Add TuxOnIce support

13 files changed, 192 insertions, 9 deletions

diff --git a/include/linux/bio.h b/include/linux/bio.h
index 5e963a6d7..9f46b625d 100644
--- a/include/linux/bio.h
+++ b/include/linux/bio.h
@@ -32,6 +32,8 @@
 /* struct bio, bio_vec and BIO_* flags are defined in blk_types.h */
 #include <linux/blk_types.h>
 
+extern int trap_non_toi_io;
+
 #define BIO_DEBUG
 
 #ifdef BIO_DEBUG
diff --git a/include/linux/blk_types.h b/include/linux/blk_types.h
index 7303b3405..74b4ec5f8 100644
--- a/include/linux/blk_types.h
+++ b/include/linux/blk_types.h
@@ -121,13 +121,14 @@ struct bio {
 #define BIO_SNAP_STABLE	7	/* bio data must be snapshotted during write */
 #define BIO_CHAIN	8	/* chained bio, ->bi_remaining in effect */
 #define BIO_REFFED	9	/* bio has elevated ->bi_cnt */
+#define BIO_TOI		10	/* bio is TuxOnIce submitted */	
 
 /*
  * Flags starting here get preserved by bio_reset() - this includes
  * BIO_POOL_IDX()
  */
-#define BIO_RESET_BITS	13
-#define BIO_OWNS_VEC	13	/* bio_free() should free bvec */
+#define BIO_RESET_BITS	14
+#define BIO_OWNS_VEC	14	/* bio_free() should free bvec */
 
 #define bio_flagged(bio, flag)	((bio)->bi_flags & (1 << (flag)))
 
diff --git a/include/linux/fs.h b/include/linux/fs.h
index 84b783f27..8826cf0ff 100644
--- a/include/linux/fs.h
+++ b/include/linux/fs.h
@@ -1258,6 +1258,8 @@ struct mm_struct;
 #define UMOUNT_NOFOLLOW	0x00000008	/* Don't follow symlink on umount */
 #define UMOUNT_UNUSED	0x80000000	/* Flag guaranteed to be unused */
 
+extern struct list_head super_blocks;
+
 /* sb->s_iflags */
 #define SB_I_CGROUPWB	0x00000001	/* cgroup-aware writeback enabled */
 
@@ -1741,6 +1743,8 @@ struct super_operations {
 #else
 #define S_DAX		0	/* Make all the DAX code disappear */
 #endif
+#define S_ATOMIC_COPY	16384	/* Pages mapped with this inode need to be
+				   atomically copied (gem) */
 
 /*
  * Note that nosuid etc flags are inode-specific: setting some file-system
@@ -2269,6 +2273,13 @@ extern struct super_block *freeze_bdev(struct block_device *);
 extern void emergency_thaw_all(void);
 extern int thaw_bdev(struct block_device *bdev, struct super_block *sb);
 extern int fsync_bdev(struct block_device *);
+extern int fsync_super(struct super_block *);
+extern int fsync_no_super(struct block_device *);
+#define FS_FREEZER_FUSE 1
+#define FS_FREEZER_NORMAL 2
+#define FS_FREEZER_ALL (FS_FREEZER_FUSE | FS_FREEZER_NORMAL)
+void freeze_filesystems(int which);
+void thaw_filesystems(int which);
 
 extern struct super_block *blockdev_superblock;
 
diff --git a/include/linux/fs_uuid.h b/include/linux/fs_uuid.h
new file mode 100644
index 000000000..3234135b5
--- /dev/null
+++ b/include/linux/fs_uuid.h
@@ -0,0 +1,19 @@
+#include <linux/device.h>
+
+struct hd_struct;
+struct block_device;
+
+struct fs_info {
+	char uuid[16];
+	dev_t dev_t;
+	char *last_mount;
+	int last_mount_size;
+};
+
+int part_matches_fs_info(struct hd_struct *part, struct fs_info *seek);
+dev_t blk_lookup_fs_info(struct fs_info *seek);
+struct fs_info *fs_info_from_block_dev(struct block_device *bdev);
+void free_fs_info(struct fs_info *fs_info);
+int bdev_matches_key(struct block_device *bdev, const char *key);
+struct block_device *next_bdev_of_type(struct block_device *last,
+	const char *key);
diff --git a/include/linux/gfp.h b/include/linux/gfp.h
index ad35f300b..d332d6165 100644
--- a/include/linux/gfp.h
+++ b/include/linux/gfp.h
@@ -35,6 +35,7 @@ struct vm_area_struct;
 #define ___GFP_NO_KSWAPD	0x400000u
 #define ___GFP_OTHER_NODE	0x800000u
 #define ___GFP_WRITE		0x1000000u
+#define ___GFP_TOI_NOTRACK      0x2000000u
 /* If the above are modified, __GFP_BITS_SHIFT may need updating */
 
 /*
@@ -94,6 +95,7 @@ struct vm_area_struct;
 #define __GFP_NO_KSWAPD	((__force gfp_t)___GFP_NO_KSWAPD)
 #define __GFP_OTHER_NODE ((__force gfp_t)___GFP_OTHER_NODE) /* On behalf of other node */
 #define __GFP_WRITE	((__force gfp_t)___GFP_WRITE)	/* Allocator intends to dirty page */
+#define __GFP_TOI_NOTRACK	((__force gfp_t)___GFP_TOI_NOTRACK)	/* Allocator wants page untracked by TOI */
 
 /*
  * This may seem redundant, but it's a way of annotating false positives vs.
@@ -101,7 +103,7 @@ struct vm_area_struct;
  */
 #define __GFP_NOTRACK_FALSE_POSITIVE (__GFP_NOTRACK)
 
-#define __GFP_BITS_SHIFT 25	/* Room for N __GFP_FOO bits */
+#define __GFP_BITS_SHIFT 26	/* Room for N __GFP_FOO bits */
 #define __GFP_BITS_MASK ((__force gfp_t)((1 << __GFP_BITS_SHIFT) - 1))
 
 /* This equals 0, but use constants in case they ever change */
diff --git a/include/linux/mm.h b/include/linux/mm.h
index bf6f117fc..ae930c67b 100644
--- a/include/linux/mm.h
+++ b/include/linux/mm.h
@@ -2137,6 +2137,7 @@ int drop_caches_sysctl_handler(struct ctl_table *, int,
 					void __user *, size_t *, loff_t *);
 #endif
 
+void drop_pagecache(void);
 void drop_slab(void);
 void drop_slab_node(int nid);
 
diff --git a/include/linux/page-flags.h b/include/linux/page-flags.h
index 41c93844f..bf858b220 100644
--- a/include/linux/page-flags.h
+++ b/include/linux/page-flags.h
@@ -109,6 +109,12 @@ enum pageflags {
 #ifdef CONFIG_TRANSPARENT_HUGEPAGE
 	PG_compound_lock,
 #endif
+#ifdef CONFIG_TOI_INCREMENTAL
+	PG_toi_untracked,	/* Don't track dirtiness of this page - assume always dirty */
+	PG_toi_ro,		/* Page was made RO by TOI */
+	PG_toi_cbw,		/* Copy the page before it is written to */
+	PG_toi_dirty,		/* Page has been modified */
+#endif
 	__NR_PAGEFLAGS,
 
 	/* Filesystems */
@@ -288,6 +294,17 @@ TESTSCFLAG(HWPoison, hwpoison)
 PAGEFLAG_FALSE(HWPoison)
 #define __PG_HWPOISON 0
 #endif
+#ifdef CONFIG_TOI_INCREMENTAL
+PAGEFLAG(TOI_RO, toi_ro)
+PAGEFLAG(TOI_Dirty, toi_dirty)
+PAGEFLAG(TOI_Untracked, toi_untracked)
+PAGEFLAG(TOI_CBW, toi_cbw)
+#else
+PAGEFLAG_FALSE(TOI_RO)
+PAGEFLAG_FALSE(TOI_Dirty)
+PAGEFLAG_FALSE(TOI_Untracked)
+PAGEFLAG_FALSE(TOI_CBW)
+#endif
 
 /*
  * On an anonymous page mapped into a user virtual memory area,
@@ -642,8 +659,12 @@ static inline void ClearPageSlabPfmemalloc(struct page *page)
  * __PG_HWPOISON is exceptional because it needs to be kept beyond page's
  * alloc-free cycle to prevent from reusing the page.
  */
-#define PAGE_FLAGS_CHECK_AT_PREP	\
-	(((1 << NR_PAGEFLAGS) - 1) & ~__PG_HWPOISON)
+#ifdef CONFIG_TOI_INCREMENTAL
+#define PAGE_FLAGS_CHECK_AT_PREP	(((1 << NR_PAGEFLAGS) - 1) & \
+        ~((1 << PG_toi_dirty) | (1 << PG_toi_ro) | ~__PG_HWPOISON))
+#else
+#define PAGE_FLAGS_CHECK_AT_PREP	(((1 << NR_PAGEFLAGS) - 1) & ~__PG_HWPOISON)
+#endif
 
 #define PAGE_FLAGS_PRIVATE				\
 	(1 << PG_private | 1 << PG_private_2)
diff --git a/include/linux/shmem_fs.h b/include/linux/shmem_fs.h
index 50777b5b1..6723fec71 100644
--- a/include/linux/shmem_fs.h
+++ b/include/linux/shmem_fs.h
@@ -48,9 +48,10 @@ static inline struct shmem_inode_info *SHMEM_I(struct inode *inode)
 extern int shmem_init(void);
 extern int shmem_fill_super(struct super_block *sb, void *data, int silent);
 extern struct file *shmem_file_setup(const char *name,
-					loff_t size, unsigned long flags);
+					loff_t size, unsigned long flags,
+					int atomic_copy);
 extern struct file *shmem_kernel_file_setup(const char *name, loff_t size,
-					    unsigned long flags);
+					    unsigned long flags, int atomic_copy);
 extern int shmem_zero_setup(struct vm_area_struct *);
 extern int shmem_lock(struct file *file, int lock, struct user_struct *user);
 extern bool shmem_mapping(struct address_space *mapping);
diff --git a/include/linux/suspend.h b/include/linux/suspend.h
index 5efe743ce..9ae75219f 100644
--- a/include/linux/suspend.h
+++ b/include/linux/suspend.h
@@ -452,6 +452,74 @@ extern bool pm_print_times_enabled;
 #define pm_print_times_enabled	(false)
 #endif
 
+enum {
+	TOI_CAN_HIBERNATE,
+	TOI_CAN_RESUME,
+	TOI_RESUME_DEVICE_OK,
+	TOI_NORESUME_SPECIFIED,
+	TOI_SANITY_CHECK_PROMPT,
+	TOI_CONTINUE_REQ,
+	TOI_RESUMED_BEFORE,
+	TOI_BOOT_TIME,
+	TOI_NOW_RESUMING,
+	TOI_IGNORE_LOGLEVEL,
+	TOI_TRYING_TO_RESUME,
+	TOI_LOADING_ALT_IMAGE,
+	TOI_STOP_RESUME,
+	TOI_IO_STOPPED,
+	TOI_NOTIFIERS_PREPARE,
+	TOI_CLUSTER_MODE,
+	TOI_BOOT_KERNEL,
+	TOI_DEVICE_HOTPLUG_LOCKED,
+};
+
+#ifdef CONFIG_TOI
+
+/* Used in init dir files */
+extern unsigned long toi_state;
+#define set_toi_state(bit) (set_bit(bit, &toi_state))
+#define clear_toi_state(bit) (clear_bit(bit, &toi_state))
+#define test_toi_state(bit) (test_bit(bit, &toi_state))
+extern int toi_running;
+
+#define test_action_state(bit) (test_bit(bit, &toi_bkd.toi_action))
+extern int try_tuxonice_hibernate(void);
+
+#else /* !CONFIG_TOI */
+
+#define toi_state		(0)
+#define set_toi_state(bit) do { } while (0)
+#define clear_toi_state(bit) do { } while (0)
+#define test_toi_state(bit) (0)
+#define toi_running (0)
+
+static inline int try_tuxonice_hibernate(void) { return 0; }
+#define test_action_state(bit) (0)
+
+#endif /* CONFIG_TOI */
+
+#ifdef CONFIG_HIBERNATION
+#ifdef CONFIG_TOI
+extern void try_tuxonice_resume(void);
+#else
+#define try_tuxonice_resume() do { } while (0)
+#endif
+
+extern int resume_attempted;
+extern int software_resume(void);
+
+static inline void check_resume_attempted(void)
+{
+	if (resume_attempted)
+		return;
+
+	software_resume();
+}
+#else
+#define check_resume_attempted() do { } while (0)
+#define resume_attempted (0)
+#endif
+
 #ifdef CONFIG_PM_AUTOSLEEP
 
 /* kernel/power/autosleep.c */
diff --git a/include/linux/swap.h b/include/linux/swap.h
index 38874729d..9cdd0be25 100644
--- a/include/linux/swap.h
+++ b/include/linux/swap.h
@@ -289,6 +289,7 @@ extern unsigned long totalram_pages;
 extern unsigned long totalreserve_pages;
 extern unsigned long dirty_balance_reserve;
 extern unsigned long nr_free_buffer_pages(void);
+extern unsigned long nr_unallocated_buffer_pages(void);
 extern unsigned long nr_free_pagecache_pages(void);
 
 /* Definition of global_page_state not available yet */
@@ -328,6 +329,8 @@ extern unsigned long mem_cgroup_shrink_node_zone(struct mem_cgroup *mem,
 						struct zone *zone,
 						unsigned long *nr_scanned);
 extern unsigned long shrink_all_memory(unsigned long nr_pages);
+extern unsigned long shrink_memory_mask(unsigned long nr_to_reclaim,
+		gfp_t mask);
 extern int vm_swappiness;
 extern int remove_mapping(struct address_space *mapping, struct page *page);
 extern unsigned long vm_total_pages;
@@ -428,13 +431,17 @@ extern void swapcache_free(swp_entry_t);
 extern int free_swap_and_cache(swp_entry_t);
 extern int swap_type_of(dev_t, sector_t, struct block_device **);
 extern unsigned int count_swap_pages(int, int);
+extern sector_t map_swap_entry(swp_entry_t entry, struct block_device **);
 extern sector_t map_swap_page(struct page *, struct block_device **);
 extern sector_t swapdev_block(int, pgoff_t);
+extern struct swap_info_struct *get_swap_info_struct(unsigned);
 extern int page_swapcount(struct page *);
 extern struct swap_info_struct *page_swap_info(struct page *);
 extern int reuse_swap_page(struct page *);
 extern int try_to_free_swap(struct page *);
 struct backing_dev_info;
+extern void get_swap_range_of_type(int type, swp_entry_t *start,
+		swp_entry_t *end, unsigned int limit);
 
 #else /* CONFIG_SWAP */
 
diff --git a/include/linux/thread_info.h b/include/linux/thread_info.h
index ff307b548..b42ba9169 100644
--- a/include/linux/thread_info.h
+++ b/include/linux/thread_info.h
@@ -56,9 +56,9 @@ extern long do_no_restart_syscall(struct restart_block *parm);
 #ifdef __KERNEL__
 
 #ifdef CONFIG_DEBUG_STACK_USAGE
-# define THREADINFO_GFP		(GFP_KERNEL | __GFP_NOTRACK | __GFP_ZERO)
+# define THREADINFO_GFP		(GFP_KERNEL | __GFP_NOTRACK | ___GFP_TOI_NOTRACK | __GFP_ZERO)
 #else
-# define THREADINFO_GFP		(GFP_KERNEL | __GFP_NOTRACK)
+# define THREADINFO_GFP		(GFP_KERNEL | __GFP_NOTRACK | ___GFP_TOI_NOTRACK)
 #endif
 
 /*
diff --git a/include/linux/tuxonice.h b/include/linux/tuxonice.h
new file mode 100644
index 000000000..67b05a750
--- /dev/null
+++ b/include/linux/tuxonice.h
@@ -0,0 +1,48 @@
+/*
+ * include/linux/tuxonice.h
+ *
+ * Copyright (C) 2015 Nigel Cunningham (nigel at tuxonice net)
+ *
+ * This file is released under the GPLv2.
+ */
+
+#ifndef INCLUDE_LINUX_TUXONICE_H
+#define INCLUDE_LINUX_TUXONICE_H
+#ifdef CONFIG_TOI_INCREMENTAL
+extern void toi_set_logbuf_untracked(void);
+extern int toi_make_writable(pgd_t *pgd, unsigned long address);
+
+static inline int toi_incremental_support(void)
+{
+    return 1;
+}
+
+/* Copy Before Write */
+struct toi_cbw {
+    unsigned long pfn;
+    void *virt;
+    struct toi_cbw *next;
+};
+
+struct toi_cbw_state {
+    bool active;            /* Is a fault handler running? */
+    bool enabled;           /* Are we doing copy before write? */
+    int size;               /* The number of pages allocated */
+    struct toi_cbw *first, *next, *last;  /* Pointers to the data structure */
+};
+
+#define CBWS_PER_PAGE (PAGE_SIZE / sizeof(struct toi_cbw))
+DECLARE_PER_CPU(struct toi_cbw_state, toi_cbw_states);
+#else
+#define toi_set_logbuf_untracked() do { } while(0)
+static inline int toi_make_writable(pgd_t *pgd, unsigned long addr)
+{
+    return 0;
+}
+
+static inline int toi_incremental_support(void)
+{
+    return 0;
+}
+#endif
+#endif
diff --git a/include/uapi/linux/netlink.h b/include/uapi/linux/netlink.h
index cf6a65ccc..20d68edd5 100644
--- a/include/uapi/linux/netlink.h
+++ b/include/uapi/linux/netlink.h
@@ -27,6 +27,8 @@
 #define NETLINK_ECRYPTFS	19
 #define NETLINK_RDMA		20
 #define NETLINK_CRYPTO		21	/* Crypto layer */
+#define NETLINK_TOI_USERUI	22	/* TuxOnIce's userui */
+#define NETLINK_TOI_USM		23	/* Userspace storage manager */
 
 #define NETLINK_INET_DIAG	NETLINK_SOCK_DIAG