + + case Opt_plink: + au_opt_set(sbinfo->si_mntflags, PLINK); + break; + case Opt_noplink: + if (au_opt_test(sbinfo->si_mntflags, PLINK)) + au_plink_put(sb, /*verbose*/1); + au_opt_clr(sbinfo->si_mntflags, PLINK); + break; + case Opt_list_plink: + if (au_opt_test(sbinfo->si_mntflags, PLINK)) + au_plink_list(sb); + break; + + case Opt_dio: + au_opt_set(sbinfo->si_mntflags, DIO); + au_fset_opts(opts->flags, REFRESH_DYAOP); + break; + case Opt_nodio: + au_opt_clr(sbinfo->si_mntflags, DIO); + au_fset_opts(opts->flags, REFRESH_DYAOP); + break; + + case Opt_fhsm_sec: + au_fhsm_set(sbinfo, opt->fhsm_second); + break; + + case Opt_diropq_a: + au_opt_set(sbinfo->si_mntflags, ALWAYS_DIROPQ); + break; + case Opt_diropq_w: + au_opt_clr(sbinfo->si_mntflags, ALWAYS_DIROPQ); + break; + + case Opt_warn_perm: + au_opt_set(sbinfo->si_mntflags, WARN_PERM); + break; + case Opt_nowarn_perm: + au_opt_clr(sbinfo->si_mntflags, WARN_PERM); + break; + + case Opt_verbose: + au_opt_set(sbinfo->si_mntflags, VERBOSE); + break; + case Opt_noverbose: + au_opt_clr(sbinfo->si_mntflags, VERBOSE); + break; + + case Opt_sum: + au_opt_set(sbinfo->si_mntflags, SUM); + break; + case Opt_wsum: + au_opt_clr(sbinfo->si_mntflags, SUM); + au_opt_set(sbinfo->si_mntflags, SUM_W); + case Opt_nosum: + au_opt_clr(sbinfo->si_mntflags, SUM); + au_opt_clr(sbinfo->si_mntflags, SUM_W); + break; + + case Opt_wbr_create: + err = au_opt_wbr_create(sb, &opt->wbr_create); + break; + case Opt_wbr_copyup: + sbinfo->si_wbr_copyup = opt->wbr_copyup; + sbinfo->si_wbr_copyup_ops = au_wbr_copyup_ops + opt->wbr_copyup; + break; + + case Opt_dirwh: + sbinfo->si_dirwh = opt->dirwh; + break; + + case Opt_rdcache: + sbinfo->si_rdcache + = msecs_to_jiffies(opt->rdcache * MSEC_PER_SEC); + break; + case Opt_rdblk: + sbinfo->si_rdblk = opt->rdblk; + break; + case Opt_rdblk_def: + sbinfo->si_rdblk = AUFS_RDBLK_DEF; + break; + case Opt_rdhash: + sbinfo->si_rdhash = opt->rdhash; + break; + case Opt_rdhash_def: + sbinfo->si_rdhash = AUFS_RDHASH_DEF; + break; + + case Opt_shwh: + au_opt_set(sbinfo->si_mntflags, SHWH); + break; + case Opt_noshwh: + au_opt_clr(sbinfo->si_mntflags, SHWH); + break; + + case Opt_dirperm1: + au_opt_set(sbinfo->si_mntflags, DIRPERM1); + break; + case Opt_nodirperm1: + au_opt_clr(sbinfo->si_mntflags, DIRPERM1); + break; + + case Opt_trunc_xino: + au_opt_set(sbinfo->si_mntflags, TRUNC_XINO); + break; + case Opt_notrunc_xino: + au_opt_clr(sbinfo->si_mntflags, TRUNC_XINO); + break; + + case Opt_trunc_xino_path: + case Opt_itrunc_xino: + err = au_xino_trunc(sb, opt->xino_itrunc.bindex); + if (!err) + err = 1; + break; + + case Opt_trunc_xib: + au_fset_opts(opts->flags, TRUNC_XIB); + break; + case Opt_notrunc_xib: + au_fclr_opts(opts->flags, TRUNC_XIB); + break; + + case Opt_acl: + sb->s_flags |= MS_POSIXACL; + break; + case Opt_noacl: + sb->s_flags &= ~MS_POSIXACL; + break; + + default: + err = 0; + break; + } + + return err; +} + +/* + * returns tri-state. + * plus: processed without an error + * zero: unprocessed + * minus: error + */ +static int au_opt_br(struct super_block *sb, struct au_opt *opt, + struct au_opts *opts) +{ + int err, do_refresh; + + err = 0; + switch (opt->type) { + case Opt_append: + opt->add.bindex = au_sbend(sb) + 1; + if (opt->add.bindex < 0) + opt->add.bindex = 0; + goto add; + case Opt_prepend: + opt->add.bindex = 0; + add: /* indented label */ + case Opt_add: + err = au_br_add(sb, &opt->add, + au_ftest_opts(opts->flags, REMOUNT)); + if (!err) { + err = 1; + au_fset_opts(opts->flags, REFRESH); + } + break; + + case Opt_del: + case Opt_idel: + err = au_br_del(sb, &opt->del, + au_ftest_opts(opts->flags, REMOUNT)); + if (!err) { + err = 1; + au_fset_opts(opts->flags, TRUNC_XIB); + au_fset_opts(opts->flags, REFRESH); + } + break; + + case Opt_mod: + case Opt_imod: + err = au_br_mod(sb, &opt->mod, + au_ftest_opts(opts->flags, REMOUNT), + &do_refresh); + if (!err) { + err = 1; + if (do_refresh) + au_fset_opts(opts->flags, REFRESH); + } + break; + } + + return err; +} + +static int au_opt_xino(struct super_block *sb, struct au_opt *opt, + struct au_opt_xino **opt_xino, + struct au_opts *opts) +{ + int err; + aufs_bindex_t bend, bindex; + struct dentry *root, *parent, *h_root; + + err = 0; + switch (opt->type) { + case Opt_xino: + err = au_xino_set(sb, &opt->xino, + !!au_ftest_opts(opts->flags, REMOUNT)); + if (unlikely(err)) + break; + + *opt_xino = &opt->xino; + au_xino_brid_set(sb, -1); + + /* safe d_parent access */ + parent = opt->xino.file->f_path.dentry->d_parent; + root = sb->s_root; + bend = au_sbend(sb); + for (bindex = 0; bindex <= bend; bindex++) { + h_root = au_h_dptr(root, bindex); + if (h_root == parent) { + au_xino_brid_set(sb, au_sbr_id(sb, bindex)); + break; + } + } + break; + + case Opt_noxino: + au_xino_clr(sb); + au_xino_brid_set(sb, -1); + *opt_xino = (void *)-1; + break; + } + + return err; +} + +int au_opts_verify(struct super_block *sb, unsigned long sb_flags, + unsigned int pending) +{ + int err, fhsm; + aufs_bindex_t bindex, bend; + unsigned char do_plink, skip, do_free, can_no_dreval; + struct au_branch *br; + struct au_wbr *wbr; + struct dentry *root, *dentry; + struct inode *dir, *h_dir; + struct au_sbinfo *sbinfo; + struct au_hinode *hdir; + + SiMustAnyLock(sb); + + sbinfo = au_sbi(sb); + AuDebugOn(!(sbinfo->si_mntflags & AuOptMask_UDBA)); + + if (!(sb_flags & MS_RDONLY)) { + if (unlikely(!au_br_writable(au_sbr_perm(sb, 0)))) + pr_warn("first branch should be rw\n"); + if (unlikely(au_opt_test(sbinfo->si_mntflags, SHWH))) + pr_warn_once("shwh should be used with ro\n"); + } + + if (au_opt_test((sbinfo->si_mntflags | pending), UDBA_HNOTIFY) + && !au_opt_test(sbinfo->si_mntflags, XINO)) + pr_warn_once("udba=*notify requires xino\n"); + + if (au_opt_test(sbinfo->si_mntflags, DIRPERM1)) + pr_warn_once("dirperm1 breaks the protection" + " by the permission bits on the lower branch\n"); + + err = 0; + fhsm = 0; + root = sb->s_root; + dir = d_inode(root); + do_plink = !!au_opt_test(sbinfo->si_mntflags, PLINK); + can_no_dreval = !!au_opt_test((sbinfo->si_mntflags | pending), + UDBA_NONE); + bend = au_sbend(sb); + for (bindex = 0; !err && bindex <= bend; bindex++) { + skip = 0; + h_dir = au_h_iptr(dir, bindex); + br = au_sbr(sb, bindex); + + if ((br->br_perm & AuBrAttr_ICEX) + && !h_dir->i_op->listxattr) + br->br_perm &= ~AuBrAttr_ICEX; +#if 0 + if ((br->br_perm & AuBrAttr_ICEX_SEC) + && (au_br_sb(br)->s_flags & MS_NOSEC)) + br->br_perm &= ~AuBrAttr_ICEX_SEC; +#endif + + do_free = 0; + wbr = br->br_wbr; + if (wbr) + wbr_wh_read_lock(wbr); + + if (!au_br_writable(br->br_perm)) { + do_free = !!wbr; + skip = (!wbr + || (!wbr->wbr_whbase + && !wbr->wbr_plink + && !wbr->wbr_orph)); + } else if (!au_br_wh_linkable(br->br_perm)) { + /* skip = (!br->br_whbase && !br->br_orph); */ + skip = (!wbr || !wbr->wbr_whbase); + if (skip && wbr) { + if (do_plink) + skip = !!wbr->wbr_plink; + else + skip = !wbr->wbr_plink; + } + } else { + /* skip = (br->br_whbase && br->br_ohph); */ + skip = (wbr && wbr->wbr_whbase); + if (skip) { + if (do_plink) + skip = !!wbr->wbr_plink; + else + skip = !wbr->wbr_plink; + } + } + if (wbr) + wbr_wh_read_unlock(wbr); + + if (can_no_dreval) { + dentry = br->br_path.dentry; + spin_lock(&dentry->d_lock); + if (dentry->d_flags & + (DCACHE_OP_REVALIDATE | DCACHE_OP_WEAK_REVALIDATE)) + can_no_dreval = 0; + spin_unlock(&dentry->d_lock); + } + + if (au_br_fhsm(br->br_perm)) { + fhsm++; + AuDebugOn(!br->br_fhsm); + } + + if (skip) + continue; + + hdir = au_hi(dir, bindex); + au_hn_imtx_lock_nested(hdir, AuLsc_I_PARENT); + if (wbr) + wbr_wh_write_lock(wbr); + err = au_wh_init(br, sb); + if (wbr) + wbr_wh_write_unlock(wbr); + au_hn_imtx_unlock(hdir); + + if (!err && do_free) { + kfree(wbr); + br->br_wbr = NULL; + } + } + + if (can_no_dreval) + au_fset_si(sbinfo, NO_DREVAL); + else + au_fclr_si(sbinfo, NO_DREVAL); + + if (fhsm >= 2) { + au_fset_si(sbinfo, FHSM); + for (bindex = bend; bindex >= 0; bindex--) { + br = au_sbr(sb, bindex); + if (au_br_fhsm(br->br_perm)) { + au_fhsm_set_bottom(sb, bindex); + break; + } + } + } else { + au_fclr_si(sbinfo, FHSM); + au_fhsm_set_bottom(sb, -1); + } + + return err; +} + +int au_opts_mount(struct super_block *sb, struct au_opts *opts) +{ + int err; + unsigned int tmp; + aufs_bindex_t bindex, bend; + struct au_opt *opt; + struct au_opt_xino *opt_xino, xino; + struct au_sbinfo *sbinfo; + struct au_branch *br; + struct inode *dir; + + SiMustWriteLock(sb); + + err = 0; + opt_xino = NULL; + opt = opts->opt; + while (err >= 0 && opt->type != Opt_tail) + err = au_opt_simple(sb, opt++, opts); + if (err > 0) + err = 0; + else if (unlikely(err < 0)) + goto out; + + /* disable xino and udba temporary */ + sbinfo = au_sbi(sb); + tmp = sbinfo->si_mntflags; + au_opt_clr(sbinfo->si_mntflags, XINO); + au_opt_set_udba(sbinfo->si_mntflags, UDBA_REVAL); + + opt = opts->opt; + while (err >= 0 && opt->type != Opt_tail) + err = au_opt_br(sb, opt++, opts); + if (err > 0) + err = 0; + else if (unlikely(err < 0)) + goto out; + + bend = au_sbend(sb); + if (unlikely(bend < 0)) { + err = -EINVAL; + pr_err("no branches\n"); + goto out; + } + + if (au_opt_test(tmp, XINO)) + au_opt_set(sbinfo->si_mntflags, XINO); + opt = opts->opt; + while (!err && opt->type != Opt_tail) + err = au_opt_xino(sb, opt++, &opt_xino, opts); + if (unlikely(err)) + goto out; + + err = au_opts_verify(sb, sb->s_flags, tmp); + if (unlikely(err)) + goto out; + + /* restore xino */ + if (au_opt_test(tmp, XINO) && !opt_xino) { + xino.file = au_xino_def(sb); + err = PTR_ERR(xino.file); + if (IS_ERR(xino.file)) + goto out; + + err = au_xino_set(sb, &xino, /*remount*/0); + fput(xino.file); + if (unlikely(err)) + goto out; + } + + /* restore udba */ + tmp &= AuOptMask_UDBA; + sbinfo->si_mntflags &= ~AuOptMask_UDBA; + sbinfo->si_mntflags |= tmp; + bend = au_sbend(sb); + for (bindex = 0; bindex <= bend; bindex++) { + br = au_sbr(sb, bindex); + err = au_hnotify_reset_br(tmp, br, br->br_perm); + if (unlikely(err)) + AuIOErr("hnotify failed on br %d, %d, ignored\n", + bindex, err); + /* go on even if err */ + } + if (au_opt_test(tmp, UDBA_HNOTIFY)) { + dir = d_inode(sb->s_root); + au_hn_reset(dir, au_hi_flags(dir, /*isdir*/1) & ~AuHi_XINO); + } + +out: + return err; +} + +int au_opts_remount(struct super_block *sb, struct au_opts *opts) +{ + int err, rerr; + unsigned char no_dreval; + struct inode *dir; + struct au_opt_xino *opt_xino; + struct au_opt *opt; + struct au_sbinfo *sbinfo; + + SiMustWriteLock(sb); + + err = 0; + dir = d_inode(sb->s_root); + sbinfo = au_sbi(sb); + opt_xino = NULL; + opt = opts->opt; + while (err >= 0 && opt->type != Opt_tail) { + err = au_opt_simple(sb, opt, opts); + if (!err) + err = au_opt_br(sb, opt, opts); + if (!err) + err = au_opt_xino(sb, opt, &opt_xino, opts); + opt++; + } + if (err > 0) + err = 0; + AuTraceErr(err); + /* go on even err */ + + no_dreval = !!au_ftest_si(sbinfo, NO_DREVAL); + rerr = au_opts_verify(sb, opts->sb_flags, /*pending*/0); + if (unlikely(rerr && !err)) + err = rerr; + + if (no_dreval != !!au_ftest_si(sbinfo, NO_DREVAL)) + au_fset_opts(opts->flags, REFRESH_IDOP); + + if (au_ftest_opts(opts->flags, TRUNC_XIB)) { + rerr = au_xib_trunc(sb); + if (unlikely(rerr && !err)) + err = rerr; + } + + /* will be handled by the caller */ + if (!au_ftest_opts(opts->flags, REFRESH) + && (opts->given_udba + || au_opt_test(sbinfo->si_mntflags, XINO) + || au_ftest_opts(opts->flags, REFRESH_IDOP) + )) + au_fset_opts(opts->flags, REFRESH); + + AuDbg("status 0x%x\n", opts->flags); + return err; +} + +/* ---------------------------------------------------------------------- */ + +unsigned int au_opt_udba(struct super_block *sb) +{ + return au_mntflags(sb) & AuOptMask_UDBA; +} diff --git a/fs/aufs/opts.h b/fs/aufs/opts.h new file mode 100644 index 000000000..0d6c2e1c7 --- /dev/null +++ b/fs/aufs/opts.h @@ -0,0 +1,198 @@ +/* + * Copyright (C) 2005-2016 Junjiro R. Okajima + */ + +/* + * mount options/flags + */ + +#ifndef __AUFS_OPTS_H__ +#define __AUFS_OPTS_H__ + +#ifdef __KERNEL__ + +#include <linux/path.h> + +struct file; +struct super_block; + +/* ---------------------------------------------------------------------- */ + +/* mount flags */ +#define AuOpt_XINO 1 /* external inode number bitmap + and translation table */ +#define AuOpt_TRUNC_XINO (1 << 1) /* truncate xino files */ +#define AuOpt_UDBA_NONE (1 << 2) /* users direct branch access */ +#define AuOpt_UDBA_REVAL (1 << 3) +#define AuOpt_UDBA_HNOTIFY (1 << 4) +#define AuOpt_SHWH (1 << 5) /* show whiteout */ +#define AuOpt_PLINK (1 << 6) /* pseudo-link */ +#define AuOpt_DIRPERM1 (1 << 7) /* ignore the lower dir's perm + bits */ +#define AuOpt_ALWAYS_DIROPQ (1 << 9) /* policy to creating diropq */ +#define AuOpt_SUM (1 << 10) /* summation for statfs(2) */ +#define AuOpt_SUM_W (1 << 11) /* unimplemented */ +#define AuOpt_WARN_PERM (1 << 12) /* warn when add-branch */ +#define AuOpt_VERBOSE (1 << 13) /* busy inode when del-branch */ +#define AuOpt_DIO (1 << 14) /* direct io */ + +#ifndef CONFIG_AUFS_HNOTIFY +#undef AuOpt_UDBA_HNOTIFY +#define AuOpt_UDBA_HNOTIFY 0 +#endif +#ifndef CONFIG_AUFS_SHWH +#undef AuOpt_SHWH +#define AuOpt_SHWH 0 +#endif + +#define AuOpt_Def (AuOpt_XINO \ + | AuOpt_UDBA_REVAL \ + | AuOpt_PLINK \ + /* | AuOpt_DIRPERM1 */ \ + | AuOpt_WARN_PERM) +#define AuOptMask_UDBA (AuOpt_UDBA_NONE \ + | AuOpt_UDBA_REVAL \ + | AuOpt_UDBA_HNOTIFY) + +#define au_opt_test(flags, name) (flags & AuOpt_##name) +#define au_opt_set(flags, name) do { \ + BUILD_BUG_ON(AuOpt_##name & AuOptMask_UDBA); \ + ((flags) |= AuOpt_##name); \ +} while (0) +#define au_opt_set_udba(flags, name) do { \ + (flags) &= ~AuOptMask_UDBA; \ + ((flags) |= AuOpt_##name); \ +} while (0) +#define au_opt_clr(flags, name) do { \ + ((flags) &= ~AuOpt_##name); \ +} while (0) + +static inline unsigned int au_opts_plink(unsigned int mntflags) +{ +#ifdef CONFIG_PROC_FS + return mntflags; +#else + return mntflags & ~AuOpt_PLINK; +#endif +} + +/* ---------------------------------------------------------------------- */ + +/* policies to select one among multiple writable branches */ +enum { + AuWbrCreate_TDP, /* top down parent */ + AuWbrCreate_RR, /* round robin */ + AuWbrCreate_MFS, /* most free space */ + AuWbrCreate_MFSV, /* mfs with seconds */ + AuWbrCreate_MFSRR, /* mfs then rr */ + AuWbrCreate_MFSRRV, /* mfs then rr with seconds */ + AuWbrCreate_PMFS, /* parent and mfs */ + AuWbrCreate_PMFSV, /* parent and mfs with seconds */ + AuWbrCreate_PMFSRR, /* parent, mfs and round-robin */ + AuWbrCreate_PMFSRRV, /* plus seconds */ + + AuWbrCreate_Def = AuWbrCreate_TDP +}; + +enum { + AuWbrCopyup_TDP, /* top down parent */ + AuWbrCopyup_BUP, /* bottom up parent */ + AuWbrCopyup_BU, /* bottom up */ + + AuWbrCopyup_Def = AuWbrCopyup_TDP +}; + +/* ---------------------------------------------------------------------- */ + +struct au_opt_add { + aufs_bindex_t bindex; + char *pathname; + int perm; + struct path path; +}; + +struct au_opt_del { + char *pathname; + struct path h_path; +}; + +struct au_opt_mod { + char *path; + int perm; + struct dentry *h_root; +}; + +struct au_opt_xino { + char *path; + struct file *file; +}; + +struct au_opt_xino_itrunc { + aufs_bindex_t bindex; +}; + +struct au_opt_wbr_create { + int wbr_create; + int mfs_second; + unsigned long long mfsrr_watermark; +}; + +struct au_opt { + int type; + union { + struct au_opt_xino xino; + struct au_opt_xino_itrunc xino_itrunc; + struct au_opt_add add; + struct au_opt_del del; + struct au_opt_mod mod; + int dirwh; + int rdcache; + unsigned int rdblk; + unsigned int rdhash; + int udba; + struct au_opt_wbr_create wbr_create; + int wbr_copyup; + unsigned int fhsm_second; + }; +}; + +/* opts flags */ +#define AuOpts_REMOUNT 1 +#define AuOpts_REFRESH (1 << 1) +#define AuOpts_TRUNC_XIB (1 << 2) +#define AuOpts_REFRESH_DYAOP (1 << 3) +#define AuOpts_REFRESH_IDOP (1 << 4) +#define au_ftest_opts(flags, name) ((flags) & AuOpts_##name) +#define au_fset_opts(flags, name) \ + do { (flags) |= AuOpts_##name; } while (0) +#define au_fclr_opts(flags, name) \ + do { (flags) &= ~AuOpts_##name; } while (0) + +struct au_opts { + struct au_opt *opt; + int max_opt; + + unsigned int given_udba; + unsigned int flags; + unsigned long sb_flags; +}; + +/* ---------------------------------------------------------------------- */ + +/* opts.c */ +void au_optstr_br_perm(au_br_perm_str_t *str, int perm); +const char *au_optstr_udba(int udba); +const char *au_optstr_wbr_copyup(int wbr_copyup); +const char *au_optstr_wbr_create(int wbr_create); + +void au_opts_free(struct au_opts *opts); +int au_opts_parse(struct super_block *sb, char *str, struct au_opts *opts); +int au_opts_verify(struct super_block *sb, unsigned long sb_flags, + unsigned int pending); +int au_opts_mount(struct super_block *sb, struct au_opts *opts); +int au_opts_remount(struct super_block *sb, struct au_opts *opts); + +unsigned int au_opt_udba(struct super_block *sb); + +#endif /* __KERNEL__ */ +#endif /* __AUFS_OPTS_H__ */ diff --git a/fs/aufs/plink.c b/fs/aufs/plink.c new file mode 100644 index 000000000..d5b36bd11 --- /dev/null +++ b/fs/aufs/plink.c @@ -0,0 +1,515 @@ +/* + * Copyright (C) 2005-2016 Junjiro R. Okajima + */ + +/* + * pseudo-link + */ + +#include "aufs.h" + +/* + * the pseudo-link maintenance mode. + * during a user process maintains the pseudo-links, + * prohibit adding a new plink and branch manipulation. + * + * Flags + * NOPLM: + * For entry functions which will handle plink, and i_mutex is already held + * in VFS. + * They cannot wait and should return an error at once. + * Callers has to check the error. + * NOPLMW: + * For entry functions which will handle plink, but i_mutex is not held + * in VFS. + * They can wait the plink maintenance mode to finish. + * + * They behave like F_SETLK and F_SETLKW. + * If the caller never handle plink, then both flags are unnecessary. + */ + +int au_plink_maint(struct super_block *sb, int flags) +{ + int err; + pid_t pid, ppid; + struct au_sbinfo *sbi; + + SiMustAnyLock(sb); + + err = 0; + if (!au_opt_test(au_mntflags(sb), PLINK)) + goto out; + + sbi = au_sbi(sb); + pid = sbi->si_plink_maint_pid; + if (!pid || pid == current->pid) + goto out; + + /* todo: it highly depends upon /sbin/mount.aufs */ + rcu_read_lock(); + ppid = task_pid_vnr(rcu_dereference(current->real_parent)); + rcu_read_unlock(); + if (pid == ppid) + goto out; + + if (au_ftest_lock(flags, NOPLMW)) { + /* if there is no i_mutex lock in VFS, we don't need to wait */ + /* AuDebugOn(!lockdep_depth(current)); */ + while (sbi->si_plink_maint_pid) { + si_read_unlock(sb); + /* gave up wake_up_bit() */ + wait_event(sbi->si_plink_wq, !sbi->si_plink_maint_pid); + + if (au_ftest_lock(flags, FLUSH)) + au_nwt_flush(&sbi->si_nowait); + si_noflush_read_lock(sb); + } + } else if (au_ftest_lock(flags, NOPLM)) { + AuDbg("ppid %d, pid %d\n", ppid, pid); + err = -EAGAIN; + } + +out: + return err; +} + +void au_plink_maint_leave(struct au_sbinfo *sbinfo) +{ + spin_lock(&sbinfo->si_plink_maint_lock); + sbinfo->si_plink_maint_pid = 0; + spin_unlock(&sbinfo->si_plink_maint_lock); + wake_up_all(&sbinfo->si_plink_wq); +} + +int au_plink_maint_enter(struct super_block *sb) +{ + int err; + struct au_sbinfo *sbinfo; + + err = 0; + sbinfo = au_sbi(sb); + /* make sure i am the only one in this fs */ + si_write_lock(sb, AuLock_FLUSH); + if (au_opt_test(au_mntflags(sb), PLINK)) { + spin_lock(&sbinfo->si_plink_maint_lock); + if (!sbinfo->si_plink_maint_pid) + sbinfo->si_plink_maint_pid = current->pid; + else + err = -EBUSY; + spin_unlock(&sbinfo->si_plink_maint_lock); + } + si_write_unlock(sb); + + return err; +} + +/* ---------------------------------------------------------------------- */ + +#ifdef CONFIG_AUFS_DEBUG +void au_plink_list(struct super_block *sb) +{ + int i; + struct au_sbinfo *sbinfo; + struct hlist_head *plink_hlist; + struct pseudo_link *plink; + + SiMustAnyLock(sb); + + sbinfo = au_sbi(sb); + AuDebugOn(!au_opt_test(au_mntflags(sb), PLINK)); + AuDebugOn(au_plink_maint(sb, AuLock_NOPLM)); + + for (i = 0; i < AuPlink_NHASH; i++) { + plink_hlist = &sbinfo->si_plink[i].head; + rcu_read_lock(); + hlist_for_each_entry_rcu(plink, plink_hlist, hlist) + AuDbg("%lu\n", plink->inode->i_ino); + rcu_read_unlock(); + } +} +#endif + +/* is the inode pseudo-linked? */ +int au_plink_test(struct inode *inode) +{ + int found, i; + struct au_sbinfo *sbinfo; + struct hlist_head *plink_hlist; + struct pseudo_link *plink; + + sbinfo = au_sbi(inode->i_sb); + AuRwMustAnyLock(&sbinfo->si_rwsem); + AuDebugOn(!au_opt_test(au_mntflags(inode->i_sb), PLINK)); + AuDebugOn(au_plink_maint(inode->i_sb, AuLock_NOPLM)); + + found = 0; + i = au_plink_hash(inode->i_ino); + plink_hlist = &sbinfo->si_plink[i].head; + rcu_read_lock(); + hlist_for_each_entry_rcu(plink, plink_hlist, hlist) + if (plink->inode == inode) { + found = 1; + break; + } + rcu_read_unlock(); + return found; +} + +/* ---------------------------------------------------------------------- */ + +/* + * generate a name for plink. + * the file will be stored under AUFS_WH_PLINKDIR. + */ +/* 20 is max digits length of ulong 64 */ +#define PLINK_NAME_LEN ((20 + 1) * 2) + +static int plink_name(char *name, int len, struct inode *inode, + aufs_bindex_t bindex) +{ + int rlen; + struct inode *h_inode; + + h_inode = au_h_iptr(inode, bindex); + rlen = snprintf(name, len, "%lu.%lu", inode->i_ino, h_inode->i_ino); + return rlen; +} + +struct au_do_plink_lkup_args { + struct dentry **errp; + struct qstr *tgtname; + struct dentry *h_parent; + struct au_branch *br; +}; + +static struct dentry *au_do_plink_lkup(struct qstr *tgtname, + struct dentry *h_parent, + struct au_branch *br) +{ + struct dentry *h_dentry; + struct inode *h_inode; + + h_inode = d_inode(h_parent); + inode_lock_nested(h_inode, AuLsc_I_CHILD2); + h_dentry = vfsub_lkup_one(tgtname, h_parent); + inode_unlock(h_inode); + return h_dentry; +} + +static void au_call_do_plink_lkup(void *args) +{ + struct au_do_plink_lkup_args *a = args; + *a->errp = au_do_plink_lkup(a->tgtname, a->h_parent, a->br); +} + +/* lookup the plink-ed @inode under the branch at @bindex */ +struct dentry *au_plink_lkup(struct inode *inode, aufs_bindex_t bindex) +{ + struct dentry *h_dentry, *h_parent; + struct au_branch *br; + int wkq_err; + char a[PLINK_NAME_LEN]; + struct qstr tgtname = QSTR_INIT(a, 0); + + AuDebugOn(au_plink_maint(inode->i_sb, AuLock_NOPLM)); + + br = au_sbr(inode->i_sb, bindex); + h_parent = br->br_wbr->wbr_plink; + tgtname.len = plink_name(a, sizeof(a), inode, bindex); + + if (!uid_eq(current_fsuid(), GLOBAL_ROOT_UID)) { + struct au_do_plink_lkup_args args = { + .errp = &h_dentry, + .tgtname = &tgtname, + .h_parent = h_parent, + .br = br + }; + + wkq_err = au_wkq_wait(au_call_do_plink_lkup, &args); + if (unlikely(wkq_err)) + h_dentry = ERR_PTR(wkq_err); + } else + h_dentry = au_do_plink_lkup(&tgtname, h_parent, br); + + return h_dentry; +} + +/* create a pseudo-link */ +static int do_whplink(struct qstr *tgt, struct dentry *h_parent, + struct dentry *h_dentry, struct au_branch *br) +{ + int err; + struct path h_path = { + .mnt = au_br_mnt(br) + }; + struct inode *h_dir, *delegated; + + h_dir = d_inode(h_parent); + inode_lock_nested(h_dir, AuLsc_I_CHILD2); +again: + h_path.dentry = vfsub_lkup_one(tgt, h_parent); + err = PTR_ERR(h_path.dentry); + if (IS_ERR(h_path.dentry)) + goto out; + + err = 0; + /* wh.plink dir is not monitored */ + /* todo: is it really safe? */ + if (d_is_positive(h_path.dentry) + && d_inode(h_path.dentry) != d_inode(h_dentry)) { + delegated = NULL; + err = vfsub_unlink(h_dir, &h_path, &delegated, /*force*/0); + if (unlikely(err == -EWOULDBLOCK)) { + pr_warn("cannot retry for NFSv4 delegation" + " for an internal unlink\n"); + iput(delegated); + } + dput(h_path.dentry); + h_path.dentry = NULL; + if (!err) + goto again; + } + if (!err && d_is_negative(h_path.dentry)) { + delegated = NULL; + err = vfsub_link(h_dentry, h_dir, &h_path, &delegated); + if (unlikely(err == -EWOULDBLOCK)) { + pr_warn("cannot retry for NFSv4 delegation" + " for an internal link\n"); + iput(delegated); + } + } + dput(h_path.dentry); + +out: + inode_unlock(h_dir); + return err; +} + +struct do_whplink_args { + int *errp; + struct qstr *tgt; + struct dentry *h_parent; + struct dentry *h_dentry; + struct au_branch *br; +}; + +static void call_do_whplink(void *args) +{ + struct do_whplink_args *a = args; + *a->errp = do_whplink(a->tgt, a->h_parent, a->h_dentry, a->br); +} + +static int whplink(struct dentry *h_dentry, struct inode *inode, + aufs_bindex_t bindex, struct au_branch *br) +{ + int err, wkq_err; + struct au_wbr *wbr; + struct dentry *h_parent; + char a[PLINK_NAME_LEN]; + struct qstr tgtname = QSTR_INIT(a, 0); + + wbr = au_sbr(inode->i_sb, bindex)->br_wbr; + h_parent = wbr->wbr_plink; + tgtname.len = plink_name(a, sizeof(a), inode, bindex); + + /* always superio. */ + if (!uid_eq(current_fsuid(), GLOBAL_ROOT_UID)) { + struct do_whplink_args args = { + .errp = &err, + .tgt = &tgtname, + .h_parent = h_parent, + .h_dentry = h_dentry, + .br = br + }; + wkq_err = au_wkq_wait(call_do_whplink, &args); + if (unlikely(wkq_err)) + err = wkq_err; + } else + err = do_whplink(&tgtname, h_parent, h_dentry, br); + + return err; +} + +/* free a single plink */ +static void do_put_plink(struct pseudo_link *plink, int do_del) +{ + if (do_del) + hlist_del(&plink->hlist); + iput(plink->inode); + kfree(plink); +} + +static void do_put_plink_rcu(struct rcu_head *rcu) +{ + struct pseudo_link *plink; + + plink = container_of(rcu, struct pseudo_link, rcu); + iput(plink->inode); + kfree(plink); +} + +/* + * create a new pseudo-link for @h_dentry on @bindex. + * the linked inode is held in aufs @inode. + */ +void au_plink_append(struct inode *inode, aufs_bindex_t bindex, + struct dentry *h_dentry) +{ + struct super_block *sb; + struct au_sbinfo *sbinfo; + struct hlist_head *plink_hlist; + struct pseudo_link *plink, *tmp; + struct au_sphlhead *sphl; + int found, err, cnt, i; + + sb = inode->i_sb; + sbinfo = au_sbi(sb); + AuDebugOn(!au_opt_test(au_mntflags(sb), PLINK)); + AuDebugOn(au_plink_maint(sb, AuLock_NOPLM)); + + found = au_plink_test(inode); + if (found) + return; + + i = au_plink_hash(inode->i_ino); + sphl = sbinfo->si_plink + i; + plink_hlist = &sphl->head; + tmp = kmalloc(sizeof(*plink), GFP_NOFS); + if (tmp) + tmp->inode = au_igrab(inode); + else { + err = -ENOMEM; + goto out; + } + + spin_lock(&sphl->spin); + hlist_for_each_entry(plink, plink_hlist, hlist) { + if (plink->inode == inode) { + found = 1; + break; + } + } + if (!found) + hlist_add_head_rcu(&tmp->hlist, plink_hlist); + spin_unlock(&sphl->spin); + if (!found) { + cnt = au_sphl_count(sphl); +#define msg "unexpectedly unblanced or too many pseudo-links" + if (cnt > AUFS_PLINK_WARN) + AuWarn1(msg ", %d\n", cnt); +#undef msg + err = whplink(h_dentry, inode, bindex, au_sbr(sb, bindex)); + } else { + do_put_plink(tmp, 0); + return; + } + +out: + if (unlikely(err)) { + pr_warn("err %d, damaged pseudo link.\n", err); + if (tmp) { + au_sphl_del_rcu(&tmp->hlist, sphl); + call_rcu(&tmp->rcu, do_put_plink_rcu); + } + } +} + +/* free all plinks */ +void au_plink_put(struct super_block *sb, int verbose) +{ + int i, warned; + struct au_sbinfo *sbinfo; + struct hlist_head *plink_hlist; + struct hlist_node *tmp; + struct pseudo_link *plink; + + SiMustWriteLock(sb); + + sbinfo = au_sbi(sb); + AuDebugOn(!au_opt_test(au_mntflags(sb), PLINK)); + AuDebugOn(au_plink_maint(sb, AuLock_NOPLM)); + + /* no spin_lock since sbinfo is write-locked */ + warned = 0; + for (i = 0; i < AuPlink_NHASH; i++) { + plink_hlist = &sbinfo->si_plink[i].head; + if (!warned && verbose && !hlist_empty(plink_hlist)) { + pr_warn("pseudo-link is not flushed"); + warned = 1; + } + hlist_for_each_entry_safe(plink, tmp, plink_hlist, hlist) + do_put_plink(plink, 0); + INIT_HLIST_HEAD(plink_hlist); + } +} + +void au_plink_clean(struct super_block *sb, int verbose) +{ + struct dentry *root; + + root = sb->s_root; + aufs_write_lock(root); + if (au_opt_test(au_mntflags(sb), PLINK)) + au_plink_put(sb, verbose); + aufs_write_unlock(root); +} + +static int au_plink_do_half_refresh(struct inode *inode, aufs_bindex_t br_id) +{ + int do_put; + aufs_bindex_t bstart, bend, bindex; + + do_put = 0; + bstart = au_ibstart(inode); + bend = au_ibend(inode); + if (bstart >= 0) { + for (bindex = bstart; bindex <= bend; bindex++) { + if (!au_h_iptr(inode, bindex) + || au_ii_br_id(inode, bindex) != br_id) + continue; + au_set_h_iptr(inode, bindex, NULL, 0); + do_put = 1; + break; + } + if (do_put) + for (bindex = bstart; bindex <= bend; bindex++) + if (au_h_iptr(inode, bindex)) { + do_put = 0; + break; + } + } else + do_put = 1; + + return do_put; +} + +/* free the plinks on a branch specified by @br_id */ +void au_plink_half_refresh(struct super_block *sb, aufs_bindex_t br_id) +{ + struct au_sbinfo *sbinfo; + struct hlist_head *plink_hlist; + struct hlist_node *tmp; + struct pseudo_link *plink; + struct inode *inode; + int i, do_put; + + SiMustWriteLock(sb); + + sbinfo = au_sbi(sb); + AuDebugOn(!au_opt_test(au_mntflags(sb), PLINK)); + AuDebugOn(au_plink_maint(sb, AuLock_NOPLM)); + + /* no spin_lock since sbinfo is write-locked */ + for (i = 0; i < AuPlink_NHASH; i++) { + plink_hlist = &sbinfo->si_plink[i].head; + hlist_for_each_entry_safe(plink, tmp, plink_hlist, hlist) { + inode = au_igrab(plink->inode); + ii_write_lock_child(inode); + do_put = au_plink_do_half_refresh(inode, br_id); + if (do_put) + do_put_plink(plink, 1); + ii_write_unlock(inode); + iput(inode); + } + } +} diff --git a/fs/aufs/poll.c b/fs/aufs/poll.c new file mode 100644 index 000000000..dd2baf5dc --- /dev/null +++ b/fs/aufs/poll.c @@ -0,0 +1,39 @@ +/* + * Copyright (C) 2005-2016 Junjiro R. Okajima + */ + +/* + * poll operation + * There is only one filesystem which implements ->poll operation, currently. + */ + +#include "aufs.h" + +unsigned int aufs_poll(struct file *file, poll_table *wait) +{ + unsigned int mask; + int err; + struct file *h_file; + struct super_block *sb; + + /* We should pretend an error happened. */ + mask = POLLERR /* | POLLIN | POLLOUT */; + sb = file->f_path.dentry->d_sb; + si_read_lock(sb, AuLock_FLUSH | AuLock_NOPLMW); + + h_file = au_read_pre(file, /*keep_fi*/0); + err = PTR_ERR(h_file); + if (IS_ERR(h_file)) + goto out; + + /* it is not an error if h_file has no operation */ + mask = DEFAULT_POLLMASK; + if (h_file->f_op->poll) + mask = h_file->f_op->poll(h_file, wait); + fput(h_file); /* instead of au_read_post() */ + +out: + si_read_unlock(sb); + AuTraceErr((int)mask); + return mask; +} diff --git a/fs/aufs/posix_acl.c b/fs/aufs/posix_acl.c new file mode 100644 index 000000000..d6eeebdec --- /dev/null +++ b/fs/aufs/posix_acl.c @@ -0,0 +1,85 @@ +/* + * Copyright (C) 2014-2016 Junjiro R. Okajima + */ + +/* + * posix acl operations + */ + +#include <linux/fs.h> +#include "aufs.h" + +struct posix_acl *aufs_get_acl(struct inode *inode, int type) +{ + struct posix_acl *acl; + int err; + aufs_bindex_t bindex; + struct inode *h_inode; + struct super_block *sb; + + acl = NULL; + sb = inode->i_sb; + si_read_lock(sb, AuLock_FLUSH); + ii_read_lock_child(inode); + if (!(sb->s_flags & MS_POSIXACL)) + goto out; + + bindex = au_ibstart(inode); + h_inode = au_h_iptr(inode, bindex); + if (unlikely(!h_inode + || ((h_inode->i_mode & S_IFMT) + != (inode->i_mode & S_IFMT)))) { + err = au_busy_or_stale(); + acl = ERR_PTR(err); + goto out; + } + + /* always topmost only */ + acl = get_acl(h_inode, type); + +out: + ii_read_unlock(inode); + si_read_unlock(sb); + + AuTraceErrPtr(acl); + return acl; +} + +int aufs_set_acl(struct inode *inode, struct posix_acl *acl, int type) +{ + int err; + ssize_t ssz; + struct dentry *dentry; + struct au_srxattr arg = { + .type = AU_ACL_SET, + .u.acl_set = { + .acl = acl, + .type = type + }, + }; + + inode_lock(inode); + if (inode->i_ino == AUFS_ROOT_INO) + dentry = dget(inode->i_sb->s_root); + else { + dentry = d_find_alias(inode); + if (!dentry) + dentry = d_find_any_alias(inode); + if (!dentry) { + pr_warn("cannot handle this inode, " + "please report to aufs-users ML\n"); + err = -ENOENT; + goto out; + } + } + + ssz = au_srxattr(dentry, &arg); + dput(dentry); + err = ssz; + if (ssz >= 0) + err = 0; + +out: + inode_unlock(inode); + return err; +} diff --git a/fs/aufs/procfs.c b/fs/aufs/procfs.c new file mode 100644 index 000000000..2c8893edf --- /dev/null +++ b/fs/aufs/procfs.c @@ -0,0 +1,156 @@ +/* + * Copyright (C) 2010-2016 Junjiro R. Okajima + */ + +/* + * procfs interfaces + */ + +#include <linux/proc_fs.h> +#include "aufs.h" + +static int au_procfs_plm_release(struct inode *inode, struct file *file) +{ + struct au_sbinfo *sbinfo; + + sbinfo = file->private_data; + if (sbinfo) { + au_plink_maint_leave(sbinfo); + kobject_put(&sbinfo->si_kobj); + } + + return 0; +} + +static void au_procfs_plm_write_clean(struct file *file) +{ + struct au_sbinfo *sbinfo; + + sbinfo = file->private_data; + if (sbinfo) + au_plink_clean(sbinfo->si_sb, /*verbose*/0); +} + +static int au_procfs_plm_write_si(struct file *file, unsigned long id) +{ + int err; + struct super_block *sb; + struct au_sbinfo *sbinfo; + + err = -EBUSY; + if (unlikely(file->private_data)) + goto out; + + sb = NULL; + /* don't use au_sbilist_lock() here */ + spin_lock(&au_sbilist.spin); + list_for_each_entry(sbinfo, &au_sbilist.head, si_list) + if (id == sysaufs_si_id(sbinfo)) { + kobject_get(&sbinfo->si_kobj); + sb = sbinfo->si_sb; + break; + } + spin_unlock(&au_sbilist.spin); + + err = -EINVAL; + if (unlikely(!sb)) + goto out; + + err = au_plink_maint_enter(sb); + if (!err) + /* keep kobject_get() */ + file->private_data = sbinfo; + else + kobject_put(&sbinfo->si_kobj); +out: + return err; +} + +/* + * Accept a valid "si=xxxx" only. + * Once it is accepted successfully, accept "clean" too. + */ +static ssize_t au_procfs_plm_write(struct file *file, const char __user *ubuf, + size_t count, loff_t *ppos) +{ + ssize_t err; + unsigned long id; + /* last newline is allowed */ + char buf[3 + sizeof(unsigned long) * 2 + 1]; + + err = -EACCES; + if (unlikely(!capable(CAP_SYS_ADMIN))) + goto out; + + err = -EINVAL; + if (unlikely(count > sizeof(buf))) + goto out; + + err = copy_from_user(buf, ubuf, count); + if (unlikely(err)) { + err = -EFAULT; + goto out; + } + buf[count] = 0; + + err = -EINVAL; + if (!strcmp("clean", buf)) { + au_procfs_plm_write_clean(file); + goto out_success; + } else if (unlikely(strncmp("si=", buf, 3))) + goto out; + + err = kstrtoul(buf + 3, 16, &id); + if (unlikely(err)) + goto out; + + err = au_procfs_plm_write_si(file, id); + if (unlikely(err)) + goto out; + +out_success: + err = count; /* success */ +out: + return err; +} + +static const struct file_operations au_procfs_plm_fop = { + .write = au_procfs_plm_write, + .release = au_procfs_plm_release, + .owner = THIS_MODULE +}; + +/* ---------------------------------------------------------------------- */ + +static struct proc_dir_entry *au_procfs_dir; + +void au_procfs_fin(void) +{ + remove_proc_entry(AUFS_PLINK_MAINT_NAME, au_procfs_dir); + remove_proc_entry(AUFS_PLINK_MAINT_DIR, NULL); +} + +int __init au_procfs_init(void) +{ + int err; + struct proc_dir_entry *entry; + + err = -ENOMEM; + au_procfs_dir = proc_mkdir(AUFS_PLINK_MAINT_DIR, NULL); + if (unlikely(!au_procfs_dir)) + goto out; + + entry = proc_create(AUFS_PLINK_MAINT_NAME, S_IFREG | S_IWUSR, + au_procfs_dir, &au_procfs_plm_fop); + if (unlikely(!entry)) + goto out_dir; + + err = 0; + goto out; /* success */ + + +out_dir: + remove_proc_entry(AUFS_PLINK_MAINT_DIR, NULL); +out: + return err; +} diff --git a/fs/aufs/rdu.c b/fs/aufs/rdu.c new file mode 100644 index 000000000..ec0844673 --- /dev/null +++ b/fs/aufs/rdu.c @@ -0,0 +1,376 @@ +/* + * Copyright (C) 2005-2016 Junjiro R. Okajima + */ + +/* + * readdir in userspace. + */ + +#include <linux/compat.h> +#include <linux/fs_stack.h> +#include <linux/security.h> +#include "aufs.h" + +/* bits for struct aufs_rdu.flags */ +#define AuRdu_CALLED 1 +#define AuRdu_CONT (1 << 1) +#define AuRdu_FULL (1 << 2) +#define au_ftest_rdu(flags, name) ((flags) & AuRdu_##name) +#define au_fset_rdu(flags, name) \ + do { (flags) |= AuRdu_##name; } while (0) +#define au_fclr_rdu(flags, name) \ + do { (flags) &= ~AuRdu_##name; } while (0) + +struct au_rdu_arg { + struct dir_context ctx; + struct aufs_rdu *rdu; + union au_rdu_ent_ul ent; + unsigned long end; + + struct super_block *sb; + int err; +}; + +static int au_rdu_fill(struct dir_context *ctx, const char *name, int nlen, + loff_t offset, u64 h_ino, unsigned int d_type) +{ + int err, len; + struct au_rdu_arg *arg = container_of(ctx, struct au_rdu_arg, ctx); + struct aufs_rdu *rdu = arg->rdu; + struct au_rdu_ent ent; + + err = 0; + arg->err = 0; + au_fset_rdu(rdu->cookie.flags, CALLED); + len = au_rdu_len(nlen); + if (arg->ent.ul + len < arg->end) { + ent.ino = h_ino; + ent.bindex = rdu->cookie.bindex; + ent.type = d_type; + ent.nlen = nlen; + if (unlikely(nlen > AUFS_MAX_NAMELEN)) + ent.type = DT_UNKNOWN; + + /* unnecessary to support mmap_sem since this is a dir */ + err = -EFAULT; + if (copy_to_user(arg->ent.e, &ent, sizeof(ent))) + goto out; + if (copy_to_user(arg->ent.e->name, name, nlen)) + goto out; + /* the terminating NULL */ + if (__put_user(0, arg->ent.e->name + nlen)) + goto out; + err = 0; + /* AuDbg("%p, %.*s\n", arg->ent.p, nlen, name); */ + arg->ent.ul += len; + rdu->rent++; + } else { + err = -EFAULT; + au_fset_rdu(rdu->cookie.flags, FULL); + rdu->full = 1; + rdu->tail = arg->ent; + } + +out: + /* AuTraceErr(err); */ + return err; +} + +static int au_rdu_do(struct file *h_file, struct au_rdu_arg *arg) +{ + int err; + loff_t offset; + struct au_rdu_cookie *cookie = &arg->rdu->cookie; + + /* we don't have to care (FMODE_32BITHASH | FMODE_64BITHASH) for ext4 */ + offset = vfsub_llseek(h_file, cookie->h_pos, SEEK_SET); + err = offset; + if (unlikely(offset != cookie->h_pos)) + goto out; + + err = 0; + do { + arg->err = 0; + au_fclr_rdu(cookie->flags, CALLED); + /* smp_mb(); */ + err = vfsub_iterate_dir(h_file, &arg->ctx); + if (err >= 0) + err = arg->err; + } while (!err + && au_ftest_rdu(cookie->flags, CALLED) + && !au_ftest_rdu(cookie->flags, FULL)); + cookie->h_pos = h_file->f_pos; + +out: + AuTraceErr(err); + return err; +} + +static int au_rdu(struct file *file, struct aufs_rdu *rdu) +{ + int err; + aufs_bindex_t bend; + struct au_rdu_arg arg = { + .ctx = { + .actor = au_rdu_fill + } + }; + struct dentry *dentry; + struct inode *inode; + struct file *h_file; + struct au_rdu_cookie *cookie = &rdu->cookie; + + err = !access_ok(VERIFY_WRITE, rdu->ent.e, rdu->sz); + if (unlikely(err)) { + err = -EFAULT; + AuTraceErr(err); + goto out; + } + rdu->rent = 0; + rdu->tail = rdu->ent; + rdu->full = 0; + arg.rdu = rdu; + arg.ent = rdu->ent; + arg.end = arg.ent.ul; + arg.end += rdu->sz; + + err = -ENOTDIR; + if (unlikely(!file->f_op->iterate)) + goto out; + + err = security_file_permission(file, MAY_READ); + AuTraceErr(err); + if (unlikely(err)) + goto out; + + dentry = file->f_path.dentry; + inode = d_inode(dentry); +#if 1 + inode_lock(inode); +#else + /* todo: create a new inline func inode_lock_killable() */ + err = mutex_lock_killable(&inode->i_mutex); + AuTraceErr(err); + if (unlikely(err)) + goto out; +#endif + + arg.sb = inode->i_sb; + err = si_read_lock(arg.sb, AuLock_FLUSH | AuLock_NOPLM); + if (unlikely(err)) + goto out_mtx; + err = au_alive_dir(dentry); + if (unlikely(err)) + goto out_si; + /* todo: reval? */ + fi_read_lock(file); + + err = -EAGAIN; + if (unlikely(au_ftest_rdu(cookie->flags, CONT) + && cookie->generation != au_figen(file))) + goto out_unlock; + + err = 0; + if (!rdu->blk) { + rdu->blk = au_sbi(arg.sb)->si_rdblk; + if (!rdu->blk) + rdu->blk = au_dir_size(file, /*dentry*/NULL); + } + bend = au_fbstart(file); + if (cookie->bindex < bend) + cookie->bindex = bend; + bend = au_fbend_dir(file); + /* AuDbg("b%d, b%d\n", cookie->bindex, bend); */ + for (; !err && cookie->bindex <= bend; + cookie->bindex++, cookie->h_pos = 0) { + h_file = au_hf_dir(file, cookie->bindex); + if (!h_file) + continue; + + au_fclr_rdu(cookie->flags, FULL); + err = au_rdu_do(h_file, &arg); + AuTraceErr(err); + if (unlikely(au_ftest_rdu(cookie->flags, FULL) || err)) + break; + } + AuDbg("rent %llu\n", rdu->rent); + + if (!err && !au_ftest_rdu(cookie->flags, CONT)) { + rdu->shwh = !!au_opt_test(au_sbi(arg.sb)->si_mntflags, SHWH); + au_fset_rdu(cookie->flags, CONT); + cookie->generation = au_figen(file); + } + + ii_read_lock_child(inode); + fsstack_copy_attr_atime(inode, au_h_iptr(inode, au_ibstart(inode))); + ii_read_unlock(inode); + +out_unlock: + fi_read_unlock(file); +out_si: + si_read_unlock(arg.sb); +out_mtx: + inode_unlock(inode); +out: + AuTraceErr(err); + return err; +} + +static int au_rdu_ino(struct file *file, struct aufs_rdu *rdu) +{ + int err; + ino_t ino; + unsigned long long nent; + union au_rdu_ent_ul *u; + struct au_rdu_ent ent; + struct super_block *sb; + + err = 0; + nent = rdu->nent; + u = &rdu->ent; + sb = file->f_path.dentry->d_sb; + si_read_lock(sb, AuLock_FLUSH); + while (nent-- > 0) { + /* unnecessary to support mmap_sem since this is a dir */ + err = copy_from_user(&ent, u->e, sizeof(ent)); + if (!err) + err = !access_ok(VERIFY_WRITE, &u->e->ino, sizeof(ino)); + if (unlikely(err)) { + err = -EFAULT; + AuTraceErr(err); + break; + } + + /* AuDbg("b%d, i%llu\n", ent.bindex, ent.ino); */ + if (!ent.wh) + err = au_ino(sb, ent.bindex, ent.ino, ent.type, &ino); + else + err = au_wh_ino(sb, ent.bindex, ent.ino, ent.type, + &ino); + if (unlikely(err)) { + AuTraceErr(err); + break; + } + + err = __put_user(ino, &u->e->ino); + if (unlikely(err)) { + err = -EFAULT; + AuTraceErr(err); + break; + } + u->ul += au_rdu_len(ent.nlen); + } + si_read_unlock(sb); + + return err; +} + +/* ---------------------------------------------------------------------- */ + +static int au_rdu_verify(struct aufs_rdu *rdu) +{ + AuDbg("rdu{%llu, %p, %u | %u | %llu, %u, %u | " + "%llu, b%d, 0x%x, g%u}\n", + rdu->sz, rdu->ent.e, rdu->verify[AufsCtlRduV_SZ], + rdu->blk, + rdu->rent, rdu->shwh, rdu->full, + rdu->cookie.h_pos, rdu->cookie.bindex, rdu->cookie.flags, + rdu->cookie.generation); + + if (rdu->verify[AufsCtlRduV_SZ] == sizeof(*rdu)) + return 0; + + AuDbg("%u:%u\n", + rdu->verify[AufsCtlRduV_SZ], (unsigned int)sizeof(*rdu)); + return -EINVAL; +} + +long au_rdu_ioctl(struct file *file, unsigned int cmd, unsigned long arg) +{ + long err, e; + struct aufs_rdu rdu; + void __user *p = (void __user *)arg; + + err = copy_from_user(&rdu, p, sizeof(rdu)); + if (unlikely(err)) { + err = -EFAULT; + AuTraceErr(err); + goto out; + } + err = au_rdu_verify(&rdu); + if (unlikely(err)) + goto out; + + switch (cmd) { + case AUFS_CTL_RDU: + err = au_rdu(file, &rdu); + if (unlikely(err)) + break; + + e = copy_to_user(p, &rdu, sizeof(rdu)); + if (unlikely(e)) { + err = -EFAULT; + AuTraceErr(err); + } + break; + case AUFS_CTL_RDU_INO: + err = au_rdu_ino(file, &rdu); + break; + + default: + /* err = -ENOTTY; */ + err = -EINVAL; + } + +out: + AuTraceErr(err); + return err; +} + +#ifdef CONFIG_COMPAT +long au_rdu_compat_ioctl(struct file *file, unsigned int cmd, unsigned long arg) +{ + long err, e; + struct aufs_rdu rdu; + void __user *p = compat_ptr(arg); + + /* todo: get_user()? */ + err = copy_from_user(&rdu, p, sizeof(rdu)); + if (unlikely(err)) { + err = -EFAULT; + AuTraceErr(err); + goto out; + } + rdu.ent.e = compat_ptr(rdu.ent.ul); + err = au_rdu_verify(&rdu); + if (unlikely(err)) + goto out; + + switch (cmd) { + case AUFS_CTL_RDU: + err = au_rdu(file, &rdu); + if (unlikely(err)) + break; + + rdu.ent.ul = ptr_to_compat(rdu.ent.e); + rdu.tail.ul = ptr_to_compat(rdu.tail.e); + e = copy_to_user(p, &rdu, sizeof(rdu)); + if (unlikely(e)) { + err = -EFAULT; + AuTraceErr(err); + } + break; + case AUFS_CTL_RDU_INO: + err = au_rdu_ino(file, &rdu); + break; + + default: + /* err = -ENOTTY; */ + err = -EINVAL; + } + +out: + AuTraceErr(err); + return err; +} +#endif diff --git a/fs/aufs/rwsem.h b/fs/aufs/rwsem.h new file mode 100644 index 000000000..ef50c2ccb --- /dev/null +++ b/fs/aufs/rwsem.h @@ -0,0 +1,178 @@ +/* + * Copyright (C) 2005-2016 Junjiro R. Okajima + */ + +/* + * simple read-write semaphore wrappers + */ + +#ifndef __AUFS_RWSEM_H__ +#define __AUFS_RWSEM_H__ + +#ifdef __KERNEL__ + +#include "debug.h" + +struct au_rwsem { + struct rw_semaphore rwsem; +#ifdef CONFIG_AUFS_DEBUG + /* just for debugging, not almighty counter */ + atomic_t rcnt, wcnt; +#endif +}; + +#ifdef CONFIG_AUFS_DEBUG +#define AuDbgCntInit(rw) do { \ + atomic_set(&(rw)->rcnt, 0); \ + atomic_set(&(rw)->wcnt, 0); \ + smp_mb(); /* atomic set */ \ +} while (0) + +#define AuDbgRcntInc(rw) atomic_inc(&(rw)->rcnt) +#define AuDbgRcntDec(rw) WARN_ON(atomic_dec_return(&(rw)->rcnt) < 0) +#define AuDbgWcntInc(rw) atomic_inc(&(rw)->wcnt) +#define AuDbgWcntDec(rw) WARN_ON(atomic_dec_return(&(rw)->wcnt) < 0) +#else +#define AuDbgCntInit(rw) do {} while (0) +#define AuDbgRcntInc(rw) do {} while (0) +#define AuDbgRcntDec(rw) do {} while (0) +#define AuDbgWcntInc(rw) do {} while (0) +#define AuDbgWcntDec(rw) do {} while (0) +#endif /* CONFIG_AUFS_DEBUG */ + +/* to debug easier, do not make them inlined functions */ +#define AuRwMustNoWaiters(rw) AuDebugOn(!list_empty(&(rw)->rwsem.wait_list)) +/* rwsem_is_locked() is unusable */ +#define AuRwMustReadLock(rw) AuDebugOn(atomic_read(&(rw)->rcnt) <= 0) +#define AuRwMustWriteLock(rw) AuDebugOn(atomic_read(&(rw)->wcnt) <= 0) +#define AuRwMustAnyLock(rw) AuDebugOn(atomic_read(&(rw)->rcnt) <= 0 \ + && atomic_read(&(rw)->wcnt) <= 0) +#define AuRwDestroy(rw) AuDebugOn(atomic_read(&(rw)->rcnt) \ + || atomic_read(&(rw)->wcnt)) + +#define au_rw_class(rw, key) lockdep_set_class(&(rw)->rwsem, key) + +static inline void au_rw_init(struct au_rwsem *rw) +{ + AuDbgCntInit(rw); + init_rwsem(&rw->rwsem); +} + +static inline void au_rw_init_wlock(struct au_rwsem *rw) +{ + au_rw_init(rw); + down_write(&rw->rwsem); + AuDbgWcntInc(rw); +} + +static inline void au_rw_init_wlock_nested(struct au_rwsem *rw, + unsigned int lsc) +{ + au_rw_init(rw); + down_write_nested(&rw->rwsem, lsc); + AuDbgWcntInc(rw); +} + +static inline void au_rw_read_lock(struct au_rwsem *rw) +{ + down_read(&rw->rwsem); + AuDbgRcntInc(rw); +} + +static inline void au_rw_read_lock_nested(struct au_rwsem *rw, unsigned int lsc) +{ + down_read_nested(&rw->rwsem, lsc); + AuDbgRcntInc(rw); +} + +static inline void au_rw_read_unlock(struct au_rwsem *rw) +{ + AuRwMustReadLock(rw); + AuDbgRcntDec(rw); + up_read(&rw->rwsem); +} + +static inline void au_rw_dgrade_lock(struct au_rwsem *rw) +{ + AuRwMustWriteLock(rw); + AuDbgRcntInc(rw); + AuDbgWcntDec(rw); + downgrade_write(&rw->rwsem); +} + +static inline void au_rw_write_lock(struct au_rwsem *rw) +{ + down_write(&rw->rwsem); + AuDbgWcntInc(rw); +} + +static inline void au_rw_write_lock_nested(struct au_rwsem *rw, + unsigned int lsc) +{ + down_write_nested(&rw->rwsem, lsc); + AuDbgWcntInc(rw); +} + +static inline void au_rw_write_unlock(struct au_rwsem *rw) +{ + AuRwMustWriteLock(rw); + AuDbgWcntDec(rw); + up_write(&rw->rwsem); +} + +/* why is not _nested version defined */ +static inline int au_rw_read_trylock(struct au_rwsem *rw) +{ + int ret; + + ret = down_read_trylock(&rw->rwsem); + if (ret) + AuDbgRcntInc(rw); + return ret; +} + +static inline int au_rw_write_trylock(struct au_rwsem *rw) +{ + int ret; + + ret = down_write_trylock(&rw->rwsem); + if (ret) + AuDbgWcntInc(rw); + return ret; +} + +#undef AuDbgCntInit +#undef AuDbgRcntInc +#undef AuDbgRcntDec +#undef AuDbgWcntInc +#undef AuDbgWcntDec + +#define AuSimpleLockRwsemFuncs(prefix, param, rwsem) \ +static inline void prefix##_read_lock(param) \ +{ au_rw_read_lock(rwsem); } \ +static inline void prefix##_write_lock(param) \ +{ au_rw_write_lock(rwsem); } \ +static inline int prefix##_read_trylock(param) \ +{ return au_rw_read_trylock(rwsem); } \ +static inline int prefix##_write_trylock(param) \ +{ return au_rw_write_trylock(rwsem); } +/* why is not _nested version defined */ +/* static inline void prefix##_read_trylock_nested(param, lsc) +{ au_rw_read_trylock_nested(rwsem, lsc)); } +static inline void prefix##_write_trylock_nestd(param, lsc) +{ au_rw_write_trylock_nested(rwsem, lsc); } */ + +#define AuSimpleUnlockRwsemFuncs(prefix, param, rwsem) \ +static inline void prefix##_read_unlock(param) \ +{ au_rw_read_unlock(rwsem); } \ +static inline void prefix##_write_unlock(param) \ +{ au_rw_write_unlock(rwsem); } \ +static inline void prefix##_downgrade_lock(param) \ +{ au_rw_dgrade_lock(rwsem); } + +#define AuSimpleRwsemFuncs(prefix, param, rwsem) \ + AuSimpleLockRwsemFuncs(prefix, param, rwsem) \ + AuSimpleUnlockRwsemFuncs(prefix, param, rwsem) + +#endif /* __KERNEL__ */ +#endif /* __AUFS_RWSEM_H__ */ diff --git a/fs/aufs/sbinfo.c b/fs/aufs/sbinfo.c new file mode 100644 index 000000000..a9d782e43 --- /dev/null +++ b/fs/aufs/sbinfo.c @@ -0,0 +1,337 @@ +/* + * Copyright (C) 2005-2016 Junjiro R. Okajima + */ + +/* + * superblock private data + */ + +#include "aufs.h" + +/* + * they are necessary regardless sysfs is disabled. + */ +void au_si_free(struct kobject *kobj) +{ + int i; + struct au_sbinfo *sbinfo; + char *locked __maybe_unused; /* debug only */ + + sbinfo = container_of(kobj, struct au_sbinfo, si_kobj); + for (i = 0; i < AuPlink_NHASH; i++) + AuDebugOn(!hlist_empty(&sbinfo->si_plink[i].head)); + AuDebugOn(atomic_read(&sbinfo->si_nowait.nw_len)); + + au_rw_write_lock(&sbinfo->si_rwsem); + au_br_free(sbinfo); + au_rw_write_unlock(&sbinfo->si_rwsem); + + kfree(sbinfo->si_branch); + for (i = 0; i < AU_NPIDMAP; i++) + kfree(sbinfo->au_si_pid.pid_bitmap[i]); + mutex_destroy(&sbinfo->au_si_pid.pid_mtx); + mutex_destroy(&sbinfo->si_xib_mtx); + AuRwDestroy(&sbinfo->si_rwsem); + + kfree(sbinfo); +} + +int au_si_alloc(struct super_block *sb) +{ + int err, i; + struct au_sbinfo *sbinfo; + static struct lock_class_key aufs_si; + + err = -ENOMEM; + sbinfo = kzalloc(sizeof(*sbinfo), GFP_NOFS); + if (unlikely(!sbinfo)) + goto out; + + /* will be reallocated separately */ + sbinfo->si_branch = kzalloc(sizeof(*sbinfo->si_branch), GFP_NOFS); + if (unlikely(!sbinfo->si_branch)) + goto out_sbinfo; + + err = sysaufs_si_init(sbinfo); + if (unlikely(err)) + goto out_br; + + au_nwt_init(&sbinfo->si_nowait); + au_rw_init_wlock(&sbinfo->si_rwsem); + au_rw_class(&sbinfo->si_rwsem, &aufs_si); + mutex_init(&sbinfo->au_si_pid.pid_mtx); + + atomic_long_set(&sbinfo->si_ninodes, 0); + atomic_long_set(&sbinfo->si_nfiles, 0); + + sbinfo->si_bend = -1; + sbinfo->si_last_br_id = AUFS_BRANCH_MAX / 2; + + sbinfo->si_wbr_copyup = AuWbrCopyup_Def; + sbinfo->si_wbr_create = AuWbrCreate_Def; + sbinfo->si_wbr_copyup_ops = au_wbr_copyup_ops + sbinfo->si_wbr_copyup; + sbinfo->si_wbr_create_ops = au_wbr_create_ops + sbinfo->si_wbr_create; + + au_fhsm_init(sbinfo); + + sbinfo->si_mntflags = au_opts_plink(AuOpt_Def); + + sbinfo->si_xino_jiffy = jiffies; + sbinfo->si_xino_expire + = msecs_to_jiffies(AUFS_XINO_DEF_SEC * MSEC_PER_SEC); + mutex_init(&sbinfo->si_xib_mtx); + sbinfo->si_xino_brid = -1; + /* leave si_xib_last_pindex and si_xib_next_bit */ + + au_sphl_init(&sbinfo->si_aopen); + + sbinfo->si_rdcache = msecs_to_jiffies(AUFS_RDCACHE_DEF * MSEC_PER_SEC); + sbinfo->si_rdblk = AUFS_RDBLK_DEF; + sbinfo->si_rdhash = AUFS_RDHASH_DEF; + sbinfo->si_dirwh = AUFS_DIRWH_DEF; + + for (i = 0; i < AuPlink_NHASH; i++) + au_sphl_init(sbinfo->si_plink + i); + init_waitqueue_head(&sbinfo->si_plink_wq); + spin_lock_init(&sbinfo->si_plink_maint_lock); + + au_sphl_init(&sbinfo->si_files); + + /* with getattr by default */ + sbinfo->si_iop_array = aufs_iop; + + /* leave other members for sysaufs and si_mnt. */ + sbinfo->si_sb = sb; + sb->s_fs_info = sbinfo; + si_pid_set(sb); + return 0; /* success */ + +out_br: + kfree(sbinfo->si_branch); +out_sbinfo: + kfree(sbinfo); +out: + return err; +} + +int au_sbr_realloc(struct au_sbinfo *sbinfo, int nbr) +{ + int err, sz; + struct au_branch **brp; + + AuRwMustWriteLock(&sbinfo->si_rwsem); + + err = -ENOMEM; + sz = sizeof(*brp) * (sbinfo->si_bend + 1); + if (unlikely(!sz)) + sz = sizeof(*brp); + brp = au_kzrealloc(sbinfo->si_branch, sz, sizeof(*brp) * nbr, GFP_NOFS); + if (brp) { + sbinfo->si_branch = brp; + err = 0; + } + + return err; +} + +/* ---------------------------------------------------------------------- */ + +unsigned int au_sigen_inc(struct super_block *sb) +{ + unsigned int gen; + struct inode *inode; + + SiMustWriteLock(sb); + + gen = ++au_sbi(sb)->si_generation; + au_update_digen(sb->s_root); + inode = d_inode(sb->s_root); + au_update_iigen(inode, /*half*/0); + inode->i_version++; + return gen; +} + +aufs_bindex_t au_new_br_id(struct super_block *sb) +{ + aufs_bindex_t br_id; + int i; + struct au_sbinfo *sbinfo; + + SiMustWriteLock(sb); + + sbinfo = au_sbi(sb); + for (i = 0; i <= AUFS_BRANCH_MAX; i++) { + br_id = ++sbinfo->si_last_br_id; + AuDebugOn(br_id < 0); + if (br_id && au_br_index(sb, br_id) < 0) + return br_id; + } + + return -1; +} + +/* ---------------------------------------------------------------------- */ + +/* it is ok that new 'nwt' tasks are appended while we are sleeping */ +int si_read_lock(struct super_block *sb, int flags) +{ + int err; + + err = 0; + if (au_ftest_lock(flags, FLUSH)) + au_nwt_flush(&au_sbi(sb)->si_nowait); + + si_noflush_read_lock(sb); + err = au_plink_maint(sb, flags); + if (unlikely(err)) + si_read_unlock(sb); + + return err; +} + +int si_write_lock(struct super_block *sb, int flags) +{ + int err; + + if (au_ftest_lock(flags, FLUSH)) + au_nwt_flush(&au_sbi(sb)->si_nowait); + + si_noflush_write_lock(sb); + err = au_plink_maint(sb, flags); + if (unlikely(err)) + si_write_unlock(sb); + + return err; +} + +/* dentry and super_block lock. call at entry point */ +int aufs_read_lock(struct dentry *dentry, int flags) +{ + int err; + struct super_block *sb; + + sb = dentry->d_sb; + err = si_read_lock(sb, flags); + if (unlikely(err)) + goto out; + + if (au_ftest_lock(flags, DW)) + di_write_lock_child(dentry); + else + di_read_lock_child(dentry, flags); + + if (au_ftest_lock(flags, GEN)) { + err = au_digen_test(dentry, au_sigen(sb)); + if (!au_opt_test(au_mntflags(sb), UDBA_NONE)) + AuDebugOn(!err && au_dbrange_test(dentry)); + else if (!err) + err = au_dbrange_test(dentry); + if (unlikely(err)) + aufs_read_unlock(dentry, flags); + } + +out: + return err; +} + +void aufs_read_unlock(struct dentry *dentry, int flags) +{ + if (au_ftest_lock(flags, DW)) + di_write_unlock(dentry); + else + di_read_unlock(dentry, flags); + si_read_unlock(dentry->d_sb); +} + +void aufs_write_lock(struct dentry *dentry) +{ + si_write_lock(dentry->d_sb, AuLock_FLUSH | AuLock_NOPLMW); + di_write_lock_child(dentry); +} + +void aufs_write_unlock(struct dentry *dentry) +{ + di_write_unlock(dentry); + si_write_unlock(dentry->d_sb); +} + +int aufs_read_and_write_lock2(struct dentry *d1, struct dentry *d2, int flags) +{ + int err; + unsigned int sigen; + struct super_block *sb; + + sb = d1->d_sb; + err = si_read_lock(sb, flags); + if (unlikely(err)) + goto out; + + di_write_lock2_child(d1, d2, au_ftest_lock(flags, DIRS)); + + if (au_ftest_lock(flags, GEN)) { + sigen = au_sigen(sb); + err = au_digen_test(d1, sigen); + AuDebugOn(!err && au_dbrange_test(d1)); + if (!err) { + err = au_digen_test(d2, sigen); + AuDebugOn(!err && au_dbrange_test(d2)); + } + if (unlikely(err)) + aufs_read_and_write_unlock2(d1, d2); + } + +out: + return err; +} + +void aufs_read_and_write_unlock2(struct dentry *d1, struct dentry *d2) +{ + di_write_unlock2(d1, d2); + si_read_unlock(d1->d_sb); +} + +/* ---------------------------------------------------------------------- */ + +static void si_pid_alloc(struct au_si_pid *au_si_pid, int idx) +{ + unsigned long *p; + + BUILD_BUG_ON(sizeof(unsigned long) != + sizeof(*au_si_pid->pid_bitmap)); + + mutex_lock(&au_si_pid->pid_mtx); + p = au_si_pid->pid_bitmap[idx]; + while (!p) { + /* + * bad approach. + * but keeping 'si_pid_set()' void is more important. + */ + p = kcalloc(BITS_TO_LONGS(AU_PIDSTEP), + sizeof(*au_si_pid->pid_bitmap), + GFP_NOFS); + if (p) + break; + cond_resched(); + } + au_si_pid->pid_bitmap[idx] = p; + mutex_unlock(&au_si_pid->pid_mtx); +} + +void si_pid_set(struct super_block *sb) +{ + pid_t bit; + int idx; + unsigned long *bitmap; + struct au_si_pid *au_si_pid; + + si_pid_idx_bit(&idx, &bit); + au_si_pid = &au_sbi(sb)->au_si_pid; + bitmap = au_si_pid->pid_bitmap[idx]; + if (!bitmap) { + si_pid_alloc(au_si_pid, idx); + bitmap = au_si_pid->pid_bitmap[idx]; + } + AuDebugOn(test_bit(bit, bitmap)); + set_bit(bit, bitmap); + /* smp_mb(); */ +} diff --git a/fs/aufs/spl.h b/fs/aufs/spl.h new file mode 100644 index 000000000..f9b528826 --- /dev/null +++ b/fs/aufs/spl.h @@ -0,0 +1,98 @@ +/* + * Copyright (C) 2005-2016 Junjiro R. Okajima + */ + +/* + * simple list protected by a spinlock + */ + +#ifndef __AUFS_SPL_H__ +#define __AUFS_SPL_H__ + +#ifdef __KERNEL__ + +struct au_splhead { + spinlock_t spin; + struct list_head head; +}; + +static inline void au_spl_init(struct au_splhead *spl) +{ + spin_lock_init(&spl->spin); + INIT_LIST_HEAD(&spl->head); +} + +static inline void au_spl_add(struct list_head *list, struct au_splhead *spl) +{ + spin_lock(&spl->spin); + list_add(list, &spl->head); + spin_unlock(&spl->spin); +} + +static inline void au_spl_del(struct list_head *list, struct au_splhead *spl) +{ + spin_lock(&spl->spin); + list_del(list); + spin_unlock(&spl->spin); +} + +static inline void au_spl_del_rcu(struct list_head *list, + struct au_splhead *spl) +{ + spin_lock(&spl->spin); + list_del_rcu(list); + spin_unlock(&spl->spin); +} + +/* ---------------------------------------------------------------------- */ + +struct au_sphlhead { + spinlock_t spin; + struct hlist_head head; +}; + +static inline void au_sphl_init(struct au_sphlhead *sphl) +{ + spin_lock_init(&sphl->spin); + INIT_HLIST_HEAD(&sphl->head); +} + +static inline void au_sphl_add(struct hlist_node *hlist, + struct au_sphlhead *sphl) +{ + spin_lock(&sphl->spin); + hlist_add_head(hlist, &sphl->head); + spin_unlock(&sphl->spin); +} + +static inline void au_sphl_del(struct hlist_node *hlist, + struct au_sphlhead *sphl) +{ + spin_lock(&sphl->spin); + hlist_del(hlist); + spin_unlock(&sphl->spin); +} + +static inline void au_sphl_del_rcu(struct hlist_node *hlist, + struct au_sphlhead *sphl) +{ + spin_lock(&sphl->spin); + hlist_del_rcu(hlist); + spin_unlock(&sphl->spin); +} + +static inline unsigned long au_sphl_count(struct au_sphlhead *sphl) +{ + unsigned long cnt; + struct hlist_node *pos; + + cnt = 0; + spin_lock(&sphl->spin); + hlist_for_each(pos, &sphl->head) + cnt++; + spin_unlock(&sphl->spin); + return cnt; +} + +#endif /* __KERNEL__ */ +#endif /* __AUFS_SPL_H__ */ diff --git a/fs/aufs/super.c b/fs/aufs/super.c new file mode 100644 index 000000000..7928a5031 --- /dev/null +++ b/fs/aufs/super.c @@ -0,0 +1,1026 @@ +/* + * Copyright (C) 2005-2016 Junjiro R. Okajima + */ + +/* + * mount and super_block operations + */ + +#include <linux/mm.h> +#include <linux/seq_file.h> +#include <linux/statfs.h> +#include <linux/vmalloc.h> +#include "aufs.h" + +/* + * super_operations + */ +static struct inode *aufs_alloc_inode(struct super_block *sb __maybe_unused) +{ + struct au_icntnr *c; + + c = au_cache_alloc_icntnr(); + if (c) { + au_icntnr_init(c); + c->vfs_inode.i_version = 1; /* sigen(sb); */ + c->iinfo.ii_hinode = NULL; + return &c->vfs_inode; + } + return NULL; +} + +static void aufs_destroy_inode_cb(struct rcu_head *head) +{ + struct inode *inode = container_of(head, struct inode, i_rcu); + + INIT_HLIST_HEAD(&inode->i_dentry); + au_cache_free_icntnr(container_of(inode, struct au_icntnr, vfs_inode)); +} + +static void aufs_destroy_inode(struct inode *inode) +{ + au_iinfo_fin(inode); + call_rcu(&inode->i_rcu, aufs_destroy_inode_cb); +} + +struct inode *au_iget_locked(struct super_block *sb, ino_t ino) +{ + struct inode *inode; + int err; + + inode = iget_locked(sb, ino); + if (unlikely(!inode)) { + inode = ERR_PTR(-ENOMEM); + goto out; + } + if (!(inode->i_state & I_NEW)) + goto out; + + err = au_xigen_new(inode); + if (!err) + err = au_iinfo_init(inode); + if (!err) + inode->i_version++; + else { + iget_failed(inode); + inode = ERR_PTR(err); + } + +out: + /* never return NULL */ + AuDebugOn(!inode); + AuTraceErrPtr(inode); + return inode; +} + +/* lock free root dinfo */ +static int au_show_brs(struct seq_file *seq, struct super_block *sb) +{ + int err; + aufs_bindex_t bindex, bend; + struct path path; + struct au_hdentry *hdp; + struct au_branch *br; + au_br_perm_str_t perm; + + err = 0; + bend = au_sbend(sb); + hdp = au_di(sb->s_root)->di_hdentry; + for (bindex = 0; !err && bindex <= bend; bindex++) { + br = au_sbr(sb, bindex); + path.mnt = au_br_mnt(br); + path.dentry = hdp[bindex].hd_dentry; + err = au_seq_path(seq, &path); + if (!err) { + au_optstr_br_perm(&perm, br->br_perm); + seq_printf(seq, "=%s", perm.a); + if (bindex != bend) + seq_putc(seq, ':'); + } + } + if (unlikely(err || seq_has_overflowed(seq))) + err = -E2BIG; + + return err; +} + +static void au_show_wbr_create(struct seq_file *m, int v, + struct au_sbinfo *sbinfo) +{ + const char *pat; + + AuRwMustAnyLock(&sbinfo->si_rwsem); + + seq_puts(m, ",create="); + pat = au_optstr_wbr_create(v); + switch (v) { + case AuWbrCreate_TDP: + case AuWbrCreate_RR: + case AuWbrCreate_MFS: + case AuWbrCreate_PMFS: + seq_puts(m, pat); + break; + case AuWbrCreate_MFSV: + seq_printf(m, /*pat*/"mfs:%lu", + jiffies_to_msecs(sbinfo->si_wbr_mfs.mfs_expire) + / MSEC_PER_SEC); + break; + case AuWbrCreate_PMFSV: + seq_printf(m, /*pat*/"pmfs:%lu", + jiffies_to_msecs(sbinfo->si_wbr_mfs.mfs_expire) + / MSEC_PER_SEC); + break; + case AuWbrCreate_MFSRR: + seq_printf(m, /*pat*/"mfsrr:%llu", + sbinfo->si_wbr_mfs.mfsrr_watermark); + break; + case AuWbrCreate_MFSRRV: + seq_printf(m, /*pat*/"mfsrr:%llu:%lu", + sbinfo->si_wbr_mfs.mfsrr_watermark, + jiffies_to_msecs(sbinfo->si_wbr_mfs.mfs_expire) + / MSEC_PER_SEC); + break; + case AuWbrCreate_PMFSRR: + seq_printf(m, /*pat*/"pmfsrr:%llu", + sbinfo->si_wbr_mfs.mfsrr_watermark); + break; + case AuWbrCreate_PMFSRRV: + seq_printf(m, /*pat*/"pmfsrr:%llu:%lu", + sbinfo->si_wbr_mfs.mfsrr_watermark, + jiffies_to_msecs(sbinfo->si_wbr_mfs.mfs_expire) + / MSEC_PER_SEC); + break; + } +} + +static int au_show_xino(struct seq_file *seq, struct super_block *sb) +{ +#ifdef CONFIG_SYSFS + return 0; +#else + int err; + const int len = sizeof(AUFS_XINO_FNAME) - 1; + aufs_bindex_t bindex, brid; + struct qstr *name; + struct file *f; + struct dentry *d, *h_root; + struct au_hdentry *hdp; + + AuRwMustAnyLock(&sbinfo->si_rwsem); + + err = 0; + f = au_sbi(sb)->si_xib; + if (!f) + goto out; + + /* stop printing the default xino path on the first writable branch */ + h_root = NULL; + brid = au_xino_brid(sb); + if (brid >= 0) { + bindex = au_br_index(sb, brid); + hdp = au_di(sb->s_root)->di_hdentry; + h_root = hdp[0 + bindex].hd_dentry; + } + d = f->f_path.dentry; + name = &d->d_name; + /* safe ->d_parent because the file is unlinked */ + if (d->d_parent == h_root + && name->len == len + && !memcmp(name->name, AUFS_XINO_FNAME, len)) + goto out; + + seq_puts(seq, ",xino="); + err = au_xino_path(seq, f); + +out: + return err; +#endif +} + +/* seq_file will re-call me in case of too long string */ +static int aufs_show_options(struct seq_file *m, struct dentry *dentry) +{ + int err; + unsigned int mnt_flags, v; + struct super_block *sb; + struct au_sbinfo *sbinfo; + +#define AuBool(name, str) do { \ + v = au_opt_test(mnt_flags, name); \ + if (v != au_opt_test(AuOpt_Def, name)) \ + seq_printf(m, ",%s" #str, v ? "" : "no"); \ +} while (0) + +#define AuStr(name, str) do { \ + v = mnt_flags & AuOptMask_##name; \ + if (v != (AuOpt_Def & AuOptMask_##name)) \ + seq_printf(m, "," #str "=%s", au_optstr_##str(v)); \ +} while (0) + +#define AuUInt(name, str, val) do { \ + if (val != AUFS_##name##_DEF) \ + seq_printf(m, "," #str "=%u", val); \ +} while (0) + + sb = dentry->d_sb; + if (sb->s_flags & MS_POSIXACL) + seq_puts(m, ",acl"); + + /* lock free root dinfo */ + si_noflush_read_lock(sb); + sbinfo = au_sbi(sb); + seq_printf(m, ",si=%lx", sysaufs_si_id(sbinfo)); + + mnt_flags = au_mntflags(sb); + if (au_opt_test(mnt_flags, XINO)) { + err = au_show_xino(m, sb); + if (unlikely(err)) + goto out; + } else + seq_puts(m, ",noxino"); + + AuBool(TRUNC_XINO, trunc_xino); + AuStr(UDBA, udba); + AuBool(SHWH, shwh); + AuBool(PLINK, plink); + AuBool(DIO, dio); + AuBool(DIRPERM1, dirperm1); + + v = sbinfo->si_wbr_create; + if (v != AuWbrCreate_Def) + au_show_wbr_create(m, v, sbinfo); + + v = sbinfo->si_wbr_copyup; + if (v != AuWbrCopyup_Def) + seq_printf(m, ",cpup=%s", au_optstr_wbr_copyup(v)); + + v = au_opt_test(mnt_flags, ALWAYS_DIROPQ); + if (v != au_opt_test(AuOpt_Def, ALWAYS_DIROPQ)) + seq_printf(m, ",diropq=%c", v ? 'a' : 'w'); + + AuUInt(DIRWH, dirwh, sbinfo->si_dirwh); + + v = jiffies_to_msecs(sbinfo->si_rdcache) / MSEC_PER_SEC; + AuUInt(RDCACHE, rdcache, v); + + AuUInt(RDBLK, rdblk, sbinfo->si_rdblk); + AuUInt(RDHASH, rdhash, sbinfo->si_rdhash); + + au_fhsm_show(m, sbinfo); + + AuBool(SUM, sum); + /* AuBool(SUM_W, wsum); */ + AuBool(WARN_PERM, warn_perm); + AuBool(VERBOSE, verbose); + +out: + /* be sure to print "br:" last */ + if (!sysaufs_brs) { + seq_puts(m, ",br:"); + au_show_brs(m, sb); + } + si_read_unlock(sb); + return 0; + +#undef AuBool +#undef AuStr +#undef AuUInt +} + +/* ---------------------------------------------------------------------- */ + +/* sum mode which returns the summation for statfs(2) */ + +static u64 au_add_till_max(u64 a, u64 b) +{ + u64 old; + + old = a; + a += b; + if (old <= a) + return a; + return ULLONG_MAX; +} + +static u64 au_mul_till_max(u64 a, long mul) +{ + u64 old; + + old = a; + a *= mul; + if (old <= a) + return a; + return ULLONG_MAX; +} + +static int au_statfs_sum(struct super_block *sb, struct kstatfs *buf) +{ + int err; + long bsize, factor; + u64 blocks, bfree, bavail, files, ffree; + aufs_bindex_t bend, bindex, i; + unsigned char shared; + struct path h_path; + struct super_block *h_sb; + + err = 0; + bsize = LONG_MAX; + files = 0; + ffree = 0; + blocks = 0; + bfree = 0; + bavail = 0; + bend = au_sbend(sb); + for (bindex = 0; bindex <= bend; bindex++) { + h_path.mnt = au_sbr_mnt(sb, bindex); + h_sb = h_path.mnt->mnt_sb; + shared = 0; + for (i = 0; !shared && i < bindex; i++) + shared = (au_sbr_sb(sb, i) == h_sb); + if (shared) + continue; + + /* sb->s_root for NFS is unreliable */ + h_path.dentry = h_path.mnt->mnt_root; + err = vfs_statfs(&h_path, buf); + if (unlikely(err)) + goto out; + + if (bsize > buf->f_bsize) { + /* + * we will reduce bsize, so we have to expand blocks + * etc. to match them again + */ + factor = (bsize / buf->f_bsize); + blocks = au_mul_till_max(blocks, factor); + bfree = au_mul_till_max(bfree, factor); + bavail = au_mul_till_max(bavail, factor); + bsize = buf->f_bsize; + } + + factor = (buf->f_bsize / bsize); + blocks = au_add_till_max(blocks, + au_mul_till_max(buf->f_blocks, factor)); + bfree = au_add_till_max(bfree, + au_mul_till_max(buf->f_bfree, factor)); + bavail = au_add_till_max(bavail, + au_mul_till_max(buf->f_bavail, factor)); + files = au_add_till_max(files, buf->f_files); + ffree = au_add_till_max(ffree, buf->f_ffree); + } + + buf->f_bsize = bsize; + buf->f_blocks = blocks; + buf->f_bfree = bfree; + buf->f_bavail = bavail; + buf->f_files = files; + buf->f_ffree = ffree; + buf->f_frsize = 0; + +out: + return err; +} + +static int aufs_statfs(struct dentry *dentry, struct kstatfs *buf) +{ + int err; + struct path h_path; + struct super_block *sb; + + /* lock free root dinfo */ + sb = dentry->d_sb; + si_noflush_read_lock(sb); + if (!au_opt_test(au_mntflags(sb), SUM)) { + /* sb->s_root for NFS is unreliable */ + h_path.mnt = au_sbr_mnt(sb, 0); + h_path.dentry = h_path.mnt->mnt_root; + err = vfs_statfs(&h_path, buf); + } else + err = au_statfs_sum(sb, buf); + si_read_unlock(sb); + + if (!err) { + buf->f_type = AUFS_SUPER_MAGIC; + buf->f_namelen = AUFS_MAX_NAMELEN; + memset(&buf->f_fsid, 0, sizeof(buf->f_fsid)); + } + /* buf->f_bsize = buf->f_blocks = buf->f_bfree = buf->f_bavail = -1; */ + + return err; +} + +/* ---------------------------------------------------------------------- */ + +static int aufs_sync_fs(struct super_block *sb, int wait) +{ + int err, e; + aufs_bindex_t bend, bindex; + struct au_branch *br; + struct super_block *h_sb; + + err = 0; + si_noflush_read_lock(sb); + bend = au_sbend(sb); + for (bindex = 0; bindex <= bend; bindex++) { + br = au_sbr(sb, bindex); + if (!au_br_writable(br->br_perm)) + continue; + + h_sb = au_sbr_sb(sb, bindex); + if (h_sb->s_op->sync_fs) { + e = h_sb->s_op->sync_fs(h_sb, wait); + if (unlikely(e && !err)) + err = e; + /* go on even if an error happens */ + } + } + si_read_unlock(sb); + + return err; +} + +/* ---------------------------------------------------------------------- */ + +/* final actions when unmounting a file system */ +static void aufs_put_super(struct super_block *sb) +{ + struct au_sbinfo *sbinfo; + + sbinfo = au_sbi(sb); + if (!sbinfo) + return; + + dbgaufs_si_fin(sbinfo); + kobject_put(&sbinfo->si_kobj); +} + +/* ---------------------------------------------------------------------- */ + +void *au_array_alloc(unsigned long long *hint, au_arraycb_t cb, + struct super_block *sb, void *arg) +{ + void *array; + unsigned long long n, sz; + + array = NULL; + n = 0; + if (!*hint) + goto out; + + if (*hint > ULLONG_MAX / sizeof(array)) { + array = ERR_PTR(-EMFILE); + pr_err("hint %llu\n", *hint); + goto out; + } + + sz = sizeof(array) * *hint; + array = kzalloc(sz, GFP_NOFS); + if (unlikely(!array)) + array = vzalloc(sz); + if (unlikely(!array)) { + array = ERR_PTR(-ENOMEM); + goto out; + } + + n = cb(sb, array, *hint, arg); + AuDebugOn(n > *hint); + +out: + *hint = n; + return array; +} + +static unsigned long long au_iarray_cb(struct super_block *sb, void *a, + unsigned long long max __maybe_unused, + void *arg) +{ + unsigned long long n; + struct inode **p, *inode; + struct list_head *head; + + n = 0; + p = a; + head = arg; + spin_lock(&sb->s_inode_list_lock); + list_for_each_entry(inode, head, i_sb_list) { + if (!is_bad_inode(inode) + && au_ii(inode)->ii_bstart >= 0) { + spin_lock(&inode->i_lock); + if (atomic_read(&inode->i_count)) { + au_igrab(inode); + *p++ = inode; + n++; + AuDebugOn(n > max); + } + spin_unlock(&inode->i_lock); + } + } + spin_unlock(&sb->s_inode_list_lock); + + return n; +} + +struct inode **au_iarray_alloc(struct super_block *sb, unsigned long long *max) +{ + *max = atomic_long_read(&au_sbi(sb)->si_ninodes); + return au_array_alloc(max, au_iarray_cb, sb, &sb->s_inodes); +} + +void au_iarray_free(struct inode **a, unsigned long long max) +{ + unsigned long long ull; + + for (ull = 0; ull < max; ull++) + iput(a[ull]); + kvfree(a); +} + +/* ---------------------------------------------------------------------- */ + +/* + * refresh dentry and inode at remount time. + */ +/* todo: consolidate with simple_reval_dpath() and au_reval_for_attr() */ +static int au_do_refresh(struct dentry *dentry, unsigned int dir_flags, + struct dentry *parent) +{ + int err; + + di_write_lock_child(dentry); + di_read_lock_parent(parent, AuLock_IR); + err = au_refresh_dentry(dentry, parent); + if (!err && dir_flags) + au_hn_reset(d_inode(dentry), dir_flags); + di_read_unlock(parent, AuLock_IR); + di_write_unlock(dentry); + + return err; +} + +static int au_do_refresh_d(struct dentry *dentry, unsigned int sigen, + struct au_sbinfo *sbinfo, + const unsigned int dir_flags, unsigned int do_idop) +{ + int err; + struct dentry *parent; + + err = 0; + parent = dget_parent(dentry); + if (!au_digen_test(parent, sigen) && au_digen_test(dentry, sigen)) { + if (d_really_is_positive(dentry)) { + if (!d_is_dir(dentry)) + err = au_do_refresh(dentry, /*dir_flags*/0, + parent); + else { + err = au_do_refresh(dentry, dir_flags, parent); + if (unlikely(err)) + au_fset_si(sbinfo, FAILED_REFRESH_DIR); + } + } else + err = au_do_refresh(dentry, /*dir_flags*/0, parent); + AuDbgDentry(dentry); + } + dput(parent); + + if (!err) { + if (do_idop) + au_refresh_dop(dentry, /*force_reval*/0); + } else + au_refresh_dop(dentry, /*force_reval*/1); + + AuTraceErr(err); + return err; +} + +static int au_refresh_d(struct super_block *sb, unsigned int do_idop) +{ + int err, i, j, ndentry, e; + unsigned int sigen; + struct au_dcsub_pages dpages; + struct au_dpage *dpage; + struct dentry **dentries, *d; + struct au_sbinfo *sbinfo; + struct dentry *root = sb->s_root; + const unsigned int dir_flags = au_hi_flags(d_inode(root), /*isdir*/1); + + if (do_idop) + au_refresh_dop(root, /*force_reval*/0); + + err = au_dpages_init(&dpages, GFP_NOFS); + if (unlikely(err)) + goto out; + err = au_dcsub_pages(&dpages, root, NULL, NULL); + if (unlikely(err)) + goto out_dpages; + + sigen = au_sigen(sb); + sbinfo = au_sbi(sb); + for (i = 0; i < dpages.ndpage; i++) { + dpage = dpages.dpages + i; + dentries = dpage->dentries; + ndentry = dpage->ndentry; + for (j = 0; j < ndentry; j++) { + d = dentries[j]; + e = au_do_refresh_d(d, sigen, sbinfo, dir_flags, + do_idop); + if (unlikely(e && !err)) + err = e; + /* go on even err */ + } + } + +out_dpages: + au_dpages_free(&dpages); +out: + return err; +} + +static int au_refresh_i(struct super_block *sb, unsigned int do_idop) +{ + int err, e; + unsigned int sigen; + unsigned long long max, ull; + struct inode *inode, **array; + + array = au_iarray_alloc(sb, &max); + err = PTR_ERR(array); + if (IS_ERR(array)) + goto out; + + err = 0; + sigen = au_sigen(sb); + for (ull = 0; ull < max; ull++) { + inode = array[ull]; + if (unlikely(!inode)) + break; + + e = 0; + ii_write_lock_child(inode); + if (au_iigen(inode, NULL) != sigen) { + e = au_refresh_hinode_self(inode); + if (unlikely(e)) { + au_refresh_iop(inode, /*force_getattr*/1); + pr_err("error %d, i%lu\n", e, inode->i_ino); + if (!err) + err = e; + /* go on even if err */ + } + } + if (!e && do_idop) + au_refresh_iop(inode, /*force_getattr*/0); + ii_write_unlock(inode); + } + + au_iarray_free(array, max); + +out: + return err; +} + +static void au_remount_refresh(struct super_block *sb, unsigned int do_idop) +{ + int err, e; + unsigned int udba; + aufs_bindex_t bindex, bend; + struct dentry *root; + struct inode *inode; + struct au_branch *br; + struct au_sbinfo *sbi; + + au_sigen_inc(sb); + sbi = au_sbi(sb); + au_fclr_si(sbi, FAILED_REFRESH_DIR); + + root = sb->s_root; + DiMustNoWaiters(root); + inode = d_inode(root); + IiMustNoWaiters(inode); + + udba = au_opt_udba(sb); + bend = au_sbend(sb); + for (bindex = 0; bindex <= bend; bindex++) { + br = au_sbr(sb, bindex); + err = au_hnotify_reset_br(udba, br, br->br_perm); + if (unlikely(err)) + AuIOErr("hnotify failed on br %d, %d, ignored\n", + bindex, err); + /* go on even if err */ + } + au_hn_reset(inode, au_hi_flags(inode, /*isdir*/1)); + + if (do_idop) { + if (au_ftest_si(sbi, NO_DREVAL)) { + AuDebugOn(sb->s_d_op == &aufs_dop_noreval); + sb->s_d_op = &aufs_dop_noreval; + AuDebugOn(sbi->si_iop_array == aufs_iop_nogetattr); + sbi->si_iop_array = aufs_iop_nogetattr; + } else { + AuDebugOn(sb->s_d_op == &aufs_dop); + sb->s_d_op = &aufs_dop; + AuDebugOn(sbi->si_iop_array == aufs_iop); + sbi->si_iop_array = aufs_iop; + } + pr_info("reset to %pf and %pf\n", + sb->s_d_op, sbi->si_iop_array); + } + + di_write_unlock(root); + err = au_refresh_d(sb, do_idop); + e = au_refresh_i(sb, do_idop); + if (unlikely(e && !err)) + err = e; + /* aufs_write_lock() calls ..._child() */ + di_write_lock_child(root); + + au_cpup_attr_all(inode, /*force*/1); + + if (unlikely(err)) + AuIOErr("refresh failed, ignored, %d\n", err); +} + +/* stop extra interpretation of errno in mount(8), and strange error messages */ +static int cvt_err(int err) +{ + AuTraceErr(err); + + switch (err) { + case -ENOENT: + case -ENOTDIR: + case -EEXIST: + case -EIO: + err = -EINVAL; + } + return err; +} + +static int aufs_remount_fs(struct super_block *sb, int *flags, char *data) +{ + int err, do_dx; + unsigned int mntflags; + struct au_opts opts = { + .opt = NULL + }; + struct dentry *root; + struct inode *inode; + struct au_sbinfo *sbinfo; + + err = 0; + root = sb->s_root; + if (!data || !*data) { + err = si_write_lock(sb, AuLock_FLUSH | AuLock_NOPLM); + if (!err) { + di_write_lock_child(root); + err = au_opts_verify(sb, *flags, /*pending*/0); + aufs_write_unlock(root); + } + goto out; + } + + err = -ENOMEM; + opts.opt = (void *)__get_free_page(GFP_NOFS); + if (unlikely(!opts.opt)) + goto out; + opts.max_opt = PAGE_SIZE / sizeof(*opts.opt); + opts.flags = AuOpts_REMOUNT; + opts.sb_flags = *flags; + + /* parse it before aufs lock */ + err = au_opts_parse(sb, data, &opts); + if (unlikely(err)) + goto out_opts; + + sbinfo = au_sbi(sb); + inode = d_inode(root); + inode_lock(inode); + err = si_write_lock(sb, AuLock_FLUSH | AuLock_NOPLM); + if (unlikely(err)) + goto out_mtx; + di_write_lock_child(root); + + /* au_opts_remount() may return an error */ + err = au_opts_remount(sb, &opts); + au_opts_free(&opts); + + if (au_ftest_opts(opts.flags, REFRESH)) + au_remount_refresh(sb, au_ftest_opts(opts.flags, REFRESH_IDOP)); + + if (au_ftest_opts(opts.flags, REFRESH_DYAOP)) { + mntflags = au_mntflags(sb); + do_dx = !!au_opt_test(mntflags, DIO); + au_dy_arefresh(do_dx); + } + + au_fhsm_wrote_all(sb, /*force*/1); /* ?? */ + aufs_write_unlock(root); + +out_mtx: + inode_unlock(inode); +out_opts: + free_page((unsigned long)opts.opt); +out: + err = cvt_err(err); + AuTraceErr(err); + return err; +} + +static const struct super_operations aufs_sop = { + .alloc_inode = aufs_alloc_inode, + .destroy_inode = aufs_destroy_inode, + /* always deleting, no clearing */ + .drop_inode = generic_delete_inode, + .show_options = aufs_show_options, + .statfs = aufs_statfs, + .put_super = aufs_put_super, + .sync_fs = aufs_sync_fs, + .remount_fs = aufs_remount_fs +}; + +/* ---------------------------------------------------------------------- */ + +static int alloc_root(struct super_block *sb) +{ + int err; + struct inode *inode; + struct dentry *root; + + err = -ENOMEM; + inode = au_iget_locked(sb, AUFS_ROOT_INO); + err = PTR_ERR(inode); + if (IS_ERR(inode)) + goto out; + + inode->i_op = aufs_iop + AuIop_DIR; /* with getattr by default */ + inode->i_fop = &aufs_dir_fop; + inode->i_mode = S_IFDIR; + set_nlink(inode, 2); + unlock_new_inode(inode); + + root = d_make_root(inode); + if (unlikely(!root)) + goto out; + err = PTR_ERR(root); + if (IS_ERR(root)) + goto out; + + err = au_di_init(root); + if (!err) { + sb->s_root = root; + return 0; /* success */ + } + dput(root); + +out: + return err; +} + +static int aufs_fill_super(struct super_block *sb, void *raw_data, + int silent __maybe_unused) +{ + int err; + struct au_opts opts = { + .opt = NULL + }; + struct au_sbinfo *sbinfo; + struct dentry *root; + struct inode *inode; + char *arg = raw_data; + + if (unlikely(!arg || !*arg)) { + err = -EINVAL; + pr_err("no arg\n"); + goto out; + } + + err = -ENOMEM; + opts.opt = (void *)__get_free_page(GFP_NOFS); + if (unlikely(!opts.opt)) + goto out; + opts.max_opt = PAGE_SIZE / sizeof(*opts.opt); + opts.sb_flags = sb->s_flags; + + err = au_si_alloc(sb); + if (unlikely(err)) + goto out_opts; + sbinfo = au_sbi(sb); + + /* all timestamps always follow the ones on the branch */ + sb->s_flags |= MS_NOATIME | MS_NODIRATIME; + sb->s_op = &aufs_sop; + sb->s_d_op = &aufs_dop; + sb->s_magic = AUFS_SUPER_MAGIC; + sb->s_maxbytes = 0; + sb->s_stack_depth = 1; + au_export_init(sb); + /* au_xattr_init(sb); */ + + err = alloc_root(sb); + if (unlikely(err)) { + si_write_unlock(sb); + goto out_info; + } + root = sb->s_root; + inode = d_inode(root); + + /* + * actually we can parse options regardless aufs lock here. + * but at remount time, parsing must be done before aufs lock. + * so we follow the same rule. + */ + ii_write_lock_parent(inode); + aufs_write_unlock(root); + err = au_opts_parse(sb, arg, &opts); + if (unlikely(err)) + goto out_root; + + /* lock vfs_inode first, then aufs. */ + inode_lock(inode); + aufs_write_lock(root); + err = au_opts_mount(sb, &opts); + au_opts_free(&opts); + if (!err && au_ftest_si(sbinfo, NO_DREVAL)) { + sb->s_d_op = &aufs_dop_noreval; + pr_info("%pf\n", sb->s_d_op); + au_refresh_dop(root, /*force_reval*/0); + sbinfo->si_iop_array = aufs_iop_nogetattr; + au_refresh_iop(inode, /*force_getattr*/0); + } + aufs_write_unlock(root); + inode_unlock(inode); + if (!err) + goto out_opts; /* success */ + +out_root: + dput(root); + sb->s_root = NULL; +out_info: + dbgaufs_si_fin(sbinfo); + kobject_put(&sbinfo->si_kobj); + sb->s_fs_info = NULL; +out_opts: + free_page((unsigned long)opts.opt); +out: + AuTraceErr(err); + err = cvt_err(err); + AuTraceErr(err); + return err; +} + +/* ---------------------------------------------------------------------- */ + +static struct dentry *aufs_mount(struct file_system_type *fs_type, int flags, + const char *dev_name __maybe_unused, + void *raw_data) +{ + struct dentry *root; + struct super_block *sb; + + /* all timestamps always follow the ones on the branch */ + /* mnt->mnt_flags |= MNT_NOATIME | MNT_NODIRATIME; */ + root = mount_nodev(fs_type, flags, raw_data, aufs_fill_super); + if (IS_ERR(root)) + goto out; + + sb = root->d_sb; + si_write_lock(sb, !AuLock_FLUSH); + sysaufs_brs_add(sb, 0); + si_write_unlock(sb); + au_sbilist_add(sb); + +out: + return root; +} + +static void aufs_kill_sb(struct super_block *sb) +{ + struct au_sbinfo *sbinfo; + + sbinfo = au_sbi(sb); + if (sbinfo) { + au_sbilist_del(sb); + aufs_write_lock(sb->s_root); + au_fhsm_fin(sb); + if (sbinfo->si_wbr_create_ops->fin) + sbinfo->si_wbr_create_ops->fin(sb); + if (au_opt_test(sbinfo->si_mntflags, UDBA_HNOTIFY)) { + au_opt_set_udba(sbinfo->si_mntflags, UDBA_NONE); + au_remount_refresh(sb, /*do_idop*/0); + } + if (au_opt_test(sbinfo->si_mntflags, PLINK)) + au_plink_put(sb, /*verbose*/1); + au_xino_clr(sb); + sbinfo->si_sb = NULL; + aufs_write_unlock(sb->s_root); + au_nwt_flush(&sbinfo->si_nowait); + } + kill_anon_super(sb); +} + +struct file_system_type aufs_fs_type = { + .name = AUFS_FSTYPE, + /* a race between rename and others */ + .fs_flags = FS_RENAME_DOES_D_MOVE, + .mount = aufs_mount, + .kill_sb = aufs_kill_sb, + /* no need to __module_get() and module_put(). */ + .owner = THIS_MODULE, +}; diff --git a/fs/aufs/super.h b/fs/aufs/super.h new file mode 100644 index 000000000..ba2855ad6 --- /dev/null +++ b/fs/aufs/super.h @@ -0,0 +1,619 @@ +/* + * Copyright (C) 2005-2016 Junjiro R. Okajima + */ + +/* + * super_block operations + */ + +#ifndef __AUFS_SUPER_H__ +#define __AUFS_SUPER_H__ + +#ifdef __KERNEL__ + +#include <linux/fs.h> +#include <linux/kobject.h> +#include "rwsem.h" +#include "spl.h" +#include "wkq.h" + +/* policies to select one among multiple writable branches */ +struct au_wbr_copyup_operations { + int (*copyup)(struct dentry *dentry); +}; + +#define AuWbr_DIR 1 /* target is a dir */ +#define AuWbr_PARENT (1 << 1) /* always require a parent */ + +#define au_ftest_wbr(flags, name) ((flags) & AuWbr_##name) +#define au_fset_wbr(flags, name) { (flags) |= AuWbr_##name; } +#define au_fclr_wbr(flags, name) { (flags) &= ~AuWbr_##name; } + +struct au_wbr_create_operations { + int (*create)(struct dentry *dentry, unsigned int flags); + int (*init)(struct super_block *sb); + int (*fin)(struct super_block *sb); +}; + +struct au_wbr_mfs { + struct mutex mfs_lock; /* protect this structure */ + unsigned long mfs_jiffy; + unsigned long mfs_expire; + aufs_bindex_t mfs_bindex; + + unsigned long long mfsrr_bytes; + unsigned long long mfsrr_watermark; +}; + +struct pseudo_link { + union { + struct hlist_node hlist; + struct rcu_head rcu; + }; + struct inode *inode; +}; + +#define AuPlink_NHASH 100 +static inline int au_plink_hash(ino_t ino) +{ + return ino % AuPlink_NHASH; +} + +/* File-based Hierarchical Storage Management */ +struct au_fhsm { +#ifdef CONFIG_AUFS_FHSM + /* allow only one process who can receive the notification */ + spinlock_t fhsm_spin; + pid_t fhsm_pid; + wait_queue_head_t fhsm_wqh; + atomic_t fhsm_readable; + + /* these are protected by si_rwsem */ + unsigned long fhsm_expire; + aufs_bindex_t fhsm_bottom; +#endif +}; + +#define AU_PIDSTEP (int)(BITS_TO_LONGS(PID_MAX_DEFAULT) * BITS_PER_LONG) +#define AU_NPIDMAP (int)DIV_ROUND_UP(PID_MAX_LIMIT, AU_PIDSTEP) +struct au_si_pid { + unsigned long *pid_bitmap[AU_NPIDMAP]; + struct mutex pid_mtx; +}; + +struct au_branch; +struct au_sbinfo { + /* nowait tasks in the system-wide workqueue */ + struct au_nowait_tasks si_nowait; + + /* + * tried sb->s_umount, but failed due to the dependecy between i_mutex. + * rwsem for au_sbinfo is necessary. + */ + struct au_rwsem si_rwsem; + + /* prevent recursive locking in deleting inode */ + struct au_si_pid au_si_pid; + + /* + * dirty approach to protect sb->sb_inodes and ->s_files (gone) from + * remount. + */ + atomic_long_t si_ninodes, si_nfiles; + + /* branch management */ + unsigned int si_generation; + + /* see AuSi_ flags */ + unsigned char au_si_status; + + aufs_bindex_t si_bend; + + /* dirty trick to keep br_id plus */ + unsigned int si_last_br_id : + sizeof(aufs_bindex_t) * BITS_PER_BYTE - 1; + struct au_branch **si_branch; + + /* policy to select a writable branch */ + unsigned char si_wbr_copyup; + unsigned char si_wbr_create; + struct au_wbr_copyup_operations *si_wbr_copyup_ops; + struct au_wbr_create_operations *si_wbr_create_ops; + + /* round robin */ + atomic_t si_wbr_rr_next; + + /* most free space */ + struct au_wbr_mfs si_wbr_mfs; + + /* File-based Hierarchical Storage Management */ + struct au_fhsm si_fhsm; + + /* mount flags */ + /* include/asm-ia64/siginfo.h defines a macro named si_flags */ + unsigned int si_mntflags; + + /* external inode number (bitmap and translation table) */ + vfs_readf_t si_xread; + vfs_writef_t si_xwrite; + struct file *si_xib; + struct mutex si_xib_mtx; /* protect xib members */ + unsigned long *si_xib_buf; + unsigned long si_xib_last_pindex; + int si_xib_next_bit; + aufs_bindex_t si_xino_brid; + unsigned long si_xino_jiffy; + unsigned long si_xino_expire; + /* reserved for future use */ + /* unsigned long long si_xib_limit; */ /* Max xib file size */ + +#ifdef CONFIG_AUFS_EXPORT + /* i_generation */ + struct file *si_xigen; + atomic_t si_xigen_next; +#endif + + /* dirty trick to suppoer atomic_open */ + struct au_sphlhead si_aopen; + + /* vdir parameters */ + unsigned long si_rdcache; /* max cache time in jiffies */ + unsigned int si_rdblk; /* deblk size */ + unsigned int si_rdhash; /* hash size */ + + /* + * If the number of whiteouts are larger than si_dirwh, leave all of + * them after au_whtmp_ren to reduce the cost of rmdir(2). + * future fsck.aufs or kernel thread will remove them later. + * Otherwise, remove all whiteouts and the dir in rmdir(2). + */ + unsigned int si_dirwh; + + /* pseudo_link list */ + struct au_sphlhead si_plink[AuPlink_NHASH]; + wait_queue_head_t si_plink_wq; + spinlock_t si_plink_maint_lock; + pid_t si_plink_maint_pid; + + /* file list */ + struct au_sphlhead si_files; + + /* with/without getattr, brother of sb->s_d_op */ + struct inode_operations *si_iop_array; + + /* + * sysfs and lifetime management. + * this is not a small structure and it may be a waste of memory in case + * of sysfs is disabled, particulary when many aufs-es are mounted. + * but using sysfs is majority. + */ + struct kobject si_kobj; +#ifdef CONFIG_DEBUG_FS + struct dentry *si_dbgaufs; + struct dentry *si_dbgaufs_plink; + struct dentry *si_dbgaufs_xib; +#ifdef CONFIG_AUFS_EXPORT + struct dentry *si_dbgaufs_xigen; +#endif +#endif + +#ifdef CONFIG_AUFS_SBILIST + struct list_head si_list; +#endif + + /* dirty, necessary for unmounting, sysfs and sysrq */ + struct super_block *si_sb; +}; + +/* sbinfo status flags */ +/* + * set true when refresh_dirs() failed at remount time. + * then try refreshing dirs at access time again. + * if it is false, refreshing dirs at access time is unnecesary + */ +#define AuSi_FAILED_REFRESH_DIR 1 +#define AuSi_FHSM (1 << 1) /* fhsm is active now */ +#define AuSi_NO_DREVAL (1 << 2) /* disable all d_revalidate */ + +#ifndef CONFIG_AUFS_FHSM +#undef AuSi_FHSM +#define AuSi_FHSM 0 +#endif + +static inline unsigned char au_do_ftest_si(struct au_sbinfo *sbi, + unsigned int flag) +{ + AuRwMustAnyLock(&sbi->si_rwsem); + return sbi->au_si_status & flag; +} +#define au_ftest_si(sbinfo, name) au_do_ftest_si(sbinfo, AuSi_##name) +#define au_fset_si(sbinfo, name) do { \ + AuRwMustWriteLock(&(sbinfo)->si_rwsem); \ + (sbinfo)->au_si_status |= AuSi_##name; \ +} while (0) +#define au_fclr_si(sbinfo, name) do { \ + AuRwMustWriteLock(&(sbinfo)->si_rwsem); \ + (sbinfo)->au_si_status &= ~AuSi_##name; \ +} while (0) + +/* ---------------------------------------------------------------------- */ + +/* policy to select one among writable branches */ +#define AuWbrCopyup(sbinfo, ...) \ + ((sbinfo)->si_wbr_copyup_ops->copyup(__VA_ARGS__)) +#define AuWbrCreate(sbinfo, ...) \ + ((sbinfo)->si_wbr_create_ops->create(__VA_ARGS__)) + +/* flags for si_read_lock()/aufs_read_lock()/di_read_lock() */ +#define AuLock_DW 1 /* write-lock dentry */ +#define AuLock_IR (1 << 1) /* read-lock inode */ +#define AuLock_IW (1 << 2) /* write-lock inode */ +#define AuLock_FLUSH (1 << 3) /* wait for 'nowait' tasks */ +#define AuLock_DIRS (1 << 4) /* target is a pair of dirs */ +#define AuLock_NOPLM (1 << 5) /* return err in plm mode */ +#define AuLock_NOPLMW (1 << 6) /* wait for plm mode ends */ +#define AuLock_GEN (1 << 7) /* test digen/iigen */ +#define au_ftest_lock(flags, name) ((flags) & AuLock_##name) +#define au_fset_lock(flags, name) \ + do { (flags) |= AuLock_##name; } while (0) +#define au_fclr_lock(flags, name) \ + do { (flags) &= ~AuLock_##name; } while (0) + +/* ---------------------------------------------------------------------- */ + +/* super.c */ +extern struct file_system_type aufs_fs_type; +struct inode *au_iget_locked(struct super_block *sb, ino_t ino); +typedef unsigned long long (*au_arraycb_t)(struct super_block *sb, void *array, + unsigned long long max, void *arg); +void *au_array_alloc(unsigned long long *hint, au_arraycb_t cb, + struct super_block *sb, void *arg); +struct inode **au_iarray_alloc(struct super_block *sb, unsigned long long *max); +void au_iarray_free(struct inode **a, unsigned long long max); + +/* sbinfo.c */ +void au_si_free(struct kobject *kobj); +int au_si_alloc(struct super_block *sb); +int au_sbr_realloc(struct au_sbinfo *sbinfo, int nbr); + +unsigned int au_sigen_inc(struct super_block *sb); +aufs_bindex_t au_new_br_id(struct super_block *sb); + +int si_read_lock(struct super_block *sb, int flags); +int si_write_lock(struct super_block *sb, int flags); +int aufs_read_lock(struct dentry *dentry, int flags); +void aufs_read_unlock(struct dentry *dentry, int flags); +void aufs_write_lock(struct dentry *dentry); +void aufs_write_unlock(struct dentry *dentry); +int aufs_read_and_write_lock2(struct dentry *d1, struct dentry *d2, int flags); +void aufs_read_and_write_unlock2(struct dentry *d1, struct dentry *d2); + +/* wbr_policy.c */ +extern struct au_wbr_copyup_operations au_wbr_copyup_ops[]; +extern struct au_wbr_create_operations au_wbr_create_ops[]; +int au_cpdown_dirs(struct dentry *dentry, aufs_bindex_t bdst); +int au_wbr_nonopq(struct dentry *dentry, aufs_bindex_t bindex); +int au_wbr_do_copyup_bu(struct dentry *dentry, aufs_bindex_t bstart); + +/* mvdown.c */ +int au_mvdown(struct dentry *dentry, struct aufs_mvdown __user *arg); + +#ifdef CONFIG_AUFS_FHSM +/* fhsm.c */ + +static inline pid_t au_fhsm_pid(struct au_fhsm *fhsm) +{ + pid_t pid; + + spin_lock(&fhsm->fhsm_spin); + pid = fhsm->fhsm_pid; + spin_unlock(&fhsm->fhsm_spin); + + return pid; +} + +void au_fhsm_wrote(struct super_block *sb, aufs_bindex_t bindex, int force); +void au_fhsm_wrote_all(struct super_block *sb, int force); +int au_fhsm_fd(struct super_block *sb, int oflags); +int au_fhsm_br_alloc(struct au_branch *br); +void au_fhsm_set_bottom(struct super_block *sb, aufs_bindex_t bindex); +void au_fhsm_fin(struct super_block *sb); +void au_fhsm_init(struct au_sbinfo *sbinfo); +void au_fhsm_set(struct au_sbinfo *sbinfo, unsigned int sec); +void au_fhsm_show(struct seq_file *seq, struct au_sbinfo *sbinfo); +#else +AuStubVoid(au_fhsm_wrote, struct super_block *sb, aufs_bindex_t bindex, + int force) +AuStubVoid(au_fhsm_wrote_all, struct super_block *sb, int force) +AuStub(int, au_fhsm_fd, return -EOPNOTSUPP, struct super_block *sb, int oflags) +AuStub(pid_t, au_fhsm_pid, return 0, struct au_fhsm *fhsm) +AuStubInt0(au_fhsm_br_alloc, struct au_branch *br) +AuStubVoid(au_fhsm_set_bottom, struct super_block *sb, aufs_bindex_t bindex) +AuStubVoid(au_fhsm_fin, struct super_block *sb) +AuStubVoid(au_fhsm_init, struct au_sbinfo *sbinfo) +AuStubVoid(au_fhsm_set, struct au_sbinfo *sbinfo, unsigned int sec) +AuStubVoid(au_fhsm_show, struct seq_file *seq, struct au_sbinfo *sbinfo) +#endif + +/* ---------------------------------------------------------------------- */ + +static inline struct au_sbinfo *au_sbi(struct super_block *sb) +{ + return sb->s_fs_info; +} + +/* ---------------------------------------------------------------------- */ + +#ifdef CONFIG_AUFS_EXPORT +int au_test_nfsd(void); +void au_export_init(struct super_block *sb); +void au_xigen_inc(struct inode *inode); +int au_xigen_new(struct inode *inode); +int au_xigen_set(struct super_block *sb, struct file *base); +void au_xigen_clr(struct super_block *sb); + +static inline int au_busy_or_stale(void) +{ + if (!au_test_nfsd()) + return -EBUSY; + return -ESTALE; +} +#else +AuStubInt0(au_test_nfsd, void) +AuStubVoid(au_export_init, struct super_block *sb) +AuStubVoid(au_xigen_inc, struct inode *inode) +AuStubInt0(au_xigen_new, struct inode *inode) +AuStubInt0(au_xigen_set, struct super_block *sb, struct file *base) +AuStubVoid(au_xigen_clr, struct super_block *sb) +AuStub(int, au_busy_or_stale, return -EBUSY, void) +#endif /* CONFIG_AUFS_EXPORT */ + +/* ---------------------------------------------------------------------- */ + +#ifdef CONFIG_AUFS_SBILIST +/* module.c */ +extern struct au_splhead au_sbilist; + +static inline void au_sbilist_init(void) +{ + au_spl_init(&au_sbilist); +} + +static inline void au_sbilist_add(struct super_block *sb) +{ + au_spl_add(&au_sbi(sb)->si_list, &au_sbilist); +} + +static inline void au_sbilist_del(struct super_block *sb) +{ + au_spl_del(&au_sbi(sb)->si_list, &au_sbilist); +} + +#ifdef CONFIG_AUFS_MAGIC_SYSRQ +static inline void au_sbilist_lock(void) +{ + spin_lock(&au_sbilist.spin); +} + +static inline void au_sbilist_unlock(void) +{ + spin_unlock(&au_sbilist.spin); +} +#define AuGFP_SBILIST GFP_ATOMIC +#else +AuStubVoid(au_sbilist_lock, void) +AuStubVoid(au_sbilist_unlock, void) +#define AuGFP_SBILIST GFP_NOFS +#endif /* CONFIG_AUFS_MAGIC_SYSRQ */ +#else +AuStubVoid(au_sbilist_init, void) +AuStubVoid(au_sbilist_add, struct super_block *sb) +AuStubVoid(au_sbilist_del, struct super_block *sb) +AuStubVoid(au_sbilist_lock, void) +AuStubVoid(au_sbilist_unlock, void) +#define AuGFP_SBILIST GFP_NOFS +#endif + +/* ---------------------------------------------------------------------- */ + +static inline void dbgaufs_si_null(struct au_sbinfo *sbinfo) +{ + /* + * This function is a dynamic '__init' function actually, + * so the tiny check for si_rwsem is unnecessary. + */ + /* AuRwMustWriteLock(&sbinfo->si_rwsem); */ +#ifdef CONFIG_DEBUG_FS + sbinfo->si_dbgaufs = NULL; + sbinfo->si_dbgaufs_plink = NULL; + sbinfo->si_dbgaufs_xib = NULL; +#ifdef CONFIG_AUFS_EXPORT + sbinfo->si_dbgaufs_xigen = NULL; +#endif +#endif +} + +/* ---------------------------------------------------------------------- */ + +static inline void si_pid_idx_bit(int *idx, pid_t *bit) +{ + /* the origin of pid is 1, but the bitmap's is 0 */ + *bit = current->pid - 1; + *idx = *bit / AU_PIDSTEP; + *bit %= AU_PIDSTEP; +} + +static inline int si_pid_test(struct super_block *sb) +{ + pid_t bit; + int idx; + unsigned long *bitmap; + + si_pid_idx_bit(&idx, &bit); + bitmap = au_sbi(sb)->au_si_pid.pid_bitmap[idx]; + if (bitmap) + return test_bit(bit, bitmap); + return 0; +} + +static inline void si_pid_clr(struct super_block *sb) +{ + pid_t bit; + int idx; + unsigned long *bitmap; + + si_pid_idx_bit(&idx, &bit); + bitmap = au_sbi(sb)->au_si_pid.pid_bitmap[idx]; + BUG_ON(!bitmap); + AuDebugOn(!test_bit(bit, bitmap)); + clear_bit(bit, bitmap); + /* smp_mb(); */ +} + +void si_pid_set(struct super_block *sb); + +/* ---------------------------------------------------------------------- */ + +/* lock superblock. mainly for entry point functions */ +/* + * __si_read_lock, __si_write_lock, + * __si_read_unlock, __si_write_unlock, __si_downgrade_lock + */ +AuSimpleRwsemFuncs(__si, struct super_block *sb, &au_sbi(sb)->si_rwsem); + +#define SiMustNoWaiters(sb) AuRwMustNoWaiters(&au_sbi(sb)->si_rwsem) +#define SiMustAnyLock(sb) AuRwMustAnyLock(&au_sbi(sb)->si_rwsem) +#define SiMustWriteLock(sb) AuRwMustWriteLock(&au_sbi(sb)->si_rwsem) + +static inline void si_noflush_read_lock(struct super_block *sb) +{ + __si_read_lock(sb); + si_pid_set(sb); +} + +static inline int si_noflush_read_trylock(struct super_block *sb) +{ + int locked; + + locked = __si_read_trylock(sb); + if (locked) + si_pid_set(sb); + return locked; +} + +static inline void si_noflush_write_lock(struct super_block *sb) +{ + __si_write_lock(sb); + si_pid_set(sb); +} + +static inline int si_noflush_write_trylock(struct super_block *sb) +{ + int locked; + + locked = __si_write_trylock(sb); + if (locked) + si_pid_set(sb); + return locked; +} + +#if 0 /* reserved */ +static inline int si_read_trylock(struct super_block *sb, int flags) +{ + if (au_ftest_lock(flags, FLUSH)) + au_nwt_flush(&au_sbi(sb)->si_nowait); + return si_noflush_read_trylock(sb); +} +#endif + +static inline void si_read_unlock(struct super_block *sb) +{ + si_pid_clr(sb); + __si_read_unlock(sb); +} + +#if 0 /* reserved */ +static inline int si_write_trylock(struct super_block *sb, int flags) +{ + if (au_ftest_lock(flags, FLUSH)) + au_nwt_flush(&au_sbi(sb)->si_nowait); + return si_noflush_write_trylock(sb); +} +#endif + +static inline void si_write_unlock(struct super_block *sb) +{ + si_pid_clr(sb); + __si_write_unlock(sb); +} + +#if 0 /* reserved */ +static inline void si_downgrade_lock(struct super_block *sb) +{ + __si_downgrade_lock(sb); +} +#endif + +/* ---------------------------------------------------------------------- */ + +static inline aufs_bindex_t au_sbend(struct super_block *sb) +{ + SiMustAnyLock(sb); + return au_sbi(sb)->si_bend; +} + +static inline unsigned int au_mntflags(struct super_block *sb) +{ + SiMustAnyLock(sb); + return au_sbi(sb)->si_mntflags; +} + +static inline unsigned int au_sigen(struct super_block *sb) +{ + SiMustAnyLock(sb); + return au_sbi(sb)->si_generation; +} + +static inline void au_ninodes_inc(struct super_block *sb) +{ + atomic_long_inc(&au_sbi(sb)->si_ninodes); +} + +static inline void au_ninodes_dec(struct super_block *sb) +{ + AuDebugOn(!atomic_long_read(&au_sbi(sb)->si_ninodes)); + atomic_long_dec(&au_sbi(sb)->si_ninodes); +} + +static inline void au_nfiles_inc(struct super_block *sb) +{ + atomic_long_inc(&au_sbi(sb)->si_nfiles); +} + +static inline void au_nfiles_dec(struct super_block *sb) +{ + AuDebugOn(!atomic_long_read(&au_sbi(sb)->si_nfiles)); + atomic_long_dec(&au_sbi(sb)->si_nfiles); +} + +static inline struct au_branch *au_sbr(struct super_block *sb, + aufs_bindex_t bindex) +{ + SiMustAnyLock(sb); + return au_sbi(sb)->si_branch[0 + bindex]; +} + +static inline void au_xino_brid_set(struct super_block *sb, aufs_bindex_t brid) +{ + SiMustWriteLock(sb); + au_sbi(sb)->si_xino_brid = brid; +} + +static inline aufs_bindex_t au_xino_brid(struct super_block *sb) +{ + SiMustAnyLock(sb); + return au_sbi(sb)->si_xino_brid; +} + +#endif /* __KERNEL__ */ +#endif /* __AUFS_SUPER_H__ */ diff --git a/fs/aufs/sysaufs.c b/fs/aufs/sysaufs.c new file mode 100644 index 000000000..8ec10fb31 --- /dev/null +++ b/fs/aufs/sysaufs.c @@ -0,0 +1,91 @@ +/* + * Copyright (C) 2005-2016 Junjiro R. Okajima + */ + +/* + * sysfs interface and lifetime management + * they are necessary regardless sysfs is disabled. + */ + +#include <linux/random.h> +#include "aufs.h" + +unsigned long sysaufs_si_mask; +struct kset *sysaufs_kset; + +#define AuSiAttr(_name) { \ + .attr = { .name = __stringify(_name), .mode = 0444 }, \ + .show = sysaufs_si_##_name, \ +} + +static struct sysaufs_si_attr sysaufs_si_attr_xi_path = AuSiAttr(xi_path); +struct attribute *sysaufs_si_attrs[] = { + &sysaufs_si_attr_xi_path.attr, + NULL, +}; + +static const struct sysfs_ops au_sbi_ops = { + .show = sysaufs_si_show +}; + +static struct kobj_type au_sbi_ktype = { + .release = au_si_free, + .sysfs_ops = &au_sbi_ops, + .default_attrs = sysaufs_si_attrs +}; + +/* ---------------------------------------------------------------------- */ + +int sysaufs_si_init(struct au_sbinfo *sbinfo) +{ + int err; + + sbinfo->si_kobj.kset = sysaufs_kset; + /* cf. sysaufs_name() */ + err = kobject_init_and_add + (&sbinfo->si_kobj, &au_sbi_ktype, /*&sysaufs_kset->kobj*/NULL, + SysaufsSiNamePrefix "%lx", sysaufs_si_id(sbinfo)); + + dbgaufs_si_null(sbinfo); + if (!err) { + err = dbgaufs_si_init(sbinfo); + if (unlikely(err)) + kobject_put(&sbinfo->si_kobj); + } + return err; +} + +void sysaufs_fin(void) +{ + dbgaufs_fin(); + sysfs_remove_group(&sysaufs_kset->kobj, sysaufs_attr_group); + kset_unregister(sysaufs_kset); +} + +int __init sysaufs_init(void) +{ + int err; + + do { + get_random_bytes(&sysaufs_si_mask, sizeof(sysaufs_si_mask)); + } while (!sysaufs_si_mask); + + err = -EINVAL; + sysaufs_kset = kset_create_and_add(AUFS_NAME, NULL, fs_kobj); + if (unlikely(!sysaufs_kset)) + goto out; + err = PTR_ERR(sysaufs_kset); + if (IS_ERR(sysaufs_kset)) + goto out; + err = sysfs_create_group(&sysaufs_kset->kobj, sysaufs_attr_group); + if (unlikely(err)) { + kset_unregister(sysaufs_kset); + goto out; + } + + err = dbgaufs_init(); + if (unlikely(err)) + sysaufs_fin(); +out: + return err; +} diff --git a/fs/aufs/sysaufs.h b/fs/aufs/sysaufs.h new file mode 100644 index 000000000..1f799835e --- /dev/null +++ b/fs/aufs/sysaufs.h @@ -0,0 +1,88 @@ +/* + * Copyright (C) 2005-2016 Junjiro R. Okajima + */ + +/* + * sysfs interface and mount lifetime management + */ + +#ifndef __SYSAUFS_H__ +#define __SYSAUFS_H__ + +#ifdef __KERNEL__ + +#include <linux/sysfs.h> +#include "module.h" + +struct super_block; +struct au_sbinfo; + +struct sysaufs_si_attr { + struct attribute attr; + int (*show)(struct seq_file *seq, struct super_block *sb); +}; + +/* ---------------------------------------------------------------------- */ + +/* sysaufs.c */ +extern unsigned long sysaufs_si_mask; +extern struct kset *sysaufs_kset; +extern struct attribute *sysaufs_si_attrs[]; +int sysaufs_si_init(struct au_sbinfo *sbinfo); +int __init sysaufs_init(void); +void sysaufs_fin(void); + +/* ---------------------------------------------------------------------- */ + +/* some people doesn't like to show a pointer in kernel */ +static inline unsigned long sysaufs_si_id(struct au_sbinfo *sbinfo) +{ + return sysaufs_si_mask ^ (unsigned long)sbinfo; +} + +#define SysaufsSiNamePrefix "si_" +#define SysaufsSiNameLen (sizeof(SysaufsSiNamePrefix) + 16) +static inline void sysaufs_name(struct au_sbinfo *sbinfo, char *name) +{ + snprintf(name, SysaufsSiNameLen, SysaufsSiNamePrefix "%lx", + sysaufs_si_id(sbinfo)); +} + +struct au_branch; +#ifdef CONFIG_SYSFS +/* sysfs.c */ +extern struct attribute_group *sysaufs_attr_group; + +int sysaufs_si_xi_path(struct seq_file *seq, struct super_block *sb); +ssize_t sysaufs_si_show(struct kobject *kobj, struct attribute *attr, + char *buf); +long au_brinfo_ioctl(struct file *file, unsigned long arg); +#ifdef CONFIG_COMPAT +long au_brinfo_compat_ioctl(struct file *file, unsigned long arg); +#endif + +void sysaufs_br_init(struct au_branch *br); +void sysaufs_brs_add(struct super_block *sb, aufs_bindex_t bindex); +void sysaufs_brs_del(struct super_block *sb, aufs_bindex_t bindex); + +#define sysaufs_brs_init() do {} while (0) + +#else +#define sysaufs_attr_group NULL + +AuStubInt0(sysaufs_si_xi_path, struct seq_file *seq, struct super_block *sb) +AuStub(ssize_t, sysaufs_si_show, return 0, struct kobject *kobj, + struct attribute *attr, char *buf) +AuStubVoid(sysaufs_br_init, struct au_branch *br) +AuStubVoid(sysaufs_brs_add, struct super_block *sb, aufs_bindex_t bindex) +AuStubVoid(sysaufs_brs_del, struct super_block *sb, aufs_bindex_t bindex) + +static inline void sysaufs_brs_init(void) +{ + sysaufs_brs = 0; +} + +#endif /* CONFIG_SYSFS */ + +#endif /* __KERNEL__ */ +#endif /* __SYSAUFS_H__ */ diff --git a/fs/aufs/sysfs.c b/fs/aufs/sysfs.c new file mode 100644 index 000000000..ed42f53d0 --- /dev/null +++ b/fs/aufs/sysfs.c @@ -0,0 +1,340 @@ +/* + * Copyright (C) 2005-2016 Junjiro R. Okajima + */ + +/* + * sysfs interface + */ + +#include <linux/compat.h> +#include <linux/seq_file.h> +#include "aufs.h" + +static struct attribute *au_attr[] = { + NULL, /* need to NULL terminate the list of attributes */ +}; + +static struct attribute_group sysaufs_attr_group_body = { + .attrs = au_attr +}; + +struct attribute_group *sysaufs_attr_group = &sysaufs_attr_group_body; + +/* ---------------------------------------------------------------------- */ + +int sysaufs_si_xi_path(struct seq_file *seq, struct super_block *sb) +{ + int err; + + SiMustAnyLock(sb); + + err = 0; + if (au_opt_test(au_mntflags(sb), XINO)) { + err = au_xino_path(seq, au_sbi(sb)->si_xib); + seq_putc(seq, '\n'); + } + return err; +} + +/* + * the lifetime of branch is independent from the entry under sysfs. + * sysfs handles the lifetime of the entry, and never call ->show() after it is + * unlinked. + */ +static int sysaufs_si_br(struct seq_file *seq, struct super_block *sb, + aufs_bindex_t bindex, int idx) +{ + int err; + struct path path; + struct dentry *root; + struct au_branch *br; + au_br_perm_str_t perm; + + AuDbg("b%d\n", bindex); + + err = 0; + root = sb->s_root; + di_read_lock_parent(root, !AuLock_IR); + br = au_sbr(sb, bindex); + + switch (idx) { + case AuBrSysfs_BR: + path.mnt = au_br_mnt(br); + path.dentry = au_h_dptr(root, bindex); + err = au_seq_path(seq, &path); + if (!err) { + au_optstr_br_perm(&perm, br->br_perm); + seq_printf(seq, "=%s\n", perm.a); + } + break; + case AuBrSysfs_BRID: + seq_printf(seq, "%d\n", br->br_id); + break; + } + di_read_unlock(root, !AuLock_IR); + if (unlikely(err || seq_has_overflowed(seq))) + err = -E2BIG; + + return err; +} + +/* ---------------------------------------------------------------------- */ + +static struct seq_file *au_seq(char *p, ssize_t len) +{ + struct seq_file *seq; + + seq = kzalloc(sizeof(*seq), GFP_NOFS); + if (seq) { + /* mutex_init(&seq.lock); */ + seq->buf = p; + seq->size = len; + return seq; /* success */ + } + + seq = ERR_PTR(-ENOMEM); + return seq; +} + +#define SysaufsBr_PREFIX "br" +#define SysaufsBrid_PREFIX "brid" + +/* todo: file size may exceed PAGE_SIZE */ +ssize_t sysaufs_si_show(struct kobject *kobj, struct attribute *attr, + char *buf) +{ + ssize_t err; + int idx; + long l; + aufs_bindex_t bend; + struct au_sbinfo *sbinfo; + struct super_block *sb; + struct seq_file *seq; + char *name; + struct attribute **cattr; + + sbinfo = container_of(kobj, struct au_sbinfo, si_kobj); + sb = sbinfo->si_sb; + + /* + * prevent a race condition between sysfs and aufs. + * for instance, sysfs_file_read() calls sysfs_get_active_two() which + * prohibits maintaining the sysfs entries. + * hew we acquire read lock after sysfs_get_active_two(). + * on the other hand, the remount process may maintain the sysfs/aufs + * entries after acquiring write lock. + * it can cause a deadlock. + * simply we gave up processing read here. + */ + err = -EBUSY; + if (unlikely(!si_noflush_read_trylock(sb))) + goto out; + + seq = au_seq(buf, PAGE_SIZE); + err = PTR_ERR(seq); + if (IS_ERR(seq)) + goto out_unlock; + + name = (void *)attr->name; + cattr = sysaufs_si_attrs; + while (*cattr) { + if (!strcmp(name, (*cattr)->name)) { + err = container_of(*cattr, struct sysaufs_si_attr, attr) + ->show(seq, sb); + goto out_seq; + } + cattr++; + } + + if (!strncmp(name, SysaufsBrid_PREFIX, + sizeof(SysaufsBrid_PREFIX) - 1)) { + idx = AuBrSysfs_BRID; + name += sizeof(SysaufsBrid_PREFIX) - 1; + } else if (!strncmp(name, SysaufsBr_PREFIX, + sizeof(SysaufsBr_PREFIX) - 1)) { + idx = AuBrSysfs_BR; + name += sizeof(SysaufsBr_PREFIX) - 1; + } else + BUG(); + + err = kstrtol(name, 10, &l); + if (!err) { + bend = au_sbend(sb); + if (l <= bend) + err = sysaufs_si_br(seq, sb, (aufs_bindex_t)l, idx); + else + err = -ENOENT; + } + +out_seq: + if (!err) { + err = seq->count; + /* sysfs limit */ + if (unlikely(err == PAGE_SIZE)) + err = -EFBIG; + } + kfree(seq); +out_unlock: + si_read_unlock(sb); +out: + return err; +} + +/* ---------------------------------------------------------------------- */ + +static int au_brinfo(struct super_block *sb, union aufs_brinfo __user *arg) +{ + int err; + int16_t brid; + aufs_bindex_t bindex, bend; + size_t sz; + char *buf; + struct seq_file *seq; + struct au_branch *br; + + si_read_lock(sb, AuLock_FLUSH); + bend = au_sbend(sb); + err = bend + 1; + if (!arg) + goto out; + + err = -ENOMEM; + buf = (void *)__get_free_page(GFP_NOFS); + if (unlikely(!buf)) + goto out; + + seq = au_seq(buf, PAGE_SIZE); + err = PTR_ERR(seq); + if (IS_ERR(seq)) + goto out_buf; + + sz = sizeof(*arg) - offsetof(union aufs_brinfo, path); + for (bindex = 0; bindex <= bend; bindex++, arg++) { + err = !access_ok(VERIFY_WRITE, arg, sizeof(*arg)); + if (unlikely(err)) + break; + + br = au_sbr(sb, bindex); + brid = br->br_id; + BUILD_BUG_ON(sizeof(brid) != sizeof(arg->id)); + err = __put_user(brid, &arg->id); + if (unlikely(err)) + break; + + BUILD_BUG_ON(sizeof(br->br_perm) != sizeof(arg->perm)); + err = __put_user(br->br_perm, &arg->perm); + if (unlikely(err)) + break; + + err = au_seq_path(seq, &br->br_path); + if (unlikely(err)) + break; + seq_putc(seq, '\0'); + if (!seq_has_overflowed(seq)) { + err = copy_to_user(arg->path, seq->buf, seq->count); + seq->count = 0; + if (unlikely(err)) + break; + } else { + err = -E2BIG; + goto out_seq; + } + } + if (unlikely(err)) + err = -EFAULT; + +out_seq: + kfree(seq); +out_buf: + free_page((unsigned long)buf); +out: + si_read_unlock(sb); + return err; +} + +long au_brinfo_ioctl(struct file *file, unsigned long arg) +{ + return au_brinfo(file->f_path.dentry->d_sb, (void __user *)arg); +} + +#ifdef CONFIG_COMPAT +long au_brinfo_compat_ioctl(struct file *file, unsigned long arg) +{ + return au_brinfo(file->f_path.dentry->d_sb, compat_ptr(arg)); +} +#endif + +/* ---------------------------------------------------------------------- */ + +void sysaufs_br_init(struct au_branch *br) +{ + int i; + struct au_brsysfs *br_sysfs; + struct attribute *attr; + + br_sysfs = br->br_sysfs; + for (i = 0; i < ARRAY_SIZE(br->br_sysfs); i++) { + attr = &br_sysfs->attr; + sysfs_attr_init(attr); + attr->name = br_sysfs->name; + attr->mode = S_IRUGO; + br_sysfs++; + } +} + +void sysaufs_brs_del(struct super_block *sb, aufs_bindex_t bindex) +{ + struct au_branch *br; + struct kobject *kobj; + struct au_brsysfs *br_sysfs; + int i; + aufs_bindex_t bend; + + dbgaufs_brs_del(sb, bindex); + + if (!sysaufs_brs) + return; + + kobj = &au_sbi(sb)->si_kobj; + bend = au_sbend(sb); + for (; bindex <= bend; bindex++) { + br = au_sbr(sb, bindex); + br_sysfs = br->br_sysfs; + for (i = 0; i < ARRAY_SIZE(br->br_sysfs); i++) { + sysfs_remove_file(kobj, &br_sysfs->attr); + br_sysfs++; + } + } +} + +void sysaufs_brs_add(struct super_block *sb, aufs_bindex_t bindex) +{ + int err, i; + aufs_bindex_t bend; + struct kobject *kobj; + struct au_branch *br; + struct au_brsysfs *br_sysfs; + + dbgaufs_brs_add(sb, bindex); + + if (!sysaufs_brs) + return; + + kobj = &au_sbi(sb)->si_kobj; + bend = au_sbend(sb); + for (; bindex <= bend; bindex++) { + br = au_sbr(sb, bindex); + br_sysfs = br->br_sysfs; + snprintf(br_sysfs[AuBrSysfs_BR].name, sizeof(br_sysfs->name), + SysaufsBr_PREFIX "%d", bindex); + snprintf(br_sysfs[AuBrSysfs_BRID].name, sizeof(br_sysfs->name), + SysaufsBrid_PREFIX "%d", bindex); + for (i = 0; i < ARRAY_SIZE(br->br_sysfs); i++) { + err = sysfs_create_file(kobj, &br_sysfs->attr); + if (unlikely(err)) + pr_warn("failed %s under sysfs(%d)\n", + br_sysfs->name, err); + br_sysfs++; + } + } +} diff --git a/fs/aufs/sysrq.c b/fs/aufs/sysrq.c new file mode 100644 index 000000000..7921ed716 --- /dev/null +++ b/fs/aufs/sysrq.c @@ -0,0 +1,144 @@ +/* + * Copyright (C) 2005-2016 Junjiro R. Okajima + */ + +/* + * magic sysrq hanlder + */ + +/* #include <linux/sysrq.h> */ +#include <linux/writeback.h> +#include "aufs.h" + +/* ---------------------------------------------------------------------- */ + +static void sysrq_sb(struct super_block *sb) +{ + char *plevel; + struct au_sbinfo *sbinfo; + struct file *file; + struct au_sphlhead *files; + struct au_finfo *finfo; + + plevel = au_plevel; + au_plevel = KERN_WARNING; + + /* since we define pr_fmt, call printk directly */ +#define pr(str) printk(KERN_WARNING AUFS_NAME ": " str) + + sbinfo = au_sbi(sb); + printk(KERN_WARNING "si=%lx\n", sysaufs_si_id(sbinfo)); + pr("superblock\n"); + au_dpri_sb(sb); + +#if 0 + pr("root dentry\n"); + au_dpri_dentry(sb->s_root); + pr("root inode\n"); + au_dpri_inode(d_inode(sb->s_root)); +#endif + +#if 0 + do { + int err, i, j, ndentry; + struct au_dcsub_pages dpages; + struct au_dpage *dpage; + + err = au_dpages_init(&dpages, GFP_ATOMIC); + if (unlikely(err)) + break; + err = au_dcsub_pages(&dpages, sb->s_root, NULL, NULL); + if (!err) + for (i = 0; i < dpages.ndpage; i++) { + dpage = dpages.dpages + i; + ndentry = dpage->ndentry; + for (j = 0; j < ndentry; j++) + au_dpri_dentry(dpage->dentries[j]); + } + au_dpages_free(&dpages); + } while (0); +#endif + +#if 1 + { + struct inode *i; + + pr("isolated inode\n"); + spin_lock(&sb->s_inode_list_lock); + list_for_each_entry(i, &sb->s_inodes, i_sb_list) { + spin_lock(&i->i_lock); + if (1 || hlist_empty(&i->i_dentry)) + au_dpri_inode(i); + spin_unlock(&i->i_lock); + } + spin_unlock(&sb->s_inode_list_lock); + } +#endif + pr("files\n"); + files = &au_sbi(sb)->si_files; + spin_lock(&files->spin); + hlist_for_each_entry(finfo, &files->head, fi_hlist) { + umode_t mode; + + file = finfo->fi_file; + mode = file_inode(file)->i_mode; + if (!special_file(mode)) + au_dpri_file(file); + } + spin_unlock(&files->spin); + pr("done\n"); + +#undef pr + au_plevel = plevel; +} + +/* ---------------------------------------------------------------------- */ + +/* module parameter */ +static char *aufs_sysrq_key = "a"; +module_param_named(sysrq, aufs_sysrq_key, charp, S_IRUGO); +MODULE_PARM_DESC(sysrq, "MagicSysRq key for " AUFS_NAME); + +static void au_sysrq(int key __maybe_unused) +{ + struct au_sbinfo *sbinfo; + + lockdep_off(); + au_sbilist_lock(); + list_for_each_entry(sbinfo, &au_sbilist.head, si_list) + sysrq_sb(sbinfo->si_sb); + au_sbilist_unlock(); + lockdep_on(); +} + +static struct sysrq_key_op au_sysrq_op = { + .handler = au_sysrq, + .help_msg = "Aufs", + .action_msg = "Aufs", + .enable_mask = SYSRQ_ENABLE_DUMP +}; + +/* ---------------------------------------------------------------------- */ + +int __init au_sysrq_init(void) +{ + int err; + char key; + + err = -1; + key = *aufs_sysrq_key; + if ('a' <= key && key <= 'z') + err = register_sysrq_key(key, &au_sysrq_op); + if (unlikely(err)) + pr_err("err %d, sysrq=%c\n", err, key); + return err; +} + +void au_sysrq_fin(void) +{ + int err; + + err = unregister_sysrq_key(*aufs_sysrq_key, &au_sysrq_op); + if (unlikely(err)) + pr_err("err %d (ignored)\n", err); +} diff --git a/fs/aufs/vdir.c b/fs/aufs/vdir.c new file mode 100644 index 000000000..f64cc2b7a --- /dev/null +++ b/fs/aufs/vdir.c @@ -0,0 +1,875 @@ +/* + * Copyright (C) 2005-2016 Junjiro R. Okajima + */ + +/* + * virtual or vertical directory + */ + +#include "aufs.h" + +static unsigned int calc_size(int nlen) +{ + return ALIGN(sizeof(struct au_vdir_de) + nlen, sizeof(ino_t)); +} + +static int set_deblk_end(union au_vdir_deblk_p *p, + union au_vdir_deblk_p *deblk_end) +{ + if (calc_size(0) <= deblk_end->deblk - p->deblk) { + p->de->de_str.len = 0; + /* smp_mb(); */ + return 0; + } + return -1; /* error */ +} + +/* returns true or false */ +static int is_deblk_end(union au_vdir_deblk_p *p, + union au_vdir_deblk_p *deblk_end) +{ + if (calc_size(0) <= deblk_end->deblk - p->deblk) + return !p->de->de_str.len; + return 1; +} + +static unsigned char *last_deblk(struct au_vdir *vdir) +{ + return vdir->vd_deblk[vdir->vd_nblk - 1]; +} + +/* ---------------------------------------------------------------------- */ + +/* estimate the appropriate size for name hash table */ +unsigned int au_rdhash_est(loff_t sz) +{ + unsigned int n; + + n = UINT_MAX; + sz >>= 10; + if (sz < n) + n = sz; + if (sz < AUFS_RDHASH_DEF) + n = AUFS_RDHASH_DEF; + /* pr_info("n %u\n", n); */ + return n; +} + +/* + * the allocated memory has to be freed by + * au_nhash_wh_free() or au_nhash_de_free(). + */ +int au_nhash_alloc(struct au_nhash *nhash, unsigned int num_hash, gfp_t gfp) +{ + struct hlist_head *head; + unsigned int u; + size_t sz; + + sz = sizeof(*nhash->nh_head) * num_hash; + head = kmalloc(sz, gfp); + if (head) { + nhash->nh_num = num_hash; + nhash->nh_head = head; + for (u = 0; u < num_hash; u++) + INIT_HLIST_HEAD(head++); + return 0; /* success */ + } + + return -ENOMEM; +} + +static void nhash_count(struct hlist_head *head) +{ +#if 0 + unsigned long n; + struct hlist_node *pos; + + n = 0; + hlist_for_each(pos, head) + n++; + pr_info("%lu\n", n); +#endif +} + +static void au_nhash_wh_do_free(struct hlist_head *head) +{ + struct au_vdir_wh *pos; + struct hlist_node *node; + + hlist_for_each_entry_safe(pos, node, head, wh_hash) + kfree(pos); +} + +static void au_nhash_de_do_free(struct hlist_head *head) +{ + struct au_vdir_dehstr *pos; + struct hlist_node *node; + + hlist_for_each_entry_safe(pos, node, head, hash) + au_cache_free_vdir_dehstr(pos); +} + +static void au_nhash_do_free(struct au_nhash *nhash, + void (*free)(struct hlist_head *head)) +{ + unsigned int n; + struct hlist_head *head; + + n = nhash->nh_num; + if (!n) + return; + + head = nhash->nh_head; + while (n-- > 0) { + nhash_count(head); + free(head++); + } + kfree(nhash->nh_head); +} + +void au_nhash_wh_free(struct au_nhash *whlist) +{ + au_nhash_do_free(whlist, au_nhash_wh_do_free); +} + +static void au_nhash_de_free(struct au_nhash *delist) +{ + au_nhash_do_free(delist, au_nhash_de_do_free); +} + +/* ---------------------------------------------------------------------- */ + +int au_nhash_test_longer_wh(struct au_nhash *whlist, aufs_bindex_t btgt, + int limit) +{ + int num; + unsigned int u, n; + struct hlist_head *head; + struct au_vdir_wh *pos; + + num = 0; + n = whlist->nh_num; + head = whlist->nh_head; + for (u = 0; u < n; u++, head++) + hlist_for_each_entry(pos, head, wh_hash) + if (pos->wh_bindex == btgt && ++num > limit) + return 1; + return 0; +} + +static struct hlist_head *au_name_hash(struct au_nhash *nhash, + unsigned char *name, + unsigned int len) +{ + unsigned int v; + /* const unsigned int magic_bit = 12; */ + + AuDebugOn(!nhash->nh_num || !nhash->nh_head); + + v = 0; + while (len--) + v += *name++; + /* v = hash_long(v, magic_bit); */ + v %= nhash->nh_num; + return nhash->nh_head + v; +} + +static int au_nhash_test_name(struct au_vdir_destr *str, const char *name, + int nlen) +{ + return str->len == nlen && !memcmp(str->name, name, nlen); +} + +/* returns found or not */ +int au_nhash_test_known_wh(struct au_nhash *whlist, char *name, int nlen) +{ + struct hlist_head *head; + struct au_vdir_wh *pos; + struct au_vdir_destr *str; + + head = au_name_hash(whlist, name, nlen); + hlist_for_each_entry(pos, head, wh_hash) { + str = &pos->wh_str; + AuDbg("%.*s\n", str->len, str->name); + if (au_nhash_test_name(str, name, nlen)) + return 1; + } + return 0; +} + +/* returns found(true) or not */ +static int test_known(struct au_nhash *delist, char *name, int nlen) +{ + struct hlist_head *head; + struct au_vdir_dehstr *pos; + struct au_vdir_destr *str; + + head = au_name_hash(delist, name, nlen); + hlist_for_each_entry(pos, head, hash) { + str = pos->str; + AuDbg("%.*s\n", str->len, str->name); + if (au_nhash_test_name(str, name, nlen)) + return 1; + } + return 0; +} + +static void au_shwh_init_wh(struct au_vdir_wh *wh, ino_t ino, + unsigned char d_type) +{ +#ifdef CONFIG_AUFS_SHWH + wh->wh_ino = ino; + wh->wh_type = d_type; +#endif +} + +/* ---------------------------------------------------------------------- */ + +int au_nhash_append_wh(struct au_nhash *whlist, char *name, int nlen, ino_t ino, + unsigned int d_type, aufs_bindex_t bindex, + unsigned char shwh) +{ + int err; + struct au_vdir_destr *str; + struct au_vdir_wh *wh; + + AuDbg("%.*s\n", nlen, name); + AuDebugOn(!whlist->nh_num || !whlist->nh_head); + + err = -ENOMEM; + wh = kmalloc(sizeof(*wh) + nlen, GFP_NOFS); + if (unlikely(!wh)) + goto out; + + err = 0; + wh->wh_bindex = bindex; + if (shwh) + au_shwh_init_wh(wh, ino, d_type); + str = &wh->wh_str; + str->len = nlen; + memcpy(str->name, name, nlen); + hlist_add_head(&wh->wh_hash, au_name_hash(whlist, name, nlen)); + /* smp_mb(); */ + +out: + return err; +} + +static int append_deblk(struct au_vdir *vdir) +{ + int err; + unsigned long ul; + const unsigned int deblk_sz = vdir->vd_deblk_sz; + union au_vdir_deblk_p p, deblk_end; + unsigned char **o; + + err = -ENOMEM; + o = krealloc(vdir->vd_deblk, sizeof(*o) * (vdir->vd_nblk + 1), + GFP_NOFS); + if (unlikely(!o)) + goto out; + + vdir->vd_deblk = o; + p.deblk = kmalloc(deblk_sz, GFP_NOFS); + if (p.deblk) { + ul = vdir->vd_nblk++; + vdir->vd_deblk[ul] = p.deblk; + vdir->vd_last.ul = ul; + vdir->vd_last.p.deblk = p.deblk; + deblk_end.deblk = p.deblk + deblk_sz; + err = set_deblk_end(&p, &deblk_end); + } + +out: + return err; +} + +static int append_de(struct au_vdir *vdir, char *name, int nlen, ino_t ino, + unsigned int d_type, struct au_nhash *delist) +{ + int err; + unsigned int sz; + const unsigned int deblk_sz = vdir->vd_deblk_sz; + union au_vdir_deblk_p p, *room, deblk_end; + struct au_vdir_dehstr *dehstr; + + p.deblk = last_deblk(vdir); + deblk_end.deblk = p.deblk + deblk_sz; + room = &vdir->vd_last.p; + AuDebugOn(room->deblk < p.deblk || deblk_end.deblk <= room->deblk + || !is_deblk_end(room, &deblk_end)); + + sz = calc_size(nlen); + if (unlikely(sz > deblk_end.deblk - room->deblk)) { + err = append_deblk(vdir); + if (unlikely(err)) + goto out; + + p.deblk = last_deblk(vdir); + deblk_end.deblk = p.deblk + deblk_sz; + /* smp_mb(); */ + AuDebugOn(room->deblk != p.deblk); + } + + err = -ENOMEM; + dehstr = au_cache_alloc_vdir_dehstr(); + if (unlikely(!dehstr)) + goto out; + + dehstr->str = &room->de->de_str; + hlist_add_head(&dehstr->hash, au_name_hash(delist, name, nlen)); + room->de->de_ino = ino; + room->de->de_type = d_type; + room->de->de_str.len = nlen; + memcpy(room->de->de_str.name, name, nlen); + + err = 0; + room->deblk += sz; + if (unlikely(set_deblk_end(room, &deblk_end))) + err = append_deblk(vdir); + /* smp_mb(); */ + +out: + return err; +} + +/* ---------------------------------------------------------------------- */ + +void au_vdir_free(struct au_vdir *vdir) +{ + unsigned char **deblk; + + deblk = vdir->vd_deblk; + while (vdir->vd_nblk--) + kfree(*deblk++); + kfree(vdir->vd_deblk); + au_cache_free_vdir(vdir); +} + +static struct au_vdir *alloc_vdir(struct file *file) +{ + struct au_vdir *vdir; + struct super_block *sb; + int err; + + sb = file->f_path.dentry->d_sb; + SiMustAnyLock(sb); + + err = -ENOMEM; + vdir = au_cache_alloc_vdir(); + if (unlikely(!vdir)) + goto out; + + vdir->vd_deblk = kzalloc(sizeof(*vdir->vd_deblk), GFP_NOFS); + if (unlikely(!vdir->vd_deblk)) + goto out_free; + + vdir->vd_deblk_sz = au_sbi(sb)->si_rdblk; + if (!vdir->vd_deblk_sz) { + /* estimate the appropriate size for deblk */ + vdir->vd_deblk_sz = au_dir_size(file, /*dentry*/NULL); + /* pr_info("vd_deblk_sz %u\n", vdir->vd_deblk_sz); */ + } + vdir->vd_nblk = 0; + vdir->vd_version = 0; + vdir->vd_jiffy = 0; + err = append_deblk(vdir); + if (!err) + return vdir; /* success */ + + kfree(vdir->vd_deblk); + +out_free: + au_cache_free_vdir(vdir); +out: + vdir = ERR_PTR(err); + return vdir; +} + +static int reinit_vdir(struct au_vdir *vdir) +{ + int err; + union au_vdir_deblk_p p, deblk_end; + + while (vdir->vd_nblk > 1) { + kfree(vdir->vd_deblk[vdir->vd_nblk - 1]); + /* vdir->vd_deblk[vdir->vd_nblk - 1] = NULL; */ + vdir->vd_nblk--; + } + p.deblk = vdir->vd_deblk[0]; + deblk_end.deblk = p.deblk + vdir->vd_deblk_sz; + err = set_deblk_end(&p, &deblk_end); + /* keep vd_dblk_sz */ + vdir->vd_last.ul = 0; + vdir->vd_last.p.deblk = vdir->vd_deblk[0]; + vdir->vd_version = 0; + vdir->vd_jiffy = 0; + /* smp_mb(); */ + return err; +} + +/* ---------------------------------------------------------------------- */ + +#define AuFillVdir_CALLED 1 +#define AuFillVdir_WHABLE (1 << 1) +#define AuFillVdir_SHWH (1 << 2) +#define au_ftest_fillvdir(flags, name) ((flags) & AuFillVdir_##name) +#define au_fset_fillvdir(flags, name) \ + do { (flags) |= AuFillVdir_##name; } while (0) +#define au_fclr_fillvdir(flags, name) \ + do { (flags) &= ~AuFillVdir_##name; } while (0) + +#ifndef CONFIG_AUFS_SHWH +#undef AuFillVdir_SHWH +#define AuFillVdir_SHWH 0 +#endif + +struct fillvdir_arg { + struct dir_context ctx; + struct file *file; + struct au_vdir *vdir; + struct au_nhash delist; + struct au_nhash whlist; + aufs_bindex_t bindex; + unsigned int flags; + int err; +}; + +static int fillvdir(struct dir_context *ctx, const char *__name, int nlen, + loff_t offset __maybe_unused, u64 h_ino, + unsigned int d_type) +{ + struct fillvdir_arg *arg = container_of(ctx, struct fillvdir_arg, ctx); + char *name = (void *)__name; + struct super_block *sb; + ino_t ino; + const unsigned char shwh = !!au_ftest_fillvdir(arg->flags, SHWH); + + arg->err = 0; + sb = arg->file->f_path.dentry->d_sb; + au_fset_fillvdir(arg->flags, CALLED); + /* smp_mb(); */ + if (nlen <= AUFS_WH_PFX_LEN + || memcmp(name, AUFS_WH_PFX, AUFS_WH_PFX_LEN)) { + if (test_known(&arg->delist, name, nlen) + || au_nhash_test_known_wh(&arg->whlist, name, nlen)) + goto out; /* already exists or whiteouted */ + + arg->err = au_ino(sb, arg->bindex, h_ino, d_type, &ino); + if (!arg->err) { + if (unlikely(nlen > AUFS_MAX_NAMELEN)) + d_type = DT_UNKNOWN; + arg->err = append_de(arg->vdir, name, nlen, ino, + d_type, &arg->delist); + } + } else if (au_ftest_fillvdir(arg->flags, WHABLE)) { + name += AUFS_WH_PFX_LEN; + nlen -= AUFS_WH_PFX_LEN; + if (au_nhash_test_known_wh(&arg->whlist, name, nlen)) + goto out; /* already whiteouted */ + + if (shwh) + arg->err = au_wh_ino(sb, arg->bindex, h_ino, d_type, + &ino); + if (!arg->err) { + if (nlen <= AUFS_MAX_NAMELEN + AUFS_WH_PFX_LEN) + d_type = DT_UNKNOWN; + arg->err = au_nhash_append_wh + (&arg->whlist, name, nlen, ino, d_type, + arg->bindex, shwh); + } + } + +out: + if (!arg->err) + arg->vdir->vd_jiffy = jiffies; + /* smp_mb(); */ + AuTraceErr(arg->err); + return arg->err; +} + +static int au_handle_shwh(struct super_block *sb, struct au_vdir *vdir, + struct au_nhash *whlist, struct au_nhash *delist) +{ +#ifdef CONFIG_AUFS_SHWH + int err; + unsigned int nh, u; + struct hlist_head *head; + struct au_vdir_wh *pos; + struct hlist_node *n; + char *p, *o; + struct au_vdir_destr *destr; + + AuDebugOn(!au_opt_test(au_mntflags(sb), SHWH)); + + err = -ENOMEM; + o = p = (void *)__get_free_page(GFP_NOFS); + if (unlikely(!p)) + goto out; + + err = 0; + nh = whlist->nh_num; + memcpy(p, AUFS_WH_PFX, AUFS_WH_PFX_LEN); + p += AUFS_WH_PFX_LEN; + for (u = 0; u < nh; u++) { + head = whlist->nh_head + u; + hlist_for_each_entry_safe(pos, n, head, wh_hash) { + destr = &pos->wh_str; + memcpy(p, destr->name, destr->len); + err = append_de(vdir, o, destr->len + AUFS_WH_PFX_LEN, + pos->wh_ino, pos->wh_type, delist); + if (unlikely(err)) + break; + } + } + + free_page((unsigned long)o); + +out: + AuTraceErr(err); + return err; +#else + return 0; +#endif +} + +static int au_do_read_vdir(struct fillvdir_arg *arg) +{ + int err; + unsigned int rdhash; + loff_t offset; + aufs_bindex_t bend, bindex, bstart; + unsigned char shwh; + struct file *hf, *file; + struct super_block *sb; + + file = arg->file; + sb = file->f_path.dentry->d_sb; + SiMustAnyLock(sb); + + rdhash = au_sbi(sb)->si_rdhash; + if (!rdhash) + rdhash = au_rdhash_est(au_dir_size(file, /*dentry*/NULL)); + err = au_nhash_alloc(&arg->delist, rdhash, GFP_NOFS); + if (unlikely(err)) + goto out; + err = au_nhash_alloc(&arg->whlist, rdhash, GFP_NOFS); + if (unlikely(err)) + goto out_delist; + + err = 0; + arg->flags = 0; + shwh = 0; + if (au_opt_test(au_mntflags(sb), SHWH)) { + shwh = 1; + au_fset_fillvdir(arg->flags, SHWH); + } + bstart = au_fbstart(file); + bend = au_fbend_dir(file); + for (bindex = bstart; !err && bindex <= bend; bindex++) { + hf = au_hf_dir(file, bindex); + if (!hf) + continue; + + offset = vfsub_llseek(hf, 0, SEEK_SET); + err = offset; + if (unlikely(offset)) + break; + + arg->bindex = bindex; + au_fclr_fillvdir(arg->flags, WHABLE); + if (shwh + || (bindex != bend + && au_br_whable(au_sbr_perm(sb, bindex)))) + au_fset_fillvdir(arg->flags, WHABLE); + do { + arg->err = 0; + au_fclr_fillvdir(arg->flags, CALLED); + /* smp_mb(); */ + err = vfsub_iterate_dir(hf, &arg->ctx); + if (err >= 0) + err = arg->err; + } while (!err && au_ftest_fillvdir(arg->flags, CALLED)); + + /* + * dir_relax() may be good for concurrency, but aufs should not + * use it since it will cause a lockdep problem. + */ + } + + if (!err && shwh) + err = au_handle_shwh(sb, arg->vdir, &arg->whlist, &arg->delist); + + au_nhash_wh_free(&arg->whlist); + +out_delist: + au_nhash_de_free(&arg->delist); +out: + return err; +} + +static int read_vdir(struct file *file, int may_read) +{ + int err; + unsigned long expire; + unsigned char do_read; + struct fillvdir_arg arg = { + .ctx = { + .actor = fillvdir + } + }; + struct inode *inode; + struct au_vdir *vdir, *allocated; + + err = 0; + inode = file_inode(file); + IMustLock(inode); + SiMustAnyLock(inode->i_sb); + + allocated = NULL; + do_read = 0; + expire = au_sbi(inode->i_sb)->si_rdcache; + vdir = au_ivdir(inode); + if (!vdir) { + do_read = 1; + vdir = alloc_vdir(file); + err = PTR_ERR(vdir); + if (IS_ERR(vdir)) + goto out; + err = 0; + allocated = vdir; + } else if (may_read + && (inode->i_version != vdir->vd_version + || time_after(jiffies, vdir->vd_jiffy + expire))) { + do_read = 1; + err = reinit_vdir(vdir); + if (unlikely(err)) + goto out; + } + + if (!do_read) + return 0; /* success */ + + arg.file = file; + arg.vdir = vdir; + err = au_do_read_vdir(&arg); + if (!err) { + /* file->f_pos = 0; */ /* todo: ctx->pos? */ + vdir->vd_version = inode->i_version; + vdir->vd_last.ul = 0; + vdir->vd_last.p.deblk = vdir->vd_deblk[0]; + if (allocated) + au_set_ivdir(inode, allocated); + } else if (allocated) + au_vdir_free(allocated); + +out: + return err; +} + +static int copy_vdir(struct au_vdir *tgt, struct au_vdir *src) +{ + int err, rerr; + unsigned long ul, n; + const unsigned int deblk_sz = src->vd_deblk_sz; + + AuDebugOn(tgt->vd_nblk != 1); + + err = -ENOMEM; + if (tgt->vd_nblk < src->vd_nblk) { + unsigned char **p; + + p = krealloc(tgt->vd_deblk, sizeof(*p) * src->vd_nblk, + GFP_NOFS); + if (unlikely(!p)) + goto out; + tgt->vd_deblk = p; + } + + if (tgt->vd_deblk_sz != deblk_sz) { + unsigned char *p; + + tgt->vd_deblk_sz = deblk_sz; + p = krealloc(tgt->vd_deblk[0], deblk_sz, GFP_NOFS); + if (unlikely(!p)) + goto out; + tgt->vd_deblk[0] = p; + } + memcpy(tgt->vd_deblk[0], src->vd_deblk[0], deblk_sz); + tgt->vd_version = src->vd_version; + tgt->vd_jiffy = src->vd_jiffy; + + n = src->vd_nblk; + for (ul = 1; ul < n; ul++) { + tgt->vd_deblk[ul] = kmemdup(src->vd_deblk[ul], deblk_sz, + GFP_NOFS); + if (unlikely(!tgt->vd_deblk[ul])) + goto out; + tgt->vd_nblk++; + } + tgt->vd_nblk = n; + tgt->vd_last.ul = tgt->vd_last.ul; + tgt->vd_last.p.deblk = tgt->vd_deblk[tgt->vd_last.ul]; + tgt->vd_last.p.deblk += src->vd_last.p.deblk + - src->vd_deblk[src->vd_last.ul]; + /* smp_mb(); */ + return 0; /* success */ + +out: + rerr = reinit_vdir(tgt); + BUG_ON(rerr); + return err; +} + +int au_vdir_init(struct file *file) +{ + int err; + struct inode *inode; + struct au_vdir *vdir_cache, *allocated; + + /* test file->f_pos here instead of ctx->pos */ + err = read_vdir(file, !file->f_pos); + if (unlikely(err)) + goto out; + + allocated = NULL; + vdir_cache = au_fvdir_cache(file); + if (!vdir_cache) { + vdir_cache = alloc_vdir(file); + err = PTR_ERR(vdir_cache); + if (IS_ERR(vdir_cache)) + goto out; + allocated = vdir_cache; + } else if (!file->f_pos && vdir_cache->vd_version != file->f_version) { + /* test file->f_pos here instead of ctx->pos */ + err = reinit_vdir(vdir_cache); + if (unlikely(err)) + goto out; + } else + return 0; /* success */ + + inode = file_inode(file); + err = copy_vdir(vdir_cache, au_ivdir(inode)); + if (!err) { + file->f_version = inode->i_version; + if (allocated) + au_set_fvdir_cache(file, allocated); + } else if (allocated) + au_vdir_free(allocated); + +out: + return err; +} + +static loff_t calc_offset(struct au_vdir *vdir) +{ + loff_t offset; + union au_vdir_deblk_p p; + + p.deblk = vdir->vd_deblk[vdir->vd_last.ul]; + offset = vdir->vd_last.p.deblk - p.deblk; + offset += vdir->vd_deblk_sz * vdir->vd_last.ul; + return offset; +} + +/* returns true or false */ +static int seek_vdir(struct file *file, struct dir_context *ctx) +{ + int valid; + unsigned int deblk_sz; + unsigned long ul, n; + loff_t offset; + union au_vdir_deblk_p p, deblk_end; + struct au_vdir *vdir_cache; + + valid = 1; + vdir_cache = au_fvdir_cache(file); + offset = calc_offset(vdir_cache); + AuDbg("offset %lld\n", offset); + if (ctx->pos == offset) + goto out; + + vdir_cache->vd_last.ul = 0; + vdir_cache->vd_last.p.deblk = vdir_cache->vd_deblk[0]; + if (!ctx->pos) + goto out; + + valid = 0; + deblk_sz = vdir_cache->vd_deblk_sz; + ul = div64_u64(ctx->pos, deblk_sz); + AuDbg("ul %lu\n", ul); + if (ul >= vdir_cache->vd_nblk) + goto out; + + n = vdir_cache->vd_nblk; + for (; ul < n; ul++) { + p.deblk = vdir_cache->vd_deblk[ul]; + deblk_end.deblk = p.deblk + deblk_sz; + offset = ul; + offset *= deblk_sz; + while (!is_deblk_end(&p, &deblk_end) && offset < ctx->pos) { + unsigned int l; + + l = calc_size(p.de->de_str.len); + offset += l; + p.deblk += l; + } + if (!is_deblk_end(&p, &deblk_end)) { + valid = 1; + vdir_cache->vd_last.ul = ul; + vdir_cache->vd_last.p = p; + break; + } + } + +out: + /* smp_mb(); */ + AuTraceErr(!valid); + return valid; +} + +int au_vdir_fill_de(struct file *file, struct dir_context *ctx) +{ + unsigned int l, deblk_sz; + union au_vdir_deblk_p deblk_end; + struct au_vdir *vdir_cache; + struct au_vdir_de *de; + + vdir_cache = au_fvdir_cache(file); + if (!seek_vdir(file, ctx)) + return 0; + + deblk_sz = vdir_cache->vd_deblk_sz; + while (1) { + deblk_end.deblk = vdir_cache->vd_deblk[vdir_cache->vd_last.ul]; + deblk_end.deblk += deblk_sz; + while (!is_deblk_end(&vdir_cache->vd_last.p, &deblk_end)) { + de = vdir_cache->vd_last.p.de; + AuDbg("%.*s, off%lld, i%lu, dt%d\n", + de->de_str.len, de->de_str.name, ctx->pos, + (unsigned long)de->de_ino, de->de_type); + if (unlikely(!dir_emit(ctx, de->de_str.name, + de->de_str.len, de->de_ino, + de->de_type))) { + /* todo: ignore the error caused by udba? */ + /* return err; */ + return 0; + } + + l = calc_size(de->de_str.len); + vdir_cache->vd_last.p.deblk += l; + ctx->pos += l; + } + if (vdir_cache->vd_last.ul < vdir_cache->vd_nblk - 1) { + vdir_cache->vd_last.ul++; + vdir_cache->vd_last.p.deblk + = vdir_cache->vd_deblk[vdir_cache->vd_last.ul]; + ctx->pos = deblk_sz * vdir_cache->vd_last.ul; + continue; + } + break; + } + + /* smp_mb(); */ + return 0; +} diff --git a/fs/aufs/vfsub.c b/fs/aufs/vfsub.c new file mode 100644 index 000000000..29e0bbc32 --- /dev/null +++ b/fs/aufs/vfsub.c @@ -0,0 +1,871 @@ +/* + * Copyright (C) 2005-2016 Junjiro R. Okajima + */ + +/* + * sub-routines for VFS + */ + +#include <linux/namei.h> +#include <linux/nsproxy.h> +#include <linux/security.h> +#include <linux/splice.h> +#include "../fs/mount.h" +#include "aufs.h" + +#ifdef CONFIG_AUFS_BR_FUSE +int vfsub_test_mntns(struct vfsmount *mnt, struct super_block *h_sb) +{ + struct nsproxy *ns; + + if (!au_test_fuse(h_sb) || !au_userns) + return 0; + + ns = current->nsproxy; + /* no {get,put}_nsproxy(ns) */ + return real_mount(mnt)->mnt_ns == ns->mnt_ns ? 0 : -EACCES; +} +#endif + +/* ---------------------------------------------------------------------- */ + +int vfsub_update_h_iattr(struct path *h_path, int *did) +{ + int err; + struct kstat st; + struct super_block *h_sb; + + /* for remote fs, leave work for its getattr or d_revalidate */ + /* for bad i_attr fs, handle them in aufs_getattr() */ + /* still some fs may acquire i_mutex. we need to skip them */ + err = 0; + if (!did) + did = &err; + h_sb = h_path->dentry->d_sb; + *did = (!au_test_fs_remote(h_sb) && au_test_fs_refresh_iattr(h_sb)); + if (*did) + err = vfs_getattr(h_path, &st); + + return err; +} + +/* ---------------------------------------------------------------------- */ + +struct file *vfsub_dentry_open(struct path *path, int flags) +{ + struct file *file; + + file = dentry_open(path, flags /* | __FMODE_NONOTIFY */, + current_cred()); + if (!IS_ERR_OR_NULL(file) + && (file->f_mode & (FMODE_READ | FMODE_WRITE)) == FMODE_READ) + i_readcount_inc(d_inode(path->dentry)); + + return file; +} + +struct file *vfsub_filp_open(const char *path, int oflags, int mode) +{ + struct file *file; + + lockdep_off(); + file = filp_open(path, + oflags /* | __FMODE_NONOTIFY */, + mode); + lockdep_on(); + if (IS_ERR(file)) + goto out; + vfsub_update_h_iattr(&file->f_path, /*did*/NULL); /*ignore*/ + +out: + return file; +} + +/* + * Ideally this function should call VFS:do_last() in order to keep all its + * checkings. But it is very hard for aufs to regenerate several VFS internal + * structure such as nameidata. This is a second (or third) best approach. + * cf. linux/fs/namei.c:do_last(), lookup_open() and atomic_open(). + */ +int vfsub_atomic_open(struct inode *dir, struct dentry *dentry, + struct vfsub_aopen_args *args, struct au_branch *br) +{ + int err; + struct file *file = args->file; + /* copied from linux/fs/namei.c:atomic_open() */ + struct dentry *const DENTRY_NOT_SET = (void *)-1UL; + + IMustLock(dir); + AuDebugOn(!dir->i_op->atomic_open); + + err = au_br_test_oflag(args->open_flag, br); + if (unlikely(err)) + goto out; + + args->file->f_path.dentry = DENTRY_NOT_SET; + args->file->f_path.mnt = au_br_mnt(br); + err = dir->i_op->atomic_open(dir, dentry, file, args->open_flag, + args->create_mode, args->opened); + if (err >= 0) { + /* some filesystems don't set FILE_CREATED while succeeded? */ + if (*args->opened & FILE_CREATED) + fsnotify_create(dir, dentry); + } else + goto out; + + + if (!err) { + /* todo: call VFS:may_open() here */ + err = open_check_o_direct(file); + /* todo: ima_file_check() too? */ + if (!err && (args->open_flag & __FMODE_EXEC)) + err = deny_write_access(file); + if (unlikely(err)) + /* note that the file is created and still opened */ + goto out; + } + + atomic_inc(&br->br_count); + fsnotify_open(file); + +out: + return err; +} + +int vfsub_kern_path(const char *name, unsigned int flags, struct path *path) +{ + int err; + + err = kern_path(name, flags, path); + if (!err && d_is_positive(path->dentry)) + vfsub_update_h_iattr(path, /*did*/NULL); /*ignore*/ + return err; +} + +struct dentry *vfsub_lookup_one_len_unlocked(const char *name, + struct dentry *parent, int len) +{ + struct path path = { + .mnt = NULL + }; + + path.dentry = lookup_one_len_unlocked(name, parent, len); + if (IS_ERR(path.dentry)) + goto out; + if (d_is_positive(path.dentry)) + vfsub_update_h_iattr(&path, /*did*/NULL); /*ignore*/ + +out: + AuTraceErrPtr(path.dentry); + return path.dentry; +} + +struct dentry *vfsub_lookup_one_len(const char *name, struct dentry *parent, + int len) +{ + struct path path = { + .mnt = NULL + }; + + /* VFS checks it too, but by WARN_ON_ONCE() */ + IMustLock(d_inode(parent)); + + path.dentry = lookup_one_len(name, parent, len); + if (IS_ERR(path.dentry)) + goto out; + if (d_is_positive(path.dentry)) + vfsub_update_h_iattr(&path, /*did*/NULL); /*ignore*/ + +out: + AuTraceErrPtr(path.dentry); + return path.dentry; +} + +void vfsub_call_lkup_one(void *args) +{ + struct vfsub_lkup_one_args *a = args; + *a->errp = vfsub_lkup_one(a->name, a->parent); +} + +/* ---------------------------------------------------------------------- */ + +struct dentry *vfsub_lock_rename(struct dentry *d1, struct au_hinode *hdir1, + struct dentry *d2, struct au_hinode *hdir2) +{ + struct dentry *d; + + lockdep_off(); + d = lock_rename(d1, d2); + lockdep_on(); + au_hn_suspend(hdir1); + if (hdir1 != hdir2) + au_hn_suspend(hdir2); + + return d; +} + +void vfsub_unlock_rename(struct dentry *d1, struct au_hinode *hdir1, + struct dentry *d2, struct au_hinode *hdir2) +{ + au_hn_resume(hdir1); + if (hdir1 != hdir2) + au_hn_resume(hdir2); + lockdep_off(); + unlock_rename(d1, d2); + lockdep_on(); +} + +/* ---------------------------------------------------------------------- */ + +int vfsub_create(struct inode *dir, struct path *path, int mode, bool want_excl) +{ + int err; + struct dentry *d; + + IMustLock(dir); + + d = path->dentry; + path->dentry = d->d_parent; + err = security_path_mknod(path, d, mode, 0); + path->dentry = d; + if (unlikely(err)) + goto out; + + lockdep_off(); + err = vfs_create(dir, path->dentry, mode, want_excl); + lockdep_on(); + if (!err) { + struct path tmp = *path; + int did; + + vfsub_update_h_iattr(&tmp, &did); + if (did) { + tmp.dentry = path->dentry->d_parent; + vfsub_update_h_iattr(&tmp, /*did*/NULL); + } + /*ignore*/ + } + +out: + return err; +} + +int vfsub_symlink(struct inode *dir, struct path *path, const char *symname) +{ + int err; + struct dentry *d; + + IMustLock(dir); + + d = path->dentry; + path->dentry = d->d_parent; + err = security_path_symlink(path, d, symname); + path->dentry = d; + if (unlikely(err)) + goto out; + + lockdep_off(); + err = vfs_symlink(dir, path->dentry, symname); + lockdep_on(); + if (!err) { + struct path tmp = *path; + int did; + + vfsub_update_h_iattr(&tmp, &did); + if (did) { + tmp.dentry = path->dentry->d_parent; + vfsub_update_h_iattr(&tmp, /*did*/NULL); + } + /*ignore*/ + } + +out: + return err; +} + +int vfsub_mknod(struct inode *dir, struct path *path, int mode, dev_t dev) +{ + int err; + struct dentry *d; + + IMustLock(dir); + + d = path->dentry; + path->dentry = d->d_parent; + err = security_path_mknod(path, d, mode, new_encode_dev(dev)); + path->dentry = d; + if (unlikely(err)) + goto out; + + lockdep_off(); + err = vfs_mknod(dir, path->dentry, mode, dev); + lockdep_on(); + if (!err) { + struct path tmp = *path; + int did; + + vfsub_update_h_iattr(&tmp, &did); + if (did) { + tmp.dentry = path->dentry->d_parent; + vfsub_update_h_iattr(&tmp, /*did*/NULL); + } + /*ignore*/ + } + +out: + return err; +} + +static int au_test_nlink(struct inode *inode) +{ + const unsigned int link_max = UINT_MAX >> 1; /* rough margin */ + + if (!au_test_fs_no_limit_nlink(inode->i_sb) + || inode->i_nlink < link_max) + return 0; + return -EMLINK; +} + +int vfsub_link(struct dentry *src_dentry, struct inode *dir, struct path *path, + struct inode **delegated_inode) +{ + int err; + struct dentry *d; + + IMustLock(dir); + + err = au_test_nlink(d_inode(src_dentry)); + if (unlikely(err)) + return err; + + /* we don't call may_linkat() */ + d = path->dentry; + path->dentry = d->d_parent; + err = security_path_link(src_dentry, path, d); + path->dentry = d; + if (unlikely(err)) + goto out; + + lockdep_off(); + err = vfs_link(src_dentry, dir, path->dentry, delegated_inode); + lockdep_on(); + if (!err) { + struct path tmp = *path; + int did; + + /* fuse has different memory inode for the same inumber */ + vfsub_update_h_iattr(&tmp, &did); + if (did) { + tmp.dentry = path->dentry->d_parent; + vfsub_update_h_iattr(&tmp, /*did*/NULL); + tmp.dentry = src_dentry; + vfsub_update_h_iattr(&tmp, /*did*/NULL); + } + /*ignore*/ + } + +out: + return err; +} + +int vfsub_rename(struct inode *src_dir, struct dentry *src_dentry, + struct inode *dir, struct path *path, + struct inode **delegated_inode) +{ + int err; + struct path tmp = { + .mnt = path->mnt + }; + struct dentry *d; + + IMustLock(dir); + IMustLock(src_dir); + + d = path->dentry; + path->dentry = d->d_parent; + tmp.dentry = src_dentry->d_parent; + err = security_path_rename(&tmp, src_dentry, path, d, /*flags*/0); + path->dentry = d; + if (unlikely(err)) + goto out; + + lockdep_off(); + err = vfs_rename(src_dir, src_dentry, dir, path->dentry, + delegated_inode, /*flags*/0); + lockdep_on(); + if (!err) { + int did; + + tmp.dentry = d->d_parent; + vfsub_update_h_iattr(&tmp, &did); + if (did) { + tmp.dentry = src_dentry; + vfsub_update_h_iattr(&tmp, /*did*/NULL); + tmp.dentry = src_dentry->d_parent; + vfsub_update_h_iattr(&tmp, /*did*/NULL); + } + /*ignore*/ + } + +out: + return err; +} + +int vfsub_mkdir(struct inode *dir, struct path *path, int mode) +{ + int err; + struct dentry *d; + + IMustLock(dir); + + d = path->dentry; + path->dentry = d->d_parent; + err = security_path_mkdir(path, d, mode); + path->dentry = d; + if (unlikely(err)) + goto out; + + lockdep_off(); + err = vfs_mkdir(dir, path->dentry, mode); + lockdep_on(); + if (!err) { + struct path tmp = *path; + int did; + + vfsub_update_h_iattr(&tmp, &did); + if (did) { + tmp.dentry = path->dentry->d_parent; + vfsub_update_h_iattr(&tmp, /*did*/NULL); + } + /*ignore*/ + } + +out: + return err; +} + +int vfsub_rmdir(struct inode *dir, struct path *path) +{ + int err; + struct dentry *d; + + IMustLock(dir); + + d = path->dentry; + path->dentry = d->d_parent; + err = security_path_rmdir(path, d); + path->dentry = d; + if (unlikely(err)) + goto out; + + lockdep_off(); + err = vfs_rmdir(dir, path->dentry); + lockdep_on(); + if (!err) { + struct path tmp = { + .dentry = path->dentry->d_parent, + .mnt = path->mnt + }; + + vfsub_update_h_iattr(&tmp, /*did*/NULL); /*ignore*/ + } + +out: + return err; +} + +/* ---------------------------------------------------------------------- */ + +/* todo: support mmap_sem? */ +ssize_t vfsub_read_u(struct file *file, char __user *ubuf, size_t count, + loff_t *ppos) +{ + ssize_t err; + + lockdep_off(); + err = vfs_read(file, ubuf, count, ppos); + lockdep_on(); + if (err >= 0) + vfsub_update_h_iattr(&file->f_path, /*did*/NULL); /*ignore*/ + return err; +} + +/* todo: kernel_read()? */ +ssize_t vfsub_read_k(struct file *file, void *kbuf, size_t count, + loff_t *ppos) +{ + ssize_t err; + mm_segment_t oldfs; + union { + void *k; + char __user *u; + } buf; + + buf.k = kbuf; + oldfs = get_fs(); + set_fs(KERNEL_DS); + err = vfsub_read_u(file, buf.u, count, ppos); + set_fs(oldfs); + return err; +} + +ssize_t vfsub_write_u(struct file *file, const char __user *ubuf, size_t count, + loff_t *ppos) +{ + ssize_t err; + + lockdep_off(); + err = vfs_write(file, ubuf, count, ppos); + lockdep_on(); + if (err >= 0) + vfsub_update_h_iattr(&file->f_path, /*did*/NULL); /*ignore*/ + return err; +} + +ssize_t vfsub_write_k(struct file *file, void *kbuf, size_t count, loff_t *ppos) +{ + ssize_t err; + mm_segment_t oldfs; + union { + void *k; + const char __user *u; + } buf; + + buf.k = kbuf; + oldfs = get_fs(); + set_fs(KERNEL_DS); + err = vfsub_write_u(file, buf.u, count, ppos); + set_fs(oldfs); + return err; +} + +int vfsub_flush(struct file *file, fl_owner_t id) +{ + int err; + + err = 0; + if (file->f_op->flush) { + if (!au_test_nfs(file->f_path.dentry->d_sb)) + err = file->f_op->flush(file, id); + else { + lockdep_off(); + err = file->f_op->flush(file, id); + lockdep_on(); + } + if (!err) + vfsub_update_h_iattr(&file->f_path, /*did*/NULL); + /*ignore*/ + } + return err; +} + +int vfsub_iterate_dir(struct file *file, struct dir_context *ctx) +{ + int err; + + AuDbg("%pD, ctx{%pf, %llu}\n", file, ctx->actor, ctx->pos); + + lockdep_off(); + err = iterate_dir(file, ctx); + lockdep_on(); + if (err >= 0) + vfsub_update_h_iattr(&file->f_path, /*did*/NULL); /*ignore*/ + return err; +} + +long vfsub_splice_to(struct file *in, loff_t *ppos, + struct pipe_inode_info *pipe, size_t len, + unsigned int flags) +{ + long err; + + lockdep_off(); + err = do_splice_to(in, ppos, pipe, len, flags); + lockdep_on(); + file_accessed(in); + if (err >= 0) + vfsub_update_h_iattr(&in->f_path, /*did*/NULL); /*ignore*/ + return err; +} + +long vfsub_splice_from(struct pipe_inode_info *pipe, struct file *out, + loff_t *ppos, size_t len, unsigned int flags) +{ + long err; + + lockdep_off(); + err = do_splice_from(pipe, out, ppos, len, flags); + lockdep_on(); + if (err >= 0) + vfsub_update_h_iattr(&out->f_path, /*did*/NULL); /*ignore*/ + return err; +} + +int vfsub_fsync(struct file *file, struct path *path, int datasync) +{ + int err; + + /* file can be NULL */ + lockdep_off(); + err = vfs_fsync(file, datasync); + lockdep_on(); + if (!err) { + if (!path) { + AuDebugOn(!file); + path = &file->f_path; + } + vfsub_update_h_iattr(path, /*did*/NULL); /*ignore*/ + } + return err; +} + +/* cf. open.c:do_sys_truncate() and do_sys_ftruncate() */ +int vfsub_trunc(struct path *h_path, loff_t length, unsigned int attr, + struct file *h_file) +{ + int err; + struct inode *h_inode; + struct super_block *h_sb; + + if (!h_file) { + err = vfsub_truncate(h_path, length); + goto out; + } + + h_inode = d_inode(h_path->dentry); + h_sb = h_inode->i_sb; + lockdep_off(); + sb_start_write(h_sb); + lockdep_on(); + err = locks_verify_truncate(h_inode, h_file, length); + if (!err) + err = security_path_truncate(h_path); + if (!err) { + lockdep_off(); + err = do_truncate(h_path->dentry, length, attr, h_file); + lockdep_on(); + } + lockdep_off(); + sb_end_write(h_sb); + lockdep_on(); + +out: + return err; +} + +/* ---------------------------------------------------------------------- */ + +struct au_vfsub_mkdir_args { + int *errp; + struct inode *dir; + struct path *path; + int mode; +}; + +static void au_call_vfsub_mkdir(void *args) +{ + struct au_vfsub_mkdir_args *a = args; + *a->errp = vfsub_mkdir(a->dir, a->path, a->mode); +} + +int vfsub_sio_mkdir(struct inode *dir, struct path *path, int mode) +{ + int err, do_sio, wkq_err; + + do_sio = au_test_h_perm_sio(dir, MAY_EXEC | MAY_WRITE); + if (!do_sio) { + lockdep_off(); + err = vfsub_mkdir(dir, path, mode); + lockdep_on(); + } else { + struct au_vfsub_mkdir_args args = { + .errp = &err, + .dir = dir, + .path = path, + .mode = mode + }; + wkq_err = au_wkq_wait(au_call_vfsub_mkdir, &args); + if (unlikely(wkq_err)) + err = wkq_err; + } + + return err; +} + +struct au_vfsub_rmdir_args { + int *errp; + struct inode *dir; + struct path *path; +}; + +static void au_call_vfsub_rmdir(void *args) +{ + struct au_vfsub_rmdir_args *a = args; + *a->errp = vfsub_rmdir(a->dir, a->path); +} + +int vfsub_sio_rmdir(struct inode *dir, struct path *path) +{ + int err, do_sio, wkq_err; + + do_sio = au_test_h_perm_sio(dir, MAY_EXEC | MAY_WRITE); + if (!do_sio) { + lockdep_off(); + err = vfsub_rmdir(dir, path); + lockdep_on(); + } else { + struct au_vfsub_rmdir_args args = { + .errp = &err, + .dir = dir, + .path = path + }; + wkq_err = au_wkq_wait(au_call_vfsub_rmdir, &args); + if (unlikely(wkq_err)) + err = wkq_err; + } + + return err; +} + +/* ---------------------------------------------------------------------- */ + +struct notify_change_args { + int *errp; + struct path *path; + struct iattr *ia; + struct inode **delegated_inode; +}; + +static void call_notify_change(void *args) +{ + struct notify_change_args *a = args; + struct inode *h_inode; + + h_inode = d_inode(a->path->dentry); + IMustLock(h_inode); + + *a->errp = -EPERM; + if (!IS_IMMUTABLE(h_inode) && !IS_APPEND(h_inode)) { + lockdep_off(); + *a->errp = notify_change(a->path->dentry, a->ia, + a->delegated_inode); + lockdep_on(); + if (!*a->errp) + vfsub_update_h_iattr(a->path, /*did*/NULL); /*ignore*/ + } + AuTraceErr(*a->errp); +} + +int vfsub_notify_change(struct path *path, struct iattr *ia, + struct inode **delegated_inode) +{ + int err; + struct notify_change_args args = { + .errp = &err, + .path = path, + .ia = ia, + .delegated_inode = delegated_inode + }; + + call_notify_change(&args); + + return err; +} + +int vfsub_sio_notify_change(struct path *path, struct iattr *ia, + struct inode **delegated_inode) +{ + int err, wkq_err; + struct notify_change_args args = { + .errp = &err, + .path = path, + .ia = ia, + .delegated_inode = delegated_inode + }; + + wkq_err = au_wkq_wait(call_notify_change, &args); + if (unlikely(wkq_err)) + err = wkq_err; + + return err; +} + +/* ---------------------------------------------------------------------- */ + +struct unlink_args { + int *errp; + struct inode *dir; + struct path *path; + struct inode **delegated_inode; +}; + +static void call_unlink(void *args) +{ + struct unlink_args *a = args; + struct dentry *d = a->path->dentry; + struct inode *h_inode; + const int stop_sillyrename = (au_test_nfs(d->d_sb) + && au_dcount(d) == 1); + + IMustLock(a->dir); + + a->path->dentry = d->d_parent; + *a->errp = security_path_unlink(a->path, d); + a->path->dentry = d; + if (unlikely(*a->errp)) + return; + + if (!stop_sillyrename) + dget(d); + h_inode = NULL; + if (d_is_positive(d)) { + h_inode = d_inode(d); + ihold(h_inode); + } + + lockdep_off(); + *a->errp = vfs_unlink(a->dir, d, a->delegated_inode); + lockdep_on(); + if (!*a->errp) { + struct path tmp = { + .dentry = d->d_parent, + .mnt = a->path->mnt + }; + vfsub_update_h_iattr(&tmp, /*did*/NULL); /*ignore*/ + } + + if (!stop_sillyrename) + dput(d); + if (h_inode) + iput(h_inode); + + AuTraceErr(*a->errp); +} + +/* + * @dir: must be locked. + * @dentry: target dentry. + */ +int vfsub_unlink(struct inode *dir, struct path *path, + struct inode **delegated_inode, int force) +{ + int err; + struct unlink_args args = { + .errp = &err, + .dir = dir, + .path = path, + .delegated_inode = delegated_inode + }; + + if (!force) + call_unlink(&args); + else { + int wkq_err; + + wkq_err = au_wkq_wait(call_unlink, &args); + if (unlikely(wkq_err)) + err = wkq_err; + } + + return err; +} diff --git a/fs/aufs/vfsub.h b/fs/aufs/vfsub.h new file mode 100644 index 000000000..4e08d33c2 --- /dev/null +++ b/fs/aufs/vfsub.h @@ -0,0 +1,297 @@ +/* + * Copyright (C) 2005-2016 Junjiro R. Okajima + */ + +/* + * sub-routines for VFS + */ + +#ifndef __AUFS_VFSUB_H__ +#define __AUFS_VFSUB_H__ + +#ifdef __KERNEL__ + +#include <linux/fs.h> +#include <linux/mount.h> +#include <linux/posix_acl.h> +#include <linux/xattr.h> +#include "debug.h" + +/* copied from linux/fs/internal.h */ +/* todo: BAD approach!! */ +extern void __mnt_drop_write(struct vfsmount *); +extern int open_check_o_direct(struct file *f); + +/* ---------------------------------------------------------------------- */ + +/* lock subclass for lower inode */ +/* default MAX_LOCKDEP_SUBCLASSES(8) is not enough */ +/* reduce? gave up. */ +enum { + AuLsc_I_Begin = I_MUTEX_PARENT2, /* 5 */ + AuLsc_I_PARENT, /* lower inode, parent first */ + AuLsc_I_PARENT2, /* copyup dirs */ + AuLsc_I_PARENT3, /* copyup wh */ + AuLsc_I_CHILD, + AuLsc_I_CHILD2, + AuLsc_I_End +}; + +/* to debug easier, do not make them inlined functions */ +#define MtxMustLock(mtx) AuDebugOn(!mutex_is_locked(mtx)) +#define IMustLock(i) AuDebugOn(!inode_is_locked(i)) + +/* ---------------------------------------------------------------------- */ + +static inline void vfsub_drop_nlink(struct inode *inode) +{ + AuDebugOn(!inode->i_nlink); + drop_nlink(inode); +} + +static inline void vfsub_dead_dir(struct inode *inode) +{ + AuDebugOn(!S_ISDIR(inode->i_mode)); + inode->i_flags |= S_DEAD; + clear_nlink(inode); +} + +static inline int vfsub_native_ro(struct inode *inode) +{ + return (inode->i_sb->s_flags & MS_RDONLY) + || IS_RDONLY(inode) + /* || IS_APPEND(inode) */ + || IS_IMMUTABLE(inode); +} + +#ifdef CONFIG_AUFS_BR_FUSE +int vfsub_test_mntns(struct vfsmount *mnt, struct super_block *h_sb); +#else +AuStubInt0(vfsub_test_mntns, struct vfsmount *mnt, struct super_block *h_sb); +#endif + +/* ---------------------------------------------------------------------- */ + +int vfsub_update_h_iattr(struct path *h_path, int *did); +struct file *vfsub_dentry_open(struct path *path, int flags); +struct file *vfsub_filp_open(const char *path, int oflags, int mode); +struct vfsub_aopen_args { + struct file *file; + unsigned int open_flag; + umode_t create_mode; + int *opened; +}; +struct au_branch; +int vfsub_atomic_open(struct inode *dir, struct dentry *dentry, + struct vfsub_aopen_args *args, struct au_branch *br); +int vfsub_kern_path(const char *name, unsigned int flags, struct path *path); + +struct dentry *vfsub_lookup_one_len_unlocked(const char *name, + struct dentry *parent, int len); +struct dentry *vfsub_lookup_one_len(const char *name, struct dentry *parent, + int len); + +struct vfsub_lkup_one_args { + struct dentry **errp; + struct qstr *name; + struct dentry *parent; +}; + +static inline struct dentry *vfsub_lkup_one(struct qstr *name, + struct dentry *parent) +{ + return vfsub_lookup_one_len(name->name, parent, name->len); +} + +void vfsub_call_lkup_one(void *args); + +/* ---------------------------------------------------------------------- */ + +static inline int vfsub_mnt_want_write(struct vfsmount *mnt) +{ + int err; + + lockdep_off(); + err = mnt_want_write(mnt); + lockdep_on(); + return err; +} + +static inline void vfsub_mnt_drop_write(struct vfsmount *mnt) +{ + lockdep_off(); + mnt_drop_write(mnt); + lockdep_on(); +} + +#if 0 /* reserved */ +static inline void vfsub_mnt_drop_write_file(struct file *file) +{ + lockdep_off(); + mnt_drop_write_file(file); + lockdep_on(); +} +#endif + +/* ---------------------------------------------------------------------- */ + +struct au_hinode; +struct dentry *vfsub_lock_rename(struct dentry *d1, struct au_hinode *hdir1, + struct dentry *d2, struct au_hinode *hdir2); +void vfsub_unlock_rename(struct dentry *d1, struct au_hinode *hdir1, + struct dentry *d2, struct au_hinode *hdir2); + +int vfsub_create(struct inode *dir, struct path *path, int mode, + bool want_excl); +int vfsub_symlink(struct inode *dir, struct path *path, + const char *symname); +int vfsub_mknod(struct inode *dir, struct path *path, int mode, dev_t dev); +int vfsub_link(struct dentry *src_dentry, struct inode *dir, + struct path *path, struct inode **delegated_inode); +int vfsub_rename(struct inode *src_hdir, struct dentry *src_dentry, + struct inode *hdir, struct path *path, + struct inode **delegated_inode); +int vfsub_mkdir(struct inode *dir, struct path *path, int mode); +int vfsub_rmdir(struct inode *dir, struct path *path); + +/* ---------------------------------------------------------------------- */ + +ssize_t vfsub_read_u(struct file *file, char __user *ubuf, size_t count, + loff_t *ppos); +ssize_t vfsub_read_k(struct file *file, void *kbuf, size_t count, + loff_t *ppos); +ssize_t vfsub_write_u(struct file *file, const char __user *ubuf, size_t count, + loff_t *ppos); +ssize_t vfsub_write_k(struct file *file, void *kbuf, size_t count, + loff_t *ppos); +int vfsub_flush(struct file *file, fl_owner_t id); +int vfsub_iterate_dir(struct file *file, struct dir_context *ctx); + +static inline loff_t vfsub_f_size_read(struct file *file) +{ + return i_size_read(file_inode(file)); +} + +static inline unsigned int vfsub_file_flags(struct file *file) +{ + unsigned int flags; + + spin_lock(&file->f_lock); + flags = file->f_flags; + spin_unlock(&file->f_lock); + + return flags; +} + +#if 0 /* reserved */ +static inline void vfsub_file_accessed(struct file *h_file) +{ + file_accessed(h_file); + vfsub_update_h_iattr(&h_file->f_path, /*did*/NULL); /*ignore*/ +} +#endif + +#if 0 /* reserved */ +static inline void vfsub_touch_atime(struct vfsmount *h_mnt, + struct dentry *h_dentry) +{ + struct path h_path = { + .dentry = h_dentry, + .mnt = h_mnt + }; + touch_atime(&h_path); + vfsub_update_h_iattr(&h_path, /*did*/NULL); /*ignore*/ +} +#endif + +static inline int vfsub_update_time(struct inode *h_inode, struct timespec *ts, + int flags) +{ + return generic_update_time(h_inode, ts, flags); + /* no vfsub_update_h_iattr() since we don't have struct path */ +} + +#ifdef CONFIG_FS_POSIX_ACL +static inline int vfsub_acl_chmod(struct inode *h_inode, umode_t h_mode) +{ + int err; + + err = posix_acl_chmod(h_inode, h_mode); + if (err == -EOPNOTSUPP) + err = 0; + return err; +} +#else +AuStubInt0(vfsub_acl_chmod, struct inode *h_inode, umode_t h_mode); +#endif + +long vfsub_splice_to(struct file *in, loff_t *ppos, + struct pipe_inode_info *pipe, size_t len, + unsigned int flags); +long vfsub_splice_from(struct pipe_inode_info *pipe, struct file *out, + loff_t *ppos, size_t len, unsigned int flags); + +static inline long vfsub_truncate(struct path *path, loff_t length) +{ + long err; + + lockdep_off(); + err = vfs_truncate(path, length); + lockdep_on(); + return err; +} + +int vfsub_trunc(struct path *h_path, loff_t length, unsigned int attr, + struct file *h_file); +int vfsub_fsync(struct file *file, struct path *path, int datasync); + +/* ---------------------------------------------------------------------- */ + +static inline loff_t vfsub_llseek(struct file *file, loff_t offset, int origin) +{ + loff_t err; + + lockdep_off(); + err = vfs_llseek(file, offset, origin); + lockdep_on(); + return err; +} + +/* ---------------------------------------------------------------------- */ + +int vfsub_sio_mkdir(struct inode *dir, struct path *path, int mode); +int vfsub_sio_rmdir(struct inode *dir, struct path *path); +int vfsub_sio_notify_change(struct path *path, struct iattr *ia, + struct inode **delegated_inode); +int vfsub_notify_change(struct path *path, struct iattr *ia, + struct inode **delegated_inode); +int vfsub_unlink(struct inode *dir, struct path *path, + struct inode **delegated_inode, int force); + +/* ---------------------------------------------------------------------- */ + +static inline int vfsub_setxattr(struct dentry *dentry, const char *name, + const void *value, size_t size, int flags) +{ + int err; + + lockdep_off(); + err = vfs_setxattr(dentry, name, value, size, flags); + lockdep_on(); + + return err; +} + +static inline int vfsub_removexattr(struct dentry *dentry, const char *name) +{ + int err; + + lockdep_off(); + err = vfs_removexattr(dentry, name); + lockdep_on(); + + return err; +} + +#endif /* __KERNEL__ */ +#endif /* __AUFS_VFSUB_H__ */ diff --git a/fs/aufs/wbr_policy.c b/fs/aufs/wbr_policy.c new file mode 100644 index 000000000..994316302 --- /dev/null +++ b/fs/aufs/wbr_policy.c @@ -0,0 +1,752 @@ +/* + * Copyright (C) 2005-2016 Junjiro R. Okajima + */ + +/* + * policies for selecting one among multiple writable branches + */ + +#include <linux/statfs.h> +#include "aufs.h" + +/* subset of cpup_attr() */ +static noinline_for_stack +int au_cpdown_attr(struct path *h_path, struct dentry *h_src) +{ + int err, sbits; + struct iattr ia; + struct inode *h_isrc; + + h_isrc = d_inode(h_src); + ia.ia_valid = ATTR_FORCE | ATTR_MODE | ATTR_UID | ATTR_GID; + ia.ia_mode = h_isrc->i_mode; + ia.ia_uid = h_isrc->i_uid; + ia.ia_gid = h_isrc->i_gid; + sbits = !!(ia.ia_mode & (S_ISUID | S_ISGID)); + au_cpup_attr_flags(d_inode(h_path->dentry), h_isrc->i_flags); + /* no delegation since it is just created */ + err = vfsub_sio_notify_change(h_path, &ia, /*delegated*/NULL); + + /* is this nfs only? */ + if (!err && sbits && au_test_nfs(h_path->dentry->d_sb)) { + ia.ia_valid = ATTR_FORCE | ATTR_MODE; + ia.ia_mode = h_isrc->i_mode; + err = vfsub_sio_notify_change(h_path, &ia, /*delegated*/NULL); + } + + return err; +} + +#define AuCpdown_PARENT_OPQ 1 +#define AuCpdown_WHED (1 << 1) +#define AuCpdown_MADE_DIR (1 << 2) +#define AuCpdown_DIROPQ (1 << 3) +#define au_ftest_cpdown(flags, name) ((flags) & AuCpdown_##name) +#define au_fset_cpdown(flags, name) \ + do { (flags) |= AuCpdown_##name; } while (0) +#define au_fclr_cpdown(flags, name) \ + do { (flags) &= ~AuCpdown_##name; } while (0) + +static int au_cpdown_dir_opq(struct dentry *dentry, aufs_bindex_t bdst, + unsigned int *flags) +{ + int err; + struct dentry *opq_dentry; + + opq_dentry = au_diropq_create(dentry, bdst); + err = PTR_ERR(opq_dentry); + if (IS_ERR(opq_dentry)) + goto out; + dput(opq_dentry); + au_fset_cpdown(*flags, DIROPQ); + +out: + return err; +} + +static int au_cpdown_dir_wh(struct dentry *dentry, struct dentry *h_parent, + struct inode *dir, aufs_bindex_t bdst) +{ + int err; + struct path h_path; + struct au_branch *br; + + br = au_sbr(dentry->d_sb, bdst); + h_path.dentry = au_wh_lkup(h_parent, &dentry->d_name, br); + err = PTR_ERR(h_path.dentry); + if (IS_ERR(h_path.dentry)) + goto out; + + err = 0; + if (d_is_positive(h_path.dentry)) { + h_path.mnt = au_br_mnt(br); + err = au_wh_unlink_dentry(au_h_iptr(dir, bdst), &h_path, + dentry); + } + dput(h_path.dentry); + +out: + return err; +} + +static int au_cpdown_dir(struct dentry *dentry, aufs_bindex_t bdst, + struct au_pin *pin, + struct dentry *h_parent, void *arg) +{ + int err, rerr; + aufs_bindex_t bopq, bstart; + struct path h_path; + struct dentry *parent; + struct inode *h_dir, *h_inode, *inode, *dir; + unsigned int *flags = arg; + + bstart = au_dbstart(dentry); + /* dentry is di-locked */ + parent = dget_parent(dentry); + dir = d_inode(parent); + h_dir = d_inode(h_parent); + AuDebugOn(h_dir != au_h_iptr(dir, bdst)); + IMustLock(h_dir); + + err = au_lkup_neg(dentry, bdst, /*wh*/0); + if (unlikely(err < 0)) + goto out; + h_path.dentry = au_h_dptr(dentry, bdst); + h_path.mnt = au_sbr_mnt(dentry->d_sb, bdst); + err = vfsub_sio_mkdir(au_h_iptr(dir, bdst), &h_path, + S_IRWXU | S_IRUGO | S_IXUGO); + if (unlikely(err)) + goto out_put; + au_fset_cpdown(*flags, MADE_DIR); + + bopq = au_dbdiropq(dentry); + au_fclr_cpdown(*flags, WHED); + au_fclr_cpdown(*flags, DIROPQ); + if (au_dbwh(dentry) == bdst) + au_fset_cpdown(*flags, WHED); + if (!au_ftest_cpdown(*flags, PARENT_OPQ) && bopq <= bdst) + au_fset_cpdown(*flags, PARENT_OPQ); + h_inode = d_inode(h_path.dentry); + inode_lock_nested(h_inode, AuLsc_I_CHILD); + if (au_ftest_cpdown(*flags, WHED)) { + err = au_cpdown_dir_opq(dentry, bdst, flags); + if (unlikely(err)) { + inode_unlock(h_inode); + goto out_dir; + } + } + + err = au_cpdown_attr(&h_path, au_h_dptr(dentry, bstart)); + inode_unlock(h_inode); + if (unlikely(err)) + goto out_opq; + + if (au_ftest_cpdown(*flags, WHED)) { + err = au_cpdown_dir_wh(dentry, h_parent, dir, bdst); + if (unlikely(err)) + goto out_opq; + } + + inode = d_inode(dentry); + if (au_ibend(inode) < bdst) + au_set_ibend(inode, bdst); + au_set_h_iptr(inode, bdst, au_igrab(h_inode), + au_hi_flags(inode, /*isdir*/1)); + au_fhsm_wrote(dentry->d_sb, bdst, /*force*/0); + goto out; /* success */ + + /* revert */ +out_opq: + if (au_ftest_cpdown(*flags, DIROPQ)) { + inode_lock_nested(h_inode, AuLsc_I_CHILD); + rerr = au_diropq_remove(dentry, bdst); + inode_unlock(h_inode); + if (unlikely(rerr)) { + AuIOErr("failed removing diropq for %pd b%d (%d)\n", + dentry, bdst, rerr); + err = -EIO; + goto out; + } + } +out_dir: + if (au_ftest_cpdown(*flags, MADE_DIR)) { + rerr = vfsub_sio_rmdir(au_h_iptr(dir, bdst), &h_path); + if (unlikely(rerr)) { + AuIOErr("failed removing %pd b%d (%d)\n", + dentry, bdst, rerr); + err = -EIO; + } + } +out_put: + au_set_h_dptr(dentry, bdst, NULL); + if (au_dbend(dentry) == bdst) + au_update_dbend(dentry); +out: + dput(parent); + return err; +} + +int au_cpdown_dirs(struct dentry *dentry, aufs_bindex_t bdst) +{ + int err; + unsigned int flags; + + flags = 0; + err = au_cp_dirs(dentry, bdst, au_cpdown_dir, &flags); + + return err; +} + +/* ---------------------------------------------------------------------- */ + +/* policies for create */ + +int au_wbr_nonopq(struct dentry *dentry, aufs_bindex_t bindex) +{ + int err, i, j, ndentry; + aufs_bindex_t bopq; + struct au_dcsub_pages dpages; + struct au_dpage *dpage; + struct dentry **dentries, *parent, *d; + + err = au_dpages_init(&dpages, GFP_NOFS); + if (unlikely(err)) + goto out; + parent = dget_parent(dentry); + err = au_dcsub_pages_rev_aufs(&dpages, parent, /*do_include*/0); + if (unlikely(err)) + goto out_free; + + err = bindex; + for (i = 0; i < dpages.ndpage; i++) { + dpage = dpages.dpages + i; + dentries = dpage->dentries; + ndentry = dpage->ndentry; + for (j = 0; j < ndentry; j++) { + d = dentries[j]; + di_read_lock_parent2(d, !AuLock_IR); + bopq = au_dbdiropq(d); + di_read_unlock(d, !AuLock_IR); + if (bopq >= 0 && bopq < err) + err = bopq; + } + } + +out_free: + dput(parent); + au_dpages_free(&dpages); +out: + return err; +} + +static int au_wbr_bu(struct super_block *sb, aufs_bindex_t bindex) +{ + for (; bindex >= 0; bindex--) + if (!au_br_rdonly(au_sbr(sb, bindex))) + return bindex; + return -EROFS; +} + +/* top down parent */ +static int au_wbr_create_tdp(struct dentry *dentry, + unsigned int flags __maybe_unused) +{ + int err; + aufs_bindex_t bstart, bindex; + struct super_block *sb; + struct dentry *parent, *h_parent; + + sb = dentry->d_sb; + bstart = au_dbstart(dentry); + err = bstart; + if (!au_br_rdonly(au_sbr(sb, bstart))) + goto out; + + err = -EROFS; + parent = dget_parent(dentry); + for (bindex = au_dbstart(parent); bindex < bstart; bindex++) { + h_parent = au_h_dptr(parent, bindex); + if (!h_parent || d_is_negative(h_parent)) + continue; + + if (!au_br_rdonly(au_sbr(sb, bindex))) { + err = bindex; + break; + } + } + dput(parent); + + /* bottom up here */ + if (unlikely(err < 0)) { + err = au_wbr_bu(sb, bstart - 1); + if (err >= 0) + err = au_wbr_nonopq(dentry, err); + } + +out: + AuDbg("b%d\n", err); + return err; +} + +/* ---------------------------------------------------------------------- */ + +/* an exception for the policy other than tdp */ +static int au_wbr_create_exp(struct dentry *dentry) +{ + int err; + aufs_bindex_t bwh, bdiropq; + struct dentry *parent; + + err = -1; + bwh = au_dbwh(dentry); + parent = dget_parent(dentry); + bdiropq = au_dbdiropq(parent); + if (bwh >= 0) { + if (bdiropq >= 0) + err = min(bdiropq, bwh); + else + err = bwh; + AuDbg("%d\n", err); + } else if (bdiropq >= 0) { + err = bdiropq; + AuDbg("%d\n", err); + } + dput(parent); + + if (err >= 0) + err = au_wbr_nonopq(dentry, err); + + if (err >= 0 && au_br_rdonly(au_sbr(dentry->d_sb, err))) + err = -1; + + AuDbg("%d\n", err); + return err; +} + +/* ---------------------------------------------------------------------- */ + +/* round robin */ +static int au_wbr_create_init_rr(struct super_block *sb) +{ + int err; + + err = au_wbr_bu(sb, au_sbend(sb)); + atomic_set(&au_sbi(sb)->si_wbr_rr_next, -err); /* less important */ + /* smp_mb(); */ + + AuDbg("b%d\n", err); + return err; +} + +static int au_wbr_create_rr(struct dentry *dentry, unsigned int flags) +{ + int err, nbr; + unsigned int u; + aufs_bindex_t bindex, bend; + struct super_block *sb; + atomic_t *next; + + err = au_wbr_create_exp(dentry); + if (err >= 0) + goto out; + + sb = dentry->d_sb; + next = &au_sbi(sb)->si_wbr_rr_next; + bend = au_sbend(sb); + nbr = bend + 1; + for (bindex = 0; bindex <= bend; bindex++) { + if (!au_ftest_wbr(flags, DIR)) { + err = atomic_dec_return(next) + 1; + /* modulo for 0 is meaningless */ + if (unlikely(!err)) + err = atomic_dec_return(next) + 1; + } else + err = atomic_read(next); + AuDbg("%d\n", err); + u = err; + err = u % nbr; + AuDbg("%d\n", err); + if (!au_br_rdonly(au_sbr(sb, err))) + break; + err = -EROFS; + } + + if (err >= 0) + err = au_wbr_nonopq(dentry, err); + +out: + AuDbg("%d\n", err); + return err; +} + +/* ---------------------------------------------------------------------- */ + +/* most free space */ +static void au_mfs(struct dentry *dentry, struct dentry *parent) +{ + struct super_block *sb; + struct au_branch *br; + struct au_wbr_mfs *mfs; + struct dentry *h_parent; + aufs_bindex_t bindex, bend; + int err; + unsigned long long b, bavail; + struct path h_path; + /* reduce the stack usage */ + struct kstatfs *st; + + st = kmalloc(sizeof(*st), GFP_NOFS); + if (unlikely(!st)) { + AuWarn1("failed updating mfs(%d), ignored\n", -ENOMEM); + return; + } + + bavail = 0; + sb = dentry->d_sb; + mfs = &au_sbi(sb)->si_wbr_mfs; + MtxMustLock(&mfs->mfs_lock); + mfs->mfs_bindex = -EROFS; + mfs->mfsrr_bytes = 0; + if (!parent) { + bindex = 0; + bend = au_sbend(sb); + } else { + bindex = au_dbstart(parent); + bend = au_dbtaildir(parent); + } + + for (; bindex <= bend; bindex++) { + if (parent) { + h_parent = au_h_dptr(parent, bindex); + if (!h_parent || d_is_negative(h_parent)) + continue; + } + br = au_sbr(sb, bindex); + if (au_br_rdonly(br)) + continue; + + /* sb->s_root for NFS is unreliable */ + h_path.mnt = au_br_mnt(br); + h_path.dentry = h_path.mnt->mnt_root; + err = vfs_statfs(&h_path, st); + if (unlikely(err)) { + AuWarn1("failed statfs, b%d, %d\n", bindex, err); + continue; + } + + /* when the available size is equal, select the lower one */ + BUILD_BUG_ON(sizeof(b) < sizeof(st->f_bavail) + || sizeof(b) < sizeof(st->f_bsize)); + b = st->f_bavail * st->f_bsize; + br->br_wbr->wbr_bytes = b; + if (b >= bavail) { + bavail = b; + mfs->mfs_bindex = bindex; + mfs->mfs_jiffy = jiffies; + } + } + + mfs->mfsrr_bytes = bavail; + AuDbg("b%d\n", mfs->mfs_bindex); + kfree(st); +} + +static int au_wbr_create_mfs(struct dentry *dentry, unsigned int flags) +{ + int err; + struct dentry *parent; + struct super_block *sb; + struct au_wbr_mfs *mfs; + + err = au_wbr_create_exp(dentry); + if (err >= 0) + goto out; + + sb = dentry->d_sb; + parent = NULL; + if (au_ftest_wbr(flags, PARENT)) + parent = dget_parent(dentry); + mfs = &au_sbi(sb)->si_wbr_mfs; + mutex_lock(&mfs->mfs_lock); + if (time_after(jiffies, mfs->mfs_jiffy + mfs->mfs_expire) + || mfs->mfs_bindex < 0 + || au_br_rdonly(au_sbr(sb, mfs->mfs_bindex))) + au_mfs(dentry, parent); + mutex_unlock(&mfs->mfs_lock); + err = mfs->mfs_bindex; + dput(parent); + + if (err >= 0) + err = au_wbr_nonopq(dentry, err); + +out: + AuDbg("b%d\n", err); + return err; +} + +static int au_wbr_create_init_mfs(struct super_block *sb) +{ + struct au_wbr_mfs *mfs; + + mfs = &au_sbi(sb)->si_wbr_mfs; + mutex_init(&mfs->mfs_lock); + mfs->mfs_jiffy = 0; + mfs->mfs_bindex = -EROFS; + + return 0; +} + +static int au_wbr_create_fin_mfs(struct super_block *sb __maybe_unused) +{ + mutex_destroy(&au_sbi(sb)->si_wbr_mfs.mfs_lock); + return 0; +} + +/* ---------------------------------------------------------------------- */ + +/* most free space and then round robin */ +static int au_wbr_create_mfsrr(struct dentry *dentry, unsigned int flags) +{ + int err; + struct au_wbr_mfs *mfs; + + err = au_wbr_create_mfs(dentry, flags); + if (err >= 0) { + mfs = &au_sbi(dentry->d_sb)->si_wbr_mfs; + mutex_lock(&mfs->mfs_lock); + if (mfs->mfsrr_bytes < mfs->mfsrr_watermark) + err = au_wbr_create_rr(dentry, flags); + mutex_unlock(&mfs->mfs_lock); + } + + AuDbg("b%d\n", err); + return err; +} + +static int au_wbr_create_init_mfsrr(struct super_block *sb) +{ + int err; + + au_wbr_create_init_mfs(sb); /* ignore */ + err = au_wbr_create_init_rr(sb); + + return err; +} + +/* ---------------------------------------------------------------------- */ + +/* top down parent and most free space */ +static int au_wbr_create_pmfs(struct dentry *dentry, unsigned int flags) +{ + int err, e2; + unsigned long long b; + aufs_bindex_t bindex, bstart, bend; + struct super_block *sb; + struct dentry *parent, *h_parent; + struct au_branch *br; + + err = au_wbr_create_tdp(dentry, flags); + if (unlikely(err < 0)) + goto out; + parent = dget_parent(dentry); + bstart = au_dbstart(parent); + bend = au_dbtaildir(parent); + if (bstart == bend) + goto out_parent; /* success */ + + e2 = au_wbr_create_mfs(dentry, flags); + if (e2 < 0) + goto out_parent; /* success */ + + /* when the available size is equal, select upper one */ + sb = dentry->d_sb; + br = au_sbr(sb, err); + b = br->br_wbr->wbr_bytes; + AuDbg("b%d, %llu\n", err, b); + + for (bindex = bstart; bindex <= bend; bindex++) { + h_parent = au_h_dptr(parent, bindex); + if (!h_parent || d_is_negative(h_parent)) + continue; + + br = au_sbr(sb, bindex); + if (!au_br_rdonly(br) && br->br_wbr->wbr_bytes > b) { + b = br->br_wbr->wbr_bytes; + err = bindex; + AuDbg("b%d, %llu\n", err, b); + } + } + + if (err >= 0) + err = au_wbr_nonopq(dentry, err); + +out_parent: + dput(parent); +out: + AuDbg("b%d\n", err); + return err; +} + +/* ---------------------------------------------------------------------- */ + +/* + * - top down parent + * - most free space with parent + * - most free space round-robin regardless parent + */ +static int au_wbr_create_pmfsrr(struct dentry *dentry, unsigned int flags) +{ + int err; + unsigned long long watermark; + struct super_block *sb; + struct au_branch *br; + struct au_wbr_mfs *mfs; + + err = au_wbr_create_pmfs(dentry, flags | AuWbr_PARENT); + if (unlikely(err < 0)) + goto out; + + sb = dentry->d_sb; + br = au_sbr(sb, err); + mfs = &au_sbi(sb)->si_wbr_mfs; + mutex_lock(&mfs->mfs_lock); + watermark = mfs->mfsrr_watermark; + mutex_unlock(&mfs->mfs_lock); + if (br->br_wbr->wbr_bytes < watermark) + /* regardless the parent dir */ + err = au_wbr_create_mfsrr(dentry, flags); + +out: + AuDbg("b%d\n", err); + return err; +} + +/* ---------------------------------------------------------------------- */ + +/* policies for copyup */ + +/* top down parent */ +static int au_wbr_copyup_tdp(struct dentry *dentry) +{ + return au_wbr_create_tdp(dentry, /*flags, anything is ok*/0); +} + +/* bottom up parent */ +static int au_wbr_copyup_bup(struct dentry *dentry) +{ + int err; + aufs_bindex_t bindex, bstart; + struct dentry *parent, *h_parent; + struct super_block *sb; + + err = -EROFS; + sb = dentry->d_sb; + parent = dget_parent(dentry); + bstart = au_dbstart(parent); + for (bindex = au_dbstart(dentry); bindex >= bstart; bindex--) { + h_parent = au_h_dptr(parent, bindex); + if (!h_parent || d_is_negative(h_parent)) + continue; + + if (!au_br_rdonly(au_sbr(sb, bindex))) { + err = bindex; + break; + } + } + dput(parent); + + /* bottom up here */ + if (unlikely(err < 0)) + err = au_wbr_bu(sb, bstart - 1); + + AuDbg("b%d\n", err); + return err; +} + +/* bottom up */ +int au_wbr_do_copyup_bu(struct dentry *dentry, aufs_bindex_t bstart) +{ + int err; + + err = au_wbr_bu(dentry->d_sb, bstart); + AuDbg("b%d\n", err); + if (err > bstart) + err = au_wbr_nonopq(dentry, err); + + AuDbg("b%d\n", err); + return err; +} + +static int au_wbr_copyup_bu(struct dentry *dentry) +{ + int err; + aufs_bindex_t bstart; + + bstart = au_dbstart(dentry); + err = au_wbr_do_copyup_bu(dentry, bstart); + return err; +} + +/* ---------------------------------------------------------------------- */ + +struct au_wbr_copyup_operations au_wbr_copyup_ops[] = { + [AuWbrCopyup_TDP] = { + .copyup = au_wbr_copyup_tdp + }, + [AuWbrCopyup_BUP] = { + .copyup = au_wbr_copyup_bup + }, + [AuWbrCopyup_BU] = { + .copyup = au_wbr_copyup_bu + } +}; + +struct au_wbr_create_operations au_wbr_create_ops[] = { + [AuWbrCreate_TDP] = { + .create = au_wbr_create_tdp + }, + [AuWbrCreate_RR] = { + .create = au_wbr_create_rr, + .init = au_wbr_create_init_rr + }, + [AuWbrCreate_MFS] = { + .create = au_wbr_create_mfs, + .init = au_wbr_create_init_mfs, + .fin = au_wbr_create_fin_mfs + }, + [AuWbrCreate_MFSV] = { + .create = au_wbr_create_mfs, + .init = au_wbr_create_init_mfs, + .fin = au_wbr_create_fin_mfs + }, + [AuWbrCreate_MFSRR] = { + .create = au_wbr_create_mfsrr, + .init = au_wbr_create_init_mfsrr, + .fin = au_wbr_create_fin_mfs + }, + [AuWbrCreate_MFSRRV] = { + .create = au_wbr_create_mfsrr, + .init = au_wbr_create_init_mfsrr, + .fin = au_wbr_create_fin_mfs + }, + [AuWbrCreate_PMFS] = { + .create = au_wbr_create_pmfs, + .init = au_wbr_create_init_mfs, + .fin = au_wbr_create_fin_mfs + }, + [AuWbrCreate_PMFSV] = { + .create = au_wbr_create_pmfs, + .init = au_wbr_create_init_mfs, + .fin = au_wbr_create_fin_mfs + }, + [AuWbrCreate_PMFSRR] = { + .create = au_wbr_create_pmfsrr, + .init = au_wbr_create_init_mfsrr, + .fin = au_wbr_create_fin_mfs + }, + [AuWbrCreate_PMFSRRV] = { + .create = au_wbr_create_pmfsrr, + .init = au_wbr_create_init_mfsrr, + .fin = au_wbr_create_fin_mfs + } +}; diff --git a/fs/aufs/whout.c b/fs/aufs/whout.c new file mode 100644 index 000000000..fdda85eef --- /dev/null +++ b/fs/aufs/whout.c @@ -0,0 +1,1047 @@ +/* + * Copyright (C) 2005-2016 Junjiro R. Okajima + */ + +/* + * whiteout for logical deletion and opaque directory + */ + +#include "aufs.h" + +#define WH_MASK S_IRUGO + +/* + * If a directory contains this file, then it is opaque. We start with the + * .wh. flag so that it is blocked by lookup. + */ +static struct qstr diropq_name = QSTR_INIT(AUFS_WH_DIROPQ, + sizeof(AUFS_WH_DIROPQ) - 1); + +/* + * generate whiteout name, which is NOT terminated by NULL. + * @name: original d_name.name + * @len: original d_name.len + * @wh: whiteout qstr + * returns zero when succeeds, otherwise error. + * succeeded value as wh->name should be freed by kfree(). + */ +int au_wh_name_alloc(struct qstr *wh, const struct qstr *name) +{ + char *p; + + if (unlikely(name->len > PATH_MAX - AUFS_WH_PFX_LEN)) + return -ENAMETOOLONG; + + wh->len = name->len + AUFS_WH_PFX_LEN; + p = kmalloc(wh->len, GFP_NOFS); + wh->name = p; + if (p) { + memcpy(p, AUFS_WH_PFX, AUFS_WH_PFX_LEN); + memcpy(p + AUFS_WH_PFX_LEN, name->name, name->len); + /* smp_mb(); */ + return 0; + } + return -ENOMEM; +} + +/* ---------------------------------------------------------------------- */ + +/* + * test if the @wh_name exists under @h_parent. + * @try_sio specifies the necessary of super-io. + */ +int au_wh_test(struct dentry *h_parent, struct qstr *wh_name, int try_sio) +{ + int err; + struct dentry *wh_dentry; + + if (!try_sio) + wh_dentry = vfsub_lkup_one(wh_name, h_parent); + else + wh_dentry = au_sio_lkup_one(wh_name, h_parent); + err = PTR_ERR(wh_dentry); + if (IS_ERR(wh_dentry)) { + if (err == -ENAMETOOLONG) + err = 0; + goto out; + } + + err = 0; + if (d_is_negative(wh_dentry)) + goto out_wh; /* success */ + + err = 1; + if (d_is_reg(wh_dentry)) + goto out_wh; /* success */ + + err = -EIO; + AuIOErr("%pd Invalid whiteout entry type 0%o.\n", + wh_dentry, d_inode(wh_dentry)->i_mode); + +out_wh: + dput(wh_dentry); +out: + return err; +} + +/* + * test if the @h_dentry sets opaque or not. + */ +int au_diropq_test(struct dentry *h_dentry) +{ + int err; + struct inode *h_dir; + + h_dir = d_inode(h_dentry); + err = au_wh_test(h_dentry, &diropq_name, + au_test_h_perm_sio(h_dir, MAY_EXEC)); + return err; +} + +/* + * returns a negative dentry whose name is unique and temporary. + */ +struct dentry *au_whtmp_lkup(struct dentry *h_parent, struct au_branch *br, + struct qstr *prefix) +{ + struct dentry *dentry; + int i; + char defname[NAME_MAX - AUFS_MAX_NAMELEN + DNAME_INLINE_LEN + 1], + *name, *p; + /* strict atomic_t is unnecessary here */ + static unsigned short cnt; + struct qstr qs; + + BUILD_BUG_ON(sizeof(cnt) * 2 > AUFS_WH_TMP_LEN); + + name = defname; + qs.len = sizeof(defname) - DNAME_INLINE_LEN + prefix->len - 1; + if (unlikely(prefix->len > DNAME_INLINE_LEN)) { + dentry = ERR_PTR(-ENAMETOOLONG); + if (unlikely(qs.len > NAME_MAX)) + goto out; + dentry = ERR_PTR(-ENOMEM); + name = kmalloc(qs.len + 1, GFP_NOFS); + if (unlikely(!name)) + goto out; + } + + /* doubly whiteout-ed */ + memcpy(name, AUFS_WH_PFX AUFS_WH_PFX, AUFS_WH_PFX_LEN * 2); + p = name + AUFS_WH_PFX_LEN * 2; + memcpy(p, prefix->name, prefix->len); + p += prefix->len; + *p++ = '.'; + AuDebugOn(name + qs.len + 1 - p <= AUFS_WH_TMP_LEN); + + qs.name = name; + for (i = 0; i < 3; i++) { + sprintf(p, "%.*x", AUFS_WH_TMP_LEN, cnt++); + dentry = au_sio_lkup_one(&qs, h_parent); + if (IS_ERR(dentry) || d_is_negative(dentry)) + goto out_name; + dput(dentry); + } + /* pr_warn("could not get random name\n"); */ + dentry = ERR_PTR(-EEXIST); + AuDbg("%.*s\n", AuLNPair(&qs)); + BUG(); + +out_name: + if (name != defname) + kfree(name); +out: + AuTraceErrPtr(dentry); + return dentry; +} + +/* + * rename the @h_dentry on @br to the whiteouted temporary name. + */ +int au_whtmp_ren(struct dentry *h_dentry, struct au_branch *br) +{ + int err; + struct path h_path = { + .mnt = au_br_mnt(br) + }; + struct inode *h_dir, *delegated; + struct dentry *h_parent; + + h_parent = h_dentry->d_parent; /* dir inode is locked */ + h_dir = d_inode(h_parent); + IMustLock(h_dir); + + h_path.dentry = au_whtmp_lkup(h_parent, br, &h_dentry->d_name); + err = PTR_ERR(h_path.dentry); + if (IS_ERR(h_path.dentry)) + goto out; + + /* under the same dir, no need to lock_rename() */ + delegated = NULL; + err = vfsub_rename(h_dir, h_dentry, h_dir, &h_path, &delegated); + AuTraceErr(err); + if (unlikely(err == -EWOULDBLOCK)) { + pr_warn("cannot retry for NFSv4 delegation" + " for an internal rename\n"); + iput(delegated); + } + dput(h_path.dentry); + +out: + AuTraceErr(err); + return err; +} + +/* ---------------------------------------------------------------------- */ +/* + * functions for removing a whiteout + */ + +static int do_unlink_wh(struct inode *h_dir, struct path *h_path) +{ + int err, force; + struct inode *delegated; + + /* + * forces superio when the dir has a sticky bit. + * this may be a violation of unix fs semantics. + */ + force = (h_dir->i_mode & S_ISVTX) + && !uid_eq(current_fsuid(), d_inode(h_path->dentry)->i_uid); + delegated = NULL; + err = vfsub_unlink(h_dir, h_path, &delegated, force); + if (unlikely(err == -EWOULDBLOCK)) { + pr_warn("cannot retry for NFSv4 delegation" + " for an internal unlink\n"); + iput(delegated); + } + return err; +} + +int au_wh_unlink_dentry(struct inode *h_dir, struct path *h_path, + struct dentry *dentry) +{ + int err; + + err = do_unlink_wh(h_dir, h_path); + if (!err && dentry) + au_set_dbwh(dentry, -1); + + return err; +} + +static int unlink_wh_name(struct dentry *h_parent, struct qstr *wh, + struct au_branch *br) +{ + int err; + struct path h_path = { + .mnt = au_br_mnt(br) + }; + + err = 0; + h_path.dentry = vfsub_lkup_one(wh, h_parent); + if (IS_ERR(h_path.dentry)) + err = PTR_ERR(h_path.dentry); + else { + if (d_is_reg(h_path.dentry)) + err = do_unlink_wh(d_inode(h_parent), &h_path); + dput(h_path.dentry); + } + + return err; +} + +/* ---------------------------------------------------------------------- */ +/* + * initialize/clean whiteout for a branch + */ + +static void au_wh_clean(struct inode *h_dir, struct path *whpath, + const int isdir) +{ + int err; + struct inode *delegated; + + if (d_is_negative(whpath->dentry)) + return; + + if (isdir) + err = vfsub_rmdir(h_dir, whpath); + else { + delegated = NULL; + err = vfsub_unlink(h_dir, whpath, &delegated, /*force*/0); + if (unlikely(err == -EWOULDBLOCK)) { + pr_warn("cannot retry for NFSv4 delegation" + " for an internal unlink\n"); + iput(delegated); + } + } + if (unlikely(err)) + pr_warn("failed removing %pd (%d), ignored.\n", + whpath->dentry, err); +} + +static int test_linkable(struct dentry *h_root) +{ + struct inode *h_dir = d_inode(h_root); + + if (h_dir->i_op->link) + return 0; + + pr_err("%pd (%s) doesn't support link(2), use noplink and rw+nolwh\n", + h_root, au_sbtype(h_root->d_sb)); + return -ENOSYS; +} + +/* todo: should this mkdir be done in /sbin/mount.aufs helper? */ +static int au_whdir(struct inode *h_dir, struct path *path) +{ + int err; + + err = -EEXIST; + if (d_is_negative(path->dentry)) { + int mode = S_IRWXU; + + if (au_test_nfs(path->dentry->d_sb)) + mode |= S_IXUGO; + err = vfsub_mkdir(h_dir, path, mode); + } else if (d_is_dir(path->dentry)) + err = 0; + else + pr_err("unknown %pd exists\n", path->dentry); + + return err; +} + +struct au_wh_base { + const struct qstr *name; + struct dentry *dentry; +}; + +static void au_wh_init_ro(struct inode *h_dir, struct au_wh_base base[], + struct path *h_path) +{ + h_path->dentry = base[AuBrWh_BASE].dentry; + au_wh_clean(h_dir, h_path, /*isdir*/0); + h_path->dentry = base[AuBrWh_PLINK].dentry; + au_wh_clean(h_dir, h_path, /*isdir*/1); + h_path->dentry = base[AuBrWh_ORPH].dentry; + au_wh_clean(h_dir, h_path, /*isdir*/1); +} + +/* + * returns tri-state, + * minus: error, caller should print the message + * zero: succuess + * plus: error, caller should NOT print the message + */ +static int au_wh_init_rw_nolink(struct dentry *h_root, struct au_wbr *wbr, + int do_plink, struct au_wh_base base[], + struct path *h_path) +{ + int err; + struct inode *h_dir; + + h_dir = d_inode(h_root); + h_path->dentry = base[AuBrWh_BASE].dentry; + au_wh_clean(h_dir, h_path, /*isdir*/0); + h_path->dentry = base[AuBrWh_PLINK].dentry; + if (do_plink) { + err = test_linkable(h_root); + if (unlikely(err)) { + err = 1; + goto out; + } + + err = au_whdir(h_dir, h_path); + if (unlikely(err)) + goto out; + wbr->wbr_plink = dget(base[AuBrWh_PLINK].dentry); + } else + au_wh_clean(h_dir, h_path, /*isdir*/1); + h_path->dentry = base[AuBrWh_ORPH].dentry; + err = au_whdir(h_dir, h_path); + if (unlikely(err)) + goto out; + wbr->wbr_orph = dget(base[AuBrWh_ORPH].dentry); + +out: + return err; +} + +/* + * for the moment, aufs supports the branch filesystem which does not support + * link(2). testing on FAT which does not support i_op->setattr() fully either, + * copyup failed. finally, such filesystem will not be used as the writable + * branch. + * + * returns tri-state, see above. + */ +static int au_wh_init_rw(struct dentry *h_root, struct au_wbr *wbr, + int do_plink, struct au_wh_base base[], + struct path *h_path) +{ + int err; + struct inode *h_dir; + + WbrWhMustWriteLock(wbr); + + err = test_linkable(h_root); + if (unlikely(err)) { + err = 1; + goto out; + } + + /* + * todo: should this create be done in /sbin/mount.aufs helper? + */ + err = -EEXIST; + h_dir = d_inode(h_root); + if (d_is_negative(base[AuBrWh_BASE].dentry)) { + h_path->dentry = base[AuBrWh_BASE].dentry; + err = vfsub_create(h_dir, h_path, WH_MASK, /*want_excl*/true); + } else if (d_is_reg(base[AuBrWh_BASE].dentry)) + err = 0; + else + pr_err("unknown %pd2 exists\n", base[AuBrWh_BASE].dentry); + if (unlikely(err)) + goto out; + + h_path->dentry = base[AuBrWh_PLINK].dentry; + if (do_plink) { + err = au_whdir(h_dir, h_path); + if (unlikely(err)) + goto out; + wbr->wbr_plink = dget(base[AuBrWh_PLINK].dentry); + } else + au_wh_clean(h_dir, h_path, /*isdir*/1); + wbr->wbr_whbase = dget(base[AuBrWh_BASE].dentry); + + h_path->dentry = base[AuBrWh_ORPH].dentry; + err = au_whdir(h_dir, h_path); + if (unlikely(err)) + goto out; + wbr->wbr_orph = dget(base[AuBrWh_ORPH].dentry); + +out: + return err; +} + +/* + * initialize the whiteout base file/dir for @br. + */ +int au_wh_init(struct au_branch *br, struct super_block *sb) +{ + int err, i; + const unsigned char do_plink + = !!au_opt_test(au_mntflags(sb), PLINK); + struct inode *h_dir; + struct path path = br->br_path; + struct dentry *h_root = path.dentry; + struct au_wbr *wbr = br->br_wbr; + static const struct qstr base_name[] = { + [AuBrWh_BASE] = QSTR_INIT(AUFS_BASE_NAME, + sizeof(AUFS_BASE_NAME) - 1), + [AuBrWh_PLINK] = QSTR_INIT(AUFS_PLINKDIR_NAME, + sizeof(AUFS_PLINKDIR_NAME) - 1), + [AuBrWh_ORPH] = QSTR_INIT(AUFS_ORPHDIR_NAME, + sizeof(AUFS_ORPHDIR_NAME) - 1) + }; + struct au_wh_base base[] = { + [AuBrWh_BASE] = { + .name = base_name + AuBrWh_BASE, + .dentry = NULL + }, + [AuBrWh_PLINK] = { + .name = base_name + AuBrWh_PLINK, + .dentry = NULL + }, + [AuBrWh_ORPH] = { + .name = base_name + AuBrWh_ORPH, + .dentry = NULL + } + }; + + if (wbr) + WbrWhMustWriteLock(wbr); + + for (i = 0; i < AuBrWh_Last; i++) { + /* doubly whiteouted */ + struct dentry *d; + + d = au_wh_lkup(h_root, (void *)base[i].name, br); + err = PTR_ERR(d); + if (IS_ERR(d)) + goto out; + + base[i].dentry = d; + AuDebugOn(wbr + && wbr->wbr_wh[i] + && wbr->wbr_wh[i] != base[i].dentry); + } + + if (wbr) + for (i = 0; i < AuBrWh_Last; i++) { + dput(wbr->wbr_wh[i]); + wbr->wbr_wh[i] = NULL; + } + + err = 0; + if (!au_br_writable(br->br_perm)) { + h_dir = d_inode(h_root); + au_wh_init_ro(h_dir, base, &path); + } else if (!au_br_wh_linkable(br->br_perm)) { + err = au_wh_init_rw_nolink(h_root, wbr, do_plink, base, &path); + if (err > 0) + goto out; + else if (err) + goto out_err; + } else { + err = au_wh_init_rw(h_root, wbr, do_plink, base, &path); + if (err > 0) + goto out; + else if (err) + goto out_err; + } + goto out; /* success */ + +out_err: + pr_err("an error(%d) on the writable branch %pd(%s)\n", + err, h_root, au_sbtype(h_root->d_sb)); +out: + for (i = 0; i < AuBrWh_Last; i++) + dput(base[i].dentry); + return err; +} + +/* ---------------------------------------------------------------------- */ +/* + * whiteouts are all hard-linked usually. + * when its link count reaches a ceiling, we create a new whiteout base + * asynchronously. + */ + +struct reinit_br_wh { + struct super_block *sb; + struct au_branch *br; +}; + +static void reinit_br_wh(void *arg) +{ + int err; + aufs_bindex_t bindex; + struct path h_path; + struct reinit_br_wh *a = arg; + struct au_wbr *wbr; + struct inode *dir, *delegated; + struct dentry *h_root; + struct au_hinode *hdir; + + err = 0; + wbr = a->br->br_wbr; + /* big aufs lock */ + si_noflush_write_lock(a->sb); + if (!au_br_writable(a->br->br_perm)) + goto out; + bindex = au_br_index(a->sb, a->br->br_id); + if (unlikely(bindex < 0)) + goto out; + + di_read_lock_parent(a->sb->s_root, AuLock_IR); + dir = d_inode(a->sb->s_root); + hdir = au_hi(dir, bindex); + h_root = au_h_dptr(a->sb->s_root, bindex); + AuDebugOn(h_root != au_br_dentry(a->br)); + + au_hn_imtx_lock_nested(hdir, AuLsc_I_PARENT); + wbr_wh_write_lock(wbr); + err = au_h_verify(wbr->wbr_whbase, au_opt_udba(a->sb), hdir->hi_inode, + h_root, a->br); + if (!err) { + h_path.dentry = wbr->wbr_whbase; + h_path.mnt = au_br_mnt(a->br); + delegated = NULL; + err = vfsub_unlink(hdir->hi_inode, &h_path, &delegated, + /*force*/0); + if (unlikely(err == -EWOULDBLOCK)) { + pr_warn("cannot retry for NFSv4 delegation" + " for an internal unlink\n"); + iput(delegated); + } + } else { + pr_warn("%pd is moved, ignored\n", wbr->wbr_whbase); + err = 0; + } + dput(wbr->wbr_whbase); + wbr->wbr_whbase = NULL; + if (!err) + err = au_wh_init(a->br, a->sb); + wbr_wh_write_unlock(wbr); + au_hn_imtx_unlock(hdir); + di_read_unlock(a->sb->s_root, AuLock_IR); + if (!err) + au_fhsm_wrote(a->sb, bindex, /*force*/0); + +out: + if (wbr) + atomic_dec(&wbr->wbr_wh_running); + atomic_dec(&a->br->br_count); + si_write_unlock(a->sb); + au_nwt_done(&au_sbi(a->sb)->si_nowait); + kfree(arg); + if (unlikely(err)) + AuIOErr("err %d\n", err); +} + +static void kick_reinit_br_wh(struct super_block *sb, struct au_branch *br) +{ + int do_dec, wkq_err; + struct reinit_br_wh *arg; + + do_dec = 1; + if (atomic_inc_return(&br->br_wbr->wbr_wh_running) != 1) + goto out; + + /* ignore ENOMEM */ + arg = kmalloc(sizeof(*arg), GFP_NOFS); + if (arg) { + /* + * dec(wh_running), kfree(arg) and dec(br_count) + * in reinit function + */ + arg->sb = sb; + arg->br = br; + atomic_inc(&br->br_count); + wkq_err = au_wkq_nowait(reinit_br_wh, arg, sb, /*flags*/0); + if (unlikely(wkq_err)) { + atomic_dec(&br->br_wbr->wbr_wh_running); + atomic_dec(&br->br_count); + kfree(arg); + } + do_dec = 0; + } + +out: + if (do_dec) + atomic_dec(&br->br_wbr->wbr_wh_running); +} + +/* ---------------------------------------------------------------------- */ + +/* + * create the whiteout @wh. + */ +static int link_or_create_wh(struct super_block *sb, aufs_bindex_t bindex, + struct dentry *wh) +{ + int err; + struct path h_path = { + .dentry = wh + }; + struct au_branch *br; + struct au_wbr *wbr; + struct dentry *h_parent; + struct inode *h_dir, *delegated; + + h_parent = wh->d_parent; /* dir inode is locked */ + h_dir = d_inode(h_parent); + IMustLock(h_dir); + + br = au_sbr(sb, bindex); + h_path.mnt = au_br_mnt(br); + wbr = br->br_wbr; + wbr_wh_read_lock(wbr); + if (wbr->wbr_whbase) { + delegated = NULL; + err = vfsub_link(wbr->wbr_whbase, h_dir, &h_path, &delegated); + if (unlikely(err == -EWOULDBLOCK)) { + pr_warn("cannot retry for NFSv4 delegation" + " for an internal link\n"); + iput(delegated); + } + if (!err || err != -EMLINK) + goto out; + + /* link count full. re-initialize br_whbase. */ + kick_reinit_br_wh(sb, br); + } + + /* return this error in this context */ + err = vfsub_create(h_dir, &h_path, WH_MASK, /*want_excl*/true); + if (!err) + au_fhsm_wrote(sb, bindex, /*force*/0); + +out: + wbr_wh_read_unlock(wbr); + return err; +} + +/* ---------------------------------------------------------------------- */ + +/* + * create or remove the diropq. + */ +static struct dentry *do_diropq(struct dentry *dentry, aufs_bindex_t bindex, + unsigned int flags) +{ + struct dentry *opq_dentry, *h_dentry; + struct super_block *sb; + struct au_branch *br; + int err; + + sb = dentry->d_sb; + br = au_sbr(sb, bindex); + h_dentry = au_h_dptr(dentry, bindex); + opq_dentry = vfsub_lkup_one(&diropq_name, h_dentry); + if (IS_ERR(opq_dentry)) + goto out; + + if (au_ftest_diropq(flags, CREATE)) { + err = link_or_create_wh(sb, bindex, opq_dentry); + if (!err) { + au_set_dbdiropq(dentry, bindex); + goto out; /* success */ + } + } else { + struct path tmp = { + .dentry = opq_dentry, + .mnt = au_br_mnt(br) + }; + err = do_unlink_wh(au_h_iptr(d_inode(dentry), bindex), &tmp); + if (!err) + au_set_dbdiropq(dentry, -1); + } + dput(opq_dentry); + opq_dentry = ERR_PTR(err); + +out: + return opq_dentry; +} + +struct do_diropq_args { + struct dentry **errp; + struct dentry *dentry; + aufs_bindex_t bindex; + unsigned int flags; +}; + +static void call_do_diropq(void *args) +{ + struct do_diropq_args *a = args; + *a->errp = do_diropq(a->dentry, a->bindex, a->flags); +} + +struct dentry *au_diropq_sio(struct dentry *dentry, aufs_bindex_t bindex, + unsigned int flags) +{ + struct dentry *diropq, *h_dentry; + + h_dentry = au_h_dptr(dentry, bindex); + if (!au_test_h_perm_sio(d_inode(h_dentry), MAY_EXEC | MAY_WRITE)) + diropq = do_diropq(dentry, bindex, flags); + else { + int wkq_err; + struct do_diropq_args args = { + .errp = &diropq, + .dentry = dentry, + .bindex = bindex, + .flags = flags + }; + + wkq_err = au_wkq_wait(call_do_diropq, &args); + if (unlikely(wkq_err)) + diropq = ERR_PTR(wkq_err); + } + + return diropq; +} + +/* ---------------------------------------------------------------------- */ + +/* + * lookup whiteout dentry. + * @h_parent: lower parent dentry which must exist and be locked + * @base_name: name of dentry which will be whiteouted + * returns dentry for whiteout. + */ +struct dentry *au_wh_lkup(struct dentry *h_parent, struct qstr *base_name, + struct au_branch *br) +{ + int err; + struct qstr wh_name; + struct dentry *wh_dentry; + + err = au_wh_name_alloc(&wh_name, base_name); + wh_dentry = ERR_PTR(err); + if (!err) { + wh_dentry = vfsub_lkup_one(&wh_name, h_parent); + kfree(wh_name.name); + } + return wh_dentry; +} + +/* + * link/create a whiteout for @dentry on @bindex. + */ +struct dentry *au_wh_create(struct dentry *dentry, aufs_bindex_t bindex, + struct dentry *h_parent) +{ + struct dentry *wh_dentry; + struct super_block *sb; + int err; + + sb = dentry->d_sb; + wh_dentry = au_wh_lkup(h_parent, &dentry->d_name, au_sbr(sb, bindex)); + if (!IS_ERR(wh_dentry) && d_is_negative(wh_dentry)) { + err = link_or_create_wh(sb, bindex, wh_dentry); + if (!err) { + au_set_dbwh(dentry, bindex); + au_fhsm_wrote(sb, bindex, /*force*/0); + } else { + dput(wh_dentry); + wh_dentry = ERR_PTR(err); + } + } + + return wh_dentry; +} + +/* ---------------------------------------------------------------------- */ + +/* Delete all whiteouts in this directory on branch bindex. */ +static int del_wh_children(struct dentry *h_dentry, struct au_nhash *whlist, + aufs_bindex_t bindex, struct au_branch *br) +{ + int err; + unsigned long ul, n; + struct qstr wh_name; + char *p; + struct hlist_head *head; + struct au_vdir_wh *pos; + struct au_vdir_destr *str; + + err = -ENOMEM; + p = (void *)__get_free_page(GFP_NOFS); + wh_name.name = p; + if (unlikely(!wh_name.name)) + goto out; + + err = 0; + memcpy(p, AUFS_WH_PFX, AUFS_WH_PFX_LEN); + p += AUFS_WH_PFX_LEN; + n = whlist->nh_num; + head = whlist->nh_head; + for (ul = 0; !err && ul < n; ul++, head++) { + hlist_for_each_entry(pos, head, wh_hash) { + if (pos->wh_bindex != bindex) + continue; + + str = &pos->wh_str; + if (str->len + AUFS_WH_PFX_LEN <= PATH_MAX) { + memcpy(p, str->name, str->len); + wh_name.len = AUFS_WH_PFX_LEN + str->len; + err = unlink_wh_name(h_dentry, &wh_name, br); + if (!err) + continue; + break; + } + AuIOErr("whiteout name too long %.*s\n", + str->len, str->name); + err = -EIO; + break; + } + } + free_page((unsigned long)wh_name.name); + +out: + return err; +} + +struct del_wh_children_args { + int *errp; + struct dentry *h_dentry; + struct au_nhash *whlist; + aufs_bindex_t bindex; + struct au_branch *br; +}; + +static void call_del_wh_children(void *args) +{ + struct del_wh_children_args *a = args; + *a->errp = del_wh_children(a->h_dentry, a->whlist, a->bindex, a->br); +} + +/* ---------------------------------------------------------------------- */ + +struct au_whtmp_rmdir *au_whtmp_rmdir_alloc(struct super_block *sb, gfp_t gfp) +{ + struct au_whtmp_rmdir *whtmp; + int err; + unsigned int rdhash; + + SiMustAnyLock(sb); + + whtmp = kzalloc(sizeof(*whtmp), gfp); + if (unlikely(!whtmp)) { + whtmp = ERR_PTR(-ENOMEM); + goto out; + } + + /* no estimation for dir size */ + rdhash = au_sbi(sb)->si_rdhash; + if (!rdhash) + rdhash = AUFS_RDHASH_DEF; + err = au_nhash_alloc(&whtmp->whlist, rdhash, gfp); + if (unlikely(err)) { + kfree(whtmp); + whtmp = ERR_PTR(err); + } + +out: + return whtmp; +} + +void au_whtmp_rmdir_free(struct au_whtmp_rmdir *whtmp) +{ + if (whtmp->br) + atomic_dec(&whtmp->br->br_count); + dput(whtmp->wh_dentry); + iput(whtmp->dir); + au_nhash_wh_free(&whtmp->whlist); + kfree(whtmp); +} + +/* + * rmdir the whiteouted temporary named dir @h_dentry. + * @whlist: whiteouted children. + */ +int au_whtmp_rmdir(struct inode *dir, aufs_bindex_t bindex, + struct dentry *wh_dentry, struct au_nhash *whlist) +{ + int err; + unsigned int h_nlink; + struct path h_tmp; + struct inode *wh_inode, *h_dir; + struct au_branch *br; + + h_dir = d_inode(wh_dentry->d_parent); /* dir inode is locked */ + IMustLock(h_dir); + + br = au_sbr(dir->i_sb, bindex); + wh_inode = d_inode(wh_dentry); + inode_lock_nested(wh_inode, AuLsc_I_CHILD); + + /* + * someone else might change some whiteouts while we were sleeping. + * it means this whlist may have an obsoleted entry. + */ + if (!au_test_h_perm_sio(wh_inode, MAY_EXEC | MAY_WRITE)) + err = del_wh_children(wh_dentry, whlist, bindex, br); + else { + int wkq_err; + struct del_wh_children_args args = { + .errp = &err, + .h_dentry = wh_dentry, + .whlist = whlist, + .bindex = bindex, + .br = br + }; + + wkq_err = au_wkq_wait(call_del_wh_children, &args); + if (unlikely(wkq_err)) + err = wkq_err; + } + inode_unlock(wh_inode); + + if (!err) { + h_tmp.dentry = wh_dentry; + h_tmp.mnt = au_br_mnt(br); + h_nlink = h_dir->i_nlink; + err = vfsub_rmdir(h_dir, &h_tmp); + /* some fs doesn't change the parent nlink in some cases */ + h_nlink -= h_dir->i_nlink; + } + + if (!err) { + if (au_ibstart(dir) == bindex) { + /* todo: dir->i_mutex is necessary */ + au_cpup_attr_timesizes(dir); + if (h_nlink) + vfsub_drop_nlink(dir); + } + return 0; /* success */ + } + + pr_warn("failed removing %pd(%d), ignored\n", wh_dentry, err); + return err; +} + +static void call_rmdir_whtmp(void *args) +{ + int err; + aufs_bindex_t bindex; + struct au_whtmp_rmdir *a = args; + struct super_block *sb; + struct dentry *h_parent; + struct inode *h_dir; + struct au_hinode *hdir; + + /* rmdir by nfsd may cause deadlock with this i_mutex */ + /* inode_lock(a->dir); */ + err = -EROFS; + sb = a->dir->i_sb; + si_read_lock(sb, !AuLock_FLUSH); + if (!au_br_writable(a->br->br_perm)) + goto out; + bindex = au_br_index(sb, a->br->br_id); + if (unlikely(bindex < 0)) + goto out; + + err = -EIO; + ii_write_lock_parent(a->dir); + h_parent = dget_parent(a->wh_dentry); + h_dir = d_inode(h_parent); + hdir = au_hi(a->dir, bindex); + err = vfsub_mnt_want_write(au_br_mnt(a->br)); + if (unlikely(err)) + goto out_mnt; + au_hn_imtx_lock_nested(hdir, AuLsc_I_PARENT); + err = au_h_verify(a->wh_dentry, au_opt_udba(sb), h_dir, h_parent, + a->br); + if (!err) + err = au_whtmp_rmdir(a->dir, bindex, a->wh_dentry, &a->whlist); + au_hn_imtx_unlock(hdir); + vfsub_mnt_drop_write(au_br_mnt(a->br)); + +out_mnt: + dput(h_parent); + ii_write_unlock(a->dir); +out: + /* inode_unlock(a->dir); */ + au_whtmp_rmdir_free(a); + si_read_unlock(sb); + au_nwt_done(&au_sbi(sb)->si_nowait); + if (unlikely(err)) + AuIOErr("err %d\n", err); +} + +void au_whtmp_kick_rmdir(struct inode *dir, aufs_bindex_t bindex, + struct dentry *wh_dentry, struct au_whtmp_rmdir *args) +{ + int wkq_err; + struct super_block *sb; + + IMustLock(dir); + + /* all post-process will be done in do_rmdir_whtmp(). */ + sb = dir->i_sb; + args->dir = au_igrab(dir); + args->br = au_sbr(sb, bindex); + atomic_inc(&args->br->br_count); + args->wh_dentry = dget(wh_dentry); + wkq_err = au_wkq_nowait(call_rmdir_whtmp, args, sb, /*flags*/0); + if (unlikely(wkq_err)) { + pr_warn("rmdir error %pd (%d), ignored\n", wh_dentry, wkq_err); + au_whtmp_rmdir_free(args); + } +} diff --git a/fs/aufs/whout.h b/fs/aufs/whout.h new file mode 100644 index 000000000..4077dd19e --- /dev/null +++ b/fs/aufs/whout.h @@ -0,0 +1,72 @@ +/* + * Copyright (C) 2005-2016 Junjiro R. Okajima + */ + +/* + * whiteout for logical deletion and opaque directory + */ + +#ifndef __AUFS_WHOUT_H__ +#define __AUFS_WHOUT_H__ + +#ifdef __KERNEL__ + +#include "dir.h" + +/* whout.c */ +int au_wh_name_alloc(struct qstr *wh, const struct qstr *name); +int au_wh_test(struct dentry *h_parent, struct qstr *wh_name, int try_sio); +int au_diropq_test(struct dentry *h_dentry); +struct au_branch; +struct dentry *au_whtmp_lkup(struct dentry *h_parent, struct au_branch *br, + struct qstr *prefix); +int au_whtmp_ren(struct dentry *h_dentry, struct au_branch *br); +int au_wh_unlink_dentry(struct inode *h_dir, struct path *h_path, + struct dentry *dentry); +int au_wh_init(struct au_branch *br, struct super_block *sb); + +/* diropq flags */ +#define AuDiropq_CREATE 1 +#define au_ftest_diropq(flags, name) ((flags) & AuDiropq_##name) +#define au_fset_diropq(flags, name) \ + do { (flags) |= AuDiropq_##name; } while (0) +#define au_fclr_diropq(flags, name) \ + do { (flags) &= ~AuDiropq_##name; } while (0) + +struct dentry *au_diropq_sio(struct dentry *dentry, aufs_bindex_t bindex, + unsigned int flags); +struct dentry *au_wh_lkup(struct dentry *h_parent, struct qstr *base_name, + struct au_branch *br); +struct dentry *au_wh_create(struct dentry *dentry, aufs_bindex_t bindex, + struct dentry *h_parent); + +/* real rmdir for the whiteout-ed dir */ +struct au_whtmp_rmdir { + struct inode *dir; + struct au_branch *br; + struct dentry *wh_dentry; + struct au_nhash whlist; +}; + +struct au_whtmp_rmdir *au_whtmp_rmdir_alloc(struct super_block *sb, gfp_t gfp); +void au_whtmp_rmdir_free(struct au_whtmp_rmdir *whtmp); +int au_whtmp_rmdir(struct inode *dir, aufs_bindex_t bindex, + struct dentry *wh_dentry, struct au_nhash *whlist); +void au_whtmp_kick_rmdir(struct inode *dir, aufs_bindex_t bindex, + struct dentry *wh_dentry, struct au_whtmp_rmdir *args); + +/* ---------------------------------------------------------------------- */ + +static inline struct dentry *au_diropq_create(struct dentry *dentry, + aufs_bindex_t bindex) +{ + return au_diropq_sio(dentry, bindex, AuDiropq_CREATE); +} + +static inline int au_diropq_remove(struct dentry *dentry, aufs_bindex_t bindex) +{ + return PTR_ERR(au_diropq_sio(dentry, bindex, !AuDiropq_CREATE)); +} + +#endif /* __KERNEL__ */ +#endif /* __AUFS_WHOUT_H__ */ diff --git a/fs/aufs/wkq.c b/fs/aufs/wkq.c new file mode 100644 index 000000000..0f1500e93 --- /dev/null +++ b/fs/aufs/wkq.c @@ -0,0 +1,200 @@ +/* + * Copyright (C) 2005-2016 Junjiro R. Okajima + */ + +/* + * workqueue for asynchronous/super-io operations + * todo: try new dredential scheme + */ + +#include <linux/module.h> +#include "aufs.h" + +/* internal workqueue named AUFS_WKQ_NAME */ + +static struct workqueue_struct *au_wkq; + +struct au_wkinfo { + struct work_struct wk; + struct kobject *kobj; + + unsigned int flags; /* see wkq.h */ + + au_wkq_func_t func; + void *args; + + struct completion *comp; +}; + +/* ---------------------------------------------------------------------- */ + +static void wkq_func(struct work_struct *wk) +{ + struct au_wkinfo *wkinfo = container_of(wk, struct au_wkinfo, wk); + + AuDebugOn(!uid_eq(current_fsuid(), GLOBAL_ROOT_UID)); + AuDebugOn(rlimit(RLIMIT_FSIZE) != RLIM_INFINITY); + + wkinfo->func(wkinfo->args); + if (au_ftest_wkq(wkinfo->flags, WAIT)) + complete(wkinfo->comp); + else { + kobject_put(wkinfo->kobj); + module_put(THIS_MODULE); /* todo: ?? */ + kfree(wkinfo); + } +} + +/* + * Since struct completion is large, try allocating it dynamically. + */ +#if 1 /* defined(CONFIG_4KSTACKS) || defined(AuTest4KSTACKS) */ +#define AuWkqCompDeclare(name) struct completion *comp = NULL + +static int au_wkq_comp_alloc(struct au_wkinfo *wkinfo, struct completion **comp) +{ + *comp = kmalloc(sizeof(**comp), GFP_NOFS); + if (*comp) { + init_completion(*comp); + wkinfo->comp = *comp; + return 0; + } + return -ENOMEM; +} + +static void au_wkq_comp_free(struct completion *comp) +{ + kfree(comp); +} + +#else + +/* no braces */ +#define AuWkqCompDeclare(name) \ + DECLARE_COMPLETION_ONSTACK(_ ## name); \ + struct completion *comp = &_ ## name + +static int au_wkq_comp_alloc(struct au_wkinfo *wkinfo, struct completion **comp) +{ + wkinfo->comp = *comp; + return 0; +} + +static void au_wkq_comp_free(struct completion *comp __maybe_unused) +{ + /* empty */ +} +#endif /* 4KSTACKS */ + +static void au_wkq_run(struct au_wkinfo *wkinfo) +{ + if (au_ftest_wkq(wkinfo->flags, NEST)) { + if (au_wkq_test()) { + AuWarn1("wkq from wkq, unless silly-rename on NFS," + " due to a dead dir by UDBA?\n"); + AuDebugOn(au_ftest_wkq(wkinfo->flags, WAIT)); + } + } else + au_dbg_verify_kthread(); + + if (au_ftest_wkq(wkinfo->flags, WAIT)) { + INIT_WORK_ONSTACK(&wkinfo->wk, wkq_func); + queue_work(au_wkq, &wkinfo->wk); + } else { + INIT_WORK(&wkinfo->wk, wkq_func); + schedule_work(&wkinfo->wk); + } +} + +/* + * Be careful. It is easy to make deadlock happen. + * processA: lock, wkq and wait + * processB: wkq and wait, lock in wkq + * --> deadlock + */ +int au_wkq_do_wait(unsigned int flags, au_wkq_func_t func, void *args) +{ + int err; + AuWkqCompDeclare(comp); + struct au_wkinfo wkinfo = { + .flags = flags, + .func = func, + .args = args + }; + + err = au_wkq_comp_alloc(&wkinfo, &comp); + if (!err) { + au_wkq_run(&wkinfo); + /* no timeout, no interrupt */ + wait_for_completion(wkinfo.comp); + au_wkq_comp_free(comp); + destroy_work_on_stack(&wkinfo.wk); + } + + return err; + +} + +/* + * Note: dget/dput() in func for aufs dentries are not supported. It will be a + * problem in a concurrent umounting. + */ +int au_wkq_nowait(au_wkq_func_t func, void *args, struct super_block *sb, + unsigned int flags) +{ + int err; + struct au_wkinfo *wkinfo; + + atomic_inc(&au_sbi(sb)->si_nowait.nw_len); + + /* + * wkq_func() must free this wkinfo. + * it highly depends upon the implementation of workqueue. + */ + err = 0; + wkinfo = kmalloc(sizeof(*wkinfo), GFP_NOFS); + if (wkinfo) { + wkinfo->kobj = &au_sbi(sb)->si_kobj; + wkinfo->flags = flags & ~AuWkq_WAIT; + wkinfo->func = func; + wkinfo->args = args; + wkinfo->comp = NULL; + kobject_get(wkinfo->kobj); + __module_get(THIS_MODULE); /* todo: ?? */ + + au_wkq_run(wkinfo); + } else { + err = -ENOMEM; + au_nwt_done(&au_sbi(sb)->si_nowait); + } + + return err; +} + +/* ---------------------------------------------------------------------- */ + +void au_nwt_init(struct au_nowait_tasks *nwt) +{ + atomic_set(&nwt->nw_len, 0); + /* smp_mb(); */ /* atomic_set */ + init_waitqueue_head(&nwt->nw_wq); +} + +void au_wkq_fin(void) +{ + destroy_workqueue(au_wkq); +} + +int __init au_wkq_init(void) +{ + int err; + + err = 0; + au_wkq = alloc_workqueue(AUFS_WKQ_NAME, 0, WQ_DFL_ACTIVE); + if (IS_ERR(au_wkq)) + err = PTR_ERR(au_wkq); + else if (!au_wkq) + err = -ENOMEM; + + return err; +} diff --git a/fs/aufs/wkq.h b/fs/aufs/wkq.h new file mode 100644 index 000000000..f6c9b9902 --- /dev/null +++ b/fs/aufs/wkq.h @@ -0,0 +1,78 @@ +/* + * Copyright (C) 2005-2016 Junjiro R. Okajima + */ + +/* + * workqueue for asynchronous/super-io operations + * todo: try new credentials management scheme + */ + +#ifndef __AUFS_WKQ_H__ +#define __AUFS_WKQ_H__ + +#ifdef __KERNEL__ + +struct super_block; + +/* ---------------------------------------------------------------------- */ + +/* + * in the next operation, wait for the 'nowait' tasks in system-wide workqueue + */ +struct au_nowait_tasks { + atomic_t nw_len; + wait_queue_head_t nw_wq; +}; + +/* ---------------------------------------------------------------------- */ + +typedef void (*au_wkq_func_t)(void *args); + +/* wkq flags */ +#define AuWkq_WAIT 1 +#define AuWkq_NEST (1 << 1) +#define au_ftest_wkq(flags, name) ((flags) & AuWkq_##name) +#define au_fset_wkq(flags, name) \ + do { (flags) |= AuWkq_##name; } while (0) +#define au_fclr_wkq(flags, name) \ + do { (flags) &= ~AuWkq_##name; } while (0) + +#ifndef CONFIG_AUFS_HNOTIFY +#undef AuWkq_NEST +#define AuWkq_NEST 0 +#endif + +/* wkq.c */ +int au_wkq_do_wait(unsigned int flags, au_wkq_func_t func, void *args); +int au_wkq_nowait(au_wkq_func_t func, void *args, struct super_block *sb, + unsigned int flags); +void au_nwt_init(struct au_nowait_tasks *nwt); +int __init au_wkq_init(void); +void au_wkq_fin(void); + +/* ---------------------------------------------------------------------- */ + +static inline int au_wkq_test(void) +{ + return current->flags & PF_WQ_WORKER; +} + +static inline int au_wkq_wait(au_wkq_func_t func, void *args) +{ + return au_wkq_do_wait(AuWkq_WAIT, func, args); +} + +static inline void au_nwt_done(struct au_nowait_tasks *nwt) +{ + if (atomic_dec_and_test(&nwt->nw_len)) + wake_up_all(&nwt->nw_wq); +} + +static inline int au_nwt_flush(struct au_nowait_tasks *nwt) +{ + wait_event(nwt->nw_wq, !atomic_read(&nwt->nw_len)); + return 0; +} + +#endif /* __KERNEL__ */ +#endif /* __AUFS_WKQ_H__ */ diff --git a/fs/aufs/xattr.c b/fs/aufs/xattr.c new file mode 100644 index 000000000..890c10795 --- /dev/null +++ b/fs/aufs/xattr.c @@ -0,0 +1,331 @@ +/* + * Copyright (C) 2014-2016 Junjiro R. Okajima + */ + +/* + * handling xattr functions + */ + +#include <linux/xattr.h> +#include "aufs.h" + +static int au_xattr_ignore(int err, char *name, unsigned int ignore_flags) +{ + if (!ignore_flags) + goto out; + switch (err) { + case -ENOMEM: + case -EDQUOT: + goto out; + } + + if ((ignore_flags & AuBrAttr_ICEX) == AuBrAttr_ICEX) { + err = 0; + goto out; + } + +#define cmp(brattr, prefix) do { \ + if (!strncmp(name, XATTR_##prefix##_PREFIX, \ + XATTR_##prefix##_PREFIX_LEN)) { \ + if (ignore_flags & AuBrAttr_ICEX_##brattr) \ + err = 0; \ + goto out; \ + } \ + } while (0) + + cmp(SEC, SECURITY); + cmp(SYS, SYSTEM); + cmp(TR, TRUSTED); + cmp(USR, USER); +#undef cmp + + if (ignore_flags & AuBrAttr_ICEX_OTH) + err = 0; + +out: + return err; +} + +static const int au_xattr_out_of_list = AuBrAttr_ICEX_OTH << 1; + +static int au_do_cpup_xattr(struct dentry *h_dst, struct dentry *h_src, + char *name, char **buf, unsigned int ignore_flags, + unsigned int verbose) +{ + int err; + ssize_t ssz; + struct inode *h_idst; + + ssz = vfs_getxattr_alloc(h_src, name, buf, 0, GFP_NOFS); + err = ssz; + if (unlikely(err <= 0)) { + if (err == -ENODATA + || (err == -EOPNOTSUPP + && ((ignore_flags & au_xattr_out_of_list) + || (au_test_nfs_noacl(d_inode(h_src)) + && (!strcmp(name, XATTR_NAME_POSIX_ACL_ACCESS) + || !strcmp(name, + XATTR_NAME_POSIX_ACL_DEFAULT)))) + )) + err = 0; + if (err && (verbose || au_debug_test())) + pr_err("%s, err %d\n", name, err); + goto out; + } + + /* unlock it temporary */ + h_idst = d_inode(h_dst); + inode_unlock(h_idst); + err = vfsub_setxattr(h_dst, name, *buf, ssz, /*flags*/0); + inode_lock_nested(h_idst, AuLsc_I_CHILD2); + if (unlikely(err)) { + if (verbose || au_debug_test()) + pr_err("%s, err %d\n", name, err); + err = au_xattr_ignore(err, name, ignore_flags); + } + +out: + return err; +} + +int au_cpup_xattr(struct dentry *h_dst, struct dentry *h_src, int ignore_flags, + unsigned int verbose) +{ + int err, unlocked, acl_access, acl_default; + ssize_t ssz; + struct inode *h_isrc, *h_idst; + char *value, *p, *o, *e; + + /* try stopping to update the source inode while we are referencing */ + /* there should not be the parent-child relationship between them */ + h_isrc = d_inode(h_src); + h_idst = d_inode(h_dst); + inode_unlock(h_idst); + inode_lock_nested(h_isrc, AuLsc_I_CHILD); + inode_lock_nested(h_idst, AuLsc_I_CHILD2); + unlocked = 0; + + /* some filesystems don't list POSIX ACL, for example tmpfs */ + ssz = vfs_listxattr(h_src, NULL, 0); + err = ssz; + if (unlikely(err < 0)) { + AuTraceErr(err); + if (err == -ENODATA + || err == -EOPNOTSUPP) + err = 0; /* ignore */ + goto out; + } + + err = 0; + p = NULL; + o = NULL; + if (ssz) { + err = -ENOMEM; + p = kmalloc(ssz, GFP_NOFS); + o = p; + if (unlikely(!p)) + goto out; + err = vfs_listxattr(h_src, p, ssz); + } + inode_unlock(h_isrc); + unlocked = 1; + AuDbg("err %d, ssz %zd\n", err, ssz); + if (unlikely(err < 0)) + goto out_free; + + err = 0; + e = p + ssz; + value = NULL; + acl_access = 0; + acl_default = 0; + while (!err && p < e) { + acl_access |= !strncmp(p, XATTR_NAME_POSIX_ACL_ACCESS, + sizeof(XATTR_NAME_POSIX_ACL_ACCESS) - 1); + acl_default |= !strncmp(p, XATTR_NAME_POSIX_ACL_DEFAULT, + sizeof(XATTR_NAME_POSIX_ACL_DEFAULT) + - 1); + err = au_do_cpup_xattr(h_dst, h_src, p, &value, ignore_flags, + verbose); + p += strlen(p) + 1; + } + AuTraceErr(err); + ignore_flags |= au_xattr_out_of_list; + if (!err && !acl_access) { + err = au_do_cpup_xattr(h_dst, h_src, + XATTR_NAME_POSIX_ACL_ACCESS, &value, + ignore_flags, verbose); + AuTraceErr(err); + } + if (!err && !acl_default) { + err = au_do_cpup_xattr(h_dst, h_src, + XATTR_NAME_POSIX_ACL_DEFAULT, &value, + ignore_flags, verbose); + AuTraceErr(err); + } + + kfree(value); + +out_free: + kfree(o); +out: + if (!unlocked) + inode_unlock(h_isrc); + AuTraceErr(err); + return err; +} + +/* ---------------------------------------------------------------------- */ + +enum { + AU_XATTR_LIST, + AU_XATTR_GET +}; + +struct au_lgxattr { + int type; + union { + struct { + char *list; + size_t size; + } list; + struct { + const char *name; + void *value; + size_t size; + } get; + } u; +}; + +static ssize_t au_lgxattr(struct dentry *dentry, struct au_lgxattr *arg) +{ + ssize_t err; + struct path h_path; + struct super_block *sb; + + sb = dentry->d_sb; + err = si_read_lock(sb, AuLock_FLUSH | AuLock_NOPLM); + if (unlikely(err)) + goto out; + err = au_h_path_getattr(dentry, /*force*/1, &h_path); + if (unlikely(err)) + goto out_si; + if (unlikely(!h_path.dentry)) + /* illegally overlapped or something */ + goto out_di; /* pretending success */ + + /* always topmost entry only */ + switch (arg->type) { + case AU_XATTR_LIST: + err = vfs_listxattr(h_path.dentry, + arg->u.list.list, arg->u.list.size); + break; + case AU_XATTR_GET: + err = vfs_getxattr(h_path.dentry, + arg->u.get.name, arg->u.get.value, + arg->u.get.size); + break; + } + +out_di: + di_read_unlock(dentry, AuLock_IR); +out_si: + si_read_unlock(sb); +out: + AuTraceErr(err); + return err; +} + +ssize_t aufs_listxattr(struct dentry *dentry, char *list, size_t size) +{ + struct au_lgxattr arg = { + .type = AU_XATTR_LIST, + .u.list = { + .list = list, + .size = size + }, + }; + + return au_lgxattr(dentry, &arg); +} + +ssize_t aufs_getxattr(struct dentry *dentry, const char *name, void *value, + size_t size) +{ + struct au_lgxattr arg = { + .type = AU_XATTR_GET, + .u.get = { + .name = name, + .value = value, + .size = size + }, + }; + + return au_lgxattr(dentry, &arg); +} + +int aufs_setxattr(struct dentry *dentry, const char *name, const void *value, + size_t size, int flags) +{ + struct au_srxattr arg = { + .type = AU_XATTR_SET, + .u.set = { + .name = name, + .value = value, + .size = size, + .flags = flags + }, + }; + + return au_srxattr(dentry, &arg); +} + +int aufs_removexattr(struct dentry *dentry, const char *name) +{ + struct au_srxattr arg = { + .type = AU_XATTR_REMOVE, + .u.remove = { + .name = name + }, + }; + + return au_srxattr(dentry, &arg); +} + +/* ---------------------------------------------------------------------- */ + +#if 0 +static size_t au_xattr_list(struct dentry *dentry, char *list, size_t list_size, + const char *name, size_t name_len, int type) +{ + return aufs_listxattr(dentry, list, list_size); +} + +static int au_xattr_get(struct dentry *dentry, const char *name, void *buffer, + size_t size, int type) +{ + return aufs_getxattr(dentry, name, buffer, size); +} + +static int au_xattr_set(struct dentry *dentry, const char *name, + const void *value, size_t size, int flags, int type) +{ + return aufs_setxattr(dentry, name, value, size, flags); +} + +static const struct xattr_handler au_xattr_handler = { + /* no prefix, no flags */ + .list = au_xattr_list, + .get = au_xattr_get, + .set = au_xattr_set + /* why no remove? */ +}; + +static const struct xattr_handler *au_xattr_handlers[] = { + &au_xattr_handler +}; + +void au_xattr_init(struct super_block *sb) +{ + /* sb->s_xattr = au_xattr_handlers; */ +} +#endif diff --git a/fs/aufs/xino.c b/fs/aufs/xino.c new file mode 100644 index 000000000..2773caedc --- /dev/null +++ b/fs/aufs/xino.c @@ -0,0 +1,1305 @@ +/* + * Copyright (C) 2005-2016 Junjiro R. Okajima + */ + +/* + * external inode number translation table and bitmap + */ + +#include <linux/seq_file.h> +#include <linux/statfs.h> +#include "aufs.h" + +/* todo: unnecessary to support mmap_sem since kernel-space? */ +ssize_t xino_fread(vfs_readf_t func, struct file *file, void *kbuf, size_t size, + loff_t *pos) +{ + ssize_t err; + mm_segment_t oldfs; + union { + void *k; + char __user *u; + } buf; + + buf.k = kbuf; + oldfs = get_fs(); + set_fs(KERNEL_DS); + do { + /* todo: signal_pending? */ + err = func(file, buf.u, size, pos); + } while (err == -EAGAIN || err == -EINTR); + set_fs(oldfs); + +#if 0 /* reserved for future use */ + if (err > 0) + fsnotify_access(file->f_path.dentry); +#endif + + return err; +} + +/* ---------------------------------------------------------------------- */ + +static ssize_t xino_fwrite_wkq(vfs_writef_t func, struct file *file, void *buf, + size_t size, loff_t *pos); + +static ssize_t do_xino_fwrite(vfs_writef_t func, struct file *file, void *kbuf, + size_t size, loff_t *pos) +{ + ssize_t err; + mm_segment_t oldfs; + union { + void *k; + const char __user *u; + } buf; + int i; + const int prevent_endless = 10; + + i = 0; + buf.k = kbuf; + oldfs = get_fs(); + set_fs(KERNEL_DS); + do { + err = func(file, buf.u, size, pos); + if (err == -EINTR + && !au_wkq_test() + && fatal_signal_pending(current)) { + set_fs(oldfs); + err = xino_fwrite_wkq(func, file, kbuf, size, pos); + BUG_ON(err == -EINTR); + oldfs = get_fs(); + set_fs(KERNEL_DS); + } + } while (i++ < prevent_endless + && (err == -EAGAIN || err == -EINTR)); + set_fs(oldfs); + +#if 0 /* reserved for future use */ + if (err > 0) + fsnotify_modify(file->f_path.dentry); +#endif + + return err; +} + +struct do_xino_fwrite_args { + ssize_t *errp; + vfs_writef_t func; + struct file *file; + void *buf; + size_t size; + loff_t *pos; +}; + +static void call_do_xino_fwrite(void *args) +{ + struct do_xino_fwrite_args *a = args; + *a->errp = do_xino_fwrite(a->func, a->file, a->buf, a->size, a->pos); +} + +static ssize_t xino_fwrite_wkq(vfs_writef_t func, struct file *file, void *buf, + size_t size, loff_t *pos) +{ + ssize_t err; + int wkq_err; + struct do_xino_fwrite_args args = { + .errp = &err, + .func = func, + .file = file, + .buf = buf, + .size = size, + .pos = pos + }; + + /* + * it breaks RLIMIT_FSIZE and normal user's limit, + * users should care about quota and real 'filesystem full.' + */ + wkq_err = au_wkq_wait(call_do_xino_fwrite, &args); + if (unlikely(wkq_err)) + err = wkq_err; + + return err; +} + +ssize_t xino_fwrite(vfs_writef_t func, struct file *file, void *buf, + size_t size, loff_t *pos) +{ + ssize_t err; + + if (rlimit(RLIMIT_FSIZE) == RLIM_INFINITY) { + lockdep_off(); + err = do_xino_fwrite(func, file, buf, size, pos); + lockdep_on(); + } else + err = xino_fwrite_wkq(func, file, buf, size, pos); + + return err; +} + +/* ---------------------------------------------------------------------- */ + +/* + * create a new xinofile at the same place/path as @base_file. + */ +struct file *au_xino_create2(struct file *base_file, struct file *copy_src) +{ + struct file *file; + struct dentry *base, *parent; + struct inode *dir, *delegated; + struct qstr *name; + struct path path; + int err; + + base = base_file->f_path.dentry; + parent = base->d_parent; /* dir inode is locked */ + dir = d_inode(parent); + IMustLock(dir); + + file = ERR_PTR(-EINVAL); + name = &base->d_name; + path.dentry = vfsub_lookup_one_len(name->name, parent, name->len); + if (IS_ERR(path.dentry)) { + file = (void *)path.dentry; + pr_err("%pd lookup err %ld\n", + base, PTR_ERR(path.dentry)); + goto out; + } + + /* no need to mnt_want_write() since we call dentry_open() later */ + err = vfs_create(dir, path.dentry, S_IRUGO | S_IWUGO, NULL); + if (unlikely(err)) { + file = ERR_PTR(err); + pr_err("%pd create err %d\n", base, err); + goto out_dput; + } + + path.mnt = base_file->f_path.mnt; + file = vfsub_dentry_open(&path, + O_RDWR | O_CREAT | O_EXCL | O_LARGEFILE + /* | __FMODE_NONOTIFY */); + if (IS_ERR(file)) { + pr_err("%pd open err %ld\n", base, PTR_ERR(file)); + goto out_dput; + } + + delegated = NULL; + err = vfsub_unlink(dir, &file->f_path, &delegated, /*force*/0); + if (unlikely(err == -EWOULDBLOCK)) { + pr_warn("cannot retry for NFSv4 delegation" + " for an internal unlink\n"); + iput(delegated); + } + if (unlikely(err)) { + pr_err("%pd unlink err %d\n", base, err); + goto out_fput; + } + + if (copy_src) { + /* no one can touch copy_src xino */ + err = au_copy_file(file, copy_src, vfsub_f_size_read(copy_src)); + if (unlikely(err)) { + pr_err("%pd copy err %d\n", base, err); + goto out_fput; + } + } + goto out_dput; /* success */ + +out_fput: + fput(file); + file = ERR_PTR(err); +out_dput: + dput(path.dentry); +out: + return file; +} + +struct au_xino_lock_dir { + struct au_hinode *hdir; + struct dentry *parent; + struct inode *dir; +}; + +static void au_xino_lock_dir(struct super_block *sb, struct file *xino, + struct au_xino_lock_dir *ldir) +{ + aufs_bindex_t brid, bindex; + + ldir->hdir = NULL; + bindex = -1; + brid = au_xino_brid(sb); + if (brid >= 0) + bindex = au_br_index(sb, brid); + if (bindex >= 0) { + ldir->hdir = au_hi(d_inode(sb->s_root), bindex); + au_hn_imtx_lock_nested(ldir->hdir, AuLsc_I_PARENT); + } else { + ldir->parent = dget_parent(xino->f_path.dentry); + ldir->dir = d_inode(ldir->parent); + inode_lock_nested(ldir->dir, AuLsc_I_PARENT); + } +} + +static void au_xino_unlock_dir(struct au_xino_lock_dir *ldir) +{ + if (ldir->hdir) + au_hn_imtx_unlock(ldir->hdir); + else { + inode_unlock(ldir->dir); + dput(ldir->parent); + } +} + +/* ---------------------------------------------------------------------- */ + +/* trucate xino files asynchronously */ + +int au_xino_trunc(struct super_block *sb, aufs_bindex_t bindex) +{ + int err; + unsigned long jiffy; + blkcnt_t blocks; + aufs_bindex_t bi, bend; + struct kstatfs *st; + struct au_branch *br; + struct file *new_xino, *file; + struct super_block *h_sb; + struct au_xino_lock_dir ldir; + + err = -ENOMEM; + st = kmalloc(sizeof(*st), GFP_NOFS); + if (unlikely(!st)) + goto out; + + err = -EINVAL; + bend = au_sbend(sb); + if (unlikely(bindex < 0 || bend < bindex)) + goto out_st; + br = au_sbr(sb, bindex); + file = br->br_xino.xi_file; + if (!file) + goto out_st; + + err = vfs_statfs(&file->f_path, st); + if (unlikely(err)) + AuErr1("statfs err %d, ignored\n", err); + jiffy = jiffies; + blocks = file_inode(file)->i_blocks; + pr_info("begin truncating xino(b%d), ib%llu, %llu/%llu free blks\n", + bindex, (u64)blocks, st->f_bfree, st->f_blocks); + + au_xino_lock_dir(sb, file, &ldir); + /* mnt_want_write() is unnecessary here */ + new_xino = au_xino_create2(file, file); + au_xino_unlock_dir(&ldir); + err = PTR_ERR(new_xino); + if (IS_ERR(new_xino)) { + pr_err("err %d, ignored\n", err); + goto out_st; + } + err = 0; + fput(file); + br->br_xino.xi_file = new_xino; + + h_sb = au_br_sb(br); + for (bi = 0; bi <= bend; bi++) { + if (unlikely(bi == bindex)) + continue; + br = au_sbr(sb, bi); + if (au_br_sb(br) != h_sb) + continue; + + fput(br->br_xino.xi_file); + br->br_xino.xi_file = new_xino; + get_file(new_xino); + } + + err = vfs_statfs(&new_xino->f_path, st); + if (!err) { + pr_info("end truncating xino(b%d), ib%llu, %llu/%llu free blks\n", + bindex, (u64)file_inode(new_xino)->i_blocks, + st->f_bfree, st->f_blocks); + if (file_inode(new_xino)->i_blocks < blocks) + au_sbi(sb)->si_xino_jiffy = jiffy; + } else + AuErr1("statfs err %d, ignored\n", err); + +out_st: + kfree(st); +out: + return err; +} + +struct xino_do_trunc_args { + struct super_block *sb; + struct au_branch *br; +}; + +static void xino_do_trunc(void *_args) +{ + struct xino_do_trunc_args *args = _args; + struct super_block *sb; + struct au_branch *br; + struct inode *dir; + int err; + aufs_bindex_t bindex; + + err = 0; + sb = args->sb; + dir = d_inode(sb->s_root); + br = args->br; + + si_noflush_write_lock(sb); + ii_read_lock_parent(dir); + bindex = au_br_index(sb, br->br_id); + err = au_xino_trunc(sb, bindex); + ii_read_unlock(dir); + if (unlikely(err)) + pr_warn("err b%d, (%d)\n", bindex, err); + atomic_dec(&br->br_xino_running); + atomic_dec(&br->br_count); + si_write_unlock(sb); + au_nwt_done(&au_sbi(sb)->si_nowait); + kfree(args); +} + +static int xino_trunc_test(struct super_block *sb, struct au_branch *br) +{ + int err; + struct kstatfs st; + struct au_sbinfo *sbinfo; + + /* todo: si_xino_expire and the ratio should be customizable */ + sbinfo = au_sbi(sb); + if (time_before(jiffies, + sbinfo->si_xino_jiffy + sbinfo->si_xino_expire)) + return 0; + + /* truncation border */ + err = vfs_statfs(&br->br_xino.xi_file->f_path, &st); + if (unlikely(err)) { + AuErr1("statfs err %d, ignored\n", err); + return 0; + } + if (div64_u64(st.f_bfree * 100, st.f_blocks) >= AUFS_XINO_DEF_TRUNC) + return 0; + + return 1; +} + +static void xino_try_trunc(struct super_block *sb, struct au_branch *br) +{ + struct xino_do_trunc_args *args; + int wkq_err; + + if (!xino_trunc_test(sb, br)) + return; + + if (atomic_inc_return(&br->br_xino_running) > 1) + goto out; + + /* lock and kfree() will be called in trunc_xino() */ + args = kmalloc(sizeof(*args), GFP_NOFS); + if (unlikely(!args)) { + AuErr1("no memory\n"); + goto out_args; + } + + atomic_inc(&br->br_count); + args->sb = sb; + args->br = br; + wkq_err = au_wkq_nowait(xino_do_trunc, args, sb, /*flags*/0); + if (!wkq_err) + return; /* success */ + + pr_err("wkq %d\n", wkq_err); + atomic_dec(&br->br_count); + +out_args: + kfree(args); +out: + atomic_dec(&br->br_xino_running); +} + +/* ---------------------------------------------------------------------- */ + +static int au_xino_do_write(vfs_writef_t write, struct file *file, + ino_t h_ino, ino_t ino) +{ + loff_t pos; + ssize_t sz; + + pos = h_ino; + if (unlikely(au_loff_max / sizeof(ino) - 1 < pos)) { + AuIOErr1("too large hi%lu\n", (unsigned long)h_ino); + return -EFBIG; + } + pos *= sizeof(ino); + sz = xino_fwrite(write, file, &ino, sizeof(ino), &pos); + if (sz == sizeof(ino)) + return 0; /* success */ + + AuIOErr("write failed (%zd)\n", sz); + return -EIO; +} + +/* + * write @ino to the xinofile for the specified branch{@sb, @bindex} + * at the position of @h_ino. + * even if @ino is zero, it is written to the xinofile and means no entry. + * if the size of the xino file on a specific filesystem exceeds the watermark, + * try truncating it. + */ +int au_xino_write(struct super_block *sb, aufs_bindex_t bindex, ino_t h_ino, + ino_t ino) +{ + int err; + unsigned int mnt_flags; + struct au_branch *br; + + BUILD_BUG_ON(sizeof(long long) != sizeof(au_loff_max) + || ((loff_t)-1) > 0); + SiMustAnyLock(sb); + + mnt_flags = au_mntflags(sb); + if (!au_opt_test(mnt_flags, XINO)) + return 0; + + br = au_sbr(sb, bindex); + err = au_xino_do_write(au_sbi(sb)->si_xwrite, br->br_xino.xi_file, + h_ino, ino); + if (!err) { + if (au_opt_test(mnt_flags, TRUNC_XINO) + && au_test_fs_trunc_xino(au_br_sb(br))) + xino_try_trunc(sb, br); + return 0; /* success */ + } + + AuIOErr("write failed (%d)\n", err); + return -EIO; +} + +/* ---------------------------------------------------------------------- */ + +/* aufs inode number bitmap */ + +static const int page_bits = (int)PAGE_SIZE * BITS_PER_BYTE; +static ino_t xib_calc_ino(unsigned long pindex, int bit) +{ + ino_t ino; + + AuDebugOn(bit < 0 || page_bits <= bit); + ino = AUFS_FIRST_INO + pindex * page_bits + bit; + return ino; +} + +static void xib_calc_bit(ino_t ino, unsigned long *pindex, int *bit) +{ + AuDebugOn(ino < AUFS_FIRST_INO); + ino -= AUFS_FIRST_INO; + *pindex = ino / page_bits; + *bit = ino % page_bits; +} + +static int xib_pindex(struct super_block *sb, unsigned long pindex) +{ + int err; + loff_t pos; + ssize_t sz; + struct au_sbinfo *sbinfo; + struct file *xib; + unsigned long *p; + + sbinfo = au_sbi(sb); + MtxMustLock(&sbinfo->si_xib_mtx); + AuDebugOn(pindex > ULONG_MAX / PAGE_SIZE + || !au_opt_test(sbinfo->si_mntflags, XINO)); + + if (pindex == sbinfo->si_xib_last_pindex) + return 0; + + xib = sbinfo->si_xib; + p = sbinfo->si_xib_buf; + pos = sbinfo->si_xib_last_pindex; + pos *= PAGE_SIZE; + sz = xino_fwrite(sbinfo->si_xwrite, xib, p, PAGE_SIZE, &pos); + if (unlikely(sz != PAGE_SIZE)) + goto out; + + pos = pindex; + pos *= PAGE_SIZE; + if (vfsub_f_size_read(xib) >= pos + PAGE_SIZE) + sz = xino_fread(sbinfo->si_xread, xib, p, PAGE_SIZE, &pos); + else { + memset(p, 0, PAGE_SIZE); + sz = xino_fwrite(sbinfo->si_xwrite, xib, p, PAGE_SIZE, &pos); + } + if (sz == PAGE_SIZE) { + sbinfo->si_xib_last_pindex = pindex; + return 0; /* success */ + } + +out: + AuIOErr1("write failed (%zd)\n", sz); + err = sz; + if (sz >= 0) + err = -EIO; + return err; +} + +/* ---------------------------------------------------------------------- */ + +static void au_xib_clear_bit(struct inode *inode) +{ + int err, bit; + unsigned long pindex; + struct super_block *sb; + struct au_sbinfo *sbinfo; + + AuDebugOn(inode->i_nlink); + + sb = inode->i_sb; + xib_calc_bit(inode->i_ino, &pindex, &bit); + AuDebugOn(page_bits <= bit); + sbinfo = au_sbi(sb); + mutex_lock(&sbinfo->si_xib_mtx); + err = xib_pindex(sb, pindex); + if (!err) { + clear_bit(bit, sbinfo->si_xib_buf); + sbinfo->si_xib_next_bit = bit; + } + mutex_unlock(&sbinfo->si_xib_mtx); +} + +/* for s_op->delete_inode() */ +void au_xino_delete_inode(struct inode *inode, const int unlinked) +{ + int err; + unsigned int mnt_flags; + aufs_bindex_t bindex, bend, bi; + unsigned char try_trunc; + struct au_iinfo *iinfo; + struct super_block *sb; + struct au_hinode *hi; + struct inode *h_inode; + struct au_branch *br; + vfs_writef_t xwrite; + + sb = inode->i_sb; + mnt_flags = au_mntflags(sb); + if (!au_opt_test(mnt_flags, XINO) + || inode->i_ino == AUFS_ROOT_INO) + return; + + if (unlinked) { + au_xigen_inc(inode); + au_xib_clear_bit(inode); + } + + iinfo = au_ii(inode); + if (!iinfo) + return; + + bindex = iinfo->ii_bstart; + if (bindex < 0) + return; + + xwrite = au_sbi(sb)->si_xwrite; + try_trunc = !!au_opt_test(mnt_flags, TRUNC_XINO); + hi = iinfo->ii_hinode + bindex; + bend = iinfo->ii_bend; + for (; bindex <= bend; bindex++, hi++) { + h_inode = hi->hi_inode; + if (!h_inode + || (!unlinked && h_inode->i_nlink)) + continue; + + /* inode may not be revalidated */ + bi = au_br_index(sb, hi->hi_id); + if (bi < 0) + continue; + + br = au_sbr(sb, bi); + err = au_xino_do_write(xwrite, br->br_xino.xi_file, + h_inode->i_ino, /*ino*/0); + if (!err && try_trunc + && au_test_fs_trunc_xino(au_br_sb(br))) + xino_try_trunc(sb, br); + } +} + +/* get an unused inode number from bitmap */ +ino_t au_xino_new_ino(struct super_block *sb) +{ + ino_t ino; + unsigned long *p, pindex, ul, pend; + struct au_sbinfo *sbinfo; + struct file *file; + int free_bit, err; + + if (!au_opt_test(au_mntflags(sb), XINO)) + return iunique(sb, AUFS_FIRST_INO); + + sbinfo = au_sbi(sb); + mutex_lock(&sbinfo->si_xib_mtx); + p = sbinfo->si_xib_buf; + free_bit = sbinfo->si_xib_next_bit; + if (free_bit < page_bits && !test_bit(free_bit, p)) + goto out; /* success */ + free_bit = find_first_zero_bit(p, page_bits); + if (free_bit < page_bits) + goto out; /* success */ + + pindex = sbinfo->si_xib_last_pindex; + for (ul = pindex - 1; ul < ULONG_MAX; ul--) { + err = xib_pindex(sb, ul); + if (unlikely(err)) + goto out_err; + free_bit = find_first_zero_bit(p, page_bits); + if (free_bit < page_bits) + goto out; /* success */ + } + + file = sbinfo->si_xib; + pend = vfsub_f_size_read(file) / PAGE_SIZE; + for (ul = pindex + 1; ul <= pend; ul++) { + err = xib_pindex(sb, ul); + if (unlikely(err)) + goto out_err; + free_bit = find_first_zero_bit(p, page_bits); + if (free_bit < page_bits) + goto out; /* success */ + } + BUG(); + +out: + set_bit(free_bit, p); + sbinfo->si_xib_next_bit = free_bit + 1; + pindex = sbinfo->si_xib_last_pindex; + mutex_unlock(&sbinfo->si_xib_mtx); + ino = xib_calc_ino(pindex, free_bit); + AuDbg("i%lu\n", (unsigned long)ino); + return ino; +out_err: + mutex_unlock(&sbinfo->si_xib_mtx); + AuDbg("i0\n"); + return 0; +} + +/* + * read @ino from xinofile for the specified branch{@sb, @bindex} + * at the position of @h_ino. + * if @ino does not exist and @do_new is true, get new one. + */ +int au_xino_read(struct super_block *sb, aufs_bindex_t bindex, ino_t h_ino, + ino_t *ino) +{ + int err; + ssize_t sz; + loff_t pos; + struct file *file; + struct au_sbinfo *sbinfo; + + *ino = 0; + if (!au_opt_test(au_mntflags(sb), XINO)) + return 0; /* no xino */ + + err = 0; + sbinfo = au_sbi(sb); + pos = h_ino; + if (unlikely(au_loff_max / sizeof(*ino) - 1 < pos)) { + AuIOErr1("too large hi%lu\n", (unsigned long)h_ino); + return -EFBIG; + } + pos *= sizeof(*ino); + + file = au_sbr(sb, bindex)->br_xino.xi_file; + if (vfsub_f_size_read(file) < pos + sizeof(*ino)) + return 0; /* no ino */ + + sz = xino_fread(sbinfo->si_xread, file, ino, sizeof(*ino), &pos); + if (sz == sizeof(*ino)) + return 0; /* success */ + + err = sz; + if (unlikely(sz >= 0)) { + err = -EIO; + AuIOErr("xino read error (%zd)\n", sz); + } + + return err; +} + +/* ---------------------------------------------------------------------- */ + +/* create and set a new xino file */ + +struct file *au_xino_create(struct super_block *sb, char *fname, int silent) +{ + struct file *file; + struct dentry *h_parent, *d; + struct inode *h_dir, *inode; + int err; + + /* + * at mount-time, and the xino file is the default path, + * hnotify is disabled so we have no notify events to ignore. + * when a user specified the xino, we cannot get au_hdir to be ignored. + */ + file = vfsub_filp_open(fname, O_RDWR | O_CREAT | O_EXCL | O_LARGEFILE + /* | __FMODE_NONOTIFY */, + S_IRUGO | S_IWUGO); + if (IS_ERR(file)) { + if (!silent) + pr_err("open %s(%ld)\n", fname, PTR_ERR(file)); + return file; + } + + /* keep file count */ + err = 0; + inode = file_inode(file); + h_parent = dget_parent(file->f_path.dentry); + h_dir = d_inode(h_parent); + inode_lock_nested(h_dir, AuLsc_I_PARENT); + /* mnt_want_write() is unnecessary here */ + /* no delegation since it is just created */ + if (inode->i_nlink) + err = vfsub_unlink(h_dir, &file->f_path, /*delegated*/NULL, + /*force*/0); + inode_unlock(h_dir); + dput(h_parent); + if (unlikely(err)) { + if (!silent) + pr_err("unlink %s(%d)\n", fname, err); + goto out; + } + + err = -EINVAL; + d = file->f_path.dentry; + if (unlikely(sb == d->d_sb)) { + if (!silent) + pr_err("%s must be outside\n", fname); + goto out; + } + if (unlikely(au_test_fs_bad_xino(d->d_sb))) { + if (!silent) + pr_err("xino doesn't support %s(%s)\n", + fname, au_sbtype(d->d_sb)); + goto out; + } + return file; /* success */ + +out: + fput(file); + file = ERR_PTR(err); + return file; +} + +/* + * find another branch who is on the same filesystem of the specified + * branch{@btgt}. search until @bend. + */ +static int is_sb_shared(struct super_block *sb, aufs_bindex_t btgt, + aufs_bindex_t bend) +{ + aufs_bindex_t bindex; + struct super_block *tgt_sb = au_sbr_sb(sb, btgt); + + for (bindex = 0; bindex < btgt; bindex++) + if (unlikely(tgt_sb == au_sbr_sb(sb, bindex))) + return bindex; + for (bindex++; bindex <= bend; bindex++) + if (unlikely(tgt_sb == au_sbr_sb(sb, bindex))) + return bindex; + return -1; +} + +/* ---------------------------------------------------------------------- */ + +/* + * initialize the xinofile for the specified branch @br + * at the place/path where @base_file indicates. + * test whether another branch is on the same filesystem or not, + * if @do_test is true. + */ +int au_xino_br(struct super_block *sb, struct au_branch *br, ino_t h_ino, + struct file *base_file, int do_test) +{ + int err; + ino_t ino; + aufs_bindex_t bend, bindex; + struct au_branch *shared_br, *b; + struct file *file; + struct super_block *tgt_sb; + + shared_br = NULL; + bend = au_sbend(sb); + if (do_test) { + tgt_sb = au_br_sb(br); + for (bindex = 0; bindex <= bend; bindex++) { + b = au_sbr(sb, bindex); + if (tgt_sb == au_br_sb(b)) { + shared_br = b; + break; + } + } + } + + if (!shared_br || !shared_br->br_xino.xi_file) { + struct au_xino_lock_dir ldir; + + au_xino_lock_dir(sb, base_file, &ldir); + /* mnt_want_write() is unnecessary here */ + file = au_xino_create2(base_file, NULL); + au_xino_unlock_dir(&ldir); + err = PTR_ERR(file); + if (IS_ERR(file)) + goto out; + br->br_xino.xi_file = file; + } else { + br->br_xino.xi_file = shared_br->br_xino.xi_file; + get_file(br->br_xino.xi_file); + } + + ino = AUFS_ROOT_INO; + err = au_xino_do_write(au_sbi(sb)->si_xwrite, br->br_xino.xi_file, + h_ino, ino); + if (unlikely(err)) { + fput(br->br_xino.xi_file); + br->br_xino.xi_file = NULL; + } + +out: + return err; +} + +/* ---------------------------------------------------------------------- */ + +/* trucate a xino bitmap file */ + +/* todo: slow */ +static int do_xib_restore(struct super_block *sb, struct file *file, void *page) +{ + int err, bit; + ssize_t sz; + unsigned long pindex; + loff_t pos, pend; + struct au_sbinfo *sbinfo; + vfs_readf_t func; + ino_t *ino; + unsigned long *p; + + err = 0; + sbinfo = au_sbi(sb); + MtxMustLock(&sbinfo->si_xib_mtx); + p = sbinfo->si_xib_buf; + func = sbinfo->si_xread; + pend = vfsub_f_size_read(file); + pos = 0; + while (pos < pend) { + sz = xino_fread(func, file, page, PAGE_SIZE, &pos); + err = sz; + if (unlikely(sz <= 0)) + goto out; + + err = 0; + for (ino = page; sz > 0; ino++, sz -= sizeof(ino)) { + if (unlikely(*ino < AUFS_FIRST_INO)) + continue; + + xib_calc_bit(*ino, &pindex, &bit); + AuDebugOn(page_bits <= bit); + err = xib_pindex(sb, pindex); + if (!err) + set_bit(bit, p); + else + goto out; + } + } + +out: + return err; +} + +static int xib_restore(struct super_block *sb) +{ + int err; + aufs_bindex_t bindex, bend; + void *page; + + err = -ENOMEM; + page = (void *)__get_free_page(GFP_NOFS); + if (unlikely(!page)) + goto out; + + err = 0; + bend = au_sbend(sb); + for (bindex = 0; !err && bindex <= bend; bindex++) + if (!bindex || is_sb_shared(sb, bindex, bindex - 1) < 0) + err = do_xib_restore + (sb, au_sbr(sb, bindex)->br_xino.xi_file, page); + else + AuDbg("b%d\n", bindex); + free_page((unsigned long)page); + +out: + return err; +} + +int au_xib_trunc(struct super_block *sb) +{ + int err; + ssize_t sz; + loff_t pos; + struct au_xino_lock_dir ldir; + struct au_sbinfo *sbinfo; + unsigned long *p; + struct file *file; + + SiMustWriteLock(sb); + + err = 0; + sbinfo = au_sbi(sb); + if (!au_opt_test(sbinfo->si_mntflags, XINO)) + goto out; + + file = sbinfo->si_xib; + if (vfsub_f_size_read(file) <= PAGE_SIZE) + goto out; + + au_xino_lock_dir(sb, file, &ldir); + /* mnt_want_write() is unnecessary here */ + file = au_xino_create2(sbinfo->si_xib, NULL); + au_xino_unlock_dir(&ldir); + err = PTR_ERR(file); + if (IS_ERR(file)) + goto out; + fput(sbinfo->si_xib); + sbinfo->si_xib = file; + + p = sbinfo->si_xib_buf; + memset(p, 0, PAGE_SIZE); + pos = 0; + sz = xino_fwrite(sbinfo->si_xwrite, sbinfo->si_xib, p, PAGE_SIZE, &pos); + if (unlikely(sz != PAGE_SIZE)) { + err = sz; + AuIOErr("err %d\n", err); + if (sz >= 0) + err = -EIO; + goto out; + } + + mutex_lock(&sbinfo->si_xib_mtx); + /* mnt_want_write() is unnecessary here */ + err = xib_restore(sb); + mutex_unlock(&sbinfo->si_xib_mtx); + +out: + return err; +} + +/* ---------------------------------------------------------------------- */ + +/* + * xino mount option handlers + */ + +/* xino bitmap */ +static void xino_clear_xib(struct super_block *sb) +{ + struct au_sbinfo *sbinfo; + + SiMustWriteLock(sb); + + sbinfo = au_sbi(sb); + sbinfo->si_xread = NULL; + sbinfo->si_xwrite = NULL; + if (sbinfo->si_xib) + fput(sbinfo->si_xib); + sbinfo->si_xib = NULL; + free_page((unsigned long)sbinfo->si_xib_buf); + sbinfo->si_xib_buf = NULL; +} + +static int au_xino_set_xib(struct super_block *sb, struct file *base) +{ + int err; + loff_t pos; + struct au_sbinfo *sbinfo; + struct file *file; + + SiMustWriteLock(sb); + + sbinfo = au_sbi(sb); + file = au_xino_create2(base, sbinfo->si_xib); + err = PTR_ERR(file); + if (IS_ERR(file)) + goto out; + if (sbinfo->si_xib) + fput(sbinfo->si_xib); + sbinfo->si_xib = file; + sbinfo->si_xread = vfs_readf(file); + sbinfo->si_xwrite = vfs_writef(file); + + err = -ENOMEM; + if (!sbinfo->si_xib_buf) + sbinfo->si_xib_buf = (void *)get_zeroed_page(GFP_NOFS); + if (unlikely(!sbinfo->si_xib_buf)) + goto out_unset; + + sbinfo->si_xib_last_pindex = 0; + sbinfo->si_xib_next_bit = 0; + if (vfsub_f_size_read(file) < PAGE_SIZE) { + pos = 0; + err = xino_fwrite(sbinfo->si_xwrite, file, sbinfo->si_xib_buf, + PAGE_SIZE, &pos); + if (unlikely(err != PAGE_SIZE)) + goto out_free; + } + err = 0; + goto out; /* success */ + +out_free: + free_page((unsigned long)sbinfo->si_xib_buf); + sbinfo->si_xib_buf = NULL; + if (err >= 0) + err = -EIO; +out_unset: + fput(sbinfo->si_xib); + sbinfo->si_xib = NULL; + sbinfo->si_xread = NULL; + sbinfo->si_xwrite = NULL; +out: + return err; +} + +/* xino for each branch */ +static void xino_clear_br(struct super_block *sb) +{ + aufs_bindex_t bindex, bend; + struct au_branch *br; + + bend = au_sbend(sb); + for (bindex = 0; bindex <= bend; bindex++) { + br = au_sbr(sb, bindex); + if (!br || !br->br_xino.xi_file) + continue; + + fput(br->br_xino.xi_file); + br->br_xino.xi_file = NULL; + } +} + +static int au_xino_set_br(struct super_block *sb, struct file *base) +{ + int err; + ino_t ino; + aufs_bindex_t bindex, bend, bshared; + struct { + struct file *old, *new; + } *fpair, *p; + struct au_branch *br; + struct inode *inode; + vfs_writef_t writef; + + SiMustWriteLock(sb); + + err = -ENOMEM; + bend = au_sbend(sb); + fpair = kcalloc(bend + 1, sizeof(*fpair), GFP_NOFS); + if (unlikely(!fpair)) + goto out; + + inode = d_inode(sb->s_root); + ino = AUFS_ROOT_INO; + writef = au_sbi(sb)->si_xwrite; + for (bindex = 0, p = fpair; bindex <= bend; bindex++, p++) { + br = au_sbr(sb, bindex); + bshared = is_sb_shared(sb, bindex, bindex - 1); + if (bshared >= 0) { + /* shared xino */ + *p = fpair[bshared]; + get_file(p->new); + } + + if (!p->new) { + /* new xino */ + p->old = br->br_xino.xi_file; + p->new = au_xino_create2(base, br->br_xino.xi_file); + err = PTR_ERR(p->new); + if (IS_ERR(p->new)) { + p->new = NULL; + goto out_pair; + } + } + + err = au_xino_do_write(writef, p->new, + au_h_iptr(inode, bindex)->i_ino, ino); + if (unlikely(err)) + goto out_pair; + } + + for (bindex = 0, p = fpair; bindex <= bend; bindex++, p++) { + br = au_sbr(sb, bindex); + if (br->br_xino.xi_file) + fput(br->br_xino.xi_file); + get_file(p->new); + br->br_xino.xi_file = p->new; + } + +out_pair: + for (bindex = 0, p = fpair; bindex <= bend; bindex++, p++) + if (p->new) + fput(p->new); + else + break; + kfree(fpair); +out: + return err; +} + +void au_xino_clr(struct super_block *sb) +{ + struct au_sbinfo *sbinfo; + + au_xigen_clr(sb); + xino_clear_xib(sb); + xino_clear_br(sb); + sbinfo = au_sbi(sb); + /* lvalue, do not call au_mntflags() */ + au_opt_clr(sbinfo->si_mntflags, XINO); +} + +int au_xino_set(struct super_block *sb, struct au_opt_xino *xino, int remount) +{ + int err, skip; + struct dentry *parent, *cur_parent; + struct qstr *dname, *cur_name; + struct file *cur_xino; + struct inode *dir; + struct au_sbinfo *sbinfo; + + SiMustWriteLock(sb); + + err = 0; + sbinfo = au_sbi(sb); + parent = dget_parent(xino->file->f_path.dentry); + if (remount) { + skip = 0; + dname = &xino->file->f_path.dentry->d_name; + cur_xino = sbinfo->si_xib; + if (cur_xino) { + cur_parent = dget_parent(cur_xino->f_path.dentry); + cur_name = &cur_xino->f_path.dentry->d_name; + skip = (cur_parent == parent + && au_qstreq(dname, cur_name)); + dput(cur_parent); + } + if (skip) + goto out; + } + + au_opt_set(sbinfo->si_mntflags, XINO); + dir = d_inode(parent); + inode_lock_nested(dir, AuLsc_I_PARENT); + /* mnt_want_write() is unnecessary here */ + err = au_xino_set_xib(sb, xino->file); + if (!err) + err = au_xigen_set(sb, xino->file); + if (!err) + err = au_xino_set_br(sb, xino->file); + inode_unlock(dir); + if (!err) + goto out; /* success */ + + /* reset all */ + AuIOErr("failed creating xino(%d).\n", err); + au_xigen_clr(sb); + xino_clear_xib(sb); + +out: + dput(parent); + return err; +} + +/* ---------------------------------------------------------------------- */ + +/* + * create a xinofile at the default place/path. + */ +struct file *au_xino_def(struct super_block *sb) +{ + struct file *file; + char *page, *p; + struct au_branch *br; + struct super_block *h_sb; + struct path path; + aufs_bindex_t bend, bindex, bwr; + + br = NULL; + bend = au_sbend(sb); + bwr = -1; + for (bindex = 0; bindex <= bend; bindex++) { + br = au_sbr(sb, bindex); + if (au_br_writable(br->br_perm) + && !au_test_fs_bad_xino(au_br_sb(br))) { + bwr = bindex; + break; + } + } + + if (bwr >= 0) { + file = ERR_PTR(-ENOMEM); + page = (void *)__get_free_page(GFP_NOFS); + if (unlikely(!page)) + goto out; + path.mnt = au_br_mnt(br); + path.dentry = au_h_dptr(sb->s_root, bwr); + p = d_path(&path, page, PATH_MAX - sizeof(AUFS_XINO_FNAME)); + file = (void *)p; + if (!IS_ERR(p)) { + strcat(p, "/" AUFS_XINO_FNAME); + AuDbg("%s\n", p); + file = au_xino_create(sb, p, /*silent*/0); + if (!IS_ERR(file)) + au_xino_brid_set(sb, br->br_id); + } + free_page((unsigned long)page); + } else { + file = au_xino_create(sb, AUFS_XINO_DEFPATH, /*silent*/0); + if (IS_ERR(file)) + goto out; + h_sb = file->f_path.dentry->d_sb; + if (unlikely(au_test_fs_bad_xino(h_sb))) { + pr_err("xino doesn't support %s(%s)\n", + AUFS_XINO_DEFPATH, au_sbtype(h_sb)); + fput(file); + file = ERR_PTR(-EINVAL); + } + if (!IS_ERR(file)) + au_xino_brid_set(sb, -1); + } + +out: + return file; +} + +/* ---------------------------------------------------------------------- */ + +int au_xino_path(struct seq_file *seq, struct file *file) +{ + int err; + + err = au_seq_path(seq, &file->f_path); + if (unlikely(err)) + goto out; + +#define Deleted "\\040(deleted)" + seq->count -= sizeof(Deleted) - 1; + AuDebugOn(memcmp(seq->buf + seq->count, Deleted, + sizeof(Deleted) - 1)); +#undef Deleted + +out: + return err; +} diff --git a/fs/coredump.c b/fs/coredump.c index 9ea87e9fd..47c32c3bf 100644 --- a/fs/coredump.c +++ b/fs/coredump.c @@ -32,6 +32,9 @@ #include <linux/pipe_fs_i.h> #include <linux/oom.h> #include <linux/compat.h> +#include <linux/sched.h> +#include <linux/fs.h> +#include <linux/path.h> #include <linux/timekeeping.h> #include <asm/uaccess.h> @@ -649,6 +652,8 @@ void do_coredump(const siginfo_t *siginfo) } } else { struct inode *inode; + int open_flags = O_CREAT | O_RDWR | O_NOFOLLOW | + O_LARGEFILE | O_EXCL; if (cprm.limit < binfmt->min_coredump) goto fail_unlock; @@ -687,10 +692,27 @@ void do_coredump(const siginfo_t *siginfo) * what matters is that at least one of the two processes * writes its coredump successfully, not which one. */ - cprm.file = filp_open(cn.corename, - O_CREAT | 2 | O_NOFOLLOW | - O_LARGEFILE | O_EXCL, - 0600); + if (need_suid_safe) { + /* + * Using user namespaces, normal user tasks can change + * their current->fs->root to point to arbitrary + * directories. Since the intention of the "only dump + * with a fully qualified path" rule is to control where + * coredumps may be placed using root privileges, + * current->fs->root must not be used. Instead, use the + * root directory of init_task. + */ + struct path root; + + task_lock(&init_task); + get_fs_root(init_task.fs, &root); + task_unlock(&init_task); + cprm.file = file_open_root(root.dentry, root.mnt, + cn.corename, open_flags, 0600); + path_put(&root); + } else { + cprm.file = filp_open(cn.corename, open_flags, 0600); + } if (IS_ERR(cprm.file)) goto fail_unlock; diff --git a/fs/dcache.c b/fs/dcache.c index 2398f9f94..72d3cc89c 100644 --- a/fs/dcache.c +++ b/fs/dcache.c @@ -1156,7 +1156,7 @@ enum d_walk_ret { * * The @enter() and @finish() callbacks are called with d_lock held. */ -static void d_walk(struct dentry *parent, void *data, +void d_walk(struct dentry *parent, void *data, enum d_walk_ret (*enter)(void *, struct dentry *), void (*finish)(void *)) { @@ -1261,6 +1261,7 @@ rename_retry: seq = 1; goto again; } +EXPORT_SYMBOL_GPL(d_walk); /* * Search for at least 1 mount point in the dentry's subdirs. @@ -105,6 +105,7 @@ bool path_noexec(const struct path *path) return (path->mnt->mnt_flags & MNT_NOEXEC) || (path->mnt->mnt_sb->s_iflags & SB_I_NOEXEC); } +EXPORT_SYMBOL_GPL(path_noexec); #ifdef CONFIG_USELIB /* diff --git a/fs/exfat/exfat_super.c b/fs/exfat/exfat_super.c index a003a7212..02d4fd2c7 100644 --- a/fs/exfat/exfat_super.c +++ b/fs/exfat/exfat_super.c @@ -1359,7 +1359,19 @@ const struct inode_operations exfat_dir_inode_operations = { /*======================================================================*/ /* File Operations */ /*======================================================================*/ -#if LINUX_VERSION_CODE > KERNEL_VERSION(4,1,0) +#if LINUX_VERSION_CODE >= KERNEL_VERSION(4,5,0) +static const char *exfat_get_link(struct dentry *dentry, struct inode *inode, struct delayed_call *done) +{ + struct exfat_inode_info *ei = EXFAT_I(inode); + if (ei->target != NULL) { + char *cookie = ei->target; + if (cookie != NULL) { + return (char *)(ei->target); + } + } + return NULL; +} +#elif LINUX_VERSION_CODE > KERNEL_VERSION(4,1,0) static const char *exfat_follow_link(struct dentry *dentry, void **cookie) { struct exfat_inode_info *ei = EXFAT_I(dentry->d_inode); @@ -1376,7 +1388,12 @@ static void *exfat_follow_link(struct dentry *dentry, struct nameidata *nd) const struct inode_operations exfat_symlink_inode_operations = { .readlink = generic_readlink, - .follow_link = exfat_follow_link, + #if LINUX_VERSION_CODE < KERNEL_VERSION(4,5,0) + .follow_link = exfat_follow_link, + #endif + #if LINUX_VERSION_CODE >= KERNEL_VERSION(4,5,0) + .get_link = exfat_get_link, + #endif }; static int exfat_file_release(struct inode *inode, struct file *filp) @@ -1970,7 +1987,7 @@ static void exfat_evict_inode(struct inode *inode) if (!inode->i_nlink) i_size_write(inode, 0); invalidate_inode_buffers(inode); -#if LINUX_VERSION_CODE < KERNEL_VERSION(3,4,80) +#if LINUX_VERSION_CODE < KERNEL_VERSION(3,5,0) end_writeback(inode); #else clear_inode(inode); diff --git a/fs/fcntl.c b/fs/fcntl.c index 350a2c8cf..04fd33cbf 100644 --- a/fs/fcntl.c +++ b/fs/fcntl.c @@ -29,7 +29,7 @@ #define SETFL_MASK (O_APPEND | O_NONBLOCK | O_NDELAY | O_DIRECT | O_NOATIME) -static int setfl(int fd, struct file * filp, unsigned long arg) +int setfl(int fd, struct file * filp, unsigned long arg) { struct inode * inode = file_inode(filp); int error = 0; @@ -60,6 +60,8 @@ static int setfl(int fd, struct file * filp, unsigned long arg) if (filp->f_op->check_flags) error = filp->f_op->check_flags(arg); + if (!error && filp->f_op->setfl) + error = filp->f_op->setfl(filp, arg); if (error) return error; @@ -80,6 +82,7 @@ static int setfl(int fd, struct file * filp, unsigned long arg) out: return error; } +EXPORT_SYMBOL_GPL(setfl); static void f_modown(struct file *filp, struct pid *pid, enum pid_type type, int force) diff --git a/fs/fhandle.c b/fs/fhandle.c index d59712dfa..ca3c3dd01 100644 --- a/fs/fhandle.c +++ b/fs/fhandle.c @@ -228,7 +228,7 @@ long do_handle_open(int mountdirfd, path_put(&path); return fd; } - file = file_open_root(path.dentry, path.mnt, "", open_flag); + file = file_open_root(path.dentry, path.mnt, "", open_flag, 0); if (IS_ERR(file)) { put_unused_fd(fd); retval = PTR_ERR(file); diff --git a/fs/file_table.c b/fs/file_table.c index ad17e05eb..ae9f2676d 100644 --- a/fs/file_table.c +++ b/fs/file_table.c @@ -147,6 +147,7 @@ over: } return ERR_PTR(-ENFILE); } +EXPORT_SYMBOL_GPL(get_empty_filp); /** * alloc_file - allocate and initialize a 'struct file' @@ -258,6 +259,7 @@ void flush_delayed_fput(void) { delayed_fput(NULL); } +EXPORT_SYMBOL_GPL(flush_delayed_fput); static DECLARE_DELAYED_WORK(delayed_fput_work, delayed_fput); @@ -300,6 +302,7 @@ void __fput_sync(struct file *file) } EXPORT_SYMBOL(fput); +EXPORT_SYMBOL_GPL(__fput_sync); void put_filp(struct file *file) { @@ -308,6 +311,7 @@ void put_filp(struct file *file) file_free(file); } } +EXPORT_SYMBOL_GPL(put_filp); void __init files_init(void) { diff --git a/fs/fs-writeback.c b/fs/fs-writeback.c index 5c46ed9f3..fee81e876 100644 --- a/fs/fs-writeback.c +++ b/fs/fs-writeback.c @@ -281,13 +281,15 @@ locked_inode_to_wb_and_lock_list(struct inode *inode) wb_get(wb); spin_unlock(&inode->i_lock); spin_lock(&wb->list_lock); - wb_put(wb); /* not gonna deref it anymore */ /* i_wb may have changed inbetween, can't use inode_to_wb() */ - if (likely(wb == inode->i_wb)) - return wb; /* @inode already has ref */ + if (likely(wb == inode->i_wb)) { + wb_put(wb); /* @inode already has ref */ + return wb; + } spin_unlock(&wb->list_lock); + wb_put(wb); cpu_relax(); spin_lock(&inode->i_lock); } @@ -1337,10 +1339,10 @@ __writeback_single_inode(struct inode *inode, struct writeback_control *wbc) * we go e.g. from filesystem. Flusher thread uses __writeback_single_inode() * and does more profound writeback list handling in writeback_sb_inodes(). */ -static int -writeback_single_inode(struct inode *inode, struct bdi_writeback *wb, - struct writeback_control *wbc) +static int writeback_single_inode(struct inode *inode, + struct writeback_control *wbc) { + struct bdi_writeback *wb; int ret = 0; spin_lock(&inode->i_lock); @@ -1378,7 +1380,8 @@ writeback_single_inode(struct inode *inode, struct bdi_writeback *wb, ret = __writeback_single_inode(inode, wbc); wbc_detach_inode(wbc); - spin_lock(&wb->list_lock); + + wb = inode_to_wb_and_lock_list(inode); spin_lock(&inode->i_lock); /* * If inode is clean, remove it from writeback lists. Otherwise don't @@ -1453,6 +1456,7 @@ static long writeback_sb_inodes(struct super_block *sb, while (!list_empty(&wb->b_io)) { struct inode *inode = wb_inode(wb->b_io.prev); + struct bdi_writeback *tmp_wb; if (inode->i_sb != sb) { if (work->sb) { @@ -1543,15 +1547,23 @@ static long writeback_sb_inodes(struct super_block *sb, cond_resched(); } - - spin_lock(&wb->list_lock); + /* + * Requeue @inode if still dirty. Be careful as @inode may + * have been switched to another wb in the meantime. + */ + tmp_wb = inode_to_wb_and_lock_list(inode); spin_lock(&inode->i_lock); if (!(inode->i_state & I_DIRTY_ALL)) wrote++; - requeue_inode(inode, wb, &wbc); + requeue_inode(inode, tmp_wb, &wbc); inode_sync_complete(inode); spin_unlock(&inode->i_lock); + if (unlikely(tmp_wb != wb)) { + spin_unlock(&tmp_wb->list_lock); + spin_lock(&wb->list_lock); + } + /* * bail out to wb_writeback() often enough to check * background threshold and other termination conditions. @@ -2338,7 +2350,6 @@ EXPORT_SYMBOL(sync_inodes_sb); */ int write_inode_now(struct inode *inode, int sync) { - struct bdi_writeback *wb = &inode_to_bdi(inode)->wb; struct writeback_control wbc = { .nr_to_write = LONG_MAX, .sync_mode = sync ? WB_SYNC_ALL : WB_SYNC_NONE, @@ -2350,7 +2361,7 @@ int write_inode_now(struct inode *inode, int sync) wbc.nr_to_write = 0; might_sleep(); - return writeback_single_inode(inode, wb, &wbc); + return writeback_single_inode(inode, &wbc); } EXPORT_SYMBOL(write_inode_now); @@ -2367,7 +2378,7 @@ EXPORT_SYMBOL(write_inode_now); */ int sync_inode(struct inode *inode, struct writeback_control *wbc) { - return writeback_single_inode(inode, &inode_to_bdi(inode)->wb, wbc); + return writeback_single_inode(inode, wbc); } EXPORT_SYMBOL(sync_inode); diff --git a/fs/fuse/cuse.c b/fs/fuse/cuse.c index 8e3ee1936..c5b6b7165 100644 --- a/fs/fuse/cuse.c +++ b/fs/fuse/cuse.c @@ -90,7 +90,7 @@ static struct list_head *cuse_conntbl_head(dev_t devt) static ssize_t cuse_read_iter(struct kiocb *kiocb, struct iov_iter *to) { - struct fuse_io_priv io = { .async = 0, .file = kiocb->ki_filp }; + struct fuse_io_priv io = FUSE_IO_PRIV_SYNC(kiocb->ki_filp); loff_t pos = 0; return fuse_direct_io(&io, to, &pos, FUSE_DIO_CUSE); @@ -98,7 +98,7 @@ static ssize_t cuse_read_iter(struct kiocb *kiocb, struct iov_iter *to) static ssize_t cuse_write_iter(struct kiocb *kiocb, struct iov_iter *from) { - struct fuse_io_priv io = { .async = 0, .file = kiocb->ki_filp }; + struct fuse_io_priv io = FUSE_IO_PRIV_SYNC(kiocb->ki_filp); loff_t pos = 0; /* * No locking or generic_write_checks(), the server is diff --git a/fs/fuse/file.c b/fs/fuse/file.c index b03d253ec..416108b42 100644 --- a/fs/fuse/file.c +++ b/fs/fuse/file.c @@ -528,6 +528,11 @@ static void fuse_release_user_pages(struct fuse_req *req, int write) } } +static void fuse_io_release(struct kref *kref) +{ + kfree(container_of(kref, struct fuse_io_priv, refcnt)); +} + static ssize_t fuse_get_res_by_io(struct fuse_io_priv *io) { if (io->err) @@ -585,8 +590,9 @@ static void fuse_aio_complete(struct fuse_io_priv *io, int err, ssize_t pos) } io->iocb->ki_complete(io->iocb, res, 0); - kfree(io); } + + kref_put(&io->refcnt, fuse_io_release); } static void fuse_aio_complete_req(struct fuse_conn *fc, struct fuse_req *req) @@ -613,6 +619,7 @@ static size_t fuse_async_req_send(struct fuse_conn *fc, struct fuse_req *req, size_t num_bytes, struct fuse_io_priv *io) { spin_lock(&io->lock); + kref_get(&io->refcnt); io->size += num_bytes; io->reqs++; spin_unlock(&io->lock); @@ -691,7 +698,7 @@ static void fuse_short_read(struct fuse_req *req, struct inode *inode, static int fuse_do_readpage(struct file *file, struct page *page) { - struct fuse_io_priv io = { .async = 0, .file = file }; + struct fuse_io_priv io = FUSE_IO_PRIV_SYNC(file); struct inode *inode = page->mapping->host; struct fuse_conn *fc = get_fuse_conn(inode); struct fuse_req *req; @@ -984,7 +991,7 @@ static size_t fuse_send_write_pages(struct fuse_req *req, struct file *file, size_t res; unsigned offset; unsigned i; - struct fuse_io_priv io = { .async = 0, .file = file }; + struct fuse_io_priv io = FUSE_IO_PRIV_SYNC(file); for (i = 0; i < req->num_pages; i++) fuse_wait_on_page_writeback(inode, req->pages[i]->index); @@ -1398,7 +1405,7 @@ static ssize_t __fuse_direct_read(struct fuse_io_priv *io, static ssize_t fuse_direct_read_iter(struct kiocb *iocb, struct iov_iter *to) { - struct fuse_io_priv io = { .async = 0, .file = iocb->ki_filp }; + struct fuse_io_priv io = FUSE_IO_PRIV_SYNC(iocb->ki_filp); return __fuse_direct_read(&io, to, &iocb->ki_pos); } @@ -1406,7 +1413,7 @@ static ssize_t fuse_direct_write_iter(struct kiocb *iocb, struct iov_iter *from) { struct file *file = iocb->ki_filp; struct inode *inode = file_inode(file); - struct fuse_io_priv io = { .async = 0, .file = file }; + struct fuse_io_priv io = FUSE_IO_PRIV_SYNC(file); ssize_t res; if (is_bad_inode(inode)) @@ -2843,6 +2850,7 @@ fuse_direct_IO(struct kiocb *iocb, struct iov_iter *iter, loff_t offset) loff_t i_size; size_t count = iov_iter_count(iter); struct fuse_io_priv *io; + bool is_sync = is_sync_kiocb(iocb); pos = offset; inode = file->f_mapping->host; @@ -2863,6 +2871,7 @@ fuse_direct_IO(struct kiocb *iocb, struct iov_iter *iter, loff_t offset) if (!io) return -ENOMEM; spin_lock_init(&io->lock); + kref_init(&io->refcnt); io->reqs = 1; io->bytes = -1; io->size = 0; @@ -2882,12 +2891,18 @@ fuse_direct_IO(struct kiocb *iocb, struct iov_iter *iter, loff_t offset) * to wait on real async I/O requests, so we must submit this request * synchronously. */ - if (!is_sync_kiocb(iocb) && (offset + count > i_size) && + if (!is_sync && (offset + count > i_size) && iov_iter_rw(iter) == WRITE) io->async = false; - if (io->async && is_sync_kiocb(iocb)) + if (io->async && is_sync) { + /* + * Additional reference to keep io around after + * calling fuse_aio_complete() + */ + kref_get(&io->refcnt); io->done = &wait; + } if (iov_iter_rw(iter) == WRITE) { ret = fuse_direct_io(io, iter, &pos, FUSE_DIO_WRITE); @@ -2900,14 +2915,14 @@ fuse_direct_IO(struct kiocb *iocb, struct iov_iter *iter, loff_t offset) fuse_aio_complete(io, ret < 0 ? ret : 0, -1); /* we have a non-extending, async request, so return */ - if (!is_sync_kiocb(iocb)) + if (!is_sync) return -EIOCBQUEUED; wait_for_completion(&wait); ret = fuse_get_res_by_io(io); } - kfree(io); + kref_put(&io->refcnt, fuse_io_release); if (iov_iter_rw(iter) == WRITE) { if (ret > 0) diff --git a/fs/fuse/fuse_i.h b/fs/fuse/fuse_i.h index ce394b5fe..eddbe02c4 100644 --- a/fs/fuse/fuse_i.h +++ b/fs/fuse/fuse_i.h @@ -22,6 +22,7 @@ #include <linux/rbtree.h> #include <linux/poll.h> #include <linux/workqueue.h> +#include <linux/kref.h> /** Max number of pages that can be used in a single read request */ #define FUSE_MAX_PAGES_PER_REQ 32 @@ -243,6 +244,7 @@ struct fuse_args { /** The request IO state (for asynchronous processing) */ struct fuse_io_priv { + struct kref refcnt; int async; spinlock_t lock; unsigned reqs; @@ -256,6 +258,13 @@ struct fuse_io_priv { struct completion *done; }; +#define FUSE_IO_PRIV_SYNC(f) \ +{ \ + .refcnt = { ATOMIC_INIT(1) }, \ + .async = 0, \ + .file = f, \ +} + /** * Request flags * diff --git a/fs/jbd2/journal.c b/fs/jbd2/journal.c index 81e622681..624a57a9c 100644 --- a/fs/jbd2/journal.c +++ b/fs/jbd2/journal.c @@ -1408,11 +1408,12 @@ out: /** * jbd2_mark_journal_empty() - Mark on disk journal as empty. * @journal: The journal to update. + * @write_op: With which operation should we write the journal sb * * Update a journal's dynamic superblock fields to show that journal is empty. * Write updated superblock to disk waiting for IO to complete. */ -static void jbd2_mark_journal_empty(journal_t *journal) +static void jbd2_mark_journal_empty(journal_t *journal, int write_op) { journal_superblock_t *sb = journal->j_superblock; @@ -1430,7 +1431,7 @@ static void jbd2_mark_journal_empty(journal_t *journal) sb->s_start = cpu_to_be32(0); read_unlock(&journal->j_state_lock); - jbd2_write_superblock(journal, WRITE_FUA); + jbd2_write_superblock(journal, write_op); /* Log is no longer empty */ write_lock(&journal->j_state_lock); @@ -1716,7 +1717,13 @@ int jbd2_journal_destroy(journal_t *journal) if (journal->j_sb_buffer) { if (!is_journal_aborted(journal)) { mutex_lock(&journal->j_checkpoint_mutex); - jbd2_mark_journal_empty(journal); + + write_lock(&journal->j_state_lock); + journal->j_tail_sequence = + ++journal->j_transaction_sequence; + write_unlock(&journal->j_state_lock); + + jbd2_mark_journal_empty(journal, WRITE_FLUSH_FUA); mutex_unlock(&journal->j_checkpoint_mutex); } else err = -EIO; @@ -1975,7 +1982,7 @@ int jbd2_journal_flush(journal_t *journal) * the magic code for a fully-recovered superblock. Any future * commits of data to the journal will restore the current * s_start value. */ - jbd2_mark_journal_empty(journal); + jbd2_mark_journal_empty(journal, WRITE_FUA); mutex_unlock(&journal->j_checkpoint_mutex); write_lock(&journal->j_state_lock); J_ASSERT(!journal->j_running_transaction); @@ -2021,7 +2028,7 @@ int jbd2_journal_wipe(journal_t *journal, int write) if (write) { /* Lock to make assertions happy... */ mutex_lock(&journal->j_checkpoint_mutex); - jbd2_mark_journal_empty(journal); + jbd2_mark_journal_empty(journal, WRITE_FUA); mutex_unlock(&journal->j_checkpoint_mutex); } diff --git a/fs/namespace.c b/fs/namespace.c index 4fb1691b4..97654d2d8 100644 --- a/fs/namespace.c +++ b/fs/namespace.c @@ -463,6 +463,7 @@ void __mnt_drop_write(struct vfsmount *mnt) mnt_dec_writers(real_mount(mnt)); preempt_enable(); } +EXPORT_SYMBOL_GPL(__mnt_drop_write); /** * mnt_drop_write - give up write access to a mount @@ -1811,6 +1812,7 @@ int iterate_mounts(int (*f)(struct vfsmount *, void *), void *arg, } return 0; } +EXPORT_SYMBOL_GPL(iterate_mounts); static void cleanup_group_ids(struct mount *mnt, struct mount *end) { diff --git a/fs/nfsd/nfs4proc.c b/fs/nfsd/nfs4proc.c index 4cba7865f..f8082c7cd 100644 --- a/fs/nfsd/nfs4proc.c +++ b/fs/nfsd/nfs4proc.c @@ -878,6 +878,7 @@ nfsd4_secinfo(struct svc_rqst *rqstp, struct nfsd4_compound_state *cstate, &exp, &dentry); if (err) return err; + fh_unlock(&cstate->current_fh); if (d_really_is_negative(dentry)) { exp_put(exp); err = nfserr_noent; diff --git a/fs/nfsd/nfs4xdr.c b/fs/nfsd/nfs4xdr.c index d6ef0955a..1600ec470 100644 --- a/fs/nfsd/nfs4xdr.c +++ b/fs/nfsd/nfs4xdr.c @@ -1072,8 +1072,9 @@ nfsd4_decode_rename(struct nfsd4_compoundargs *argp, struct nfsd4_rename *rename READ_BUF(4); rename->rn_snamelen = be32_to_cpup(p++); - READ_BUF(rename->rn_snamelen + 4); + READ_BUF(rename->rn_snamelen); SAVEMEM(rename->rn_sname, rename->rn_snamelen); + READ_BUF(4); rename->rn_tnamelen = be32_to_cpup(p++); READ_BUF(rename->rn_tnamelen); SAVEMEM(rename->rn_tname, rename->rn_tnamelen); @@ -1155,13 +1156,14 @@ nfsd4_decode_setclientid(struct nfsd4_compoundargs *argp, struct nfsd4_setclient READ_BUF(8); setclientid->se_callback_prog = be32_to_cpup(p++); setclientid->se_callback_netid_len = be32_to_cpup(p++); - - READ_BUF(setclientid->se_callback_netid_len + 4); + READ_BUF(setclientid->se_callback_netid_len); SAVEMEM(setclientid->se_callback_netid_val, setclientid->se_callback_netid_len); + READ_BUF(4); setclientid->se_callback_addr_len = be32_to_cpup(p++); - READ_BUF(setclientid->se_callback_addr_len + 4); + READ_BUF(setclientid->se_callback_addr_len); SAVEMEM(setclientid->se_callback_addr_val, setclientid->se_callback_addr_len); + READ_BUF(4); setclientid->se_callback_ident = be32_to_cpup(p++); DECODE_TAIL; @@ -1835,8 +1837,9 @@ nfsd4_decode_compound(struct nfsd4_compoundargs *argp) READ_BUF(4); argp->taglen = be32_to_cpup(p++); - READ_BUF(argp->taglen + 8); + READ_BUF(argp->taglen); SAVEMEM(argp->tag, argp->taglen); + READ_BUF(8); argp->minorversion = be32_to_cpup(p++); argp->opcnt = be32_to_cpup(p++); max_reply += 4 + (XDR_QUADLEN(argp->taglen) << 2); diff --git a/fs/notify/group.c b/fs/notify/group.c index d16b62cb2..53e45b61d 100644 --- a/fs/notify/group.c +++ b/fs/notify/group.c @@ -22,6 +22,7 @@ #include <linux/srcu.h> #include <linux/rculist.h> #include <linux/wait.h> +#include <linux/module.h> #include <linux/fsnotify_backend.h> #include "fsnotify.h" @@ -72,6 +73,7 @@ void fsnotify_get_group(struct fsnotify_group *group) { atomic_inc(&group->refcnt); } +EXPORT_SYMBOL_GPL(fsnotify_get_group); /* * Drop a reference to a group. Free it if it's through. @@ -81,6 +83,7 @@ void fsnotify_put_group(struct fsnotify_group *group) if (atomic_dec_and_test(&group->refcnt)) fsnotify_final_destroy_group(group); } +EXPORT_SYMBOL_GPL(fsnotify_put_group); /* * Create a new fsnotify_group and hold a reference for the group returned. @@ -109,6 +112,7 @@ struct fsnotify_group *fsnotify_alloc_group(const struct fsnotify_ops *ops) return group; } +EXPORT_SYMBOL_GPL(fsnotify_alloc_group); int fsnotify_fasync(int fd, struct file *file, int on) { diff --git a/fs/notify/mark.c b/fs/notify/mark.c index 7115c5d7d..ac2bd697f 100644 --- a/fs/notify/mark.c +++ b/fs/notify/mark.c @@ -113,6 +113,7 @@ void fsnotify_put_mark(struct fsnotify_mark *mark) mark->free_mark(mark); } } +EXPORT_SYMBOL_GPL(fsnotify_put_mark); /* Calculate mask of events for a list of marks */ u32 fsnotify_recalc_mask(struct hlist_head *head) @@ -213,6 +214,7 @@ void fsnotify_destroy_mark(struct fsnotify_mark *mark, mutex_unlock(&group->mark_mutex); fsnotify_free_mark(mark); } +EXPORT_SYMBOL_GPL(fsnotify_destroy_mark); void fsnotify_destroy_marks(struct hlist_head *head, spinlock_t *lock) { @@ -398,6 +400,7 @@ err: return ret; } +EXPORT_SYMBOL_GPL(fsnotify_add_mark); int fsnotify_add_mark(struct fsnotify_mark *mark, struct fsnotify_group *group, struct inode *inode, struct vfsmount *mnt, int allow_dups) @@ -498,6 +501,7 @@ void fsnotify_init_mark(struct fsnotify_mark *mark, atomic_set(&mark->refcnt, 1); mark->free_mark = free_mark; } +EXPORT_SYMBOL_GPL(fsnotify_init_mark); static void fsnotify_mark_destroy(struct work_struct *work) { diff --git a/fs/ocfs2/cluster/heartbeat.c b/fs/ocfs2/cluster/heartbeat.c index a76b9ea77..a2370e2c7 100644 --- a/fs/ocfs2/cluster/heartbeat.c +++ b/fs/ocfs2/cluster/heartbeat.c @@ -1445,8 +1445,8 @@ static void o2hb_region_release(struct config_item *item) debugfs_remove(reg->hr_debug_dir); kfree(reg->hr_db_livenodes); kfree(reg->hr_db_regnum); - kfree(reg->hr_debug_elapsed_time); - kfree(reg->hr_debug_pinned); + kfree(reg->hr_db_elapsed_time); + kfree(reg->hr_db_pinned); spin_lock(&o2hb_live_lock); list_del(®->hr_all_item); diff --git a/fs/ocfs2/dlm/dlmconvert.c b/fs/ocfs2/dlm/dlmconvert.c index e36d63ff1..f90931335 100644 --- a/fs/ocfs2/dlm/dlmconvert.c +++ b/fs/ocfs2/dlm/dlmconvert.c @@ -262,6 +262,7 @@ enum dlm_status dlmconvert_remote(struct dlm_ctxt *dlm, struct dlm_lock *lock, int flags, int type) { enum dlm_status status; + u8 old_owner = res->owner; mlog(0, "type=%d, convert_type=%d, busy=%d\n", lock->ml.type, lock->ml.convert_type, res->state & DLM_LOCK_RES_IN_PROGRESS); @@ -287,6 +288,19 @@ enum dlm_status dlmconvert_remote(struct dlm_ctxt *dlm, status = DLM_DENIED; goto bail; } + + if (lock->ml.type == type && lock->ml.convert_type == LKM_IVMODE) { + mlog(0, "last convert request returned DLM_RECOVERING, but " + "owner has already queued and sent ast to me. res %.*s, " + "(cookie=%u:%llu, type=%d, conv=%d)\n", + res->lockname.len, res->lockname.name, + dlm_get_lock_cookie_node(be64_to_cpu(lock->ml.cookie)), + dlm_get_lock_cookie_seq(be64_to_cpu(lock->ml.cookie)), + lock->ml.type, lock->ml.convert_type); + status = DLM_NORMAL; + goto bail; + } + res->state |= DLM_LOCK_RES_IN_PROGRESS; /* move lock to local convert queue */ /* do not alter lock refcount. switching lists. */ @@ -316,11 +330,19 @@ enum dlm_status dlmconvert_remote(struct dlm_ctxt *dlm, spin_lock(&res->spinlock); res->state &= ~DLM_LOCK_RES_IN_PROGRESS; lock->convert_pending = 0; - /* if it failed, move it back to granted queue */ + /* if it failed, move it back to granted queue. + * if master returns DLM_NORMAL and then down before sending ast, + * it may have already been moved to granted queue, reset to + * DLM_RECOVERING and retry convert */ if (status != DLM_NORMAL) { if (status != DLM_NOTQUEUED) dlm_error(status); dlm_revert_pending_convert(res, lock); + } else if ((res->state & DLM_LOCK_RES_RECOVERING) || + (old_owner != res->owner)) { + mlog(0, "res %.*s is in recovering or has been recovered.\n", + res->lockname.len, res->lockname.name); + status = DLM_RECOVERING; } bail: spin_unlock(&res->spinlock); diff --git a/fs/ocfs2/dlm/dlmrecovery.c b/fs/ocfs2/dlm/dlmrecovery.c index b94a425f0..23d0ab881 100644 --- a/fs/ocfs2/dlm/dlmrecovery.c +++ b/fs/ocfs2/dlm/dlmrecovery.c @@ -2071,7 +2071,6 @@ void dlm_move_lockres_to_recovery_list(struct dlm_ctxt *dlm, dlm_lock_get(lock); if (lock->convert_pending) { /* move converting lock back to granted */ - BUG_ON(i != DLM_CONVERTING_LIST); mlog(0, "node died with convert pending " "on %.*s. move back to granted list.\n", res->lockname.len, res->lockname.name); @@ -67,6 +67,7 @@ int do_truncate(struct dentry *dentry, loff_t length, unsigned int time_attrs, inode_unlock(dentry->d_inode); return ret; } +EXPORT_SYMBOL_GPL(do_truncate); long vfs_truncate(struct path *path, loff_t length) { @@ -681,6 +682,7 @@ int open_check_o_direct(struct file *f) } return 0; } +EXPORT_SYMBOL_GPL(open_check_o_direct); static int do_dentry_open(struct file *f, struct inode *inode, @@ -995,14 +997,12 @@ struct file *filp_open(const char *filename, int flags, umode_t mode) EXPORT_SYMBOL(filp_open); struct file *file_open_root(struct dentry *dentry, struct vfsmount *mnt, - const char *filename, int flags) + const char *filename, int flags, umode_t mode) { struct open_flags op; - int err = build_open_flags(flags, 0, &op); + int err = build_open_flags(flags, mode, &op); if (err) return ERR_PTR(err); - if (flags & O_CREAT) - return ERR_PTR(-EINVAL); return do_file_open_root(dentry, mnt, filename, &op); } EXPORT_SYMBOL(file_open_root); diff --git a/fs/proc/base.c b/fs/proc/base.c index 4f764c2ac..11aba74cb 100644 --- a/fs/proc/base.c +++ b/fs/proc/base.c @@ -505,7 +505,7 @@ static int proc_pid_schedstat(struct seq_file *m, struct pid_namespace *ns, seq_printf(m, "0 0 0\n"); else seq_printf(m, "%llu %llu %lu\n", - (unsigned long long)task->se.sum_exec_runtime, + (unsigned long long)tsk_seruntime(task), (unsigned long long)task->sched_info.run_delay, task->sched_info.pcount); @@ -1933,7 +1933,7 @@ static int map_files_get_link(struct dentry *dentry, struct path *path) down_read(&mm->mmap_sem); vma = find_exact_vma(mm, vm_start, vm_end); if (vma && vma->vm_file) { - *path = vma->vm_file->f_path; + *path = vma_pr_or_file(vma)->f_path; path_get(path); rc = 0; } diff --git a/fs/proc/nommu.c b/fs/proc/nommu.c index f8595e8b5..cb8eda060 100644 --- a/fs/proc/nommu.c +++ b/fs/proc/nommu.c @@ -45,7 +45,10 @@ static int nommu_region_show(struct seq_file *m, struct vm_region *region) file = region->vm_file; if (file) { - struct inode *inode = file_inode(region->vm_file); + struct inode *inode; + + file = vmr_pr_or_file(region); + inode = file_inode(file); dev = inode->i_sb->s_dev; ino = inode->i_ino; } diff --git a/fs/proc/task_mmu.c b/fs/proc/task_mmu.c index fa95ab2d3..d4403544a 100644 --- a/fs/proc/task_mmu.c +++ b/fs/proc/task_mmu.c @@ -298,7 +298,10 @@ show_map_vma(struct seq_file *m, struct vm_area_struct *vma, int is_pid) const char *name = NULL; if (file) { - struct inode *inode = file_inode(vma->vm_file); + struct inode *inode; + + file = vma_pr_or_file(vma); + inode = file_inode(file); dev = inode->i_sb->s_dev; ino = inode->i_ino; pgoff = ((loff_t)vma->vm_pgoff) << PAGE_SHIFT; @@ -1576,7 +1579,7 @@ static int show_numa_map(struct seq_file *m, void *v, int is_pid) struct proc_maps_private *proc_priv = &numa_priv->proc_maps; struct vm_area_struct *vma = v; struct numa_maps *md = &numa_priv->md; - struct file *file = vma->vm_file; + struct file *file = vma_pr_or_file(vma); struct mm_struct *mm = vma->vm_mm; struct mm_walk walk = { .hugetlb_entry = gather_hugetlb_stats, diff --git a/fs/proc/task_nommu.c b/fs/proc/task_nommu.c index faacb0c0d..17b43be13 100644 --- a/fs/proc/task_nommu.c +++ b/fs/proc/task_nommu.c @@ -163,7 +163,10 @@ static int nommu_vma_show(struct seq_file *m, struct vm_area_struct *vma, file = vma->vm_file; if (file) { - struct inode *inode = file_inode(vma->vm_file); + struct inode *inode; + + file = vma_pr_or_file(vma); + inode = file_inode(file); dev = inode->i_sb->s_dev; ino = inode->i_ino; pgoff = (loff_t)vma->vm_pgoff << PAGE_SHIFT; diff --git a/fs/proc_namespace.c b/fs/proc_namespace.c index 2256e7e23..3f1190d18 100644 --- a/fs/proc_namespace.c +++ b/fs/proc_namespace.c @@ -199,6 +199,8 @@ static int show_vfsstat(struct seq_file *m, struct vfsmount *mnt) if (sb->s_op->show_devname) { seq_puts(m, "device "); err = sb->s_op->show_devname(m, mnt_path.dentry); + if (err) + goto out; } else { if (r->mnt_devname) { seq_puts(m, "device "); diff --git a/fs/quota/dquot.c b/fs/quota/dquot.c index 3c3b81bb6..850d17fa0 100644 --- a/fs/quota/dquot.c +++ b/fs/quota/dquot.c @@ -1398,7 +1398,7 @@ static int dquot_active(const struct inode *inode) static int __dquot_initialize(struct inode *inode, int type) { int cnt, init_needed = 0; - struct dquot **dquots, *got[MAXQUOTAS]; + struct dquot **dquots, *got[MAXQUOTAS] = {}; struct super_block *sb = inode->i_sb; qsize_t rsv; int ret = 0; @@ -1415,7 +1415,6 @@ static int __dquot_initialize(struct inode *inode, int type) int rc; struct dquot *dquot; - got[cnt] = NULL; if (type != -1 && cnt != type) continue; /* diff --git a/fs/read_write.c b/fs/read_write.c index dadf24e5c..9461d1264 100644 --- a/fs/read_write.c +++ b/fs/read_write.c @@ -534,6 +534,30 @@ ssize_t __vfs_write(struct file *file, const char __user *p, size_t count, } EXPORT_SYMBOL(__vfs_write); +vfs_readf_t vfs_readf(struct file *file) +{ + const struct file_operations *fop = file->f_op; + + if (fop->read) + return fop->read; + if (fop->read_iter) + return new_sync_read; + return ERR_PTR(-ENOSYS); +} +EXPORT_SYMBOL_GPL(vfs_readf); + +vfs_writef_t vfs_writef(struct file *file) +{ + const struct file_operations *fop = file->f_op; + + if (fop->write) + return fop->write; + if (fop->write_iter) + return new_sync_write; + return ERR_PTR(-ENOSYS); +} +EXPORT_SYMBOL_GPL(vfs_writef); + ssize_t __kernel_write(struct file *file, const char *buf, size_t count, loff_t *pos) { mm_segment_t old_fs; diff --git a/fs/splice.c b/fs/splice.c index 82bc0d64f..3d8407c0d 100644 --- a/fs/splice.c +++ b/fs/splice.c @@ -185,6 +185,9 @@ ssize_t splice_to_pipe(struct pipe_inode_info *pipe, unsigned int spd_pages = spd->nr_pages; int ret, do_wakeup, page_nr; + if (!spd_pages) + return 0; + ret = 0; do_wakeup = 0; page_nr = 0; @@ -1108,8 +1111,8 @@ EXPORT_SYMBOL(generic_splice_sendpage); /* * Attempt to initiate a splice from pipe to file. */ -static long do_splice_from(struct pipe_inode_info *pipe, struct file *out, - loff_t *ppos, size_t len, unsigned int flags) +long do_splice_from(struct pipe_inode_info *pipe, struct file *out, + loff_t *ppos, size_t len, unsigned int flags) { ssize_t (*splice_write)(struct pipe_inode_info *, struct file *, loff_t *, size_t, unsigned int); @@ -1121,13 +1124,14 @@ static long do_splice_from(struct pipe_inode_info *pipe, struct file *out, return splice_write(pipe, out, ppos, len, flags); } +EXPORT_SYMBOL_GPL(do_splice_from); /* * Attempt to initiate a splice from a file to a pipe. */ -static long do_splice_to(struct file *in, loff_t *ppos, - struct pipe_inode_info *pipe, size_t len, - unsigned int flags) +long do_splice_to(struct file *in, loff_t *ppos, + struct pipe_inode_info *pipe, size_t len, + unsigned int flags) { ssize_t (*splice_read)(struct file *, loff_t *, struct pipe_inode_info *, size_t, unsigned int); @@ -1147,6 +1151,7 @@ static long do_splice_to(struct file *in, loff_t *ppos, return splice_read(in, ppos, pipe, len, flags); } +EXPORT_SYMBOL_GPL(do_splice_to); /** * splice_direct_to_actor - splices data directly between two non-pipes diff --git a/fs/xattr.c b/fs/xattr.c index 4861322e2..c4bb039ac 100644 --- a/fs/xattr.c +++ b/fs/xattr.c @@ -207,6 +207,7 @@ vfs_getxattr_alloc(struct dentry *dentry, const char *name, char **xattr_value, *xattr_value = value; return error; } +EXPORT_SYMBOL_GPL(vfs_getxattr_alloc); ssize_t vfs_getxattr(struct dentry *dentry, const char *name, void *value, size_t size) diff --git a/fs/xfs/xfs_attr_list.c b/fs/xfs/xfs_attr_list.c index 0ef7c2ed3..4fa14820e 100644 --- a/fs/xfs/xfs_attr_list.c +++ b/fs/xfs/xfs_attr_list.c @@ -202,8 +202,10 @@ xfs_attr_shortform_list(xfs_attr_list_context_t *context) sbp->namelen, sbp->valuelen, &sbp->name[sbp->namelen]); - if (error) + if (error) { + kmem_free(sbuf); return error; + } if (context->seen_enough) break; cursor->offset++; @@ -454,14 +456,13 @@ xfs_attr3_leaf_list_int( args.rmtblkcnt = xfs_attr3_rmt_blocks( args.dp->i_mount, valuelen); retval = xfs_attr_rmtval_get(&args); - if (retval) - return retval; - retval = context->put_listent(context, - entry->flags, - name_rmt->name, - (int)name_rmt->namelen, - valuelen, - args.value); + if (!retval) + retval = context->put_listent(context, + entry->flags, + name_rmt->name, + (int)name_rmt->namelen, + valuelen, + args.value); kmem_free(args.value); } else { retval = context->put_listent(context, diff --git a/include/asm-generic/bitops/lock.h b/include/asm-generic/bitops/lock.h index c30266e94..8ef0ccbf8 100644 --- a/include/asm-generic/bitops/lock.h +++ b/include/asm-generic/bitops/lock.h @@ -29,16 +29,16 @@ do { \ * @nr: the bit to set * @addr: the address to start counting from * - * This operation is like clear_bit_unlock, however it is not atomic. - * It does provide release barrier semantics so it can be used to unlock - * a bit lock, however it would only be used if no other CPU can modify - * any bits in the memory until the lock is released (a good example is - * if the bit lock itself protects access to the other bits in the word). + * A weaker form of clear_bit_unlock() as used by __bit_lock_unlock(). If all + * the bits in the word are protected by this lock some archs can use weaker + * ops to safely unlock. + * + * See for example x86's implementation. */ #define __clear_bit_unlock(nr, addr) \ do { \ - smp_mb(); \ - __clear_bit(nr, addr); \ + smp_mb__before_atomic(); \ + clear_bit(nr, addr); \ } while (0) #endif /* _ASM_GENERIC_BITOPS_LOCK_H_ */ diff --git a/include/linux/cgroup-defs.h b/include/linux/cgroup-defs.h index 789471dba..89d944b25 100644 --- a/include/linux/cgroup-defs.h +++ b/include/linux/cgroup-defs.h @@ -210,6 +210,9 @@ struct css_set { /* all css_task_iters currently walking this cset */ struct list_head task_iters; + /* dead and being drained, ignore for migration */ + bool dead; + /* For RCU-protected deletion */ struct rcu_head rcu_head; }; diff --git a/include/linux/device-mapper.h b/include/linux/device-mapper.h index ec1c61c87..899ab9f85 100644 --- a/include/linux/device-mapper.h +++ b/include/linux/device-mapper.h @@ -124,6 +124,8 @@ struct dm_dev { char name[16]; }; +dev_t dm_get_dev_t(const char *path); + /* * Constructors should call these functions to ensure destination devices * are opened/closed correctly. diff --git a/include/linux/file.h b/include/linux/file.h index f87d30882..9a290b36b 100644 --- a/include/linux/file.h +++ b/include/linux/file.h @@ -19,6 +19,7 @@ struct dentry; struct path; extern struct file *alloc_file(struct path *, fmode_t mode, const struct file_operations *fop); +extern struct file *get_empty_filp(void); static inline void fput_light(struct file *file, int fput_needed) { diff --git a/include/linux/fs.h b/include/linux/fs.h index cf958ec45..e35963a6f 100644 --- a/include/linux/fs.h +++ b/include/linux/fs.h @@ -1260,6 +1260,7 @@ extern void fasync_free(struct fasync_struct *); /* can be called from interrupts */ extern void kill_fasync(struct fasync_struct **, int, int); +extern int setfl(int fd, struct file * filp, unsigned long arg); extern void __f_setown(struct file *filp, struct pid *, enum pid_type, int force); extern void f_setown(struct file *filp, unsigned long arg, int force); extern void f_delown(struct file *filp); @@ -1648,6 +1649,7 @@ struct file_operations { ssize_t (*sendpage) (struct file *, struct page *, int, size_t, loff_t *, int); unsigned long (*get_unmapped_area)(struct file *, unsigned long, unsigned long, unsigned long, unsigned long); int (*check_flags)(int); + int (*setfl)(struct file *, unsigned long); int (*flock) (struct file *, int, struct file_lock *); ssize_t (*splice_write)(struct pipe_inode_info *, struct file *, loff_t *, size_t, unsigned int); ssize_t (*splice_read)(struct file *, loff_t *, struct pipe_inode_info *, size_t, unsigned int); @@ -1706,6 +1708,12 @@ ssize_t rw_copy_check_uvector(int type, const struct iovec __user * uvector, struct iovec *fast_pointer, struct iovec **ret_pointer); +typedef ssize_t (*vfs_readf_t)(struct file *, char __user *, size_t, loff_t *); +typedef ssize_t (*vfs_writef_t)(struct file *, const char __user *, size_t, + loff_t *); +vfs_readf_t vfs_readf(struct file *file); +vfs_writef_t vfs_writef(struct file *file); + extern ssize_t __vfs_read(struct file *, char __user *, size_t, loff_t *); extern ssize_t __vfs_write(struct file *, const char __user *, size_t, loff_t *); extern ssize_t vfs_read(struct file *, char __user *, size_t, loff_t *); @@ -2263,7 +2271,7 @@ extern long do_sys_open(int dfd, const char __user *filename, int flags, extern struct file *file_open_name(struct filename *, int, umode_t); extern struct file *filp_open(const char *, int, umode_t); extern struct file *file_open_root(struct dentry *, struct vfsmount *, - const char *, int); + const char *, int, umode_t); extern struct file * dentry_open(const struct path *, int, const struct cred *); extern int filp_close(struct file *, fl_owner_t id); diff --git a/include/linux/init_task.h b/include/linux/init_task.h index f2cb8d455..056d3227d 100644 --- a/include/linux/init_task.h +++ b/include/linux/init_task.h @@ -157,8 +157,6 @@ extern struct task_group root_task_group; # define INIT_VTIME(tsk) #endif -#define INIT_TASK_COMM "swapper" - #ifdef CONFIG_RT_MUTEXES # define INIT_RT_MUTEXES(tsk) \ .pi_waiters = RB_ROOT, \ @@ -187,6 +185,78 @@ extern struct task_group root_task_group; * INIT_TASK is used to set up the first task table, touch at * your own risk!. Base=0, limit=0x1fffff (=2MB) */ +#ifdef CONFIG_SCHED_BFS +#define INIT_TASK_COMM "BFS" +#define INIT_TASK(tsk) \ +{ \ + .state = 0, \ + .stack = &init_thread_info, \ + .usage = ATOMIC_INIT(2), \ + .flags = PF_KTHREAD, \ + .prio = NORMAL_PRIO, \ + .static_prio = MAX_PRIO-20, \ + .normal_prio = NORMAL_PRIO, \ + .deadline = 0, \ + .policy = SCHED_NORMAL, \ + .cpus_allowed = CPU_MASK_ALL, \ + .mm = NULL, \ + .active_mm = &init_mm, \ + .restart_block = { \ + .fn = do_no_restart_syscall, \ + }, \ + .run_list = LIST_HEAD_INIT(tsk.run_list), \ + .time_slice = HZ, \ + .tasks = LIST_HEAD_INIT(tsk.tasks), \ + INIT_PUSHABLE_TASKS(tsk) \ + .ptraced = LIST_HEAD_INIT(tsk.ptraced), \ + .ptrace_entry = LIST_HEAD_INIT(tsk.ptrace_entry), \ + .real_parent = &tsk, \ + .parent = &tsk, \ + .children = LIST_HEAD_INIT(tsk.children), \ + .sibling = LIST_HEAD_INIT(tsk.sibling), \ + .group_leader = &tsk, \ + RCU_POINTER_INITIALIZER(real_cred, &init_cred), \ + RCU_POINTER_INITIALIZER(cred, &init_cred), \ + .comm = INIT_TASK_COMM, \ + .thread = INIT_THREAD, \ + .fs = &init_fs, \ + .files = &init_files, \ + .signal = &init_signals, \ + .sighand = &init_sighand, \ + .nsproxy = &init_nsproxy, \ + .pending = { \ + .list = LIST_HEAD_INIT(tsk.pending.list), \ + .signal = {{0}}}, \ + .blocked = {{0}}, \ + .alloc_lock = __SPIN_LOCK_UNLOCKED(tsk.alloc_lock), \ + .journal_info = NULL, \ + .cpu_timers = INIT_CPU_TIMERS(tsk.cpu_timers), \ + .pi_lock = __RAW_SPIN_LOCK_UNLOCKED(tsk.pi_lock), \ + .timer_slack_ns = 50000, /* 50 usec default slack */ \ + .pids = { \ + [PIDTYPE_PID] = INIT_PID_LINK(PIDTYPE_PID), \ + [PIDTYPE_PGID] = INIT_PID_LINK(PIDTYPE_PGID), \ + [PIDTYPE_SID] = INIT_PID_LINK(PIDTYPE_SID), \ + }, \ + .thread_group = LIST_HEAD_INIT(tsk.thread_group), \ + .thread_node = LIST_HEAD_INIT(init_signals.thread_head), \ + INIT_IDS \ + INIT_PERF_EVENTS(tsk) \ + INIT_TRACE_IRQFLAGS \ + INIT_LOCKDEP \ + INIT_FTRACE_GRAPH \ + INIT_TRACE_RECURSION \ + INIT_TASK_RCU_PREEMPT(tsk) \ + INIT_TASK_RCU_TASKS(tsk) \ + INIT_CPUSET_SEQ(tsk) \ + INIT_RT_MUTEXES(tsk) \ + INIT_PREV_CPUTIME(tsk) \ + INIT_VTIME(tsk) \ + INIT_NUMA_BALANCING(tsk) \ + INIT_KASAN(tsk) \ +} +#else /* CONFIG_SCHED_BFS */ +#define INIT_TASK_COMM "swapper" #define INIT_TASK(tsk) \ { \ .state = 0, \ @@ -261,7 +331,7 @@ extern struct task_group root_task_group; INIT_NUMA_BALANCING(tsk) \ INIT_KASAN(tsk) \ } - +#endif /* CONFIG_SCHED_BFS */ #define INIT_CPU_TIMERS(cpu_timers) \ { \ diff --git a/include/linux/ioprio.h b/include/linux/ioprio.h index beb9ce1c2..ce2fc3c74 100644 --- a/include/linux/ioprio.h +++ b/include/linux/ioprio.h @@ -52,6 +52,8 @@ enum { */ static inline int task_nice_ioprio(struct task_struct *task) { + if (iso_task(task)) + return 0; return (task_nice(task) + 20) / 5; } diff --git a/include/linux/jiffies.h b/include/linux/jiffies.h index 5fdc55312..9384572f6 100644 --- a/include/linux/jiffies.h +++ b/include/linux/jiffies.h @@ -164,7 +164,11 @@ static inline u64 get_jiffies_64(void) * Have the 32 bit jiffies value wrap 5 minutes after boot * so jiffies wrap bugs show up earlier. */ +#ifdef CONFIG_SCHED_BFS +#define INITIAL_JIFFIES ((unsigned long)(unsigned int) (-10*HZ)) +#else #define INITIAL_JIFFIES ((unsigned long)(unsigned int) (-300*HZ)) +#endif /* * Change timeval to jiffies, trying to avoid the diff --git a/include/linux/kernel.h b/include/linux/kernel.h index f31638c6e..95452f723 100644 --- a/include/linux/kernel.h +++ b/include/linux/kernel.h @@ -635,7 +635,7 @@ do { \ #define do_trace_printk(fmt, args...) \ do { \ - static const char *trace_printk_fmt \ + static const char *trace_printk_fmt __used \ __attribute__((section("__trace_printk_fmt"))) = \ __builtin_constant_p(fmt) ? fmt : NULL; \ \ @@ -679,7 +679,7 @@ int __trace_printk(unsigned long ip, const char *fmt, ...); */ #define trace_puts(str) ({ \ - static const char *trace_printk_fmt \ + static const char *trace_printk_fmt __used \ __attribute__((section("__trace_printk_fmt"))) = \ __builtin_constant_p(str) ? str : NULL; \ \ @@ -701,7 +701,7 @@ extern void trace_dump_stack(int skip); #define ftrace_vprintk(fmt, vargs) \ do { \ if (__builtin_constant_p(fmt)) { \ - static const char *trace_printk_fmt \ + static const char *trace_printk_fmt __used \ __attribute__((section("__trace_printk_fmt"))) = \ __builtin_constant_p(fmt) ? fmt : NULL; \ \ diff --git a/include/linux/mm.h b/include/linux/mm.h index ccc992c6a..1043097e9 100644 --- a/include/linux/mm.h +++ b/include/linux/mm.h @@ -1217,6 +1217,28 @@ static inline int fixup_user_fault(struct task_struct *tsk, } #endif +extern void vma_do_file_update_time(struct vm_area_struct *, const char[], int); +extern struct file *vma_do_pr_or_file(struct vm_area_struct *, const char[], + int); +extern void vma_do_get_file(struct vm_area_struct *, const char[], int); +extern void vma_do_fput(struct vm_area_struct *, const char[], int); + +#define vma_file_update_time(vma) vma_do_file_update_time(vma, __func__, \ + __LINE__) +#define vma_pr_or_file(vma) vma_do_pr_or_file(vma, __func__, \ + __LINE__) +#define vma_get_file(vma) vma_do_get_file(vma, __func__, __LINE__) +#define vma_fput(vma) vma_do_fput(vma, __func__, __LINE__) + +#ifndef CONFIG_MMU +extern struct file *vmr_do_pr_or_file(struct vm_region *, const char[], int); +extern void vmr_do_fput(struct vm_region *, const char[], int); + +#define vmr_pr_or_file(region) vmr_do_pr_or_file(region, __func__, \ + __LINE__) +#define vmr_fput(region) vmr_do_fput(region, __func__, __LINE__) +#endif /* !CONFIG_MMU */ + extern int access_process_vm(struct task_struct *tsk, unsigned long addr, void *buf, int len, int write); extern int access_remote_vm(struct mm_struct *mm, unsigned long addr, void *buf, int len, int write); diff --git a/include/linux/mm_types.h b/include/linux/mm_types.h index 624b78b84..1be91c585 100644 --- a/include/linux/mm_types.h +++ b/include/linux/mm_types.h @@ -269,6 +269,7 @@ struct vm_region { unsigned long vm_top; /* region allocated to here */ unsigned long vm_pgoff; /* the offset in vm_file corresponding to vm_start */ struct file *vm_file; /* the backing file or NULL */ + struct file *vm_prfile; /* the virtual backing file or NULL */ int vm_usage; /* region usage count (access under nommu_region_sem) */ bool vm_icache_flushed : 1; /* true if the icache has been flushed for @@ -343,6 +344,7 @@ struct vm_area_struct { unsigned long vm_pgoff; /* Offset (within vm_file) in PAGE_SIZE units, *not* PAGE_CACHE_SIZE */ struct file * vm_file; /* File we map to (can be NULL). */ + struct file *vm_prfile; /* shadow of vm_file */ void * vm_private_data; /* was vm_pte (shared mem) */ #ifndef CONFIG_MMU diff --git a/include/linux/pci.h b/include/linux/pci.h index 27716254d..60042ab5d 100644 --- a/include/linux/pci.h +++ b/include/linux/pci.h @@ -359,6 +359,7 @@ struct pci_dev { unsigned int io_window_1k:1; /* Intel P2P bridge 1K I/O windows */ unsigned int irq_managed:1; unsigned int has_secondary_link:1; + unsigned int non_compliant_bars:1; /* broken BARs; ignore them */ pci_dev_flags_t dev_flags; atomic_t enable_cnt; /* pci_enable_device has been called */ diff --git a/include/linux/sched.h b/include/linux/sched.h index 838a89a78..a1aec10f1 100644 --- a/include/linux/sched.h +++ b/include/linux/sched.h @@ -176,7 +176,7 @@ extern void get_iowait_load(unsigned long *nr_waiters, unsigned long *load); extern void calc_global_load(unsigned long ticks); -#if defined(CONFIG_SMP) && defined(CONFIG_NO_HZ_COMMON) +#if defined(CONFIG_SMP) && defined(CONFIG_NO_HZ_COMMON) && !defined(CONFIG_SCHED_BFS) extern void update_cpu_load_nohz(int active); #else static inline void update_cpu_load_nohz(int active) { } @@ -337,8 +337,6 @@ extern void init_idle_bootup_task(struct task_struct *idle); extern cpumask_var_t cpu_isolated_map; -extern int runqueue_is_locked(int cpu); - #if defined(CONFIG_SMP) && defined(CONFIG_NO_HZ_COMMON) extern void nohz_balance_enter_idle(int cpu); extern void set_cpu_sd_state_idle(void); @@ -1393,9 +1391,11 @@ struct task_struct { unsigned int flags; /* per process flags, defined below */ unsigned int ptrace; +#if defined(CONFIG_SMP) || defined(CONFIG_SCHED_BFS) + int on_cpu; +#endif #ifdef CONFIG_SMP struct llist_node wake_entry; - int on_cpu; unsigned int wakee_flips; unsigned long wakee_flip_decay_ts; struct task_struct *last_wakee; @@ -1403,12 +1403,29 @@ struct task_struct { int wake_cpu; #endif int on_rq; - int prio, static_prio, normal_prio; unsigned int rt_priority; +#ifdef CONFIG_SCHED_BFS + int time_slice; + u64 deadline; + struct list_head run_list; + u64 last_ran; + u64 sched_time; /* sched_clock time spent running */ +#ifdef CONFIG_SMT_NICE + int smt_bias; /* Policy/nice level bias across smt siblings */ +#endif +#ifdef CONFIG_SMP + bool sticky; /* Soft affined flag */ +#endif +#ifdef CONFIG_HOTPLUG_CPU + bool zerobound; /* Bound to CPU0 for hotplug */ +#endif + unsigned long rt_timeout; +#else /* CONFIG_SCHED_BFS */ const struct sched_class *sched_class; struct sched_entity se; struct sched_rt_entity rt; +#endif #ifdef CONFIG_CGROUP_SCHED struct task_group *sched_task_group; #endif @@ -1528,6 +1545,9 @@ struct task_struct { int __user *clear_child_tid; /* CLONE_CHILD_CLEARTID */ cputime_t utime, stime, utimescaled, stimescaled; +#ifdef CONFIG_SCHED_BFS + unsigned long utime_pc, stime_pc; +#endif cputime_t gtime; struct prev_cputime prev_cputime; #ifdef CONFIG_VIRT_CPU_ACCOUNTING_GEN @@ -1846,6 +1866,63 @@ extern int arch_task_struct_size __read_mostly; # define arch_task_struct_size (sizeof(struct task_struct)) #endif +#ifdef CONFIG_SCHED_BFS +bool grunqueue_is_locked(void); +void grq_unlock_wait(void); +void cpu_scaling(int cpu); +void cpu_nonscaling(int cpu); +#define tsk_seruntime(t) ((t)->sched_time) +#define tsk_rttimeout(t) ((t)->rt_timeout) + +static inline void tsk_cpus_current(struct task_struct *p) +{ +} + +static inline int runqueue_is_locked(int cpu) +{ + return grunqueue_is_locked(); +} + +void print_scheduler_version(void); + +static inline bool iso_task(struct task_struct *p) +{ + return (p->policy == SCHED_ISO); +} +#else /* CFS */ +extern int runqueue_is_locked(int cpu); +static inline void cpu_scaling(int cpu) +{ +} + +static inline void cpu_nonscaling(int cpu) +{ +} +#define tsk_seruntime(t) ((t)->se.sum_exec_runtime) +#define tsk_rttimeout(t) ((t)->rt.timeout) + +static inline void tsk_cpus_current(struct task_struct *p) +{ + p->nr_cpus_allowed = current->nr_cpus_allowed; +} + +static inline void print_scheduler_version(void) +{ + printk(KERN_INFO"CFS CPU scheduler.\n"); +} + +static inline bool iso_task(struct task_struct *p) +{ + return false; +} + +/* Anyone feel like implementing this? */ +static inline bool above_background_load(void) +{ + return false; +} +#endif /* CONFIG_SCHED_BFS */ + /* Future-safe accessor for struct task_struct's cpus_allowed. */ #define tsk_cpus_allowed(tsk) (&(tsk)->cpus_allowed) @@ -2261,7 +2338,7 @@ static inline int set_cpus_allowed_ptr(struct task_struct *p, } #endif -#ifdef CONFIG_NO_HZ_COMMON +#if defined(CONFIG_NO_HZ_COMMON) && !defined(CONFIG_SCHED_BFS) void calc_load_enter_idle(void); void calc_load_exit_idle(void); #else @@ -2334,7 +2411,7 @@ extern unsigned long long task_sched_runtime(struct task_struct *task); /* sched_exec is called by processes performing an exec */ -#ifdef CONFIG_SMP +#if defined(CONFIG_SMP) && !defined(CONFIG_SCHED_BFS) extern void sched_exec(void); #else #define sched_exec() {} diff --git a/include/linux/sched/prio.h b/include/linux/sched/prio.h index d9cf5a576..7d5d0b861 100644 --- a/include/linux/sched/prio.h +++ b/include/linux/sched/prio.h @@ -19,8 +19,20 @@ */ #define MAX_USER_RT_PRIO 100 + +#ifdef CONFIG_SCHED_BFS +/* Note different MAX_RT_PRIO */ +#define MAX_RT_PRIO (MAX_USER_RT_PRIO + 1) + +#define ISO_PRIO (MAX_RT_PRIO) +#define NORMAL_PRIO (MAX_RT_PRIO + 1) +#define IDLE_PRIO (MAX_RT_PRIO + 2) +#define PRIO_LIMIT ((IDLE_PRIO) + 1) +#else /* CONFIG_SCHED_BFS */ #define MAX_RT_PRIO MAX_USER_RT_PRIO +#endif /* CONFIG_SCHED_BFS */ + #define MAX_PRIO (MAX_RT_PRIO + NICE_WIDTH) #define DEFAULT_PRIO (MAX_RT_PRIO + NICE_WIDTH / 2) diff --git a/include/linux/splice.h b/include/linux/splice.h index da2751d3b..2e0fca67c 100644 --- a/include/linux/splice.h +++ b/include/linux/splice.h @@ -83,4 +83,10 @@ extern void splice_shrink_spd(struct splice_pipe_desc *); extern void spd_release_page(struct splice_pipe_desc *, unsigned int); extern const struct pipe_buf_operations page_cache_pipe_buf_ops; + +extern long do_splice_from(struct pipe_inode_info *pipe, struct file *out, + loff_t *ppos, size_t len, unsigned int flags); +extern long do_splice_to(struct file *in, loff_t *ppos, + struct pipe_inode_info *pipe, size_t len, + unsigned int flags); #endif diff --git a/include/linux/thermal.h b/include/linux/thermal.h index e13a1ace5..4a849f19e 100644 --- a/include/linux/thermal.h +++ b/include/linux/thermal.h @@ -156,6 +156,7 @@ struct thermal_attr { * @trip_hyst_attrs: attributes for trip points for sysfs: trip hysteresis * @devdata: private pointer for device private data * @trips: number of trip points the thermal zone supports + * @trips_disabled; bitmap for disabled trips * @passive_delay: number of milliseconds to wait between polls when * performing passive cooling. * @polling_delay: number of milliseconds to wait between polls when @@ -191,6 +192,7 @@ struct thermal_zone_device { struct thermal_attr *trip_hyst_attrs; void *devdata; int trips; + unsigned long trips_disabled; /* bitmap for disabled trips */ int passive_delay; int polling_delay; int temperature; diff --git a/include/linux/tty.h b/include/linux/tty.h index d9fb4b043..19199c267 100644 --- a/include/linux/tty.h +++ b/include/linux/tty.h @@ -589,7 +589,7 @@ static inline int tty_ldisc_receive_buf(struct tty_ldisc *ld, unsigned char *p, count = ld->ops->receive_buf2(ld->tty, p, f, count); else { count = min_t(int, count, ld->tty->receive_room); - if (count) + if (count && ld->ops->receive_buf) ld->ops->receive_buf(ld->tty, p, f, count); } return count; diff --git a/include/sound/hdaudio.h b/include/sound/hdaudio.h index c21c38ce7..93e63c56f 100644 --- a/include/sound/hdaudio.h +++ b/include/sound/hdaudio.h @@ -168,11 +168,13 @@ int snd_hdac_power_up(struct hdac_device *codec); int snd_hdac_power_down(struct hdac_device *codec); int snd_hdac_power_up_pm(struct hdac_device *codec); int snd_hdac_power_down_pm(struct hdac_device *codec); +int snd_hdac_keep_power_up(struct hdac_device *codec); #else static inline int snd_hdac_power_up(struct hdac_device *codec) { return 0; } static inline int snd_hdac_power_down(struct hdac_device *codec) { return 0; } static inline int snd_hdac_power_up_pm(struct hdac_device *codec) { return 0; } static inline int snd_hdac_power_down_pm(struct hdac_device *codec) { return 0; } +static inline int snd_hdac_keep_power_up(struct hdac_device *codec) { return 0; } #endif /* diff --git a/include/uapi/linux/Kbuild b/include/uapi/linux/Kbuild index ebd10e624..32152e771 100644 --- a/include/uapi/linux/Kbuild +++ b/include/uapi/linux/Kbuild @@ -59,6 +59,7 @@ header-y += atmsvc.h header-y += atm_tcp.h header-y += atm_zatm.h header-y += audit.h +header-y += aufs_type.h header-y += auto_fs4.h header-y += auto_fs.h header-y += auxvec.h diff --git a/include/uapi/linux/aufs_type.h b/include/uapi/linux/aufs_type.h new file mode 100644 index 000000000..e783d44e2 --- /dev/null +++ b/include/uapi/linux/aufs_type.h @@ -0,0 +1,406 @@ +/* + * Copyright (C) 2005-2016 Junjiro R. Okajima + */ + +#ifndef __AUFS_TYPE_H__ +#define __AUFS_TYPE_H__ + +#define AUFS_NAME "aufs" + +#ifdef __KERNEL__ +/* + * define it before including all other headers. + * sched.h may use pr_* macros before defining "current", so define the + * no-current version first, and re-define later. + */ +#define pr_fmt(fmt) AUFS_NAME " %s:%d: " fmt, __func__, __LINE__ +#include <linux/sched.h> +#undef pr_fmt +#define pr_fmt(fmt) \ + AUFS_NAME " %s:%d:%.*s[%d]: " fmt, __func__, __LINE__, \ + (int)sizeof(current->comm), current->comm, current->pid +#else +#include <stdint.h> +#include <sys/types.h> +#endif /* __KERNEL__ */ + +#include <linux/limits.h> + +#define AUFS_VERSION "4.5" + +/* todo? move this to linux-2.6.19/include/magic.h */ +#define AUFS_SUPER_MAGIC ('a' << 24 | 'u' << 16 | 'f' << 8 | 's') + +/* ---------------------------------------------------------------------- */ + +#ifdef CONFIG_AUFS_BRANCH_MAX_127 +typedef int8_t aufs_bindex_t; +#define AUFS_BRANCH_MAX 127 +#else +typedef int16_t aufs_bindex_t; +#ifdef CONFIG_AUFS_BRANCH_MAX_511 +#define AUFS_BRANCH_MAX 511 +#elif defined(CONFIG_AUFS_BRANCH_MAX_1023) +#define AUFS_BRANCH_MAX 1023 +#elif defined(CONFIG_AUFS_BRANCH_MAX_32767) +#define AUFS_BRANCH_MAX 32767 +#endif +#endif + +#ifdef __KERNEL__ +#ifndef AUFS_BRANCH_MAX +#error unknown CONFIG_AUFS_BRANCH_MAX value +#endif +#endif /* __KERNEL__ */ + +/* ---------------------------------------------------------------------- */ + +#define AUFS_FSTYPE AUFS_NAME + +#define AUFS_ROOT_INO 2 +#define AUFS_FIRST_INO 11 + +#define AUFS_WH_PFX ".wh." +#define AUFS_WH_PFX_LEN ((int)sizeof(AUFS_WH_PFX) - 1) +#define AUFS_WH_TMP_LEN 4 +/* a limit for rmdir/rename a dir and copyup */ +#define AUFS_MAX_NAMELEN (NAME_MAX \ + - AUFS_WH_PFX_LEN * 2 /* doubly whiteouted */\ + - 1 /* dot */\ + - AUFS_WH_TMP_LEN) /* hex */ +#define AUFS_XINO_FNAME "." AUFS_NAME ".xino" +#define AUFS_XINO_DEFPATH "/tmp/" AUFS_XINO_FNAME +#define AUFS_XINO_DEF_SEC 30 /* seconds */ +#define AUFS_XINO_DEF_TRUNC 45 /* percentage */ +#define AUFS_DIRWH_DEF 3 +#define AUFS_RDCACHE_DEF 10 /* seconds */ +#define AUFS_RDCACHE_MAX 3600 /* seconds */ +#define AUFS_RDBLK_DEF 512 /* bytes */ +#define AUFS_RDHASH_DEF 32 +#define AUFS_WKQ_NAME AUFS_NAME "d" +#define AUFS_MFS_DEF_SEC 30 /* seconds */ +#define AUFS_MFS_MAX_SEC 3600 /* seconds */ +#define AUFS_FHSM_CACHE_DEF_SEC 30 /* seconds */ +#define AUFS_PLINK_WARN 50 /* number of plinks in a single bucket */ + +/* pseudo-link maintenace under /proc */ +#define AUFS_PLINK_MAINT_NAME "plink_maint" +#define AUFS_PLINK_MAINT_DIR "fs/" AUFS_NAME +#define AUFS_PLINK_MAINT_PATH AUFS_PLINK_MAINT_DIR "/" AUFS_PLINK_MAINT_NAME + +#define AUFS_DIROPQ_NAME AUFS_WH_PFX ".opq" /* whiteouted doubly */ +#define AUFS_WH_DIROPQ AUFS_WH_PFX AUFS_DIROPQ_NAME + +#define AUFS_BASE_NAME AUFS_WH_PFX AUFS_NAME +#define AUFS_PLINKDIR_NAME AUFS_WH_PFX "plnk" +#define AUFS_ORPHDIR_NAME AUFS_WH_PFX "orph" + +/* doubly whiteouted */ +#define AUFS_WH_BASE AUFS_WH_PFX AUFS_BASE_NAME +#define AUFS_WH_PLINKDIR AUFS_WH_PFX AUFS_PLINKDIR_NAME +#define AUFS_WH_ORPHDIR AUFS_WH_PFX AUFS_ORPHDIR_NAME + +/* branch permissions and attributes */ +#define AUFS_BRPERM_RW "rw" +#define AUFS_BRPERM_RO "ro" +#define AUFS_BRPERM_RR "rr" +#define AUFS_BRATTR_COO_REG "coo_reg" +#define AUFS_BRATTR_COO_ALL "coo_all" +#define AUFS_BRATTR_FHSM "fhsm" +#define AUFS_BRATTR_UNPIN "unpin" +#define AUFS_BRATTR_ICEX "icex" +#define AUFS_BRATTR_ICEX_SEC "icexsec" +#define AUFS_BRATTR_ICEX_SYS "icexsys" +#define AUFS_BRATTR_ICEX_TR "icextr" +#define AUFS_BRATTR_ICEX_USR "icexusr" +#define AUFS_BRATTR_ICEX_OTH "icexoth" +#define AUFS_BRRATTR_WH "wh" +#define AUFS_BRWATTR_NLWH "nolwh" +#define AUFS_BRWATTR_MOO "moo" + +#define AuBrPerm_RW 1 /* writable, hardlinkable wh */ +#define AuBrPerm_RO (1 << 1) /* readonly */ +#define AuBrPerm_RR (1 << 2) /* natively readonly */ +#define AuBrPerm_Mask (AuBrPerm_RW | AuBrPerm_RO | AuBrPerm_RR) + +#define AuBrAttr_COO_REG (1 << 3) /* copy-up on open */ +#define AuBrAttr_COO_ALL (1 << 4) +#define AuBrAttr_COO_Mask (AuBrAttr_COO_REG | AuBrAttr_COO_ALL) + +#define AuBrAttr_FHSM (1 << 5) /* file-based hsm */ +#define AuBrAttr_UNPIN (1 << 6) /* rename-able top dir of + branch. meaningless since + linux-3.18-rc1 */ + +/* ignore error in copying XATTR */ +#define AuBrAttr_ICEX_SEC (1 << 7) +#define AuBrAttr_ICEX_SYS (1 << 8) +#define AuBrAttr_ICEX_TR (1 << 9) +#define AuBrAttr_ICEX_USR (1 << 10) +#define AuBrAttr_ICEX_OTH (1 << 11) +#define AuBrAttr_ICEX (AuBrAttr_ICEX_SEC \ + | AuBrAttr_ICEX_SYS \ + | AuBrAttr_ICEX_TR \ + | AuBrAttr_ICEX_USR \ + | AuBrAttr_ICEX_OTH) + +#define AuBrRAttr_WH (1 << 12) /* whiteout-able */ +#define AuBrRAttr_Mask AuBrRAttr_WH + +#define AuBrWAttr_NoLinkWH (1 << 13) /* un-hardlinkable whiteouts */ +#define AuBrWAttr_MOO (1 << 14) /* move-up on open */ +#define AuBrWAttr_Mask (AuBrWAttr_NoLinkWH | AuBrWAttr_MOO) + +#define AuBrAttr_CMOO_Mask (AuBrAttr_COO_Mask | AuBrWAttr_MOO) + +/* #warning test userspace */ +#ifdef __KERNEL__ +#ifndef CONFIG_AUFS_FHSM +#undef AuBrAttr_FHSM +#define AuBrAttr_FHSM 0 +#endif +#ifndef CONFIG_AUFS_XATTR +#undef AuBrAttr_ICEX +#define AuBrAttr_ICEX 0 +#undef AuBrAttr_ICEX_SEC +#define AuBrAttr_ICEX_SEC 0 +#undef AuBrAttr_ICEX_SYS +#define AuBrAttr_ICEX_SYS 0 +#undef AuBrAttr_ICEX_TR +#define AuBrAttr_ICEX_TR 0 +#undef AuBrAttr_ICEX_USR +#define AuBrAttr_ICEX_USR 0 +#undef AuBrAttr_ICEX_OTH +#define AuBrAttr_ICEX_OTH 0 +#endif +#endif + +/* the longest combination */ +/* AUFS_BRATTR_ICEX and AUFS_BRATTR_ICEX_TR don't affect here */ +#define AuBrPermStrSz sizeof(AUFS_BRPERM_RW \ + "+" AUFS_BRATTR_COO_REG \ + "+" AUFS_BRATTR_FHSM \ + "+" AUFS_BRATTR_UNPIN \ + "+" AUFS_BRATTR_ICEX_SEC \ + "+" AUFS_BRATTR_ICEX_SYS \ + "+" AUFS_BRATTR_ICEX_USR \ + "+" AUFS_BRATTR_ICEX_OTH \ + "+" AUFS_BRWATTR_NLWH) + +typedef struct { + char a[AuBrPermStrSz]; +} au_br_perm_str_t; + +static inline int au_br_writable(int brperm) +{ + return brperm & AuBrPerm_RW; +} + +static inline int au_br_whable(int brperm) +{ + return brperm & (AuBrPerm_RW | AuBrRAttr_WH); +} + +static inline int au_br_wh_linkable(int brperm) +{ + return !(brperm & AuBrWAttr_NoLinkWH); +} + +static inline int au_br_cmoo(int brperm) +{ + return brperm & AuBrAttr_CMOO_Mask; +} + +static inline int au_br_fhsm(int brperm) +{ + return brperm & AuBrAttr_FHSM; +} + +/* ---------------------------------------------------------------------- */ + +/* ioctl */ +enum { + /* readdir in userspace */ + AuCtl_RDU, + AuCtl_RDU_INO, + + AuCtl_WBR_FD, /* pathconf wrapper */ + AuCtl_IBUSY, /* busy inode */ + AuCtl_MVDOWN, /* move-down */ + AuCtl_BR, /* info about branches */ + AuCtl_FHSM_FD /* connection for fhsm */ +}; + +/* borrowed from linux/include/linux/kernel.h */ +#ifndef ALIGN +#define ALIGN(x, a) __ALIGN_MASK(x, (typeof(x))(a)-1) +#define __ALIGN_MASK(x, mask) (((x)+(mask))&~(mask)) +#endif + +/* borrowed from linux/include/linux/compiler-gcc3.h */ +#ifndef __aligned +#define __aligned(x) __attribute__((aligned(x))) +#endif + +#ifdef __KERNEL__ +#ifndef __packed +#define __packed __attribute__((packed)) +#endif +#endif + +struct au_rdu_cookie { + uint64_t h_pos; + int16_t bindex; + uint8_t flags; + uint8_t pad; + uint32_t generation; +} __aligned(8); + +struct au_rdu_ent { + uint64_t ino; + int16_t bindex; + uint8_t type; + uint8_t nlen; + uint8_t wh; + char name[0]; +} __aligned(8); + +static inline int au_rdu_len(int nlen) +{ + /* include the terminating NULL */ + return ALIGN(sizeof(struct au_rdu_ent) + nlen + 1, + sizeof(uint64_t)); +} + +union au_rdu_ent_ul { + struct au_rdu_ent __user *e; + uint64_t ul; +}; + +enum { + AufsCtlRduV_SZ, + AufsCtlRduV_End +}; + +struct aufs_rdu { + /* input */ + union { + uint64_t sz; /* AuCtl_RDU */ + uint64_t nent; /* AuCtl_RDU_INO */ + }; + union au_rdu_ent_ul ent; + uint16_t verify[AufsCtlRduV_End]; + + /* input/output */ + uint32_t blk; + + /* output */ + union au_rdu_ent_ul tail; + /* number of entries which were added in a single call */ + uint64_t rent; + uint8_t full; + uint8_t shwh; + + struct au_rdu_cookie cookie; +} __aligned(8); + +/* ---------------------------------------------------------------------- */ + +struct aufs_wbr_fd { + uint32_t oflags; + int16_t brid; +} __aligned(8); + +/* ---------------------------------------------------------------------- */ + +struct aufs_ibusy { + uint64_t ino, h_ino; + int16_t bindex; +} __aligned(8); + +/* ---------------------------------------------------------------------- */ + +/* error code for move-down */ +/* the actual message strings are implemented in aufs-util.git */ +enum { + EAU_MVDOWN_OPAQUE = 1, + EAU_MVDOWN_WHITEOUT, + EAU_MVDOWN_UPPER, + EAU_MVDOWN_BOTTOM, + EAU_MVDOWN_NOUPPER, + EAU_MVDOWN_NOLOWERBR, + EAU_Last +}; + +/* flags for move-down */ +#define AUFS_MVDOWN_DMSG 1 +#define AUFS_MVDOWN_OWLOWER (1 << 1) /* overwrite lower */ +#define AUFS_MVDOWN_KUPPER (1 << 2) /* keep upper */ +#define AUFS_MVDOWN_ROLOWER (1 << 3) /* do even if lower is RO */ +#define AUFS_MVDOWN_ROLOWER_R (1 << 4) /* did on lower RO */ +#define AUFS_MVDOWN_ROUPPER (1 << 5) /* do even if upper is RO */ +#define AUFS_MVDOWN_ROUPPER_R (1 << 6) /* did on upper RO */ +#define AUFS_MVDOWN_BRID_UPPER (1 << 7) /* upper brid */ +#define AUFS_MVDOWN_BRID_LOWER (1 << 8) /* lower brid */ +#define AUFS_MVDOWN_FHSM_LOWER (1 << 9) /* find fhsm attr for lower */ +#define AUFS_MVDOWN_STFS (1 << 10) /* req. stfs */ +#define AUFS_MVDOWN_STFS_FAILED (1 << 11) /* output: stfs is unusable */ +#define AUFS_MVDOWN_BOTTOM (1 << 12) /* output: no more lowers */ + +/* index for move-down */ +enum { + AUFS_MVDOWN_UPPER, + AUFS_MVDOWN_LOWER, + AUFS_MVDOWN_NARRAY +}; + +/* + * additional info of move-down + * number of free blocks and inodes. + * subset of struct kstatfs, but smaller and always 64bit. + */ +struct aufs_stfs { + uint64_t f_blocks; + uint64_t f_bavail; + uint64_t f_files; + uint64_t f_ffree; +}; + +struct aufs_stbr { + int16_t brid; /* optional input */ + int16_t bindex; /* output */ + struct aufs_stfs stfs; /* output when AUFS_MVDOWN_STFS set */ +} __aligned(8); + +struct aufs_mvdown { + uint32_t flags; /* input/output */ + struct aufs_stbr stbr[AUFS_MVDOWN_NARRAY]; /* input/output */ + int8_t au_errno; /* output */ +} __aligned(8); + +/* ---------------------------------------------------------------------- */ + +union aufs_brinfo { + /* PATH_MAX may differ between kernel-space and user-space */ + char _spacer[4096]; + struct { + int16_t id; + int perm; + char path[0]; + }; +} __aligned(8); + +/* ---------------------------------------------------------------------- */ + +#define AuCtlType 'A' +#define AUFS_CTL_RDU _IOWR(AuCtlType, AuCtl_RDU, struct aufs_rdu) +#define AUFS_CTL_RDU_INO _IOWR(AuCtlType, AuCtl_RDU_INO, struct aufs_rdu) +#define AUFS_CTL_WBR_FD _IOW(AuCtlType, AuCtl_WBR_FD, \ + struct aufs_wbr_fd) +#define AUFS_CTL_IBUSY _IOWR(AuCtlType, AuCtl_IBUSY, struct aufs_ibusy) +#define AUFS_CTL_MVDOWN _IOWR(AuCtlType, AuCtl_MVDOWN, \ + struct aufs_mvdown) +#define AUFS_CTL_BRINFO _IOW(AuCtlType, AuCtl_BR, union aufs_brinfo) +#define AUFS_CTL_FHSM_FD _IOW(AuCtlType, AuCtl_FHSM_FD, int) + +#endif /* __AUFS_TYPE_H__ */ diff --git a/include/uapi/linux/sched.h b/include/uapi/linux/sched.h index cc89ddefa..f63e1cd82 100644 --- a/include/uapi/linux/sched.h +++ b/include/uapi/linux/sched.h @@ -37,9 +37,16 @@ #define SCHED_FIFO 1 #define SCHED_RR 2 #define SCHED_BATCH 3 -/* SCHED_ISO: reserved but not implemented yet */ +/* SCHED_ISO: Implemented on BFS only */ #define SCHED_IDLE 5 +#ifdef CONFIG_SCHED_BFS +#define SCHED_ISO 4 +#define SCHED_IDLEPRIO SCHED_IDLE +#define SCHED_MAX (SCHED_IDLEPRIO) +#define SCHED_RANGE(policy) ((policy) <= SCHED_MAX) +#else /* CONFIG_SCHED_BFS */ #define SCHED_DEADLINE 6 +#endif /* CONFIG_SCHED_BFS */ /* Can be ORed in to make sure the process is reverted back to SCHED_NORMAL on fork */ #define SCHED_RESET_ON_FORK 0x40000000 diff --git a/init/Kconfig b/init/Kconfig index ec251fadb..0d6bb5cf4 100644 --- a/init/Kconfig +++ b/init/Kconfig @@ -44,7 +44,7 @@ config PCK_INTERACTIVE NCQ Queue Depth................: 31 -> 8 Block Layer Queue Depth........: 128 -> 16 - --- CPU Scheduler -------------------------------------- + --- CPU Scheduler (CFS) -------------------------------- Scheduling latency.............: 6 -> 3 ms Minimal granularity............: 0.75 -> 0.3 ms @@ -52,10 +52,38 @@ config PCK_INTERACTIVE CPU migration cost.............: 0.5 -> 0.25 ms Bandwidth slice size...........: 5 -> 3 ms + --- CPU Scheduler (BFS) -------------------------------- + + Scheduling interval............: 6 -> 3 ms + ISO task max realtime use......: 70 % -> 25 % + --- CPU Frequency Scaling ------------------------------ Ondemand down scaling factor...: 1 -> 10 +config SCHED_BFS + bool "BFS cpu scheduler" + default n + help + The Brain Fuck CPU Scheduler for excellent interactivity and + responsiveness on the desktop and solid scalability on normal + hardware and commodity servers. Not recommended for 4096 CPUs. + + Currently incompatible with the Group CPU scheduler, and RCU TORTURE + TEST so these options are disabled. + +config SCHED_BFS_AUTOISO + bool "Automatically use SCHED_ISO policy for X" + depends on SCHED_BFS + default n + help + Selecting this option will automatically use the SCHED_ISO scheduling + policy for X, resulting in an interactivity boost. This *may* cause + things like skipping sound on audio applications that are not run + as SCHED_ISO. + + Tasks (including X) can be run as sched_iso manually using schedtool. + config BROKEN bool @@ -363,7 +391,7 @@ choice # Kind of a stub config for the pure tick based cputime accounting config TICK_CPU_ACCOUNTING bool "Simple tick based cputime accounting" - depends on !S390 && !NO_HZ_FULL + depends on !S390 && !NO_HZ_FULL && !SCHED_BFS help This is the basic tick based cputime accounting that maintains statistics about user, system and idle time spent on per jiffies @@ -388,6 +416,7 @@ config VIRT_CPU_ACCOUNTING_GEN bool "Full dynticks CPU time accounting" depends on HAVE_CONTEXT_TRACKING depends on HAVE_VIRT_CPU_ACCOUNTING_GEN + depends on !SCHED_BFS select VIRT_CPU_ACCOUNTING select CONTEXT_TRACKING help @@ -722,6 +751,7 @@ config RCU_NOCB_CPU bool "Offload RCU callback processing from boot-selected CPUs" depends on TREE_RCU || PREEMPT_RCU depends on RCU_EXPERT || NO_HZ_FULL + depends on !SCHED_BFS default n help Use this option to reduce OS jitter for aggressive HPC or @@ -932,6 +962,7 @@ config NUMA_BALANCING depends on ARCH_SUPPORTS_NUMA_BALANCING depends on !ARCH_WANT_NUMA_VARIABLE_LOCALITY depends on SMP && NUMA && MIGRATION + depends on !SCHED_BFS help This option adds support for automatic NUMA aware memory/task placement. The mechanism is quite primitive and is based on migrating memory when @@ -1030,6 +1061,7 @@ config CGROUP_WRITEBACK menuconfig CGROUP_SCHED bool "CPU controller" + depends on !SCHED_BFS default n help This feature lets CPU scheduler recognize task groups and control CPU @@ -1132,6 +1164,7 @@ config CGROUP_DEVICE config CGROUP_CPUACCT bool "Simple CPU accounting controller" + depends on !SCHED_BFS help Provides a simple controller for monitoring the total CPU consumed by the tasks in a cgroup. @@ -1230,6 +1263,7 @@ endif # NAMESPACES config SCHED_AUTOGROUP bool "Automatic process group scheduling" + depends on !SCHED_BFS select CGROUPS select CGROUP_SCHED select FAIR_GROUP_SCHED @@ -1744,6 +1778,7 @@ choice This option allows to select a slab allocator. config SLAB + depends on !SCHED_BFS bool "SLAB" help The regular slab allocator that is established and known to work @@ -1761,7 +1796,7 @@ config SLUB a slab allocator. config SLOB - depends on EXPERT + depends on EXPERT && !SCHED_BFS bool "SLOB (Simple Allocator)" help SLOB replaces the stock allocator with a drastically simpler diff --git a/init/main.c b/init/main.c index 58c9e3747..055106098 100644 --- a/init/main.c +++ b/init/main.c @@ -808,7 +808,6 @@ int __init_or_module do_one_initcall(initcall_t fn) return ret; } - extern initcall_t __initcall_start[]; extern initcall_t __initcall0_start[]; extern initcall_t __initcall1_start[]; @@ -945,6 +944,8 @@ static int __ref kernel_init(void *unused) rcu_end_inkernel_boot(); + print_scheduler_version(); + if (ramdisk_execute_command) { ret = run_init_process(ramdisk_execute_command); if (!ret) diff --git a/kernel/cgroup.c b/kernel/cgroup.c index d27904c19..6a498daf2 100644 --- a/kernel/cgroup.c +++ b/kernel/cgroup.c @@ -2474,6 +2474,14 @@ static void cgroup_migrate_add_src(struct css_set *src_cset, lockdep_assert_held(&cgroup_mutex); lockdep_assert_held(&css_set_lock); + /* + * If ->dead, @src_set is associated with one or more dead cgroups + * and doesn't contain any migratable tasks. Ignore it early so + * that the rest of migration path doesn't get confused by it. + */ + if (src_cset->dead) + return; + src_cgrp = cset_cgroup_from_root(src_cset, dst_cgrp->root); if (!list_empty(&src_cset->mg_preload_node)) @@ -5114,6 +5122,7 @@ static int cgroup_destroy_locked(struct cgroup *cgrp) __releases(&cgroup_mutex) __acquires(&cgroup_mutex) { struct cgroup_subsys_state *css; + struct cgrp_cset_link *link; int ssid; lockdep_assert_held(&cgroup_mutex); @@ -5134,11 +5143,18 @@ static int cgroup_destroy_locked(struct cgroup *cgrp) return -EBUSY; /* - * Mark @cgrp dead. This prevents further task migration and child - * creation by disabling cgroup_lock_live_group(). + * Mark @cgrp and the associated csets dead. The former prevents + * further task migration and child creation by disabling + * cgroup_lock_live_group(). The latter makes the csets ignored by + * the migration path. */ cgrp->self.flags &= ~CSS_ONLINE; + spin_lock_bh(&css_set_lock); + list_for_each_entry(link, &cgrp->cset_links, cset_link) + link->cset->dead = true; + spin_unlock_bh(&css_set_lock); + /* initiate massacre of all css's */ for_each_css(css, ssid, cgrp) kill_css(css); diff --git a/kernel/delayacct.c b/kernel/delayacct.c index 435c14a45..a80d56dc7 100644 --- a/kernel/delayacct.c +++ b/kernel/delayacct.c @@ -104,7 +104,7 @@ int __delayacct_add_tsk(struct taskstats *d, struct task_struct *tsk) */ t1 = tsk->sched_info.pcount; t2 = tsk->sched_info.run_delay; - t3 = tsk->se.sum_exec_runtime; + t3 = tsk_seruntime(tsk); d->cpu_count += t1; diff --git a/kernel/events/core.c b/kernel/events/core.c index 614614821..f0b4b328d 100644 --- a/kernel/events/core.c +++ b/kernel/events/core.c @@ -8001,6 +8001,9 @@ perf_event_alloc(struct perf_event_attr *attr, int cpu, } } + /* symmetric to unaccount_event() in _free_event() */ + account_event(event); + return event; err_per_task: @@ -8364,8 +8367,6 @@ SYSCALL_DEFINE5(perf_event_open, } } - account_event(event); - /* * Special case software events and allow them to be part of * any hardware group. @@ -8662,8 +8663,6 @@ perf_event_create_kernel_counter(struct perf_event_attr *attr, int cpu, /* Mark owner so we could distinguish it from user events. */ event->owner = TASK_TOMBSTONE; - account_event(event); - ctx = find_get_context(event->pmu, task, event); if (IS_ERR(ctx)) { err = PTR_ERR(ctx); diff --git a/kernel/exit.c b/kernel/exit.c index 10e088237..ac8f9b7d6 100644 --- a/kernel/exit.c +++ b/kernel/exit.c @@ -133,7 +133,7 @@ static void __exit_signal(struct task_struct *tsk) sig->inblock += task_io_get_inblock(tsk); sig->oublock += task_io_get_oublock(tsk); task_io_accounting_add(&sig->ioac, &tsk->ioac); - sig->sum_sched_runtime += tsk->se.sum_exec_runtime; + sig->sum_sched_runtime += tsk_seruntime(tsk); sig->nr_threads--; __unhash_process(tsk, group_dead); write_sequnlock(&sig->stats_lock); diff --git a/kernel/fork.c b/kernel/fork.c index bb061d0ea..8fbec37c5 100644 --- a/kernel/fork.c +++ b/kernel/fork.c @@ -464,7 +464,7 @@ static int dup_mmap(struct mm_struct *mm, struct mm_struct *oldmm) struct inode *inode = file_inode(file); struct address_space *mapping = file->f_mapping; - get_file(file); + vma_get_file(tmp); if (tmp->vm_flags & VM_DENYWRITE) atomic_dec(&inode->i_writecount); i_mmap_lock_write(mapping); diff --git a/kernel/power/hibernate.c b/kernel/power/hibernate.c index 153e51db5..ba2dda520 100644 --- a/kernel/power/hibernate.c +++ b/kernel/power/hibernate.c @@ -339,6 +339,7 @@ int hibernation_snapshot(int platform_mode) pm_message_t msg; int error; + pm_suspend_clear_flags(); error = platform_begin(platform_mode); if (error) goto Close; diff --git a/kernel/sched/Makefile b/kernel/sched/Makefile index 7d4cba227..15e398808 100644 --- a/kernel/sched/Makefile +++ b/kernel/sched/Makefile @@ -11,11 +11,16 @@ ifneq ($(CONFIG_SCHED_OMIT_FRAME_POINTER),y) CFLAGS_core.o := $(PROFILING) -fno-omit-frame-pointer endif +ifdef CONFIG_SCHED_BFS +obj-y += bfs.o clock.o +else obj-y += core.o loadavg.o clock.o cputime.o obj-y += idle_task.o fair.o rt.o deadline.o stop_task.o -obj-y += wait.o swait.o completion.o idle.o -obj-$(CONFIG_SMP) += cpupri.o cpudeadline.o +obj-$(CONFIG_SMP) += cpudeadline.o obj-$(CONFIG_SCHED_AUTOGROUP) += auto_group.o -obj-$(CONFIG_SCHEDSTATS) += stats.o obj-$(CONFIG_SCHED_DEBUG) += debug.o obj-$(CONFIG_CGROUP_CPUACCT) += cpuacct.o +endif +obj-y += wait.o swait.o completion.o idle.o +obj-$(CONFIG_SMP) += cpupri.o +obj-$(CONFIG_SCHEDSTATS) += stats.o diff --git a/kernel/sched/bfs.c b/kernel/sched/bfs.c new file mode 100644 index 000000000..698007a9f --- /dev/null +++ b/kernel/sched/bfs.c @@ -0,0 +1,7609 @@ +/* + * kernel/sched/bfs.c, was kernel/sched.c + * + * Kernel scheduler and related syscalls + * + * Copyright (C) 1991-2002 Linus Torvalds + * + * 1996-12-23 Modified by Dave Grothe to fix bugs in semaphores and + * make semaphores SMP safe + * 1998-11-19 Implemented schedule_timeout() and related stuff + * by Andrea Arcangeli + * 2002-01-04 New ultra-scalable O(1) scheduler by Ingo Molnar: + * hybrid priority-list and round-robin design with + * an array-switch method of distributing timeslices + * and per-CPU runqueues. Cleanups and useful suggestions + * by Davide Libenzi, preemptible kernel bits by Robert Love. + * 2003-09-03 Interactivity tuning by Con Kolivas. + * 2004-04-02 Scheduler domains code by Nick Piggin + * 2007-04-15 Work begun on replacing all interactivity tuning with a + * fair scheduling design by Con Kolivas. + * 2007-05-05 Load balancing (smp-nice) and other improvements + * by Peter Williams + * 2007-05-06 Interactivity improvements to CFS by Mike Galbraith + * 2007-07-01 Group scheduling enhancements by Srivatsa Vaddagiri + * 2007-11-29 RT balancing improvements by Steven Rostedt, Gregory Haskins, + * Thomas Gleixner, Mike Kravetz + * now Brainfuck deadline scheduling policy by Con Kolivas deletes + * a whole lot of those previous things. + */ + +#include <linux/kasan.h> +#include <linux/mm.h> +#include <linux/module.h> +#include <linux/nmi.h> +#include <linux/init.h> +#include <asm/uaccess.h> +#include <linux/highmem.h> +#include <asm/mmu_context.h> +#include <linux/interrupt.h> +#include <linux/capability.h> +#include <linux/completion.h> +#include <linux/kernel_stat.h> +#include <linux/debug_locks.h> +#include <linux/perf_event.h> +#include <linux/security.h> +#include <linux/notifier.h> +#include <linux/profile.h> +#include <linux/freezer.h> +#include <linux/vmalloc.h> +#include <linux/blkdev.h> +#include <linux/delay.h> +#include <linux/smp.h> +#include <linux/threads.h> +#include <linux/timer.h> +#include <linux/rcupdate.h> +#include <linux/cpu.h> +#include <linux/cpuset.h> +#include <linux/cpumask.h> +#include <linux/percpu.h> +#include <linux/proc_fs.h> +#include <linux/seq_file.h> +#include <linux/syscalls.h> +#include <linux/sched/sysctl.h> +#include <linux/times.h> +#include <linux/tsacct_kern.h> +#include <linux/kprobes.h> +#include <linux/delayacct.h> +#include <linux/log2.h> +#include <linux/bootmem.h> +#include <linux/ftrace.h> +#include <linux/slab.h> +#include <linux/init_task.h> +#include <linux/binfmts.h> +#include <linux/context_tracking.h> +#include <linux/sched/prio.h> +#include <linux/tick.h> + +#include <asm/irq_regs.h> +#include <asm/switch_to.h> +#include <asm/tlb.h> +#include <asm/unistd.h> +#include <asm/mutex.h> +#ifdef CONFIG_PARAVIRT +#include <asm/paravirt.h> +#endif + +#include "cpupri.h" +#include "../workqueue_internal.h" +#include "../smpboot.h" + +#define CREATE_TRACE_POINTS +#include <trace/events/sched.h> + +#include "bfs_sched.h" + +#define rt_prio(prio) unlikely((prio) < MAX_RT_PRIO) +#define rt_task(p) rt_prio((p)->prio) +#define rt_queue(rq) rt_prio((rq)->rq_prio) +#define batch_task(p) (unlikely((p)->policy == SCHED_BATCH)) +#define is_rt_policy(policy) ((policy) == SCHED_FIFO || \ + (policy) == SCHED_RR) +#define has_rt_policy(p) unlikely(is_rt_policy((p)->policy)) + +#define is_idle_policy(policy) ((policy) == SCHED_IDLEPRIO) +#define idleprio_task(p) unlikely(is_idle_policy((p)->policy)) +#define task_running_idle(p) unlikely((p)->prio == IDLE_PRIO) +#define idle_queue(rq) (unlikely(is_idle_policy((rq)->rq_policy))) + +#define is_iso_policy(policy) ((policy) == SCHED_ISO) +#define iso_task(p) unlikely(is_iso_policy((p)->policy)) +#define iso_queue(rq) unlikely(is_iso_policy((rq)->rq_policy)) +#define task_running_iso(p) unlikely((p)->prio == ISO_PRIO) +#define rq_running_iso(rq) ((rq)->rq_prio == ISO_PRIO) + +#define rq_idle(rq) ((rq)->rq_prio == PRIO_LIMIT) + +#define ISO_PERIOD ((5 * HZ * grq.noc) + 1) + +#define SCHED_PRIO(p) ((p) + MAX_RT_PRIO) +#define STOP_PRIO (MAX_RT_PRIO - 1) + +/* + * Some helpers for converting to/from various scales. Use shifts to get + * approximate multiples of ten for less overhead. + */ +#define JIFFIES_TO_NS(TIME) ((TIME) * (1000000000 / HZ)) +#define JIFFY_NS (1000000000 / HZ) +#define HALF_JIFFY_NS (1000000000 / HZ / 2) +#define HALF_JIFFY_US (1000000 / HZ / 2) +#define MS_TO_NS(TIME) ((TIME) << 20) +#define MS_TO_US(TIME) ((TIME) << 10) +#define NS_TO_MS(TIME) ((TIME) >> 20) +#define NS_TO_US(TIME) ((TIME) >> 10) + +#define RESCHED_US (100) /* Reschedule if less than this many μs left */ + +void print_scheduler_version(void) +{ + printk(KERN_INFO "BFS CPU scheduler v0.469 by Con Kolivas.\n"); +} + +/* + * This is the time all tasks within the same priority round robin. + * Value is in ms and set to a minimum of 6ms. Scales with number of cpus. + * Tunable via /proc interface. + */ +#ifdef CONFIG_PCK_INTERACTIVE +int rr_interval __read_mostly = 3; +#else +int rr_interval __read_mostly = 6; +#endif + +/* Tunable to choose whether to prioritise latency or throughput, simple + * binary yes or no */ + +int sched_interactive __read_mostly = 1; + +/* + * sched_iso_cpu - sysctl which determines the cpu percentage SCHED_ISO tasks + * are allowed to run five seconds as real time tasks. This is the total over + * all online cpus. + */ +#ifdef CONFIG_PCK_INTERACTIVE +int sched_iso_cpu __read_mostly = 25; +#else +int sched_iso_cpu __read_mostly = 70; +#endif + +/* + * The relative length of deadline for each priority(nice) level. + */ +static int prio_ratios[NICE_WIDTH] __read_mostly; + +/* + * The quota handed out to tasks of all priority levels when refilling their + * time_slice. + */ +static inline int timeslice(void) +{ + return MS_TO_US(rr_interval); +} + +/* + * The global runqueue data that all CPUs work off. Data is protected either + * by the global grq lock, or the discrete lock that precedes the data in this + * struct. + */ +struct global_rq { + raw_spinlock_t lock; + unsigned long nr_running; + unsigned long nr_uninterruptible; + unsigned long long nr_switches; + struct list_head queue[PRIO_LIMIT]; + DECLARE_BITMAP(prio_bitmap, PRIO_LIMIT + 1); + unsigned long qnr; /* queued not running */ +#ifdef CONFIG_SMP + cpumask_t cpu_idle_map; + bool idle_cpus; +#endif + int noc; /* num_online_cpus stored and updated when it changes */ + u64 niffies; /* Nanosecond jiffies */ + unsigned long last_jiffy; /* Last jiffy we updated niffies */ + + raw_spinlock_t iso_lock; + int iso_ticks; + bool iso_refractory; +}; + +#ifdef CONFIG_SMP +/* + * We add the notion of a root-domain which will be used to define per-domain + * variables. Each exclusive cpuset essentially defines an island domain by + * fully partitioning the member cpus from any other cpuset. Whenever a new + * exclusive cpuset is created, we also create and attach a new root-domain + * object. + * + */ +struct root_domain { + atomic_t refcount; + atomic_t rto_count; + struct rcu_head rcu; + cpumask_var_t span; + cpumask_var_t online; + + /* + * The "RT overload" flag: it gets set if a CPU has more than + * one runnable RT task. + */ + cpumask_var_t rto_mask; + struct cpupri cpupri; +}; + +/* + * By default the system creates a single root-domain with all cpus as + * members (mimicking the global state we have today). + */ +static struct root_domain def_root_domain; + +#endif /* CONFIG_SMP */ + +/* There can be only one */ +static struct global_rq grq; + +static DEFINE_MUTEX(sched_hotcpu_mutex); + +/* cpus with isolated domains */ +cpumask_var_t cpu_isolated_map; + +DEFINE_PER_CPU_SHARED_ALIGNED(struct rq, runqueues); +#ifdef CONFIG_SMP +struct rq *cpu_rq(int cpu) +{ + return &per_cpu(runqueues, (cpu)); +} +#define task_rq(p) cpu_rq(task_cpu(p)) +#define cpu_curr(cpu) (cpu_rq(cpu)->curr) +/* + * sched_domains_mutex serialises calls to init_sched_domains, + * detach_destroy_domains and partition_sched_domains. + */ +DEFINE_MUTEX(sched_domains_mutex); + +/* + * By default the system creates a single root-domain with all cpus as + * members (mimicking the global state we have today). + */ +static struct root_domain def_root_domain; + +int __weak arch_sd_sibling_asym_packing(void) +{ + return 0*SD_ASYM_PACKING; +} +#else +struct rq *uprq; +#endif /* CONFIG_SMP */ + +static inline void update_rq_clock(struct rq *rq); + +/* + * Sanity check should sched_clock return bogus values. We make sure it does + * not appear to go backwards, and use jiffies to determine the maximum and + * minimum it could possibly have increased, and round down to the nearest + * jiffy when it falls outside this. + */ +static inline void niffy_diff(s64 *niff_diff, int jiff_diff) +{ + unsigned long min_diff, max_diff; + + if (jiff_diff > 1) + min_diff = JIFFIES_TO_NS(jiff_diff - 1); + else + min_diff = 1; + /* Round up to the nearest tick for maximum */ + max_diff = JIFFIES_TO_NS(jiff_diff + 1); + + if (unlikely(*niff_diff < min_diff || *niff_diff > max_diff)) + *niff_diff = min_diff; +} + +#ifdef CONFIG_SMP +static inline int cpu_of(struct rq *rq) +{ + return rq->cpu; +} + +/* + * Niffies are a globally increasing nanosecond counter. Whenever a runqueue + * clock is updated with the grq.lock held, it is an opportunity to update the + * niffies value. Any CPU can update it by adding how much its clock has + * increased since it last updated niffies, minus any added niffies by other + * CPUs. + */ +static inline void update_clocks(struct rq *rq) +{ + s64 ndiff; + long jdiff; + + update_rq_clock(rq); + ndiff = rq->clock - rq->old_clock; + /* old_clock is only updated when we are updating niffies */ + rq->old_clock = rq->clock; + ndiff -= grq.niffies - rq->last_niffy; + jdiff = jiffies - grq.last_jiffy; + niffy_diff(&ndiff, jdiff); + grq.last_jiffy += jdiff; + grq.niffies += ndiff; + rq->last_niffy = grq.niffies; +} +#else /* CONFIG_SMP */ +static inline int cpu_of(struct rq *rq) +{ + return 0; +} + +static inline void update_clocks(struct rq *rq) +{ + s64 ndiff; + long jdiff; + + update_rq_clock(rq); + ndiff = rq->clock - rq->old_clock; + rq->old_clock = rq->clock; + jdiff = jiffies - grq.last_jiffy; + niffy_diff(&ndiff, jdiff); + grq.last_jiffy += jdiff; + grq.niffies += ndiff; +} +#endif + +#include "stats.h" + +#ifndef prepare_arch_switch +# define prepare_arch_switch(next) do { } while (0) +#endif +#ifndef finish_arch_switch +# define finish_arch_switch(prev) do { } while (0) +#endif +#ifndef finish_arch_post_lock_switch +# define finish_arch_post_lock_switch() do { } while (0) +#endif + +/* + * All common locking functions performed on grq.lock. rq->clock is local to + * the CPU accessing it so it can be modified just with interrupts disabled + * when we're not updating niffies. + * Looking up task_rq must be done under grq.lock to be safe. + */ +static void update_rq_clock_task(struct rq *rq, s64 delta); + +static inline void update_rq_clock(struct rq *rq) +{ + s64 delta = sched_clock_cpu(cpu_of(rq)) - rq->clock; + + if (unlikely(delta < 0)) + return; + rq->clock += delta; + update_rq_clock_task(rq, delta); +} + +static inline bool task_running(struct task_struct *p) +{ + return p->on_cpu; +} + +static inline void grq_lock(void) + __acquires(grq.lock) +{ + raw_spin_lock(&grq.lock); +} + +static inline void grq_unlock(void) + __releases(grq.lock) +{ + raw_spin_unlock(&grq.lock); +} + +static inline void grq_lock_irq(void) + __acquires(grq.lock) +{ + raw_spin_lock_irq(&grq.lock); +} + +static inline void time_lock_grq(struct rq *rq) + __acquires(grq.lock) +{ + grq_lock(); + update_clocks(rq); +} + +static inline void grq_unlock_irq(void) + __releases(grq.lock) +{ + raw_spin_unlock_irq(&grq.lock); +} + +static inline void grq_lock_irqsave(unsigned long *flags) + __acquires(grq.lock) +{ + raw_spin_lock_irqsave(&grq.lock, *flags); +} + +static inline void grq_unlock_irqrestore(unsigned long *flags) + __releases(grq.lock) +{ + raw_spin_unlock_irqrestore(&grq.lock, *flags); +} + +static inline struct rq +*task_grq_lock(struct task_struct *p, unsigned long *flags) + __acquires(grq.lock) +{ + grq_lock_irqsave(flags); + return task_rq(p); +} + +static inline struct rq +*time_task_grq_lock(struct task_struct *p, unsigned long *flags) + __acquires(grq.lock) +{ + struct rq *rq = task_grq_lock(p, flags); + update_clocks(rq); + return rq; +} + +static inline struct rq *task_grq_lock_irq(struct task_struct *p) + __acquires(grq.lock) +{ + grq_lock_irq(); + return task_rq(p); +} + +static inline void time_task_grq_lock_irq(struct task_struct *p) + __acquires(grq.lock) +{ + struct rq *rq = task_grq_lock_irq(p); + update_clocks(rq); +} + +static inline void task_grq_unlock_irq(void) + __releases(grq.lock) +{ + grq_unlock_irq(); +} + +static inline void task_grq_unlock(unsigned long *flags) + __releases(grq.lock) +{ + grq_unlock_irqrestore(flags); +} + +/** + * grunqueue_is_locked + * + * Returns true if the global runqueue is locked. + * This interface allows printk to be called with the runqueue lock + * held and know whether or not it is OK to wake up the klogd. + */ +bool grunqueue_is_locked(void) +{ + return raw_spin_is_locked(&grq.lock); +} + +void grq_unlock_wait(void) + __releases(grq.lock) +{ + smp_mb(); /* spin-unlock-wait is not a full memory barrier */ + raw_spin_unlock_wait(&grq.lock); +} + +static inline void time_grq_lock(struct rq *rq, unsigned long *flags) + __acquires(grq.lock) +{ + local_irq_save(*flags); + time_lock_grq(rq); +} + +static inline struct rq *__task_grq_lock(struct task_struct *p) + __acquires(grq.lock) +{ + grq_lock(); + return task_rq(p); +} + +static inline void __task_grq_unlock(void) + __releases(grq.lock) +{ + grq_unlock(); +} + +static inline void prepare_lock_switch(struct rq *rq, struct task_struct *next) +{ +} + +static inline void finish_lock_switch(struct rq *rq, struct task_struct *prev) +{ +#ifdef CONFIG_DEBUG_SPINLOCK + /* this is a valid case when another task releases the spinlock */ + grq.lock.owner = current; +#endif + /* + * If we are tracking spinlock dependencies then we have to + * fix up the runqueue lock - which gets 'carried over' from + * prev into current: + */ + spin_acquire(&grq.lock.dep_map, 0, 0, _THIS_IP_); + + grq_unlock_irq(); +} + +static inline bool deadline_before(u64 deadline, u64 time) +{ + return (deadline < time); +} + +static inline bool deadline_after(u64 deadline, u64 time) +{ + return (deadline > time); +} + +/* + * A task that is queued but not running will be on the grq run list. + * A task that is not running or queued will not be on the grq run list. + * A task that is currently running will have ->on_cpu set but not on the + * grq run list. + */ +static inline bool task_queued(struct task_struct *p) +{ + return (!list_empty(&p->run_list)); +} + +/* + * Removing from the global runqueue. Enter with grq locked. + */ +static void dequeue_task(struct task_struct *p) +{ + list_del_init(&p->run_list); + if (list_empty(grq.queue + p->prio)) + __clear_bit(p->prio, grq.prio_bitmap); + sched_info_dequeued(task_rq(p), p); +} + +/* + * To determine if it's safe for a task of SCHED_IDLEPRIO to actually run as + * an idle task, we ensure none of the following conditions are met. + */ +static bool idleprio_suitable(struct task_struct *p) +{ + return (!freezing(p) && !signal_pending(p) && + !(task_contributes_to_load(p)) && !(p->flags & (PF_EXITING))); +} + +/* + * To determine if a task of SCHED_ISO can run in pseudo-realtime, we check + * that the iso_refractory flag is not set. + */ +static bool isoprio_suitable(void) +{ + return !grq.iso_refractory; +} + +/* + * Adding to the global runqueue. Enter with grq locked. + */ +static void enqueue_task(struct task_struct *p, struct rq *rq) +{ + if (!rt_task(p)) { + /* Check it hasn't gotten rt from PI */ + if ((idleprio_task(p) && idleprio_suitable(p)) || + (iso_task(p) && isoprio_suitable())) + p->prio = p->normal_prio; + else + p->prio = NORMAL_PRIO; + } + __set_bit(p->prio, grq.prio_bitmap); + list_add_tail(&p->run_list, grq.queue + p->prio); + sched_info_queued(rq, p); +} + +static inline void requeue_task(struct task_struct *p) +{ + sched_info_queued(task_rq(p), p); +} + +/* + * Returns the relative length of deadline all compared to the shortest + * deadline which is that of nice -20. + */ +static inline int task_prio_ratio(struct task_struct *p) +{ + return prio_ratios[TASK_USER_PRIO(p)]; +} + +/* + * task_timeslice - all tasks of all priorities get the exact same timeslice + * length. CPU distribution is handled by giving different deadlines to + * tasks of different priorities. Use 128 as the base value for fast shifts. + */ +static inline int task_timeslice(struct task_struct *p) +{ + return (rr_interval * task_prio_ratio(p) / 128); +} + +static void resched_task(struct task_struct *p); + +static inline void resched_curr(struct rq *rq) +{ + resched_task(rq->curr); +} + +/* + * qnr is the "queued but not running" count which is the total number of + * tasks on the global runqueue list waiting for cpu time but not actually + * currently running on a cpu. + */ +static inline void inc_qnr(void) +{ + grq.qnr++; +} + +static inline void dec_qnr(void) +{ + grq.qnr--; +} + +static inline int queued_notrunning(void) +{ + return grq.qnr; +} + +#ifdef CONFIG_SMP +/* + * The cpu_idle_map stores a bitmap of all the CPUs currently idle to + * allow easy lookup of whether any suitable idle CPUs are available. + * It's cheaper to maintain a binary yes/no if there are any idle CPUs on the + * idle_cpus variable than to do a full bitmask check when we are busy. + */ +static inline void set_cpuidle_map(int cpu) +{ + if (likely(cpu_online(cpu))) { + cpumask_set_cpu(cpu, &grq.cpu_idle_map); + grq.idle_cpus = true; + } +} + +static inline void clear_cpuidle_map(int cpu) +{ + cpumask_clear_cpu(cpu, &grq.cpu_idle_map); + if (cpumask_empty(&grq.cpu_idle_map)) + grq.idle_cpus = false; +} + +static bool suitable_idle_cpus(struct task_struct *p) +{ + if (!grq.idle_cpus) + return false; + return (cpumask_intersects(&p->cpus_allowed, &grq.cpu_idle_map)); +} + +#define CPUIDLE_DIFF_THREAD (1) +#define CPUIDLE_DIFF_CORE (2) +#define CPUIDLE_CACHE_BUSY (4) +#define CPUIDLE_DIFF_CPU (8) +#define CPUIDLE_THREAD_BUSY (16) +#define CPUIDLE_THROTTLED (32) +#define CPUIDLE_DIFF_NODE (64) + +static inline bool scaling_rq(struct rq *rq); + +/* + * The best idle CPU is chosen according to the CPUIDLE ranking above where the + * lowest value would give the most suitable CPU to schedule p onto next. The + * order works out to be the following: + * + * Same core, idle or busy cache, idle or busy threads + * Other core, same cache, idle or busy cache, idle threads. + * Same node, other CPU, idle cache, idle threads. + * Same node, other CPU, busy cache, idle threads. + * Other core, same cache, busy threads. + * Same node, other CPU, busy threads. + * Other node, other CPU, idle cache, idle threads. + * Other node, other CPU, busy cache, idle threads. + * Other node, other CPU, busy threads. + */ +static int best_mask_cpu(int best_cpu, struct rq *rq, cpumask_t *tmpmask) +{ + int best_ranking = CPUIDLE_DIFF_NODE | CPUIDLE_THROTTLED | + CPUIDLE_THREAD_BUSY | CPUIDLE_DIFF_CPU | CPUIDLE_CACHE_BUSY | + CPUIDLE_DIFF_CORE | CPUIDLE_DIFF_THREAD; + int cpu_tmp; + + if (cpumask_test_cpu(best_cpu, tmpmask)) + goto out; + + for_each_cpu(cpu_tmp, tmpmask) { + int ranking, locality; + struct rq *tmp_rq; + + ranking = 0; + tmp_rq = cpu_rq(cpu_tmp); + + locality = rq->cpu_locality[cpu_tmp]; +#ifdef CONFIG_NUMA + if (locality > 3) + ranking |= CPUIDLE_DIFF_NODE; + else +#endif + if (locality > 2) + ranking |= CPUIDLE_DIFF_CPU; +#ifdef CONFIG_SCHED_MC + else if (locality == 2) + ranking |= CPUIDLE_DIFF_CORE; + if (!(tmp_rq->cache_idle(cpu_tmp))) + ranking |= CPUIDLE_CACHE_BUSY; +#endif +#ifdef CONFIG_SCHED_SMT + if (locality == 1) + ranking |= CPUIDLE_DIFF_THREAD; + if (!(tmp_rq->siblings_idle(cpu_tmp))) + ranking |= CPUIDLE_THREAD_BUSY; +#endif + if (scaling_rq(tmp_rq)) + ranking |= CPUIDLE_THROTTLED; + + if (ranking < best_ranking) { + best_cpu = cpu_tmp; + best_ranking = ranking; + } + } +out: + return best_cpu; +} + +static void resched_best_mask(int best_cpu, struct rq *rq, cpumask_t *tmpmask) +{ + best_cpu = best_mask_cpu(best_cpu, rq, tmpmask); + resched_curr(cpu_rq(best_cpu)); +} + +bool cpus_share_cache(int this_cpu, int that_cpu) +{ + struct rq *this_rq = cpu_rq(this_cpu); + + return (this_rq->cpu_locality[that_cpu] < 3); +} + +#ifdef CONFIG_SCHED_SMT +#ifdef CONFIG_SMT_NICE +static const cpumask_t *thread_cpumask(int cpu); + +/* Find the best real time priority running on any SMT siblings of cpu and if + * none are running, the static priority of the best deadline task running. + * The lookups to the other runqueues is done lockless as the occasional wrong + * value would be harmless. */ +static int best_smt_bias(int cpu) +{ + int other_cpu, best_bias = 0; + + for_each_cpu(other_cpu, thread_cpumask(cpu)) { + struct rq *rq; + + if (other_cpu == cpu) + continue; + rq = cpu_rq(other_cpu); + if (rq_idle(rq)) + continue; + if (!rq->online) + continue; + if (!rq->rq_mm) + continue; + if (likely(rq->rq_smt_bias > best_bias)) + best_bias = rq->rq_smt_bias; + } + return best_bias; +} + +static int task_prio_bias(struct task_struct *p) +{ + if (rt_task(p)) + return 1 << 30; + else if (task_running_iso(p)) + return 1 << 29; + else if (task_running_idle(p)) + return 0; + return MAX_PRIO - p->static_prio; +} + +/* We've already decided p can run on CPU, now test if it shouldn't for SMT + * nice reasons. */ +static bool smt_should_schedule(struct task_struct *p, int cpu) +{ + int best_bias, task_bias; + + /* Kernel threads always run */ + if (unlikely(!p->mm)) + return true; + if (rt_task(p)) + return true; + if (!idleprio_suitable(p)) + return true; + best_bias = best_smt_bias(cpu); + /* The smt siblings are all idle or running IDLEPRIO */ + if (best_bias < 1) + return true; + task_bias = task_prio_bias(p); + if (task_bias < 1) + return false; + if (task_bias >= best_bias) + return true; + /* Dither 25% cpu of normal tasks regardless of nice difference */ + if (best_bias % 4 == 1) + return true; + /* Sorry, you lose */ + return false; +} +#endif +#endif + +static bool resched_best_idle(struct task_struct *p) +{ + cpumask_t tmpmask; + int best_cpu; + + cpumask_and(&tmpmask, &p->cpus_allowed, &grq.cpu_idle_map); + best_cpu = best_mask_cpu(task_cpu(p), task_rq(p), &tmpmask); +#ifdef CONFIG_SMT_NICE + if (!smt_should_schedule(p, best_cpu)) + return false; +#endif + resched_curr(cpu_rq(best_cpu)); + return true; +} + +static inline void resched_suitable_idle(struct task_struct *p) +{ + if (suitable_idle_cpus(p)) + resched_best_idle(p); +} +/* + * Flags to tell us whether this CPU is running a CPU frequency governor that + * has slowed its speed or not. No locking required as the very rare wrongly + * read value would be harmless. + */ +void cpu_scaling(int cpu) +{ + cpu_rq(cpu)->scaling = true; +} + +void cpu_nonscaling(int cpu) +{ + cpu_rq(cpu)->scaling = false; +} + +static inline bool scaling_rq(struct rq *rq) +{ + return rq->scaling; +} + +static inline int locality_diff(int cpu, struct rq *rq) +{ + return rq->cpu_locality[cpu]; +} +#else /* CONFIG_SMP */ +static inline void set_cpuidle_map(int cpu) +{ +} + +static inline void clear_cpuidle_map(int cpu) +{ +} + +static inline bool suitable_idle_cpus(struct task_struct *p) +{ + return uprq->curr == uprq->idle; +} + +static inline void resched_suitable_idle(struct task_struct *p) +{ +} + +void cpu_scaling(int __unused) +{ +} + +void cpu_nonscaling(int __unused) +{ +} + +/* + * Although CPUs can scale in UP, there is nowhere else for tasks to go so this + * always returns 0. + */ +static inline bool scaling_rq(struct rq *rq) +{ + return false; +} + +static inline int locality_diff(int cpu, struct rq *rq) +{ + return 0; +} +#endif /* CONFIG_SMP */ +EXPORT_SYMBOL_GPL(cpu_scaling); +EXPORT_SYMBOL_GPL(cpu_nonscaling); + +static inline int normal_prio(struct task_struct *p) +{ + if (has_rt_policy(p)) + return MAX_RT_PRIO - 1 - p->rt_priority; + if (idleprio_task(p)) + return IDLE_PRIO; + if (iso_task(p)) + return ISO_PRIO; + return NORMAL_PRIO; +} + +/* + * Calculate the current priority, i.e. the priority + * taken into account by the scheduler. This value might + * be boosted by RT tasks as it will be RT if the task got + * RT-boosted. If not then it returns p->normal_prio. + */ +static int effective_prio(struct task_struct *p) +{ + p->normal_prio = normal_prio(p); + /* + * If we are RT tasks or we were boosted to RT priority, + * keep the priority unchanged. Otherwise, update priority + * to the normal priority: + */ + if (!rt_prio(p->prio)) + return p->normal_prio; + return p->prio; +} + +/* + * activate_task - move a task to the runqueue. Enter with grq locked. + */ +static void activate_task(struct task_struct *p, struct rq *rq) +{ + update_clocks(rq); + + /* + * Sleep time is in units of nanosecs, so shift by 20 to get a + * milliseconds-range estimation of the amount of time that the task + * spent sleeping: + */ + if (unlikely(prof_on == SLEEP_PROFILING)) { + if (p->state == TASK_UNINTERRUPTIBLE) + profile_hits(SLEEP_PROFILING, (void *)get_wchan(p), + (rq->clock_task - p->last_ran) >> 20); + } + + p->prio = effective_prio(p); + if (task_contributes_to_load(p)) + grq.nr_uninterruptible--; + enqueue_task(p, rq); + rq->soft_affined++; + p->on_rq = 1; + grq.nr_running++; + inc_qnr(); +} + +static inline void clear_sticky(struct task_struct *p); + +/* + * deactivate_task - If it's running, it's not on the grq and we can just + * decrement the nr_running. Enter with grq locked. + */ +static inline void deactivate_task(struct task_struct *p, struct rq *rq) +{ + if (task_contributes_to_load(p)) + grq.nr_uninterruptible++; + rq->soft_affined--; + p->on_rq = 0; + grq.nr_running--; + clear_sticky(p); +} + +#ifdef CONFIG_SMP +void set_task_cpu(struct task_struct *p, unsigned int cpu) +{ +#ifdef CONFIG_LOCKDEP + /* + * The caller should hold grq lock. + */ + WARN_ON_ONCE(debug_locks && !lockdep_is_held(&grq.lock)); +#endif + if (task_cpu(p) == cpu) + return; + trace_sched_migrate_task(p, cpu); + perf_event_task_migrate(p); + + /* + * After ->cpu is set up to a new value, task_grq_lock(p, ...) can be + * successfully executed on another CPU. We must ensure that updates of + * per-task data have been completed by this moment. + */ + smp_wmb(); + if (p->on_rq) { + task_rq(p)->soft_affined--; + cpu_rq(cpu)->soft_affined++; + } + task_thread_info(p)->cpu = cpu; +} + +static inline void clear_sticky(struct task_struct *p) +{ + p->sticky = false; +} + +static inline bool task_sticky(struct task_struct *p) +{ + return p->sticky; +} + +/* Reschedule the best idle CPU that is not this one. */ +static void +resched_closest_idle(struct rq *rq, int cpu, struct task_struct *p) +{ + cpumask_t tmpmask; + + cpumask_and(&tmpmask, &p->cpus_allowed, &grq.cpu_idle_map); + cpumask_clear_cpu(cpu, &tmpmask); + if (cpumask_empty(&tmpmask)) + return; + resched_best_mask(cpu, rq, &tmpmask); +} + +/* + * We set the sticky flag on a task that is descheduled involuntarily meaning + * it is awaiting further CPU time. If the last sticky task is still sticky + * but unlucky enough to not be the next task scheduled, we unstick it and try + * to find it an idle CPU. Realtime tasks do not stick to minimise their + * latency at all times. + */ +static inline void +swap_sticky(struct rq *rq, int cpu, struct task_struct *p) +{ + if (rq->sticky_task) { + if (rq->sticky_task == p) { + p->sticky = true; + return; + } + if (task_sticky(rq->sticky_task)) { + clear_sticky(rq->sticky_task); + resched_closest_idle(rq, cpu, rq->sticky_task); + } + } + if (!rt_task(p)) { + p->sticky = true; + rq->sticky_task = p; + } else { + resched_closest_idle(rq, cpu, p); + rq->sticky_task = NULL; + } +} + +static inline void unstick_task(struct rq *rq, struct task_struct *p) +{ + rq->sticky_task = NULL; + clear_sticky(p); +} +#else +static inline void clear_sticky(struct task_struct *p) +{ +} + +static inline bool task_sticky(struct task_struct *p) +{ + return false; +} + +static inline void +swap_sticky(struct rq *rq, int cpu, struct task_struct *p) +{ +} + +static inline void unstick_task(struct rq *rq, struct task_struct *p) +{ +} +#endif + +/* + * Move a task off the global queue and take it to a cpu for it will + * become the running task. + */ +static inline void take_task(int cpu, struct task_struct *p) +{ + set_task_cpu(p, cpu); + dequeue_task(p); + clear_sticky(p); + dec_qnr(); +} + +/* + * Returns a descheduling task to the grq runqueue unless it is being + * deactivated. + */ +static inline void return_task(struct task_struct *p, struct rq *rq, bool deactivate) +{ + if (deactivate) + deactivate_task(p, rq); + else { + inc_qnr(); + enqueue_task(p, rq); + } +} + +/* Enter with grq lock held. We know p is on the local cpu */ +static inline void __set_tsk_resched(struct task_struct *p) +{ + set_tsk_need_resched(p); + set_preempt_need_resched(); +} + +/* + * resched_task - mark a task 'to be rescheduled now'. + * + * On UP this means the setting of the need_resched flag, on SMP it + * might also involve a cross-CPU call to trigger the scheduler on + * the target CPU. + */ +void resched_task(struct task_struct *p) +{ + int cpu; + + lockdep_assert_held(&grq.lock); + + if (test_tsk_need_resched(p)) + return; + + set_tsk_need_resched(p); + + cpu = task_cpu(p); + if (cpu == smp_processor_id()) { + set_preempt_need_resched(); + return; + } + + smp_send_reschedule(cpu); +} + +/** + * task_curr - is this task currently executing on a CPU? + * @p: the task in question. + * + * Return: 1 if the task is currently executing. 0 otherwise. + */ +inline int task_curr(const struct task_struct *p) +{ + return cpu_curr(task_cpu(p)) == p; +} + +#ifdef CONFIG_SMP +struct migration_req { + struct task_struct *task; + int dest_cpu; +}; + +/* + * wait_task_inactive - wait for a thread to unschedule. + * + * If @match_state is nonzero, it's the @p->state value just checked and + * not expected to change. If it changes, i.e. @p might have woken up, + * then return zero. When we succeed in waiting for @p to be off its CPU, + * we return a positive number (its total switch count). If a second call + * a short while later returns the same number, the caller can be sure that + * @p has remained unscheduled the whole time. + * + * The caller must ensure that the task *will* unschedule sometime soon, + * else this function might spin for a *long* time. This function can't + * be called with interrupts off, or it may introduce deadlock with + * smp_call_function() if an IPI is sent by the same process we are + * waiting to become inactive. + */ +unsigned long wait_task_inactive(struct task_struct *p, long match_state) +{ + unsigned long flags; + bool running, on_rq; + unsigned long ncsw; + struct rq *rq; + + for (;;) { + rq = task_rq(p); + + /* + * If the task is actively running on another CPU + * still, just relax and busy-wait without holding + * any locks. + * + * NOTE! Since we don't hold any locks, it's not + * even sure that "rq" stays as the right runqueue! + * But we don't care, since this will return false + * if the runqueue has changed and p is actually now + * running somewhere else! + */ + while (task_running(p) && p == rq->curr) { + if (match_state && unlikely(p->state != match_state)) + return 0; + cpu_relax(); + } + + /* + * Ok, time to look more closely! We need the grq + * lock now, to be *sure*. If we're wrong, we'll + * just go back and repeat. + */ + rq = task_grq_lock(p, &flags); + trace_sched_wait_task(p); + running = task_running(p); + on_rq = p->on_rq; + ncsw = 0; + if (!match_state || p->state == match_state) + ncsw = p->nvcsw | LONG_MIN; /* sets MSB */ + task_grq_unlock(&flags); + + /* + * If it changed from the expected state, bail out now. + */ + if (unlikely(!ncsw)) + break; + + /* + * Was it really running after all now that we + * checked with the proper locks actually held? + * + * Oops. Go back and try again.. + */ + if (unlikely(running)) { + cpu_relax(); + continue; + } + + /* + * It's not enough that it's not actively running, + * it must be off the runqueue _entirely_, and not + * preempted! + * + * So if it was still runnable (but just not actively + * running right now), it's preempted, and we should + * yield - it could be a while. + */ + if (unlikely(on_rq)) { + ktime_t to = ktime_set(0, NSEC_PER_SEC / HZ); + + set_current_state(TASK_UNINTERRUPTIBLE); + schedule_hrtimeout(&to, HRTIMER_MODE_REL); + continue; + } + + /* + * Ahh, all good. It wasn't running, and it wasn't + * runnable, which means that it will never become + * running in the future either. We're all done! + */ + break; + } + + return ncsw; +} + +/*** + * kick_process - kick a running thread to enter/exit the kernel + * @p: the to-be-kicked thread + * + * Cause a process which is running on another CPU to enter + * kernel-mode, without any delay. (to get signals handled.) + * + * NOTE: this function doesn't have to take the runqueue lock, + * because all it wants to ensure is that the remote task enters + * the kernel. If the IPI races and the task has been migrated + * to another CPU then no harm is done and the purpose has been + * achieved as well. + */ +void kick_process(struct task_struct *p) +{ + int cpu; + + preempt_disable(); + cpu = task_cpu(p); + if ((cpu != smp_processor_id()) && task_curr(p)) + smp_send_reschedule(cpu); + preempt_enable(); +} +EXPORT_SYMBOL_GPL(kick_process); +#endif + +/* + * RT tasks preempt purely on priority. SCHED_NORMAL tasks preempt on the + * basis of earlier deadlines. SCHED_IDLEPRIO don't preempt anything else or + * between themselves, they cooperatively multitask. An idle rq scores as + * prio PRIO_LIMIT so it is always preempted. + */ +static inline bool +can_preempt(struct task_struct *p, int prio, u64 deadline) +{ + /* Better static priority RT task or better policy preemption */ + if (p->prio < prio) + return true; + if (p->prio > prio) + return false; + /* SCHED_NORMAL, BATCH and ISO will preempt based on deadline */ + if (!deadline_before(p->deadline, deadline)) + return false; + return true; +} + +#ifdef CONFIG_SMP +#define cpu_online_map (*(cpumask_t *)cpu_online_mask) +#ifdef CONFIG_HOTPLUG_CPU +/* + * Check to see if there is a task that is affined only to offline CPUs but + * still wants runtime. This happens to kernel threads during suspend/halt and + * disabling of CPUs. + */ +static inline bool online_cpus(struct task_struct *p) +{ + return (likely(cpumask_intersects(&cpu_online_map, &p->cpus_allowed))); +} +#else /* CONFIG_HOTPLUG_CPU */ +/* All available CPUs are always online without hotplug. */ +static inline bool online_cpus(struct task_struct *p) +{ + return true; +} +#endif + +/* + * Check to see if p can run on cpu, and if not, whether there are any online + * CPUs it can run on instead. + */ +static inline bool needs_other_cpu(struct task_struct *p, int cpu) +{ + if (unlikely(!cpumask_test_cpu(cpu, &p->cpus_allowed))) + return true; + return false; +} + +/* + * When all else is equal, still prefer this_rq. + */ +static void try_preempt(struct task_struct *p, struct rq *this_rq) +{ + struct rq *highest_prio_rq = NULL; + int cpu, highest_prio; + u64 latest_deadline; + cpumask_t tmp; + + /* + * We clear the sticky flag here because for a task to have called + * try_preempt with the sticky flag enabled means some complicated + * re-scheduling has occurred and we should ignore the sticky flag. + */ + clear_sticky(p); + + if (suitable_idle_cpus(p) && resched_best_idle(p)) + return; + + /* IDLEPRIO tasks never preempt anything but idle */ + if (p->policy == SCHED_IDLEPRIO) + return; + + if (likely(online_cpus(p))) + cpumask_and(&tmp, &cpu_online_map, &p->cpus_allowed); + else + return; + + highest_prio = latest_deadline = 0; + + for_each_cpu(cpu, &tmp) { + struct rq *rq; + int rq_prio; + + rq = cpu_rq(cpu); + rq_prio = rq->rq_prio; + if (rq_prio < highest_prio) + continue; + + if (rq_prio > highest_prio || + deadline_after(rq->rq_deadline, latest_deadline)) { + latest_deadline = rq->rq_deadline; + highest_prio = rq_prio; + highest_prio_rq = rq; + } + } + + if (likely(highest_prio_rq)) { +#ifdef CONFIG_SMT_NICE + cpu = cpu_of(highest_prio_rq); + if (!smt_should_schedule(p, cpu)) + return; +#endif + if (can_preempt(p, highest_prio, highest_prio_rq->rq_deadline)) + resched_curr(highest_prio_rq); + } +} +static int __set_cpus_allowed_ptr(struct task_struct *p, + const struct cpumask *new_mask, bool check); +#else /* CONFIG_SMP */ +static inline bool needs_other_cpu(struct task_struct *p, int cpu) +{ + return false; +} + +static void try_preempt(struct task_struct *p, struct rq *this_rq) +{ + if (p->policy == SCHED_IDLEPRIO) + return; + if (can_preempt(p, uprq->rq_prio, uprq->rq_deadline)) + resched_curr(uprq); +} + +static inline int __set_cpus_allowed_ptr(struct task_struct *p, + const struct cpumask *new_mask, bool check) +{ + return set_cpus_allowed_ptr(p, new_mask); +} +#endif /* CONFIG_SMP */ + +static void +ttwu_stat(struct task_struct *p, int cpu, int wake_flags) +{ +#ifdef CONFIG_SCHEDSTATS + struct rq *rq = this_rq(); + +#ifdef CONFIG_SMP + int this_cpu = smp_processor_id(); + + if (cpu == this_cpu) + schedstat_inc(rq, ttwu_local); + else { + struct sched_domain *sd; + + rcu_read_lock(); + for_each_domain(this_cpu, sd) { + if (cpumask_test_cpu(cpu, sched_domain_span(sd))) { + schedstat_inc(sd, ttwu_wake_remote); + break; + } + } + rcu_read_unlock(); + } + +#endif /* CONFIG_SMP */ + + schedstat_inc(rq, ttwu_count); +#endif /* CONFIG_SCHEDSTATS */ +} + +void wake_up_if_idle(int cpu) +{ + struct rq *rq = cpu_rq(cpu); + unsigned long flags; + + rcu_read_lock(); + + if (!is_idle_task(rcu_dereference(rq->curr))) + goto out; + + grq_lock_irqsave(&flags); + if (likely(is_idle_task(rq->curr))) + smp_send_reschedule(cpu); + /* Else cpu is not in idle, do nothing here */ + grq_unlock_irqrestore(&flags); + +out: + rcu_read_unlock(); +} + +#ifdef CONFIG_SMP +void scheduler_ipi(void) +{ + /* + * Fold TIF_NEED_RESCHED into the preempt_count; anybody setting + * TIF_NEED_RESCHED remotely (for the first time) will also send + * this IPI. + */ + preempt_fold_need_resched(); +} +#endif + +static inline void ttwu_activate(struct task_struct *p, struct rq *rq, + bool is_sync) +{ + activate_task(p, rq); + + /* + * Sync wakeups (i.e. those types of wakeups where the waker + * has indicated that it will leave the CPU in short order) + * don't trigger a preemption if there are no idle cpus, + * instead waiting for current to deschedule. + */ + if (!is_sync || suitable_idle_cpus(p)) + try_preempt(p, rq); +} + +static inline void ttwu_post_activation(struct task_struct *p, struct rq *rq, + bool success) +{ + trace_sched_wakeup(p); + p->state = TASK_RUNNING; + + /* + * if a worker is waking up, notify workqueue. Note that on BFS, we + * don't really know what cpu it will be, so we fake it for + * wq_worker_waking_up :/ + */ + if ((p->flags & PF_WQ_WORKER) && success) + wq_worker_waking_up(p, cpu_of(rq)); +} + +/* + * wake flags + */ +#define WF_SYNC 0x01 /* waker goes to sleep after wakeup */ +#define WF_FORK 0x02 /* child wakeup after fork */ +#define WF_MIGRATED 0x4 /* internal use, task got migrated */ + +/*** + * try_to_wake_up - wake up a thread + * @p: the thread to be awakened + * @state: the mask of task states that can be woken + * @wake_flags: wake modifier flags (WF_*) + * + * Put it on the run-queue if it's not already there. The "current" + * thread is always on the run-queue (except when the actual + * re-schedule is in progress), and as such you're allowed to do + * the simpler "current->state = TASK_RUNNING" to mark yourself + * runnable without the overhead of this. + * + * Return: %true if @p was woken up, %false if it was already running. + * or @state didn't match @p's state. + */ +static bool try_to_wake_up(struct task_struct *p, unsigned int state, + int wake_flags) +{ + bool success = false; + unsigned long flags; + struct rq *rq; + int cpu; + + get_cpu(); + + /* + * If we are going to wake up a thread waiting for CONDITION we + * need to ensure that CONDITION=1 done by the caller can not be + * reordered with p->state check below. This pairs with mb() in + * set_current_state() the waiting thread does. + */ + smp_mb__before_spinlock(); + + /* + * No need to do time_lock_grq as we only need to update the rq clock + * if we activate the task + */ + rq = task_grq_lock(p, &flags); + cpu = task_cpu(p); + + /* state is a volatile long, どうして、分からない */ + if (!((unsigned int)p->state & state)) + goto out_unlock; + + trace_sched_waking(p); + + if (task_queued(p) || task_running(p)) + goto out_running; + + ttwu_activate(p, rq, wake_flags & WF_SYNC); + success = true; + +out_running: + ttwu_post_activation(p, rq, success); +out_unlock: + task_grq_unlock(&flags); + + ttwu_stat(p, cpu, wake_flags); + + put_cpu(); + + return success; +} + +/** + * try_to_wake_up_local - try to wake up a local task with grq lock held + * @p: the thread to be awakened + * + * Put @p on the run-queue if it's not already there. The caller must + * ensure that grq is locked and, @p is not the current task. + * grq stays locked over invocation. + */ +static void try_to_wake_up_local(struct task_struct *p) +{ + struct rq *rq = task_rq(p); + bool success = false; + + lockdep_assert_held(&grq.lock); + + if (!(p->state & TASK_NORMAL)) + return; + + trace_sched_waking(p); + + if (!task_queued(p)) { + if (likely(!task_running(p))) { + schedstat_inc(rq, ttwu_count); + schedstat_inc(rq, ttwu_local); + } + ttwu_activate(p, rq, false); + ttwu_stat(p, smp_processor_id(), 0); + success = true; + } + ttwu_post_activation(p, rq, success); +} + +/** + * wake_up_process - Wake up a specific process + * @p: The process to be woken up. + * + * Attempt to wake up the nominated process and move it to the set of runnable + * processes. + * + * Return: 1 if the process was woken up, 0 if it was already running. + * + * It may be assumed that this function implies a write memory barrier before + * changing the task state if and only if any tasks are woken up. + */ +int wake_up_process(struct task_struct *p) +{ + return try_to_wake_up(p, TASK_NORMAL, 0); +} +EXPORT_SYMBOL(wake_up_process); + +int wake_up_state(struct task_struct *p, unsigned int state) +{ + return try_to_wake_up(p, state, 0); +} + +static void time_slice_expired(struct task_struct *p); + +/* + * Perform scheduler related setup for a newly forked process p. + * p is forked by current. + */ +int sched_fork(unsigned long __maybe_unused clone_flags, struct task_struct *p) +{ +#ifdef CONFIG_PREEMPT_NOTIFIERS + INIT_HLIST_HEAD(&p->preempt_notifiers); +#endif + /* + * The process state is set to the same value of the process executing + * do_fork() code. That is running. This guarantees that nobody will + * actually run it, and a signal or other external event cannot wake + * it up and insert it on the runqueue either. + */ + + /* Should be reset in fork.c but done here for ease of bfs patching */ + p->on_rq = + p->utime = + p->stime = + p->utimescaled = + p->stimescaled = + p->sched_time = + p->stime_pc = + p->utime_pc = 0; + + /* + * Revert to default priority/policy on fork if requested. + */ + if (unlikely(p->sched_reset_on_fork)) { + if (p->policy == SCHED_FIFO || p->policy == SCHED_RR) { + p->policy = SCHED_NORMAL; + p->normal_prio = normal_prio(p); + } + + if (PRIO_TO_NICE(p->static_prio) < 0) { + p->static_prio = NICE_TO_PRIO(0); + p->normal_prio = p->static_prio; + } + + /* + * We don't need the reset flag anymore after the fork. It has + * fulfilled its duty: + */ + p->sched_reset_on_fork = 0; + } + + INIT_LIST_HEAD(&p->run_list); +#ifdef CONFIG_SCHED_INFO + if (unlikely(sched_info_on())) + memset(&p->sched_info, 0, sizeof(p->sched_info)); +#endif + p->on_cpu = false; + clear_sticky(p); + init_task_preempt_count(p); + return 0; +} + +/* + * wake_up_new_task - wake up a newly created task for the first time. + * + * This function will do some initial scheduler statistics housekeeping + * that must be done for every newly created context, then puts the task + * on the runqueue and wakes it. + */ +void wake_up_new_task(struct task_struct *p) +{ + struct task_struct *parent; + unsigned long flags; + struct rq *rq; + + parent = p->parent; + rq = task_grq_lock(p, &flags); + + /* + * Reinit new task deadline as its creator deadline could have changed + * since call to dup_task_struct(). + */ + p->deadline = rq->rq_deadline; + + /* + * If the task is a new process, current and parent are the same. If + * the task is a new thread in the thread group, it will have much more + * in common with current than with the parent. + */ + set_task_cpu(p, task_cpu(rq->curr)); + + /* + * Make sure we do not leak PI boosting priority to the child. + */ + p->prio = rq->curr->normal_prio; + + activate_task(p, rq); + trace_sched_wakeup_new(p); + if (unlikely(p->policy == SCHED_FIFO)) + goto after_ts_init; + + /* + * Share the timeslice between parent and child, thus the + * total amount of pending timeslices in the system doesn't change, + * resulting in more scheduling fairness. If it's negative, it won't + * matter since that's the same as being 0. current's time_slice is + * actually in rq_time_slice when it's running, as is its last_ran + * value. rq->rq_deadline is only modified within schedule() so it + * is always equal to current->deadline. + */ + p->last_ran = rq->rq_last_ran; + if (likely(rq->rq_time_slice >= RESCHED_US * 2)) { + rq->rq_time_slice /= 2; + p->time_slice = rq->rq_time_slice; +after_ts_init: + if (rq->curr == parent && !suitable_idle_cpus(p)) { + /* + * The VM isn't cloned, so we're in a good position to + * do child-runs-first in anticipation of an exec. This + * usually avoids a lot of COW overhead. + */ + __set_tsk_resched(parent); + } else + try_preempt(p, rq); + } else { + if (rq->curr == parent) { + /* + * Forking task has run out of timeslice. Reschedule it and + * start its child with a new time slice and deadline. The + * child will end up running first because its deadline will + * be slightly earlier. + */ + rq->rq_time_slice = 0; + __set_tsk_resched(parent); + } + time_slice_expired(p); + } + task_grq_unlock(&flags); +} + +#ifdef CONFIG_PREEMPT_NOTIFIERS + +static struct static_key preempt_notifier_key = STATIC_KEY_INIT_FALSE; + +void preempt_notifier_inc(void) +{ + static_key_slow_inc(&preempt_notifier_key); +} +EXPORT_SYMBOL_GPL(preempt_notifier_inc); + +void preempt_notifier_dec(void) +{ + static_key_slow_dec(&preempt_notifier_key); +} +EXPORT_SYMBOL_GPL(preempt_notifier_dec); + +/** + * preempt_notifier_register - tell me when current is being preempted & rescheduled + * @notifier: notifier struct to register + */ +void preempt_notifier_register(struct preempt_notifier *notifier) +{ + if (!static_key_false(&preempt_notifier_key)) + WARN(1, "registering preempt_notifier while notifiers disabled\n"); + + hlist_add_head(¬ifier->link, ¤t->preempt_notifiers); +} +EXPORT_SYMBOL_GPL(preempt_notifier_register); + +/** + * preempt_notifier_unregister - no longer interested in preemption notifications + * @notifier: notifier struct to unregister + * + * This is *not* safe to call from within a preemption notifier. + */ +void preempt_notifier_unregister(struct preempt_notifier *notifier) +{ + hlist_del(¬ifier->link); +} +EXPORT_SYMBOL_GPL(preempt_notifier_unregister); + +static void __fire_sched_in_preempt_notifiers(struct task_struct *curr) +{ + struct preempt_notifier *notifier; + + hlist_for_each_entry(notifier, &curr->preempt_notifiers, link) + notifier->ops->sched_in(notifier, raw_smp_processor_id()); +} + +static __always_inline void fire_sched_in_preempt_notifiers(struct task_struct *curr) +{ + if (static_key_false(&preempt_notifier_key)) + __fire_sched_in_preempt_notifiers(curr); +} + +static void +__fire_sched_out_preempt_notifiers(struct task_struct *curr, + struct task_struct *next) +{ + struct preempt_notifier *notifier; + + hlist_for_each_entry(notifier, &curr->preempt_notifiers, link) + notifier->ops->sched_out(notifier, next); +} + +static __always_inline void +fire_sched_out_preempt_notifiers(struct task_struct *curr, + struct task_struct *next) +{ + if (static_key_false(&preempt_notifier_key)) + __fire_sched_out_preempt_notifiers(curr, next); +} + +#else /* !CONFIG_PREEMPT_NOTIFIERS */ + +static inline void fire_sched_in_preempt_notifiers(struct task_struct *curr) +{ +} + +static inline void +fire_sched_out_preempt_notifiers(struct task_struct *curr, + struct task_struct *next) +{ +} + +#endif /* CONFIG_PREEMPT_NOTIFIERS */ + +/** + * prepare_task_switch - prepare to switch tasks + * @rq: the runqueue preparing to switch + * @next: the task we are going to switch to. + * + * This is called with the rq lock held and interrupts off. It must + * be paired with a subsequent finish_task_switch after the context + * switch. + * + * prepare_task_switch sets up locking and calls architecture specific + * hooks. + */ +static inline void +prepare_task_switch(struct rq *rq, struct task_struct *prev, + struct task_struct *next) +{ + sched_info_switch(rq, prev, next); + perf_event_task_sched_out(prev, next); + fire_sched_out_preempt_notifiers(prev, next); + prepare_lock_switch(rq, next); + prepare_arch_switch(next); +} + +/** + * finish_task_switch - clean up after a task-switch + * @rq: runqueue associated with task-switch + * @prev: the thread we just switched away from. + * + * finish_task_switch must be called after the context switch, paired + * with a prepare_task_switch call before the context switch. + * finish_task_switch will reconcile locking set up by prepare_task_switch, + * and do any other architecture-specific cleanup actions. + * + * Note that we may have delayed dropping an mm in context_switch(). If + * so, we finish that here outside of the runqueue lock. (Doing it + * with the lock held can cause deadlocks; see schedule() for + * details.) + * + * The context switch have flipped the stack from under us and restored the + * local variables which were saved when this task called schedule() in the + * past. prev == current is still correct but we need to recalculate this_rq + * because prev may have moved to another CPU. + */ +static struct rq *finish_task_switch(struct task_struct *prev) + __releases(grq.lock) +{ + struct rq *rq = this_rq(); + struct mm_struct *mm = rq->prev_mm; + long prev_state; + + /* + * The previous task will have left us with a preempt_count of 2 + * because it left us after: + * + * schedule() + * preempt_disable(); // 1 + * __schedule() + * raw_spin_lock_irq(&rq->lock) // 2 + * + * Also, see FORK_PREEMPT_COUNT. + */ + if (WARN_ONCE(preempt_count() != 2*PREEMPT_DISABLE_OFFSET, + "corrupted preempt_count: %s/%d/0x%x\n", + current->comm, current->pid, preempt_count())) + preempt_count_set(FORK_PREEMPT_COUNT); + + rq->prev_mm = NULL; + + /* + * A task struct has one reference for the use as "current". + * If a task dies, then it sets TASK_DEAD in tsk->state and calls + * schedule one last time. The schedule call will never return, and + * the scheduled task must drop that reference. + * + * We must observe prev->state before clearing prev->on_cpu (in + * finish_lock_switch), otherwise a concurrent wakeup can get prev + * running on another CPU and we could rave with its RUNNING -> DEAD + * transition, resulting in a double drop. + */ + prev_state = prev->state; + vtime_task_switch(prev); + perf_event_task_sched_in(prev, current); + finish_lock_switch(rq, prev); + finish_arch_post_lock_switch(); + + fire_sched_in_preempt_notifiers(current); + if (mm) + mmdrop(mm); + if (unlikely(prev_state == TASK_DEAD)) { + /* + * Remove function-return probe instances associated with this + * task and put them back on the free list. + */ + kprobe_flush_task(prev); + put_task_struct(prev); + } + return rq; +} + +/** + * schedule_tail - first thing a freshly forked thread must call. + * @prev: the thread we just switched away from. + */ +asmlinkage __visible void schedule_tail(struct task_struct *prev) + __releases(grq.lock) +{ + struct rq *rq; + + /* + * New tasks start with FORK_PREEMPT_COUNT, see there and + * finish_task_switch() for details. + * + * finish_task_switch() will drop rq->lock() and lower preempt_count + * and the preempt_enable() will end up enabling preemption (on + * PREEMPT_COUNT kernels). + */ + + rq = finish_task_switch(prev); + preempt_enable(); + + if (current->set_child_tid) + put_user(task_pid_vnr(current), current->set_child_tid); +} + +/* + * context_switch - switch to the new MM and the new thread's register state. + */ +static inline struct rq * +context_switch(struct rq *rq, struct task_struct *prev, + struct task_struct *next) +{ + struct mm_struct *mm, *oldmm; + + prepare_task_switch(rq, prev, next); + + mm = next->mm; + oldmm = prev->active_mm; + /* + * For paravirt, this is coupled with an exit in switch_to to + * combine the page table reload and the switch backend into + * one hypercall. + */ + arch_start_context_switch(prev); + + if (!mm) { + next->active_mm = oldmm; + atomic_inc(&oldmm->mm_count); + enter_lazy_tlb(oldmm, next); + } else + switch_mm(oldmm, mm, next); + + if (!prev->mm) { + prev->active_mm = NULL; + rq->prev_mm = oldmm; + } + /* + * Since the runqueue lock will be released by the next + * task (which is an invalid locking op but in the case + * of the scheduler it's an obvious special-case), so we + * do an early lockdep release here: + */ + spin_release(&grq.lock.dep_map, 1, _THIS_IP_); + + /* Here we just switch the register state and the stack. */ + switch_to(prev, next, prev); + barrier(); + + return finish_task_switch(prev); +} + +/* + * nr_running, nr_uninterruptible and nr_context_switches: + * + * externally visible scheduler statistics: current number of runnable + * threads, total number of context switches performed since bootup. All are + * measured without grabbing the grq lock but the occasional inaccurate result + * doesn't matter so long as it's positive. + */ +unsigned long nr_running(void) +{ + long nr = grq.nr_running; + + if (unlikely(nr < 0)) + nr = 0; + return (unsigned long)nr; +} + +static unsigned long nr_uninterruptible(void) +{ + long nu = grq.nr_uninterruptible; + + if (unlikely(nu < 0)) + nu = 0; + return nu; +} + +/* + * Check if only the current task is running on the cpu. + * + * Caution: this function does not check that the caller has disabled + * preemption, thus the result might have a time-of-check-to-time-of-use + * race. The caller is responsible to use it correctly, for example: + * + * - from a non-preemptable section (of course) + * + * - from a thread that is bound to a single CPU + * + * - in a loop with very short iterations (e.g. a polling loop) + */ +bool single_task_running(void) +{ + if (cpu_rq(smp_processor_id())->soft_affined == 1) + return true; + else + return false; +} +EXPORT_SYMBOL(single_task_running); + +unsigned long long nr_context_switches(void) +{ + long long ns = grq.nr_switches; + + /* This is of course impossible */ + if (unlikely(ns < 0)) + ns = 1; + return (unsigned long long)ns; +} + +unsigned long nr_iowait(void) +{ + unsigned long i, sum = 0; + + for_each_possible_cpu(i) + sum += atomic_read(&cpu_rq(i)->nr_iowait); + + return sum; +} + +unsigned long nr_iowait_cpu(int cpu) +{ + struct rq *this = cpu_rq(cpu); + return atomic_read(&this->nr_iowait); +} + +unsigned long nr_active(void) +{ + return nr_running() + nr_uninterruptible(); +} + +/* Beyond a task running on this CPU, load is equal everywhere on BFS, so we + * base it on the number of running or queued tasks with their ->rq pointer + * set to this cpu as being the CPU they're more likely to run on. */ +void get_iowait_load(unsigned long *nr_waiters, unsigned long *load) +{ + struct rq *rq = this_rq(); + + *nr_waiters = atomic_read(&rq->nr_iowait); + *load = rq->soft_affined; +} + +/* Variables and functions for calc_load */ +static unsigned long calc_load_update; +unsigned long avenrun[3]; +EXPORT_SYMBOL(avenrun); + +/** + * get_avenrun - get the load average array + * @loads: pointer to dest load array + * @offset: offset to add + * @shift: shift count to shift the result left + * + * These values are estimates at best, so no need for locking. + */ +void get_avenrun(unsigned long *loads, unsigned long offset, int shift) +{ + loads[0] = (avenrun[0] + offset) << shift; + loads[1] = (avenrun[1] + offset) << shift; + loads[2] = (avenrun[2] + offset) << shift; +} + +static unsigned long +calc_load(unsigned long load, unsigned long exp, unsigned long active) +{ + load *= exp; + load += active * (FIXED_1 - exp); + return load >> FSHIFT; +} + +/* + * calc_load - update the avenrun load estimates every LOAD_FREQ seconds. + */ +void calc_global_load(unsigned long ticks) +{ + long active; + + if (time_before(jiffies, calc_load_update)) + return; + active = nr_active() * FIXED_1; + + avenrun[0] = calc_load(avenrun[0], EXP_1, active); + avenrun[1] = calc_load(avenrun[1], EXP_5, active); + avenrun[2] = calc_load(avenrun[2], EXP_15, active); + + calc_load_update = jiffies + LOAD_FREQ; +} + +DEFINE_PER_CPU(struct kernel_stat, kstat); +DEFINE_PER_CPU(struct kernel_cpustat, kernel_cpustat); + +EXPORT_PER_CPU_SYMBOL(kstat); +EXPORT_PER_CPU_SYMBOL(kernel_cpustat); + +#ifdef CONFIG_IRQ_TIME_ACCOUNTING + +/* + * There are no locks covering percpu hardirq/softirq time. + * They are only modified in account_system_vtime, on corresponding CPU + * with interrupts disabled. So, writes are safe. + * They are read and saved off onto struct rq in update_rq_clock(). + * This may result in other CPU reading this CPU's irq time and can + * race with irq/account_system_vtime on this CPU. We would either get old + * or new value with a side effect of accounting a slice of irq time to wrong + * task when irq is in progress while we read rq->clock. That is a worthy + * compromise in place of having locks on each irq in account_system_time. + */ +static DEFINE_PER_CPU(u64, cpu_hardirq_time); +static DEFINE_PER_CPU(u64, cpu_softirq_time); + +static DEFINE_PER_CPU(u64, irq_start_time); +static int sched_clock_irqtime; + +void enable_sched_clock_irqtime(void) +{ + sched_clock_irqtime = 1; +} + +void disable_sched_clock_irqtime(void) +{ + sched_clock_irqtime = 0; +} + +#ifndef CONFIG_64BIT +static DEFINE_PER_CPU(seqcount_t, irq_time_seq); + +static inline void irq_time_write_begin(void) +{ + __this_cpu_inc(irq_time_seq.sequence); + smp_wmb(); +} + +static inline void irq_time_write_end(void) +{ + smp_wmb(); + __this_cpu_inc(irq_time_seq.sequence); +} + +static inline u64 irq_time_read(int cpu) +{ + u64 irq_time; + unsigned seq; + + do { + seq = read_seqcount_begin(&per_cpu(irq_time_seq, cpu)); + irq_time = per_cpu(cpu_softirq_time, cpu) + + per_cpu(cpu_hardirq_time, cpu); + } while (read_seqcount_retry(&per_cpu(irq_time_seq, cpu), seq)); + + return irq_time; +} +#else /* CONFIG_64BIT */ +static inline void irq_time_write_begin(void) +{ +} + +static inline void irq_time_write_end(void) +{ +} + +static inline u64 irq_time_read(int cpu) +{ + return per_cpu(cpu_softirq_time, cpu) + per_cpu(cpu_hardirq_time, cpu); +} +#endif /* CONFIG_64BIT */ + +/* + * Called before incrementing preempt_count on {soft,}irq_enter + * and before decrementing preempt_count on {soft,}irq_exit. + */ +void irqtime_account_irq(struct task_struct *curr) +{ + unsigned long flags; + s64 delta; + int cpu; + + if (!sched_clock_irqtime) + return; + + local_irq_save(flags); + + cpu = smp_processor_id(); + delta = sched_clock_cpu(cpu) - __this_cpu_read(irq_start_time); + __this_cpu_add(irq_start_time, delta); + + irq_time_write_begin(); + /* + * We do not account for softirq time from ksoftirqd here. + * We want to continue accounting softirq time to ksoftirqd thread + * in that case, so as not to confuse scheduler with a special task + * that do not consume any time, but still wants to run. + */ + if (hardirq_count()) + __this_cpu_add(cpu_hardirq_time, delta); + else if (in_serving_softirq() && curr != this_cpu_ksoftirqd()) + __this_cpu_add(cpu_softirq_time, delta); + + irq_time_write_end(); + local_irq_restore(flags); +} +EXPORT_SYMBOL_GPL(irqtime_account_irq); + +#endif /* CONFIG_IRQ_TIME_ACCOUNTING */ + +#ifdef CONFIG_PARAVIRT +static inline u64 steal_ticks(u64 steal) +{ + if (unlikely(steal > NSEC_PER_SEC)) + return div_u64(steal, TICK_NSEC); + + return __iter_div_u64_rem(steal, TICK_NSEC, &steal); +} +#endif + +static void update_rq_clock_task(struct rq *rq, s64 delta) +{ +/* + * In theory, the compile should just see 0 here, and optimize out the call + * to sched_rt_avg_update. But I don't trust it... + */ +#ifdef CONFIG_IRQ_TIME_ACCOUNTING + s64 irq_delta = irq_time_read(cpu_of(rq)) - rq->prev_irq_time; + + /* + * Since irq_time is only updated on {soft,}irq_exit, we might run into + * this case when a previous update_rq_clock() happened inside a + * {soft,}irq region. + * + * When this happens, we stop ->clock_task and only update the + * prev_irq_time stamp to account for the part that fit, so that a next + * update will consume the rest. This ensures ->clock_task is + * monotonic. + * + * It does however cause some slight miss-attribution of {soft,}irq + * time, a more accurate solution would be to update the irq_time using + * the current rq->clock timestamp, except that would require using + * atomic ops. + */ + if (irq_delta > delta) + irq_delta = delta; + + rq->prev_irq_time += irq_delta; + delta -= irq_delta; +#endif +#ifdef CONFIG_PARAVIRT_TIME_ACCOUNTING + if (static_key_false((¶virt_steal_rq_enabled))) { + s64 steal = paravirt_steal_clock(cpu_of(rq)); + + steal -= rq->prev_steal_time_rq; + + if (unlikely(steal > delta)) + steal = delta; + + rq->prev_steal_time_rq += steal; + + delta -= steal; + } +#endif + + rq->clock_task += delta; +} + +#ifndef nsecs_to_cputime +# define nsecs_to_cputime(__nsecs) nsecs_to_jiffies(__nsecs) +#endif + +#ifdef CONFIG_IRQ_TIME_ACCOUNTING +static void irqtime_account_hi_si(void) +{ + u64 *cpustat = kcpustat_this_cpu->cpustat; + u64 latest_ns; + + latest_ns = nsecs_to_cputime64(this_cpu_read(cpu_hardirq_time)); + if (latest_ns > cpustat[CPUTIME_IRQ]) + cpustat[CPUTIME_IRQ] += (__force u64)cputime_one_jiffy; + + latest_ns = nsecs_to_cputime64(this_cpu_read(cpu_softirq_time)); + if (latest_ns > cpustat[CPUTIME_SOFTIRQ]) + cpustat[CPUTIME_SOFTIRQ] += (__force u64)cputime_one_jiffy; +} +#else /* CONFIG_IRQ_TIME_ACCOUNTING */ + +#define sched_clock_irqtime (0) + +static inline void irqtime_account_hi_si(void) +{ +} +#endif /* CONFIG_IRQ_TIME_ACCOUNTING */ + +static __always_inline bool steal_account_process_tick(void) +{ +#ifdef CONFIG_PARAVIRT + if (static_key_false(¶virt_steal_enabled)) { + u64 steal; + cputime_t steal_ct; + + steal = paravirt_steal_clock(smp_processor_id()); + steal -= this_rq()->prev_steal_time; + + /* + * cputime_t may be less precise than nsecs (eg: if it's + * based on jiffies). Lets cast the result to cputime + * granularity and account the rest on the next rounds. + */ + steal_ct = nsecs_to_cputime(steal); + this_rq()->prev_steal_time += cputime_to_nsecs(steal_ct); + + account_steal_time(steal_ct); + return steal_ct; + } +#endif + return false; +} + +/* + * Accumulate raw cputime values of dead tasks (sig->[us]time) and live + * tasks (sum on group iteration) belonging to @tsk's group. + */ +void thread_group_cputime(struct task_struct *tsk, struct task_cputime *times) +{ + struct signal_struct *sig = tsk->signal; + cputime_t utime, stime; + struct task_struct *t; + unsigned int seq, nextseq; + unsigned long flags; + + rcu_read_lock(); + /* Attempt a lockless read on the first round. */ + nextseq = 0; + do { + seq = nextseq; + flags = read_seqbegin_or_lock_irqsave(&sig->stats_lock, &seq); + times->utime = sig->utime; + times->stime = sig->stime; + times->sum_exec_runtime = sig->sum_sched_runtime; + + for_each_thread(tsk, t) { + task_cputime(t, &utime, &stime); + times->utime += utime; + times->stime += stime; + times->sum_exec_runtime += task_sched_runtime(t); + } + /* If lockless access failed, take the lock. */ + nextseq = 1; + } while (need_seqretry(&sig->stats_lock, seq)); + done_seqretry_irqrestore(&sig->stats_lock, seq, flags); + rcu_read_unlock(); +} + +/* + * On each tick, see what percentage of that tick was attributed to each + * component and add the percentage to the _pc values. Once a _pc value has + * accumulated one tick's worth, account for that. This means the total + * percentage of load components will always be 128 (pseudo 100) per tick. + */ +static void pc_idle_time(struct rq *rq, struct task_struct *idle, unsigned long pc) +{ + u64 *cpustat = kcpustat_this_cpu->cpustat; + + if (atomic_read(&rq->nr_iowait) > 0) { + rq->iowait_pc += pc; + if (rq->iowait_pc >= 128) { + cpustat[CPUTIME_IOWAIT] += (__force u64)cputime_one_jiffy * rq->iowait_pc / 128; + rq->iowait_pc %= 128; + } + } else { + rq->idle_pc += pc; + if (rq->idle_pc >= 128) { + cpustat[CPUTIME_IDLE] += (__force u64)cputime_one_jiffy * rq->idle_pc / 128; + rq->idle_pc %= 128; + } + } + acct_update_integrals(idle); +} + +static void +pc_system_time(struct rq *rq, struct task_struct *p, int hardirq_offset, + unsigned long pc, unsigned long ns) +{ + u64 *cpustat = kcpustat_this_cpu->cpustat; + cputime_t one_jiffy_scaled = cputime_to_scaled(cputime_one_jiffy); + + p->stime_pc += pc; + if (p->stime_pc >= 128) { + int jiffs = p->stime_pc / 128; + + p->stime_pc %= 128; + p->stime += (__force u64)cputime_one_jiffy * jiffs; + p->stimescaled += one_jiffy_scaled * jiffs; + account_group_system_time(p, cputime_one_jiffy * jiffs); + } + p->sched_time += ns; + account_group_exec_runtime(p, ns); + + if (hardirq_count() - hardirq_offset) { + rq->irq_pc += pc; + if (rq->irq_pc >= 128) { + cpustat[CPUTIME_IRQ] += (__force u64)cputime_one_jiffy * rq->irq_pc / 128; + rq->irq_pc %= 128; + } + } else if (in_serving_softirq()) { + rq->softirq_pc += pc; + if (rq->softirq_pc >= 128) { + cpustat[CPUTIME_SOFTIRQ] += (__force u64)cputime_one_jiffy * rq->softirq_pc / 128; + rq->softirq_pc %= 128; + } + } else { + rq->system_pc += pc; + if (rq->system_pc >= 128) { + cpustat[CPUTIME_SYSTEM] += (__force u64)cputime_one_jiffy * rq->system_pc / 128; + rq->system_pc %= 128; + } + } + acct_update_integrals(p); +} + +static void pc_user_time(struct rq *rq, struct task_struct *p, + unsigned long pc, unsigned long ns) +{ + u64 *cpustat = kcpustat_this_cpu->cpustat; + cputime_t one_jiffy_scaled = cputime_to_scaled(cputime_one_jiffy); + + p->utime_pc += pc; + if (p->utime_pc >= 128) { + int jiffs = p->utime_pc / 128; + + p->utime_pc %= 128; + p->utime += (__force u64)cputime_one_jiffy * jiffs; + p->utimescaled += one_jiffy_scaled * jiffs; + account_group_user_time(p, cputime_one_jiffy * jiffs); + } + p->sched_time += ns; + account_group_exec_runtime(p, ns); + + if (this_cpu_ksoftirqd() == p) { + /* + * ksoftirqd time do not get accounted in cpu_softirq_time. + * So, we have to handle it separately here. + */ + rq->softirq_pc += pc; + if (rq->softirq_pc >= 128) { + cpustat[CPUTIME_SOFTIRQ] += (__force u64)cputime_one_jiffy * rq->softirq_pc / 128; + rq->softirq_pc %= 128; + } + } + + if (task_nice(p) > 0 || idleprio_task(p)) { + rq->nice_pc += pc; + if (rq->nice_pc >= 128) { + cpustat[CPUTIME_NICE] += (__force u64)cputime_one_jiffy * rq->nice_pc / 128; + rq->nice_pc %= 128; + } + } else { + rq->user_pc += pc; + if (rq->user_pc >= 128) { + cpustat[CPUTIME_USER] += (__force u64)cputime_one_jiffy * rq->user_pc / 128; + rq->user_pc %= 128; + } + } + acct_update_integrals(p); +} + +/* + * Convert nanoseconds to pseudo percentage of one tick. Use 128 for fast + * shifts instead of 100 + */ +#define NS_TO_PC(NS) (NS * 128 / JIFFY_NS) + +/* + * This is called on clock ticks. + * Bank in p->sched_time the ns elapsed since the last tick or switch. + * CPU scheduler quota accounting is also performed here in microseconds. + */ +static void +update_cpu_clock_tick(struct rq *rq, struct task_struct *p) +{ + long account_ns = rq->clock_task - rq->rq_last_ran; + struct task_struct *idle = rq->idle; + unsigned long account_pc; + + if (unlikely(account_ns < 0) || steal_account_process_tick()) + goto ts_account; + + account_pc = NS_TO_PC(account_ns); + + /* Accurate tick timekeeping */ + if (user_mode(get_irq_regs())) + pc_user_time(rq, p, account_pc, account_ns); + else if (p != idle || (irq_count() != HARDIRQ_OFFSET)) + pc_system_time(rq, p, HARDIRQ_OFFSET, + account_pc, account_ns); + else + pc_idle_time(rq, idle, account_pc); + + if (sched_clock_irqtime) + irqtime_account_hi_si(); + +ts_account: + /* time_slice accounting is done in usecs to avoid overflow on 32bit */ + if (rq->rq_policy != SCHED_FIFO && p != idle) { + s64 time_diff = rq->clock - rq->timekeep_clock; + + niffy_diff(&time_diff, 1); + rq->rq_time_slice -= NS_TO_US(time_diff); + } + + rq->rq_last_ran = rq->clock_task; + rq->timekeep_clock = rq->clock; +} + +/* + * This is called on context switches. + * Bank in p->sched_time the ns elapsed since the last tick or switch. + * CPU scheduler quota accounting is also performed here in microseconds. + */ +static void +update_cpu_clock_switch(struct rq *rq, struct task_struct *p) +{ + long account_ns = rq->clock_task - rq->rq_last_ran; + struct task_struct *idle = rq->idle; + unsigned long account_pc; + + if (unlikely(account_ns < 0)) + goto ts_account; + + account_pc = NS_TO_PC(account_ns); + + /* Accurate subtick timekeeping */ + if (p != idle) { + pc_user_time(rq, p, account_pc, account_ns); + } + else + pc_idle_time(rq, idle, account_pc); + +ts_account: + /* time_slice accounting is done in usecs to avoid overflow on 32bit */ + if (rq->rq_policy != SCHED_FIFO && p != idle) { + s64 time_diff = rq->clock - rq->timekeep_clock; + + niffy_diff(&time_diff, 1); + rq->rq_time_slice -= NS_TO_US(time_diff); + } + + rq->rq_last_ran = rq->clock_task; + rq->timekeep_clock = rq->clock; +} + +/* + * Return any ns on the sched_clock that have not yet been accounted in + * @p in case that task is currently running. + * + * Called with task_grq_lock() held. + */ +static inline u64 do_task_delta_exec(struct task_struct *p, struct rq *rq) +{ + u64 ns = 0; + + /* + * Must be ->curr _and_ ->on_rq. If dequeued, we would + * project cycles that may never be accounted to this + * thread, breaking clock_gettime(). + */ + if (p == rq->curr && p->on_rq) { + update_clocks(rq); + ns = rq->clock_task - rq->rq_last_ran; + if (unlikely((s64)ns < 0)) + ns = 0; + } + + return ns; +} + +/* + * Return accounted runtime for the task. + * Return separately the current's pending runtime that have not been + * accounted yet. + * + */ +unsigned long long task_sched_runtime(struct task_struct *p) +{ + unsigned long flags; + struct rq *rq; + u64 ns; + +#if defined(CONFIG_64BIT) && defined(CONFIG_SMP) + /* + * 64-bit doesn't need locks to atomically read a 64bit value. + * So we have a optimization chance when the task's delta_exec is 0. + * Reading ->on_cpu is racy, but this is ok. + * + * If we race with it leaving cpu, we'll take a lock. So we're correct. + * If we race with it entering cpu, unaccounted time is 0. This is + * indistinguishable from the read occurring a few cycles earlier. + * If we see ->on_cpu without ->on_rq, the task is leaving, and has + * been accounted, so we're correct here as well. + */ + if (!p->on_cpu || !p->on_rq) + return tsk_seruntime(p); +#endif + + rq = task_grq_lock(p, &flags); + ns = p->sched_time + do_task_delta_exec(p, rq); + task_grq_unlock(&flags); + + return ns; +} + +/* Compatibility crap */ +void account_user_time(struct task_struct *p, cputime_t cputime, + cputime_t cputime_scaled) +{ +} + +void account_idle_time(cputime_t cputime) +{ +} + +/* + * Account guest cpu time to a process. + * @p: the process that the cpu time gets accounted to + * @cputime: the cpu time spent in virtual machine since the last update + * @cputime_scaled: cputime scaled by cpu frequency + */ +static void account_guest_time(struct task_struct *p, cputime_t cputime, + cputime_t cputime_scaled) +{ + u64 *cpustat = kcpustat_this_cpu->cpustat; + + /* Add guest time to process. */ + p->utime += (__force u64)cputime; + p->utimescaled += (__force u64)cputime_scaled; + account_group_user_time(p, cputime); + p->gtime += (__force u64)cputime; + + /* Add guest time to cpustat. */ + if (task_nice(p) > 0) { + cpustat[CPUTIME_NICE] += (__force u64)cputime; + cpustat[CPUTIME_GUEST_NICE] += (__force u64)cputime; + } else { + cpustat[CPUTIME_USER] += (__force u64)cputime; + cpustat[CPUTIME_GUEST] += (__force u64)cputime; + } +} + +/* + * Account system cpu time to a process and desired cpustat field + * @p: the process that the cpu time gets accounted to + * @cputime: the cpu time spent in kernel space since the last update + * @cputime_scaled: cputime scaled by cpu frequency + * @target_cputime64: pointer to cpustat field that has to be updated + */ +static inline +void __account_system_time(struct task_struct *p, cputime_t cputime, + cputime_t cputime_scaled, cputime64_t *target_cputime64) +{ + /* Add system time to process. */ + p->stime += (__force u64)cputime; + p->stimescaled += (__force u64)cputime_scaled; + account_group_system_time(p, cputime); + + /* Add system time to cpustat. */ + *target_cputime64 += (__force u64)cputime; + + /* Account for system time used */ + acct_update_integrals(p); +} + +/* + * Account system cpu time to a process. + * @p: the process that the cpu time gets accounted to + * @hardirq_offset: the offset to subtract from hardirq_count() + * @cputime: the cpu time spent in kernel space since the last update + * @cputime_scaled: cputime scaled by cpu frequency + * This is for guest only now. + */ +void account_system_time(struct task_struct *p, int hardirq_offset, + cputime_t cputime, cputime_t cputime_scaled) +{ + + if ((p->flags & PF_VCPU) && (irq_count() - hardirq_offset == 0)) + account_guest_time(p, cputime, cputime_scaled); +} + +/* + * Account for involuntary wait time. + * @steal: the cpu time spent in involuntary wait + */ +void account_steal_time(cputime_t cputime) +{ + u64 *cpustat = kcpustat_this_cpu->cpustat; + + cpustat[CPUTIME_STEAL] += (__force u64)cputime; +} + +/* + * Account for idle time. + * @cputime: the cpu time spent in idle wait + */ +static void account_idle_times(cputime_t cputime) +{ + u64 *cpustat = kcpustat_this_cpu->cpustat; + struct rq *rq = this_rq(); + + if (atomic_read(&rq->nr_iowait) > 0) + cpustat[CPUTIME_IOWAIT] += (__force u64)cputime; + else + cpustat[CPUTIME_IDLE] += (__force u64)cputime; +} + +#ifndef CONFIG_VIRT_CPU_ACCOUNTING_NATIVE + +void account_process_tick(struct task_struct *p, int user_tick) +{ +} + +/* + * Account multiple ticks of steal time. + * @p: the process from which the cpu time has been stolen + * @ticks: number of stolen ticks + */ +void account_steal_ticks(unsigned long ticks) +{ + account_steal_time(jiffies_to_cputime(ticks)); +} + +/* + * Account multiple ticks of idle time. + * @ticks: number of stolen ticks + */ +void account_idle_ticks(unsigned long ticks) +{ + account_idle_times(jiffies_to_cputime(ticks)); +} +#endif + +static inline void grq_iso_lock(void) + __acquires(grq.iso_lock) +{ + raw_spin_lock(&grq.iso_lock); +} + +static inline void grq_iso_unlock(void) + __releases(grq.iso_lock) +{ + raw_spin_unlock(&grq.iso_lock); +} + +/* + * Functions to test for when SCHED_ISO tasks have used their allocated + * quota as real time scheduling and convert them back to SCHED_NORMAL. + * Where possible, the data is tested lockless, to avoid grabbing iso_lock + * because the occasional inaccurate result won't matter. However the + * tick data is only ever modified under lock. iso_refractory is only simply + * set to 0 or 1 so it's not worth grabbing the lock yet again for that. + */ +static bool set_iso_refractory(void) +{ + grq.iso_refractory = true; + return grq.iso_refractory; +} + +static bool clear_iso_refractory(void) +{ + grq.iso_refractory = false; + return grq.iso_refractory; +} + +/* + * Test if SCHED_ISO tasks have run longer than their alloted period as RT + * tasks and set the refractory flag if necessary. There is 10% hysteresis + * for unsetting the flag. 115/128 is ~90/100 as a fast shift instead of a + * slow division. + */ +static bool test_ret_isorefractory(struct rq *rq) +{ + if (likely(!grq.iso_refractory)) { + if (grq.iso_ticks > ISO_PERIOD * sched_iso_cpu) + return set_iso_refractory(); + } else { + if (grq.iso_ticks < ISO_PERIOD * (sched_iso_cpu * 115 / 128)) + return clear_iso_refractory(); + } + return grq.iso_refractory; +} + +static void iso_tick(void) +{ + grq_iso_lock(); + grq.iso_ticks += 100; + grq_iso_unlock(); +} + +/* No SCHED_ISO task was running so decrease rq->iso_ticks */ +static inline void no_iso_tick(void) +{ + if (grq.iso_ticks) { + grq_iso_lock(); + grq.iso_ticks -= grq.iso_ticks / ISO_PERIOD + 1; + if (unlikely(grq.iso_refractory && grq.iso_ticks < + ISO_PERIOD * (sched_iso_cpu * 115 / 128))) + clear_iso_refractory(); + grq_iso_unlock(); + } +} + +/* This manages tasks that have run out of timeslice during a scheduler_tick */ +static void task_running_tick(struct rq *rq) +{ + struct task_struct *p; + + /* + * If a SCHED_ISO task is running we increment the iso_ticks. In + * order to prevent SCHED_ISO tasks from causing starvation in the + * presence of true RT tasks we account those as iso_ticks as well. + */ + if ((rt_queue(rq) || (iso_queue(rq) && !grq.iso_refractory))) { + if (grq.iso_ticks <= (ISO_PERIOD * 128) - 128) + iso_tick(); + } else + no_iso_tick(); + + if (iso_queue(rq)) { + if (unlikely(test_ret_isorefractory(rq))) { + if (rq_running_iso(rq)) { + /* + * SCHED_ISO task is running as RT and limit + * has been hit. Force it to reschedule as + * SCHED_NORMAL by zeroing its time_slice + */ + rq->rq_time_slice = 0; + } + } + } + + /* SCHED_FIFO tasks never run out of timeslice. */ + if (rq->rq_policy == SCHED_FIFO) + return; + /* + * Tasks that were scheduled in the first half of a tick are not + * allowed to run into the 2nd half of the next tick if they will + * run out of time slice in the interim. Otherwise, if they have + * less than RESCHED_US μs of time slice left they will be rescheduled. + */ + if (rq->dither) { + if (rq->rq_time_slice > HALF_JIFFY_US) + return; + else + rq->rq_time_slice = 0; + } else if (rq->rq_time_slice >= RESCHED_US) + return; + + /* p->time_slice < RESCHED_US. We only modify task_struct under grq lock */ + p = rq->curr; + + grq_lock(); + requeue_task(p); + __set_tsk_resched(p); + grq_unlock(); +} + +/* + * This function gets called by the timer code, with HZ frequency. + * We call it with interrupts disabled. The data modified is all + * local to struct rq so we don't need to grab grq lock. + */ +void scheduler_tick(void) +{ + int cpu __maybe_unused = smp_processor_id(); + struct rq *rq = cpu_rq(cpu); + + sched_clock_tick(); + /* grq lock not grabbed, so only update rq clock */ + update_rq_clock(rq); + update_cpu_clock_tick(rq, rq->curr); + if (!rq_idle(rq)) + task_running_tick(rq); + else + no_iso_tick(); + rq->last_tick = rq->clock; + perf_event_task_tick(); +} + +notrace unsigned long get_parent_ip(unsigned long addr) +{ + if (in_lock_functions(addr)) { + addr = CALLER_ADDR2; + if (in_lock_functions(addr)) + addr = CALLER_ADDR3; + } + return addr; +} + +#if defined(CONFIG_PREEMPT) && (defined(CONFIG_DEBUG_PREEMPT) || \ + defined(CONFIG_PREEMPT_TRACER)) +void preempt_count_add(int val) +{ +#ifdef CONFIG_DEBUG_PREEMPT + /* + * Underflow? + */ + if (DEBUG_LOCKS_WARN_ON((preempt_count() < 0))) + return; +#endif + __preempt_count_add(val); +#ifdef CONFIG_DEBUG_PREEMPT + /* + * Spinlock count overflowing soon? + */ + DEBUG_LOCKS_WARN_ON((preempt_count() & PREEMPT_MASK) >= + PREEMPT_MASK - 10); +#endif + if (preempt_count() == val) { + unsigned long ip = get_parent_ip(CALLER_ADDR1); +#ifdef CONFIG_DEBUG_PREEMPT + current->preempt_disable_ip = ip; +#endif + trace_preempt_off(CALLER_ADDR0, ip); + } +} +EXPORT_SYMBOL(preempt_count_add); +NOKPROBE_SYMBOL(preempt_count_add); + +void preempt_count_sub(int val) +{ +#ifdef CONFIG_DEBUG_PREEMPT + /* + * Underflow? + */ + if (DEBUG_LOCKS_WARN_ON(val > preempt_count())) + return; + /* + * Is the spinlock portion underflowing? + */ + if (DEBUG_LOCKS_WARN_ON((val < PREEMPT_MASK) && + !(preempt_count() & PREEMPT_MASK))) + return; +#endif + + if (preempt_count() == val) + trace_preempt_on(CALLER_ADDR0, get_parent_ip(CALLER_ADDR1)); + __preempt_count_sub(val); +} +EXPORT_SYMBOL(preempt_count_sub); +NOKPROBE_SYMBOL(preempt_count_sub); +#endif + +/* + * Deadline is "now" in niffies + (offset by priority). Setting the deadline + * is the key to everything. It distributes cpu fairly amongst tasks of the + * same nice value, it proportions cpu according to nice level, it means the + * task that last woke up the longest ago has the earliest deadline, thus + * ensuring that interactive tasks get low latency on wake up. The CPU + * proportion works out to the square of the virtual deadline difference, so + * this equation will give nice 19 3% CPU compared to nice 0. + */ +static inline u64 prio_deadline_diff(int user_prio) +{ + return (prio_ratios[user_prio] * rr_interval * (MS_TO_NS(1) / 128)); +} + +static inline u64 task_deadline_diff(struct task_struct *p) +{ + return prio_deadline_diff(TASK_USER_PRIO(p)); +} + +static inline u64 static_deadline_diff(int static_prio) +{ + return prio_deadline_diff(USER_PRIO(static_prio)); +} + +static inline int longest_deadline_diff(void) +{ + return prio_deadline_diff(39); +} + +static inline int ms_longest_deadline_diff(void) +{ + return NS_TO_MS(longest_deadline_diff()); +} + +/* + * The time_slice is only refilled when it is empty and that is when we set a + * new deadline. + */ +static void time_slice_expired(struct task_struct *p) +{ + p->time_slice = timeslice(); + p->deadline = grq.niffies + task_deadline_diff(p); +#ifdef CONFIG_SMT_NICE + if (!p->mm) + p->smt_bias = 0; + else if (rt_task(p)) + p->smt_bias = 1 << 30; + else if (task_running_iso(p)) + p->smt_bias = 1 << 29; + else if (idleprio_task(p)) { + if (task_running_idle(p)) + p->smt_bias = 0; + else + p->smt_bias = 1; + } else if (--p->smt_bias < 1) + p->smt_bias = MAX_PRIO - p->static_prio; +#endif +} + +/* + * Timeslices below RESCHED_US are considered as good as expired as there's no + * point rescheduling when there's so little time left. SCHED_BATCH tasks + * have been flagged be not latency sensitive and likely to be fully CPU + * bound so every time they're rescheduled they have their time_slice + * refilled, but get a new later deadline to have little effect on + * SCHED_NORMAL tasks. + + */ +static inline void check_deadline(struct task_struct *p) +{ + if (p->time_slice < RESCHED_US || batch_task(p)) + time_slice_expired(p); +} + +#define BITOP_WORD(nr) ((nr) / BITS_PER_LONG) + +/* + * Scheduler queue bitmap specific find next bit. + */ +static inline unsigned long +next_sched_bit(const unsigned long *addr, unsigned long offset) +{ + const unsigned long *p; + unsigned long result; + unsigned long size; + unsigned long tmp; + + size = PRIO_LIMIT; + if (offset >= size) + return size; + + p = addr + BITOP_WORD(offset); + result = offset & ~(BITS_PER_LONG-1); + size -= result; + offset %= BITS_PER_LONG; + if (offset) { + tmp = *(p++); + tmp &= (~0UL << offset); + if (size < BITS_PER_LONG) + goto found_first; + if (tmp) + goto found_middle; + size -= BITS_PER_LONG; + result += BITS_PER_LONG; + } + while (size & ~(BITS_PER_LONG-1)) { + if ((tmp = *(p++))) + goto found_middle; + result += BITS_PER_LONG; + size -= BITS_PER_LONG; + } + if (!size) + return result; + tmp = *p; + +found_first: + tmp &= (~0UL >> (BITS_PER_LONG - size)); + if (tmp == 0UL) /* Are any bits set? */ + return result + size; /* Nope. */ +found_middle: + return result + __ffs(tmp); +} + +/* + * O(n) lookup of all tasks in the global runqueue. The real brainfuck + * of lock contention and O(n). It's not really O(n) as only the queued, + * but not running tasks are scanned, and is O(n) queued in the worst case + * scenario only because the right task can be found before scanning all of + * them. + * Tasks are selected in this order: + * Real time tasks are selected purely by their static priority and in the + * order they were queued, so the lowest value idx, and the first queued task + * of that priority value is chosen. + * If no real time tasks are found, the SCHED_ISO priority is checked, and + * all SCHED_ISO tasks have the same priority value, so they're selected by + * the earliest deadline value. + * If no SCHED_ISO tasks are found, SCHED_NORMAL tasks are selected by the + * earliest deadline. + * Finally if no SCHED_NORMAL tasks are found, SCHED_IDLEPRIO tasks are + * selected by the earliest deadline. + */ +static inline struct +task_struct *earliest_deadline_task(struct rq *rq, int cpu, struct task_struct *idle) +{ + struct task_struct *edt = NULL; + unsigned long idx = -1; + + do { + struct list_head *queue; + struct task_struct *p; + u64 earliest_deadline; + + idx = next_sched_bit(grq.prio_bitmap, ++idx); + if (idx >= PRIO_LIMIT) + return idle; + queue = grq.queue + idx; + + if (idx < MAX_RT_PRIO) { + /* We found an rt task */ + list_for_each_entry(p, queue, run_list) { + /* Make sure cpu affinity is ok */ + if (needs_other_cpu(p, cpu)) + continue; + edt = p; + goto out_take; + } + /* + * None of the RT tasks at this priority can run on + * this cpu + */ + continue; + } + + /* + * No rt tasks. Find the earliest deadline task. Now we're in + * O(n) territory. + */ + earliest_deadline = ~0ULL; + list_for_each_entry(p, queue, run_list) { + u64 dl; + + /* Make sure cpu affinity is ok */ + if (needs_other_cpu(p, cpu)) + continue; + +#ifdef CONFIG_SMT_NICE + if (!smt_should_schedule(p, cpu)) + continue; +#endif + /* + * Soft affinity happens here by not scheduling a task + * with its sticky flag set that ran on a different CPU + * last when the CPU is scaling, or by greatly biasing + * against its deadline when not, based on cpu cache + * locality. + */ + if (sched_interactive) + dl = p->deadline; + else { + int tcpu = task_cpu(p); + + if (tcpu != cpu && task_sticky(p) && scaling_rq(rq)) + continue; + dl = p->deadline << locality_diff(tcpu, rq); + } + + if (deadline_before(dl, earliest_deadline)) { + earliest_deadline = dl; + edt = p; + } + } + } while (!edt); + +out_take: + take_task(cpu, edt); + return edt; +} + + +/* + * Print scheduling while atomic bug: + */ +static noinline void __schedule_bug(struct task_struct *prev) +{ + if (oops_in_progress) + return; + + printk(KERN_ERR "BUG: scheduling while atomic: %s/%d/0x%08x\n", + prev->comm, prev->pid, preempt_count()); + + debug_show_held_locks(prev); + print_modules(); + if (irqs_disabled()) + print_irqtrace_events(prev); +#ifdef CONFIG_DEBUG_PREEMPT + if (in_atomic_preempt_off()) { + pr_err("Preemption disabled at:"); + print_ip_sym(current->preempt_disable_ip); + pr_cont("\n"); + } +#endif + dump_stack(); + add_taint(TAINT_WARN, LOCKDEP_STILL_OK); +} + +/* + * Various schedule()-time debugging checks and statistics: + */ +static inline void schedule_debug(struct task_struct *prev) +{ +#ifdef CONFIG_SCHED_STACK_END_CHECK + BUG_ON(task_stack_end_corrupted(prev)); +#endif + + if (unlikely(in_atomic_preempt_off())) { + __schedule_bug(prev); + preempt_count_set(PREEMPT_DISABLED); + } + rcu_sleep_check(); + + profile_hit(SCHED_PROFILING, __builtin_return_address(0)); + + schedstat_inc(this_rq(), sched_count); +} + +/* + * The currently running task's information is all stored in rq local data + * which is only modified by the local CPU, thereby allowing the data to be + * changed without grabbing the grq lock. + */ +static inline void set_rq_task(struct rq *rq, struct task_struct *p) +{ + rq->rq_time_slice = p->time_slice; + rq->rq_deadline = p->deadline; + rq->rq_last_ran = p->last_ran = rq->clock_task; + rq->rq_policy = p->policy; + rq->rq_prio = p->prio; +#ifdef CONFIG_SMT_NICE + rq->rq_mm = p->mm; + rq->rq_smt_bias = p->smt_bias; +#endif + if (p != rq->idle) + rq->rq_running = true; + else + rq->rq_running = false; +} + +static void reset_rq_task(struct rq *rq, struct task_struct *p) +{ + rq->rq_policy = p->policy; + rq->rq_prio = p->prio; +#ifdef CONFIG_SMT_NICE + rq->rq_smt_bias = p->smt_bias; +#endif +} + +#ifdef CONFIG_SMT_NICE +/* Iterate over smt siblings when we've scheduled a process on cpu and decide + * whether they should continue running or be descheduled. */ +static void check_smt_siblings(int cpu) +{ + int other_cpu; + + for_each_cpu(other_cpu, thread_cpumask(cpu)) { + struct task_struct *p; + struct rq *rq; + + if (other_cpu == cpu) + continue; + rq = cpu_rq(other_cpu); + if (rq_idle(rq)) + continue; + if (!rq->online) + continue; + p = rq->curr; + if (!smt_should_schedule(p, cpu)) { + set_tsk_need_resched(p); + smp_send_reschedule(other_cpu); + } + } +} + +static void wake_smt_siblings(int cpu) +{ + int other_cpu; + + if (!queued_notrunning()) + return; + + for_each_cpu(other_cpu, thread_cpumask(cpu)) { + struct rq *rq; + + if (other_cpu == cpu) + continue; + rq = cpu_rq(other_cpu); + if (rq_idle(rq)) { + struct task_struct *p = rq->curr; + + set_tsk_need_resched(p); + smp_send_reschedule(other_cpu); + } + } +} +#else +static void check_smt_siblings(int __maybe_unused cpu) {} +static void wake_smt_siblings(int __maybe_unused cpu) {} +#endif + +/* + * schedule() is the main scheduler function. + * + * The main means of driving the scheduler and thus entering this function are: + * + * 1. Explicit blocking: mutex, semaphore, waitqueue, etc. + * + * 2. TIF_NEED_RESCHED flag is checked on interrupt and userspace return + * paths. For example, see arch/x86/entry_64.S. + * + * To drive preemption between tasks, the scheduler sets the flag in timer + * interrupt handler scheduler_tick(). + * + * 3. Wakeups don't really cause entry into schedule(). They add a + * task to the run-queue and that's it. + * + * Now, if the new task added to the run-queue preempts the current + * task, then the wakeup sets TIF_NEED_RESCHED and schedule() gets + * called on the nearest possible occasion: + * + * - If the kernel is preemptible (CONFIG_PREEMPT=y): + * + * - in syscall or exception context, at the next outmost + * preempt_enable(). (this might be as soon as the wake_up()'s + * spin_unlock()!) + * + * - in IRQ context, return from interrupt-handler to + * preemptible context + * + * - If the kernel is not preemptible (CONFIG_PREEMPT is not set) + * then at the next: + * + * - cond_resched() call + * - explicit schedule() call + * - return from syscall or exception to user-space + * - return from interrupt-handler to user-space + * + * WARNING: must be called with preemption disabled! + */ +static void __sched notrace __schedule(bool preempt) +{ + struct task_struct *prev, *next, *idle; + unsigned long *switch_count; + bool deactivate = false; + struct rq *rq; + int cpu; + + cpu = smp_processor_id(); + rq = cpu_rq(cpu); + prev = rq->curr; + + /* + * do_exit() calls schedule() with preemption disabled as an exception; + * however we must fix that up, otherwise the next task will see an + * inconsistent (higher) preempt count. + * + * It also avoids the below schedule_debug() test from complaining + * about this. + */ + if (unlikely(prev->state == TASK_DEAD)) + preempt_enable_no_resched_notrace(); + + schedule_debug(prev); + + local_irq_disable(); + rcu_note_context_switch(); + + /* + * Make sure that signal_pending_state()->signal_pending() below + * can't be reordered with __set_current_state(TASK_INTERRUPTIBLE) + * done by the caller to avoid the race with signal_wake_up(). + */ + smp_mb__before_spinlock(); + grq_lock(); + + switch_count = &prev->nivcsw; + if (!preempt && prev->state) { + if (unlikely(signal_pending_state(prev->state, prev))) { + prev->state = TASK_RUNNING; + } else { + deactivate = true; + prev->on_rq = 0; + + /* + * If a worker is going to sleep, notify and + * ask workqueue whether it wants to wake up a + * task to maintain concurrency. If so, wake + * up the task. + */ + if (prev->flags & PF_WQ_WORKER) { + struct task_struct *to_wakeup; + + to_wakeup = wq_worker_sleeping(prev, cpu); + if (to_wakeup) { + /* This shouldn't happen, but does */ + if (unlikely(to_wakeup == prev)) + deactivate = false; + else + try_to_wake_up_local(to_wakeup); + } + } + } + switch_count = &prev->nvcsw; + } + + update_clocks(rq); + update_cpu_clock_switch(rq, prev); + if (rq->clock - rq->last_tick > HALF_JIFFY_NS) + rq->dither = false; + else + rq->dither = true; + + clear_tsk_need_resched(prev); + clear_preempt_need_resched(); + + idle = rq->idle; + if (idle != prev) { + /* Update all the information stored on struct rq */ + prev->time_slice = rq->rq_time_slice; + prev->deadline = rq->rq_deadline; + check_deadline(prev); + prev->last_ran = rq->clock_task; + + /* Task changed affinity off this CPU */ + if (likely(!needs_other_cpu(prev, cpu))) { + if (!deactivate) { + if (!queued_notrunning()) { + /* + * We now know prev is the only thing that is + * awaiting CPU so we can bypass rechecking for + * the earliest deadline task and just run it + * again. + */ + set_rq_task(rq, prev); + check_smt_siblings(cpu); + grq_unlock_irq(); + goto rerun_prev_unlocked; + } else + swap_sticky(rq, cpu, prev); + } + } + return_task(prev, rq, deactivate); + } + + if (unlikely(!queued_notrunning())) { + /* + * This CPU is now truly idle as opposed to when idle is + * scheduled as a high priority task in its own right. + */ + next = idle; + schedstat_inc(rq, sched_goidle); + set_cpuidle_map(cpu); + } else { + next = earliest_deadline_task(rq, cpu, idle); + if (likely(next->prio != PRIO_LIMIT)) + clear_cpuidle_map(cpu); + else + set_cpuidle_map(cpu); + } + + if (likely(prev != next)) { + /* + * Don't reschedule an idle task or deactivated tasks + */ + if (prev != idle && !deactivate) + resched_suitable_idle(prev); + /* + * Don't stick tasks when a real time task is going to run as + * they may literally get stuck. + */ + if (rt_task(next)) + unstick_task(rq, prev); + set_rq_task(rq, next); + if (next != idle) + check_smt_siblings(cpu); + else + wake_smt_siblings(cpu); + grq.nr_switches++; + prev->on_cpu = false; + next->on_cpu = true; + rq->curr = next; + ++*switch_count; + + trace_sched_switch(preempt, prev, next); + rq = context_switch(rq, prev, next); /* unlocks the grq */ + cpu = cpu_of(rq); + idle = rq->idle; + } else { + check_smt_siblings(cpu); + grq_unlock_irq(); + } + +rerun_prev_unlocked: + return; +} + +static inline void sched_submit_work(struct task_struct *tsk) +{ + if (!tsk->state || tsk_is_pi_blocked(tsk) || + preempt_count() || + signal_pending_state(tsk->state, tsk)) + return; + + /* + * If we are going to sleep and we have plugged IO queued, + * make sure to submit it to avoid deadlocks. + */ + if (blk_needs_flush_plug(tsk)) + blk_schedule_flush_plug(tsk); +} + +asmlinkage __visible void __sched schedule(void) +{ + struct task_struct *tsk = current; + + sched_submit_work(tsk); + do { + preempt_disable(); + __schedule(false); + sched_preempt_enable_no_resched(); + } while (need_resched()); +} + +EXPORT_SYMBOL(schedule); + +#ifdef CONFIG_CONTEXT_TRACKING +asmlinkage __visible void __sched schedule_user(void) +{ + /* + * If we come here after a random call to set_need_resched(), + * or we have been woken up remotely but the IPI has not yet arrived, + * we haven't yet exited the RCU idle mode. Do it here manually until + * we find a better solution. + * + * NB: There are buggy callers of this function. Ideally we + * should warn if prev_state != IN_USER, but that will trigger + * too frequently to make sense yet. + */ + enum ctx_state prev_state = exception_enter(); + schedule(); + exception_exit(prev_state); +} +#endif + +/** + * schedule_preempt_disabled - called with preemption disabled + * + * Returns with preemption disabled. Note: preempt_count must be 1 + */ +void __sched schedule_preempt_disabled(void) +{ + sched_preempt_enable_no_resched(); + schedule(); + preempt_disable(); +} + +static void __sched notrace preempt_schedule_common(void) +{ + do { + preempt_disable_notrace(); + __schedule(true); + preempt_enable_no_resched_notrace(); + + /* + * Check again in case we missed a preemption opportunity + * between schedule and now. + */ + } while (need_resched()); +} + +#ifdef CONFIG_PREEMPT +/* + * this is the entry point to schedule() from in-kernel preemption + * off of preempt_enable. Kernel preemptions off return from interrupt + * occur there and call schedule directly. + */ +asmlinkage __visible void __sched notrace preempt_schedule(void) +{ + /* + * If there is a non-zero preempt_count or interrupts are disabled, + * we do not want to preempt the current task. Just return.. + */ + if (likely(!preemptible())) + return; + + preempt_schedule_common(); +} +NOKPROBE_SYMBOL(preempt_schedule); +EXPORT_SYMBOL(preempt_schedule); + +/** + * preempt_schedule_notrace - preempt_schedule called by tracing + * + * The tracing infrastructure uses preempt_enable_notrace to prevent + * recursion and tracing preempt enabling caused by the tracing + * infrastructure itself. But as tracing can happen in areas coming + * from userspace or just about to enter userspace, a preempt enable + * can occur before user_exit() is called. This will cause the scheduler + * to be called when the system is still in usermode. + * + * To prevent this, the preempt_enable_notrace will use this function + * instead of preempt_schedule() to exit user context if needed before + * calling the scheduler. + */ +asmlinkage __visible void __sched notrace preempt_schedule_notrace(void) +{ + enum ctx_state prev_ctx; + + if (likely(!preemptible())) + return; + + do { + preempt_disable_notrace(); + /* + * Needs preempt disabled in case user_exit() is traced + * and the tracer calls preempt_enable_notrace() causing + * an infinite recursion. + */ + prev_ctx = exception_enter(); + __schedule(true); + exception_exit(prev_ctx); + + preempt_enable_no_resched_notrace(); + } while (need_resched()); +} +EXPORT_SYMBOL_GPL(preempt_schedule_notrace); + +#endif /* CONFIG_PREEMPT */ + +/* + * this is the entry point to schedule() from kernel preemption + * off of irq context. + * Note, that this is called and return with irqs disabled. This will + * protect us against recursive calling from irq. + */ +asmlinkage __visible void __sched preempt_schedule_irq(void) +{ + enum ctx_state prev_state; + + /* Catch callers which need to be fixed */ + BUG_ON(preempt_count() || !irqs_disabled()); + + prev_state = exception_enter(); + + do { + preempt_disable(); + local_irq_enable(); + __schedule(true); + local_irq_disable(); + sched_preempt_enable_no_resched(); + } while (need_resched()); + + exception_exit(prev_state); +} + +int default_wake_function(wait_queue_t *curr, unsigned mode, int wake_flags, + void *key) +{ + return try_to_wake_up(curr->private, mode, wake_flags); +} +EXPORT_SYMBOL(default_wake_function); + +#ifdef CONFIG_RT_MUTEXES + +/* + * rt_mutex_setprio - set the current priority of a task + * @p: task + * @prio: prio value (kernel-internal form) + * + * This function changes the 'effective' priority of a task. It does + * not touch ->normal_prio like __setscheduler(). + * + * Used by the rt_mutex code to implement priority inheritance + * logic. Call site only calls if the priority of the task changed. + */ +void rt_mutex_setprio(struct task_struct *p, int prio) +{ + unsigned long flags; + int queued, oldprio; + struct rq *rq; + + BUG_ON(prio < 0 || prio > MAX_PRIO); + + rq = task_grq_lock(p, &flags); + + /* + * Idle task boosting is a nono in general. There is one + * exception, when PREEMPT_RT and NOHZ is active: + * + * The idle task calls get_next_timer_interrupt() and holds + * the timer wheel base->lock on the CPU and another CPU wants + * to access the timer (probably to cancel it). We can safely + * ignore the boosting request, as the idle CPU runs this code + * with interrupts disabled and will complete the lock + * protected section without being interrupted. So there is no + * real need to boost. + */ + if (unlikely(p == rq->idle)) { + WARN_ON(p != rq->curr); + WARN_ON(p->pi_blocked_on); + goto out_unlock; + } + + trace_sched_pi_setprio(p, prio); + oldprio = p->prio; + queued = task_queued(p); + if (queued) + dequeue_task(p); + p->prio = prio; + if (task_running(p) && prio > oldprio) + resched_task(p); + if (queued) { + enqueue_task(p, rq); + try_preempt(p, rq); + } + +out_unlock: + task_grq_unlock(&flags); +} + +#endif + +/* + * Adjust the deadline for when the priority is to change, before it's + * changed. + */ +static inline void adjust_deadline(struct task_struct *p, int new_prio) +{ + p->deadline += static_deadline_diff(new_prio) - task_deadline_diff(p); +} + +void set_user_nice(struct task_struct *p, long nice) +{ + int queued, new_static, old_static; + unsigned long flags; + struct rq *rq; + + if (task_nice(p) == nice || nice < MIN_NICE || nice > MAX_NICE) + return; + new_static = NICE_TO_PRIO(nice); + /* + * We have to be careful, if called from sys_setpriority(), + * the task might be in the middle of scheduling on another CPU. + */ + rq = time_task_grq_lock(p, &flags); + /* + * The RT priorities are set via sched_setscheduler(), but we still + * allow the 'normal' nice value to be set - but as expected + * it wont have any effect on scheduling until the task is + * not SCHED_NORMAL/SCHED_BATCH: + */ + if (has_rt_policy(p)) { + p->static_prio = new_static; + goto out_unlock; + } + queued = task_queued(p); + if (queued) + dequeue_task(p); + + adjust_deadline(p, new_static); + old_static = p->static_prio; + p->static_prio = new_static; + p->prio = effective_prio(p); + + if (queued) { + enqueue_task(p, rq); + if (new_static < old_static) + try_preempt(p, rq); + } else if (task_running(p)) { + reset_rq_task(rq, p); + if (old_static < new_static) + resched_task(p); + } +out_unlock: + task_grq_unlock(&flags); +} +EXPORT_SYMBOL(set_user_nice); + +/* + * can_nice - check if a task can reduce its nice value + * @p: task + * @nice: nice value + */ +int can_nice(const struct task_struct *p, const int nice) +{ + /* convert nice value [19,-20] to rlimit style value [1,40] */ + int nice_rlim = nice_to_rlimit(nice); + + return (nice_rlim <= task_rlimit(p, RLIMIT_NICE) || + capable(CAP_SYS_NICE)); +} + +#ifdef __ARCH_WANT_SYS_NICE + +/* + * sys_nice - change the priority of the current process. + * @increment: priority increment + * + * sys_setpriority is a more generic, but much slower function that + * does similar things. + */ +SYSCALL_DEFINE1(nice, int, increment) +{ + long nice, retval; + + /* + * Setpriority might change our priority at the same moment. + * We don't have to worry. Conceptually one call occurs first + * and we have a single winner. + */ + + increment = clamp(increment, -NICE_WIDTH, NICE_WIDTH); + nice = task_nice(current) + increment; + + nice = clamp_val(nice, MIN_NICE, MAX_NICE); + if (increment < 0 && !can_nice(current, nice)) + return -EPERM; + + retval = security_task_setnice(current, nice); + if (retval) + return retval; + + set_user_nice(current, nice); + return 0; +} + +#endif + +/** + * task_prio - return the priority value of a given task. + * @p: the task in question. + * + * Return: The priority value as seen by users in /proc. + * RT tasks are offset by -100. Normal tasks are centered around 1, value goes + * from 0 (SCHED_ISO) up to 82 (nice +19 SCHED_IDLEPRIO). + */ +int task_prio(const struct task_struct *p) +{ + int delta, prio = p->prio - MAX_RT_PRIO; + + /* rt tasks and iso tasks */ + if (prio <= 0) + goto out; + + /* Convert to ms to avoid overflows */ + delta = NS_TO_MS(p->deadline - grq.niffies); + delta = delta * 40 / ms_longest_deadline_diff(); + if (delta > 0 && delta <= 80) + prio += delta; + if (idleprio_task(p)) + prio += 40; +out: + return prio; +} + +/** + * idle_cpu - is a given cpu idle currently? + * @cpu: the processor in question. + * + * Return: 1 if the CPU is currently idle. 0 otherwise. + */ +int idle_cpu(int cpu) +{ + return cpu_curr(cpu) == cpu_rq(cpu)->idle; +} + +/** + * idle_task - return the idle task for a given cpu. + * @cpu: the processor in question. + * + * Return: The idle task for the cpu @cpu. + */ +struct task_struct *idle_task(int cpu) +{ + return cpu_rq(cpu)->idle; +} + +/** + * find_process_by_pid - find a process with a matching PID value. + * @pid: the pid in question. + * + * The task of @pid, if found. %NULL otherwise. + */ +static inline struct task_struct *find_process_by_pid(pid_t pid) +{ + return pid ? find_task_by_vpid(pid) : current; +} + +/* Actually do priority change: must hold grq lock. */ +static void __setscheduler(struct task_struct *p, struct rq *rq, int policy, + int prio, bool keep_boost) +{ + int oldrtprio, oldprio; + + p->policy = policy; + oldrtprio = p->rt_priority; + p->rt_priority = prio; + p->normal_prio = normal_prio(p); + oldprio = p->prio; + /* + * Keep a potential priority boosting if called from + * sched_setscheduler(). + */ + if (keep_boost) { + /* + * Take priority boosted tasks into account. If the new + * effective priority is unchanged, we just store the new + * normal parameters and do not touch the scheduler class and + * the runqueue. This will be done when the task deboost + * itself. + */ + p->prio = rt_mutex_get_effective_prio(p, p->normal_prio); + } else + p->prio = p->normal_prio; + if (task_running(p)) { + reset_rq_task(rq, p); + /* Resched only if we might now be preempted */ + if (p->prio > oldprio || p->rt_priority > oldrtprio) + resched_task(p); + } +} + +/* + * check the target process has a UID that matches the current process's + */ +static bool check_same_owner(struct task_struct *p) +{ + const struct cred *cred = current_cred(), *pcred; + bool match; + + rcu_read_lock(); + pcred = __task_cred(p); + match = (uid_eq(cred->euid, pcred->euid) || + uid_eq(cred->euid, pcred->uid)); + rcu_read_unlock(); + return match; +} + +static int +__sched_setscheduler(struct task_struct *p, int policy, + const struct sched_param *param, bool user, bool pi) +{ + struct sched_param zero_param = { .sched_priority = 0 }; + int queued, retval, oldpolicy = -1; + unsigned long flags, rlim_rtprio = 0; + int reset_on_fork; + struct rq *rq; + + /* may grab non-irq protected spin_locks */ + BUG_ON(in_interrupt()); + + if (is_rt_policy(policy) && !capable(CAP_SYS_NICE)) { + unsigned long lflags; + + if (!lock_task_sighand(p, &lflags)) + return -ESRCH; + rlim_rtprio = task_rlimit(p, RLIMIT_RTPRIO); + unlock_task_sighand(p, &lflags); + if (rlim_rtprio) + goto recheck; + /* + * If the caller requested an RT policy without having the + * necessary rights, we downgrade the policy to SCHED_ISO. + * We also set the parameter to zero to pass the checks. + */ + policy = SCHED_ISO; + param = &zero_param; + } +recheck: + /* double check policy once rq lock held */ + if (policy < 0) { + reset_on_fork = p->sched_reset_on_fork; + policy = oldpolicy = p->policy; + } else { + reset_on_fork = !!(policy & SCHED_RESET_ON_FORK); + policy &= ~SCHED_RESET_ON_FORK; + + if (!SCHED_RANGE(policy)) + return -EINVAL; + } + + /* + * Valid priorities for SCHED_FIFO and SCHED_RR are + * 1..MAX_USER_RT_PRIO-1, valid priority for SCHED_NORMAL and + * SCHED_BATCH is 0. + */ + if (param->sched_priority < 0 || + (p->mm && param->sched_priority > MAX_USER_RT_PRIO - 1) || + (!p->mm && param->sched_priority > MAX_RT_PRIO - 1)) + return -EINVAL; + if (is_rt_policy(policy) != (param->sched_priority != 0)) + return -EINVAL; + + /* + * Allow unprivileged RT tasks to decrease priority: + */ + if (user && !capable(CAP_SYS_NICE)) { + if (is_rt_policy(policy)) { + unsigned long rlim_rtprio = + task_rlimit(p, RLIMIT_RTPRIO); + + /* can't set/change the rt policy */ + if (policy != p->policy && !rlim_rtprio) + return -EPERM; + + /* can't increase priority */ + if (param->sched_priority > p->rt_priority && + param->sched_priority > rlim_rtprio) + return -EPERM; + } else { + switch (p->policy) { + /* + * Can only downgrade policies but not back to + * SCHED_NORMAL + */ + case SCHED_ISO: + if (policy == SCHED_ISO) + goto out; + if (policy == SCHED_NORMAL) + return -EPERM; + break; + case SCHED_BATCH: + if (policy == SCHED_BATCH) + goto out; + if (policy != SCHED_IDLEPRIO) + return -EPERM; + break; + case SCHED_IDLEPRIO: + if (policy == SCHED_IDLEPRIO) + goto out; + return -EPERM; + default: + break; + } + } + + /* can't change other user's priorities */ + if (!check_same_owner(p)) + return -EPERM; + + /* Normal users shall not reset the sched_reset_on_fork flag */ + if (p->sched_reset_on_fork && !reset_on_fork) + return -EPERM; + } + + if (user) { + retval = security_task_setscheduler(p); + if (retval) + return retval; + } + + /* + * make sure no PI-waiters arrive (or leave) while we are + * changing the priority of the task: + */ + raw_spin_lock_irqsave(&p->pi_lock, flags); + /* + * To be able to change p->policy safely, the grunqueue lock must be + * held. + */ + rq = __task_grq_lock(p); + + /* + * Changing the policy of the stop threads its a very bad idea + */ + if (p == rq->stop) { + __task_grq_unlock(); + raw_spin_unlock_irqrestore(&p->pi_lock, flags); + return -EINVAL; + } + + /* + * If not changing anything there's no need to proceed further: + */ + if (unlikely(policy == p->policy && (!is_rt_policy(policy) || + param->sched_priority == p->rt_priority))) { + + __task_grq_unlock(); + raw_spin_unlock_irqrestore(&p->pi_lock, flags); + return 0; + } + + /* recheck policy now with rq lock held */ + if (unlikely(oldpolicy != -1 && oldpolicy != p->policy)) { + policy = oldpolicy = -1; + __task_grq_unlock(); + raw_spin_unlock_irqrestore(&p->pi_lock, flags); + goto recheck; + } + update_clocks(rq); + p->sched_reset_on_fork = reset_on_fork; + + queued = task_queued(p); + if (queued) + dequeue_task(p); + __setscheduler(p, rq, policy, param->sched_priority, pi); + if (queued) { + enqueue_task(p, rq); + try_preempt(p, rq); + } + __task_grq_unlock(); + raw_spin_unlock_irqrestore(&p->pi_lock, flags); + + if (pi) + rt_mutex_adjust_pi(p); +out: + return 0; +} + +/** + * sched_setscheduler - change the scheduling policy and/or RT priority of a thread. + * @p: the task in question. + * @policy: new policy. + * @param: structure containing the new RT priority. + * + * Return: 0 on success. An error code otherwise. + * + * NOTE that the task may be already dead. + */ +int sched_setscheduler(struct task_struct *p, int policy, + const struct sched_param *param) +{ + return __sched_setscheduler(p, policy, param, true, true); +} + +EXPORT_SYMBOL_GPL(sched_setscheduler); + +int sched_setattr(struct task_struct *p, const struct sched_attr *attr) +{ + const struct sched_param param = { .sched_priority = attr->sched_priority }; + int policy = attr->sched_policy; + + return __sched_setscheduler(p, policy, ¶m, true, true); +} +EXPORT_SYMBOL_GPL(sched_setattr); + +/** + * sched_setscheduler_nocheck - change the scheduling policy and/or RT priority of a thread from kernelspace. + * @p: the task in question. + * @policy: new policy. + * @param: structure containing the new RT priority. + * + * Just like sched_setscheduler, only don't bother checking if the + * current context has permission. For example, this is needed in + * stop_machine(): we create temporary high priority worker threads, + * but our caller might not have that capability. + * + * Return: 0 on success. An error code otherwise. + */ +int sched_setscheduler_nocheck(struct task_struct *p, int policy, + const struct sched_param *param) +{ + return __sched_setscheduler(p, policy, param, false, true); +} +EXPORT_SYMBOL_GPL(sched_setscheduler_nocheck); + +static int +do_sched_setscheduler(pid_t pid, int policy, struct sched_param __user *param) +{ + struct sched_param lparam; + struct task_struct *p; + int retval; + + if (!param || pid < 0) + return -EINVAL; + if (copy_from_user(&lparam, param, sizeof(struct sched_param))) + return -EFAULT; + + rcu_read_lock(); + retval = -ESRCH; + p = find_process_by_pid(pid); + if (p != NULL) + retval = sched_setscheduler(p, policy, &lparam); + rcu_read_unlock(); + + return retval; +} + +/* + * Mimics kernel/events/core.c perf_copy_attr(). + */ +static int sched_copy_attr(struct sched_attr __user *uattr, + struct sched_attr *attr) +{ + u32 size; + int ret; + + if (!access_ok(VERIFY_WRITE, uattr, SCHED_ATTR_SIZE_VER0)) + return -EFAULT; + + /* + * zero the full structure, so that a short copy will be nice. + */ + memset(attr, 0, sizeof(*attr)); + + ret = get_user(size, &uattr->size); + if (ret) + return ret; + + if (size > PAGE_SIZE) /* silly large */ + goto err_size; + + if (!size) /* abi compat */ + size = SCHED_ATTR_SIZE_VER0; + + if (size < SCHED_ATTR_SIZE_VER0) + goto err_size; + + /* + * If we're handed a bigger struct than we know of, + * ensure all the unknown bits are 0 - i.e. new + * user-space does not rely on any kernel feature + * extensions we dont know about yet. + */ + if (size > sizeof(*attr)) { + unsigned char __user *addr; + unsigned char __user *end; + unsigned char val; + + addr = (void __user *)uattr + sizeof(*attr); + end = (void __user *)uattr + size; + + for (; addr < end; addr++) { + ret = get_user(val, addr); + if (ret) + return ret; + if (val) + goto err_size; + } + size = sizeof(*attr); + } + + ret = copy_from_user(attr, uattr, size); + if (ret) + return -EFAULT; + + /* + * XXX: do we want to be lenient like existing syscalls; or do we want + * to be strict and return an error on out-of-bounds values? + */ + attr->sched_nice = clamp(attr->sched_nice, -20, 19); + + /* sched/core.c uses zero here but we already know ret is zero */ + return 0; + +err_size: + put_user(sizeof(*attr), &uattr->size); + return -E2BIG; +} + +/** + * sys_sched_setscheduler - set/change the scheduler policy and RT priority + * @pid: the pid in question. + * @policy: new policy. + * + * Return: 0 on success. An error code otherwise. + * @param: structure containing the new RT priority. + */ +asmlinkage long sys_sched_setscheduler(pid_t pid, int policy, + struct sched_param __user *param) +{ + /* negative values for policy are not valid */ + if (policy < 0) + return -EINVAL; + + return do_sched_setscheduler(pid, policy, param); +} + +/* + * sched_setparam() passes in -1 for its policy, to let the functions + * it calls know not to change it. + */ +#define SETPARAM_POLICY -1 + +/** + * sys_sched_setparam - set/change the RT priority of a thread + * @pid: the pid in question. + * @param: structure containing the new RT priority. + * + * Return: 0 on success. An error code otherwise. + */ +SYSCALL_DEFINE2(sched_setparam, pid_t, pid, struct sched_param __user *, param) +{ + return do_sched_setscheduler(pid, SETPARAM_POLICY, param); +} + +/** + * sys_sched_setattr - same as above, but with extended sched_attr + * @pid: the pid in question. + * @uattr: structure containing the extended parameters. + */ +SYSCALL_DEFINE3(sched_setattr, pid_t, pid, struct sched_attr __user *, uattr, + unsigned int, flags) +{ + struct sched_attr attr; + struct task_struct *p; + int retval; + + if (!uattr || pid < 0 || flags) + return -EINVAL; + + retval = sched_copy_attr(uattr, &attr); + if (retval) + return retval; + + if ((int)attr.sched_policy < 0) + return -EINVAL; + + rcu_read_lock(); + retval = -ESRCH; + p = find_process_by_pid(pid); + if (p != NULL) + retval = sched_setattr(p, &attr); + rcu_read_unlock(); + + return retval; +} + +/** + * sys_sched_getscheduler - get the policy (scheduling class) of a thread + * @pid: the pid in question. + * + * Return: On success, the policy of the thread. Otherwise, a negative error + * code. + */ +SYSCALL_DEFINE1(sched_getscheduler, pid_t, pid) +{ + struct task_struct *p; + int retval = -EINVAL; + + if (pid < 0) + goto out_nounlock; + + retval = -ESRCH; + rcu_read_lock(); + p = find_process_by_pid(pid); + if (p) { + retval = security_task_getscheduler(p); + if (!retval) + retval = p->policy; + } + rcu_read_unlock(); + +out_nounlock: + return retval; +} + +/** + * sys_sched_getscheduler - get the RT priority of a thread + * @pid: the pid in question. + * @param: structure containing the RT priority. + * + * Return: On success, 0 and the RT priority is in @param. Otherwise, an error + * code. + */ +SYSCALL_DEFINE2(sched_getparam, pid_t, pid, struct sched_param __user *, param) +{ + struct sched_param lp = { .sched_priority = 0 }; + struct task_struct *p; + int retval = -EINVAL; + + if (!param || pid < 0) + goto out_nounlock; + + rcu_read_lock(); + p = find_process_by_pid(pid); + retval = -ESRCH; + if (!p) + goto out_unlock; + + retval = security_task_getscheduler(p); + if (retval) + goto out_unlock; + + if (has_rt_policy(p)) + lp.sched_priority = p->rt_priority; + rcu_read_unlock(); + + /* + * This one might sleep, we cannot do it with a spinlock held ... + */ + retval = copy_to_user(param, &lp, sizeof(*param)) ? -EFAULT : 0; + +out_nounlock: + return retval; + +out_unlock: + rcu_read_unlock(); + return retval; +} + +static int sched_read_attr(struct sched_attr __user *uattr, + struct sched_attr *attr, + unsigned int usize) +{ + int ret; + + if (!access_ok(VERIFY_WRITE, uattr, usize)) + return -EFAULT; + + /* + * If we're handed a smaller struct than we know of, + * ensure all the unknown bits are 0 - i.e. old + * user-space does not get uncomplete information. + */ + if (usize < sizeof(*attr)) { + unsigned char *addr; + unsigned char *end; + + addr = (void *)attr + usize; + end = (void *)attr + sizeof(*attr); + + for (; addr < end; addr++) { + if (*addr) + return -EFBIG; + } + + attr->size = usize; + } + + ret = copy_to_user(uattr, attr, attr->size); + if (ret) + return -EFAULT; + + /* sched/core.c uses zero here but we already know ret is zero */ + return ret; +} + +/** + * sys_sched_getattr - similar to sched_getparam, but with sched_attr + * @pid: the pid in question. + * @uattr: structure containing the extended parameters. + * @size: sizeof(attr) for fwd/bwd comp. + * @flags: for future extension. + */ +SYSCALL_DEFINE4(sched_getattr, pid_t, pid, struct sched_attr __user *, uattr, + unsigned int, size, unsigned int, flags) +{ + struct sched_attr attr = { + .size = sizeof(struct sched_attr), + }; + struct task_struct *p; + int retval; + + if (!uattr || pid < 0 || size > PAGE_SIZE || + size < SCHED_ATTR_SIZE_VER0 || flags) + return -EINVAL; + + rcu_read_lock(); + p = find_process_by_pid(pid); + retval = -ESRCH; + if (!p) + goto out_unlock; + + retval = security_task_getscheduler(p); + if (retval) + goto out_unlock; + + attr.sched_policy = p->policy; + if (rt_task(p)) + attr.sched_priority = p->rt_priority; + else + attr.sched_nice = task_nice(p); + + rcu_read_unlock(); + + retval = sched_read_attr(uattr, &attr, size); + return retval; + +out_unlock: + rcu_read_unlock(); + return retval; +} + +long sched_setaffinity(pid_t pid, const struct cpumask *in_mask) +{ + cpumask_var_t cpus_allowed, new_mask; + struct task_struct *p; + int retval; + + get_online_cpus(); + rcu_read_lock(); + + p = find_process_by_pid(pid); + if (!p) { + rcu_read_unlock(); + put_online_cpus(); + return -ESRCH; + } + + /* Prevent p going away */ + get_task_struct(p); + rcu_read_unlock(); + + if (p->flags & PF_NO_SETAFFINITY) { + retval = -EINVAL; + goto out_put_task; + } + if (!alloc_cpumask_var(&cpus_allowed, GFP_KERNEL)) { + retval = -ENOMEM; + goto out_put_task; + } + if (!alloc_cpumask_var(&new_mask, GFP_KERNEL)) { + retval = -ENOMEM; + goto out_free_cpus_allowed; + } + retval = -EPERM; + if (!check_same_owner(p)) { + rcu_read_lock(); + if (!ns_capable(__task_cred(p)->user_ns, CAP_SYS_NICE)) { + rcu_read_unlock(); + goto out_unlock; + } + rcu_read_unlock(); + } + + retval = security_task_setscheduler(p); + if (retval) + goto out_unlock; + + cpuset_cpus_allowed(p, cpus_allowed); + cpumask_and(new_mask, in_mask, cpus_allowed); +again: + retval = __set_cpus_allowed_ptr(p, new_mask, true); + + if (!retval) { + cpuset_cpus_allowed(p, cpus_allowed); + if (!cpumask_subset(new_mask, cpus_allowed)) { + /* + * We must have raced with a concurrent cpuset + * update. Just reset the cpus_allowed to the + * cpuset's cpus_allowed + */ + cpumask_copy(new_mask, cpus_allowed); + goto again; + } + } +out_unlock: + free_cpumask_var(new_mask); +out_free_cpus_allowed: + free_cpumask_var(cpus_allowed); +out_put_task: + put_task_struct(p); + put_online_cpus(); + return retval; +} + +static int get_user_cpu_mask(unsigned long __user *user_mask_ptr, unsigned len, + cpumask_t *new_mask) +{ + if (len < sizeof(cpumask_t)) { + memset(new_mask, 0, sizeof(cpumask_t)); + } else if (len > sizeof(cpumask_t)) { + len = sizeof(cpumask_t); + } + return copy_from_user(new_mask, user_mask_ptr, len) ? -EFAULT : 0; +} + + +/** + * sys_sched_setaffinity - set the cpu affinity of a process + * @pid: pid of the process + * @len: length in bytes of the bitmask pointed to by user_mask_ptr + * @user_mask_ptr: user-space pointer to the new cpu mask + * + * Return: 0 on success. An error code otherwise. + */ +SYSCALL_DEFINE3(sched_setaffinity, pid_t, pid, unsigned int, len, + unsigned long __user *, user_mask_ptr) +{ + cpumask_var_t new_mask; + int retval; + + if (!alloc_cpumask_var(&new_mask, GFP_KERNEL)) + return -ENOMEM; + + retval = get_user_cpu_mask(user_mask_ptr, len, new_mask); + if (retval == 0) + retval = sched_setaffinity(pid, new_mask); + free_cpumask_var(new_mask); + return retval; +} + +long sched_getaffinity(pid_t pid, cpumask_t *mask) +{ + struct task_struct *p; + unsigned long flags; + int retval; + + get_online_cpus(); + rcu_read_lock(); + + retval = -ESRCH; + p = find_process_by_pid(pid); + if (!p) + goto out_unlock; + + retval = security_task_getscheduler(p); + if (retval) + goto out_unlock; + + grq_lock_irqsave(&flags); + cpumask_and(mask, tsk_cpus_allowed(p), cpu_active_mask); + grq_unlock_irqrestore(&flags); + +out_unlock: + rcu_read_unlock(); + put_online_cpus(); + + return retval; +} + +/** + * sys_sched_getaffinity - get the cpu affinity of a process + * @pid: pid of the process + * @len: length in bytes of the bitmask pointed to by user_mask_ptr + * @user_mask_ptr: user-space pointer to hold the current cpu mask + * + * Return: 0 on success. An error code otherwise. + */ +SYSCALL_DEFINE3(sched_getaffinity, pid_t, pid, unsigned int, len, + unsigned long __user *, user_mask_ptr) +{ + int ret; + cpumask_var_t mask; + + if ((len * BITS_PER_BYTE) < nr_cpu_ids) + return -EINVAL; + if (len & (sizeof(unsigned long)-1)) + return -EINVAL; + + if (!alloc_cpumask_var(&mask, GFP_KERNEL)) + return -ENOMEM; + + ret = sched_getaffinity(pid, mask); + if (ret == 0) { + size_t retlen = min_t(size_t, len, cpumask_size()); + + if (copy_to_user(user_mask_ptr, mask, retlen)) + ret = -EFAULT; + else + ret = retlen; + } + free_cpumask_var(mask); + + return ret; +} + +/** + * sys_sched_yield - yield the current processor to other threads. + * + * This function yields the current CPU to other tasks. It does this by + * scheduling away the current task. If it still has the earliest deadline + * it will be scheduled again as the next task. + * + * Return: 0. + */ +SYSCALL_DEFINE0(sched_yield) +{ + struct task_struct *p; + + p = current; + grq_lock_irq(); + schedstat_inc(task_rq(p), yld_count); + requeue_task(p); + + /* + * Since we are going to call schedule() anyway, there's + * no need to preempt or enable interrupts: + */ + __release(grq.lock); + spin_release(&grq.lock.dep_map, 1, _THIS_IP_); + do_raw_spin_unlock(&grq.lock); + sched_preempt_enable_no_resched(); + + schedule(); + + return 0; +} + +int __sched _cond_resched(void) +{ + if (should_resched(0)) { + preempt_schedule_common(); + return 1; + } + return 0; +} +EXPORT_SYMBOL(_cond_resched); + +/* + * __cond_resched_lock() - if a reschedule is pending, drop the given lock, + * call schedule, and on return reacquire the lock. + * + * This works OK both with and without CONFIG_PREEMPT. We do strange low-level + * operations here to prevent schedule() from being called twice (once via + * spin_unlock(), once by hand). + */ +int __cond_resched_lock(spinlock_t *lock) +{ + int resched = should_resched(PREEMPT_LOCK_OFFSET); + int ret = 0; + + lockdep_assert_held(lock); + + if (spin_needbreak(lock) || resched) { + spin_unlock(lock); + if (resched) + preempt_schedule_common(); + else + cpu_relax(); + ret = 1; + spin_lock(lock); + } + return ret; +} +EXPORT_SYMBOL(__cond_resched_lock); + +int __sched __cond_resched_softirq(void) +{ + BUG_ON(!in_softirq()); + + if (should_resched(SOFTIRQ_DISABLE_OFFSET)) { + local_bh_enable(); + preempt_schedule_common(); + local_bh_disable(); + return 1; + } + return 0; +} +EXPORT_SYMBOL(__cond_resched_softirq); + +/** + * yield - yield the current processor to other threads. + * + * Do not ever use this function, there's a 99% chance you're doing it wrong. + * + * The scheduler is at all times free to pick the calling task as the most + * eligible task to run, if removing the yield() call from your code breaks + * it, its already broken. + * + * Typical broken usage is: + * + * while (!event) + * yield(); + * + * where one assumes that yield() will let 'the other' process run that will + * make event true. If the current task is a SCHED_FIFO task that will never + * happen. Never use yield() as a progress guarantee!! + * + * If you want to use yield() to wait for something, use wait_event(). + * If you want to use yield() to be 'nice' for others, use cond_resched(). + * If you still want to use yield(), do not! + */ +void __sched yield(void) +{ + set_current_state(TASK_RUNNING); + sys_sched_yield(); +} +EXPORT_SYMBOL(yield); + +/** + * yield_to - yield the current processor to another thread in + * your thread group, or accelerate that thread toward the + * processor it's on. + * @p: target task + * @preempt: whether task preemption is allowed or not + * + * It's the caller's job to ensure that the target task struct + * can't go away on us before we can do any checks. + * + * Return: + * true (>0) if we indeed boosted the target task. + * false (0) if we failed to boost the target. + * -ESRCH if there's no task to yield to. + */ +int __sched yield_to(struct task_struct *p, bool preempt) +{ + struct rq *rq, *p_rq; + unsigned long flags; + int yielded = 0; + + rq = this_rq(); + grq_lock_irqsave(&flags); + if (task_running(p) || p->state) { + yielded = -ESRCH; + goto out_unlock; + } + + p_rq = task_rq(p); + yielded = 1; + if (p->deadline > rq->rq_deadline) + p->deadline = rq->rq_deadline; + p->time_slice += rq->rq_time_slice; + rq->rq_time_slice = 0; + if (p->time_slice > timeslice()) + p->time_slice = timeslice(); + if (preempt && rq != p_rq) + resched_curr(p_rq); +out_unlock: + grq_unlock_irqrestore(&flags); + + if (yielded > 0) + schedule(); + return yielded; +} +EXPORT_SYMBOL_GPL(yield_to); + +/* + * This task is about to go to sleep on IO. Increment rq->nr_iowait so + * that process accounting knows that this is a task in IO wait state. + * + * But don't do that if it is a deliberate, throttling IO wait (this task + * has set its backing_dev_info: the queue against which it should throttle) + */ + +long __sched io_schedule_timeout(long timeout) +{ + int old_iowait = current->in_iowait; + struct rq *rq; + long ret; + + current->in_iowait = 1; + blk_schedule_flush_plug(current); + + delayacct_blkio_start(); + rq = raw_rq(); + atomic_inc(&rq->nr_iowait); + ret = schedule_timeout(timeout); + current->in_iowait = old_iowait; + atomic_dec(&rq->nr_iowait); + delayacct_blkio_end(); + + return ret; +} +EXPORT_SYMBOL(io_schedule_timeout); + +/** + * sys_sched_get_priority_max - return maximum RT priority. + * @policy: scheduling class. + * + * Return: On success, this syscall returns the maximum + * rt_priority that can be used by a given scheduling class. + * On failure, a negative error code is returned. + */ +SYSCALL_DEFINE1(sched_get_priority_max, int, policy) +{ + int ret = -EINVAL; + + switch (policy) { + case SCHED_FIFO: + case SCHED_RR: + ret = MAX_USER_RT_PRIO-1; + break; + case SCHED_NORMAL: + case SCHED_BATCH: + case SCHED_ISO: + case SCHED_IDLEPRIO: + ret = 0; + break; + } + return ret; +} + +/** + * sys_sched_get_priority_min - return minimum RT priority. + * @policy: scheduling class. + * + * Return: On success, this syscall returns the minimum + * rt_priority that can be used by a given scheduling class. + * On failure, a negative error code is returned. + */ +SYSCALL_DEFINE1(sched_get_priority_min, int, policy) +{ + int ret = -EINVAL; + + switch (policy) { + case SCHED_FIFO: + case SCHED_RR: + ret = 1; + break; + case SCHED_NORMAL: + case SCHED_BATCH: + case SCHED_ISO: + case SCHED_IDLEPRIO: + ret = 0; + break; + } + return ret; +} + +/** + * sys_sched_rr_get_interval - return the default timeslice of a process. + * @pid: pid of the process. + * @interval: userspace pointer to the timeslice value. + * + * + * Return: On success, 0 and the timeslice is in @interval. Otherwise, + * an error code. + */ +SYSCALL_DEFINE2(sched_rr_get_interval, pid_t, pid, + struct timespec __user *, interval) +{ + struct task_struct *p; + unsigned int time_slice; + unsigned long flags; + int retval; + struct timespec t; + + if (pid < 0) + return -EINVAL; + + retval = -ESRCH; + rcu_read_lock(); + p = find_process_by_pid(pid); + if (!p) + goto out_unlock; + + retval = security_task_getscheduler(p); + if (retval) + goto out_unlock; + + grq_lock_irqsave(&flags); + time_slice = p->policy == SCHED_FIFO ? 0 : MS_TO_NS(task_timeslice(p)); + grq_unlock_irqrestore(&flags); + + rcu_read_unlock(); + t = ns_to_timespec(time_slice); + retval = copy_to_user(interval, &t, sizeof(t)) ? -EFAULT : 0; + return retval; + +out_unlock: + rcu_read_unlock(); + return retval; +} + +static const char stat_nam[] = TASK_STATE_TO_CHAR_STR; + +void sched_show_task(struct task_struct *p) +{ + unsigned long free = 0; + int ppid; + unsigned long state = p->state; + + if (state) + state = __ffs(state) + 1; + printk(KERN_INFO "%-15.15s %c", p->comm, + state < sizeof(stat_nam) - 1 ? stat_nam[state] : '?'); +#if BITS_PER_LONG == 32 + if (state == TASK_RUNNING) + printk(KERN_CONT " running "); + else + printk(KERN_CONT " %08lx ", thread_saved_pc(p)); +#else + if (state == TASK_RUNNING) + printk(KERN_CONT " running task "); + else + printk(KERN_CONT " %016lx ", thread_saved_pc(p)); +#endif +#ifdef CONFIG_DEBUG_STACK_USAGE + free = stack_not_used(p); +#endif + ppid = 0; + rcu_read_lock(); + if (pid_alive(p)) + ppid = task_pid_nr(rcu_dereference(p->real_parent)); + rcu_read_unlock(); + printk(KERN_CONT "%5lu %5d %6d 0x%08lx\n", free, + task_pid_nr(p), ppid, + (unsigned long)task_thread_info(p)->flags); + + print_worker_info(KERN_INFO, p); + show_stack(p, NULL); +} + +void show_state_filter(unsigned long state_filter) +{ + struct task_struct *g, *p; + +#if BITS_PER_LONG == 32 + printk(KERN_INFO + " task PC stack pid father\n"); +#else + printk(KERN_INFO + " task PC stack pid father\n"); +#endif + rcu_read_lock(); + for_each_process_thread(g, p) { + /* + * reset the NMI-timeout, listing all files on a slow + * console might take a lot of time: + */ + touch_nmi_watchdog(); + if (!state_filter || (p->state & state_filter)) + sched_show_task(p); + } + + touch_all_softlockup_watchdogs(); + + rcu_read_unlock(); + /* + * Only show locks if all tasks are dumped: + */ + if (!state_filter) + debug_show_all_locks(); +} + +void dump_cpu_task(int cpu) +{ + pr_info("Task dump for CPU %d:\n", cpu); + sched_show_task(cpu_curr(cpu)); +} + +#ifdef CONFIG_SMP +void set_cpus_allowed_common(struct task_struct *p, const struct cpumask *new_mask) +{ + cpumask_copy(&p->cpus_allowed, new_mask); + p->nr_cpus_allowed = cpumask_weight(new_mask); +} + +void do_set_cpus_allowed(struct task_struct *p, const struct cpumask *new_mask) +{ + cpumask_copy(tsk_cpus_allowed(p), new_mask); +} +#endif + +/** + * init_idle - set up an idle thread for a given CPU + * @idle: task in question + * @cpu: cpu the idle task belongs to + * + * NOTE: this function does not set the idle thread's NEED_RESCHED + * flag, to make booting more robust. + */ +void init_idle(struct task_struct *idle, int cpu) +{ + struct rq *rq = cpu_rq(cpu); + unsigned long flags; + + raw_spin_lock_irqsave(&idle->pi_lock, flags); + time_lock_grq(rq); + idle->last_ran = rq->clock_task; + idle->state = TASK_RUNNING; + /* Setting prio to illegal value shouldn't matter when never queued */ + idle->prio = PRIO_LIMIT; + + kasan_unpoison_task_stack(idle); + +#ifdef CONFIG_SMP + /* + * It's possible that init_idle() gets called multiple times on a task, + * in that case do_set_cpus_allowed() will not do the right thing. + * + * And since this is boot we can forgo the serialisation. + */ + set_cpus_allowed_common(idle, cpumask_of(cpu)); +#ifdef CONFIG_SMT_NICE + idle->smt_bias = 0; +#endif +#endif + set_rq_task(rq, idle); + + /* Silence PROVE_RCU */ + rcu_read_lock(); + set_task_cpu(idle, cpu); + rcu_read_unlock(); + + rq->curr = rq->idle = idle; + idle->on_cpu = 1; + grq_unlock(); + raw_spin_unlock_irqrestore(&idle->pi_lock, flags); + + /* Set the preempt count _outside_ the spinlocks! */ + init_idle_preempt_count(idle, cpu); + + ftrace_graph_init_idle_task(idle, cpu); +#ifdef CONFIG_SMP + sprintf(idle->comm, "%s/%d", INIT_TASK_COMM, cpu); +#endif +} + +int cpuset_cpumask_can_shrink(const struct cpumask __maybe_unused *cur, + const struct cpumask __maybe_unused *trial) +{ + return 1; +} + +int task_can_attach(struct task_struct *p, + const struct cpumask *cs_cpus_allowed) +{ + int ret = 0; + + /* + * Kthreads which disallow setaffinity shouldn't be moved + * to a new cpuset; we don't want to change their cpu + * affinity and isolating such threads by their set of + * allowed nodes is unnecessary. Thus, cpusets are not + * applicable for such threads. This prevents checking for + * success of set_cpus_allowed_ptr() on all attached tasks + * before cpus_allowed may be changed. + */ + if (p->flags & PF_NO_SETAFFINITY) + ret = -EINVAL; + + return ret; +} + +void wake_q_add(struct wake_q_head *head, struct task_struct *task) +{ + struct wake_q_node *node = &task->wake_q; + + /* + * Atomically grab the task, if ->wake_q is !nil already it means + * its already queued (either by us or someone else) and will get the + * wakeup due to that. + * + * This cmpxchg() implies a full barrier, which pairs with the write + * barrier implied by the wakeup in wake_up_list(). + */ + if (cmpxchg(&node->next, NULL, WAKE_Q_TAIL)) + return; + + get_task_struct(task); + + /* + * The head is context local, there can be no concurrency. + */ + *head->lastp = node; + head->lastp = &node->next; +} + +void wake_up_q(struct wake_q_head *head) +{ + struct wake_q_node *node = head->first; + + while (node != WAKE_Q_TAIL) { + struct task_struct *task; + + task = container_of(node, struct task_struct, wake_q); + BUG_ON(!task); + /* task can safely be re-inserted now */ + node = node->next; + task->wake_q.next = NULL; + + /* + * wake_up_process() implies a wmb() to pair with the queueing + * in wake_q_add() so as not to miss wakeups. + */ + wake_up_process(task); + put_task_struct(task); + } +} + +void resched_cpu(int cpu) +{ + unsigned long flags; + + grq_lock_irqsave(&flags); + resched_task(cpu_curr(cpu)); + grq_unlock_irqrestore(&flags); +} + +#ifdef CONFIG_SMP +#ifdef CONFIG_NO_HZ_COMMON +void nohz_balance_enter_idle(int cpu) +{ +} + +void select_nohz_load_balancer(int stop_tick) +{ +} + +void set_cpu_sd_state_idle(void) {} +#if defined(CONFIG_SCHED_MC) || defined(CONFIG_SCHED_SMT) +/** + * lowest_flag_domain - Return lowest sched_domain containing flag. + * @cpu: The cpu whose lowest level of sched domain is to + * be returned. + * @flag: The flag to check for the lowest sched_domain + * for the given cpu. + * + * Returns the lowest sched_domain of a cpu which contains the given flag. + */ +static inline struct sched_domain *lowest_flag_domain(int cpu, int flag) +{ + struct sched_domain *sd; + + for_each_domain(cpu, sd) + if (sd && (sd->flags & flag)) + break; + + return sd; +} + +/** + * for_each_flag_domain - Iterates over sched_domains containing the flag. + * @cpu: The cpu whose domains we're iterating over. + * @sd: variable holding the value of the power_savings_sd + * for cpu. + * @flag: The flag to filter the sched_domains to be iterated. + * + * Iterates over all the scheduler domains for a given cpu that has the 'flag' + * set, starting from the lowest sched_domain to the highest. + */ +#define for_each_flag_domain(cpu, sd, flag) \ + for (sd = lowest_flag_domain(cpu, flag); \ + (sd && (sd->flags & flag)); sd = sd->parent) + +#endif /* (CONFIG_SCHED_MC || CONFIG_SCHED_SMT) */ + +/* + * In the semi idle case, use the nearest busy cpu for migrating timers + * from an idle cpu. This is good for power-savings. + * + * We don't do similar optimization for completely idle system, as + * selecting an idle cpu will add more delays to the timers than intended + * (as that cpu's timer base may not be uptodate wrt jiffies etc). + */ +int get_nohz_timer_target(void) +{ + int i, cpu = smp_processor_id(); + struct sched_domain *sd; + + if (!idle_cpu(cpu) && is_housekeeping_cpu(cpu)) + return cpu; + + rcu_read_lock(); + for_each_domain(cpu, sd) { + for_each_cpu(i, sched_domain_span(sd)) { + if (!idle_cpu(i) && is_housekeeping_cpu(cpu)) { + cpu = i; + goto unlock; + } + } + } + + if (!is_housekeeping_cpu(cpu)) + cpu = housekeeping_any_cpu(); +unlock: + rcu_read_unlock(); + return cpu; +} + +/* + * When add_timer_on() enqueues a timer into the timer wheel of an + * idle CPU then this timer might expire before the next timer event + * which is scheduled to wake up that CPU. In case of a completely + * idle system the next event might even be infinite time into the + * future. wake_up_idle_cpu() ensures that the CPU is woken up and + * leaves the inner idle loop so the newly added timer is taken into + * account when the CPU goes back to idle and evaluates the timer + * wheel for the next timer event. + */ +void wake_up_idle_cpu(int cpu) +{ + if (cpu == smp_processor_id()) + return; + + set_tsk_need_resched(cpu_rq(cpu)->idle); + smp_send_reschedule(cpu); +} + +void wake_up_nohz_cpu(int cpu) +{ + wake_up_idle_cpu(cpu); +} +#endif /* CONFIG_NO_HZ_COMMON */ + +/* + * Change a given task's CPU affinity. Migrate the thread to a + * proper CPU and schedule it away if the CPU it's executing on + * is removed from the allowed bitmask. + * + * NOTE: the caller must have a valid reference to the task, the + * task must not exit() & deallocate itself prematurely. The + * call is not atomic; no spinlocks may be held. + */ +static int __set_cpus_allowed_ptr(struct task_struct *p, + const struct cpumask *new_mask, bool check) +{ + bool running_wrong = false; + bool queued = false; + unsigned long flags; + struct rq *rq; + int ret = 0; + + rq = task_grq_lock(p, &flags); + + /* + * Must re-check here, to close a race against __kthread_bind(), + * sched_setaffinity() is not guaranteed to observe the flag. + */ + if (check && (p->flags & PF_NO_SETAFFINITY)) { + ret = -EINVAL; + goto out; + } + + if (cpumask_equal(tsk_cpus_allowed(p), new_mask)) + goto out; + + if (!cpumask_intersects(new_mask, cpu_active_mask)) { + ret = -EINVAL; + goto out; + } + + queued = task_queued(p); + + do_set_cpus_allowed(p, new_mask); + + /* Can the task run on the task's current CPU? If so, we're done */ + if (cpumask_test_cpu(task_cpu(p), new_mask)) + goto out; + + if (task_running(p)) { + /* Task is running on the wrong cpu now, reschedule it. */ + if (rq == this_rq()) { + set_tsk_need_resched(p); + running_wrong = true; + } else + resched_task(p); + } else + set_task_cpu(p, cpumask_any_and(cpu_active_mask, new_mask)); + +out: + if (queued) + try_preempt(p, rq); + task_grq_unlock(&flags); + + if (running_wrong) + preempt_schedule_common(); + + return ret; +} + +int set_cpus_allowed_ptr(struct task_struct *p, const struct cpumask *new_mask) +{ + return __set_cpus_allowed_ptr(p, new_mask, false); +} +EXPORT_SYMBOL_GPL(set_cpus_allowed_ptr); + +#ifdef CONFIG_HOTPLUG_CPU +/* Run through task list and find tasks affined to the dead cpu, then remove + * that cpu from the list, enable cpu0 and set the zerobound flag. */ +static void bind_zero(int src_cpu) +{ + struct task_struct *p, *t; + int bound = 0; + + if (src_cpu == 0) + return; + + do_each_thread(t, p) { + if (cpumask_test_cpu(src_cpu, tsk_cpus_allowed(p))) { + cpumask_clear_cpu(src_cpu, tsk_cpus_allowed(p)); + cpumask_set_cpu(0, tsk_cpus_allowed(p)); + p->zerobound = true; + bound++; + } + clear_sticky(p); + } while_each_thread(t, p); + + if (bound) { + printk(KERN_INFO "Removed affinity for %d processes to cpu %d\n", + bound, src_cpu); + } +} + +/* Find processes with the zerobound flag and reenable their affinity for the + * CPU coming alive. */ +static void unbind_zero(int src_cpu) +{ + int unbound = 0, zerobound = 0; + struct task_struct *p, *t; + + if (src_cpu == 0) + return; + + do_each_thread(t, p) { + if (!p->mm) + p->zerobound = false; + if (p->zerobound) { + unbound++; + cpumask_set_cpu(src_cpu, tsk_cpus_allowed(p)); + /* Once every CPU affinity has been re-enabled, remove + * the zerobound flag */ + if (cpumask_subset(cpu_possible_mask, tsk_cpus_allowed(p))) { + p->zerobound = false; + zerobound++; + } + } + } while_each_thread(t, p); + + if (unbound) { + printk(KERN_INFO "Added affinity for %d processes to cpu %d\n", + unbound, src_cpu); + } + if (zerobound) { + printk(KERN_INFO "Released forced binding to cpu0 for %d processes\n", + zerobound); + } +} + +/* + * Ensures that the idle task is using init_mm right before its cpu goes + * offline. + */ +void idle_task_exit(void) +{ + struct mm_struct *mm = current->active_mm; + + BUG_ON(cpu_online(smp_processor_id())); + + if (mm != &init_mm) { + switch_mm(mm, &init_mm, current); + finish_arch_post_lock_switch(); + } + mmdrop(mm); +} +#else /* CONFIG_HOTPLUG_CPU */ +static void unbind_zero(int src_cpu) {} +#endif /* CONFIG_HOTPLUG_CPU */ + +void sched_set_stop_task(int cpu, struct task_struct *stop) +{ + struct sched_param stop_param = { .sched_priority = STOP_PRIO }; + struct sched_param start_param = { .sched_priority = 0 }; + struct task_struct *old_stop = cpu_rq(cpu)->stop; + + if (stop) { + /* + * Make it appear like a SCHED_FIFO task, its something + * userspace knows about and won't get confused about. + * + * Also, it will make PI more or less work without too + * much confusion -- but then, stop work should not + * rely on PI working anyway. + */ + sched_setscheduler_nocheck(stop, SCHED_FIFO, &stop_param); + } + + cpu_rq(cpu)->stop = stop; + + if (old_stop) { + /* + * Reset it back to a normal scheduling policy so that + * it can die in pieces. + */ + sched_setscheduler_nocheck(old_stop, SCHED_NORMAL, &start_param); + } +} + + +#if defined(CONFIG_SCHED_DEBUG) && defined(CONFIG_SYSCTL) + +static struct ctl_table sd_ctl_dir[] = { + { + .procname = "sched_domain", + .mode = 0555, + }, + {} +}; + +static struct ctl_table sd_ctl_root[] = { + { + .procname = "kernel", + .mode = 0555, + .child = sd_ctl_dir, + }, + {} +}; + +static struct ctl_table *sd_alloc_ctl_entry(int n) +{ + struct ctl_table *entry = + kcalloc(n, sizeof(struct ctl_table), GFP_KERNEL); + + return entry; +} + +static void sd_free_ctl_entry(struct ctl_table **tablep) +{ + struct ctl_table *entry; + + /* + * In the intermediate directories, both the child directory and + * procname are dynamically allocated and could fail but the mode + * will always be set. In the lowest directory the names are + * static strings and all have proc handlers. + */ + for (entry = *tablep; entry->mode; entry++) { + if (entry->child) + sd_free_ctl_entry(&entry->child); + if (entry->proc_handler == NULL) + kfree(entry->procname); + } + + kfree(*tablep); + *tablep = NULL; +} + +static void +set_table_entry(struct ctl_table *entry, + const char *procname, void *data, int maxlen, + mode_t mode, proc_handler *proc_handler) +{ + entry->procname = procname; + entry->data = data; + entry->maxlen = maxlen; + entry->mode = mode; + entry->proc_handler = proc_handler; +} + +static struct ctl_table * +sd_alloc_ctl_domain_table(struct sched_domain *sd) +{ + struct ctl_table *table = sd_alloc_ctl_entry(14); + + if (table == NULL) + return NULL; + + set_table_entry(&table[0], "min_interval", &sd->min_interval, + sizeof(long), 0644, proc_doulongvec_minmax); + set_table_entry(&table[1], "max_interval", &sd->max_interval, + sizeof(long), 0644, proc_doulongvec_minmax); + set_table_entry(&table[2], "busy_idx", &sd->busy_idx, + sizeof(int), 0644, proc_dointvec_minmax); + set_table_entry(&table[3], "idle_idx", &sd->idle_idx, + sizeof(int), 0644, proc_dointvec_minmax); + set_table_entry(&table[4], "newidle_idx", &sd->newidle_idx, + sizeof(int), 0644, proc_dointvec_minmax); + set_table_entry(&table[5], "wake_idx", &sd->wake_idx, + sizeof(int), 0644, proc_dointvec_minmax); + set_table_entry(&table[6], "forkexec_idx", &sd->forkexec_idx, + sizeof(int), 0644, proc_dointvec_minmax); + set_table_entry(&table[7], "busy_factor", &sd->busy_factor, + sizeof(int), 0644, proc_dointvec_minmax); + set_table_entry(&table[8], "imbalance_pct", &sd->imbalance_pct, + sizeof(int), 0644, proc_dointvec_minmax); + set_table_entry(&table[9], "cache_nice_tries", + &sd->cache_nice_tries, + sizeof(int), 0644, proc_dointvec_minmax); + set_table_entry(&table[10], "flags", &sd->flags, + sizeof(int), 0644, proc_dointvec_minmax); + set_table_entry(&table[11], "max_newidle_lb_cost", + &sd->max_newidle_lb_cost, + sizeof(long), 0644, proc_doulongvec_minmax); + set_table_entry(&table[12], "name", sd->name, + CORENAME_MAX_SIZE, 0444, proc_dostring); + /* &table[13] is terminator */ + + return table; +} + +static struct ctl_table *sd_alloc_ctl_cpu_table(int cpu) +{ + struct ctl_table *entry, *table; + struct sched_domain *sd; + int domain_num = 0, i; + char buf[32]; + + for_each_domain(cpu, sd) + domain_num++; + entry = table = sd_alloc_ctl_entry(domain_num + 1); + if (table == NULL) + return NULL; + + i = 0; + for_each_domain(cpu, sd) { + snprintf(buf, 32, "domain%d", i); + entry->procname = kstrdup(buf, GFP_KERNEL); + entry->mode = 0555; + entry->child = sd_alloc_ctl_domain_table(sd); + entry++; + i++; + } + return table; +} + +static struct ctl_table_header *sd_sysctl_header; +static void register_sched_domain_sysctl(void) +{ + int i, cpu_num = num_possible_cpus(); + struct ctl_table *entry = sd_alloc_ctl_entry(cpu_num + 1); + char buf[32]; + + WARN_ON(sd_ctl_dir[0].child); + sd_ctl_dir[0].child = entry; + + if (entry == NULL) + return; + + for_each_possible_cpu(i) { + snprintf(buf, 32, "cpu%d", i); + entry->procname = kstrdup(buf, GFP_KERNEL); + entry->mode = 0555; + entry->child = sd_alloc_ctl_cpu_table(i); + entry++; + } + + WARN_ON(sd_sysctl_header); + sd_sysctl_header = register_sysctl_table(sd_ctl_root); +} + +/* may be called multiple times per register */ +static void unregister_sched_domain_sysctl(void) +{ + unregister_sysctl_table(sd_sysctl_header); + sd_sysctl_header = NULL; + if (sd_ctl_dir[0].child) + sd_free_ctl_entry(&sd_ctl_dir[0].child); +} +#else /* CONFIG_SCHED_DEBUG && CONFIG_SYSCTL */ +static void register_sched_domain_sysctl(void) +{ +} +static void unregister_sched_domain_sysctl(void) +{ +} +#endif /* CONFIG_SCHED_DEBUG && CONFIG_SYSCTL */ + +static void set_rq_online(struct rq *rq) +{ + if (!rq->online) { + cpumask_set_cpu(cpu_of(rq), rq->rd->online); + rq->online = true; + } +} + +static void set_rq_offline(struct rq *rq) +{ + if (rq->online) { + cpumask_clear_cpu(cpu_of(rq), rq->rd->online); + rq->online = false; + } +} + +/* + * migration_call - callback that gets triggered when a CPU is added. + */ +static int +migration_call(struct notifier_block *nfb, unsigned long action, void *hcpu) +{ + int cpu = (long)hcpu; + unsigned long flags; + struct rq *rq = cpu_rq(cpu); +#ifdef CONFIG_HOTPLUG_CPU + struct task_struct *idle = rq->idle; +#endif + + switch (action & ~CPU_TASKS_FROZEN) { + case CPU_STARTING: + return NOTIFY_OK; + case CPU_UP_PREPARE: + break; + + case CPU_ONLINE: + /* Update our root-domain */ + grq_lock_irqsave(&flags); + if (rq->rd) { + BUG_ON(!cpumask_test_cpu(cpu, rq->rd->span)); + + set_rq_online(rq); + } + unbind_zero(cpu); + grq.noc = num_online_cpus(); + grq_unlock_irqrestore(&flags); + break; + +#ifdef CONFIG_HOTPLUG_CPU + case CPU_DEAD: + grq_lock_irq(); + set_rq_task(rq, idle); + update_clocks(rq); + grq_unlock_irq(); + break; + + case CPU_DYING: + /* Update our root-domain */ + grq_lock_irqsave(&flags); + if (rq->rd) { + BUG_ON(!cpumask_test_cpu(cpu, rq->rd->span)); + set_rq_offline(rq); + } + bind_zero(cpu); + grq.noc = num_online_cpus(); + grq_unlock_irqrestore(&flags); + break; +#endif + } + return NOTIFY_OK; +} + +/* + * Register at high priority so that task migration (migrate_all_tasks) + * happens before everything else. This has to be lower priority than + * the notifier in the perf_counter subsystem, though. + */ +static struct notifier_block migration_notifier = { + .notifier_call = migration_call, + .priority = CPU_PRI_MIGRATION, +}; + +static int sched_cpu_active(struct notifier_block *nfb, + unsigned long action, void *hcpu) +{ + int cpu = (long)hcpu; + + switch (action & ~CPU_TASKS_FROZEN) { + case CPU_STARTING: + return NOTIFY_OK; + case CPU_ONLINE: + /* + * At this point a starting CPU has marked itself as online via + * set_cpu_online(). But it might not yet have marked itself + * as active, which is essential from here on. + */ + set_cpu_active(cpu, true); + stop_machine_unpark(cpu); + return NOTIFY_OK; + + case CPU_DOWN_FAILED: + set_cpu_active(cpu, true); + return NOTIFY_OK; + default: + return NOTIFY_DONE; + } +} + +static int sched_cpu_inactive(struct notifier_block *nfb, + unsigned long action, void *hcpu) +{ + switch (action & ~CPU_TASKS_FROZEN) { + case CPU_DOWN_PREPARE: + set_cpu_active((long)hcpu, false); + return NOTIFY_OK; + default: + return NOTIFY_DONE; + } +} + +int __init migration_init(void) +{ + void *cpu = (void *)(long)smp_processor_id(); + int err; + + /* Initialise migration for the boot CPU */ + err = migration_call(&migration_notifier, CPU_UP_PREPARE, cpu); + BUG_ON(err == NOTIFY_BAD); + migration_call(&migration_notifier, CPU_ONLINE, cpu); + register_cpu_notifier(&migration_notifier); + + /* Register cpu active notifiers */ + cpu_notifier(sched_cpu_active, CPU_PRI_SCHED_ACTIVE); + cpu_notifier(sched_cpu_inactive, CPU_PRI_SCHED_INACTIVE); + + return 0; +} +early_initcall(migration_init); + +static cpumask_var_t sched_domains_tmpmask; /* sched_domains_mutex */ + +#ifdef CONFIG_SCHED_DEBUG + +static __read_mostly int sched_debug_enabled; + +static int __init sched_debug_setup(char *str) +{ + sched_debug_enabled = 1; + + return 0; +} +early_param("sched_debug", sched_debug_setup); + +static inline bool sched_debug(void) +{ + return sched_debug_enabled; +} + +static int sched_domain_debug_one(struct sched_domain *sd, int cpu, int level, + struct cpumask *groupmask) +{ + cpumask_clear(groupmask); + + printk(KERN_DEBUG "%*s domain %d: ", level, "", level); + + if (!(sd->flags & SD_LOAD_BALANCE)) { + printk("does not load-balance\n"); + if (sd->parent) + printk(KERN_ERR "ERROR: !SD_LOAD_BALANCE domain" + " has parent"); + return -1; + } + + printk(KERN_CONT "span %*pbl level %s\n", + cpumask_pr_args(sched_domain_span(sd)), sd->name); + + if (!cpumask_test_cpu(cpu, sched_domain_span(sd))) { + printk(KERN_ERR "ERROR: domain->span does not contain " + "CPU%d\n", cpu); + } + + printk(KERN_CONT "\n"); + + if (!cpumask_equal(sched_domain_span(sd), groupmask)) + printk(KERN_ERR "ERROR: groups don't span domain->span\n"); + + if (sd->parent && + !cpumask_subset(groupmask, sched_domain_span(sd->parent))) + printk(KERN_ERR "ERROR: parent span is not a superset " + "of domain->span\n"); + return 0; +} + +static void sched_domain_debug(struct sched_domain *sd, int cpu) +{ + int level = 0; + + if (!sched_debug_enabled) + return; + + if (!sd) { + printk(KERN_DEBUG "CPU%d attaching NULL sched-domain.\n", cpu); + return; + } + + printk(KERN_DEBUG "CPU%d attaching sched-domain:\n", cpu); + + for (;;) { + if (sched_domain_debug_one(sd, cpu, level, sched_domains_tmpmask)) + break; + level++; + sd = sd->parent; + if (!sd) + break; + } +} +#else /* !CONFIG_SCHED_DEBUG */ +# define sched_domain_debug(sd, cpu) do { } while (0) +static inline bool sched_debug(void) +{ + return false; +} +#endif /* CONFIG_SCHED_DEBUG */ + +static int sd_degenerate(struct sched_domain *sd) +{ + if (cpumask_weight(sched_domain_span(sd)) == 1) + return 1; + + /* Following flags don't use groups */ + if (sd->flags & (SD_WAKE_AFFINE)) + return 0; + + return 1; +} + +static int +sd_parent_degenerate(struct sched_domain *sd, struct sched_domain *parent) +{ + unsigned long cflags = sd->flags, pflags = parent->flags; + + if (sd_degenerate(parent)) + return 1; + + if (!cpumask_equal(sched_domain_span(sd), sched_domain_span(parent))) + return 0; + + if (~cflags & pflags) + return 0; + + return 1; +} + +static void free_rootdomain(struct rcu_head *rcu) +{ + struct root_domain *rd = container_of(rcu, struct root_domain, rcu); + + cpupri_cleanup(&rd->cpupri); + free_cpumask_var(rd->rto_mask); + free_cpumask_var(rd->online); + free_cpumask_var(rd->span); + kfree(rd); +} + +static void rq_attach_root(struct rq *rq, struct root_domain *rd) +{ + struct root_domain *old_rd = NULL; + unsigned long flags; + + grq_lock_irqsave(&flags); + + if (rq->rd) { + old_rd = rq->rd; + + if (cpumask_test_cpu(rq->cpu, old_rd->online)) + set_rq_offline(rq); + + cpumask_clear_cpu(rq->cpu, old_rd->span); + + /* + * If we dont want to free the old_rd yet then + * set old_rd to NULL to skip the freeing later + * in this function: + */ + if (!atomic_dec_and_test(&old_rd->refcount)) + old_rd = NULL; + } + + atomic_inc(&rd->refcount); + rq->rd = rd; + + cpumask_set_cpu(rq->cpu, rd->span); + if (cpumask_test_cpu(rq->cpu, cpu_active_mask)) + set_rq_online(rq); + + grq_unlock_irqrestore(&flags); + + if (old_rd) + call_rcu_sched(&old_rd->rcu, free_rootdomain); +} + +static int init_rootdomain(struct root_domain *rd) +{ + memset(rd, 0, sizeof(*rd)); + + if (!zalloc_cpumask_var(&rd->span, GFP_KERNEL)) + goto out; + if (!zalloc_cpumask_var(&rd->online, GFP_KERNEL)) + goto free_span; + if (!zalloc_cpumask_var(&rd->rto_mask, GFP_KERNEL)) + goto free_online; + + if (cpupri_init(&rd->cpupri) != 0) + goto free_rto_mask; + return 0; + +free_rto_mask: + free_cpumask_var(rd->rto_mask); +free_online: + free_cpumask_var(rd->online); +free_span: + free_cpumask_var(rd->span); +out: + return -ENOMEM; +} + +static void init_defrootdomain(void) +{ + init_rootdomain(&def_root_domain); + + atomic_set(&def_root_domain.refcount, 1); +} + +static struct root_domain *alloc_rootdomain(void) +{ + struct root_domain *rd; + + rd = kmalloc(sizeof(*rd), GFP_KERNEL); + if (!rd) + return NULL; + + if (init_rootdomain(rd) != 0) { + kfree(rd); + return NULL; + } + + return rd; +} + +static void free_sched_domain(struct rcu_head *rcu) +{ + struct sched_domain *sd = container_of(rcu, struct sched_domain, rcu); + + kfree(sd); +} + +static void destroy_sched_domain(struct sched_domain *sd, int cpu) +{ + call_rcu(&sd->rcu, free_sched_domain); +} + +static void destroy_sched_domains(struct sched_domain *sd, int cpu) +{ + for (; sd; sd = sd->parent) + destroy_sched_domain(sd, cpu); +} + +/* + * Attach the domain 'sd' to 'cpu' as its base domain. Callers must + * hold the hotplug lock. + */ +static void +cpu_attach_domain(struct sched_domain *sd, struct root_domain *rd, int cpu) +{ + struct rq *rq = cpu_rq(cpu); + struct sched_domain *tmp; + + /* Remove the sched domains which do not contribute to scheduling. */ + for (tmp = sd; tmp; ) { + struct sched_domain *parent = tmp->parent; + if (!parent) + break; + + if (sd_parent_degenerate(tmp, parent)) { + tmp->parent = parent->parent; + if (parent->parent) + parent->parent->child = tmp; + /* + * Transfer SD_PREFER_SIBLING down in case of a + * degenerate parent; the spans match for this + * so the property transfers. + */ + if (parent->flags & SD_PREFER_SIBLING) + tmp->flags |= SD_PREFER_SIBLING; + destroy_sched_domain(parent, cpu); + } else + tmp = tmp->parent; + } + + if (sd && sd_degenerate(sd)) { + tmp = sd; + sd = sd->parent; + destroy_sched_domain(tmp, cpu); + if (sd) + sd->child = NULL; + } + + sched_domain_debug(sd, cpu); + + rq_attach_root(rq, rd); + tmp = rq->sd; + rcu_assign_pointer(rq->sd, sd); + destroy_sched_domains(tmp, cpu); +} + +/* Setup the mask of cpus configured for isolated domains */ +static int __init isolated_cpu_setup(char *str) +{ + alloc_bootmem_cpumask_var(&cpu_isolated_map); + cpulist_parse(str, cpu_isolated_map); + return 1; +} + +__setup("isolcpus=", isolated_cpu_setup); + +struct s_data { + struct sched_domain ** __percpu sd; + struct root_domain *rd; +}; + +enum s_alloc { + sa_rootdomain, + sa_sd, + sa_sd_storage, + sa_none, +}; + +/* + * Initializers for schedule domains + * Non-inlined to reduce accumulated stack pressure in build_sched_domains() + */ + +static int default_relax_domain_level = -1; +int sched_domain_level_max; + +static int __init setup_relax_domain_level(char *str) +{ + if (kstrtoint(str, 0, &default_relax_domain_level)) + pr_warn("Unable to set relax_domain_level\n"); + + return 1; +} +__setup("relax_domain_level=", setup_relax_domain_level); + +static void set_domain_attribute(struct sched_domain *sd, + struct sched_domain_attr *attr) +{ + int request; + + if (!attr || attr->relax_domain_level < 0) { + if (default_relax_domain_level < 0) + return; + else + request = default_relax_domain_level; + } else + request = attr->relax_domain_level; + if (request < sd->level) { + /* turn off idle balance on this domain */ + sd->flags &= ~(SD_BALANCE_WAKE|SD_BALANCE_NEWIDLE); + } else { + /* turn on idle balance on this domain */ + sd->flags |= (SD_BALANCE_WAKE|SD_BALANCE_NEWIDLE); + } +} + +static void __sdt_free(const struct cpumask *cpu_map); +static int __sdt_alloc(const struct cpumask *cpu_map); + +static void __free_domain_allocs(struct s_data *d, enum s_alloc what, + const struct cpumask *cpu_map) +{ + switch (what) { + case sa_rootdomain: + if (!atomic_read(&d->rd->refcount)) + free_rootdomain(&d->rd->rcu); /* fall through */ + case sa_sd: + free_percpu(d->sd); /* fall through */ + case sa_sd_storage: + __sdt_free(cpu_map); /* fall through */ + case sa_none: + break; + } +} + +static enum s_alloc __visit_domain_allocation_hell(struct s_data *d, + const struct cpumask *cpu_map) +{ + memset(d, 0, sizeof(*d)); + + if (__sdt_alloc(cpu_map)) + return sa_sd_storage; + d->sd = alloc_percpu(struct sched_domain *); + if (!d->sd) + return sa_sd_storage; + d->rd = alloc_rootdomain(); + if (!d->rd) + return sa_sd; + return sa_rootdomain; +} + +/* + * NULL the sd_data elements we've used to build the sched_domain + * structure so that the subsequent __free_domain_allocs() + * will not free the data we're using. + */ +static void claim_allocations(int cpu, struct sched_domain *sd) +{ + struct sd_data *sdd = sd->private; + + WARN_ON_ONCE(*per_cpu_ptr(sdd->sd, cpu) != sd); + *per_cpu_ptr(sdd->sd, cpu) = NULL; +} + +#ifdef CONFIG_NUMA +static int sched_domains_numa_levels; +static int *sched_domains_numa_distance; +static struct cpumask ***sched_domains_numa_masks; +static int sched_domains_curr_level; +#endif + +/* + * SD_flags allowed in topology descriptions. + * + * SD_SHARE_CPUCAPACITY - describes SMT topologies + * SD_SHARE_PKG_RESOURCES - describes shared caches + * SD_NUMA - describes NUMA topologies + * SD_SHARE_POWERDOMAIN - describes shared power domain + * + * Odd one out: + * SD_ASYM_PACKING - describes SMT quirks + */ +#define TOPOLOGY_SD_FLAGS \ + (SD_SHARE_CPUCAPACITY | \ + SD_SHARE_PKG_RESOURCES | \ + SD_NUMA | \ + SD_ASYM_PACKING | \ + SD_SHARE_POWERDOMAIN) + +static struct sched_domain * +sd_init(struct sched_domain_topology_level *tl, int cpu) +{ + struct sched_domain *sd = *per_cpu_ptr(tl->data.sd, cpu); + int sd_weight, sd_flags = 0; + +#ifdef CONFIG_NUMA + /* + * Ugly hack to pass state to sd_numa_mask()... + */ + sched_domains_curr_level = tl->numa_level; +#endif + + sd_weight = cpumask_weight(tl->mask(cpu)); + + if (tl->sd_flags) + sd_flags = (*tl->sd_flags)(); + if (WARN_ONCE(sd_flags & ~TOPOLOGY_SD_FLAGS, + "wrong sd_flags in topology description\n")) + sd_flags &= ~TOPOLOGY_SD_FLAGS; + + *sd = (struct sched_domain){ + .min_interval = sd_weight, + .max_interval = 2*sd_weight, + .busy_factor = 32, + .imbalance_pct = 125, + + .cache_nice_tries = 0, + .busy_idx = 0, + .idle_idx = 0, + .newidle_idx = 0, + .wake_idx = 0, + .forkexec_idx = 0, + + .flags = 1*SD_LOAD_BALANCE + | 1*SD_BALANCE_NEWIDLE + | 1*SD_BALANCE_EXEC + | 1*SD_BALANCE_FORK + | 0*SD_BALANCE_WAKE + | 1*SD_WAKE_AFFINE + | 0*SD_SHARE_CPUCAPACITY + | 0*SD_SHARE_PKG_RESOURCES + | 0*SD_SERIALIZE + | 0*SD_PREFER_SIBLING + | 0*SD_NUMA + | sd_flags + , + + .last_balance = jiffies, + .balance_interval = sd_weight, + .smt_gain = 0, + .max_newidle_lb_cost = 0, + .next_decay_max_lb_cost = jiffies, +#ifdef CONFIG_SCHED_DEBUG + .name = tl->name, +#endif + }; + + /* + * Convert topological properties into behaviour. + */ + + if (sd->flags & SD_SHARE_CPUCAPACITY) { + sd->flags |= SD_PREFER_SIBLING; + sd->imbalance_pct = 110; + sd->smt_gain = 1178; /* ~15% */ + + } else if (sd->flags & SD_SHARE_PKG_RESOURCES) { + sd->imbalance_pct = 117; + sd->cache_nice_tries = 1; + sd->busy_idx = 2; + +#ifdef CONFIG_NUMA + } else if (sd->flags & SD_NUMA) { + sd->cache_nice_tries = 2; + sd->busy_idx = 3; + sd->idle_idx = 2; + + sd->flags |= SD_SERIALIZE; + if (sched_domains_numa_distance[tl->numa_level] > RECLAIM_DISTANCE) { + sd->flags &= ~(SD_BALANCE_EXEC | + SD_BALANCE_FORK | + SD_WAKE_AFFINE); + } + +#endif + } else { + sd->flags |= SD_PREFER_SIBLING; + sd->cache_nice_tries = 1; + sd->busy_idx = 2; + sd->idle_idx = 1; + } + + sd->private = &tl->data; + + return sd; +} + +/* + * Topology list, bottom-up. + */ +static struct sched_domain_topology_level default_topology[] = { +#ifdef CONFIG_SCHED_SMT + { cpu_smt_mask, cpu_smt_flags, SD_INIT_NAME(SMT) }, +#endif +#ifdef CONFIG_SCHED_MC + { cpu_coregroup_mask, cpu_core_flags, SD_INIT_NAME(MC) }, +#endif + { cpu_cpu_mask, SD_INIT_NAME(DIE) }, + { NULL, }, +}; + +static struct sched_domain_topology_level *sched_domain_topology = + default_topology; + +#define for_each_sd_topology(tl) \ + for (tl = sched_domain_topology; tl->mask; tl++) + +void set_sched_topology(struct sched_domain_topology_level *tl) +{ + sched_domain_topology = tl; +} + +#ifdef CONFIG_NUMA + +static const struct cpumask *sd_numa_mask(int cpu) +{ + return sched_domains_numa_masks[sched_domains_curr_level][cpu_to_node(cpu)]; +} + +static void sched_numa_warn(const char *str) +{ + static int done = false; + int i,j; + + if (done) + return; + + done = true; + + printk(KERN_WARNING "ERROR: %s\n\n", str); + + for (i = 0; i < nr_node_ids; i++) { + printk(KERN_WARNING " "); + for (j = 0; j < nr_node_ids; j++) + printk(KERN_CONT "%02d ", node_distance(i,j)); + printk(KERN_CONT "\n"); + } + printk(KERN_WARNING "\n"); +} + +static bool find_numa_distance(int distance) +{ + int i; + + if (distance == node_distance(0, 0)) + return true; + + for (i = 0; i < sched_domains_numa_levels; i++) { + if (sched_domains_numa_distance[i] == distance) + return true; + } + + return false; +} + +static void sched_init_numa(void) +{ + int next_distance, curr_distance = node_distance(0, 0); + struct sched_domain_topology_level *tl; + int level = 0; + int i, j, k; + + sched_domains_numa_distance = kzalloc(sizeof(int) * nr_node_ids, GFP_KERNEL); + if (!sched_domains_numa_distance) + return; + + /* + * O(nr_nodes^2) deduplicating selection sort -- in order to find the + * unique distances in the node_distance() table. + * + * Assumes node_distance(0,j) includes all distances in + * node_distance(i,j) in order to avoid cubic time. + */ + next_distance = curr_distance; + for (i = 0; i < nr_node_ids; i++) { + for (j = 0; j < nr_node_ids; j++) { + for (k = 0; k < nr_node_ids; k++) { + int distance = node_distance(i, k); + + if (distance > curr_distance && + (distance < next_distance || + next_distance == curr_distance)) + next_distance = distance; + + /* + * While not a strong assumption it would be nice to know + * about cases where if node A is connected to B, B is not + * equally connected to A. + */ + if (sched_debug() && node_distance(k, i) != distance) + sched_numa_warn("Node-distance not symmetric"); + + if (sched_debug() && i && !find_numa_distance(distance)) + sched_numa_warn("Node-0 not representative"); + } + if (next_distance != curr_distance) { + sched_domains_numa_distance[level++] = next_distance; + sched_domains_numa_levels = level; + curr_distance = next_distance; + } else break; + } + + /* + * In case of sched_debug() we verify the above assumption. + */ + if (!sched_debug()) + break; + } + /* + * 'level' contains the number of unique distances, excluding the + * identity distance node_distance(i,i). + * + * The sched_domains_numa_distance[] array includes the actual distance + * numbers. + */ + + /* + * Here, we should temporarily reset sched_domains_numa_levels to 0. + * If it fails to allocate memory for array sched_domains_numa_masks[][], + * the array will contain less then 'level' members. This could be + * dangerous when we use it to iterate array sched_domains_numa_masks[][] + * in other functions. + * + * We reset it to 'level' at the end of this function. + */ + sched_domains_numa_levels = 0; + + sched_domains_numa_masks = kzalloc(sizeof(void *) * level, GFP_KERNEL); + if (!sched_domains_numa_masks) + return; + + /* + * Now for each level, construct a mask per node which contains all + * cpus of nodes that are that many hops away from us. + */ + for (i = 0; i < level; i++) { + sched_domains_numa_masks[i] = + kzalloc(nr_node_ids * sizeof(void *), GFP_KERNEL); + if (!sched_domains_numa_masks[i]) + return; + + for (j = 0; j < nr_node_ids; j++) { + struct cpumask *mask = kzalloc(cpumask_size(), GFP_KERNEL); + if (!mask) + return; + + sched_domains_numa_masks[i][j] = mask; + + for_each_node(k) { + if (node_distance(j, k) > sched_domains_numa_distance[i]) + continue; + + cpumask_or(mask, mask, cpumask_of_node(k)); + } + } + } + + /* Compute default topology size */ + for (i = 0; sched_domain_topology[i].mask; i++); + + tl = kzalloc((i + level + 1) * + sizeof(struct sched_domain_topology_level), GFP_KERNEL); + if (!tl) + return; + + /* + * Copy the default topology bits.. + */ + for (i = 0; sched_domain_topology[i].mask; i++) + tl[i] = sched_domain_topology[i]; + + /* + * .. and append 'j' levels of NUMA goodness. + */ + for (j = 0; j < level; i++, j++) { + tl[i] = (struct sched_domain_topology_level){ + .mask = sd_numa_mask, + .sd_flags = cpu_numa_flags, + .flags = SDTL_OVERLAP, + .numa_level = j, + SD_INIT_NAME(NUMA) + }; + } + + sched_domain_topology = tl; + + sched_domains_numa_levels = level; +} + +static void sched_domains_numa_masks_set(int cpu) +{ + int i, j; + int node = cpu_to_node(cpu); + + for (i = 0; i < sched_domains_numa_levels; i++) { + for (j = 0; j < nr_node_ids; j++) { + if (node_distance(j, node) <= sched_domains_numa_distance[i]) + cpumask_set_cpu(cpu, sched_domains_numa_masks[i][j]); + } + } +} + +static void sched_domains_numa_masks_clear(int cpu) +{ + int i, j; + for (i = 0; i < sched_domains_numa_levels; i++) { + for (j = 0; j < nr_node_ids; j++) + cpumask_clear_cpu(cpu, sched_domains_numa_masks[i][j]); + } +} + +/* + * Update sched_domains_numa_masks[level][node] array when new cpus + * are onlined. + */ +static int sched_domains_numa_masks_update(struct notifier_block *nfb, + unsigned long action, + void *hcpu) +{ + int cpu = (long)hcpu; + + switch (action & ~CPU_TASKS_FROZEN) { + case CPU_ONLINE: + sched_domains_numa_masks_set(cpu); + break; + + case CPU_DEAD: + sched_domains_numa_masks_clear(cpu); + break; + + default: + return NOTIFY_DONE; + } + + return NOTIFY_OK; +} +#else +static inline void sched_init_numa(void) +{ +} + +static int sched_domains_numa_masks_update(struct notifier_block *nfb, + unsigned long action, + void *hcpu) +{ + return 0; +} +#endif /* CONFIG_NUMA */ + +static int __sdt_alloc(const struct cpumask *cpu_map) +{ + struct sched_domain_topology_level *tl; + int j; + + for_each_sd_topology(tl) { + struct sd_data *sdd = &tl->data; + + sdd->sd = alloc_percpu(struct sched_domain *); + if (!sdd->sd) + return -ENOMEM; + + for_each_cpu(j, cpu_map) { + struct sched_domain *sd; + + sd = kzalloc_node(sizeof(struct sched_domain) + cpumask_size(), + GFP_KERNEL, cpu_to_node(j)); + if (!sd) + return -ENOMEM; + + *per_cpu_ptr(sdd->sd, j) = sd; + } + } + + return 0; +} + +static void __sdt_free(const struct cpumask *cpu_map) +{ + struct sched_domain_topology_level *tl; + int j; + + for_each_sd_topology(tl) { + struct sd_data *sdd = &tl->data; + + for_each_cpu(j, cpu_map) { + struct sched_domain *sd; + + if (sdd->sd) { + sd = *per_cpu_ptr(sdd->sd, j); + kfree(*per_cpu_ptr(sdd->sd, j)); + } + } + free_percpu(sdd->sd); + sdd->sd = NULL; + } +} + +struct sched_domain *build_sched_domain(struct sched_domain_topology_level *tl, + const struct cpumask *cpu_map, struct sched_domain_attr *attr, + struct sched_domain *child, int cpu) +{ + struct sched_domain *sd = sd_init(tl, cpu); + if (!sd) + return child; + + cpumask_and(sched_domain_span(sd), cpu_map, tl->mask(cpu)); + if (child) { + sd->level = child->level + 1; + sched_domain_level_max = max(sched_domain_level_max, sd->level); + child->parent = sd; + sd->child = child; + + if (!cpumask_subset(sched_domain_span(child), + sched_domain_span(sd))) { + pr_err("BUG: arch topology borken\n"); +#ifdef CONFIG_SCHED_DEBUG + pr_err(" the %s domain not a subset of the %s domain\n", + child->name, sd->name); +#endif + /* Fixup, ensure @sd has at least @child cpus. */ + cpumask_or(sched_domain_span(sd), + sched_domain_span(sd), + sched_domain_span(child)); + } + + } + set_domain_attribute(sd, attr); + + return sd; +} + +/* + * Build sched domains for a given set of cpus and attach the sched domains + * to the individual cpus + */ +static int build_sched_domains(const struct cpumask *cpu_map, + struct sched_domain_attr *attr) +{ + enum s_alloc alloc_state; + struct sched_domain *sd; + struct s_data d; + int i, ret = -ENOMEM; + + alloc_state = __visit_domain_allocation_hell(&d, cpu_map); + if (alloc_state != sa_rootdomain) + goto error; + + /* Set up domains for cpus specified by the cpu_map. */ + for_each_cpu(i, cpu_map) { + struct sched_domain_topology_level *tl; + + sd = NULL; + for_each_sd_topology(tl) { + sd = build_sched_domain(tl, cpu_map, attr, sd, i); + if (tl == sched_domain_topology) + *per_cpu_ptr(d.sd, i) = sd; + if (tl->flags & SDTL_OVERLAP) + sd->flags |= SD_OVERLAP; + if (cpumask_equal(cpu_map, sched_domain_span(sd))) + break; + } + } + + /* Calculate CPU capacity for physical packages and nodes */ + for (i = nr_cpumask_bits-1; i >= 0; i--) { + if (!cpumask_test_cpu(i, cpu_map)) + continue; + + for (sd = *per_cpu_ptr(d.sd, i); sd; sd = sd->parent) { + claim_allocations(i, sd); + } + } + + /* Attach the domains */ + rcu_read_lock(); + for_each_cpu(i, cpu_map) { + sd = *per_cpu_ptr(d.sd, i); + cpu_attach_domain(sd, d.rd, i); + } + rcu_read_unlock(); + + ret = 0; +error: + __free_domain_allocs(&d, alloc_state, cpu_map); + return ret; +} + +static cpumask_var_t *doms_cur; /* current sched domains */ +static int ndoms_cur; /* number of sched domains in 'doms_cur' */ +static struct sched_domain_attr *dattr_cur; + /* attribues of custom domains in 'doms_cur' */ + +/* + * Special case: If a kmalloc of a doms_cur partition (array of + * cpumask) fails, then fallback to a single sched domain, + * as determined by the single cpumask fallback_doms. + */ +static cpumask_var_t fallback_doms; + +/* + * arch_update_cpu_topology lets virtualized architectures update the + * cpu core maps. It is supposed to return 1 if the topology changed + * or 0 if it stayed the same. + */ +int __weak arch_update_cpu_topology(void) +{ + return 0; +} + +cpumask_var_t *alloc_sched_domains(unsigned int ndoms) +{ + int i; + cpumask_var_t *doms; + + doms = kmalloc(sizeof(*doms) * ndoms, GFP_KERNEL); + if (!doms) + return NULL; + for (i = 0; i < ndoms; i++) { + if (!alloc_cpumask_var(&doms[i], GFP_KERNEL)) { + free_sched_domains(doms, i); + return NULL; + } + } + return doms; +} + +void free_sched_domains(cpumask_var_t doms[], unsigned int ndoms) +{ + unsigned int i; + for (i = 0; i < ndoms; i++) + free_cpumask_var(doms[i]); + kfree(doms); +} + +/* + * Set up scheduler domains and groups. Callers must hold the hotplug lock. + * For now this just excludes isolated cpus, but could be used to + * exclude other special cases in the future. + */ +static int init_sched_domains(const struct cpumask *cpu_map) +{ + int err; + + arch_update_cpu_topology(); + ndoms_cur = 1; + doms_cur = alloc_sched_domains(ndoms_cur); + if (!doms_cur) + doms_cur = &fallback_doms; + cpumask_andnot(doms_cur[0], cpu_map, cpu_isolated_map); + err = build_sched_domains(doms_cur[0], NULL); + register_sched_domain_sysctl(); + + return err; +} + +/* + * Detach sched domains from a group of cpus specified in cpu_map + * These cpus will now be attached to the NULL domain + */ +static void detach_destroy_domains(const struct cpumask *cpu_map) +{ + int i; + + rcu_read_lock(); + for_each_cpu(i, cpu_map) + cpu_attach_domain(NULL, &def_root_domain, i); + rcu_read_unlock(); +} + +/* handle null as "default" */ +static int dattrs_equal(struct sched_domain_attr *cur, int idx_cur, + struct sched_domain_attr *new, int idx_new) +{ + struct sched_domain_attr tmp; + + /* fast path */ + if (!new && !cur) + return 1; + + tmp = SD_ATTR_INIT; + return !memcmp(cur ? (cur + idx_cur) : &tmp, + new ? (new + idx_new) : &tmp, + sizeof(struct sched_domain_attr)); +} + +/* + * Partition sched domains as specified by the 'ndoms_new' + * cpumasks in the array doms_new[] of cpumasks. This compares + * doms_new[] to the current sched domain partitioning, doms_cur[]. + * It destroys each deleted domain and builds each new domain. + * + * 'doms_new' is an array of cpumask_var_t's of length 'ndoms_new'. + * The masks don't intersect (don't overlap.) We should setup one + * sched domain for each mask. CPUs not in any of the cpumasks will + * not be load balanced. If the same cpumask appears both in the + * current 'doms_cur' domains and in the new 'doms_new', we can leave + * it as it is. + * + * The passed in 'doms_new' should be allocated using + * alloc_sched_domains. This routine takes ownership of it and will + * free_sched_domains it when done with it. If the caller failed the + * alloc call, then it can pass in doms_new == NULL && ndoms_new == 1, + * and partition_sched_domains() will fallback to the single partition + * 'fallback_doms', it also forces the domains to be rebuilt. + * + * If doms_new == NULL it will be replaced with cpu_online_mask. + * ndoms_new == 0 is a special case for destroying existing domains, + * and it will not create the default domain. + * + * Call with hotplug lock held + */ +void partition_sched_domains(int ndoms_new, cpumask_var_t doms_new[], + struct sched_domain_attr *dattr_new) +{ + int i, j, n; + int new_topology; + + mutex_lock(&sched_domains_mutex); + + /* always unregister in case we don't destroy any domains */ + unregister_sched_domain_sysctl(); + + /* Let architecture update cpu core mappings. */ + new_topology = arch_update_cpu_topology(); + + n = doms_new ? ndoms_new : 0; + + /* Destroy deleted domains */ + for (i = 0; i < ndoms_cur; i++) { + for (j = 0; j < n && !new_topology; j++) { + if (cpumask_equal(doms_cur[i], doms_new[j]) + && dattrs_equal(dattr_cur, i, dattr_new, j)) + goto match1; + } + /* no match - a current sched domain not in new doms_new[] */ + detach_destroy_domains(doms_cur[i]); +match1: + ; + } + + n = ndoms_cur; + if (doms_new == NULL) { + n = 0; + doms_new = &fallback_doms; + cpumask_andnot(doms_new[0], cpu_active_mask, cpu_isolated_map); + WARN_ON_ONCE(dattr_new); + } + + /* Build new domains */ + for (i = 0; i < ndoms_new; i++) { + for (j = 0; j < n && !new_topology; j++) { + if (cpumask_equal(doms_new[i], doms_cur[j]) + && dattrs_equal(dattr_new, i, dattr_cur, j)) + goto match2; + } + /* no match - add a new doms_new */ + build_sched_domains(doms_new[i], dattr_new ? dattr_new + i : NULL); +match2: + ; + } + + /* Remember the new sched domains */ + if (doms_cur != &fallback_doms) + free_sched_domains(doms_cur, ndoms_cur); + kfree(dattr_cur); /* kfree(NULL) is safe */ + doms_cur = doms_new; + dattr_cur = dattr_new; + ndoms_cur = ndoms_new; + + register_sched_domain_sysctl(); + + mutex_unlock(&sched_domains_mutex); +} + +static int num_cpus_frozen; /* used to mark begin/end of suspend/resume */ + +/* + * Update cpusets according to cpu_active mask. If cpusets are + * disabled, cpuset_update_active_cpus() becomes a simple wrapper + * around partition_sched_domains(). + * + * If we come here as part of a suspend/resume, don't touch cpusets because we + * want to restore it back to its original state upon resume anyway. + */ +static int cpuset_cpu_active(struct notifier_block *nfb, unsigned long action, + void *hcpu) +{ + switch (action) { + case CPU_ONLINE_FROZEN: + case CPU_DOWN_FAILED_FROZEN: + + /* + * num_cpus_frozen tracks how many CPUs are involved in suspend + * resume sequence. As long as this is not the last online + * operation in the resume sequence, just build a single sched + * domain, ignoring cpusets. + */ + num_cpus_frozen--; + if (likely(num_cpus_frozen)) { + partition_sched_domains(1, NULL, NULL); + break; + } + + /* + * This is the last CPU online operation. So fall through and + * restore the original sched domains by considering the + * cpuset configurations. + */ + + case CPU_ONLINE: + cpuset_update_active_cpus(true); + break; + default: + return NOTIFY_DONE; + } + return NOTIFY_OK; +} + +static int cpuset_cpu_inactive(struct notifier_block *nfb, unsigned long action, + void *hcpu) +{ + switch (action) { + case CPU_DOWN_PREPARE: + cpuset_update_active_cpus(false); + break; + case CPU_DOWN_PREPARE_FROZEN: + num_cpus_frozen++; + partition_sched_domains(1, NULL, NULL); + break; + default: + return NOTIFY_DONE; + } + return NOTIFY_OK; +} + +#if defined(CONFIG_SCHED_SMT) || defined(CONFIG_SCHED_MC) +/* + * Cheaper version of the below functions in case support for SMT and MC is + * compiled in but CPUs have no siblings. + */ +static bool sole_cpu_idle(int cpu) +{ + return rq_idle(cpu_rq(cpu)); +} +#endif +#ifdef CONFIG_SCHED_SMT +static const cpumask_t *thread_cpumask(int cpu) +{ + return topology_sibling_cpumask(cpu); +} +/* All this CPU's SMT siblings are idle */ +static bool siblings_cpu_idle(int cpu) +{ + return cpumask_subset(thread_cpumask(cpu), &grq.cpu_idle_map); +} +#endif +#ifdef CONFIG_SCHED_MC +static const cpumask_t *core_cpumask(int cpu) +{ + return topology_core_cpumask(cpu); +} +/* All this CPU's shared cache siblings are idle */ +static bool cache_cpu_idle(int cpu) +{ + return cpumask_subset(core_cpumask(cpu), &grq.cpu_idle_map); +} +#endif + +enum sched_domain_level { + SD_LV_NONE = 0, + SD_LV_SIBLING, + SD_LV_MC, + SD_LV_BOOK, + SD_LV_CPU, + SD_LV_NODE, + SD_LV_ALLNODES, + SD_LV_MAX +}; + +void __init sched_init_smp(void) +{ + struct sched_domain *sd; + int cpu, other_cpu; + + cpumask_var_t non_isolated_cpus; + + alloc_cpumask_var(&non_isolated_cpus, GFP_KERNEL); + alloc_cpumask_var(&fallback_doms, GFP_KERNEL); + + sched_init_numa(); + + /* + * There's no userspace yet to cause hotplug operations; hence all the + * cpu masks are stable and all blatant races in the below code cannot + * happen. + */ + mutex_lock(&sched_domains_mutex); + init_sched_domains(cpu_active_mask); + cpumask_andnot(non_isolated_cpus, cpu_possible_mask, cpu_isolated_map); + if (cpumask_empty(non_isolated_cpus)) + cpumask_set_cpu(smp_processor_id(), non_isolated_cpus); + mutex_unlock(&sched_domains_mutex); + + hotcpu_notifier(sched_domains_numa_masks_update, CPU_PRI_SCHED_ACTIVE); + hotcpu_notifier(cpuset_cpu_active, CPU_PRI_CPUSET_ACTIVE); + hotcpu_notifier(cpuset_cpu_inactive, CPU_PRI_CPUSET_INACTIVE); + + /* Move init over to a non-isolated CPU */ + if (set_cpus_allowed_ptr(current, non_isolated_cpus) < 0) + BUG(); + free_cpumask_var(non_isolated_cpus); + + mutex_lock(&sched_domains_mutex); + grq_lock_irq(); + /* + * Set up the relative cache distance of each online cpu from each + * other in a simple array for quick lookup. Locality is determined + * by the closest sched_domain that CPUs are separated by. CPUs with + * shared cache in SMT and MC are treated as local. Separate CPUs + * (within the same package or physically) within the same node are + * treated as not local. CPUs not even in the same domain (different + * nodes) are treated as very distant. + */ + for_each_online_cpu(cpu) { + struct rq *rq = cpu_rq(cpu); + + /* First check if this cpu is in the same node */ + for_each_domain(cpu, sd) { + if (sd->level > SD_LV_NODE) + continue; + /* Set locality to local node if not already found lower */ + for_each_cpu(other_cpu, sched_domain_span(sd)) { + if (rq->cpu_locality[other_cpu] > 3) + rq->cpu_locality[other_cpu] = 3; + } + } + + /* + * Each runqueue has its own function in case it doesn't have + * siblings of its own allowing mixed topologies. + */ +#ifdef CONFIG_SCHED_MC + for_each_cpu(other_cpu, core_cpumask(cpu)) { + if (rq->cpu_locality[other_cpu] > 2) + rq->cpu_locality[other_cpu] = 2; + } + if (cpumask_weight(core_cpumask(cpu)) > 1) + rq->cache_idle = cache_cpu_idle; +#endif +#ifdef CONFIG_SCHED_SMT + for_each_cpu(other_cpu, thread_cpumask(cpu)) + rq->cpu_locality[other_cpu] = 1; + if (cpumask_weight(thread_cpumask(cpu)) > 1) + rq->siblings_idle = siblings_cpu_idle; +#endif + } + grq_unlock_irq(); + mutex_unlock(&sched_domains_mutex); + + for_each_online_cpu(cpu) { + struct rq *rq = cpu_rq(cpu); + for_each_online_cpu(other_cpu) { + if (other_cpu <= cpu) + continue; + printk(KERN_DEBUG "BFS LOCALITY CPU %d to %d: %d\n", cpu, other_cpu, rq->cpu_locality[other_cpu]); + } + } +} +#else +void __init sched_init_smp(void) +{ +} +#endif /* CONFIG_SMP */ + +int in_sched_functions(unsigned long addr) +{ + return in_lock_functions(addr) || + (addr >= (unsigned long)__sched_text_start + && addr < (unsigned long)__sched_text_end); +} + +void __init sched_init(void) +{ +#ifdef CONFIG_SMP + int cpu_ids; +#endif + int i; + struct rq *rq; + + prio_ratios[0] = 128; + for (i = 1 ; i < NICE_WIDTH ; i++) + prio_ratios[i] = prio_ratios[i - 1] * 11 / 10; + + raw_spin_lock_init(&grq.lock); + grq.nr_running = grq.nr_uninterruptible = grq.nr_switches = 0; + grq.niffies = 0; + grq.last_jiffy = jiffies; + raw_spin_lock_init(&grq.iso_lock); + grq.iso_ticks = 0; + grq.iso_refractory = false; + grq.noc = 1; +#ifdef CONFIG_SMP + init_defrootdomain(); + grq.qnr = grq.idle_cpus = 0; + cpumask_clear(&grq.cpu_idle_map); +#else + uprq = &per_cpu(runqueues, 0); +#endif + for_each_possible_cpu(i) { + rq = cpu_rq(i); + rq->grq_lock = &grq.lock; + rq->user_pc = rq->nice_pc = rq->softirq_pc = rq->system_pc = + rq->iowait_pc = rq->idle_pc = 0; + rq->dither = false; +#ifdef CONFIG_SMP + rq->sticky_task = NULL; + rq->last_niffy = 0; + rq->sd = NULL; + rq->rd = NULL; + rq->online = false; + rq->cpu = i; + rq_attach_root(rq, &def_root_domain); +#endif + atomic_set(&rq->nr_iowait, 0); + } + +#ifdef CONFIG_SMP + cpu_ids = i; + /* + * Set the base locality for cpu cache distance calculation to + * "distant" (3). Make sure the distance from a CPU to itself is 0. + */ + for_each_possible_cpu(i) { + int j; + + rq = cpu_rq(i); +#ifdef CONFIG_SCHED_SMT + rq->siblings_idle = sole_cpu_idle; +#endif +#ifdef CONFIG_SCHED_MC + rq->cache_idle = sole_cpu_idle; +#endif + rq->cpu_locality = kmalloc(cpu_ids * sizeof(int *), GFP_ATOMIC); + for_each_possible_cpu(j) { + if (i == j) + rq->cpu_locality[j] = 0; + else + rq->cpu_locality[j] = 4; + } + } +#endif + + for (i = 0; i < PRIO_LIMIT; i++) + INIT_LIST_HEAD(grq.queue + i); + /* delimiter for bitsearch */ + __set_bit(PRIO_LIMIT, grq.prio_bitmap); + +#ifdef CONFIG_PREEMPT_NOTIFIERS + INIT_HLIST_HEAD(&init_task.preempt_notifiers); +#endif + + /* + * The boot idle thread does lazy MMU switching as well: + */ + atomic_inc(&init_mm.mm_count); + enter_lazy_tlb(&init_mm, current); + + /* + * Make us the idle thread. Technically, schedule() should not be + * called from this thread, however somewhere below it might be, + * but because we are the idle thread, we just pick up running again + * when this runqueue becomes "idle". + */ + init_idle(current, smp_processor_id()); + +#ifdef CONFIG_SMP + zalloc_cpumask_var(&sched_domains_tmpmask, GFP_NOWAIT); + /* May be allocated at isolcpus cmdline parse time */ + if (cpu_isolated_map == NULL) + zalloc_cpumask_var(&cpu_isolated_map, GFP_NOWAIT); + idle_thread_set_boot_cpu(); +#endif /* SMP */ +} + +#ifdef CONFIG_DEBUG_ATOMIC_SLEEP +static inline int preempt_count_equals(int preempt_offset) +{ + int nested = preempt_count() + rcu_preempt_depth(); + + return (nested == preempt_offset); +} + +void __might_sleep(const char *file, int line, int preempt_offset) +{ + /* + * Blocking primitives will set (and therefore destroy) current->state, + * since we will exit with TASK_RUNNING make sure we enter with it, + * otherwise we will destroy state. + */ + WARN_ONCE(current->state != TASK_RUNNING && current->task_state_change, + "do not call blocking ops when !TASK_RUNNING; " + "state=%lx set at [<%p>] %pS\n", + current->state, + (void *)current->task_state_change, + (void *)current->task_state_change); + + ___might_sleep(file, line, preempt_offset); +} +EXPORT_SYMBOL(__might_sleep); + +void ___might_sleep(const char *file, int line, int preempt_offset) +{ + static unsigned long prev_jiffy; /* ratelimiting */ + + rcu_sleep_check(); /* WARN_ON_ONCE() by default, no rate limit reqd. */ + if ((preempt_count_equals(preempt_offset) && !irqs_disabled() && + !is_idle_task(current)) || + system_state != SYSTEM_RUNNING || oops_in_progress) + return; + if (time_before(jiffies, prev_jiffy + HZ) && prev_jiffy) + return; + prev_jiffy = jiffies; + + printk(KERN_ERR + "BUG: sleeping function called from invalid context at %s:%d\n", + file, line); + printk(KERN_ERR + "in_atomic(): %d, irqs_disabled(): %d, pid: %d, name: %s\n", + in_atomic(), irqs_disabled(), + current->pid, current->comm); + + if (task_stack_end_corrupted(current)) + printk(KERN_EMERG "Thread overran stack, or stack corrupted\n"); + + debug_show_held_locks(current); + if (irqs_disabled()) + print_irqtrace_events(current); +#ifdef CONFIG_DEBUG_PREEMPT + if (!preempt_count_equals(preempt_offset)) { + pr_err("Preemption disabled at:"); + print_ip_sym(current->preempt_disable_ip); + pr_cont("\n"); + } +#endif + dump_stack(); +} +EXPORT_SYMBOL(___might_sleep); +#endif + +#ifdef CONFIG_MAGIC_SYSRQ +static inline void normalise_rt_tasks(void) +{ + struct task_struct *g, *p; + unsigned long flags; + struct rq *rq; + int queued; + + read_lock(&tasklist_lock); + for_each_process_thread(g, p) { + /* + * Only normalize user tasks: + */ + if (p->flags & PF_KTHREAD) + continue; + + if (!rt_task(p) && !iso_task(p)) + continue; + + rq = task_grq_lock(p, &flags); + queued = task_queued(p); + if (queued) + dequeue_task(p); + __setscheduler(p, rq, SCHED_NORMAL, 0, false); + if (queued) { + enqueue_task(p, rq); + try_preempt(p, rq); + } + + task_grq_unlock(&flags); + } + read_unlock(&tasklist_lock); +} + +void normalize_rt_tasks(void) +{ + normalise_rt_tasks(); +} +#endif /* CONFIG_MAGIC_SYSRQ */ + +#if defined(CONFIG_IA64) || defined(CONFIG_KGDB_KDB) +/* + * These functions are only useful for the IA64 MCA handling, or kdb. + * + * They can only be called when the whole system has been + * stopped - every CPU needs to be quiescent, and no scheduling + * activity can take place. Using them for anything else would + * be a serious bug, and as a result, they aren't even visible + * under any other configuration. + */ + +/** + * curr_task - return the current task for a given cpu. + * @cpu: the processor in question. + * + * ONLY VALID WHEN THE WHOLE SYSTEM IS STOPPED! + * + * Return: The current task for @cpu. + */ +struct task_struct *curr_task(int cpu) +{ + return cpu_curr(cpu); +} + +#endif /* defined(CONFIG_IA64) || defined(CONFIG_KGDB_KDB) */ + +#ifdef CONFIG_IA64 +/** + * set_curr_task - set the current task for a given cpu. + * @cpu: the processor in question. + * @p: the task pointer to set. + * + * Description: This function must only be used when non-maskable interrupts + * are serviced on a separate stack. It allows the architecture to switch the + * notion of the current task on a cpu in a non-blocking manner. This function + * must be called with all CPU's synchronised, and interrupts disabled, the + * and caller must save the original value of the current task (see + * curr_task() above) and restore that value before reenabling interrupts and + * re-starting the system. + * + * ONLY VALID WHEN THE WHOLE SYSTEM IS STOPPED! + */ +void set_curr_task(int cpu, struct task_struct *p) +{ + cpu_curr(cpu) = p; +} + +#endif + +/* + * Use precise platform statistics if available: + */ +#ifdef CONFIG_VIRT_CPU_ACCOUNTING_NATIVE +void task_cputime_adjusted(struct task_struct *p, cputime_t *ut, cputime_t *st) +{ + *ut = p->utime; + *st = p->stime; +} +EXPORT_SYMBOL_GPL(task_cputime_adjusted); + +void thread_group_cputime_adjusted(struct task_struct *p, cputime_t *ut, cputime_t *st) +{ + struct task_cputime cputime; + + thread_group_cputime(p, &cputime); + + *ut = cputime.utime; + *st = cputime.stime; +} + +void vtime_account_system_irqsafe(struct task_struct *tsk) +{ + unsigned long flags; + + local_irq_save(flags); + vtime_account_system(tsk); + local_irq_restore(flags); +} +EXPORT_SYMBOL_GPL(vtime_account_system_irqsafe); + +#ifndef __ARCH_HAS_VTIME_TASK_SWITCH +void vtime_task_switch(struct task_struct *prev) +{ + if (is_idle_task(prev)) + vtime_account_idle(prev); + else + vtime_account_system(prev); + + vtime_account_user(prev); + arch_vtime_task_switch(prev); +} +#endif + +#else +/* + * Perform (stime * rtime) / total, but avoid multiplication overflow by + * losing precision when the numbers are big. + */ +static cputime_t scale_stime(u64 stime, u64 rtime, u64 total) +{ + u64 scaled; + + for (;;) { + /* Make sure "rtime" is the bigger of stime/rtime */ + if (stime > rtime) { + u64 tmp = rtime; rtime = stime; stime = tmp; + } + + /* Make sure 'total' fits in 32 bits */ + if (total >> 32) + goto drop_precision; + + /* Does rtime (and thus stime) fit in 32 bits? */ + if (!(rtime >> 32)) + break; + + /* Can we just balance rtime/stime rather than dropping bits? */ + if (stime >> 31) + goto drop_precision; + + /* We can grow stime and shrink rtime and try to make them both fit */ + stime <<= 1; + rtime >>= 1; + continue; + +drop_precision: + /* We drop from rtime, it has more bits than stime */ + rtime >>= 1; + total >>= 1; + } + + /* + * Make sure gcc understands that this is a 32x32->64 multiply, + * followed by a 64/32->64 divide. + */ + scaled = div_u64((u64) (u32) stime * (u64) (u32) rtime, (u32)total); + return (__force cputime_t) scaled; +} + +/* + * Adjust tick based cputime random precision against scheduler + * runtime accounting. + */ +static void cputime_adjust(struct task_cputime *curr, + struct prev_cputime *prev, + cputime_t *ut, cputime_t *st) +{ + cputime_t rtime, stime, utime, total; + + stime = curr->stime; + total = stime + curr->utime; + + /* + * Tick based cputime accounting depend on random scheduling + * timeslices of a task to be interrupted or not by the timer. + * Depending on these circumstances, the number of these interrupts + * may be over or under-optimistic, matching the real user and system + * cputime with a variable precision. + * + * Fix this by scaling these tick based values against the total + * runtime accounted by the CFS scheduler. + */ + rtime = nsecs_to_cputime(curr->sum_exec_runtime); + + /* + * Update userspace visible utime/stime values only if actual execution + * time is bigger than already exported. Note that can happen, that we + * provided bigger values due to scaling inaccuracy on big numbers. + */ + if (prev->stime + prev->utime >= rtime) + goto out; + + if (total) { + stime = scale_stime((__force u64)stime, + (__force u64)rtime, (__force u64)total); + utime = rtime - stime; + } else { + stime = rtime; + utime = 0; + } + + /* + * If the tick based count grows faster than the scheduler one, + * the result of the scaling may go backward. + * Let's enforce monotonicity. + */ + prev->stime = max(prev->stime, stime); + prev->utime = max(prev->utime, utime); + +out: + *ut = prev->utime; + *st = prev->stime; +} + +void task_cputime_adjusted(struct task_struct *p, cputime_t *ut, cputime_t *st) +{ + struct task_cputime cputime = { + .sum_exec_runtime = tsk_seruntime(p), + }; + + task_cputime(p, &cputime.utime, &cputime.stime); + cputime_adjust(&cputime, &p->prev_cputime, ut, st); +} +EXPORT_SYMBOL_GPL(task_cputime_adjusted); + +/* + * Must be called with siglock held. + */ +void thread_group_cputime_adjusted(struct task_struct *p, cputime_t *ut, cputime_t *st) +{ + struct task_cputime cputime; + + thread_group_cputime(p, &cputime); + cputime_adjust(&cputime, &p->signal->prev_cputime, ut, st); +} +#endif + +void init_idle_bootup_task(struct task_struct *idle) +{} + +#ifdef CONFIG_SCHED_DEBUG +void proc_sched_show_task(struct task_struct *p, struct seq_file *m) +{} + +void proc_sched_set_task(struct task_struct *p) +{} +#endif + +#ifdef CONFIG_SMP +#define SCHED_LOAD_SHIFT (10) +#define SCHED_LOAD_SCALE (1L << SCHED_LOAD_SHIFT) + +unsigned long default_scale_freq_power(struct sched_domain *sd, int cpu) +{ + return SCHED_LOAD_SCALE; +} + +unsigned long default_scale_smt_power(struct sched_domain *sd, int cpu) +{ + unsigned long weight = cpumask_weight(sched_domain_span(sd)); + unsigned long smt_gain = sd->smt_gain; + + smt_gain /= weight; + + return smt_gain; +} +#endif diff --git a/kernel/sched/bfs_sched.h b/kernel/sched/bfs_sched.h new file mode 100644 index 000000000..5d97f7919 --- /dev/null +++ b/kernel/sched/bfs_sched.h @@ -0,0 +1,181 @@ +#include <linux/sched.h> +#include <linux/cpuidle.h> +#include <linux/stop_machine.h> + +#ifndef BFS_SCHED_H +#define BFS_SCHED_H + +/* + * This is the main, per-CPU runqueue data structure. + * This data should only be modified by the local cpu. + */ +struct rq { + struct task_struct *curr, *idle, *stop; + struct mm_struct *prev_mm; + + /* Pointer to grq spinlock */ + raw_spinlock_t *grq_lock; + + /* Stored data about rq->curr to work outside grq lock */ + u64 rq_deadline; + unsigned int rq_policy; + int rq_time_slice; + u64 rq_last_ran; + int rq_prio; + bool rq_running; /* There is a task running */ + int soft_affined; /* Running or queued tasks with this set as their rq */ +#ifdef CONFIG_SMT_NICE + struct mm_struct *rq_mm; + int rq_smt_bias; /* Policy/nice level bias across smt siblings */ +#endif + /* Accurate timekeeping data */ + u64 timekeep_clock; + unsigned long user_pc, nice_pc, irq_pc, softirq_pc, system_pc, + iowait_pc, idle_pc; + atomic_t nr_iowait; + +#ifdef CONFIG_SMP + int cpu; /* cpu of this runqueue */ + bool online; + bool scaling; /* This CPU is managed by a scaling CPU freq governor */ + struct task_struct *sticky_task; + + struct root_domain *rd; + struct sched_domain *sd; + int *cpu_locality; /* CPU relative cache distance */ +#ifdef CONFIG_SCHED_SMT + bool (*siblings_idle)(int cpu); + /* See if all smt siblings are idle */ +#endif /* CONFIG_SCHED_SMT */ +#ifdef CONFIG_SCHED_MC + bool (*cache_idle)(int cpu); + /* See if all cache siblings are idle */ +#endif /* CONFIG_SCHED_MC */ + u64 last_niffy; /* Last time this RQ updated grq.niffies */ +#endif /* CONFIG_SMP */ +#ifdef CONFIG_IRQ_TIME_ACCOUNTING + u64 prev_irq_time; +#endif /* CONFIG_IRQ_TIME_ACCOUNTING */ +#ifdef CONFIG_PARAVIRT + u64 prev_steal_time; +#endif /* CONFIG_PARAVIRT */ +#ifdef CONFIG_PARAVIRT_TIME_ACCOUNTING + u64 prev_steal_time_rq; +#endif /* CONFIG_PARAVIRT_TIME_ACCOUNTING */ + + u64 clock, old_clock, last_tick; + u64 clock_task; + bool dither; + +#ifdef CONFIG_SCHEDSTATS + + /* latency stats */ + struct sched_info rq_sched_info; + unsigned long long rq_cpu_time; + /* could above be rq->cfs_rq.exec_clock + rq->rt_rq.rt_runtime ? */ + + /* sys_sched_yield() stats */ + unsigned int yld_count; + + /* schedule() stats */ + unsigned int sched_switch; + unsigned int sched_count; + unsigned int sched_goidle; + + /* try_to_wake_up() stats */ + unsigned int ttwu_count; + unsigned int ttwu_local; +#endif /* CONFIG_SCHEDSTATS */ +#ifdef CONFIG_CPU_IDLE + /* Must be inspected within a rcu lock section */ + struct cpuidle_state *idle_state; +#endif +}; + +#ifdef CONFIG_SMP +struct rq *cpu_rq(int cpu); +#endif + +#ifndef CONFIG_SMP +extern struct rq *uprq; +#define cpu_rq(cpu) (uprq) +#define this_rq() (uprq) +#define raw_rq() (uprq) +#define task_rq(p) (uprq) +#define cpu_curr(cpu) ((uprq)->curr) +#else /* CONFIG_SMP */ +DECLARE_PER_CPU_SHARED_ALIGNED(struct rq, runqueues); +#define this_rq() this_cpu_ptr(&runqueues) +#define raw_rq() raw_cpu_ptr(&runqueues) +#endif /* CONFIG_SMP */ + +static inline u64 __rq_clock_broken(struct rq *rq) +{ + return READ_ONCE(rq->clock); +} + +static inline u64 rq_clock(struct rq *rq) +{ + lockdep_assert_held(rq->grq_lock); + return rq->clock; +} + +static inline u64 rq_clock_task(struct rq *rq) +{ + lockdep_assert_held(rq->grq_lock); + return rq->clock_task; +} + +extern struct mutex sched_domains_mutex; + +#define rcu_dereference_check_sched_domain(p) \ + rcu_dereference_check((p), \ + lockdep_is_held(&sched_domains_mutex)) + +/* + * The domain tree (rq->sd) is protected by RCU's quiescent state transition. + * See detach_destroy_domains: synchronize_sched for details. + * + * The domain tree of any CPU may only be accessed from within + * preempt-disabled sections. + */ +#define for_each_domain(cpu, __sd) \ + for (__sd = rcu_dereference_check_sched_domain(cpu_rq(cpu)->sd); __sd; __sd = __sd->parent) + +static inline void sched_ttwu_pending(void) { } + +static inline int task_on_rq_queued(struct task_struct *p) +{ + return p->on_rq; +} + +#ifdef CONFIG_SMP + +extern void set_cpus_allowed_common(struct task_struct *p, const struct cpumask *new_mask); + +#endif + +#ifdef CONFIG_CPU_IDLE +static inline void idle_set_state(struct rq *rq, + struct cpuidle_state *idle_state) +{ + rq->idle_state = idle_state; +} + +static inline struct cpuidle_state *idle_get_state(struct rq *rq) +{ + WARN_ON(!rcu_read_lock_held()); + return rq->idle_state; +} +#else +static inline void idle_set_state(struct rq *rq, + struct cpuidle_state *idle_state) +{ +} + +static inline struct cpuidle_state *idle_get_state(struct rq *rq) +{ + return NULL; +} +#endif +#endif /* BFS_SCHED_H */ diff --git a/kernel/sched/core.c b/kernel/sched/core.c index 05114b15b..544c98aac 100644 --- a/kernel/sched/core.c +++ b/kernel/sched/core.c @@ -5378,6 +5378,7 @@ migration_call(struct notifier_block *nfb, unsigned long action, void *hcpu) case CPU_UP_PREPARE: rq->calc_load_update = calc_load_update; + account_reset_rq(rq); break; case CPU_ONLINE: diff --git a/kernel/sched/idle.c b/kernel/sched/idle.c index 544a7133c..bbf813929 100644 --- a/kernel/sched/idle.c +++ b/kernel/sched/idle.c @@ -13,7 +13,11 @@ #include <trace/events/power.h> +#ifdef CONFIG_SCHED_BFS +#include "bfs_sched.h" +#else #include "sched.h" +#endif /** * sched_idle_set_state - Record idle state for the current CPU. diff --git a/kernel/sched/sched.h b/kernel/sched/sched.h index ef5875fff..60c5b6ba4 100644 --- a/kernel/sched/sched.h +++ b/kernel/sched/sched.h @@ -1773,3 +1773,16 @@ static inline u64 irq_time_read(int cpu) } #endif /* CONFIG_64BIT */ #endif /* CONFIG_IRQ_TIME_ACCOUNTING */ + +static inline void account_reset_rq(struct rq *rq) +{ +#ifdef CONFIG_IRQ_TIME_ACCOUNTING + rq->prev_irq_time = 0; +#endif +#ifdef CONFIG_PARAVIRT + rq->prev_steal_time = 0; +#endif +#ifdef CONFIG_PARAVIRT_TIME_ACCOUNTING + rq->prev_steal_time_rq = 0; +#endif +} diff --git a/kernel/sched/stats.c b/kernel/sched/stats.c index 87e2c9f0c..7466a0bb2 100644 --- a/kernel/sched/stats.c +++ b/kernel/sched/stats.c @@ -4,7 +4,11 @@ #include <linux/seq_file.h> #include <linux/proc_fs.h> +#ifndef CONFIG_SCHED_BFS #include "sched.h" +#else +#include "bfs_sched.h" +#endif /* * bump this up when changing the output format or the meaning of an existing diff --git a/kernel/sysctl.c b/kernel/sysctl.c index f5102fabe..c4887773a 100644 --- a/kernel/sysctl.c +++ b/kernel/sysctl.c @@ -125,7 +125,13 @@ static int __maybe_unused one = 1; static int __maybe_unused two = 2; static int __maybe_unused four = 4; static unsigned long one_ul = 1; -static int one_hundred = 100; +static int __maybe_unused one_hundred = 100; +#ifdef CONFIG_SCHED_BFS +extern int rr_interval; +extern int sched_interactive; +extern int sched_iso_cpu; +static int __read_mostly one_thousand = 1000; +#endif #ifdef CONFIG_PRINTK static int ten_thousand = 10000; #endif @@ -260,7 +266,7 @@ static struct ctl_table sysctl_base_table[] = { { } }; -#ifdef CONFIG_SCHED_DEBUG +#if defined(CONFIG_SCHED_DEBUG) && !defined(CONFIG_SCHED_BFS) static int min_sched_granularity_ns = 100000; /* 100 usecs */ static int max_sched_granularity_ns = NSEC_PER_SEC; /* 1 second */ static int min_wakeup_granularity_ns; /* 0 usecs */ @@ -277,6 +283,7 @@ static int max_extfrag_threshold = 1000; #endif static struct ctl_table kern_table[] = { +#ifndef CONFIG_SCHED_BFS { .procname = "sched_child_runs_first", .data = &sysctl_sched_child_runs_first, @@ -445,6 +452,7 @@ static struct ctl_table kern_table[] = { .extra1 = &one, }, #endif +#endif /* !CONFIG_SCHED_BFS */ #ifdef CONFIG_PROVE_LOCKING { .procname = "prove_locking", @@ -1002,6 +1010,35 @@ static struct ctl_table kern_table[] = { .proc_handler = proc_dointvec, }, #endif +#ifdef CONFIG_SCHED_BFS + { + .procname = "rr_interval", + .data = &rr_interval, + .maxlen = sizeof (int), + .mode = 0644, + .proc_handler = &proc_dointvec_minmax, + .extra1 = &one, + .extra2 = &one_thousand, + }, + { + .procname = "interactive", + .data = &sched_interactive, + .maxlen = sizeof(int), + .mode = 0644, + .proc_handler = &proc_dointvec_minmax, + .extra1 = &zero, + .extra2 = &one, + }, + { + .procname = "iso_cpu", + .data = &sched_iso_cpu, + .maxlen = sizeof (int), + .mode = 0644, + .proc_handler = &proc_dointvec_minmax, + .extra1 = &zero, + .extra2 = &one_hundred, + }, +#endif #if defined(CONFIG_S390) && defined(CONFIG_SMP) { .procname = "spin_retry", diff --git a/kernel/sysctl_binary.c b/kernel/sysctl_binary.c index 7e7746a42..10a1d7dc9 100644 --- a/kernel/sysctl_binary.c +++ b/kernel/sysctl_binary.c @@ -1321,7 +1321,7 @@ static ssize_t binary_sysctl(const int *name, int nlen, } mnt = task_active_pid_ns(current)->proc_mnt; - file = file_open_root(mnt->mnt_root, mnt, pathname, flags); + file = file_open_root(mnt->mnt_root, mnt, pathname, flags, 0); result = PTR_ERR(file); if (IS_ERR(file)) goto out_putname; diff --git a/kernel/task_work.c b/kernel/task_work.c index 53fa971d0..bce3211e7 100644 --- a/kernel/task_work.c +++ b/kernel/task_work.c @@ -118,3 +118,4 @@ void task_work_run(void) } while (work); } } +EXPORT_SYMBOL_GPL(task_work_run); diff --git a/kernel/time/Kconfig b/kernel/time/Kconfig index 4008d9f95..6931b6e3c 100644 --- a/kernel/time/Kconfig +++ b/kernel/time/Kconfig @@ -89,7 +89,7 @@ config NO_HZ_IDLE config NO_HZ_FULL bool "Full dynticks system (tickless)" # NO_HZ_COMMON dependency - depends on !ARCH_USES_GETTIMEOFFSET && GENERIC_CLOCKEVENTS + depends on !ARCH_USES_GETTIMEOFFSET && GENERIC_CLOCKEVENTS && !SCHED_BFS # We need at least one periodic CPU for timekeeping depends on SMP depends on HAVE_CONTEXT_TRACKING diff --git a/kernel/time/posix-cpu-timers.c b/kernel/time/posix-cpu-timers.c index f5e86d282..7a48442ed 100644 --- a/kernel/time/posix-cpu-timers.c +++ b/kernel/time/posix-cpu-timers.c @@ -448,7 +448,7 @@ static void cleanup_timers(struct list_head *head) */ void posix_cpu_timers_exit(struct task_struct *tsk) { - add_device_randomness((const void*) &tsk->se.sum_exec_runtime, + add_device_randomness((const void*) &tsk_seruntime(tsk), sizeof(unsigned long long)); cleanup_timers(tsk->cpu_timers); @@ -878,7 +878,7 @@ static void check_thread_timers(struct task_struct *tsk, tsk_expires->virt_exp = expires_to_cputime(expires); tsk_expires->sched_exp = check_timers_list(++timers, firing, - tsk->se.sum_exec_runtime); + tsk_seruntime(tsk)); /* * Check for the special case thread timers. @@ -889,7 +889,7 @@ static void check_thread_timers(struct task_struct *tsk, READ_ONCE(sig->rlim[RLIMIT_RTTIME].rlim_max); if (hard != RLIM_INFINITY && - tsk->rt.timeout > DIV_ROUND_UP(hard, USEC_PER_SEC/HZ)) { + tsk_rttimeout(tsk) > DIV_ROUND_UP(hard, USEC_PER_SEC/HZ)) { /* * At the hard limit, we just die. * No need to calculate anything else now. @@ -897,7 +897,7 @@ static void check_thread_timers(struct task_struct *tsk, __group_send_sig_info(SIGKILL, SEND_SIG_PRIV, tsk); return; } - if (tsk->rt.timeout > DIV_ROUND_UP(soft, USEC_PER_SEC/HZ)) { + if (tsk_rttimeout(tsk) > DIV_ROUND_UP(soft, USEC_PER_SEC/HZ)) { /* * At the soft limit, send a SIGXCPU every second. */ @@ -1144,7 +1144,7 @@ static inline int fastpath_timer_check(struct task_struct *tsk) struct task_cputime task_sample; task_cputime(tsk, &task_sample.utime, &task_sample.stime); - task_sample.sum_exec_runtime = tsk->se.sum_exec_runtime; + task_sample.sum_exec_runtime = tsk_seruntime(tsk); if (task_cputime_expired(&task_sample, &tsk->cputime_expires)) return 1; } diff --git a/kernel/trace/trace.c b/kernel/trace/trace.c index d9293402e..8305cbb2d 100644 --- a/kernel/trace/trace.c +++ b/kernel/trace/trace.c @@ -4949,7 +4949,10 @@ static ssize_t tracing_splice_read_pipe(struct file *filp, spd.nr_pages = i; - ret = splice_to_pipe(pipe, &spd); + if (i) + ret = splice_to_pipe(pipe, &spd); + else + ret = 0; out: splice_shrink_spd(&spd); return ret; diff --git a/kernel/trace/trace_irqsoff.c b/kernel/trace/trace_irqsoff.c index e4e56589e..be3222b7d 100644 --- a/kernel/trace/trace_irqsoff.c +++ b/kernel/trace/trace_irqsoff.c @@ -109,8 +109,12 @@ static int func_prolog_dec(struct trace_array *tr, return 0; local_save_flags(*flags); - /* slight chance to get a false positive on tracing_cpu */ - if (!irqs_disabled_flags(*flags)) + /* + * Slight chance to get a false positive on tracing_cpu, + * although I'm starting to think there isn't a chance. + * Leave this for now just to be paranoid. + */ + if (!irqs_disabled_flags(*flags) && !preempt_count()) return 0; *data = per_cpu_ptr(tr->trace_buffer.data, cpu); diff --git a/kernel/trace/trace_printk.c b/kernel/trace/trace_printk.c index 060df67db..f96f0383f 100644 --- a/kernel/trace/trace_printk.c +++ b/kernel/trace/trace_printk.c @@ -296,6 +296,9 @@ static int t_show(struct seq_file *m, void *v) const char *str = *fmt; int i; + if (!*fmt) + return 0; + seq_printf(m, "0x%lx : \"", *(unsigned long *)fmt); /* diff --git a/kernel/trace/trace_selftest.c b/kernel/trace/trace_selftest.c index b0f86ea77..287cf721c 100644 --- a/kernel/trace/trace_selftest.c +++ b/kernel/trace/trace_selftest.c @@ -1039,10 +1039,15 @@ static int trace_wakeup_test_thread(void *data) { /* Make this a -deadline thread */ static const struct sched_attr attr = { +#ifdef CONFIG_SCHED_BFS + /* No deadline on BFS, use RR */ + .sched_policy = SCHED_RR, +#else .sched_policy = SCHED_DEADLINE, .sched_runtime = 100000ULL, .sched_deadline = 10000000ULL, .sched_period = 10000000ULL +#endif }; struct wakeup_test_data *x = data; diff --git a/kernel/watchdog.c b/kernel/watchdog.c index b3ace6ebb..9acb29f28 100644 --- a/kernel/watchdog.c +++ b/kernel/watchdog.c @@ -923,6 +923,9 @@ static int proc_watchdog_common(int which, struct ctl_table *table, int write, * both lockup detectors are disabled if proc_watchdog_update() * returns an error. */ + if (old == new) + goto out; + err = proc_watchdog_update(); } out: @@ -967,7 +970,7 @@ int proc_soft_watchdog(struct ctl_table *table, int write, int proc_watchdog_thresh(struct ctl_table *table, int write, void __user *buffer, size_t *lenp, loff_t *ppos) { - int err, old; + int err, old, new; get_online_cpus(); mutex_lock(&watchdog_proc_mutex); @@ -987,6 +990,10 @@ int proc_watchdog_thresh(struct ctl_table *table, int write, /* * Update the sample period. Restore on failure. */ + new = ACCESS_ONCE(watchdog_thresh); + if (old == new) + goto out; + set_sample_period(); err = proc_watchdog_update(); if (err) { diff --git a/lib/Kconfig.debug b/lib/Kconfig.debug index 8bfd1aca7..897979c8e 100644 --- a/lib/Kconfig.debug +++ b/lib/Kconfig.debug @@ -1258,7 +1258,7 @@ config TORTURE_TEST config RCU_TORTURE_TEST tristate "torture tests for RCU" - depends on DEBUG_KERNEL + depends on DEBUG_KERNEL && !SCHED_BFS select TORTURE_TEST select SRCU select TASKS_RCU diff --git a/mm/Makefile b/mm/Makefile index 2ed43191f..e3a53f5ba 100644 --- a/mm/Makefile +++ b/mm/Makefile @@ -21,7 +21,7 @@ obj-y := filemap.o mempool.o oom_kill.o \ mm_init.o mmu_context.o percpu.o slab_common.o \ compaction.o vmacache.o \ interval_tree.o list_lru.o workingset.o \ - debug.o $(mmu-y) + prfile.o debug.o $(mmu-y) obj-y += init-mm.o diff --git a/mm/filemap.c b/mm/filemap.c index da7a35d83..f800f8749 100644 --- a/mm/filemap.c +++ b/mm/filemap.c @@ -2229,7 +2229,7 @@ int filemap_page_mkwrite(struct vm_area_struct *vma, struct vm_fault *vmf) int ret = VM_FAULT_LOCKED; sb_start_pagefault(inode->i_sb); - file_update_time(vma->vm_file); + vma_file_update_time(vma); lock_page(page); if (page->mapping != inode->i_mapping) { unlock_page(page); diff --git a/mm/memcontrol.c b/mm/memcontrol.c index d06cae2de..caf3bf73b 100644 --- a/mm/memcontrol.c +++ b/mm/memcontrol.c @@ -1262,7 +1262,7 @@ static unsigned long mem_cgroup_get_limit(struct mem_cgroup *memcg) return limit; } -static void mem_cgroup_out_of_memory(struct mem_cgroup *memcg, gfp_t gfp_mask, +static bool mem_cgroup_out_of_memory(struct mem_cgroup *memcg, gfp_t gfp_mask, int order) { struct oom_control oc = { @@ -1340,6 +1340,7 @@ static void mem_cgroup_out_of_memory(struct mem_cgroup *memcg, gfp_t gfp_mask, } unlock: mutex_unlock(&oom_lock); + return chosen; } #if MAX_NUMNODES > 1 @@ -5051,6 +5052,7 @@ static ssize_t memory_high_write(struct kernfs_open_file *of, char *buf, size_t nbytes, loff_t off) { struct mem_cgroup *memcg = mem_cgroup_from_css(of_css(of)); + unsigned long nr_pages; unsigned long high; int err; @@ -5061,6 +5063,11 @@ static ssize_t memory_high_write(struct kernfs_open_file *of, memcg->high = high; + nr_pages = page_counter_read(&memcg->memory); + if (nr_pages > high) + try_to_free_mem_cgroup_pages(memcg, nr_pages - high, + GFP_KERNEL, true); + memcg_wb_domain_size_changed(memcg); return nbytes; } @@ -5082,6 +5089,8 @@ static ssize_t memory_max_write(struct kernfs_open_file *of, char *buf, size_t nbytes, loff_t off) { struct mem_cgroup *memcg = mem_cgroup_from_css(of_css(of)); + unsigned int nr_reclaims = MEM_CGROUP_RECLAIM_RETRIES; + bool drained = false; unsigned long max; int err; @@ -5090,9 +5099,36 @@ static ssize_t memory_max_write(struct kernfs_open_file *of, if (err) return err; - err = mem_cgroup_resize_limit(memcg, max); - if (err) - return err; + xchg(&memcg->memory.limit, max); + + for (;;) { + unsigned long nr_pages = page_counter_read(&memcg->memory); + + if (nr_pages <= max) + break; + + if (signal_pending(current)) { + err = -EINTR; + break; + } + + if (!drained) { + drain_all_stock(memcg); + drained = true; + continue; + } + + if (nr_reclaims) { + if (!try_to_free_mem_cgroup_pages(memcg, nr_pages - max, + GFP_KERNEL, true)) + nr_reclaims--; + continue; + } + + mem_cgroup_events(memcg, MEMCG_OOM, 1); + if (!mem_cgroup_out_of_memory(memcg, GFP_KERNEL, 0)) + break; + } memcg_wb_domain_size_changed(memcg); return nbytes; diff --git a/mm/memory.c b/mm/memory.c index 8132787ae..3f7de66ed 100644 --- a/mm/memory.c +++ b/mm/memory.c @@ -2042,7 +2042,7 @@ static inline int wp_page_reuse(struct mm_struct *mm, } if (!page_mkwrite) - file_update_time(vma->vm_file); + vma_file_update_time(vma); } return VM_FAULT_WRITE; @@ -290,7 +290,7 @@ static struct vm_area_struct *remove_vma(struct vm_area_struct *vma) if (vma->vm_ops && vma->vm_ops->close) vma->vm_ops->close(vma); if (vma->vm_file) - fput(vma->vm_file); + vma_fput(vma); mpol_put(vma_policy(vma)); kmem_cache_free(vm_area_cachep, vma); return next; @@ -909,7 +909,7 @@ again: remove_next = 1 + (end > next->vm_end); if (remove_next) { if (file) { uprobe_munmap(next, next->vm_start, next->vm_end); - fput(file); + vma_fput(vma); } if (next->anon_vma) anon_vma_merge(vma, next); @@ -1683,8 +1683,8 @@ out: return addr; unmap_and_free_vma: + vma_fput(vma); vma->vm_file = NULL; - fput(file); /* Undo any partial mapping done by a device driver. */ unmap_region(mm, vma, prev, vma->vm_start, vma->vm_end); @@ -2479,7 +2479,7 @@ static int __split_vma(struct mm_struct *mm, struct vm_area_struct *vma, goto out_free_mpol; if (new->vm_file) - get_file(new->vm_file); + vma_get_file(new); if (new->vm_ops && new->vm_ops->open) new->vm_ops->open(new); @@ -2498,7 +2498,7 @@ static int __split_vma(struct mm_struct *mm, struct vm_area_struct *vma, if (new->vm_ops && new->vm_ops->close) new->vm_ops->close(new); if (new->vm_file) - fput(new->vm_file); + vma_fput(new); unlink_anon_vmas(new); out_free_mpol: mpol_put(vma_policy(new)); @@ -2640,7 +2640,6 @@ SYSCALL_DEFINE5(remap_file_pages, unsigned long, start, unsigned long, size, struct vm_area_struct *vma; unsigned long populate = 0; unsigned long ret = -EINVAL; - struct file *file; pr_warn_once("%s (%d) uses deprecated remap_file_pages() syscall. " "See Documentation/vm/remap_file_pages.txt.\n", @@ -2708,10 +2707,10 @@ SYSCALL_DEFINE5(remap_file_pages, unsigned long, start, unsigned long, size, } } - file = get_file(vma->vm_file); + vma_get_file(vma); ret = do_mmap_pgoff(vma->vm_file, start, size, prot, flags, pgoff, &populate); - fput(file); + vma_fput(vma); out: up_write(&mm->mmap_sem); if (populate) @@ -2982,7 +2981,7 @@ struct vm_area_struct *copy_vma(struct vm_area_struct **vmap, if (anon_vma_clone(new_vma, vma)) goto out_free_mempol; if (new_vma->vm_file) - get_file(new_vma->vm_file); + vma_get_file(new_vma); if (new_vma->vm_ops && new_vma->vm_ops->open) new_vma->vm_ops->open(new_vma); vma_link(mm, new_vma, prev, rb_link, rb_parent); diff --git a/mm/nommu.c b/mm/nommu.c index fbf6f0f1d..1a4a06d09 100644 --- a/mm/nommu.c +++ b/mm/nommu.c @@ -671,7 +671,7 @@ static void __put_nommu_region(struct vm_region *region) up_write(&nommu_region_sem); if (region->vm_file) - fput(region->vm_file); + vmr_fput(region); /* IO memory and memory shared directly out of the pagecache * from ramfs/tmpfs mustn't be released here */ @@ -829,7 +829,7 @@ static void delete_vma(struct mm_struct *mm, struct vm_area_struct *vma) if (vma->vm_ops && vma->vm_ops->close) vma->vm_ops->close(vma); if (vma->vm_file) - fput(vma->vm_file); + vma_fput(vma); put_nommu_region(vma->vm_region); kmem_cache_free(vm_area_cachep, vma); } @@ -1355,7 +1355,7 @@ unsigned long do_mmap(struct file *file, goto error_just_free; } } - fput(region->vm_file); + vmr_fput(region); kmem_cache_free(vm_region_jar, region); region = pregion; result = start; @@ -1430,10 +1430,10 @@ error_just_free: up_write(&nommu_region_sem); error: if (region->vm_file) - fput(region->vm_file); + vmr_fput(region); kmem_cache_free(vm_region_jar, region); if (vma->vm_file) - fput(vma->vm_file); + vma_fput(vma); kmem_cache_free(vm_area_cachep, vma); return ret; diff --git a/mm/page_alloc.c b/mm/page_alloc.c index 58bc48250..497866021 100644 --- a/mm/page_alloc.c +++ b/mm/page_alloc.c @@ -661,34 +661,28 @@ static inline void __free_one_page(struct page *page, unsigned long combined_idx; unsigned long uninitialized_var(buddy_idx); struct page *buddy; - unsigned int max_order = MAX_ORDER; + unsigned int max_order; + + max_order = min_t(unsigned int, MAX_ORDER, pageblock_order + 1); VM_BUG_ON(!zone_is_initialized(zone)); VM_BUG_ON_PAGE(page->flags & PAGE_FLAGS_CHECK_AT_PREP, page); VM_BUG_ON(migratetype == -1); - if (is_migrate_isolate(migratetype)) { - /* - * We restrict max order of merging to prevent merge - * between freepages on isolate pageblock and normal - * pageblock. Without this, pageblock isolation - * could cause incorrect freepage accounting. - */ - max_order = min_t(unsigned int, MAX_ORDER, pageblock_order + 1); - } else { + if (likely(!is_migrate_isolate(migratetype))) __mod_zone_freepage_state(zone, 1 << order, migratetype); - } - page_idx = pfn & ((1 << max_order) - 1); + page_idx = pfn & ((1 << MAX_ORDER) - 1); VM_BUG_ON_PAGE(page_idx & ((1 << order) - 1), page); VM_BUG_ON_PAGE(bad_range(zone, page), page); +continue_merging: while (order < max_order - 1) { buddy_idx = __find_buddy_index(page_idx, order); buddy = page + (buddy_idx - page_idx); if (!page_is_buddy(page, buddy, order)) - break; + goto done_merging; /* * Our buddy is free or it is CONFIG_DEBUG_PAGEALLOC guard page, * merge with it and move up one order. @@ -705,6 +699,32 @@ static inline void __free_one_page(struct page *page, page_idx = combined_idx; order++; } + if (max_order < MAX_ORDER) { + /* If we are here, it means order is >= pageblock_order. + * We want to prevent merge between freepages on isolate + * pageblock and normal pageblock. Without this, pageblock + * isolation could cause incorrect freepage or CMA accounting. + * + * We don't want to hit this code for the more frequent + * low-order merging. + */ + if (unlikely(has_isolate_pageblock(zone))) { + int buddy_mt; + + buddy_idx = __find_buddy_index(page_idx, order); + buddy = page + (buddy_idx - page_idx); + buddy_mt = get_pageblock_migratetype(buddy); + + if (migratetype != buddy_mt + && (is_migrate_isolate(migratetype) || + is_migrate_isolate(buddy_mt))) + goto done_merging; + } + max_order++; + goto continue_merging; + } + +done_merging: set_page_order(page, order); /* diff --git a/mm/prfile.c b/mm/prfile.c new file mode 100644 index 000000000..b323b8af9 --- /dev/null +++ b/mm/prfile.c @@ -0,0 +1,86 @@ +/* + * Mainly for aufs which mmap(2) diffrent file and wants to print different path + * in /proc/PID/maps. + * Call these functions via macros defined in linux/mm.h. + * + * See Documentation/filesystems/aufs/design/06mmap.txt + * + * Copyright (c) 2014 Junjro R. Okajima + * Copyright (c) 2014 Ian Campbell + */ + +#include <linux/mm.h> +#include <linux/file.h> +#include <linux/fs.h> + +/* #define PRFILE_TRACE */ +static inline void prfile_trace(struct file *f, struct file *pr, + const char func[], int line, const char func2[]) +{ +#ifdef PRFILE_TRACE + if (pr) + pr_info("%s:%d: %s, %s\n", func, line, func2, + f ? (char *)f->f_path.dentry->d_name.name : "(null)"); +#endif +} + +void vma_do_file_update_time(struct vm_area_struct *vma, const char func[], + int line) +{ + struct file *f = vma->vm_file, *pr = vma->vm_prfile; + + prfile_trace(f, pr, func, line, __func__); + file_update_time(f); + if (f && pr) + file_update_time(pr); +} + +struct file *vma_do_pr_or_file(struct vm_area_struct *vma, const char func[], + int line) +{ + struct file *f = vma->vm_file, *pr = vma->vm_prfile; + + prfile_trace(f, pr, func, line, __func__); + return (f && pr) ? pr : f; +} + +void vma_do_get_file(struct vm_area_struct *vma, const char func[], int line) +{ + struct file *f = vma->vm_file, *pr = vma->vm_prfile; + + prfile_trace(f, pr, func, line, __func__); + get_file(f); + if (f && pr) + get_file(pr); +} + +void vma_do_fput(struct vm_area_struct *vma, const char func[], int line) +{ + struct file *f = vma->vm_file, *pr = vma->vm_prfile; + + prfile_trace(f, pr, func, line, __func__); + fput(f); + if (f && pr) + fput(pr); +} + +#ifndef CONFIG_MMU +struct file *vmr_do_pr_or_file(struct vm_region *region, const char func[], + int line) +{ + struct file *f = region->vm_file, *pr = region->vm_prfile; + + prfile_trace(f, pr, func, line, __func__); + return (f && pr) ? pr : f; +} + +void vmr_do_fput(struct vm_region *region, const char func[], int line) +{ + struct file *f = region->vm_file, *pr = region->vm_prfile; + + prfile_trace(f, pr, func, line, __func__); + fput(f); + if (f && pr) + fput(pr); +} +#endif /* !CONFIG_MMU */ diff --git a/net/bluetooth/mgmt.c b/net/bluetooth/mgmt.c index 5a5089cb6..1363b8ffd 100644 --- a/net/bluetooth/mgmt.c +++ b/net/bluetooth/mgmt.c @@ -5979,6 +5979,10 @@ static int add_advertising(struct sock *sk, struct hci_dev *hdev, return mgmt_cmd_status(sk, hdev->id, MGMT_OP_ADD_ADVERTISING, MGMT_STATUS_INVALID_PARAMS); + if (data_len != sizeof(*cp) + cp->adv_data_len + cp->scan_rsp_len) + return mgmt_cmd_status(sk, hdev->id, MGMT_OP_ADD_ADVERTISING, + MGMT_STATUS_INVALID_PARAMS); + flags = __le32_to_cpu(cp->flags); timeout = __le16_to_cpu(cp->timeout); duration = __le16_to_cpu(cp->duration); diff --git a/scripts/coccinelle/iterators/use_after_iter.cocci b/scripts/coccinelle/iterators/use_after_iter.cocci index f085f5968..ce8cc9c00 100644 --- a/scripts/coccinelle/iterators/use_after_iter.cocci +++ b/scripts/coccinelle/iterators/use_after_iter.cocci @@ -123,7 +123,7 @@ list_remove_head(x,c,...) | sizeof(<+...c...+>) | -&c->member + &c->member | c = E | diff --git a/scripts/gdb/linux/modules.py b/scripts/gdb/linux/modules.py index 25db8cff4..0a35d6dbf 100644 --- a/scripts/gdb/linux/modules.py +++ b/scripts/gdb/linux/modules.py @@ -73,10 +73,11 @@ class LxLsmod(gdb.Command): " " if utils.get_long_type().sizeof == 8 else "")) for module in module_list(): + layout = module['core_layout'] gdb.write("{address} {name:<19} {size:>8} {ref}".format( - address=str(module['module_core']).split()[0], + address=str(layout['base']).split()[0], name=module['name'].string(), - size=str(module['core_size']), + size=str(layout['size']), ref=str(module['refcnt']['counter']))) source_list = module['source_list'] diff --git a/scripts/gdb/linux/symbols.py b/scripts/gdb/linux/symbols.py index 627750cb4..9a0f8923f 100644 --- a/scripts/gdb/linux/symbols.py +++ b/scripts/gdb/linux/symbols.py @@ -108,7 +108,7 @@ lx-symbols command.""" def load_module_symbols(self, module): module_name = module['name'].string() - module_addr = str(module['module_core']).split()[0] + module_addr = str(module['core_layout']['base']).split()[0] module_file = self._get_module_file(module_name) if not module_file and not self.module_files_updated: diff --git a/scripts/kconfig/Makefile b/scripts/kconfig/Makefile index d79cba4ce..ebced77de 100644 --- a/scripts/kconfig/Makefile +++ b/scripts/kconfig/Makefile @@ -96,13 +96,15 @@ savedefconfig: $(obj)/conf defconfig: $(obj)/conf ifeq ($(KBUILD_DEFCONFIG),) $< $(silent) --defconfig $(Kconfig) -else ifneq ($(wildcard $(srctree)/arch/$(SRCARCH)/configs/$(KBUILD_DEFCONFIG)),) +else +ifneq ($(wildcard $(srctree)/arch/$(SRCARCH)/configs/$(KBUILD_DEFCONFIG)),) @$(kecho) "*** Default configuration is based on '$(KBUILD_DEFCONFIG)'" $(Q)$< $(silent) --defconfig=arch/$(SRCARCH)/configs/$(KBUILD_DEFCONFIG) $(Kconfig) else @$(kecho) "*** Default configuration is based on target '$(KBUILD_DEFCONFIG)'" $(Q)$(MAKE) -f $(srctree)/Makefile $(KBUILD_DEFCONFIG) endif +endif %_defconfig: $(obj)/conf $(Q)$< $(silent) --defconfig=arch/$(SRCARCH)/configs/$@ $(Kconfig) diff --git a/scripts/package/mkspec b/scripts/package/mkspec index 71004daef..fe44d68e9 100755 --- a/scripts/package/mkspec +++ b/scripts/package/mkspec @@ -131,11 +131,11 @@ echo 'rm -rf $RPM_BUILD_ROOT' echo "" echo "%post" echo "if [ -x /sbin/installkernel -a -r /boot/vmlinuz-$KERNELRELEASE -a -r /boot/System.map-$KERNELRELEASE ]; then" -echo "cp /boot/vmlinuz-$KERNELRELEASE /boot/vmlinuz-$KERNELRELEASE-rpm" -echo "cp /boot/System.map-$KERNELRELEASE /boot/System.map-$KERNELRELEASE-rpm" +echo "cp /boot/vmlinuz-$KERNELRELEASE /boot/.vmlinuz-$KERNELRELEASE-rpm" +echo "cp /boot/System.map-$KERNELRELEASE /boot/.System.map-$KERNELRELEASE-rpm" echo "rm -f /boot/vmlinuz-$KERNELRELEASE /boot/System.map-$KERNELRELEASE" -echo "/sbin/installkernel $KERNELRELEASE /boot/vmlinuz-$KERNELRELEASE-rpm /boot/System.map-$KERNELRELEASE-rpm" -echo "rm -f /boot/vmlinuz-$KERNELRELEASE-rpm /boot/System.map-$KERNELRELEASE-rpm" +echo "/sbin/installkernel $KERNELRELEASE /boot/.vmlinuz-$KERNELRELEASE-rpm /boot/.System.map-$KERNELRELEASE-rpm" +echo "rm -f /boot/.vmlinuz-$KERNELRELEASE-rpm /boot/.System.map-$KERNELRELEASE-rpm" echo "fi" echo "" echo "%files" diff --git a/security/commoncap.c b/security/commoncap.c index 48071ed7c..50a1a40f0 100644 --- a/security/commoncap.c +++ b/security/commoncap.c @@ -1058,12 +1058,14 @@ int cap_mmap_addr(unsigned long addr) } return ret; } +EXPORT_SYMBOL_GPL(cap_mmap_addr); int cap_mmap_file(struct file *file, unsigned long reqprot, unsigned long prot, unsigned long flags) { return 0; } +EXPORT_SYMBOL_GPL(cap_mmap_file); #ifdef CONFIG_SECURITY diff --git a/security/device_cgroup.c b/security/device_cgroup.c index 03c1652c9..f88c84bf1 100644 --- a/security/device_cgroup.c +++ b/security/device_cgroup.c @@ -7,6 +7,7 @@ #include <linux/device_cgroup.h> #include <linux/cgroup.h> #include <linux/ctype.h> +#include <linux/export.h> #include <linux/list.h> #include <linux/uaccess.h> #include <linux/seq_file.h> @@ -849,6 +850,7 @@ int __devcgroup_inode_permission(struct inode *inode, int mask) return __devcgroup_check_permission(type, imajor(inode), iminor(inode), access); } +EXPORT_SYMBOL_GPL(__devcgroup_inode_permission); int devcgroup_inode_mknod(int mode, dev_t dev) { diff --git a/security/security.c b/security/security.c index e8ffd92ae..51fa026d4 100644 --- a/security/security.c +++ b/security/security.c @@ -433,6 +433,7 @@ int security_path_rmdir(struct path *dir, struct dentry *dentry) return 0; return call_int_hook(path_rmdir, 0, dir, dentry); } +EXPORT_SYMBOL_GPL(security_path_rmdir); int security_path_unlink(struct path *dir, struct dentry *dentry) { @@ -449,6 +450,7 @@ int security_path_symlink(struct path *dir, struct dentry *dentry, return 0; return call_int_hook(path_symlink, 0, dir, dentry, old_name); } +EXPORT_SYMBOL_GPL(security_path_symlink); int security_path_link(struct dentry *old_dentry, struct path *new_dir, struct dentry *new_dentry) @@ -457,6 +459,7 @@ int security_path_link(struct dentry *old_dentry, struct path *new_dir, return 0; return call_int_hook(path_link, 0, old_dentry, new_dir, new_dentry); } +EXPORT_SYMBOL_GPL(security_path_link); int security_path_rename(struct path *old_dir, struct dentry *old_dentry, struct path *new_dir, struct dentry *new_dentry, @@ -484,6 +487,7 @@ int security_path_truncate(struct path *path) return 0; return call_int_hook(path_truncate, 0, path); } +EXPORT_SYMBOL_GPL(security_path_truncate); int security_path_chmod(struct path *path, umode_t mode) { @@ -491,6 +495,7 @@ int security_path_chmod(struct path *path, umode_t mode) return 0; return call_int_hook(path_chmod, 0, path, mode); } +EXPORT_SYMBOL_GPL(security_path_chmod); int security_path_chown(struct path *path, kuid_t uid, kgid_t gid) { @@ -498,6 +503,7 @@ int security_path_chown(struct path *path, kuid_t uid, kgid_t gid) return 0; return call_int_hook(path_chown, 0, path, uid, gid); } +EXPORT_SYMBOL_GPL(security_path_chown); int security_path_chroot(struct path *path) { @@ -583,6 +589,7 @@ int security_inode_readlink(struct dentry *dentry) return 0; return call_int_hook(inode_readlink, 0, dentry); } +EXPORT_SYMBOL_GPL(security_inode_readlink); int security_inode_follow_link(struct dentry *dentry, struct inode *inode, bool rcu) @@ -598,6 +605,7 @@ int security_inode_permission(struct inode *inode, int mask) return 0; return call_int_hook(inode_permission, 0, inode, mask); } +EXPORT_SYMBOL_GPL(security_inode_permission); int security_inode_setattr(struct dentry *dentry, struct iattr *attr) { @@ -736,6 +744,7 @@ int security_file_permission(struct file *file, int mask) return fsnotify_perm(file, mask); } +EXPORT_SYMBOL_GPL(security_file_permission); int security_file_alloc(struct file *file) { @@ -795,6 +804,7 @@ int security_mmap_file(struct file *file, unsigned long prot, return ret; return ima_file_mmap(file, prot); } +EXPORT_SYMBOL_GPL(security_mmap_file); int security_mmap_addr(unsigned long addr) { diff --git a/sound/core/pcm_lib.c b/sound/core/pcm_lib.c index 6b5a811e0..3a9b66c6e 100644 --- a/sound/core/pcm_lib.c +++ b/sound/core/pcm_lib.c @@ -322,7 +322,7 @@ static int snd_pcm_update_hw_ptr0(struct snd_pcm_substream *substream, char name[16]; snd_pcm_debug_name(substream, name, sizeof(name)); pcm_err(substream->pcm, - "BUG: %s, pos = %ld, buffer size = %ld, period size = %ld\n", + "invalid position: %s, pos = %ld, buffer size = %ld, period size = %ld\n", name, pos, runtime->buffer_size, runtime->period_size); } diff --git a/sound/hda/hdac_device.c b/sound/hda/hdac_device.c index e361024ea..d1a4d6973 100644 --- a/sound/hda/hdac_device.c +++ b/sound/hda/hdac_device.c @@ -611,6 +611,22 @@ int snd_hdac_power_up_pm(struct hdac_device *codec) } EXPORT_SYMBOL_GPL(snd_hdac_power_up_pm); +/* like snd_hdac_power_up_pm(), but only increment the pm count when + * already powered up. Returns -1 if not powered up, 1 if incremented + * or 0 if unchanged. Only used in hdac_regmap.c + */ +int snd_hdac_keep_power_up(struct hdac_device *codec) +{ + if (!atomic_inc_not_zero(&codec->in_pm)) { + int ret = pm_runtime_get_if_in_use(&codec->dev); + if (!ret) + return -1; + if (ret < 0) + return 0; + } + return 1; +} + /** * snd_hdac_power_down_pm - power down the codec * @codec: the codec object diff --git a/sound/hda/hdac_regmap.c b/sound/hda/hdac_regmap.c index eb8f7c30c..bdbcd6b75 100644 --- a/sound/hda/hdac_regmap.c +++ b/sound/hda/hdac_regmap.c @@ -21,13 +21,16 @@ #include <sound/hdaudio.h> #include <sound/hda_regmap.h> -#ifdef CONFIG_PM -#define codec_is_running(codec) \ - (atomic_read(&(codec)->in_pm) || \ - !pm_runtime_suspended(&(codec)->dev)) -#else -#define codec_is_running(codec) true -#endif +static int codec_pm_lock(struct hdac_device *codec) +{ + return snd_hdac_keep_power_up(codec); +} + +static void codec_pm_unlock(struct hdac_device *codec, int lock) +{ + if (lock == 1) + snd_hdac_power_down_pm(codec); +} #define get_verb(reg) (((reg) >> 8) & 0xfff) @@ -238,20 +241,28 @@ static int hda_reg_read(void *context, unsigned int reg, unsigned int *val) struct hdac_device *codec = context; int verb = get_verb(reg); int err; + int pm_lock = 0; - if (!codec_is_running(codec) && verb != AC_VERB_GET_POWER_STATE) - return -EAGAIN; + if (verb != AC_VERB_GET_POWER_STATE) { + pm_lock = codec_pm_lock(codec); + if (pm_lock < 0) + return -EAGAIN; + } reg |= (codec->addr << 28); - if (is_stereo_amp_verb(reg)) - return hda_reg_read_stereo_amp(codec, reg, val); - if (verb == AC_VERB_GET_PROC_COEF) - return hda_reg_read_coef(codec, reg, val); + if (is_stereo_amp_verb(reg)) { + err = hda_reg_read_stereo_amp(codec, reg, val); + goto out; + } + if (verb == AC_VERB_GET_PROC_COEF) { + err = hda_reg_read_coef(codec, reg, val); + goto out; + } if ((verb & 0x700) == AC_VERB_SET_AMP_GAIN_MUTE) reg &= ~AC_AMP_FAKE_MUTE; err = snd_hdac_exec_verb(codec, reg, 0, val); if (err < 0) - return err; + goto out; /* special handling for asymmetric reads */ if (verb == AC_VERB_GET_POWER_STATE) { if (*val & AC_PWRST_ERROR) @@ -259,7 +270,9 @@ static int hda_reg_read(void *context, unsigned int reg, unsigned int *val) else /* take only the actual state */ *val = (*val >> 4) & 0x0f; } - return 0; + out: + codec_pm_unlock(codec, pm_lock); + return err; } static int hda_reg_write(void *context, unsigned int reg, unsigned int val) @@ -267,6 +280,7 @@ static int hda_reg_write(void *context, unsigned int reg, unsigned int val) struct hdac_device *codec = context; unsigned int verb; int i, bytes, err; + int pm_lock = 0; if (codec->caps_overwriting) return 0; @@ -275,14 +289,21 @@ static int hda_reg_write(void *context, unsigned int reg, unsigned int val) reg |= (codec->addr << 28); verb = get_verb(reg); - if (!codec_is_running(codec) && verb != AC_VERB_SET_POWER_STATE) - return codec->lazy_cache ? 0 : -EAGAIN; + if (verb != AC_VERB_SET_POWER_STATE) { + pm_lock = codec_pm_lock(codec); + if (pm_lock < 0) + return codec->lazy_cache ? 0 : -EAGAIN; + } - if (is_stereo_amp_verb(reg)) - return hda_reg_write_stereo_amp(codec, reg, val); + if (is_stereo_amp_verb(reg)) { + err = hda_reg_write_stereo_amp(codec, reg, val); + goto out; + } - if (verb == AC_VERB_SET_PROC_COEF) - return hda_reg_write_coef(codec, reg, val); + if (verb == AC_VERB_SET_PROC_COEF) { + err = hda_reg_write_coef(codec, reg, val); + goto out; + } switch (verb & 0xf00) { case AC_VERB_SET_AMP_GAIN_MUTE: @@ -319,10 +340,12 @@ static int hda_reg_write(void *context, unsigned int reg, unsigned int val) reg |= (verb + i) << 8 | ((val >> (8 * i)) & 0xff); err = snd_hdac_exec_verb(codec, reg, 0, NULL); if (err < 0) - return err; + goto out; } - return 0; + out: + codec_pm_unlock(codec, pm_lock); + return err; } static const struct regmap_config hda_regmap_cfg = { diff --git a/sound/pci/hda/patch_cirrus.c b/sound/pci/hda/patch_cirrus.c index c1c855a6c..a47e8ae0e 100644 --- a/sound/pci/hda/patch_cirrus.c +++ b/sound/pci/hda/patch_cirrus.c @@ -174,8 +174,12 @@ static void cs_automute(struct hda_codec *codec) snd_hda_gen_update_outputs(codec); if (spec->gpio_eapd_hp || spec->gpio_eapd_speaker) { - spec->gpio_data = spec->gen.hp_jack_present ? - spec->gpio_eapd_hp : spec->gpio_eapd_speaker; + if (spec->gen.automute_speaker) + spec->gpio_data = spec->gen.hp_jack_present ? + spec->gpio_eapd_hp : spec->gpio_eapd_speaker; + else + spec->gpio_data = + spec->gpio_eapd_hp | spec->gpio_eapd_speaker; snd_hda_codec_write(codec, 0x01, 0, AC_VERB_SET_GPIO_DATA, spec->gpio_data); } diff --git a/sound/pci/hda/patch_conexant.c b/sound/pci/hda/patch_conexant.c index 6122b8ca8..56fefbd85 100644 --- a/sound/pci/hda/patch_conexant.c +++ b/sound/pci/hda/patch_conexant.c @@ -204,8 +204,13 @@ static void cx_auto_reboot_notify(struct hda_codec *codec) { struct conexant_spec *spec = codec->spec; - if (codec->core.vendor_id != 0x14f150f2) + switch (codec->core.vendor_id) { + case 0x14f150f2: /* CX20722 */ + case 0x14f150f4: /* CX20724 */ + break; + default: return; + } /* Turn the CX20722 codec into D3 to avoid spurious noises from the internal speaker during (and after) reboot */ diff --git a/sound/pci/hda/patch_hdmi.c b/sound/pci/hda/patch_hdmi.c index bcbc4ee10..e68fa449e 100644 --- a/sound/pci/hda/patch_hdmi.c +++ b/sound/pci/hda/patch_hdmi.c @@ -152,13 +152,17 @@ struct hdmi_spec { struct hda_pcm_stream pcm_playback; /* i915/powerwell (Haswell+/Valleyview+) specific */ + bool use_acomp_notifier; /* use i915 eld_notify callback for hotplug */ struct i915_audio_component_audio_ops i915_audio_ops; bool i915_bound; /* was i915 bound in this driver? */ }; #ifdef CONFIG_SND_HDA_I915 -#define codec_has_acomp(codec) \ - ((codec)->bus->core.audio_component != NULL) +static inline bool codec_has_acomp(struct hda_codec *codec) +{ + struct hdmi_spec *spec = codec->spec; + return spec->use_acomp_notifier; +} #else #define codec_has_acomp(codec) false #endif @@ -1562,6 +1566,7 @@ static void update_eld(struct hda_codec *codec, eld->eld_size) != 0) eld_changed = true; + pin_eld->monitor_present = eld->monitor_present; pin_eld->eld_valid = eld->eld_valid; pin_eld->eld_size = eld->eld_size; if (eld->eld_valid) @@ -1665,11 +1670,10 @@ static void sync_eld_via_acomp(struct hda_codec *codec, int size; mutex_lock(&per_pin->lock); + eld->monitor_present = false; size = snd_hdac_acomp_get_eld(&codec->bus->core, per_pin->pin_nid, &eld->monitor_present, eld->eld_buffer, ELD_MAX_SIZE); - if (size < 0) - goto unlock; if (size > 0) { size = min(size, ELD_MAX_SIZE); if (snd_hdmi_parse_eld(codec, &eld->info, @@ -1873,7 +1877,8 @@ static int generic_hdmi_playback_pcm_prepare(struct hda_pcm_stream *hinfo, /* Call sync_audio_rate to set the N/CTS/M manually if necessary */ /* Todo: add DP1.2 MST audio support later */ - snd_hdac_sync_audio_rate(&codec->bus->core, pin_nid, runtime->rate); + if (codec_has_acomp(codec)) + snd_hdac_sync_audio_rate(&codec->bus->core, pin_nid, runtime->rate); non_pcm = check_non_pcm_per_cvt(codec, cvt_nid); mutex_lock(&per_pin->lock); @@ -2432,6 +2437,10 @@ static void intel_pin_eld_notify(void *audio_ptr, int port) struct hda_codec *codec = audio_ptr; int pin_nid = port + 0x04; + /* we assume only from port-B to port-D */ + if (port < 1 || port > 3) + return; + /* skip notification during system suspend (but not in runtime PM); * the state will be updated at resume */ @@ -2456,11 +2465,24 @@ static int patch_generic_hdmi(struct hda_codec *codec) codec->spec = spec; hdmi_array_init(spec, 4); - /* Try to bind with i915 for any Intel codecs (if not done yet) */ - if (!codec_has_acomp(codec) && - (codec->core.vendor_id >> 16) == 0x8086) - if (!snd_hdac_i915_init(&codec->bus->core)) - spec->i915_bound = true; +#ifdef CONFIG_SND_HDA_I915 + /* Try to bind with i915 for Intel HSW+ codecs (if not done yet) */ + if ((codec->core.vendor_id >> 16) == 0x8086 && + is_haswell_plus(codec)) { +#if 0 + /* on-demand binding leads to an unbalanced refcount when + * both i915 and hda drivers are probed concurrently; + * disabled temporarily for now + */ + if (!codec->bus->core.audio_component) + if (!snd_hdac_i915_init(&codec->bus->core)) + spec->i915_bound = true; +#endif + /* use i915 audio component notifier for hotplug */ + if (codec->bus->core.audio_component) + spec->use_acomp_notifier = true; + } +#endif if (is_haswell_plus(codec)) { intel_haswell_enable_all_pins(codec, true); @@ -3659,6 +3681,7 @@ HDA_CODEC_ENTRY(0x10de0070, "GPU 70 HDMI/DP", patch_nvhdmi), HDA_CODEC_ENTRY(0x10de0071, "GPU 71 HDMI/DP", patch_nvhdmi), HDA_CODEC_ENTRY(0x10de0072, "GPU 72 HDMI/DP", patch_nvhdmi), HDA_CODEC_ENTRY(0x10de007d, "GPU 7d HDMI/DP", patch_nvhdmi), +HDA_CODEC_ENTRY(0x10de0082, "GPU 82 HDMI/DP", patch_nvhdmi), HDA_CODEC_ENTRY(0x10de0083, "GPU 83 HDMI/DP", patch_nvhdmi), HDA_CODEC_ENTRY(0x10de8001, "MCP73 HDMI", patch_nvhdmi_2ch), HDA_CODEC_ENTRY(0x11069f80, "VX900 HDMI/DP", patch_via_hdmi), diff --git a/sound/pci/hda/patch_realtek.c b/sound/pci/hda/patch_realtek.c index 93d2156b6..4f5ca0b9c 100644 --- a/sound/pci/hda/patch_realtek.c +++ b/sound/pci/hda/patch_realtek.c @@ -5556,6 +5556,7 @@ static const struct snd_pci_quirk alc269_fixup_tbl[] = { SND_PCI_QUIRK(0x17aa, 0x2226, "ThinkPad X250", ALC292_FIXUP_TPT440_DOCK), SND_PCI_QUIRK(0x17aa, 0x2233, "Thinkpad", ALC293_FIXUP_LENOVO_SPK_NOISE), SND_PCI_QUIRK(0x17aa, 0x30bb, "ThinkCentre AIO", ALC233_FIXUP_LENOVO_LINE2_MIC_HOTKEY), + SND_PCI_QUIRK(0x17aa, 0x30e2, "ThinkCentre AIO", ALC233_FIXUP_LENOVO_LINE2_MIC_HOTKEY), SND_PCI_QUIRK(0x17aa, 0x3902, "Lenovo E50-80", ALC269_FIXUP_DMIC_THINKPAD_ACPI), SND_PCI_QUIRK(0x17aa, 0x3977, "IdeaPad S210", ALC283_FIXUP_INT_MIC), SND_PCI_QUIRK(0x17aa, 0x3978, "IdeaPad Y410P", ALC269_FIXUP_NO_SHUTUP), diff --git a/sound/pci/intel8x0.c b/sound/pci/intel8x0.c index 42bcbac80..ccdab29a8 100644 --- a/sound/pci/intel8x0.c +++ b/sound/pci/intel8x0.c @@ -2879,6 +2879,7 @@ static void intel8x0_measure_ac97_clock(struct intel8x0 *chip) static struct snd_pci_quirk intel8x0_clock_list[] = { SND_PCI_QUIRK(0x0e11, 0x008a, "AD1885", 41000), + SND_PCI_QUIRK(0x1014, 0x0581, "AD1981B", 48000), SND_PCI_QUIRK(0x1028, 0x00be, "AD1885", 44100), SND_PCI_QUIRK(0x1028, 0x0177, "AD1980", 48000), SND_PCI_QUIRK(0x1028, 0x01ad, "AD1981B", 48000), diff --git a/sound/usb/clock.c b/sound/usb/clock.c index 2ed260b10..7ccbcaf6a 100644 --- a/sound/usb/clock.c +++ b/sound/usb/clock.c @@ -285,6 +285,8 @@ static int set_sample_rate_v1(struct snd_usb_audio *chip, int iface, unsigned char data[3]; int err, crate; + if (get_iface_desc(alts)->bNumEndpoints < 1) + return -EINVAL; ep = get_endpoint(alts, 0)->bEndpointAddress; /* if endpoint doesn't have sampling rate control, bail out */ diff --git a/sound/usb/endpoint.c b/sound/usb/endpoint.c index 7b1cb365f..c07a7eda4 100644 --- a/sound/usb/endpoint.c +++ b/sound/usb/endpoint.c @@ -438,6 +438,9 @@ exit_clear: * * New endpoints will be added to chip->ep_list and must be freed by * calling snd_usb_endpoint_free(). + * + * For SND_USB_ENDPOINT_TYPE_SYNC, the caller needs to guarantee that + * bNumEndpoints > 1 beforehand. */ struct snd_usb_endpoint *snd_usb_add_endpoint(struct snd_usb_audio *chip, struct usb_host_interface *alts, diff --git a/sound/usb/mixer_quirks.c b/sound/usb/mixer_quirks.c index 279025650..f6c3bf79a 100644 --- a/sound/usb/mixer_quirks.c +++ b/sound/usb/mixer_quirks.c @@ -1519,7 +1519,11 @@ static int snd_microii_spdif_default_get(struct snd_kcontrol *kcontrol, /* use known values for that card: interface#1 altsetting#1 */ iface = usb_ifnum_to_if(chip->dev, 1); + if (!iface || iface->num_altsetting < 2) + return -EINVAL; alts = &iface->altsetting[1]; + if (get_iface_desc(alts)->bNumEndpoints < 1) + return -EINVAL; ep = get_endpoint(alts, 0)->bEndpointAddress; err = snd_usb_ctl_msg(chip->dev, diff --git a/sound/usb/pcm.c b/sound/usb/pcm.c index 9245f52d4..44d178ee9 100644 --- a/sound/usb/pcm.c +++ b/sound/usb/pcm.c @@ -159,6 +159,8 @@ static int init_pitch_v1(struct snd_usb_audio *chip, int iface, unsigned char data[1]; int err; + if (get_iface_desc(alts)->bNumEndpoints < 1) + return -EINVAL; ep = get_endpoint(alts, 0)->bEndpointAddress; data[0] = 1; diff --git a/sound/usb/quirks.c b/sound/usb/quirks.c index c458d60d5..cd7eac28e 100644 --- a/sound/usb/quirks.c +++ b/sound/usb/quirks.c @@ -150,6 +150,7 @@ static int create_fixed_stream_quirk(struct snd_usb_audio *chip, usb_audio_err(chip, "cannot memdup\n"); return -ENOMEM; } + INIT_LIST_HEAD(&fp->list); if (fp->nr_rates > MAX_NR_RATES) { kfree(fp); return -EINVAL; @@ -167,19 +168,20 @@ static int create_fixed_stream_quirk(struct snd_usb_audio *chip, stream = (fp->endpoint & USB_DIR_IN) ? SNDRV_PCM_STREAM_CAPTURE : SNDRV_PCM_STREAM_PLAYBACK; err = snd_usb_add_audio_stream(chip, stream, fp); - if (err < 0) { - kfree(fp); - kfree(rate_table); - return err; - } + if (err < 0) + goto error; if (fp->iface != get_iface_desc(&iface->altsetting[0])->bInterfaceNumber || fp->altset_idx >= iface->num_altsetting) { - kfree(fp); - kfree(rate_table); - return -EINVAL; + err = -EINVAL; + goto error; } alts = &iface->altsetting[fp->altset_idx]; altsd = get_iface_desc(alts); + if (altsd->bNumEndpoints < 1) { + err = -EINVAL; + goto error; + } + fp->protocol = altsd->bInterfaceProtocol; if (fp->datainterval == 0) @@ -190,6 +192,12 @@ static int create_fixed_stream_quirk(struct snd_usb_audio *chip, snd_usb_init_pitch(chip, fp->iface, alts, fp); snd_usb_init_sample_rate(chip, fp->iface, alts, fp, fp->rate_max); return 0; + + error: + list_del(&fp->list); /* unlink for avoiding double-free */ + kfree(fp); + kfree(rate_table); + return err; } static int create_auto_pcm_quirk(struct snd_usb_audio *chip, @@ -462,6 +470,7 @@ static int create_uaxx_quirk(struct snd_usb_audio *chip, fp->ep_attr = get_endpoint(alts, 0)->bmAttributes; fp->datainterval = 0; fp->maxpacksize = le16_to_cpu(get_endpoint(alts, 0)->wMaxPacketSize); + INIT_LIST_HEAD(&fp->list); switch (fp->maxpacksize) { case 0x120: @@ -485,6 +494,7 @@ static int create_uaxx_quirk(struct snd_usb_audio *chip, ? SNDRV_PCM_STREAM_CAPTURE : SNDRV_PCM_STREAM_PLAYBACK; err = snd_usb_add_audio_stream(chip, stream, fp); if (err < 0) { + list_del(&fp->list); /* unlink for avoiding double-free */ kfree(fp); return err; } @@ -1121,6 +1131,7 @@ bool snd_usb_get_sample_rate_quirk(struct snd_usb_audio *chip) switch (chip->usb_id) { case USB_ID(0x045E, 0x075D): /* MS Lifecam Cinema */ case USB_ID(0x045E, 0x076D): /* MS Lifecam HD-5000 */ + case USB_ID(0x045E, 0x076E): /* MS Lifecam HD-5001 */ case USB_ID(0x045E, 0x076F): /* MS Lifecam HD-6000 */ case USB_ID(0x045E, 0x0772): /* MS Lifecam Studio */ case USB_ID(0x045E, 0x0779): /* MS Lifecam HD-3000 */ diff --git a/sound/usb/stream.c b/sound/usb/stream.c index c4dc577ab..8e9548bc1 100644 --- a/sound/usb/stream.c +++ b/sound/usb/stream.c @@ -314,7 +314,9 @@ static struct snd_pcm_chmap_elem *convert_chmap(int channels, unsigned int bits, /* * add this endpoint to the chip instance. * if a stream with the same endpoint already exists, append to it. - * if not, create a new pcm stream. + * if not, create a new pcm stream. note, fp is added to the substream + * fmt_list and will be freed on the chip instance release. do not free + * fp or do remove it from the substream fmt_list to avoid double-free. */ int snd_usb_add_audio_stream(struct snd_usb_audio *chip, int stream, @@ -675,6 +677,7 @@ int snd_usb_parse_audio_interface(struct snd_usb_audio *chip, int iface_no) * (fp->maxpacksize & 0x7ff); fp->attributes = parse_uac_endpoint_attributes(chip, alts, protocol, iface_no); fp->clock = clock; + INIT_LIST_HEAD(&fp->list); /* some quirks for attributes here */ @@ -723,6 +726,7 @@ int snd_usb_parse_audio_interface(struct snd_usb_audio *chip, int iface_no) dev_dbg(&dev->dev, "%u:%d: add audio endpoint %#x\n", iface_no, altno, fp->endpoint); err = snd_usb_add_audio_stream(chip, stream, fp); if (err < 0) { + list_del(&fp->list); /* unlink for avoiding double-free */ kfree(fp->rate_table); kfree(fp->chmap); kfree(fp); diff --git a/tools/hv/Makefile b/tools/hv/Makefile index a8ab79556..a8c464402 100644 --- a/tools/hv/Makefile +++ b/tools/hv/Makefile @@ -5,6 +5,8 @@ PTHREAD_LIBS = -lpthread WARNINGS = -Wall -Wextra CFLAGS = $(WARNINGS) -g $(PTHREAD_LIBS) $(shell getconf LFS_CFLAGS) +CFLAGS += -D__EXPORTED_HEADERS__ -I../../include/uapi -I../../include + all: hv_kvp_daemon hv_vss_daemon hv_fcopy_daemon %: %.c $(CC) $(CFLAGS) -o $@ $^ diff --git a/tools/perf/util/parse-events.c b/tools/perf/util/parse-events.c index 813d9b272..48a1c5e7d 100644 --- a/tools/perf/util/parse-events.c +++ b/tools/perf/util/parse-events.c @@ -2101,11 +2101,11 @@ char *parse_events_formats_error_string(char *additional_terms) /* valid terms */ if (additional_terms) { - if (!asprintf(&str, "valid terms: %s,%s", - additional_terms, static_terms)) + if (asprintf(&str, "valid terms: %s,%s", + additional_terms, static_terms) < 0) goto fail; } else { - if (!asprintf(&str, "valid terms: %s", static_terms)) + if (asprintf(&str, "valid terms: %s", static_terms) < 0) goto fail; } return str; diff --git a/tools/perf/util/pmu.c b/tools/perf/util/pmu.c index b597bcc8f..37b4f5070 100644 --- a/tools/perf/util/pmu.c +++ b/tools/perf/util/pmu.c @@ -284,13 +284,12 @@ static int pmu_aliases_parse(char *dir, struct list_head *head) { struct dirent *evt_ent; DIR *event_dir; - int ret = 0; event_dir = opendir(dir); if (!event_dir) return -EINVAL; - while (!ret && (evt_ent = readdir(event_dir))) { + while ((evt_ent = readdir(event_dir))) { char path[PATH_MAX]; char *name = evt_ent->d_name; FILE *file; @@ -306,17 +305,19 @@ static int pmu_aliases_parse(char *dir, struct list_head *head) snprintf(path, PATH_MAX, "%s/%s", dir, name); - ret = -EINVAL; file = fopen(path, "r"); - if (!file) - break; + if (!file) { + pr_debug("Cannot open %s\n", path); + continue; + } - ret = perf_pmu__new_alias(head, dir, name, file); + if (perf_pmu__new_alias(head, dir, name, file) < 0) + pr_debug("Cannot set up %s\n", name); fclose(file); } closedir(event_dir); - return ret; + return 0; } /* diff --git a/tools/perf/util/setup.py b/tools/perf/util/setup.py index 183310376..c8680984d 100644 --- a/tools/perf/util/setup.py +++ b/tools/perf/util/setup.py @@ -22,6 +22,7 @@ cflags = getenv('CFLAGS', '').split() # switch off several checks (need to be at the end of cflags list) cflags += ['-fno-strict-aliasing', '-Wno-write-strings', '-Wno-unused-parameter' ] +src_perf = getenv('srctree') + '/tools/perf' build_lib = getenv('PYTHON_EXTBUILD_LIB') build_tmp = getenv('PYTHON_EXTBUILD_TMP') libtraceevent = getenv('LIBTRACEEVENT') @@ -30,6 +31,9 @@ libapikfs = getenv('LIBAPI') ext_sources = [f.strip() for f in file('util/python-ext-sources') if len(f.strip()) > 0 and f[0] != '#'] +# use full paths with source files +ext_sources = map(lambda x: '%s/%s' % (src_perf, x) , ext_sources) + perf = Extension('perf', sources = ext_sources, include_dirs = ['util/include'], diff --git a/virt/kvm/kvm_main.c b/virt/kvm/kvm_main.c index 5af50c3dd..a84014e2e 100644 --- a/virt/kvm/kvm_main.c +++ b/virt/kvm/kvm_main.c @@ -536,6 +536,16 @@ static struct kvm *kvm_create_vm(unsigned long type) if (!kvm) return ERR_PTR(-ENOMEM); + spin_lock_init(&kvm->mmu_lock); + atomic_inc(¤t->mm->mm_count); + kvm->mm = current->mm; + kvm_eventfd_init(kvm); + mutex_init(&kvm->lock); + mutex_init(&kvm->irq_lock); + mutex_init(&kvm->slots_lock); + atomic_set(&kvm->users_count, 1); + INIT_LIST_HEAD(&kvm->devices); + r = kvm_arch_init_vm(kvm, type); if (r) goto out_err_no_disable; @@ -568,16 +578,6 @@ static struct kvm *kvm_create_vm(unsigned long type) goto out_err; } - spin_lock_init(&kvm->mmu_lock); - kvm->mm = current->mm; - atomic_inc(&kvm->mm->mm_count); - kvm_eventfd_init(kvm); - mutex_init(&kvm->lock); - mutex_init(&kvm->irq_lock); - mutex_init(&kvm->slots_lock); - atomic_set(&kvm->users_count, 1); - INIT_LIST_HEAD(&kvm->devices); - r = kvm_init_mmu_notifier(kvm); if (r) goto out_err; @@ -602,6 +602,7 @@ out_err_no_disable: for (i = 0; i < KVM_ADDRESS_SPACE_NUM; i++) kvm_free_memslots(kvm, kvm->memslots[i]); kvm_arch_free_vm(kvm); + mmdrop(current->mm); return ERR_PTR(r); } |