summaryrefslogtreecommitdiff
AgeCommit message (Collapse)Author
2014-10-20man: document sd_machine_get_class() and sd_machine_get_ifindices()Lennart Poettering
2014-10-20sysctl.d: default to fq_codel, fight bufferbloatMichal Schmidt
Quoting from Jon Corbet's report of Stephen Hemminger's talk at Linux Plumbers Conference 2014 (https://lwn.net/Articles/616241/): [...] So Stephen encouraged everybody to run a command like: sysctl -w net.core.default_qdisc=fq_codel That will cause fq_codel to be used for all future connections [Qdiscs apply to interfaces, not connections. Pointed out by TomH in the article comments. -- mschmidt] (up to the next reboot). Unfortunately, the default queuing discipline cannot be changed, since it will certainly disturb some user's workload somewhere. Let's have the recommended default in systemd. Thanks to Dave Täht for advice and the summary at https://lists.bufferbloat.net/pipermail/cerowrt-devel/2014-October/003701.html
2014-10-20sd-bus: sync kdbus.h (ABI break)Daniel Mack
kdbus.h now has KDBUS_ATTACH_COMM split into KDBUS_ATTACH_TID_COMM and KDBUS_ATTACH_PID_COMM. The items were split already, so the change in systemd is easy.
2014-10-19shell-completion/bash: add add-wants and add-requiresLukas Nykryn
2014-10-18sd-bus: sync kdbus.h (ABI break)Daniel Mack
In kdbus.h, the 'features' field has been dropped again. Instead of negotiating features that way, we decided to make the kernel return the set of supported flags in each ioctl struct's .flags field, in both the success and error cases.
2014-10-17systemd: try harder to bind to notify socketZbigniew Jędrzejewski-Szmek
Without the socket open we are going to crash and burn. If for whatever reason we fail during deserialization we will fail when trying to open the socket. In this case it is better to unlink the old socket and maybe lose some messages, than to continue without the notification socket. Of course this situation should not happen, but we should handle it as gracefully as possible anyway. https://bugzilla.redhat.com/show_bug.cgi?id=1099299
2014-10-17systemd: log deserialization errors as warningsZbigniew Jędrzejewski-Szmek
If we failed to parse something that we wrote ourselves, things are seriously off. This is also likely to lead to problems futher on.
2014-10-17systemd: continue switch-root even if umount failsZbigniew Jędrzejewski-Szmek
Leaving the old root around seems better than aborting the switch.
2014-10-17nspawn: fix DeviceAllow listDaniel Mack
Commit 864e17068 ("nspawn: actually allow access to /dev/net/tun in the container") added "/dev/net/tun" to the list of allowed devices but forgot to tweak the array length, which caused "/dev/kdbus/*" to be missed.
2014-10-17environment: append unit_id to error messages regarding EnvironmentFileLukas Nykryn
2014-10-17missing: remove fanotifyZbigniew Jędrzejewski-Szmek
It was only used in readahead.
2014-10-17execute: downgrade namespace error to "warning"Lennart Poettering
Also, extend the printed warning a bit, explaining the situation more verbosely.
2014-10-17update TODOLennart Poettering
2014-10-17namespace: add missing 'const' to parametersLennart Poettering
2014-10-17systemctl: by default interactively ask for polkit authorization, if possibleLennart Poettering
Also, allow using --no-ask-password to turn off ineractive polkit authorization.
2014-10-17execute: don't fail child when we don't have privileges to setup namespacesMichal Sekletar
If we don't have privileges to setup the namespaces then we are most likely running inside some sort of unprivileged container, hence not being able to create namespace is not a problem because spawned service can't access host system anyway.
2014-10-17Revert "sd-bus: sync kdbus.h (ABI break)"Kay Sievers
This reverts commit 1a2409e262da65a4b0ca8ab18fcf5eabd2d404ca. Support from the kdbus interface was removed. We require memfds to be supported by all clients.
2014-10-17resolve: add missing headerLukasz Skalski
Change-Id: I64f7c6b446f6d92057c35cc3d4e29bd2bad8f75b
2014-10-16terminal: log unsupported ANSI or DEC mode changesTom Gundersen
2014-10-16terminal: split ANSI from DEC mode changesTom Gundersen
2014-10-15man: fix copy/paste error in CPUQuota= section of resource-controlDavid Strauss
2014-10-15man: fix typosJan Synacek
2014-10-15selinux: fix potential double free crash in child processMichal Sekletar
Before returning from function we should reset ret to NULL, thus cleanup function is nop. Also context_str() returns pointer to a string containing context but not a copy, hence we must make copy it explicitly.
2014-10-15man: we don't have 'Wanted' dependencyLukas Nykryn
2014-10-15bus-proxyd: add missing flag translation for RequestNameLukasz Skalski
2014-10-14shell-completion: propose templates for disable/[re]enable/[re]startZbigniew Jędrzejewski-Szmek
Templates can be [re]enabled, on their own if the have DefaultInstance set, and with an instance suffix in all cases. Propose just the template name ending in @, to underline the instance suffix may have to be appended. Likewise for start/restart. This means that sometimes superflous units that one will not really want to operate on will be proposed, but this seems better than proposing a very incomplete set of names. https://bugs.freedesktop.org/show_bug.cgi?id=66912
2014-10-14shell-completion: fix completion of inactive unitsZbigniew Jędrzejewski-Szmek
Units which not loaded were not proposed properly. OTOH, we should filter units from get-unit-files by their state if they are currently loaded. Bring zsh completions in line with bash completion, the same logic should be used in both implementations. https://bugzilla.redhat.com/show_bug.cgi?id=1024379 https://bugzilla.redhat.com/show_bug.cgi?id=790768 https://bugs.freedesktop.org/show_bug.cgi?id=84720
2014-10-14journal-remote: fix mem leak on errorZbigniew Jędrzejewski-Szmek
2014-10-14sd-bus: sync kdbus.h (ABI break)Daniel Mack
kdbus learned KDBUS_HELLO_ACCEPT_MEMFD as new connection negotiation flag. Set it by default in systemd for now.
2014-10-14bus-proxyd: improve compatibility with dbus-1Lukasz Skalski
'GetConnectionUnixProcessID', 'GetConnectionUnixUser' and 'GetConnectionSELinuxSecurityContext' methods should return 'NameHasNoOwner' error (if chosen name is not available on bus) with more detailed description - like dbus-1: Could not get PID of name 'org.freedesktop.test': no such name. Could not get UID of name 'org.freedesktop.test': no such name. Could not get security context of name 'org.freedesktop.test': no such name. Otherwise we have only laconic message without proper dbus error: Error System.Error.ENXIO: No such device or address
2014-10-14python-systemd: avoid hitting assert in __exit__Dave Reisner
Reader_close() asserts that 'args' is always NULL, but the __exit__ function forwards a non-NULL args.
2014-10-13resolved: apply label to /run/systemd/resolve/resolv.confZbigniew Jędrzejewski-Szmek
Under an SELinux system, we want the file that is created to have a proper context, different from the default for files in /run. This is so that the policy can give access to almost everyone to this file.
2014-10-13resolved: split out writing of resolv.confZbigniew Jędrzejewski-Szmek
manager_write_resolv_conf() is quite long anyway.
2014-10-13bus: check allow-interactive-auhtorization flag when doing polkitLennart Poettering
2014-10-13sd-bus: add support for new allow-interactive-authorization message header flagLennart Poettering
2014-10-13polkit: as we dropped support for the polkit bus transport, also drop its ↵Lennart Poettering
polkit policy
2014-10-13update TODOLennart Poettering
2014-10-13systemctl: when mangle unit names for the "isolate", suffix with ".target" ↵Lennart Poettering
rather than ".service" by default After all, we set AllowIsolate exclusively for target units so far, and this is more or less the only thing tht makes sense, hence also use ".target" as completion suffix by default.
2014-10-13systemctl: move unit action table closer to the functions that use themLennart Poettering
All other static tables we placed next to the functions that access them, so let's do this for the unit action table too.
2014-10-13fi-latin1 console keymap is no moreMarko Myllynen
http://git.altlinux.org/people/legion/packages/kbd.git?p=kbd.git;a=commitdiff;h=a88711232f3216debbb02d0171187442863ccedb
2014-10-13test-dhcp-client: add more assertsTom Gundersen
2014-10-12tmpfiles: compare return against correct errnoDave Reisner
name_to_handle_at returns -EOPNOTSUPP, not -ENOTSUP.
2014-10-12man: join tables in systemd-detect-virt(1)Zbigniew Jędrzejewski-Szmek
I think it is more readable and nicer if everything is in one table. Also, describe what the return value, since it seems awkward to describe the change in behaviour with --quiet before describing what the default is.
2014-10-12tests: fix some tests to pass in kojiZbigniew Jędrzejewski-Szmek
FAIL: test-engine ================= Cannot find cgroup mount point: No such file or directory Assertion 'r >= 0' failed at ../src/test/test-engine.c:46, function main(). Aborting. FAIL: test-sched-prio ===================== Cannot find cgroup mount point: No such file or directory Assertion 'r >= 0' failed at ../src/test/test-sched-prio.c:42, function main(). Aborting. FAIL: test-bus-cleanup ====================== Assertion 'sd_bus_open_system(&bus) >= 0' failed at ../src/libsystemd/sd-bus/test-bus-cleanup.c:40, function test_bus_open(). Aborting. after new: refcount 1
2014-10-12Set $NOTIFY_SOCKET for control procs if NotifyAccess=allBrandon L Black
2014-10-12ModernizationZbigniew Jędrzejewski-Szmek
2014-10-11NEWS: don't mention noinst_* debug programsDavid Herrmann
systemd-evcat and systemd-modeset are not installed systemd wide and should really just be used by developers to test the libsystemd-terminal features. No reason to mention it in NEWS.. despite the undulating rainbows..
2014-10-11terminal/grdrm: force deep modeset on enterDavid Herrmann
Usually, when our session is activated (or re-configurated) we should be able to try a page-flip to our buffer. The kernel driver should reject it if it is incompatible. As it turns out, drivers don't do this. Therefore, we now force a deep modeset if we're not sure what mode is set. This has the side-effect that we might get glitches on session-switches (depending on driver behavior). However, there's no way around this and it is what everyone does so far. Most drivers still detect if we keep the mode and so don't touch the clocks. Therefore, we just get a regular async flip.
2014-10-11sd-dhcp-client: support non-Ethernet hardware addressesDan Williams
Like Infiniband. See RFC 4390 section 2.1 for details on DHCP and Infiniband; chaddr is zeroed, hlen is set to 0, and htype is set to ARPHRD_INFINIBAND because IB hardware addresses are 20 bytes in length.
2014-10-11Report aa_change_onexec error codeMichael Scherer
Since aa_change_onexec return the error code in errno, and return -1, the current code do not give any useful information when something fail. This make apparmor easier to debug, as seen on https://bugs.debian.org/cgi-bin/bugreport.cgi?bug=760526