linux/drivers/md/dm-snap.c

2505 lines
58 KiB
C
Raw Normal View History

/*
* dm-snapshot.c
*
* Copyright (C) 2001-2002 Sistina Software (UK) Limited.
*
* This file is released under the GPL.
*/
#include <linux/blkdev.h>
#include <linux/device-mapper.h>
#include <linux/delay.h>
#include <linux/fs.h>
#include <linux/init.h>
#include <linux/kdev_t.h>
#include <linux/list.h>
#include <linux/mempool.h>
#include <linux/module.h>
#include <linux/slab.h>
#include <linux/vmalloc.h>
#include <linux/log2.h>
#include <linux/dm-kcopyd.h>
dm snapshot: suspend origin when doing exception handover In the function snapshot_resume we perform exception store handover. If there is another active snapshot target, the exception store is moved from this target to the target that is being resumed. The problem is that if there is some pending exception, it will point to an incorrect exception store after that handover, causing a crash due to dm-snap-persistent.c:get_exception()'s BUG_ON. This bug can be triggered by repeatedly changing snapshot permissions with "lvchange -p r" and "lvchange -p rw" while there are writes on the associated origin device. To fix this bug, we must suspend the origin device when doing the exception store handover to make sure that there are no pending exceptions: - introduce _origin_hash that keeps track of dm_origin structures. - introduce functions __lookup_dm_origin, __insert_dm_origin and __remove_dm_origin that manipulate the origin hash. - modify snapshot_resume so that it calls dm_internal_suspend_fast() and dm_internal_resume_fast() on the origin device. NOTE to stable@ people: When backporting to kernels 3.12-3.18, use dm_internal_suspend and dm_internal_resume instead of dm_internal_suspend_fast and dm_internal_resume_fast. When backporting to kernels older than 3.12, you need to pick functions dm_internal_suspend and dm_internal_resume from the commit fd2ed4d252701d3bbed4cd3e3d267ad469bb832a. Signed-off-by: Mikulas Patocka <mpatocka@redhat.com> Signed-off-by: Mike Snitzer <snitzer@redhat.com> Cc: stable@vger.kernel.org
2015-02-27 00:40:35 +08:00
#include "dm.h"
#include "dm-exception-store.h"
#define DM_MSG_PREFIX "snapshots"
static const char dm_snapshot_merge_target_name[] = "snapshot-merge";
#define dm_target_is_snapshot_merge(ti) \
((ti)->type->name == dm_snapshot_merge_target_name)
/*
* The size of the mempool used to track chunks in use.
*/
#define MIN_IOS 256
#define DM_TRACKED_CHUNK_HASH_SIZE 16
#define DM_TRACKED_CHUNK_HASH(x) ((unsigned long)(x) & \
(DM_TRACKED_CHUNK_HASH_SIZE - 1))
struct dm_exception_table {
uint32_t hash_mask;
unsigned hash_shift;
struct list_head *table;
};
struct dm_snapshot {
struct mutex lock;
struct dm_dev *origin;
struct dm_dev *cow;
struct dm_target *ti;
/* List of snapshots per Origin */
struct list_head list;
/*
* You can't use a snapshot if this is 0 (e.g. if full).
* A snapshot-merge target never clears this.
*/
int valid;
/*
* The snapshot overflowed because of a write to the snapshot device.
* We don't have to invalidate the snapshot in this case, but we need
* to prevent further writes.
*/
int snapshot_overflowed;
/* Origin writes don't trigger exceptions until this is set */
int active;
atomic_t pending_exceptions_count;
dm snapshot: avoid snapshot space leak on crash There is a possible leak of snapshot space in case of crash. The reason for space leaking is that chunks in the snapshot device are allocated sequentially, but they are finished (and stored in the metadata) out of order, depending on the order in which copying finished. For example, supposed that the metadata contains the following records SUPERBLOCK METADATA (blocks 0 ... 250) DATA 0 DATA 1 DATA 2 ... DATA 250 Now suppose that you allocate 10 new data blocks 251-260. Suppose that copying of these blocks finish out of order (block 260 finished first and the block 251 finished last). Now, the snapshot device looks like this: SUPERBLOCK METADATA (blocks 0 ... 250, 260, 259, 258, 257, 256) DATA 0 DATA 1 DATA 2 ... DATA 250 DATA 251 DATA 252 DATA 253 DATA 254 DATA 255 METADATA (blocks 255, 254, 253, 252, 251) DATA 256 DATA 257 DATA 258 DATA 259 DATA 260 Now, if the machine crashes after writing the first metadata block but before writing the second metadata block, the space for areas DATA 250-255 is leaked, it contains no valid data and it will never be used in the future. This patch makes dm-snapshot complete exceptions in the same order they were allocated, thus fixing this bug. Note: when backporting this patch to the stable kernel, change the version field in the following way: * if version in the stable kernel is {1, 11, 1}, change it to {1, 12, 0} * if version in the stable kernel is {1, 10, 0} or {1, 10, 1}, change it to {1, 10, 2} Userspace reads the version to determine if the bug was fixed, so the version change is needed. Signed-off-by: Mikulas Patocka <mpatocka@redhat.com> Signed-off-by: Mike Snitzer <snitzer@redhat.com> Cc: stable@vger.kernel.org
2013-11-30 07:13:37 +08:00
/* Protected by "lock" */
sector_t exception_start_sequence;
/* Protected by kcopyd single-threaded callback */
sector_t exception_complete_sequence;
/*
* A list of pending exceptions that completed out of order.
* Protected by kcopyd single-threaded callback.
*/
struct rb_root out_of_order_tree;
dm snapshot: avoid snapshot space leak on crash There is a possible leak of snapshot space in case of crash. The reason for space leaking is that chunks in the snapshot device are allocated sequentially, but they are finished (and stored in the metadata) out of order, depending on the order in which copying finished. For example, supposed that the metadata contains the following records SUPERBLOCK METADATA (blocks 0 ... 250) DATA 0 DATA 1 DATA 2 ... DATA 250 Now suppose that you allocate 10 new data blocks 251-260. Suppose that copying of these blocks finish out of order (block 260 finished first and the block 251 finished last). Now, the snapshot device looks like this: SUPERBLOCK METADATA (blocks 0 ... 250, 260, 259, 258, 257, 256) DATA 0 DATA 1 DATA 2 ... DATA 250 DATA 251 DATA 252 DATA 253 DATA 254 DATA 255 METADATA (blocks 255, 254, 253, 252, 251) DATA 256 DATA 257 DATA 258 DATA 259 DATA 260 Now, if the machine crashes after writing the first metadata block but before writing the second metadata block, the space for areas DATA 250-255 is leaked, it contains no valid data and it will never be used in the future. This patch makes dm-snapshot complete exceptions in the same order they were allocated, thus fixing this bug. Note: when backporting this patch to the stable kernel, change the version field in the following way: * if version in the stable kernel is {1, 11, 1}, change it to {1, 12, 0} * if version in the stable kernel is {1, 10, 0} or {1, 10, 1}, change it to {1, 10, 2} Userspace reads the version to determine if the bug was fixed, so the version change is needed. Signed-off-by: Mikulas Patocka <mpatocka@redhat.com> Signed-off-by: Mike Snitzer <snitzer@redhat.com> Cc: stable@vger.kernel.org
2013-11-30 07:13:37 +08:00
mempool_t pending_pool;
struct dm_exception_table pending;
struct dm_exception_table complete;
/*
* pe_lock protects all pending_exception operations and access
* as well as the snapshot_bios list.
*/
spinlock_t pe_lock;
/* Chunks with outstanding reads */
spinlock_t tracked_chunk_lock;
struct hlist_head tracked_chunk_hash[DM_TRACKED_CHUNK_HASH_SIZE];
/* The on disk metadata handler */
struct dm_exception_store *store;
struct dm_kcopyd_client *kcopyd_client;
/* Wait for events based on state_bits */
unsigned long state_bits;
/* Range of chunks currently being merged. */
chunk_t first_merging_chunk;
int num_merging_chunks;
/*
* The merge operation failed if this flag is set.
* Failure modes are handled as follows:
* - I/O error reading the header
* => don't load the target; abort.
* - Header does not have "valid" flag set
* => use the origin; forget about the snapshot.
* - I/O error when reading exceptions
* => don't load the target; abort.
* (We can't use the intermediate origin state.)
* - I/O error while merging
* => stop merging; set merge_failed; process I/O normally.
*/
int merge_failed;
/*
* Incoming bios that overlap with chunks being merged must wait
* for them to be committed.
*/
struct bio_list bios_queued_during_merge;
};
/*
* state_bits:
* RUNNING_MERGE - Merge operation is in progress.
* SHUTDOWN_MERGE - Set to signal that merge needs to be stopped;
* cleared afterwards.
*/
#define RUNNING_MERGE 0
#define SHUTDOWN_MERGE 1
DECLARE_DM_KCOPYD_THROTTLE_WITH_MODULE_PARM(snapshot_copy_throttle,
"A percentage of time allocated for copy on write");
struct dm_dev *dm_snap_origin(struct dm_snapshot *s)
{
return s->origin;
}
EXPORT_SYMBOL(dm_snap_origin);
struct dm_dev *dm_snap_cow(struct dm_snapshot *s)
{
return s->cow;
}
EXPORT_SYMBOL(dm_snap_cow);
static sector_t chunk_to_sector(struct dm_exception_store *store,
chunk_t chunk)
{
return chunk << store->chunk_shift;
}
static int bdev_equal(struct block_device *lhs, struct block_device *rhs)
{
/*
* There is only ever one instance of a particular block
* device so we can compare pointers safely.
*/
return lhs == rhs;
}
struct dm_snap_pending_exception {
struct dm_exception e;
/*
* Origin buffers waiting for this to complete are held
* in a bio list
*/
struct bio_list origin_bios;
struct bio_list snapshot_bios;
/* Pointer back to snapshot context */
struct dm_snapshot *snap;
/*
* 1 indicates the exception has already been sent to
* kcopyd.
*/
int started;
dm snapshot: avoid snapshot space leak on crash There is a possible leak of snapshot space in case of crash. The reason for space leaking is that chunks in the snapshot device are allocated sequentially, but they are finished (and stored in the metadata) out of order, depending on the order in which copying finished. For example, supposed that the metadata contains the following records SUPERBLOCK METADATA (blocks 0 ... 250) DATA 0 DATA 1 DATA 2 ... DATA 250 Now suppose that you allocate 10 new data blocks 251-260. Suppose that copying of these blocks finish out of order (block 260 finished first and the block 251 finished last). Now, the snapshot device looks like this: SUPERBLOCK METADATA (blocks 0 ... 250, 260, 259, 258, 257, 256) DATA 0 DATA 1 DATA 2 ... DATA 250 DATA 251 DATA 252 DATA 253 DATA 254 DATA 255 METADATA (blocks 255, 254, 253, 252, 251) DATA 256 DATA 257 DATA 258 DATA 259 DATA 260 Now, if the machine crashes after writing the first metadata block but before writing the second metadata block, the space for areas DATA 250-255 is leaked, it contains no valid data and it will never be used in the future. This patch makes dm-snapshot complete exceptions in the same order they were allocated, thus fixing this bug. Note: when backporting this patch to the stable kernel, change the version field in the following way: * if version in the stable kernel is {1, 11, 1}, change it to {1, 12, 0} * if version in the stable kernel is {1, 10, 0} or {1, 10, 1}, change it to {1, 10, 2} Userspace reads the version to determine if the bug was fixed, so the version change is needed. Signed-off-by: Mikulas Patocka <mpatocka@redhat.com> Signed-off-by: Mike Snitzer <snitzer@redhat.com> Cc: stable@vger.kernel.org
2013-11-30 07:13:37 +08:00
/* There was copying error. */
int copy_error;
/* A sequence number, it is used for in-order completion. */
sector_t exception_sequence;
struct rb_node out_of_order_node;
dm snapshot: avoid snapshot space leak on crash There is a possible leak of snapshot space in case of crash. The reason for space leaking is that chunks in the snapshot device are allocated sequentially, but they are finished (and stored in the metadata) out of order, depending on the order in which copying finished. For example, supposed that the metadata contains the following records SUPERBLOCK METADATA (blocks 0 ... 250) DATA 0 DATA 1 DATA 2 ... DATA 250 Now suppose that you allocate 10 new data blocks 251-260. Suppose that copying of these blocks finish out of order (block 260 finished first and the block 251 finished last). Now, the snapshot device looks like this: SUPERBLOCK METADATA (blocks 0 ... 250, 260, 259, 258, 257, 256) DATA 0 DATA 1 DATA 2 ... DATA 250 DATA 251 DATA 252 DATA 253 DATA 254 DATA 255 METADATA (blocks 255, 254, 253, 252, 251) DATA 256 DATA 257 DATA 258 DATA 259 DATA 260 Now, if the machine crashes after writing the first metadata block but before writing the second metadata block, the space for areas DATA 250-255 is leaked, it contains no valid data and it will never be used in the future. This patch makes dm-snapshot complete exceptions in the same order they were allocated, thus fixing this bug. Note: when backporting this patch to the stable kernel, change the version field in the following way: * if version in the stable kernel is {1, 11, 1}, change it to {1, 12, 0} * if version in the stable kernel is {1, 10, 0} or {1, 10, 1}, change it to {1, 10, 2} Userspace reads the version to determine if the bug was fixed, so the version change is needed. Signed-off-by: Mikulas Patocka <mpatocka@redhat.com> Signed-off-by: Mike Snitzer <snitzer@redhat.com> Cc: stable@vger.kernel.org
2013-11-30 07:13:37 +08:00
/*
* For writing a complete chunk, bypassing the copy.
*/
struct bio *full_bio;
bio_end_io_t *full_bio_end_io;
};
/*
* Hash table mapping origin volumes to lists of snapshots and
* a lock to protect it
*/
static struct kmem_cache *exception_cache;
static struct kmem_cache *pending_cache;
struct dm_snap_tracked_chunk {
struct hlist_node node;
chunk_t chunk;
};
static void init_tracked_chunk(struct bio *bio)
{
struct dm_snap_tracked_chunk *c = dm_per_bio_data(bio, sizeof(struct dm_snap_tracked_chunk));
INIT_HLIST_NODE(&c->node);
}
static bool is_bio_tracked(struct bio *bio)
{
struct dm_snap_tracked_chunk *c = dm_per_bio_data(bio, sizeof(struct dm_snap_tracked_chunk));
return !hlist_unhashed(&c->node);
}
static void track_chunk(struct dm_snapshot *s, struct bio *bio, chunk_t chunk)
{
struct dm_snap_tracked_chunk *c = dm_per_bio_data(bio, sizeof(struct dm_snap_tracked_chunk));
c->chunk = chunk;
spin_lock_irq(&s->tracked_chunk_lock);
hlist_add_head(&c->node,
&s->tracked_chunk_hash[DM_TRACKED_CHUNK_HASH(chunk)]);
spin_unlock_irq(&s->tracked_chunk_lock);
}
static void stop_tracking_chunk(struct dm_snapshot *s, struct bio *bio)
{
struct dm_snap_tracked_chunk *c = dm_per_bio_data(bio, sizeof(struct dm_snap_tracked_chunk));
unsigned long flags;
spin_lock_irqsave(&s->tracked_chunk_lock, flags);
hlist_del(&c->node);
spin_unlock_irqrestore(&s->tracked_chunk_lock, flags);
}
static int __chunk_is_tracked(struct dm_snapshot *s, chunk_t chunk)
{
struct dm_snap_tracked_chunk *c;
int found = 0;
spin_lock_irq(&s->tracked_chunk_lock);
hlist: drop the node parameter from iterators I'm not sure why, but the hlist for each entry iterators were conceived list_for_each_entry(pos, head, member) The hlist ones were greedy and wanted an extra parameter: hlist_for_each_entry(tpos, pos, head, member) Why did they need an extra pos parameter? I'm not quite sure. Not only they don't really need it, it also prevents the iterator from looking exactly like the list iterator, which is unfortunate. Besides the semantic patch, there was some manual work required: - Fix up the actual hlist iterators in linux/list.h - Fix up the declaration of other iterators based on the hlist ones. - A very small amount of places were using the 'node' parameter, this was modified to use 'obj->member' instead. - Coccinelle didn't handle the hlist_for_each_entry_safe iterator properly, so those had to be fixed up manually. The semantic patch which is mostly the work of Peter Senna Tschudin is here: @@ iterator name hlist_for_each_entry, hlist_for_each_entry_continue, hlist_for_each_entry_from, hlist_for_each_entry_rcu, hlist_for_each_entry_rcu_bh, hlist_for_each_entry_continue_rcu_bh, for_each_busy_worker, ax25_uid_for_each, ax25_for_each, inet_bind_bucket_for_each, sctp_for_each_hentry, sk_for_each, sk_for_each_rcu, sk_for_each_from, sk_for_each_safe, sk_for_each_bound, hlist_for_each_entry_safe, hlist_for_each_entry_continue_rcu, nr_neigh_for_each, nr_neigh_for_each_safe, nr_node_for_each, nr_node_for_each_safe, for_each_gfn_indirect_valid_sp, for_each_gfn_sp, for_each_host; type T; expression a,c,d,e; identifier b; statement S; @@ -T b; <+... when != b ( hlist_for_each_entry(a, - b, c, d) S | hlist_for_each_entry_continue(a, - b, c) S | hlist_for_each_entry_from(a, - b, c) S | hlist_for_each_entry_rcu(a, - b, c, d) S | hlist_for_each_entry_rcu_bh(a, - b, c, d) S | hlist_for_each_entry_continue_rcu_bh(a, - b, c) S | for_each_busy_worker(a, c, - b, d) S | ax25_uid_for_each(a, - b, c) S | ax25_for_each(a, - b, c) S | inet_bind_bucket_for_each(a, - b, c) S | sctp_for_each_hentry(a, - b, c) S | sk_for_each(a, - b, c) S | sk_for_each_rcu(a, - b, c) S | sk_for_each_from -(a, b) +(a) S + sk_for_each_from(a) S | sk_for_each_safe(a, - b, c, d) S | sk_for_each_bound(a, - b, c) S | hlist_for_each_entry_safe(a, - b, c, d, e) S | hlist_for_each_entry_continue_rcu(a, - b, c) S | nr_neigh_for_each(a, - b, c) S | nr_neigh_for_each_safe(a, - b, c, d) S | nr_node_for_each(a, - b, c) S | nr_node_for_each_safe(a, - b, c, d) S | - for_each_gfn_sp(a, c, d, b) S + for_each_gfn_sp(a, c, d) S | - for_each_gfn_indirect_valid_sp(a, c, d, b) S + for_each_gfn_indirect_valid_sp(a, c, d) S | for_each_host(a, - b, c) S | for_each_host_safe(a, - b, c, d) S | for_each_mesh_entry(a, - b, c, d) S ) ...+> [akpm@linux-foundation.org: drop bogus change from net/ipv4/raw.c] [akpm@linux-foundation.org: drop bogus hunk from net/ipv6/raw.c] [akpm@linux-foundation.org: checkpatch fixes] [akpm@linux-foundation.org: fix warnings] [akpm@linux-foudnation.org: redo intrusive kvm changes] Tested-by: Peter Senna Tschudin <peter.senna@gmail.com> Acked-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com> Signed-off-by: Sasha Levin <sasha.levin@oracle.com> Cc: Wu Fengguang <fengguang.wu@intel.com> Cc: Marcelo Tosatti <mtosatti@redhat.com> Cc: Gleb Natapov <gleb@redhat.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2013-02-28 09:06:00 +08:00
hlist_for_each_entry(c,
&s->tracked_chunk_hash[DM_TRACKED_CHUNK_HASH(chunk)], node) {
if (c->chunk == chunk) {
found = 1;
break;
}
}
spin_unlock_irq(&s->tracked_chunk_lock);
return found;
}
/*
* This conflicting I/O is extremely improbable in the caller,
* so msleep(1) is sufficient and there is no need for a wait queue.
*/
static void __check_for_conflicting_io(struct dm_snapshot *s, chunk_t chunk)
{
while (__chunk_is_tracked(s, chunk))
msleep(1);
}
/*
* One of these per registered origin, held in the snapshot_origins hash
*/
struct origin {
/* The origin device */
struct block_device *bdev;
struct list_head hash_list;
/* List of snapshots for this origin */
struct list_head snapshots;
};
dm snapshot: suspend origin when doing exception handover In the function snapshot_resume we perform exception store handover. If there is another active snapshot target, the exception store is moved from this target to the target that is being resumed. The problem is that if there is some pending exception, it will point to an incorrect exception store after that handover, causing a crash due to dm-snap-persistent.c:get_exception()'s BUG_ON. This bug can be triggered by repeatedly changing snapshot permissions with "lvchange -p r" and "lvchange -p rw" while there are writes on the associated origin device. To fix this bug, we must suspend the origin device when doing the exception store handover to make sure that there are no pending exceptions: - introduce _origin_hash that keeps track of dm_origin structures. - introduce functions __lookup_dm_origin, __insert_dm_origin and __remove_dm_origin that manipulate the origin hash. - modify snapshot_resume so that it calls dm_internal_suspend_fast() and dm_internal_resume_fast() on the origin device. NOTE to stable@ people: When backporting to kernels 3.12-3.18, use dm_internal_suspend and dm_internal_resume instead of dm_internal_suspend_fast and dm_internal_resume_fast. When backporting to kernels older than 3.12, you need to pick functions dm_internal_suspend and dm_internal_resume from the commit fd2ed4d252701d3bbed4cd3e3d267ad469bb832a. Signed-off-by: Mikulas Patocka <mpatocka@redhat.com> Signed-off-by: Mike Snitzer <snitzer@redhat.com> Cc: stable@vger.kernel.org
2015-02-27 00:40:35 +08:00
/*
* This structure is allocated for each origin target
*/
struct dm_origin {
struct dm_dev *dev;
struct dm_target *ti;
unsigned split_boundary;
struct list_head hash_list;
};
/*
* Size of the hash table for origin volumes. If we make this
* the size of the minors list then it should be nearly perfect
*/
#define ORIGIN_HASH_SIZE 256
#define ORIGIN_MASK 0xFF
static struct list_head *_origins;
dm snapshot: suspend origin when doing exception handover In the function snapshot_resume we perform exception store handover. If there is another active snapshot target, the exception store is moved from this target to the target that is being resumed. The problem is that if there is some pending exception, it will point to an incorrect exception store after that handover, causing a crash due to dm-snap-persistent.c:get_exception()'s BUG_ON. This bug can be triggered by repeatedly changing snapshot permissions with "lvchange -p r" and "lvchange -p rw" while there are writes on the associated origin device. To fix this bug, we must suspend the origin device when doing the exception store handover to make sure that there are no pending exceptions: - introduce _origin_hash that keeps track of dm_origin structures. - introduce functions __lookup_dm_origin, __insert_dm_origin and __remove_dm_origin that manipulate the origin hash. - modify snapshot_resume so that it calls dm_internal_suspend_fast() and dm_internal_resume_fast() on the origin device. NOTE to stable@ people: When backporting to kernels 3.12-3.18, use dm_internal_suspend and dm_internal_resume instead of dm_internal_suspend_fast and dm_internal_resume_fast. When backporting to kernels older than 3.12, you need to pick functions dm_internal_suspend and dm_internal_resume from the commit fd2ed4d252701d3bbed4cd3e3d267ad469bb832a. Signed-off-by: Mikulas Patocka <mpatocka@redhat.com> Signed-off-by: Mike Snitzer <snitzer@redhat.com> Cc: stable@vger.kernel.org
2015-02-27 00:40:35 +08:00
static struct list_head *_dm_origins;
static struct rw_semaphore _origins_lock;
static DECLARE_WAIT_QUEUE_HEAD(_pending_exceptions_done);
static DEFINE_SPINLOCK(_pending_exceptions_done_spinlock);
static uint64_t _pending_exceptions_done_count;
static int init_origin_hash(void)
{
int i;
treewide: kmalloc() -> kmalloc_array() The kmalloc() function has a 2-factor argument form, kmalloc_array(). This patch replaces cases of: kmalloc(a * b, gfp) with: kmalloc_array(a * b, gfp) as well as handling cases of: kmalloc(a * b * c, gfp) with: kmalloc(array3_size(a, b, c), gfp) as it's slightly less ugly than: kmalloc_array(array_size(a, b), c, gfp) This does, however, attempt to ignore constant size factors like: kmalloc(4 * 1024, gfp) though any constants defined via macros get caught up in the conversion. Any factors with a sizeof() of "unsigned char", "char", and "u8" were dropped, since they're redundant. The tools/ directory was manually excluded, since it has its own implementation of kmalloc(). The Coccinelle script used for this was: // Fix redundant parens around sizeof(). @@ type TYPE; expression THING, E; @@ ( kmalloc( - (sizeof(TYPE)) * E + sizeof(TYPE) * E , ...) | kmalloc( - (sizeof(THING)) * E + sizeof(THING) * E , ...) ) // Drop single-byte sizes and redundant parens. @@ expression COUNT; typedef u8; typedef __u8; @@ ( kmalloc( - sizeof(u8) * (COUNT) + COUNT , ...) | kmalloc( - sizeof(__u8) * (COUNT) + COUNT , ...) | kmalloc( - sizeof(char) * (COUNT) + COUNT , ...) | kmalloc( - sizeof(unsigned char) * (COUNT) + COUNT , ...) | kmalloc( - sizeof(u8) * COUNT + COUNT , ...) | kmalloc( - sizeof(__u8) * COUNT + COUNT , ...) | kmalloc( - sizeof(char) * COUNT + COUNT , ...) | kmalloc( - sizeof(unsigned char) * COUNT + COUNT , ...) ) // 2-factor product with sizeof(type/expression) and identifier or constant. @@ type TYPE; expression THING; identifier COUNT_ID; constant COUNT_CONST; @@ ( - kmalloc + kmalloc_array ( - sizeof(TYPE) * (COUNT_ID) + COUNT_ID, sizeof(TYPE) , ...) | - kmalloc + kmalloc_array ( - sizeof(TYPE) * COUNT_ID + COUNT_ID, sizeof(TYPE) , ...) | - kmalloc + kmalloc_array ( - sizeof(TYPE) * (COUNT_CONST) + COUNT_CONST, sizeof(TYPE) , ...) | - kmalloc + kmalloc_array ( - sizeof(TYPE) * COUNT_CONST + COUNT_CONST, sizeof(TYPE) , ...) | - kmalloc + kmalloc_array ( - sizeof(THING) * (COUNT_ID) + COUNT_ID, sizeof(THING) , ...) | - kmalloc + kmalloc_array ( - sizeof(THING) * COUNT_ID + COUNT_ID, sizeof(THING) , ...) | - kmalloc + kmalloc_array ( - sizeof(THING) * (COUNT_CONST) + COUNT_CONST, sizeof(THING) , ...) | - kmalloc + kmalloc_array ( - sizeof(THING) * COUNT_CONST + COUNT_CONST, sizeof(THING) , ...) ) // 2-factor product, only identifiers. @@ identifier SIZE, COUNT; @@ - kmalloc + kmalloc_array ( - SIZE * COUNT + COUNT, SIZE , ...) // 3-factor product with 1 sizeof(type) or sizeof(expression), with // redundant parens removed. @@ expression THING; identifier STRIDE, COUNT; type TYPE; @@ ( kmalloc( - sizeof(TYPE) * (COUNT) * (STRIDE) + array3_size(COUNT, STRIDE, sizeof(TYPE)) , ...) | kmalloc( - sizeof(TYPE) * (COUNT) * STRIDE + array3_size(COUNT, STRIDE, sizeof(TYPE)) , ...) | kmalloc( - sizeof(TYPE) * COUNT * (STRIDE) + array3_size(COUNT, STRIDE, sizeof(TYPE)) , ...) | kmalloc( - sizeof(TYPE) * COUNT * STRIDE + array3_size(COUNT, STRIDE, sizeof(TYPE)) , ...) | kmalloc( - sizeof(THING) * (COUNT) * (STRIDE) + array3_size(COUNT, STRIDE, sizeof(THING)) , ...) | kmalloc( - sizeof(THING) * (COUNT) * STRIDE + array3_size(COUNT, STRIDE, sizeof(THING)) , ...) | kmalloc( - sizeof(THING) * COUNT * (STRIDE) + array3_size(COUNT, STRIDE, sizeof(THING)) , ...) | kmalloc( - sizeof(THING) * COUNT * STRIDE + array3_size(COUNT, STRIDE, sizeof(THING)) , ...) ) // 3-factor product with 2 sizeof(variable), with redundant parens removed. @@ expression THING1, THING2; identifier COUNT; type TYPE1, TYPE2; @@ ( kmalloc( - sizeof(TYPE1) * sizeof(TYPE2) * COUNT + array3_size(COUNT, sizeof(TYPE1), sizeof(TYPE2)) , ...) | kmalloc( - sizeof(TYPE1) * sizeof(THING2) * (COUNT) + array3_size(COUNT, sizeof(TYPE1), sizeof(TYPE2)) , ...) | kmalloc( - sizeof(THING1) * sizeof(THING2) * COUNT + array3_size(COUNT, sizeof(THING1), sizeof(THING2)) , ...) | kmalloc( - sizeof(THING1) * sizeof(THING2) * (COUNT) + array3_size(COUNT, sizeof(THING1), sizeof(THING2)) , ...) | kmalloc( - sizeof(TYPE1) * sizeof(THING2) * COUNT + array3_size(COUNT, sizeof(TYPE1), sizeof(THING2)) , ...) | kmalloc( - sizeof(TYPE1) * sizeof(THING2) * (COUNT) + array3_size(COUNT, sizeof(TYPE1), sizeof(THING2)) , ...) ) // 3-factor product, only identifiers, with redundant parens removed. @@ identifier STRIDE, SIZE, COUNT; @@ ( kmalloc( - (COUNT) * STRIDE * SIZE + array3_size(COUNT, STRIDE, SIZE) , ...) | kmalloc( - COUNT * (STRIDE) * SIZE + array3_size(COUNT, STRIDE, SIZE) , ...) | kmalloc( - COUNT * STRIDE * (SIZE) + array3_size(COUNT, STRIDE, SIZE) , ...) | kmalloc( - (COUNT) * (STRIDE) * SIZE + array3_size(COUNT, STRIDE, SIZE) , ...) | kmalloc( - COUNT * (STRIDE) * (SIZE) + array3_size(COUNT, STRIDE, SIZE) , ...) | kmalloc( - (COUNT) * STRIDE * (SIZE) + array3_size(COUNT, STRIDE, SIZE) , ...) | kmalloc( - (COUNT) * (STRIDE) * (SIZE) + array3_size(COUNT, STRIDE, SIZE) , ...) | kmalloc( - COUNT * STRIDE * SIZE + array3_size(COUNT, STRIDE, SIZE) , ...) ) // Any remaining multi-factor products, first at least 3-factor products, // when they're not all constants... @@ expression E1, E2, E3; constant C1, C2, C3; @@ ( kmalloc(C1 * C2 * C3, ...) | kmalloc( - (E1) * E2 * E3 + array3_size(E1, E2, E3) , ...) | kmalloc( - (E1) * (E2) * E3 + array3_size(E1, E2, E3) , ...) | kmalloc( - (E1) * (E2) * (E3) + array3_size(E1, E2, E3) , ...) | kmalloc( - E1 * E2 * E3 + array3_size(E1, E2, E3) , ...) ) // And then all remaining 2 factors products when they're not all constants, // keeping sizeof() as the second factor argument. @@ expression THING, E1, E2; type TYPE; constant C1, C2, C3; @@ ( kmalloc(sizeof(THING) * C2, ...) | kmalloc(sizeof(TYPE) * C2, ...) | kmalloc(C1 * C2 * C3, ...) | kmalloc(C1 * C2, ...) | - kmalloc + kmalloc_array ( - sizeof(TYPE) * (E2) + E2, sizeof(TYPE) , ...) | - kmalloc + kmalloc_array ( - sizeof(TYPE) * E2 + E2, sizeof(TYPE) , ...) | - kmalloc + kmalloc_array ( - sizeof(THING) * (E2) + E2, sizeof(THING) , ...) | - kmalloc + kmalloc_array ( - sizeof(THING) * E2 + E2, sizeof(THING) , ...) | - kmalloc + kmalloc_array ( - (E1) * E2 + E1, E2 , ...) | - kmalloc + kmalloc_array ( - (E1) * (E2) + E1, E2 , ...) | - kmalloc + kmalloc_array ( - E1 * E2 + E1, E2 , ...) ) Signed-off-by: Kees Cook <keescook@chromium.org>
2018-06-13 04:55:00 +08:00
_origins = kmalloc_array(ORIGIN_HASH_SIZE, sizeof(struct list_head),
GFP_KERNEL);
if (!_origins) {
dm snapshot: suspend origin when doing exception handover In the function snapshot_resume we perform exception store handover. If there is another active snapshot target, the exception store is moved from this target to the target that is being resumed. The problem is that if there is some pending exception, it will point to an incorrect exception store after that handover, causing a crash due to dm-snap-persistent.c:get_exception()'s BUG_ON. This bug can be triggered by repeatedly changing snapshot permissions with "lvchange -p r" and "lvchange -p rw" while there are writes on the associated origin device. To fix this bug, we must suspend the origin device when doing the exception store handover to make sure that there are no pending exceptions: - introduce _origin_hash that keeps track of dm_origin structures. - introduce functions __lookup_dm_origin, __insert_dm_origin and __remove_dm_origin that manipulate the origin hash. - modify snapshot_resume so that it calls dm_internal_suspend_fast() and dm_internal_resume_fast() on the origin device. NOTE to stable@ people: When backporting to kernels 3.12-3.18, use dm_internal_suspend and dm_internal_resume instead of dm_internal_suspend_fast and dm_internal_resume_fast. When backporting to kernels older than 3.12, you need to pick functions dm_internal_suspend and dm_internal_resume from the commit fd2ed4d252701d3bbed4cd3e3d267ad469bb832a. Signed-off-by: Mikulas Patocka <mpatocka@redhat.com> Signed-off-by: Mike Snitzer <snitzer@redhat.com> Cc: stable@vger.kernel.org
2015-02-27 00:40:35 +08:00
DMERR("unable to allocate memory for _origins");
return -ENOMEM;
}
for (i = 0; i < ORIGIN_HASH_SIZE; i++)
INIT_LIST_HEAD(_origins + i);
dm snapshot: suspend origin when doing exception handover In the function snapshot_resume we perform exception store handover. If there is another active snapshot target, the exception store is moved from this target to the target that is being resumed. The problem is that if there is some pending exception, it will point to an incorrect exception store after that handover, causing a crash due to dm-snap-persistent.c:get_exception()'s BUG_ON. This bug can be triggered by repeatedly changing snapshot permissions with "lvchange -p r" and "lvchange -p rw" while there are writes on the associated origin device. To fix this bug, we must suspend the origin device when doing the exception store handover to make sure that there are no pending exceptions: - introduce _origin_hash that keeps track of dm_origin structures. - introduce functions __lookup_dm_origin, __insert_dm_origin and __remove_dm_origin that manipulate the origin hash. - modify snapshot_resume so that it calls dm_internal_suspend_fast() and dm_internal_resume_fast() on the origin device. NOTE to stable@ people: When backporting to kernels 3.12-3.18, use dm_internal_suspend and dm_internal_resume instead of dm_internal_suspend_fast and dm_internal_resume_fast. When backporting to kernels older than 3.12, you need to pick functions dm_internal_suspend and dm_internal_resume from the commit fd2ed4d252701d3bbed4cd3e3d267ad469bb832a. Signed-off-by: Mikulas Patocka <mpatocka@redhat.com> Signed-off-by: Mike Snitzer <snitzer@redhat.com> Cc: stable@vger.kernel.org
2015-02-27 00:40:35 +08:00
treewide: kmalloc() -> kmalloc_array() The kmalloc() function has a 2-factor argument form, kmalloc_array(). This patch replaces cases of: kmalloc(a * b, gfp) with: kmalloc_array(a * b, gfp) as well as handling cases of: kmalloc(a * b * c, gfp) with: kmalloc(array3_size(a, b, c), gfp) as it's slightly less ugly than: kmalloc_array(array_size(a, b), c, gfp) This does, however, attempt to ignore constant size factors like: kmalloc(4 * 1024, gfp) though any constants defined via macros get caught up in the conversion. Any factors with a sizeof() of "unsigned char", "char", and "u8" were dropped, since they're redundant. The tools/ directory was manually excluded, since it has its own implementation of kmalloc(). The Coccinelle script used for this was: // Fix redundant parens around sizeof(). @@ type TYPE; expression THING, E; @@ ( kmalloc( - (sizeof(TYPE)) * E + sizeof(TYPE) * E , ...) | kmalloc( - (sizeof(THING)) * E + sizeof(THING) * E , ...) ) // Drop single-byte sizes and redundant parens. @@ expression COUNT; typedef u8; typedef __u8; @@ ( kmalloc( - sizeof(u8) * (COUNT) + COUNT , ...) | kmalloc( - sizeof(__u8) * (COUNT) + COUNT , ...) | kmalloc( - sizeof(char) * (COUNT) + COUNT , ...) | kmalloc( - sizeof(unsigned char) * (COUNT) + COUNT , ...) | kmalloc( - sizeof(u8) * COUNT + COUNT , ...) | kmalloc( - sizeof(__u8) * COUNT + COUNT , ...) | kmalloc( - sizeof(char) * COUNT + COUNT , ...) | kmalloc( - sizeof(unsigned char) * COUNT + COUNT , ...) ) // 2-factor product with sizeof(type/expression) and identifier or constant. @@ type TYPE; expression THING; identifier COUNT_ID; constant COUNT_CONST; @@ ( - kmalloc + kmalloc_array ( - sizeof(TYPE) * (COUNT_ID) + COUNT_ID, sizeof(TYPE) , ...) | - kmalloc + kmalloc_array ( - sizeof(TYPE) * COUNT_ID + COUNT_ID, sizeof(TYPE) , ...) | - kmalloc + kmalloc_array ( - sizeof(TYPE) * (COUNT_CONST) + COUNT_CONST, sizeof(TYPE) , ...) | - kmalloc + kmalloc_array ( - sizeof(TYPE) * COUNT_CONST + COUNT_CONST, sizeof(TYPE) , ...) | - kmalloc + kmalloc_array ( - sizeof(THING) * (COUNT_ID) + COUNT_ID, sizeof(THING) , ...) | - kmalloc + kmalloc_array ( - sizeof(THING) * COUNT_ID + COUNT_ID, sizeof(THING) , ...) | - kmalloc + kmalloc_array ( - sizeof(THING) * (COUNT_CONST) + COUNT_CONST, sizeof(THING) , ...) | - kmalloc + kmalloc_array ( - sizeof(THING) * COUNT_CONST + COUNT_CONST, sizeof(THING) , ...) ) // 2-factor product, only identifiers. @@ identifier SIZE, COUNT; @@ - kmalloc + kmalloc_array ( - SIZE * COUNT + COUNT, SIZE , ...) // 3-factor product with 1 sizeof(type) or sizeof(expression), with // redundant parens removed. @@ expression THING; identifier STRIDE, COUNT; type TYPE; @@ ( kmalloc( - sizeof(TYPE) * (COUNT) * (STRIDE) + array3_size(COUNT, STRIDE, sizeof(TYPE)) , ...) | kmalloc( - sizeof(TYPE) * (COUNT) * STRIDE + array3_size(COUNT, STRIDE, sizeof(TYPE)) , ...) | kmalloc( - sizeof(TYPE) * COUNT * (STRIDE) + array3_size(COUNT, STRIDE, sizeof(TYPE)) , ...) | kmalloc( - sizeof(TYPE) * COUNT * STRIDE + array3_size(COUNT, STRIDE, sizeof(TYPE)) , ...) | kmalloc( - sizeof(THING) * (COUNT) * (STRIDE) + array3_size(COUNT, STRIDE, sizeof(THING)) , ...) | kmalloc( - sizeof(THING) * (COUNT) * STRIDE + array3_size(COUNT, STRIDE, sizeof(THING)) , ...) | kmalloc( - sizeof(THING) * COUNT * (STRIDE) + array3_size(COUNT, STRIDE, sizeof(THING)) , ...) | kmalloc( - sizeof(THING) * COUNT * STRIDE + array3_size(COUNT, STRIDE, sizeof(THING)) , ...) ) // 3-factor product with 2 sizeof(variable), with redundant parens removed. @@ expression THING1, THING2; identifier COUNT; type TYPE1, TYPE2; @@ ( kmalloc( - sizeof(TYPE1) * sizeof(TYPE2) * COUNT + array3_size(COUNT, sizeof(TYPE1), sizeof(TYPE2)) , ...) | kmalloc( - sizeof(TYPE1) * sizeof(THING2) * (COUNT) + array3_size(COUNT, sizeof(TYPE1), sizeof(TYPE2)) , ...) | kmalloc( - sizeof(THING1) * sizeof(THING2) * COUNT + array3_size(COUNT, sizeof(THING1), sizeof(THING2)) , ...) | kmalloc( - sizeof(THING1) * sizeof(THING2) * (COUNT) + array3_size(COUNT, sizeof(THING1), sizeof(THING2)) , ...) | kmalloc( - sizeof(TYPE1) * sizeof(THING2) * COUNT + array3_size(COUNT, sizeof(TYPE1), sizeof(THING2)) , ...) | kmalloc( - sizeof(TYPE1) * sizeof(THING2) * (COUNT) + array3_size(COUNT, sizeof(TYPE1), sizeof(THING2)) , ...) ) // 3-factor product, only identifiers, with redundant parens removed. @@ identifier STRIDE, SIZE, COUNT; @@ ( kmalloc( - (COUNT) * STRIDE * SIZE + array3_size(COUNT, STRIDE, SIZE) , ...) | kmalloc( - COUNT * (STRIDE) * SIZE + array3_size(COUNT, STRIDE, SIZE) , ...) | kmalloc( - COUNT * STRIDE * (SIZE) + array3_size(COUNT, STRIDE, SIZE) , ...) | kmalloc( - (COUNT) * (STRIDE) * SIZE + array3_size(COUNT, STRIDE, SIZE) , ...) | kmalloc( - COUNT * (STRIDE) * (SIZE) + array3_size(COUNT, STRIDE, SIZE) , ...) | kmalloc( - (COUNT) * STRIDE * (SIZE) + array3_size(COUNT, STRIDE, SIZE) , ...) | kmalloc( - (COUNT) * (STRIDE) * (SIZE) + array3_size(COUNT, STRIDE, SIZE) , ...) | kmalloc( - COUNT * STRIDE * SIZE + array3_size(COUNT, STRIDE, SIZE) , ...) ) // Any remaining multi-factor products, first at least 3-factor products, // when they're not all constants... @@ expression E1, E2, E3; constant C1, C2, C3; @@ ( kmalloc(C1 * C2 * C3, ...) | kmalloc( - (E1) * E2 * E3 + array3_size(E1, E2, E3) , ...) | kmalloc( - (E1) * (E2) * E3 + array3_size(E1, E2, E3) , ...) | kmalloc( - (E1) * (E2) * (E3) + array3_size(E1, E2, E3) , ...) | kmalloc( - E1 * E2 * E3 + array3_size(E1, E2, E3) , ...) ) // And then all remaining 2 factors products when they're not all constants, // keeping sizeof() as the second factor argument. @@ expression THING, E1, E2; type TYPE; constant C1, C2, C3; @@ ( kmalloc(sizeof(THING) * C2, ...) | kmalloc(sizeof(TYPE) * C2, ...) | kmalloc(C1 * C2 * C3, ...) | kmalloc(C1 * C2, ...) | - kmalloc + kmalloc_array ( - sizeof(TYPE) * (E2) + E2, sizeof(TYPE) , ...) | - kmalloc + kmalloc_array ( - sizeof(TYPE) * E2 + E2, sizeof(TYPE) , ...) | - kmalloc + kmalloc_array ( - sizeof(THING) * (E2) + E2, sizeof(THING) , ...) | - kmalloc + kmalloc_array ( - sizeof(THING) * E2 + E2, sizeof(THING) , ...) | - kmalloc + kmalloc_array ( - (E1) * E2 + E1, E2 , ...) | - kmalloc + kmalloc_array ( - (E1) * (E2) + E1, E2 , ...) | - kmalloc + kmalloc_array ( - E1 * E2 + E1, E2 , ...) ) Signed-off-by: Kees Cook <keescook@chromium.org>
2018-06-13 04:55:00 +08:00
_dm_origins = kmalloc_array(ORIGIN_HASH_SIZE,
sizeof(struct list_head),
GFP_KERNEL);
dm snapshot: suspend origin when doing exception handover In the function snapshot_resume we perform exception store handover. If there is another active snapshot target, the exception store is moved from this target to the target that is being resumed. The problem is that if there is some pending exception, it will point to an incorrect exception store after that handover, causing a crash due to dm-snap-persistent.c:get_exception()'s BUG_ON. This bug can be triggered by repeatedly changing snapshot permissions with "lvchange -p r" and "lvchange -p rw" while there are writes on the associated origin device. To fix this bug, we must suspend the origin device when doing the exception store handover to make sure that there are no pending exceptions: - introduce _origin_hash that keeps track of dm_origin structures. - introduce functions __lookup_dm_origin, __insert_dm_origin and __remove_dm_origin that manipulate the origin hash. - modify snapshot_resume so that it calls dm_internal_suspend_fast() and dm_internal_resume_fast() on the origin device. NOTE to stable@ people: When backporting to kernels 3.12-3.18, use dm_internal_suspend and dm_internal_resume instead of dm_internal_suspend_fast and dm_internal_resume_fast. When backporting to kernels older than 3.12, you need to pick functions dm_internal_suspend and dm_internal_resume from the commit fd2ed4d252701d3bbed4cd3e3d267ad469bb832a. Signed-off-by: Mikulas Patocka <mpatocka@redhat.com> Signed-off-by: Mike Snitzer <snitzer@redhat.com> Cc: stable@vger.kernel.org
2015-02-27 00:40:35 +08:00
if (!_dm_origins) {
DMERR("unable to allocate memory for _dm_origins");
kfree(_origins);
return -ENOMEM;
}
for (i = 0; i < ORIGIN_HASH_SIZE; i++)
INIT_LIST_HEAD(_dm_origins + i);
init_rwsem(&_origins_lock);
return 0;
}
static void exit_origin_hash(void)
{
kfree(_origins);
dm snapshot: suspend origin when doing exception handover In the function snapshot_resume we perform exception store handover. If there is another active snapshot target, the exception store is moved from this target to the target that is being resumed. The problem is that if there is some pending exception, it will point to an incorrect exception store after that handover, causing a crash due to dm-snap-persistent.c:get_exception()'s BUG_ON. This bug can be triggered by repeatedly changing snapshot permissions with "lvchange -p r" and "lvchange -p rw" while there are writes on the associated origin device. To fix this bug, we must suspend the origin device when doing the exception store handover to make sure that there are no pending exceptions: - introduce _origin_hash that keeps track of dm_origin structures. - introduce functions __lookup_dm_origin, __insert_dm_origin and __remove_dm_origin that manipulate the origin hash. - modify snapshot_resume so that it calls dm_internal_suspend_fast() and dm_internal_resume_fast() on the origin device. NOTE to stable@ people: When backporting to kernels 3.12-3.18, use dm_internal_suspend and dm_internal_resume instead of dm_internal_suspend_fast and dm_internal_resume_fast. When backporting to kernels older than 3.12, you need to pick functions dm_internal_suspend and dm_internal_resume from the commit fd2ed4d252701d3bbed4cd3e3d267ad469bb832a. Signed-off-by: Mikulas Patocka <mpatocka@redhat.com> Signed-off-by: Mike Snitzer <snitzer@redhat.com> Cc: stable@vger.kernel.org
2015-02-27 00:40:35 +08:00
kfree(_dm_origins);
}
static unsigned origin_hash(struct block_device *bdev)
{
return bdev->bd_dev & ORIGIN_MASK;
}
static struct origin *__lookup_origin(struct block_device *origin)
{
struct list_head *ol;
struct origin *o;
ol = &_origins[origin_hash(origin)];
list_for_each_entry (o, ol, hash_list)
if (bdev_equal(o->bdev, origin))
return o;
return NULL;
}
static void __insert_origin(struct origin *o)
{
struct list_head *sl = &_origins[origin_hash(o->bdev)];
list_add_tail(&o->hash_list, sl);
}
dm snapshot: suspend origin when doing exception handover In the function snapshot_resume we perform exception store handover. If there is another active snapshot target, the exception store is moved from this target to the target that is being resumed. The problem is that if there is some pending exception, it will point to an incorrect exception store after that handover, causing a crash due to dm-snap-persistent.c:get_exception()'s BUG_ON. This bug can be triggered by repeatedly changing snapshot permissions with "lvchange -p r" and "lvchange -p rw" while there are writes on the associated origin device. To fix this bug, we must suspend the origin device when doing the exception store handover to make sure that there are no pending exceptions: - introduce _origin_hash that keeps track of dm_origin structures. - introduce functions __lookup_dm_origin, __insert_dm_origin and __remove_dm_origin that manipulate the origin hash. - modify snapshot_resume so that it calls dm_internal_suspend_fast() and dm_internal_resume_fast() on the origin device. NOTE to stable@ people: When backporting to kernels 3.12-3.18, use dm_internal_suspend and dm_internal_resume instead of dm_internal_suspend_fast and dm_internal_resume_fast. When backporting to kernels older than 3.12, you need to pick functions dm_internal_suspend and dm_internal_resume from the commit fd2ed4d252701d3bbed4cd3e3d267ad469bb832a. Signed-off-by: Mikulas Patocka <mpatocka@redhat.com> Signed-off-by: Mike Snitzer <snitzer@redhat.com> Cc: stable@vger.kernel.org
2015-02-27 00:40:35 +08:00
static struct dm_origin *__lookup_dm_origin(struct block_device *origin)
{
struct list_head *ol;
struct dm_origin *o;
ol = &_dm_origins[origin_hash(origin)];
list_for_each_entry (o, ol, hash_list)
if (bdev_equal(o->dev->bdev, origin))
return o;
return NULL;
}
static void __insert_dm_origin(struct dm_origin *o)
{
struct list_head *sl = &_dm_origins[origin_hash(o->dev->bdev)];
list_add_tail(&o->hash_list, sl);
}
static void __remove_dm_origin(struct dm_origin *o)
{
list_del(&o->hash_list);
}
/*
* _origins_lock must be held when calling this function.
* Returns number of snapshots registered using the supplied cow device, plus:
* snap_src - a snapshot suitable for use as a source of exception handover
* snap_dest - a snapshot capable of receiving exception handover.
* snap_merge - an existing snapshot-merge target linked to the same origin.
* There can be at most one snapshot-merge target. The parameter is optional.
*
* Possible return values and states of snap_src and snap_dest.
* 0: NULL, NULL - first new snapshot
* 1: snap_src, NULL - normal snapshot
* 2: snap_src, snap_dest - waiting for handover
* 2: snap_src, NULL - handed over, waiting for old to be deleted
* 1: NULL, snap_dest - source got destroyed without handover
*/
static int __find_snapshots_sharing_cow(struct dm_snapshot *snap,
struct dm_snapshot **snap_src,
struct dm_snapshot **snap_dest,
struct dm_snapshot **snap_merge)
{
struct dm_snapshot *s;
struct origin *o;
int count = 0;
int active;
o = __lookup_origin(snap->origin->bdev);
if (!o)
goto out;
list_for_each_entry(s, &o->snapshots, list) {
if (dm_target_is_snapshot_merge(s->ti) && snap_merge)
*snap_merge = s;
if (!bdev_equal(s->cow->bdev, snap->cow->bdev))
continue;
mutex_lock(&s->lock);
active = s->active;
mutex_unlock(&s->lock);
if (active) {
if (snap_src)
*snap_src = s;
} else if (snap_dest)
*snap_dest = s;
count++;
}
out:
return count;
}
/*
* On success, returns 1 if this snapshot is a handover destination,
* otherwise returns 0.
*/
static int __validate_exception_handover(struct dm_snapshot *snap)
{
struct dm_snapshot *snap_src = NULL, *snap_dest = NULL;
struct dm_snapshot *snap_merge = NULL;
/* Does snapshot need exceptions handed over to it? */
if ((__find_snapshots_sharing_cow(snap, &snap_src, &snap_dest,
&snap_merge) == 2) ||
snap_dest) {
snap->ti->error = "Snapshot cow pairing for exception "
"table handover failed";
return -EINVAL;
}
/*
* If no snap_src was found, snap cannot become a handover
* destination.
*/
if (!snap_src)
return 0;
/*
* Non-snapshot-merge handover?
*/
if (!dm_target_is_snapshot_merge(snap->ti))
return 1;
/*
* Do not allow more than one merging snapshot.
*/
if (snap_merge) {
snap->ti->error = "A snapshot is already merging.";
return -EINVAL;
}
if (!snap_src->store->type->prepare_merge ||
!snap_src->store->type->commit_merge) {
snap->ti->error = "Snapshot exception store does not "
"support snapshot-merge.";
return -EINVAL;
}
return 1;
}
static void __insert_snapshot(struct origin *o, struct dm_snapshot *s)
{
struct dm_snapshot *l;
/* Sort the list according to chunk size, largest-first smallest-last */
list_for_each_entry(l, &o->snapshots, list)
if (l->store->chunk_size < s->store->chunk_size)
break;
list_add_tail(&s->list, &l->list);
}
/*
* Make a note of the snapshot and its origin so we can look it
* up when the origin has a write on it.
*
* Also validate snapshot exception store handovers.
* On success, returns 1 if this registration is a handover destination,
* otherwise returns 0.
*/
static int register_snapshot(struct dm_snapshot *snap)
{
struct origin *o, *new_o = NULL;
struct block_device *bdev = snap->origin->bdev;
int r = 0;
new_o = kmalloc(sizeof(*new_o), GFP_KERNEL);
if (!new_o)
return -ENOMEM;
down_write(&_origins_lock);
r = __validate_exception_handover(snap);
if (r < 0) {
kfree(new_o);
goto out;
}
o = __lookup_origin(bdev);
if (o)
kfree(new_o);
else {
/* New origin */
o = new_o;
/* Initialise the struct */
INIT_LIST_HEAD(&o->snapshots);
o->bdev = bdev;
__insert_origin(o);
}
__insert_snapshot(o, snap);
out:
up_write(&_origins_lock);
return r;
}
/*
* Move snapshot to correct place in list according to chunk size.
*/
static void reregister_snapshot(struct dm_snapshot *s)
{
struct block_device *bdev = s->origin->bdev;
down_write(&_origins_lock);
list_del(&s->list);
__insert_snapshot(__lookup_origin(bdev), s);
up_write(&_origins_lock);
}
static void unregister_snapshot(struct dm_snapshot *s)
{
struct origin *o;
down_write(&_origins_lock);
o = __lookup_origin(s->origin->bdev);
list_del(&s->list);
if (o && list_empty(&o->snapshots)) {
list_del(&o->hash_list);
kfree(o);
}
up_write(&_origins_lock);
}
/*
* Implementation of the exception hash tables.
* The lowest hash_shift bits of the chunk number are ignored, allowing
* some consecutive chunks to be grouped together.
*/
static int dm_exception_table_init(struct dm_exception_table *et,
uint32_t size, unsigned hash_shift)
{
unsigned int i;
et->hash_shift = hash_shift;
et->hash_mask = size - 1;
et->table = dm_vcalloc(size, sizeof(struct list_head));
if (!et->table)
return -ENOMEM;
for (i = 0; i < size; i++)
INIT_LIST_HEAD(et->table + i);
return 0;
}
static void dm_exception_table_exit(struct dm_exception_table *et,
struct kmem_cache *mem)
{
struct list_head *slot;
struct dm_exception *ex, *next;
int i, size;
size = et->hash_mask + 1;
for (i = 0; i < size; i++) {
slot = et->table + i;
list_for_each_entry_safe (ex, next, slot, hash_list)
kmem_cache_free(mem, ex);
}
vfree(et->table);
}
static uint32_t exception_hash(struct dm_exception_table *et, chunk_t chunk)
{
return (chunk >> et->hash_shift) & et->hash_mask;
}
static void dm_remove_exception(struct dm_exception *e)
{
list_del(&e->hash_list);
}
/*
* Return the exception data for a sector, or NULL if not
* remapped.
*/
static struct dm_exception *dm_lookup_exception(struct dm_exception_table *et,
chunk_t chunk)
{
struct list_head *slot;
struct dm_exception *e;
slot = &et->table[exception_hash(et, chunk)];
list_for_each_entry (e, slot, hash_list)
if (chunk >= e->old_chunk &&
chunk <= e->old_chunk + dm_consecutive_chunk_count(e))
return e;
return NULL;
}
static struct dm_exception *alloc_completed_exception(gfp_t gfp)
{
struct dm_exception *e;
e = kmem_cache_alloc(exception_cache, gfp);
if (!e && gfp == GFP_NOIO)
e = kmem_cache_alloc(exception_cache, GFP_ATOMIC);
return e;
}
static void free_completed_exception(struct dm_exception *e)
{
kmem_cache_free(exception_cache, e);
}
static struct dm_snap_pending_exception *alloc_pending_exception(struct dm_snapshot *s)
{
struct dm_snap_pending_exception *pe = mempool_alloc(&s->pending_pool,
GFP_NOIO);
dm snapshot: wait for chunks in destructor If there are several snapshots sharing an origin and one is removed while the origin is being written to, the snapshot's mempool may get deleted while elements are still referenced. Prior to dm-snapshot-use-per-device-mempools.patch the pending exceptions may still have been referenced after the snapshot was destroyed, but this was not a problem because the shared mempool was still there. This patch fixes the problem by tracking the number of mempool elements in use. The scenario: - You have an origin and two snapshots 1 and 2. - Someone writes to the origin. - It creates two exceptions in the snapshots, snapshot 1 will be primary exception, snapshot 2's pending_exception->primary_pe will point to the exception in snapshot 1. - The exceptions are being relocated, relocation of exception 1 finishes (but it's pending_exception is still allocated, because it is referenced by an exception from snapshot 2) - The user lvremoves snapshot 1 --- it calls just suspend (does nothing) and destructor. md->pending is zero (there is no I/O submitted to the snapshot by md layer), so it won't help us. - The destructor waits for kcopyd jobs to finish on snapshot 1 --- but there are none. - The destructor on snapshot 1 cleans up everything. - The relocation of exception on snapshot 2 finishes, it drops reference on primary_pe. This frees its primary_pe pointer. Primary_pe points to pending exception created for snapshot 1. So it frees memory into non-existing mempool. Signed-off-by: Mikulas Patocka <mpatocka@redhat.com> Signed-off-by: Alasdair G Kergon <agk@redhat.com>
2008-10-30 21:33:16 +08:00
atomic_inc(&s->pending_exceptions_count);
pe->snap = s;
return pe;
}
static void free_pending_exception(struct dm_snap_pending_exception *pe)
{
dm snapshot: wait for chunks in destructor If there are several snapshots sharing an origin and one is removed while the origin is being written to, the snapshot's mempool may get deleted while elements are still referenced. Prior to dm-snapshot-use-per-device-mempools.patch the pending exceptions may still have been referenced after the snapshot was destroyed, but this was not a problem because the shared mempool was still there. This patch fixes the problem by tracking the number of mempool elements in use. The scenario: - You have an origin and two snapshots 1 and 2. - Someone writes to the origin. - It creates two exceptions in the snapshots, snapshot 1 will be primary exception, snapshot 2's pending_exception->primary_pe will point to the exception in snapshot 1. - The exceptions are being relocated, relocation of exception 1 finishes (but it's pending_exception is still allocated, because it is referenced by an exception from snapshot 2) - The user lvremoves snapshot 1 --- it calls just suspend (does nothing) and destructor. md->pending is zero (there is no I/O submitted to the snapshot by md layer), so it won't help us. - The destructor waits for kcopyd jobs to finish on snapshot 1 --- but there are none. - The destructor on snapshot 1 cleans up everything. - The relocation of exception on snapshot 2 finishes, it drops reference on primary_pe. This frees its primary_pe pointer. Primary_pe points to pending exception created for snapshot 1. So it frees memory into non-existing mempool. Signed-off-by: Mikulas Patocka <mpatocka@redhat.com> Signed-off-by: Alasdair G Kergon <agk@redhat.com>
2008-10-30 21:33:16 +08:00
struct dm_snapshot *s = pe->snap;
mempool_free(pe, &s->pending_pool);
smp_mb__before_atomic();
dm snapshot: wait for chunks in destructor If there are several snapshots sharing an origin and one is removed while the origin is being written to, the snapshot's mempool may get deleted while elements are still referenced. Prior to dm-snapshot-use-per-device-mempools.patch the pending exceptions may still have been referenced after the snapshot was destroyed, but this was not a problem because the shared mempool was still there. This patch fixes the problem by tracking the number of mempool elements in use. The scenario: - You have an origin and two snapshots 1 and 2. - Someone writes to the origin. - It creates two exceptions in the snapshots, snapshot 1 will be primary exception, snapshot 2's pending_exception->primary_pe will point to the exception in snapshot 1. - The exceptions are being relocated, relocation of exception 1 finishes (but it's pending_exception is still allocated, because it is referenced by an exception from snapshot 2) - The user lvremoves snapshot 1 --- it calls just suspend (does nothing) and destructor. md->pending is zero (there is no I/O submitted to the snapshot by md layer), so it won't help us. - The destructor waits for kcopyd jobs to finish on snapshot 1 --- but there are none. - The destructor on snapshot 1 cleans up everything. - The relocation of exception on snapshot 2 finishes, it drops reference on primary_pe. This frees its primary_pe pointer. Primary_pe points to pending exception created for snapshot 1. So it frees memory into non-existing mempool. Signed-off-by: Mikulas Patocka <mpatocka@redhat.com> Signed-off-by: Alasdair G Kergon <agk@redhat.com>
2008-10-30 21:33:16 +08:00
atomic_dec(&s->pending_exceptions_count);
}
static void dm_insert_exception(struct dm_exception_table *eh,
struct dm_exception *new_e)
{
struct list_head *l;
struct dm_exception *e = NULL;
l = &eh->table[exception_hash(eh, new_e->old_chunk)];
/* Add immediately if this table doesn't support consecutive chunks */
if (!eh->hash_shift)
goto out;
/* List is ordered by old_chunk */
list_for_each_entry_reverse(e, l, hash_list) {
/* Insert after an existing chunk? */
if (new_e->old_chunk == (e->old_chunk +
dm_consecutive_chunk_count(e) + 1) &&
new_e->new_chunk == (dm_chunk_number(e->new_chunk) +
dm_consecutive_chunk_count(e) + 1)) {
dm_consecutive_chunk_count_inc(e);
free_completed_exception(new_e);
return;
}
/* Insert before an existing chunk? */
if (new_e->old_chunk == (e->old_chunk - 1) &&
new_e->new_chunk == (dm_chunk_number(e->new_chunk) - 1)) {
dm_consecutive_chunk_count_inc(e);
e->old_chunk--;
e->new_chunk--;
free_completed_exception(new_e);
return;
}
if (new_e->old_chunk > e->old_chunk)
break;
}
out:
list_add(&new_e->hash_list, e ? &e->hash_list : l);
}
/*
* Callback used by the exception stores to load exceptions when
* initialising.
*/
static int dm_add_exception(void *context, chunk_t old, chunk_t new)
{
struct dm_snapshot *s = context;
struct dm_exception *e;
e = alloc_completed_exception(GFP_KERNEL);
if (!e)
return -ENOMEM;
e->old_chunk = old;
/* Consecutive_count is implicitly initialised to zero */
e->new_chunk = new;
dm_insert_exception(&s->complete, e);
return 0;
}
/*
* Return a minimum chunk size of all snapshots that have the specified origin.
* Return zero if the origin has no snapshots.
*/
static uint32_t __minimum_chunk_size(struct origin *o)
{
struct dm_snapshot *snap;
unsigned chunk_size = 0;
if (o)
list_for_each_entry(snap, &o->snapshots, list)
chunk_size = min_not_zero(chunk_size,
snap->store->chunk_size);
return (uint32_t) chunk_size;
}
/*
* Hard coded magic.
*/
static int calc_max_buckets(void)
{
/* use a fixed size of 2MB */
unsigned long mem = 2 * 1024 * 1024;
mem /= sizeof(struct list_head);
return mem;
}
/*
* Allocate room for a suitable hash table.
*/
static int init_hash_tables(struct dm_snapshot *s)
{
sector_t hash_size, cow_dev_size, max_buckets;
/*
* Calculate based on the size of the original volume or
* the COW volume...
*/
cow_dev_size = get_dev_size(s->cow->bdev);
max_buckets = calc_max_buckets();
hash_size = cow_dev_size >> s->store->chunk_shift;
hash_size = min(hash_size, max_buckets);
if (hash_size < 64)
hash_size = 64;
hash_size = rounddown_pow_of_two(hash_size);
if (dm_exception_table_init(&s->complete, hash_size,
DM_CHUNK_CONSECUTIVE_BITS))
return -ENOMEM;
/*
* Allocate hash table for in-flight exceptions
* Make this smaller than the real hash table
*/
hash_size >>= 3;
if (hash_size < 64)
hash_size = 64;
if (dm_exception_table_init(&s->pending, hash_size, 0)) {
dm_exception_table_exit(&s->complete, exception_cache);
return -ENOMEM;
}
return 0;
}
static void merge_shutdown(struct dm_snapshot *s)
{
clear_bit_unlock(RUNNING_MERGE, &s->state_bits);
smp_mb__after_atomic();
wake_up_bit(&s->state_bits, RUNNING_MERGE);
}
static struct bio *__release_queued_bios_after_merge(struct dm_snapshot *s)
{
s->first_merging_chunk = 0;
s->num_merging_chunks = 0;
return bio_list_get(&s->bios_queued_during_merge);
}
/*
* Remove one chunk from the index of completed exceptions.
*/
static int __remove_single_exception_chunk(struct dm_snapshot *s,
chunk_t old_chunk)
{
struct dm_exception *e;
e = dm_lookup_exception(&s->complete, old_chunk);
if (!e) {
DMERR("Corruption detected: exception for block %llu is "
"on disk but not in memory",
(unsigned long long)old_chunk);
return -EINVAL;
}
/*
* If this is the only chunk using this exception, remove exception.
*/
if (!dm_consecutive_chunk_count(e)) {
dm_remove_exception(e);
free_completed_exception(e);
return 0;
}
/*
* The chunk may be either at the beginning or the end of a
* group of consecutive chunks - never in the middle. We are
* removing chunks in the opposite order to that in which they
* were added, so this should always be true.
* Decrement the consecutive chunk counter and adjust the
* starting point if necessary.
*/
if (old_chunk == e->old_chunk) {
e->old_chunk++;
e->new_chunk++;
} else if (old_chunk != e->old_chunk +
dm_consecutive_chunk_count(e)) {
DMERR("Attempt to merge block %llu from the "
"middle of a chunk range [%llu - %llu]",
(unsigned long long)old_chunk,
(unsigned long long)e->old_chunk,
(unsigned long long)
e->old_chunk + dm_consecutive_chunk_count(e));
return -EINVAL;
}
dm_consecutive_chunk_count_dec(e);
return 0;
}
static void flush_bios(struct bio *bio);
static int remove_single_exception_chunk(struct dm_snapshot *s)
{
struct bio *b = NULL;
int r;
chunk_t old_chunk = s->first_merging_chunk + s->num_merging_chunks - 1;
mutex_lock(&s->lock);
/*
* Process chunks (and associated exceptions) in reverse order
* so that dm_consecutive_chunk_count_dec() accounting works.
*/
do {
r = __remove_single_exception_chunk(s, old_chunk);
if (r)
goto out;
} while (old_chunk-- > s->first_merging_chunk);
b = __release_queued_bios_after_merge(s);
out:
mutex_unlock(&s->lock);
if (b)
flush_bios(b);
return r;
}
static int origin_write_extent(struct dm_snapshot *merging_snap,
sector_t sector, unsigned chunk_size);
static void merge_callback(int read_err, unsigned long write_err,
void *context);
static uint64_t read_pending_exceptions_done_count(void)
{
uint64_t pending_exceptions_done;
spin_lock(&_pending_exceptions_done_spinlock);
pending_exceptions_done = _pending_exceptions_done_count;
spin_unlock(&_pending_exceptions_done_spinlock);
return pending_exceptions_done;
}
static void increment_pending_exceptions_done_count(void)
{
spin_lock(&_pending_exceptions_done_spinlock);
_pending_exceptions_done_count++;
spin_unlock(&_pending_exceptions_done_spinlock);
wake_up_all(&_pending_exceptions_done);
}
static void snapshot_merge_next_chunks(struct dm_snapshot *s)
{
int i, linear_chunks;
chunk_t old_chunk, new_chunk;
struct dm_io_region src, dest;
sector_t io_size;
uint64_t previous_count;
BUG_ON(!test_bit(RUNNING_MERGE, &s->state_bits));
if (unlikely(test_bit(SHUTDOWN_MERGE, &s->state_bits)))
goto shut;
/*
* valid flag never changes during merge, so no lock required.
*/
if (!s->valid) {
DMERR("Snapshot is invalid: can't merge");
goto shut;
}
linear_chunks = s->store->type->prepare_merge(s->store, &old_chunk,
&new_chunk);
if (linear_chunks <= 0) {
if (linear_chunks < 0) {
DMERR("Read error in exception store: "
"shutting down merge");
mutex_lock(&s->lock);
s->merge_failed = 1;
mutex_unlock(&s->lock);
}
goto shut;
}
/* Adjust old_chunk and new_chunk to reflect start of linear region */
old_chunk = old_chunk + 1 - linear_chunks;
new_chunk = new_chunk + 1 - linear_chunks;
/*
* Use one (potentially large) I/O to copy all 'linear_chunks'
* from the exception store to the origin
*/
io_size = linear_chunks * s->store->chunk_size;
dest.bdev = s->origin->bdev;
dest.sector = chunk_to_sector(s->store, old_chunk);
dest.count = min(io_size, get_dev_size(dest.bdev) - dest.sector);
src.bdev = s->cow->bdev;
src.sector = chunk_to_sector(s->store, new_chunk);
src.count = dest.count;
/*
* Reallocate any exceptions needed in other snapshots then
* wait for the pending exceptions to complete.
* Each time any pending exception (globally on the system)
* completes we are woken and repeat the process to find out
* if we can proceed. While this may not seem a particularly
* efficient algorithm, it is not expected to have any
* significant impact on performance.
*/
previous_count = read_pending_exceptions_done_count();
while (origin_write_extent(s, dest.sector, io_size)) {
wait_event(_pending_exceptions_done,
(read_pending_exceptions_done_count() !=
previous_count));
/* Retry after the wait, until all exceptions are done. */
previous_count = read_pending_exceptions_done_count();
}
mutex_lock(&s->lock);
s->first_merging_chunk = old_chunk;
s->num_merging_chunks = linear_chunks;
mutex_unlock(&s->lock);
/* Wait until writes to all 'linear_chunks' drain */
for (i = 0; i < linear_chunks; i++)
__check_for_conflicting_io(s, old_chunk + i);
dm_kcopyd_copy(s->kcopyd_client, &src, 1, &dest, 0, merge_callback, s);
return;
shut:
merge_shutdown(s);
}
static void error_bios(struct bio *bio);
static void merge_callback(int read_err, unsigned long write_err, void *context)
{
struct dm_snapshot *s = context;
struct bio *b = NULL;
if (read_err || write_err) {
if (read_err)
DMERR("Read error: shutting down merge.");
else
DMERR("Write error: shutting down merge.");
goto shut;
}
if (s->store->type->commit_merge(s->store,
s->num_merging_chunks) < 0) {
DMERR("Write error in exception store: shutting down merge");
goto shut;
}
if (remove_single_exception_chunk(s) < 0)
goto shut;
snapshot_merge_next_chunks(s);
return;
shut:
mutex_lock(&s->lock);
s->merge_failed = 1;
b = __release_queued_bios_after_merge(s);
mutex_unlock(&s->lock);
error_bios(b);
merge_shutdown(s);
}
static void start_merge(struct dm_snapshot *s)
{
if (!test_and_set_bit(RUNNING_MERGE, &s->state_bits))
snapshot_merge_next_chunks(s);
}
/*
* Stop the merging process and wait until it finishes.
*/
static void stop_merge(struct dm_snapshot *s)
{
set_bit(SHUTDOWN_MERGE, &s->state_bits);
sched: Remove proliferation of wait_on_bit() action functions The current "wait_on_bit" interface requires an 'action' function to be provided which does the actual waiting. There are over 20 such functions, many of them identical. Most cases can be satisfied by one of just two functions, one which uses io_schedule() and one which just uses schedule(). So: Rename wait_on_bit and wait_on_bit_lock to wait_on_bit_action and wait_on_bit_lock_action to make it explicit that they need an action function. Introduce new wait_on_bit{,_lock} and wait_on_bit{,_lock}_io which are *not* given an action function but implicitly use a standard one. The decision to error-out if a signal is pending is now made based on the 'mode' argument rather than being encoded in the action function. All instances of the old wait_on_bit and wait_on_bit_lock which can use the new version have been changed accordingly and their action functions have been discarded. wait_on_bit{_lock} does not return any specific error code in the event of a signal so the caller must check for non-zero and interpolate their own error code as appropriate. The wait_on_bit() call in __fscache_wait_on_invalidate() was ambiguous as it specified TASK_UNINTERRUPTIBLE but used fscache_wait_bit_interruptible as an action function. David Howells confirms this should be uniformly "uninterruptible" The main remaining user of wait_on_bit{,_lock}_action is NFS which needs to use a freezer-aware schedule() call. A comment in fs/gfs2/glock.c notes that having multiple 'action' functions is useful as they display differently in the 'wchan' field of 'ps'. (and /proc/$PID/wchan). As the new bit_wait{,_io} functions are tagged "__sched", they will not show up at all, but something higher in the stack. So the distinction will still be visible, only with different function names (gds2_glock_wait versus gfs2_glock_dq_wait in the gfs2/glock.c case). Since first version of this patch (against 3.15) two new action functions appeared, on in NFS and one in CIFS. CIFS also now uses an action function that makes the same freezer aware schedule call as NFS. Signed-off-by: NeilBrown <neilb@suse.de> Acked-by: David Howells <dhowells@redhat.com> (fscache, keys) Acked-by: Steven Whitehouse <swhiteho@redhat.com> (gfs2) Acked-by: Peter Zijlstra <peterz@infradead.org> Cc: Oleg Nesterov <oleg@redhat.com> Cc: Steve French <sfrench@samba.org> Cc: Linus Torvalds <torvalds@linux-foundation.org> Link: http://lkml.kernel.org/r/20140707051603.28027.72349.stgit@notabene.brown Signed-off-by: Ingo Molnar <mingo@kernel.org>
2014-07-07 13:16:04 +08:00
wait_on_bit(&s->state_bits, RUNNING_MERGE, TASK_UNINTERRUPTIBLE);
clear_bit(SHUTDOWN_MERGE, &s->state_bits);
}
/*
* Construct a snapshot mapping: <origin_dev> <COW-dev> <p|po|n> <chunk-size>
*/
static int snapshot_ctr(struct dm_target *ti, unsigned int argc, char **argv)
{
struct dm_snapshot *s;
int i;
int r = -EINVAL;
char *origin_path, *cow_path;
dm snapshot: disallow the COW and origin devices from being identical Otherwise loading a "snapshot" table using the same device for the origin and COW devices, e.g.: echo "0 20971520 snapshot 253:3 253:3 P 8" | dmsetup create snap will trigger: BUG: unable to handle kernel NULL pointer dereference at 0000000000000098 [ 1958.979934] IP: [<ffffffffa040efba>] dm_exception_store_set_chunk_size+0x7a/0x110 [dm_snapshot] [ 1958.989655] PGD 0 [ 1958.991903] Oops: 0000 [#1] SMP ... [ 1959.059647] CPU: 9 PID: 3556 Comm: dmsetup Tainted: G IO 4.5.0-rc5.snitm+ #150 ... [ 1959.083517] task: ffff8800b9660c80 ti: ffff88032a954000 task.ti: ffff88032a954000 [ 1959.091865] RIP: 0010:[<ffffffffa040efba>] [<ffffffffa040efba>] dm_exception_store_set_chunk_size+0x7a/0x110 [dm_snapshot] [ 1959.104295] RSP: 0018:ffff88032a957b30 EFLAGS: 00010246 [ 1959.110219] RAX: 0000000000000000 RBX: 0000000000000008 RCX: 0000000000000001 [ 1959.118180] RDX: 0000000000000000 RSI: 0000000000000008 RDI: ffff880329334a00 [ 1959.126141] RBP: ffff88032a957b50 R08: 0000000000000000 R09: 0000000000000001 [ 1959.134102] R10: 000000000000000a R11: f000000000000000 R12: ffff880330884d80 [ 1959.142061] R13: 0000000000000008 R14: ffffc90001c13088 R15: ffff880330884d80 [ 1959.150021] FS: 00007f8926ba3840(0000) GS:ffff880333440000(0000) knlGS:0000000000000000 [ 1959.159047] CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033 [ 1959.165456] CR2: 0000000000000098 CR3: 000000032f48b000 CR4: 00000000000006e0 [ 1959.173415] Stack: [ 1959.175656] ffffc90001c13040 ffff880329334a00 ffff880330884ed0 ffff88032a957bdc [ 1959.183946] ffff88032a957bb8 ffffffffa040f225 ffff880329334a30 ffff880300000000 [ 1959.192233] ffffffffa04133e0 ffff880329334b30 0000000830884d58 00000000569c58cf [ 1959.200521] Call Trace: [ 1959.203248] [<ffffffffa040f225>] dm_exception_store_create+0x1d5/0x240 [dm_snapshot] [ 1959.211986] [<ffffffffa040d310>] snapshot_ctr+0x140/0x630 [dm_snapshot] [ 1959.219469] [<ffffffffa0005c44>] ? dm_split_args+0x64/0x150 [dm_mod] [ 1959.226656] [<ffffffffa0005ea7>] dm_table_add_target+0x177/0x440 [dm_mod] [ 1959.234328] [<ffffffffa0009203>] table_load+0x143/0x370 [dm_mod] [ 1959.241129] [<ffffffffa00090c0>] ? retrieve_status+0x1b0/0x1b0 [dm_mod] [ 1959.248607] [<ffffffffa0009e35>] ctl_ioctl+0x255/0x4d0 [dm_mod] [ 1959.255307] [<ffffffff813304e2>] ? memzero_explicit+0x12/0x20 [ 1959.261816] [<ffffffffa000a0c3>] dm_ctl_ioctl+0x13/0x20 [dm_mod] [ 1959.268615] [<ffffffff81215eb6>] do_vfs_ioctl+0xa6/0x5c0 [ 1959.274637] [<ffffffff81120d2f>] ? __audit_syscall_entry+0xaf/0x100 [ 1959.281726] [<ffffffff81003176>] ? do_audit_syscall_entry+0x66/0x70 [ 1959.288814] [<ffffffff81216449>] SyS_ioctl+0x79/0x90 [ 1959.294450] [<ffffffff8167e4ae>] entry_SYSCALL_64_fastpath+0x12/0x71 ... [ 1959.323277] RIP [<ffffffffa040efba>] dm_exception_store_set_chunk_size+0x7a/0x110 [dm_snapshot] [ 1959.333090] RSP <ffff88032a957b30> [ 1959.336978] CR2: 0000000000000098 [ 1959.344121] ---[ end trace b049991ccad1169e ]--- Fixes: https://bugzilla.redhat.com/show_bug.cgi?id=1195899 Cc: stable@vger.kernel.org Signed-off-by: Ding Xiang <dingxiang@huawei.com> Signed-off-by: Mike Snitzer <snitzer@redhat.com>
2016-02-02 12:29:18 +08:00
dev_t origin_dev, cow_dev;
unsigned args_used, num_flush_bios = 1;
fmode_t origin_mode = FMODE_READ;
if (argc != 4) {
ti->error = "requires exactly 4 arguments";
r = -EINVAL;
goto bad;
}
if (dm_target_is_snapshot_merge(ti)) {
num_flush_bios = 2;
origin_mode = FMODE_WRITE;
}
s = kzalloc(sizeof(*s), GFP_KERNEL);
if (!s) {
ti->error = "Cannot allocate private snapshot structure";
r = -ENOMEM;
goto bad;
}
origin_path = argv[0];
argv++;
argc--;
r = dm_get_device(ti, origin_path, origin_mode, &s->origin);
if (r) {
ti->error = "Cannot get origin device";
goto bad_origin;
}
dm snapshot: disallow the COW and origin devices from being identical Otherwise loading a "snapshot" table using the same device for the origin and COW devices, e.g.: echo "0 20971520 snapshot 253:3 253:3 P 8" | dmsetup create snap will trigger: BUG: unable to handle kernel NULL pointer dereference at 0000000000000098 [ 1958.979934] IP: [<ffffffffa040efba>] dm_exception_store_set_chunk_size+0x7a/0x110 [dm_snapshot] [ 1958.989655] PGD 0 [ 1958.991903] Oops: 0000 [#1] SMP ... [ 1959.059647] CPU: 9 PID: 3556 Comm: dmsetup Tainted: G IO 4.5.0-rc5.snitm+ #150 ... [ 1959.083517] task: ffff8800b9660c80 ti: ffff88032a954000 task.ti: ffff88032a954000 [ 1959.091865] RIP: 0010:[<ffffffffa040efba>] [<ffffffffa040efba>] dm_exception_store_set_chunk_size+0x7a/0x110 [dm_snapshot] [ 1959.104295] RSP: 0018:ffff88032a957b30 EFLAGS: 00010246 [ 1959.110219] RAX: 0000000000000000 RBX: 0000000000000008 RCX: 0000000000000001 [ 1959.118180] RDX: 0000000000000000 RSI: 0000000000000008 RDI: ffff880329334a00 [ 1959.126141] RBP: ffff88032a957b50 R08: 0000000000000000 R09: 0000000000000001 [ 1959.134102] R10: 000000000000000a R11: f000000000000000 R12: ffff880330884d80 [ 1959.142061] R13: 0000000000000008 R14: ffffc90001c13088 R15: ffff880330884d80 [ 1959.150021] FS: 00007f8926ba3840(0000) GS:ffff880333440000(0000) knlGS:0000000000000000 [ 1959.159047] CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033 [ 1959.165456] CR2: 0000000000000098 CR3: 000000032f48b000 CR4: 00000000000006e0 [ 1959.173415] Stack: [ 1959.175656] ffffc90001c13040 ffff880329334a00 ffff880330884ed0 ffff88032a957bdc [ 1959.183946] ffff88032a957bb8 ffffffffa040f225 ffff880329334a30 ffff880300000000 [ 1959.192233] ffffffffa04133e0 ffff880329334b30 0000000830884d58 00000000569c58cf [ 1959.200521] Call Trace: [ 1959.203248] [<ffffffffa040f225>] dm_exception_store_create+0x1d5/0x240 [dm_snapshot] [ 1959.211986] [<ffffffffa040d310>] snapshot_ctr+0x140/0x630 [dm_snapshot] [ 1959.219469] [<ffffffffa0005c44>] ? dm_split_args+0x64/0x150 [dm_mod] [ 1959.226656] [<ffffffffa0005ea7>] dm_table_add_target+0x177/0x440 [dm_mod] [ 1959.234328] [<ffffffffa0009203>] table_load+0x143/0x370 [dm_mod] [ 1959.241129] [<ffffffffa00090c0>] ? retrieve_status+0x1b0/0x1b0 [dm_mod] [ 1959.248607] [<ffffffffa0009e35>] ctl_ioctl+0x255/0x4d0 [dm_mod] [ 1959.255307] [<ffffffff813304e2>] ? memzero_explicit+0x12/0x20 [ 1959.261816] [<ffffffffa000a0c3>] dm_ctl_ioctl+0x13/0x20 [dm_mod] [ 1959.268615] [<ffffffff81215eb6>] do_vfs_ioctl+0xa6/0x5c0 [ 1959.274637] [<ffffffff81120d2f>] ? __audit_syscall_entry+0xaf/0x100 [ 1959.281726] [<ffffffff81003176>] ? do_audit_syscall_entry+0x66/0x70 [ 1959.288814] [<ffffffff81216449>] SyS_ioctl+0x79/0x90 [ 1959.294450] [<ffffffff8167e4ae>] entry_SYSCALL_64_fastpath+0x12/0x71 ... [ 1959.323277] RIP [<ffffffffa040efba>] dm_exception_store_set_chunk_size+0x7a/0x110 [dm_snapshot] [ 1959.333090] RSP <ffff88032a957b30> [ 1959.336978] CR2: 0000000000000098 [ 1959.344121] ---[ end trace b049991ccad1169e ]--- Fixes: https://bugzilla.redhat.com/show_bug.cgi?id=1195899 Cc: stable@vger.kernel.org Signed-off-by: Ding Xiang <dingxiang@huawei.com> Signed-off-by: Mike Snitzer <snitzer@redhat.com>
2016-02-02 12:29:18 +08:00
origin_dev = s->origin->bdev->bd_dev;
cow_path = argv[0];
argv++;
argc--;
dm snapshot: disallow the COW and origin devices from being identical Otherwise loading a "snapshot" table using the same device for the origin and COW devices, e.g.: echo "0 20971520 snapshot 253:3 253:3 P 8" | dmsetup create snap will trigger: BUG: unable to handle kernel NULL pointer dereference at 0000000000000098 [ 1958.979934] IP: [<ffffffffa040efba>] dm_exception_store_set_chunk_size+0x7a/0x110 [dm_snapshot] [ 1958.989655] PGD 0 [ 1958.991903] Oops: 0000 [#1] SMP ... [ 1959.059647] CPU: 9 PID: 3556 Comm: dmsetup Tainted: G IO 4.5.0-rc5.snitm+ #150 ... [ 1959.083517] task: ffff8800b9660c80 ti: ffff88032a954000 task.ti: ffff88032a954000 [ 1959.091865] RIP: 0010:[<ffffffffa040efba>] [<ffffffffa040efba>] dm_exception_store_set_chunk_size+0x7a/0x110 [dm_snapshot] [ 1959.104295] RSP: 0018:ffff88032a957b30 EFLAGS: 00010246 [ 1959.110219] RAX: 0000000000000000 RBX: 0000000000000008 RCX: 0000000000000001 [ 1959.118180] RDX: 0000000000000000 RSI: 0000000000000008 RDI: ffff880329334a00 [ 1959.126141] RBP: ffff88032a957b50 R08: 0000000000000000 R09: 0000000000000001 [ 1959.134102] R10: 000000000000000a R11: f000000000000000 R12: ffff880330884d80 [ 1959.142061] R13: 0000000000000008 R14: ffffc90001c13088 R15: ffff880330884d80 [ 1959.150021] FS: 00007f8926ba3840(0000) GS:ffff880333440000(0000) knlGS:0000000000000000 [ 1959.159047] CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033 [ 1959.165456] CR2: 0000000000000098 CR3: 000000032f48b000 CR4: 00000000000006e0 [ 1959.173415] Stack: [ 1959.175656] ffffc90001c13040 ffff880329334a00 ffff880330884ed0 ffff88032a957bdc [ 1959.183946] ffff88032a957bb8 ffffffffa040f225 ffff880329334a30 ffff880300000000 [ 1959.192233] ffffffffa04133e0 ffff880329334b30 0000000830884d58 00000000569c58cf [ 1959.200521] Call Trace: [ 1959.203248] [<ffffffffa040f225>] dm_exception_store_create+0x1d5/0x240 [dm_snapshot] [ 1959.211986] [<ffffffffa040d310>] snapshot_ctr+0x140/0x630 [dm_snapshot] [ 1959.219469] [<ffffffffa0005c44>] ? dm_split_args+0x64/0x150 [dm_mod] [ 1959.226656] [<ffffffffa0005ea7>] dm_table_add_target+0x177/0x440 [dm_mod] [ 1959.234328] [<ffffffffa0009203>] table_load+0x143/0x370 [dm_mod] [ 1959.241129] [<ffffffffa00090c0>] ? retrieve_status+0x1b0/0x1b0 [dm_mod] [ 1959.248607] [<ffffffffa0009e35>] ctl_ioctl+0x255/0x4d0 [dm_mod] [ 1959.255307] [<ffffffff813304e2>] ? memzero_explicit+0x12/0x20 [ 1959.261816] [<ffffffffa000a0c3>] dm_ctl_ioctl+0x13/0x20 [dm_mod] [ 1959.268615] [<ffffffff81215eb6>] do_vfs_ioctl+0xa6/0x5c0 [ 1959.274637] [<ffffffff81120d2f>] ? __audit_syscall_entry+0xaf/0x100 [ 1959.281726] [<ffffffff81003176>] ? do_audit_syscall_entry+0x66/0x70 [ 1959.288814] [<ffffffff81216449>] SyS_ioctl+0x79/0x90 [ 1959.294450] [<ffffffff8167e4ae>] entry_SYSCALL_64_fastpath+0x12/0x71 ... [ 1959.323277] RIP [<ffffffffa040efba>] dm_exception_store_set_chunk_size+0x7a/0x110 [dm_snapshot] [ 1959.333090] RSP <ffff88032a957b30> [ 1959.336978] CR2: 0000000000000098 [ 1959.344121] ---[ end trace b049991ccad1169e ]--- Fixes: https://bugzilla.redhat.com/show_bug.cgi?id=1195899 Cc: stable@vger.kernel.org Signed-off-by: Ding Xiang <dingxiang@huawei.com> Signed-off-by: Mike Snitzer <snitzer@redhat.com>
2016-02-02 12:29:18 +08:00
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";
goto bad_cow;
}
r = dm_exception_store_create(ti, argc, argv, s, &args_used, &s->store);
if (r) {
ti->error = "Couldn't create exception store";
r = -EINVAL;
goto bad_store;
}
argv += args_used;
argc -= args_used;
s->ti = ti;
s->valid = 1;
s->snapshot_overflowed = 0;
s->active = 0;
dm snapshot: wait for chunks in destructor If there are several snapshots sharing an origin and one is removed while the origin is being written to, the snapshot's mempool may get deleted while elements are still referenced. Prior to dm-snapshot-use-per-device-mempools.patch the pending exceptions may still have been referenced after the snapshot was destroyed, but this was not a problem because the shared mempool was still there. This patch fixes the problem by tracking the number of mempool elements in use. The scenario: - You have an origin and two snapshots 1 and 2. - Someone writes to the origin. - It creates two exceptions in the snapshots, snapshot 1 will be primary exception, snapshot 2's pending_exception->primary_pe will point to the exception in snapshot 1. - The exceptions are being relocated, relocation of exception 1 finishes (but it's pending_exception is still allocated, because it is referenced by an exception from snapshot 2) - The user lvremoves snapshot 1 --- it calls just suspend (does nothing) and destructor. md->pending is zero (there is no I/O submitted to the snapshot by md layer), so it won't help us. - The destructor waits for kcopyd jobs to finish on snapshot 1 --- but there are none. - The destructor on snapshot 1 cleans up everything. - The relocation of exception on snapshot 2 finishes, it drops reference on primary_pe. This frees its primary_pe pointer. Primary_pe points to pending exception created for snapshot 1. So it frees memory into non-existing mempool. Signed-off-by: Mikulas Patocka <mpatocka@redhat.com> Signed-off-by: Alasdair G Kergon <agk@redhat.com>
2008-10-30 21:33:16 +08:00
atomic_set(&s->pending_exceptions_count, 0);
dm snapshot: avoid snapshot space leak on crash There is a possible leak of snapshot space in case of crash. The reason for space leaking is that chunks in the snapshot device are allocated sequentially, but they are finished (and stored in the metadata) out of order, depending on the order in which copying finished. For example, supposed that the metadata contains the following records SUPERBLOCK METADATA (blocks 0 ... 250) DATA 0 DATA 1 DATA 2 ... DATA 250 Now suppose that you allocate 10 new data blocks 251-260. Suppose that copying of these blocks finish out of order (block 260 finished first and the block 251 finished last). Now, the snapshot device looks like this: SUPERBLOCK METADATA (blocks 0 ... 250, 260, 259, 258, 257, 256) DATA 0 DATA 1 DATA 2 ... DATA 250 DATA 251 DATA 252 DATA 253 DATA 254 DATA 255 METADATA (blocks 255, 254, 253, 252, 251) DATA 256 DATA 257 DATA 258 DATA 259 DATA 260 Now, if the machine crashes after writing the first metadata block but before writing the second metadata block, the space for areas DATA 250-255 is leaked, it contains no valid data and it will never be used in the future. This patch makes dm-snapshot complete exceptions in the same order they were allocated, thus fixing this bug. Note: when backporting this patch to the stable kernel, change the version field in the following way: * if version in the stable kernel is {1, 11, 1}, change it to {1, 12, 0} * if version in the stable kernel is {1, 10, 0} or {1, 10, 1}, change it to {1, 10, 2} Userspace reads the version to determine if the bug was fixed, so the version change is needed. Signed-off-by: Mikulas Patocka <mpatocka@redhat.com> Signed-off-by: Mike Snitzer <snitzer@redhat.com> Cc: stable@vger.kernel.org
2013-11-30 07:13:37 +08:00
s->exception_start_sequence = 0;
s->exception_complete_sequence = 0;
s->out_of_order_tree = RB_ROOT;
mutex_init(&s->lock);
INIT_LIST_HEAD(&s->list);
spin_lock_init(&s->pe_lock);
s->state_bits = 0;
s->merge_failed = 0;
s->first_merging_chunk = 0;
s->num_merging_chunks = 0;
bio_list_init(&s->bios_queued_during_merge);
/* Allocate hash table for COW data */
if (init_hash_tables(s)) {
ti->error = "Unable to allocate hash table space";
r = -ENOMEM;
goto bad_hash_tables;
}
s->kcopyd_client = dm_kcopyd_client_create(&dm_kcopyd_throttle);
if (IS_ERR(s->kcopyd_client)) {
r = PTR_ERR(s->kcopyd_client);
ti->error = "Could not create kcopyd client";
goto bad_kcopyd;
}
r = mempool_init_slab_pool(&s->pending_pool, MIN_IOS, pending_cache);
if (r) {
ti->error = "Could not allocate mempool for pending exceptions";
goto bad_pending_pool;
}
for (i = 0; i < DM_TRACKED_CHUNK_HASH_SIZE; i++)
INIT_HLIST_HEAD(&s->tracked_chunk_hash[i]);
spin_lock_init(&s->tracked_chunk_lock);
ti->private = s;
ti->num_flush_bios = num_flush_bios;
ti->per_io_data_size = sizeof(struct dm_snap_tracked_chunk);
/* Add snapshot to the list of snapshots for this origin */
/* Exceptions aren't triggered till snapshot_resume() is called */
r = register_snapshot(s);
if (r == -ENOMEM) {
ti->error = "Snapshot origin struct allocation failed";
goto bad_load_and_register;
} else if (r < 0) {
/* invalid handover, register_snapshot has set ti->error */
goto bad_load_and_register;
}
/*
* Metadata must only be loaded into one table at once, so skip this
* if metadata will be handed over during resume.
* Chunk size will be set during the handover - set it to zero to
* ensure it's ignored.
*/
if (r > 0) {
s->store->chunk_size = 0;
return 0;
}
r = s->store->type->read_metadata(s->store, dm_add_exception,
(void *)s);
if (r < 0) {
ti->error = "Failed to read snapshot metadata";
goto bad_read_metadata;
} else if (r > 0) {
s->valid = 0;
DMWARN("Snapshot is marked invalid.");
}
if (!s->store->chunk_size) {
ti->error = "Chunk size not set";
goto bad_read_metadata;
}
r = dm_set_target_max_io_len(ti, s->store->chunk_size);
if (r)
goto bad_read_metadata;
return 0;
bad_read_metadata:
unregister_snapshot(s);
bad_load_and_register:
mempool_exit(&s->pending_pool);
bad_pending_pool:
dm_kcopyd_client_destroy(s->kcopyd_client);
bad_kcopyd:
dm_exception_table_exit(&s->pending, pending_cache);
dm_exception_table_exit(&s->complete, exception_cache);
bad_hash_tables:
dm_exception_store_destroy(s->store);
bad_store:
dm_put_device(ti, s->cow);
bad_cow:
dm_put_device(ti, s->origin);
bad_origin:
kfree(s);
bad:
return r;
}
static void __free_exceptions(struct dm_snapshot *s)
{
dm_kcopyd_client_destroy(s->kcopyd_client);
s->kcopyd_client = NULL;
dm_exception_table_exit(&s->pending, pending_cache);
dm_exception_table_exit(&s->complete, exception_cache);
}
static void __handover_exceptions(struct dm_snapshot *snap_src,
struct dm_snapshot *snap_dest)
{
union {
struct dm_exception_table table_swap;
struct dm_exception_store *store_swap;
} u;
/*
* Swap all snapshot context information between the two instances.
*/
u.table_swap = snap_dest->complete;
snap_dest->complete = snap_src->complete;
snap_src->complete = u.table_swap;
u.store_swap = snap_dest->store;
snap_dest->store = snap_src->store;
snap_dest->store->userspace_supports_overflow = u.store_swap->userspace_supports_overflow;
snap_src->store = u.store_swap;
snap_dest->store->snap = snap_dest;
snap_src->store->snap = snap_src;
snap_dest->ti->max_io_len = snap_dest->store->chunk_size;
snap_dest->valid = snap_src->valid;
snap_dest->snapshot_overflowed = snap_src->snapshot_overflowed;
/*
* Set source invalid to ensure it receives no further I/O.
*/
snap_src->valid = 0;
}
static void snapshot_dtr(struct dm_target *ti)
{
#ifdef CONFIG_DM_DEBUG
int i;
#endif
struct dm_snapshot *s = ti->private;
struct dm_snapshot *snap_src = NULL, *snap_dest = NULL;
down_read(&_origins_lock);
/* Check whether exception handover must be cancelled */
(void) __find_snapshots_sharing_cow(s, &snap_src, &snap_dest, NULL);
if (snap_src && snap_dest && (s == snap_src)) {
mutex_lock(&snap_dest->lock);
snap_dest->valid = 0;
mutex_unlock(&snap_dest->lock);
DMERR("Cancelling snapshot handover.");
}
up_read(&_origins_lock);
if (dm_target_is_snapshot_merge(ti))
stop_merge(s);
/* Prevent further origin writes from using this snapshot. */
/* After this returns there can be no new kcopyd jobs. */
unregister_snapshot(s);
dm snapshot: wait for chunks in destructor If there are several snapshots sharing an origin and one is removed while the origin is being written to, the snapshot's mempool may get deleted while elements are still referenced. Prior to dm-snapshot-use-per-device-mempools.patch the pending exceptions may still have been referenced after the snapshot was destroyed, but this was not a problem because the shared mempool was still there. This patch fixes the problem by tracking the number of mempool elements in use. The scenario: - You have an origin and two snapshots 1 and 2. - Someone writes to the origin. - It creates two exceptions in the snapshots, snapshot 1 will be primary exception, snapshot 2's pending_exception->primary_pe will point to the exception in snapshot 1. - The exceptions are being relocated, relocation of exception 1 finishes (but it's pending_exception is still allocated, because it is referenced by an exception from snapshot 2) - The user lvremoves snapshot 1 --- it calls just suspend (does nothing) and destructor. md->pending is zero (there is no I/O submitted to the snapshot by md layer), so it won't help us. - The destructor waits for kcopyd jobs to finish on snapshot 1 --- but there are none. - The destructor on snapshot 1 cleans up everything. - The relocation of exception on snapshot 2 finishes, it drops reference on primary_pe. This frees its primary_pe pointer. Primary_pe points to pending exception created for snapshot 1. So it frees memory into non-existing mempool. Signed-off-by: Mikulas Patocka <mpatocka@redhat.com> Signed-off-by: Alasdair G Kergon <agk@redhat.com>
2008-10-30 21:33:16 +08:00
while (atomic_read(&s->pending_exceptions_count))
msleep(1);
dm snapshot: wait for chunks in destructor If there are several snapshots sharing an origin and one is removed while the origin is being written to, the snapshot's mempool may get deleted while elements are still referenced. Prior to dm-snapshot-use-per-device-mempools.patch the pending exceptions may still have been referenced after the snapshot was destroyed, but this was not a problem because the shared mempool was still there. This patch fixes the problem by tracking the number of mempool elements in use. The scenario: - You have an origin and two snapshots 1 and 2. - Someone writes to the origin. - It creates two exceptions in the snapshots, snapshot 1 will be primary exception, snapshot 2's pending_exception->primary_pe will point to the exception in snapshot 1. - The exceptions are being relocated, relocation of exception 1 finishes (but it's pending_exception is still allocated, because it is referenced by an exception from snapshot 2) - The user lvremoves snapshot 1 --- it calls just suspend (does nothing) and destructor. md->pending is zero (there is no I/O submitted to the snapshot by md layer), so it won't help us. - The destructor waits for kcopyd jobs to finish on snapshot 1 --- but there are none. - The destructor on snapshot 1 cleans up everything. - The relocation of exception on snapshot 2 finishes, it drops reference on primary_pe. This frees its primary_pe pointer. Primary_pe points to pending exception created for snapshot 1. So it frees memory into non-existing mempool. Signed-off-by: Mikulas Patocka <mpatocka@redhat.com> Signed-off-by: Alasdair G Kergon <agk@redhat.com>
2008-10-30 21:33:16 +08:00
/*
* Ensure instructions in mempool_exit aren't reordered
dm snapshot: wait for chunks in destructor If there are several snapshots sharing an origin and one is removed while the origin is being written to, the snapshot's mempool may get deleted while elements are still referenced. Prior to dm-snapshot-use-per-device-mempools.patch the pending exceptions may still have been referenced after the snapshot was destroyed, but this was not a problem because the shared mempool was still there. This patch fixes the problem by tracking the number of mempool elements in use. The scenario: - You have an origin and two snapshots 1 and 2. - Someone writes to the origin. - It creates two exceptions in the snapshots, snapshot 1 will be primary exception, snapshot 2's pending_exception->primary_pe will point to the exception in snapshot 1. - The exceptions are being relocated, relocation of exception 1 finishes (but it's pending_exception is still allocated, because it is referenced by an exception from snapshot 2) - The user lvremoves snapshot 1 --- it calls just suspend (does nothing) and destructor. md->pending is zero (there is no I/O submitted to the snapshot by md layer), so it won't help us. - The destructor waits for kcopyd jobs to finish on snapshot 1 --- but there are none. - The destructor on snapshot 1 cleans up everything. - The relocation of exception on snapshot 2 finishes, it drops reference on primary_pe. This frees its primary_pe pointer. Primary_pe points to pending exception created for snapshot 1. So it frees memory into non-existing mempool. Signed-off-by: Mikulas Patocka <mpatocka@redhat.com> Signed-off-by: Alasdair G Kergon <agk@redhat.com>
2008-10-30 21:33:16 +08:00
* before atomic_read.
*/
smp_mb();
#ifdef CONFIG_DM_DEBUG
for (i = 0; i < DM_TRACKED_CHUNK_HASH_SIZE; i++)
BUG_ON(!hlist_empty(&s->tracked_chunk_hash[i]));
#endif
__free_exceptions(s);
mempool_exit(&s->pending_pool);
dm_exception_store_destroy(s->store);
mutex_destroy(&s->lock);
dm_put_device(ti, s->cow);
dm_put_device(ti, s->origin);
kfree(s);
}
/*
* Flush a list of buffers.
*/
static void flush_bios(struct bio *bio)
{
struct bio *n;
while (bio) {
n = bio->bi_next;
bio->bi_next = NULL;
generic_make_request(bio);
bio = n;
}
}
static int do_origin(struct dm_dev *origin, struct bio *bio);
/*
* Flush a list of buffers.
*/
static void retry_origin_bios(struct dm_snapshot *s, struct bio *bio)
{
struct bio *n;
int r;
while (bio) {
n = bio->bi_next;
bio->bi_next = NULL;
r = do_origin(s->origin, bio);
if (r == DM_MAPIO_REMAPPED)
generic_make_request(bio);
bio = n;
}
}
/*
* Error a list of buffers.
*/
static void error_bios(struct bio *bio)
{
struct bio *n;
while (bio) {
n = bio->bi_next;
bio->bi_next = NULL;
bio_io_error(bio);
bio = n;
}
}
static void __invalidate_snapshot(struct dm_snapshot *s, int err)
{
if (!s->valid)
return;
if (err == -EIO)
DMERR("Invalidating snapshot: Error reading/writing.");
else if (err == -ENOMEM)
DMERR("Invalidating snapshot: Unable to allocate exception.");
if (s->store->type->drop_snapshot)
s->store->type->drop_snapshot(s->store);
s->valid = 0;
dm_table_event(s->ti->table);
}
static void pending_complete(void *context, int success)
{
struct dm_snap_pending_exception *pe = context;
struct dm_exception *e;
struct dm_snapshot *s = pe->snap;
struct bio *origin_bios = NULL;
struct bio *snapshot_bios = NULL;
struct bio *full_bio = NULL;
int error = 0;
if (!success) {
/* Read/write error - snapshot is unusable */
mutex_lock(&s->lock);
__invalidate_snapshot(s, -EIO);
error = 1;
goto out;
}
e = alloc_completed_exception(GFP_NOIO);
if (!e) {
mutex_lock(&s->lock);
__invalidate_snapshot(s, -ENOMEM);
error = 1;
goto out;
}
*e = pe->e;
mutex_lock(&s->lock);
if (!s->valid) {
free_completed_exception(e);
error = 1;
goto out;
}
/* Check for conflicting reads */
__check_for_conflicting_io(s, pe->e.old_chunk);
/*
* Add a proper exception, and remove the
* in-flight exception from the list.
*/
dm_insert_exception(&s->complete, e);
out:
dm_remove_exception(&pe->e);
snapshot_bios = bio_list_get(&pe->snapshot_bios);
origin_bios = bio_list_get(&pe->origin_bios);
full_bio = pe->full_bio;
if (full_bio)
full_bio->bi_end_io = pe->full_bio_end_io;
increment_pending_exceptions_done_count();
mutex_unlock(&s->lock);
/* Submit any pending write bios */
if (error) {
if (full_bio)
bio_io_error(full_bio);
error_bios(snapshot_bios);
} else {
if (full_bio)
bio_endio(full_bio);
flush_bios(snapshot_bios);
}
retry_origin_bios(s, origin_bios);
free_pending_exception(pe);
}
dm snapshot: avoid snapshot space leak on crash There is a possible leak of snapshot space in case of crash. The reason for space leaking is that chunks in the snapshot device are allocated sequentially, but they are finished (and stored in the metadata) out of order, depending on the order in which copying finished. For example, supposed that the metadata contains the following records SUPERBLOCK METADATA (blocks 0 ... 250) DATA 0 DATA 1 DATA 2 ... DATA 250 Now suppose that you allocate 10 new data blocks 251-260. Suppose that copying of these blocks finish out of order (block 260 finished first and the block 251 finished last). Now, the snapshot device looks like this: SUPERBLOCK METADATA (blocks 0 ... 250, 260, 259, 258, 257, 256) DATA 0 DATA 1 DATA 2 ... DATA 250 DATA 251 DATA 252 DATA 253 DATA 254 DATA 255 METADATA (blocks 255, 254, 253, 252, 251) DATA 256 DATA 257 DATA 258 DATA 259 DATA 260 Now, if the machine crashes after writing the first metadata block but before writing the second metadata block, the space for areas DATA 250-255 is leaked, it contains no valid data and it will never be used in the future. This patch makes dm-snapshot complete exceptions in the same order they were allocated, thus fixing this bug. Note: when backporting this patch to the stable kernel, change the version field in the following way: * if version in the stable kernel is {1, 11, 1}, change it to {1, 12, 0} * if version in the stable kernel is {1, 10, 0} or {1, 10, 1}, change it to {1, 10, 2} Userspace reads the version to determine if the bug was fixed, so the version change is needed. Signed-off-by: Mikulas Patocka <mpatocka@redhat.com> Signed-off-by: Mike Snitzer <snitzer@redhat.com> Cc: stable@vger.kernel.org
2013-11-30 07:13:37 +08:00
static void complete_exception(struct dm_snap_pending_exception *pe)
{
struct dm_snapshot *s = pe->snap;
/* Update the metadata if we are persistent */
s->store->type->commit_exception(s->store, &pe->e, !pe->copy_error,
pending_complete, pe);
dm snapshot: avoid snapshot space leak on crash There is a possible leak of snapshot space in case of crash. The reason for space leaking is that chunks in the snapshot device are allocated sequentially, but they are finished (and stored in the metadata) out of order, depending on the order in which copying finished. For example, supposed that the metadata contains the following records SUPERBLOCK METADATA (blocks 0 ... 250) DATA 0 DATA 1 DATA 2 ... DATA 250 Now suppose that you allocate 10 new data blocks 251-260. Suppose that copying of these blocks finish out of order (block 260 finished first and the block 251 finished last). Now, the snapshot device looks like this: SUPERBLOCK METADATA (blocks 0 ... 250, 260, 259, 258, 257, 256) DATA 0 DATA 1 DATA 2 ... DATA 250 DATA 251 DATA 252 DATA 253 DATA 254 DATA 255 METADATA (blocks 255, 254, 253, 252, 251) DATA 256 DATA 257 DATA 258 DATA 259 DATA 260 Now, if the machine crashes after writing the first metadata block but before writing the second metadata block, the space for areas DATA 250-255 is leaked, it contains no valid data and it will never be used in the future. This patch makes dm-snapshot complete exceptions in the same order they were allocated, thus fixing this bug. Note: when backporting this patch to the stable kernel, change the version field in the following way: * if version in the stable kernel is {1, 11, 1}, change it to {1, 12, 0} * if version in the stable kernel is {1, 10, 0} or {1, 10, 1}, change it to {1, 10, 2} Userspace reads the version to determine if the bug was fixed, so the version change is needed. Signed-off-by: Mikulas Patocka <mpatocka@redhat.com> Signed-off-by: Mike Snitzer <snitzer@redhat.com> Cc: stable@vger.kernel.org
2013-11-30 07:13:37 +08:00
}
/*
* Called when the copy I/O has finished. kcopyd actually runs
* this code so don't block.
*/
static void copy_callback(int read_err, unsigned long write_err, void *context)
{
struct dm_snap_pending_exception *pe = context;
struct dm_snapshot *s = pe->snap;
dm snapshot: avoid snapshot space leak on crash There is a possible leak of snapshot space in case of crash. The reason for space leaking is that chunks in the snapshot device are allocated sequentially, but they are finished (and stored in the metadata) out of order, depending on the order in which copying finished. For example, supposed that the metadata contains the following records SUPERBLOCK METADATA (blocks 0 ... 250) DATA 0 DATA 1 DATA 2 ... DATA 250 Now suppose that you allocate 10 new data blocks 251-260. Suppose that copying of these blocks finish out of order (block 260 finished first and the block 251 finished last). Now, the snapshot device looks like this: SUPERBLOCK METADATA (blocks 0 ... 250, 260, 259, 258, 257, 256) DATA 0 DATA 1 DATA 2 ... DATA 250 DATA 251 DATA 252 DATA 253 DATA 254 DATA 255 METADATA (blocks 255, 254, 253, 252, 251) DATA 256 DATA 257 DATA 258 DATA 259 DATA 260 Now, if the machine crashes after writing the first metadata block but before writing the second metadata block, the space for areas DATA 250-255 is leaked, it contains no valid data and it will never be used in the future. This patch makes dm-snapshot complete exceptions in the same order they were allocated, thus fixing this bug. Note: when backporting this patch to the stable kernel, change the version field in the following way: * if version in the stable kernel is {1, 11, 1}, change it to {1, 12, 0} * if version in the stable kernel is {1, 10, 0} or {1, 10, 1}, change it to {1, 10, 2} Userspace reads the version to determine if the bug was fixed, so the version change is needed. Signed-off-by: Mikulas Patocka <mpatocka@redhat.com> Signed-off-by: Mike Snitzer <snitzer@redhat.com> Cc: stable@vger.kernel.org
2013-11-30 07:13:37 +08:00
pe->copy_error = read_err || write_err;
dm snapshot: avoid snapshot space leak on crash There is a possible leak of snapshot space in case of crash. The reason for space leaking is that chunks in the snapshot device are allocated sequentially, but they are finished (and stored in the metadata) out of order, depending on the order in which copying finished. For example, supposed that the metadata contains the following records SUPERBLOCK METADATA (blocks 0 ... 250) DATA 0 DATA 1 DATA 2 ... DATA 250 Now suppose that you allocate 10 new data blocks 251-260. Suppose that copying of these blocks finish out of order (block 260 finished first and the block 251 finished last). Now, the snapshot device looks like this: SUPERBLOCK METADATA (blocks 0 ... 250, 260, 259, 258, 257, 256) DATA 0 DATA 1 DATA 2 ... DATA 250 DATA 251 DATA 252 DATA 253 DATA 254 DATA 255 METADATA (blocks 255, 254, 253, 252, 251) DATA 256 DATA 257 DATA 258 DATA 259 DATA 260 Now, if the machine crashes after writing the first metadata block but before writing the second metadata block, the space for areas DATA 250-255 is leaked, it contains no valid data and it will never be used in the future. This patch makes dm-snapshot complete exceptions in the same order they were allocated, thus fixing this bug. Note: when backporting this patch to the stable kernel, change the version field in the following way: * if version in the stable kernel is {1, 11, 1}, change it to {1, 12, 0} * if version in the stable kernel is {1, 10, 0} or {1, 10, 1}, change it to {1, 10, 2} Userspace reads the version to determine if the bug was fixed, so the version change is needed. Signed-off-by: Mikulas Patocka <mpatocka@redhat.com> Signed-off-by: Mike Snitzer <snitzer@redhat.com> Cc: stable@vger.kernel.org
2013-11-30 07:13:37 +08:00
if (pe->exception_sequence == s->exception_complete_sequence) {
struct rb_node *next;
dm snapshot: avoid snapshot space leak on crash There is a possible leak of snapshot space in case of crash. The reason for space leaking is that chunks in the snapshot device are allocated sequentially, but they are finished (and stored in the metadata) out of order, depending on the order in which copying finished. For example, supposed that the metadata contains the following records SUPERBLOCK METADATA (blocks 0 ... 250) DATA 0 DATA 1 DATA 2 ... DATA 250 Now suppose that you allocate 10 new data blocks 251-260. Suppose that copying of these blocks finish out of order (block 260 finished first and the block 251 finished last). Now, the snapshot device looks like this: SUPERBLOCK METADATA (blocks 0 ... 250, 260, 259, 258, 257, 256) DATA 0 DATA 1 DATA 2 ... DATA 250 DATA 251 DATA 252 DATA 253 DATA 254 DATA 255 METADATA (blocks 255, 254, 253, 252, 251) DATA 256 DATA 257 DATA 258 DATA 259 DATA 260 Now, if the machine crashes after writing the first metadata block but before writing the second metadata block, the space for areas DATA 250-255 is leaked, it contains no valid data and it will never be used in the future. This patch makes dm-snapshot complete exceptions in the same order they were allocated, thus fixing this bug. Note: when backporting this patch to the stable kernel, change the version field in the following way: * if version in the stable kernel is {1, 11, 1}, change it to {1, 12, 0} * if version in the stable kernel is {1, 10, 0} or {1, 10, 1}, change it to {1, 10, 2} Userspace reads the version to determine if the bug was fixed, so the version change is needed. Signed-off-by: Mikulas Patocka <mpatocka@redhat.com> Signed-off-by: Mike Snitzer <snitzer@redhat.com> Cc: stable@vger.kernel.org
2013-11-30 07:13:37 +08:00
s->exception_complete_sequence++;
complete_exception(pe);
next = rb_first(&s->out_of_order_tree);
while (next) {
pe = rb_entry(next, struct dm_snap_pending_exception,
out_of_order_node);
dm snapshot: avoid snapshot space leak on crash There is a possible leak of snapshot space in case of crash. The reason for space leaking is that chunks in the snapshot device are allocated sequentially, but they are finished (and stored in the metadata) out of order, depending on the order in which copying finished. For example, supposed that the metadata contains the following records SUPERBLOCK METADATA (blocks 0 ... 250) DATA 0 DATA 1 DATA 2 ... DATA 250 Now suppose that you allocate 10 new data blocks 251-260. Suppose that copying of these blocks finish out of order (block 260 finished first and the block 251 finished last). Now, the snapshot device looks like this: SUPERBLOCK METADATA (blocks 0 ... 250, 260, 259, 258, 257, 256) DATA 0 DATA 1 DATA 2 ... DATA 250 DATA 251 DATA 252 DATA 253 DATA 254 DATA 255 METADATA (blocks 255, 254, 253, 252, 251) DATA 256 DATA 257 DATA 258 DATA 259 DATA 260 Now, if the machine crashes after writing the first metadata block but before writing the second metadata block, the space for areas DATA 250-255 is leaked, it contains no valid data and it will never be used in the future. This patch makes dm-snapshot complete exceptions in the same order they were allocated, thus fixing this bug. Note: when backporting this patch to the stable kernel, change the version field in the following way: * if version in the stable kernel is {1, 11, 1}, change it to {1, 12, 0} * if version in the stable kernel is {1, 10, 0} or {1, 10, 1}, change it to {1, 10, 2} Userspace reads the version to determine if the bug was fixed, so the version change is needed. Signed-off-by: Mikulas Patocka <mpatocka@redhat.com> Signed-off-by: Mike Snitzer <snitzer@redhat.com> Cc: stable@vger.kernel.org
2013-11-30 07:13:37 +08:00
if (pe->exception_sequence != s->exception_complete_sequence)
break;
next = rb_next(next);
dm snapshot: avoid snapshot space leak on crash There is a possible leak of snapshot space in case of crash. The reason for space leaking is that chunks in the snapshot device are allocated sequentially, but they are finished (and stored in the metadata) out of order, depending on the order in which copying finished. For example, supposed that the metadata contains the following records SUPERBLOCK METADATA (blocks 0 ... 250) DATA 0 DATA 1 DATA 2 ... DATA 250 Now suppose that you allocate 10 new data blocks 251-260. Suppose that copying of these blocks finish out of order (block 260 finished first and the block 251 finished last). Now, the snapshot device looks like this: SUPERBLOCK METADATA (blocks 0 ... 250, 260, 259, 258, 257, 256) DATA 0 DATA 1 DATA 2 ... DATA 250 DATA 251 DATA 252 DATA 253 DATA 254 DATA 255 METADATA (blocks 255, 254, 253, 252, 251) DATA 256 DATA 257 DATA 258 DATA 259 DATA 260 Now, if the machine crashes after writing the first metadata block but before writing the second metadata block, the space for areas DATA 250-255 is leaked, it contains no valid data and it will never be used in the future. This patch makes dm-snapshot complete exceptions in the same order they were allocated, thus fixing this bug. Note: when backporting this patch to the stable kernel, change the version field in the following way: * if version in the stable kernel is {1, 11, 1}, change it to {1, 12, 0} * if version in the stable kernel is {1, 10, 0} or {1, 10, 1}, change it to {1, 10, 2} Userspace reads the version to determine if the bug was fixed, so the version change is needed. Signed-off-by: Mikulas Patocka <mpatocka@redhat.com> Signed-off-by: Mike Snitzer <snitzer@redhat.com> Cc: stable@vger.kernel.org
2013-11-30 07:13:37 +08:00
s->exception_complete_sequence++;
rb_erase(&pe->out_of_order_node, &s->out_of_order_tree);
dm snapshot: avoid snapshot space leak on crash There is a possible leak of snapshot space in case of crash. The reason for space leaking is that chunks in the snapshot device are allocated sequentially, but they are finished (and stored in the metadata) out of order, depending on the order in which copying finished. For example, supposed that the metadata contains the following records SUPERBLOCK METADATA (blocks 0 ... 250) DATA 0 DATA 1 DATA 2 ... DATA 250 Now suppose that you allocate 10 new data blocks 251-260. Suppose that copying of these blocks finish out of order (block 260 finished first and the block 251 finished last). Now, the snapshot device looks like this: SUPERBLOCK METADATA (blocks 0 ... 250, 260, 259, 258, 257, 256) DATA 0 DATA 1 DATA 2 ... DATA 250 DATA 251 DATA 252 DATA 253 DATA 254 DATA 255 METADATA (blocks 255, 254, 253, 252, 251) DATA 256 DATA 257 DATA 258 DATA 259 DATA 260 Now, if the machine crashes after writing the first metadata block but before writing the second metadata block, the space for areas DATA 250-255 is leaked, it contains no valid data and it will never be used in the future. This patch makes dm-snapshot complete exceptions in the same order they were allocated, thus fixing this bug. Note: when backporting this patch to the stable kernel, change the version field in the following way: * if version in the stable kernel is {1, 11, 1}, change it to {1, 12, 0} * if version in the stable kernel is {1, 10, 0} or {1, 10, 1}, change it to {1, 10, 2} Userspace reads the version to determine if the bug was fixed, so the version change is needed. Signed-off-by: Mikulas Patocka <mpatocka@redhat.com> Signed-off-by: Mike Snitzer <snitzer@redhat.com> Cc: stable@vger.kernel.org
2013-11-30 07:13:37 +08:00
complete_exception(pe);
cond_resched();
dm snapshot: avoid snapshot space leak on crash There is a possible leak of snapshot space in case of crash. The reason for space leaking is that chunks in the snapshot device are allocated sequentially, but they are finished (and stored in the metadata) out of order, depending on the order in which copying finished. For example, supposed that the metadata contains the following records SUPERBLOCK METADATA (blocks 0 ... 250) DATA 0 DATA 1 DATA 2 ... DATA 250 Now suppose that you allocate 10 new data blocks 251-260. Suppose that copying of these blocks finish out of order (block 260 finished first and the block 251 finished last). Now, the snapshot device looks like this: SUPERBLOCK METADATA (blocks 0 ... 250, 260, 259, 258, 257, 256) DATA 0 DATA 1 DATA 2 ... DATA 250 DATA 251 DATA 252 DATA 253 DATA 254 DATA 255 METADATA (blocks 255, 254, 253, 252, 251) DATA 256 DATA 257 DATA 258 DATA 259 DATA 260 Now, if the machine crashes after writing the first metadata block but before writing the second metadata block, the space for areas DATA 250-255 is leaked, it contains no valid data and it will never be used in the future. This patch makes dm-snapshot complete exceptions in the same order they were allocated, thus fixing this bug. Note: when backporting this patch to the stable kernel, change the version field in the following way: * if version in the stable kernel is {1, 11, 1}, change it to {1, 12, 0} * if version in the stable kernel is {1, 10, 0} or {1, 10, 1}, change it to {1, 10, 2} Userspace reads the version to determine if the bug was fixed, so the version change is needed. Signed-off-by: Mikulas Patocka <mpatocka@redhat.com> Signed-off-by: Mike Snitzer <snitzer@redhat.com> Cc: stable@vger.kernel.org
2013-11-30 07:13:37 +08:00
}
} else {
struct rb_node *parent = NULL;
struct rb_node **p = &s->out_of_order_tree.rb_node;
dm snapshot: avoid snapshot space leak on crash There is a possible leak of snapshot space in case of crash. The reason for space leaking is that chunks in the snapshot device are allocated sequentially, but they are finished (and stored in the metadata) out of order, depending on the order in which copying finished. For example, supposed that the metadata contains the following records SUPERBLOCK METADATA (blocks 0 ... 250) DATA 0 DATA 1 DATA 2 ... DATA 250 Now suppose that you allocate 10 new data blocks 251-260. Suppose that copying of these blocks finish out of order (block 260 finished first and the block 251 finished last). Now, the snapshot device looks like this: SUPERBLOCK METADATA (blocks 0 ... 250, 260, 259, 258, 257, 256) DATA 0 DATA 1 DATA 2 ... DATA 250 DATA 251 DATA 252 DATA 253 DATA 254 DATA 255 METADATA (blocks 255, 254, 253, 252, 251) DATA 256 DATA 257 DATA 258 DATA 259 DATA 260 Now, if the machine crashes after writing the first metadata block but before writing the second metadata block, the space for areas DATA 250-255 is leaked, it contains no valid data and it will never be used in the future. This patch makes dm-snapshot complete exceptions in the same order they were allocated, thus fixing this bug. Note: when backporting this patch to the stable kernel, change the version field in the following way: * if version in the stable kernel is {1, 11, 1}, change it to {1, 12, 0} * if version in the stable kernel is {1, 10, 0} or {1, 10, 1}, change it to {1, 10, 2} Userspace reads the version to determine if the bug was fixed, so the version change is needed. Signed-off-by: Mikulas Patocka <mpatocka@redhat.com> Signed-off-by: Mike Snitzer <snitzer@redhat.com> Cc: stable@vger.kernel.org
2013-11-30 07:13:37 +08:00
struct dm_snap_pending_exception *pe2;
while (*p) {
pe2 = rb_entry(*p, struct dm_snap_pending_exception, out_of_order_node);
parent = *p;
BUG_ON(pe->exception_sequence == pe2->exception_sequence);
if (pe->exception_sequence < pe2->exception_sequence)
p = &((*p)->rb_left);
else
p = &((*p)->rb_right);
dm snapshot: avoid snapshot space leak on crash There is a possible leak of snapshot space in case of crash. The reason for space leaking is that chunks in the snapshot device are allocated sequentially, but they are finished (and stored in the metadata) out of order, depending on the order in which copying finished. For example, supposed that the metadata contains the following records SUPERBLOCK METADATA (blocks 0 ... 250) DATA 0 DATA 1 DATA 2 ... DATA 250 Now suppose that you allocate 10 new data blocks 251-260. Suppose that copying of these blocks finish out of order (block 260 finished first and the block 251 finished last). Now, the snapshot device looks like this: SUPERBLOCK METADATA (blocks 0 ... 250, 260, 259, 258, 257, 256) DATA 0 DATA 1 DATA 2 ... DATA 250 DATA 251 DATA 252 DATA 253 DATA 254 DATA 255 METADATA (blocks 255, 254, 253, 252, 251) DATA 256 DATA 257 DATA 258 DATA 259 DATA 260 Now, if the machine crashes after writing the first metadata block but before writing the second metadata block, the space for areas DATA 250-255 is leaked, it contains no valid data and it will never be used in the future. This patch makes dm-snapshot complete exceptions in the same order they were allocated, thus fixing this bug. Note: when backporting this patch to the stable kernel, change the version field in the following way: * if version in the stable kernel is {1, 11, 1}, change it to {1, 12, 0} * if version in the stable kernel is {1, 10, 0} or {1, 10, 1}, change it to {1, 10, 2} Userspace reads the version to determine if the bug was fixed, so the version change is needed. Signed-off-by: Mikulas Patocka <mpatocka@redhat.com> Signed-off-by: Mike Snitzer <snitzer@redhat.com> Cc: stable@vger.kernel.org
2013-11-30 07:13:37 +08:00
}
rb_link_node(&pe->out_of_order_node, parent, p);
rb_insert_color(&pe->out_of_order_node, &s->out_of_order_tree);
dm snapshot: avoid snapshot space leak on crash There is a possible leak of snapshot space in case of crash. The reason for space leaking is that chunks in the snapshot device are allocated sequentially, but they are finished (and stored in the metadata) out of order, depending on the order in which copying finished. For example, supposed that the metadata contains the following records SUPERBLOCK METADATA (blocks 0 ... 250) DATA 0 DATA 1 DATA 2 ... DATA 250 Now suppose that you allocate 10 new data blocks 251-260. Suppose that copying of these blocks finish out of order (block 260 finished first and the block 251 finished last). Now, the snapshot device looks like this: SUPERBLOCK METADATA (blocks 0 ... 250, 260, 259, 258, 257, 256) DATA 0 DATA 1 DATA 2 ... DATA 250 DATA 251 DATA 252 DATA 253 DATA 254 DATA 255 METADATA (blocks 255, 254, 253, 252, 251) DATA 256 DATA 257 DATA 258 DATA 259 DATA 260 Now, if the machine crashes after writing the first metadata block but before writing the second metadata block, the space for areas DATA 250-255 is leaked, it contains no valid data and it will never be used in the future. This patch makes dm-snapshot complete exceptions in the same order they were allocated, thus fixing this bug. Note: when backporting this patch to the stable kernel, change the version field in the following way: * if version in the stable kernel is {1, 11, 1}, change it to {1, 12, 0} * if version in the stable kernel is {1, 10, 0} or {1, 10, 1}, change it to {1, 10, 2} Userspace reads the version to determine if the bug was fixed, so the version change is needed. Signed-off-by: Mikulas Patocka <mpatocka@redhat.com> Signed-off-by: Mike Snitzer <snitzer@redhat.com> Cc: stable@vger.kernel.org
2013-11-30 07:13:37 +08:00
}
}
/*
* Dispatches the copy operation to kcopyd.
*/
static void start_copy(struct dm_snap_pending_exception *pe)
{
struct dm_snapshot *s = pe->snap;
struct dm_io_region src, dest;
struct block_device *bdev = s->origin->bdev;
sector_t dev_size;
dev_size = get_dev_size(bdev);
src.bdev = bdev;
src.sector = chunk_to_sector(s->store, pe->e.old_chunk);
src.count = min((sector_t)s->store->chunk_size, dev_size - src.sector);
dest.bdev = s->cow->bdev;
dest.sector = chunk_to_sector(s->store, pe->e.new_chunk);
dest.count = src.count;
/* Hand over to kcopyd */
dm_kcopyd_copy(s->kcopyd_client, &src, 1, &dest, 0, copy_callback, pe);
}
static void full_bio_end_io(struct bio *bio)
{
void *callback_data = bio->bi_private;
dm_kcopyd_do_callback(callback_data, 0, bio->bi_status ? 1 : 0);
}
static void start_full_bio(struct dm_snap_pending_exception *pe,
struct bio *bio)
{
struct dm_snapshot *s = pe->snap;
void *callback_data;
pe->full_bio = bio;
pe->full_bio_end_io = bio->bi_end_io;
callback_data = dm_kcopyd_prepare_callback(s->kcopyd_client,
copy_callback, pe);
bio->bi_end_io = full_bio_end_io;
bio->bi_private = callback_data;
generic_make_request(bio);
}
static struct dm_snap_pending_exception *
__lookup_pending_exception(struct dm_snapshot *s, chunk_t chunk)
{
struct dm_exception *e = dm_lookup_exception(&s->pending, chunk);
if (!e)
return NULL;
return container_of(e, struct dm_snap_pending_exception, e);
}
/*
* Looks to see if this snapshot already has a pending exception
* for this chunk, otherwise it allocates a new one and inserts
* it into the pending table.
*
* NOTE: a write lock must be held on snap->lock before calling
* this.
*/
static struct dm_snap_pending_exception *
__find_pending_exception(struct dm_snapshot *s,
struct dm_snap_pending_exception *pe, chunk_t chunk)
{
struct dm_snap_pending_exception *pe2;
pe2 = __lookup_pending_exception(s, chunk);
if (pe2) {
free_pending_exception(pe);
return pe2;
}
pe->e.old_chunk = chunk;
bio_list_init(&pe->origin_bios);
bio_list_init(&pe->snapshot_bios);
pe->started = 0;
pe->full_bio = NULL;
if (s->store->type->prepare_exception(s->store, &pe->e)) {
free_pending_exception(pe);
return NULL;
}
dm snapshot: avoid snapshot space leak on crash There is a possible leak of snapshot space in case of crash. The reason for space leaking is that chunks in the snapshot device are allocated sequentially, but they are finished (and stored in the metadata) out of order, depending on the order in which copying finished. For example, supposed that the metadata contains the following records SUPERBLOCK METADATA (blocks 0 ... 250) DATA 0 DATA 1 DATA 2 ... DATA 250 Now suppose that you allocate 10 new data blocks 251-260. Suppose that copying of these blocks finish out of order (block 260 finished first and the block 251 finished last). Now, the snapshot device looks like this: SUPERBLOCK METADATA (blocks 0 ... 250, 260, 259, 258, 257, 256) DATA 0 DATA 1 DATA 2 ... DATA 250 DATA 251 DATA 252 DATA 253 DATA 254 DATA 255 METADATA (blocks 255, 254, 253, 252, 251) DATA 256 DATA 257 DATA 258 DATA 259 DATA 260 Now, if the machine crashes after writing the first metadata block but before writing the second metadata block, the space for areas DATA 250-255 is leaked, it contains no valid data and it will never be used in the future. This patch makes dm-snapshot complete exceptions in the same order they were allocated, thus fixing this bug. Note: when backporting this patch to the stable kernel, change the version field in the following way: * if version in the stable kernel is {1, 11, 1}, change it to {1, 12, 0} * if version in the stable kernel is {1, 10, 0} or {1, 10, 1}, change it to {1, 10, 2} Userspace reads the version to determine if the bug was fixed, so the version change is needed. Signed-off-by: Mikulas Patocka <mpatocka@redhat.com> Signed-off-by: Mike Snitzer <snitzer@redhat.com> Cc: stable@vger.kernel.org
2013-11-30 07:13:37 +08:00
pe->exception_sequence = s->exception_start_sequence++;
dm_insert_exception(&s->pending, &pe->e);
return pe;
}
static void remap_exception(struct dm_snapshot *s, struct dm_exception *e,
struct bio *bio, chunk_t chunk)
{
bio_set_dev(bio, s->cow->bdev);
block: Abstract out bvec iterator Immutable biovecs are going to require an explicit iterator. To implement immutable bvecs, a later patch is going to add a bi_bvec_done member to this struct; for now, this patch effectively just renames things. Signed-off-by: Kent Overstreet <kmo@daterainc.com> Cc: Jens Axboe <axboe@kernel.dk> Cc: Geert Uytterhoeven <geert@linux-m68k.org> Cc: Benjamin Herrenschmidt <benh@kernel.crashing.org> Cc: Paul Mackerras <paulus@samba.org> Cc: "Ed L. Cashin" <ecashin@coraid.com> Cc: Nick Piggin <npiggin@kernel.dk> Cc: Lars Ellenberg <drbd-dev@lists.linbit.com> Cc: Jiri Kosina <jkosina@suse.cz> Cc: Matthew Wilcox <willy@linux.intel.com> Cc: Geoff Levand <geoff@infradead.org> Cc: Yehuda Sadeh <yehuda@inktank.com> Cc: Sage Weil <sage@inktank.com> Cc: Alex Elder <elder@inktank.com> Cc: ceph-devel@vger.kernel.org Cc: Joshua Morris <josh.h.morris@us.ibm.com> Cc: Philip Kelleher <pjk1939@linux.vnet.ibm.com> Cc: Rusty Russell <rusty@rustcorp.com.au> Cc: "Michael S. Tsirkin" <mst@redhat.com> Cc: Konrad Rzeszutek Wilk <konrad.wilk@oracle.com> Cc: Jeremy Fitzhardinge <jeremy@goop.org> Cc: Neil Brown <neilb@suse.de> Cc: Alasdair Kergon <agk@redhat.com> Cc: Mike Snitzer <snitzer@redhat.com> Cc: dm-devel@redhat.com Cc: Martin Schwidefsky <schwidefsky@de.ibm.com> Cc: Heiko Carstens <heiko.carstens@de.ibm.com> Cc: linux390@de.ibm.com Cc: Boaz Harrosh <bharrosh@panasas.com> Cc: Benny Halevy <bhalevy@tonian.com> Cc: "James E.J. Bottomley" <JBottomley@parallels.com> Cc: Greg Kroah-Hartman <gregkh@linuxfoundation.org> Cc: "Nicholas A. Bellinger" <nab@linux-iscsi.org> Cc: Alexander Viro <viro@zeniv.linux.org.uk> Cc: Chris Mason <chris.mason@fusionio.com> Cc: "Theodore Ts'o" <tytso@mit.edu> Cc: Andreas Dilger <adilger.kernel@dilger.ca> Cc: Jaegeuk Kim <jaegeuk.kim@samsung.com> Cc: Steven Whitehouse <swhiteho@redhat.com> Cc: Dave Kleikamp <shaggy@kernel.org> Cc: Joern Engel <joern@logfs.org> Cc: Prasad Joshi <prasadjoshi.linux@gmail.com> Cc: Trond Myklebust <Trond.Myklebust@netapp.com> Cc: KONISHI Ryusuke <konishi.ryusuke@lab.ntt.co.jp> Cc: Mark Fasheh <mfasheh@suse.com> Cc: Joel Becker <jlbec@evilplan.org> Cc: Ben Myers <bpm@sgi.com> Cc: xfs@oss.sgi.com Cc: Steven Rostedt <rostedt@goodmis.org> Cc: Frederic Weisbecker <fweisbec@gmail.com> Cc: Ingo Molnar <mingo@redhat.com> Cc: Len Brown <len.brown@intel.com> Cc: Pavel Machek <pavel@ucw.cz> Cc: "Rafael J. Wysocki" <rjw@sisk.pl> Cc: Herton Ronaldo Krzesinski <herton.krzesinski@canonical.com> Cc: Ben Hutchings <ben@decadent.org.uk> Cc: Andrew Morton <akpm@linux-foundation.org> Cc: Guo Chao <yan@linux.vnet.ibm.com> Cc: Tejun Heo <tj@kernel.org> Cc: Asai Thambi S P <asamymuthupa@micron.com> Cc: Selvan Mani <smani@micron.com> Cc: Sam Bradshaw <sbradshaw@micron.com> Cc: Wei Yongjun <yongjun_wei@trendmicro.com.cn> Cc: "Roger Pau Monné" <roger.pau@citrix.com> Cc: Jan Beulich <jbeulich@suse.com> Cc: Stefano Stabellini <stefano.stabellini@eu.citrix.com> Cc: Ian Campbell <Ian.Campbell@citrix.com> Cc: Sebastian Ott <sebott@linux.vnet.ibm.com> Cc: Christian Borntraeger <borntraeger@de.ibm.com> Cc: Minchan Kim <minchan@kernel.org> Cc: Jiang Liu <jiang.liu@huawei.com> Cc: Nitin Gupta <ngupta@vflare.org> Cc: Jerome Marchand <jmarchand@redhat.com> Cc: Joe Perches <joe@perches.com> Cc: Peng Tao <tao.peng@emc.com> Cc: Andy Adamson <andros@netapp.com> Cc: fanchaoting <fanchaoting@cn.fujitsu.com> Cc: Jie Liu <jeff.liu@oracle.com> Cc: Sunil Mushran <sunil.mushran@gmail.com> Cc: "Martin K. Petersen" <martin.petersen@oracle.com> Cc: Namjae Jeon <namjae.jeon@samsung.com> Cc: Pankaj Kumar <pankaj.km@samsung.com> Cc: Dan Magenheimer <dan.magenheimer@oracle.com> Cc: Mel Gorman <mgorman@suse.de>6
2013-10-12 06:44:27 +08:00
bio->bi_iter.bi_sector =
chunk_to_sector(s->store, dm_chunk_number(e->new_chunk) +
(chunk - e->old_chunk)) +
(bio->bi_iter.bi_sector & s->store->chunk_mask);
}
static int snapshot_map(struct dm_target *ti, struct bio *bio)
{
struct dm_exception *e;
struct dm_snapshot *s = ti->private;
int r = DM_MAPIO_REMAPPED;
chunk_t chunk;
struct dm_snap_pending_exception *pe = NULL;
init_tracked_chunk(bio);
if (bio->bi_opf & REQ_PREFLUSH) {
bio_set_dev(bio, s->cow->bdev);
return DM_MAPIO_REMAPPED;
}
block: Abstract out bvec iterator Immutable biovecs are going to require an explicit iterator. To implement immutable bvecs, a later patch is going to add a bi_bvec_done member to this struct; for now, this patch effectively just renames things. Signed-off-by: Kent Overstreet <kmo@daterainc.com> Cc: Jens Axboe <axboe@kernel.dk> Cc: Geert Uytterhoeven <geert@linux-m68k.org> Cc: Benjamin Herrenschmidt <benh@kernel.crashing.org> Cc: Paul Mackerras <paulus@samba.org> Cc: "Ed L. Cashin" <ecashin@coraid.com> Cc: Nick Piggin <npiggin@kernel.dk> Cc: Lars Ellenberg <drbd-dev@lists.linbit.com> Cc: Jiri Kosina <jkosina@suse.cz> Cc: Matthew Wilcox <willy@linux.intel.com> Cc: Geoff Levand <geoff@infradead.org> Cc: Yehuda Sadeh <yehuda@inktank.com> Cc: Sage Weil <sage@inktank.com> Cc: Alex Elder <elder@inktank.com> Cc: ceph-devel@vger.kernel.org Cc: Joshua Morris <josh.h.morris@us.ibm.com> Cc: Philip Kelleher <pjk1939@linux.vnet.ibm.com> Cc: Rusty Russell <rusty@rustcorp.com.au> Cc: "Michael S. Tsirkin" <mst@redhat.com> Cc: Konrad Rzeszutek Wilk <konrad.wilk@oracle.com> Cc: Jeremy Fitzhardinge <jeremy@goop.org> Cc: Neil Brown <neilb@suse.de> Cc: Alasdair Kergon <agk@redhat.com> Cc: Mike Snitzer <snitzer@redhat.com> Cc: dm-devel@redhat.com Cc: Martin Schwidefsky <schwidefsky@de.ibm.com> Cc: Heiko Carstens <heiko.carstens@de.ibm.com> Cc: linux390@de.ibm.com Cc: Boaz Harrosh <bharrosh@panasas.com> Cc: Benny Halevy <bhalevy@tonian.com> Cc: "James E.J. Bottomley" <JBottomley@parallels.com> Cc: Greg Kroah-Hartman <gregkh@linuxfoundation.org> Cc: "Nicholas A. Bellinger" <nab@linux-iscsi.org> Cc: Alexander Viro <viro@zeniv.linux.org.uk> Cc: Chris Mason <chris.mason@fusionio.com> Cc: "Theodore Ts'o" <tytso@mit.edu> Cc: Andreas Dilger <adilger.kernel@dilger.ca> Cc: Jaegeuk Kim <jaegeuk.kim@samsung.com> Cc: Steven Whitehouse <swhiteho@redhat.com> Cc: Dave Kleikamp <shaggy@kernel.org> Cc: Joern Engel <joern@logfs.org> Cc: Prasad Joshi <prasadjoshi.linux@gmail.com> Cc: Trond Myklebust <Trond.Myklebust@netapp.com> Cc: KONISHI Ryusuke <konishi.ryusuke@lab.ntt.co.jp> Cc: Mark Fasheh <mfasheh@suse.com> Cc: Joel Becker <jlbec@evilplan.org> Cc: Ben Myers <bpm@sgi.com> Cc: xfs@oss.sgi.com Cc: Steven Rostedt <rostedt@goodmis.org> Cc: Frederic Weisbecker <fweisbec@gmail.com> Cc: Ingo Molnar <mingo@redhat.com> Cc: Len Brown <len.brown@intel.com> Cc: Pavel Machek <pavel@ucw.cz> Cc: "Rafael J. Wysocki" <rjw@sisk.pl> Cc: Herton Ronaldo Krzesinski <herton.krzesinski@canonical.com> Cc: Ben Hutchings <ben@decadent.org.uk> Cc: Andrew Morton <akpm@linux-foundation.org> Cc: Guo Chao <yan@linux.vnet.ibm.com> Cc: Tejun Heo <tj@kernel.org> Cc: Asai Thambi S P <asamymuthupa@micron.com> Cc: Selvan Mani <smani@micron.com> Cc: Sam Bradshaw <sbradshaw@micron.com> Cc: Wei Yongjun <yongjun_wei@trendmicro.com.cn> Cc: "Roger Pau Monné" <roger.pau@citrix.com> Cc: Jan Beulich <jbeulich@suse.com> Cc: Stefano Stabellini <stefano.stabellini@eu.citrix.com> Cc: Ian Campbell <Ian.Campbell@citrix.com> Cc: Sebastian Ott <sebott@linux.vnet.ibm.com> Cc: Christian Borntraeger <borntraeger@de.ibm.com> Cc: Minchan Kim <minchan@kernel.org> Cc: Jiang Liu <jiang.liu@huawei.com> Cc: Nitin Gupta <ngupta@vflare.org> Cc: Jerome Marchand <jmarchand@redhat.com> Cc: Joe Perches <joe@perches.com> Cc: Peng Tao <tao.peng@emc.com> Cc: Andy Adamson <andros@netapp.com> Cc: fanchaoting <fanchaoting@cn.fujitsu.com> Cc: Jie Liu <jeff.liu@oracle.com> Cc: Sunil Mushran <sunil.mushran@gmail.com> Cc: "Martin K. Petersen" <martin.petersen@oracle.com> Cc: Namjae Jeon <namjae.jeon@samsung.com> Cc: Pankaj Kumar <pankaj.km@samsung.com> Cc: Dan Magenheimer <dan.magenheimer@oracle.com> Cc: Mel Gorman <mgorman@suse.de>6
2013-10-12 06:44:27 +08:00
chunk = sector_to_chunk(s->store, bio->bi_iter.bi_sector);
/* Full snapshots are not usable */
/* To get here the table must be live so s->active is always set. */
if (!s->valid)
return DM_MAPIO_KILL;
mutex_lock(&s->lock);
if (!s->valid || (unlikely(s->snapshot_overflowed) &&
bio_data_dir(bio) == WRITE)) {
r = DM_MAPIO_KILL;
goto out_unlock;
}
/* If the block is already remapped - use that, else remap it */
e = dm_lookup_exception(&s->complete, chunk);
if (e) {
remap_exception(s, e, bio, chunk);
goto out_unlock;
}
/*
* Write to snapshot - higher level takes care of RW/RO
* flags so we should only get this if we are
* writeable.
*/
if (bio_data_dir(bio) == WRITE) {
pe = __lookup_pending_exception(s, chunk);
if (!pe) {
mutex_unlock(&s->lock);
pe = alloc_pending_exception(s);
mutex_lock(&s->lock);
if (!s->valid || s->snapshot_overflowed) {
free_pending_exception(pe);
r = DM_MAPIO_KILL;
goto out_unlock;
}
e = dm_lookup_exception(&s->complete, chunk);
if (e) {
free_pending_exception(pe);
remap_exception(s, e, bio, chunk);
goto out_unlock;
}
pe = __find_pending_exception(s, pe, chunk);
if (!pe) {
if (s->store->userspace_supports_overflow) {
s->snapshot_overflowed = 1;
DMERR("Snapshot overflowed: Unable to allocate exception.");
} else
__invalidate_snapshot(s, -ENOMEM);
r = DM_MAPIO_KILL;
goto out_unlock;
}
}
remap_exception(s, &pe->e, bio, chunk);
r = DM_MAPIO_SUBMITTED;
if (!pe->started &&
block: Abstract out bvec iterator Immutable biovecs are going to require an explicit iterator. To implement immutable bvecs, a later patch is going to add a bi_bvec_done member to this struct; for now, this patch effectively just renames things. Signed-off-by: Kent Overstreet <kmo@daterainc.com> Cc: Jens Axboe <axboe@kernel.dk> Cc: Geert Uytterhoeven <geert@linux-m68k.org> Cc: Benjamin Herrenschmidt <benh@kernel.crashing.org> Cc: Paul Mackerras <paulus@samba.org> Cc: "Ed L. Cashin" <ecashin@coraid.com> Cc: Nick Piggin <npiggin@kernel.dk> Cc: Lars Ellenberg <drbd-dev@lists.linbit.com> Cc: Jiri Kosina <jkosina@suse.cz> Cc: Matthew Wilcox <willy@linux.intel.com> Cc: Geoff Levand <geoff@infradead.org> Cc: Yehuda Sadeh <yehuda@inktank.com> Cc: Sage Weil <sage@inktank.com> Cc: Alex Elder <elder@inktank.com> Cc: ceph-devel@vger.kernel.org Cc: Joshua Morris <josh.h.morris@us.ibm.com> Cc: Philip Kelleher <pjk1939@linux.vnet.ibm.com> Cc: Rusty Russell <rusty@rustcorp.com.au> Cc: "Michael S. Tsirkin" <mst@redhat.com> Cc: Konrad Rzeszutek Wilk <konrad.wilk@oracle.com> Cc: Jeremy Fitzhardinge <jeremy@goop.org> Cc: Neil Brown <neilb@suse.de> Cc: Alasdair Kergon <agk@redhat.com> Cc: Mike Snitzer <snitzer@redhat.com> Cc: dm-devel@redhat.com Cc: Martin Schwidefsky <schwidefsky@de.ibm.com> Cc: Heiko Carstens <heiko.carstens@de.ibm.com> Cc: linux390@de.ibm.com Cc: Boaz Harrosh <bharrosh@panasas.com> Cc: Benny Halevy <bhalevy@tonian.com> Cc: "James E.J. Bottomley" <JBottomley@parallels.com> Cc: Greg Kroah-Hartman <gregkh@linuxfoundation.org> Cc: "Nicholas A. Bellinger" <nab@linux-iscsi.org> Cc: Alexander Viro <viro@zeniv.linux.org.uk> Cc: Chris Mason <chris.mason@fusionio.com> Cc: "Theodore Ts'o" <tytso@mit.edu> Cc: Andreas Dilger <adilger.kernel@dilger.ca> Cc: Jaegeuk Kim <jaegeuk.kim@samsung.com> Cc: Steven Whitehouse <swhiteho@redhat.com> Cc: Dave Kleikamp <shaggy@kernel.org> Cc: Joern Engel <joern@logfs.org> Cc: Prasad Joshi <prasadjoshi.linux@gmail.com> Cc: Trond Myklebust <Trond.Myklebust@netapp.com> Cc: KONISHI Ryusuke <konishi.ryusuke@lab.ntt.co.jp> Cc: Mark Fasheh <mfasheh@suse.com> Cc: Joel Becker <jlbec@evilplan.org> Cc: Ben Myers <bpm@sgi.com> Cc: xfs@oss.sgi.com Cc: Steven Rostedt <rostedt@goodmis.org> Cc: Frederic Weisbecker <fweisbec@gmail.com> Cc: Ingo Molnar <mingo@redhat.com> Cc: Len Brown <len.brown@intel.com> Cc: Pavel Machek <pavel@ucw.cz> Cc: "Rafael J. Wysocki" <rjw@sisk.pl> Cc: Herton Ronaldo Krzesinski <herton.krzesinski@canonical.com> Cc: Ben Hutchings <ben@decadent.org.uk> Cc: Andrew Morton <akpm@linux-foundation.org> Cc: Guo Chao <yan@linux.vnet.ibm.com> Cc: Tejun Heo <tj@kernel.org> Cc: Asai Thambi S P <asamymuthupa@micron.com> Cc: Selvan Mani <smani@micron.com> Cc: Sam Bradshaw <sbradshaw@micron.com> Cc: Wei Yongjun <yongjun_wei@trendmicro.com.cn> Cc: "Roger Pau Monné" <roger.pau@citrix.com> Cc: Jan Beulich <jbeulich@suse.com> Cc: Stefano Stabellini <stefano.stabellini@eu.citrix.com> Cc: Ian Campbell <Ian.Campbell@citrix.com> Cc: Sebastian Ott <sebott@linux.vnet.ibm.com> Cc: Christian Borntraeger <borntraeger@de.ibm.com> Cc: Minchan Kim <minchan@kernel.org> Cc: Jiang Liu <jiang.liu@huawei.com> Cc: Nitin Gupta <ngupta@vflare.org> Cc: Jerome Marchand <jmarchand@redhat.com> Cc: Joe Perches <joe@perches.com> Cc: Peng Tao <tao.peng@emc.com> Cc: Andy Adamson <andros@netapp.com> Cc: fanchaoting <fanchaoting@cn.fujitsu.com> Cc: Jie Liu <jeff.liu@oracle.com> Cc: Sunil Mushran <sunil.mushran@gmail.com> Cc: "Martin K. Petersen" <martin.petersen@oracle.com> Cc: Namjae Jeon <namjae.jeon@samsung.com> Cc: Pankaj Kumar <pankaj.km@samsung.com> Cc: Dan Magenheimer <dan.magenheimer@oracle.com> Cc: Mel Gorman <mgorman@suse.de>6
2013-10-12 06:44:27 +08:00
bio->bi_iter.bi_size ==
(s->store->chunk_size << SECTOR_SHIFT)) {
pe->started = 1;
mutex_unlock(&s->lock);
start_full_bio(pe, bio);
goto out;
}
bio_list_add(&pe->snapshot_bios, bio);
if (!pe->started) {
/* this is protected by snap->lock */
pe->started = 1;
mutex_unlock(&s->lock);
start_copy(pe);
goto out;
}
} else {
bio_set_dev(bio, s->origin->bdev);
track_chunk(s, bio, chunk);
}
out_unlock:
mutex_unlock(&s->lock);
out:
return r;
}
/*
* A snapshot-merge target behaves like a combination of a snapshot
* target and a snapshot-origin target. It only generates new
* exceptions in other snapshots and not in the one that is being
* merged.
*
* For each chunk, if there is an existing exception, it is used to
* redirect I/O to the cow device. Otherwise I/O is sent to the origin,
* which in turn might generate exceptions in other snapshots.
* If merging is currently taking place on the chunk in question, the
* I/O is deferred by adding it to s->bios_queued_during_merge.
*/
static int snapshot_merge_map(struct dm_target *ti, struct bio *bio)
{
struct dm_exception *e;
struct dm_snapshot *s = ti->private;
int r = DM_MAPIO_REMAPPED;
chunk_t chunk;
init_tracked_chunk(bio);
if (bio->bi_opf & REQ_PREFLUSH) {
if (!dm_bio_get_target_bio_nr(bio))
bio_set_dev(bio, s->origin->bdev);
else
bio_set_dev(bio, s->cow->bdev);
return DM_MAPIO_REMAPPED;
}
block: Abstract out bvec iterator Immutable biovecs are going to require an explicit iterator. To implement immutable bvecs, a later patch is going to add a bi_bvec_done member to this struct; for now, this patch effectively just renames things. Signed-off-by: Kent Overstreet <kmo@daterainc.com> Cc: Jens Axboe <axboe@kernel.dk> Cc: Geert Uytterhoeven <geert@linux-m68k.org> Cc: Benjamin Herrenschmidt <benh@kernel.crashing.org> Cc: Paul Mackerras <paulus@samba.org> Cc: "Ed L. Cashin" <ecashin@coraid.com> Cc: Nick Piggin <npiggin@kernel.dk> Cc: Lars Ellenberg <drbd-dev@lists.linbit.com> Cc: Jiri Kosina <jkosina@suse.cz> Cc: Matthew Wilcox <willy@linux.intel.com> Cc: Geoff Levand <geoff@infradead.org> Cc: Yehuda Sadeh <yehuda@inktank.com> Cc: Sage Weil <sage@inktank.com> Cc: Alex Elder <elder@inktank.com> Cc: ceph-devel@vger.kernel.org Cc: Joshua Morris <josh.h.morris@us.ibm.com> Cc: Philip Kelleher <pjk1939@linux.vnet.ibm.com> Cc: Rusty Russell <rusty@rustcorp.com.au> Cc: "Michael S. Tsirkin" <mst@redhat.com> Cc: Konrad Rzeszutek Wilk <konrad.wilk@oracle.com> Cc: Jeremy Fitzhardinge <jeremy@goop.org> Cc: Neil Brown <neilb@suse.de> Cc: Alasdair Kergon <agk@redhat.com> Cc: Mike Snitzer <snitzer@redhat.com> Cc: dm-devel@redhat.com Cc: Martin Schwidefsky <schwidefsky@de.ibm.com> Cc: Heiko Carstens <heiko.carstens@de.ibm.com> Cc: linux390@de.ibm.com Cc: Boaz Harrosh <bharrosh@panasas.com> Cc: Benny Halevy <bhalevy@tonian.com> Cc: "James E.J. Bottomley" <JBottomley@parallels.com> Cc: Greg Kroah-Hartman <gregkh@linuxfoundation.org> Cc: "Nicholas A. Bellinger" <nab@linux-iscsi.org> Cc: Alexander Viro <viro@zeniv.linux.org.uk> Cc: Chris Mason <chris.mason@fusionio.com> Cc: "Theodore Ts'o" <tytso@mit.edu> Cc: Andreas Dilger <adilger.kernel@dilger.ca> Cc: Jaegeuk Kim <jaegeuk.kim@samsung.com> Cc: Steven Whitehouse <swhiteho@redhat.com> Cc: Dave Kleikamp <shaggy@kernel.org> Cc: Joern Engel <joern@logfs.org> Cc: Prasad Joshi <prasadjoshi.linux@gmail.com> Cc: Trond Myklebust <Trond.Myklebust@netapp.com> Cc: KONISHI Ryusuke <konishi.ryusuke@lab.ntt.co.jp> Cc: Mark Fasheh <mfasheh@suse.com> Cc: Joel Becker <jlbec@evilplan.org> Cc: Ben Myers <bpm@sgi.com> Cc: xfs@oss.sgi.com Cc: Steven Rostedt <rostedt@goodmis.org> Cc: Frederic Weisbecker <fweisbec@gmail.com> Cc: Ingo Molnar <mingo@redhat.com> Cc: Len Brown <len.brown@intel.com> Cc: Pavel Machek <pavel@ucw.cz> Cc: "Rafael J. Wysocki" <rjw@sisk.pl> Cc: Herton Ronaldo Krzesinski <herton.krzesinski@canonical.com> Cc: Ben Hutchings <ben@decadent.org.uk> Cc: Andrew Morton <akpm@linux-foundation.org> Cc: Guo Chao <yan@linux.vnet.ibm.com> Cc: Tejun Heo <tj@kernel.org> Cc: Asai Thambi S P <asamymuthupa@micron.com> Cc: Selvan Mani <smani@micron.com> Cc: Sam Bradshaw <sbradshaw@micron.com> Cc: Wei Yongjun <yongjun_wei@trendmicro.com.cn> Cc: "Roger Pau Monné" <roger.pau@citrix.com> Cc: Jan Beulich <jbeulich@suse.com> Cc: Stefano Stabellini <stefano.stabellini@eu.citrix.com> Cc: Ian Campbell <Ian.Campbell@citrix.com> Cc: Sebastian Ott <sebott@linux.vnet.ibm.com> Cc: Christian Borntraeger <borntraeger@de.ibm.com> Cc: Minchan Kim <minchan@kernel.org> Cc: Jiang Liu <jiang.liu@huawei.com> Cc: Nitin Gupta <ngupta@vflare.org> Cc: Jerome Marchand <jmarchand@redhat.com> Cc: Joe Perches <joe@perches.com> Cc: Peng Tao <tao.peng@emc.com> Cc: Andy Adamson <andros@netapp.com> Cc: fanchaoting <fanchaoting@cn.fujitsu.com> Cc: Jie Liu <jeff.liu@oracle.com> Cc: Sunil Mushran <sunil.mushran@gmail.com> Cc: "Martin K. Petersen" <martin.petersen@oracle.com> Cc: Namjae Jeon <namjae.jeon@samsung.com> Cc: Pankaj Kumar <pankaj.km@samsung.com> Cc: Dan Magenheimer <dan.magenheimer@oracle.com> Cc: Mel Gorman <mgorman@suse.de>6
2013-10-12 06:44:27 +08:00
chunk = sector_to_chunk(s->store, bio->bi_iter.bi_sector);
mutex_lock(&s->lock);
/* Full merging snapshots are redirected to the origin */
if (!s->valid)
goto redirect_to_origin;
/* If the block is already remapped - use that */
e = dm_lookup_exception(&s->complete, chunk);
if (e) {
/* Queue writes overlapping with chunks being merged */
if (bio_data_dir(bio) == WRITE &&
chunk >= s->first_merging_chunk &&
chunk < (s->first_merging_chunk +
s->num_merging_chunks)) {
bio_set_dev(bio, s->origin->bdev);
bio_list_add(&s->bios_queued_during_merge, bio);
r = DM_MAPIO_SUBMITTED;
goto out_unlock;
}
remap_exception(s, e, bio, chunk);
if (bio_data_dir(bio) == WRITE)
track_chunk(s, bio, chunk);
goto out_unlock;
}
redirect_to_origin:
bio_set_dev(bio, s->origin->bdev);
if (bio_data_dir(bio) == WRITE) {
mutex_unlock(&s->lock);
return do_origin(s->origin, bio);
}
out_unlock:
mutex_unlock(&s->lock);
return r;
}
static int snapshot_end_io(struct dm_target *ti, struct bio *bio,
blk_status_t *error)
{
struct dm_snapshot *s = ti->private;
if (is_bio_tracked(bio))
stop_tracking_chunk(s, bio);
return DM_ENDIO_DONE;
}
static void snapshot_merge_presuspend(struct dm_target *ti)
{
struct dm_snapshot *s = ti->private;
stop_merge(s);
}
static int snapshot_preresume(struct dm_target *ti)
{
int r = 0;
struct dm_snapshot *s = ti->private;
struct dm_snapshot *snap_src = NULL, *snap_dest = NULL;
down_read(&_origins_lock);
(void) __find_snapshots_sharing_cow(s, &snap_src, &snap_dest, NULL);
if (snap_src && snap_dest) {
mutex_lock(&snap_src->lock);
if (s == snap_src) {
DMERR("Unable to resume snapshot source until "
"handover completes.");
r = -EINVAL;
} else if (!dm_suspended(snap_src->ti)) {
DMERR("Unable to perform snapshot handover until "
"source is suspended.");
r = -EINVAL;
}
mutex_unlock(&snap_src->lock);
}
up_read(&_origins_lock);
return r;
}
static void snapshot_resume(struct dm_target *ti)
{
struct dm_snapshot *s = ti->private;
struct dm_snapshot *snap_src = NULL, *snap_dest = NULL, *snap_merging = NULL;
dm snapshot: suspend origin when doing exception handover In the function snapshot_resume we perform exception store handover. If there is another active snapshot target, the exception store is moved from this target to the target that is being resumed. The problem is that if there is some pending exception, it will point to an incorrect exception store after that handover, causing a crash due to dm-snap-persistent.c:get_exception()'s BUG_ON. This bug can be triggered by repeatedly changing snapshot permissions with "lvchange -p r" and "lvchange -p rw" while there are writes on the associated origin device. To fix this bug, we must suspend the origin device when doing the exception store handover to make sure that there are no pending exceptions: - introduce _origin_hash that keeps track of dm_origin structures. - introduce functions __lookup_dm_origin, __insert_dm_origin and __remove_dm_origin that manipulate the origin hash. - modify snapshot_resume so that it calls dm_internal_suspend_fast() and dm_internal_resume_fast() on the origin device. NOTE to stable@ people: When backporting to kernels 3.12-3.18, use dm_internal_suspend and dm_internal_resume instead of dm_internal_suspend_fast and dm_internal_resume_fast. When backporting to kernels older than 3.12, you need to pick functions dm_internal_suspend and dm_internal_resume from the commit fd2ed4d252701d3bbed4cd3e3d267ad469bb832a. Signed-off-by: Mikulas Patocka <mpatocka@redhat.com> Signed-off-by: Mike Snitzer <snitzer@redhat.com> Cc: stable@vger.kernel.org
2015-02-27 00:40:35 +08:00
struct dm_origin *o;
struct mapped_device *origin_md = NULL;
bool must_restart_merging = false;
down_read(&_origins_lock);
dm snapshot: suspend origin when doing exception handover In the function snapshot_resume we perform exception store handover. If there is another active snapshot target, the exception store is moved from this target to the target that is being resumed. The problem is that if there is some pending exception, it will point to an incorrect exception store after that handover, causing a crash due to dm-snap-persistent.c:get_exception()'s BUG_ON. This bug can be triggered by repeatedly changing snapshot permissions with "lvchange -p r" and "lvchange -p rw" while there are writes on the associated origin device. To fix this bug, we must suspend the origin device when doing the exception store handover to make sure that there are no pending exceptions: - introduce _origin_hash that keeps track of dm_origin structures. - introduce functions __lookup_dm_origin, __insert_dm_origin and __remove_dm_origin that manipulate the origin hash. - modify snapshot_resume so that it calls dm_internal_suspend_fast() and dm_internal_resume_fast() on the origin device. NOTE to stable@ people: When backporting to kernels 3.12-3.18, use dm_internal_suspend and dm_internal_resume instead of dm_internal_suspend_fast and dm_internal_resume_fast. When backporting to kernels older than 3.12, you need to pick functions dm_internal_suspend and dm_internal_resume from the commit fd2ed4d252701d3bbed4cd3e3d267ad469bb832a. Signed-off-by: Mikulas Patocka <mpatocka@redhat.com> Signed-off-by: Mike Snitzer <snitzer@redhat.com> Cc: stable@vger.kernel.org
2015-02-27 00:40:35 +08:00
o = __lookup_dm_origin(s->origin->bdev);
if (o)
origin_md = dm_table_get_md(o->ti->table);
if (!origin_md) {
(void) __find_snapshots_sharing_cow(s, NULL, NULL, &snap_merging);
if (snap_merging)
origin_md = dm_table_get_md(snap_merging->ti->table);
}
dm snapshot: suspend origin when doing exception handover In the function snapshot_resume we perform exception store handover. If there is another active snapshot target, the exception store is moved from this target to the target that is being resumed. The problem is that if there is some pending exception, it will point to an incorrect exception store after that handover, causing a crash due to dm-snap-persistent.c:get_exception()'s BUG_ON. This bug can be triggered by repeatedly changing snapshot permissions with "lvchange -p r" and "lvchange -p rw" while there are writes on the associated origin device. To fix this bug, we must suspend the origin device when doing the exception store handover to make sure that there are no pending exceptions: - introduce _origin_hash that keeps track of dm_origin structures. - introduce functions __lookup_dm_origin, __insert_dm_origin and __remove_dm_origin that manipulate the origin hash. - modify snapshot_resume so that it calls dm_internal_suspend_fast() and dm_internal_resume_fast() on the origin device. NOTE to stable@ people: When backporting to kernels 3.12-3.18, use dm_internal_suspend and dm_internal_resume instead of dm_internal_suspend_fast and dm_internal_resume_fast. When backporting to kernels older than 3.12, you need to pick functions dm_internal_suspend and dm_internal_resume from the commit fd2ed4d252701d3bbed4cd3e3d267ad469bb832a. Signed-off-by: Mikulas Patocka <mpatocka@redhat.com> Signed-off-by: Mike Snitzer <snitzer@redhat.com> Cc: stable@vger.kernel.org
2015-02-27 00:40:35 +08:00
if (origin_md == dm_table_get_md(ti->table))
origin_md = NULL;
if (origin_md) {
if (dm_hold(origin_md))
origin_md = NULL;
}
dm snapshot: suspend origin when doing exception handover In the function snapshot_resume we perform exception store handover. If there is another active snapshot target, the exception store is moved from this target to the target that is being resumed. The problem is that if there is some pending exception, it will point to an incorrect exception store after that handover, causing a crash due to dm-snap-persistent.c:get_exception()'s BUG_ON. This bug can be triggered by repeatedly changing snapshot permissions with "lvchange -p r" and "lvchange -p rw" while there are writes on the associated origin device. To fix this bug, we must suspend the origin device when doing the exception store handover to make sure that there are no pending exceptions: - introduce _origin_hash that keeps track of dm_origin structures. - introduce functions __lookup_dm_origin, __insert_dm_origin and __remove_dm_origin that manipulate the origin hash. - modify snapshot_resume so that it calls dm_internal_suspend_fast() and dm_internal_resume_fast() on the origin device. NOTE to stable@ people: When backporting to kernels 3.12-3.18, use dm_internal_suspend and dm_internal_resume instead of dm_internal_suspend_fast and dm_internal_resume_fast. When backporting to kernels older than 3.12, you need to pick functions dm_internal_suspend and dm_internal_resume from the commit fd2ed4d252701d3bbed4cd3e3d267ad469bb832a. Signed-off-by: Mikulas Patocka <mpatocka@redhat.com> Signed-off-by: Mike Snitzer <snitzer@redhat.com> Cc: stable@vger.kernel.org
2015-02-27 00:40:35 +08:00
up_read(&_origins_lock);
if (origin_md) {
dm snapshot: suspend origin when doing exception handover In the function snapshot_resume we perform exception store handover. If there is another active snapshot target, the exception store is moved from this target to the target that is being resumed. The problem is that if there is some pending exception, it will point to an incorrect exception store after that handover, causing a crash due to dm-snap-persistent.c:get_exception()'s BUG_ON. This bug can be triggered by repeatedly changing snapshot permissions with "lvchange -p r" and "lvchange -p rw" while there are writes on the associated origin device. To fix this bug, we must suspend the origin device when doing the exception store handover to make sure that there are no pending exceptions: - introduce _origin_hash that keeps track of dm_origin structures. - introduce functions __lookup_dm_origin, __insert_dm_origin and __remove_dm_origin that manipulate the origin hash. - modify snapshot_resume so that it calls dm_internal_suspend_fast() and dm_internal_resume_fast() on the origin device. NOTE to stable@ people: When backporting to kernels 3.12-3.18, use dm_internal_suspend and dm_internal_resume instead of dm_internal_suspend_fast and dm_internal_resume_fast. When backporting to kernels older than 3.12, you need to pick functions dm_internal_suspend and dm_internal_resume from the commit fd2ed4d252701d3bbed4cd3e3d267ad469bb832a. Signed-off-by: Mikulas Patocka <mpatocka@redhat.com> Signed-off-by: Mike Snitzer <snitzer@redhat.com> Cc: stable@vger.kernel.org
2015-02-27 00:40:35 +08:00
dm_internal_suspend_fast(origin_md);
if (snap_merging && test_bit(RUNNING_MERGE, &snap_merging->state_bits)) {
must_restart_merging = true;
stop_merge(snap_merging);
}
}
down_read(&_origins_lock);
dm snapshot: suspend origin when doing exception handover In the function snapshot_resume we perform exception store handover. If there is another active snapshot target, the exception store is moved from this target to the target that is being resumed. The problem is that if there is some pending exception, it will point to an incorrect exception store after that handover, causing a crash due to dm-snap-persistent.c:get_exception()'s BUG_ON. This bug can be triggered by repeatedly changing snapshot permissions with "lvchange -p r" and "lvchange -p rw" while there are writes on the associated origin device. To fix this bug, we must suspend the origin device when doing the exception store handover to make sure that there are no pending exceptions: - introduce _origin_hash that keeps track of dm_origin structures. - introduce functions __lookup_dm_origin, __insert_dm_origin and __remove_dm_origin that manipulate the origin hash. - modify snapshot_resume so that it calls dm_internal_suspend_fast() and dm_internal_resume_fast() on the origin device. NOTE to stable@ people: When backporting to kernels 3.12-3.18, use dm_internal_suspend and dm_internal_resume instead of dm_internal_suspend_fast and dm_internal_resume_fast. When backporting to kernels older than 3.12, you need to pick functions dm_internal_suspend and dm_internal_resume from the commit fd2ed4d252701d3bbed4cd3e3d267ad469bb832a. Signed-off-by: Mikulas Patocka <mpatocka@redhat.com> Signed-off-by: Mike Snitzer <snitzer@redhat.com> Cc: stable@vger.kernel.org
2015-02-27 00:40:35 +08:00
(void) __find_snapshots_sharing_cow(s, &snap_src, &snap_dest, NULL);
if (snap_src && snap_dest) {
mutex_lock(&snap_src->lock);
mutex_lock_nested(&snap_dest->lock, SINGLE_DEPTH_NESTING);
__handover_exceptions(snap_src, snap_dest);
mutex_unlock(&snap_dest->lock);
mutex_unlock(&snap_src->lock);
}
dm snapshot: suspend origin when doing exception handover In the function snapshot_resume we perform exception store handover. If there is another active snapshot target, the exception store is moved from this target to the target that is being resumed. The problem is that if there is some pending exception, it will point to an incorrect exception store after that handover, causing a crash due to dm-snap-persistent.c:get_exception()'s BUG_ON. This bug can be triggered by repeatedly changing snapshot permissions with "lvchange -p r" and "lvchange -p rw" while there are writes on the associated origin device. To fix this bug, we must suspend the origin device when doing the exception store handover to make sure that there are no pending exceptions: - introduce _origin_hash that keeps track of dm_origin structures. - introduce functions __lookup_dm_origin, __insert_dm_origin and __remove_dm_origin that manipulate the origin hash. - modify snapshot_resume so that it calls dm_internal_suspend_fast() and dm_internal_resume_fast() on the origin device. NOTE to stable@ people: When backporting to kernels 3.12-3.18, use dm_internal_suspend and dm_internal_resume instead of dm_internal_suspend_fast and dm_internal_resume_fast. When backporting to kernels older than 3.12, you need to pick functions dm_internal_suspend and dm_internal_resume from the commit fd2ed4d252701d3bbed4cd3e3d267ad469bb832a. Signed-off-by: Mikulas Patocka <mpatocka@redhat.com> Signed-off-by: Mike Snitzer <snitzer@redhat.com> Cc: stable@vger.kernel.org
2015-02-27 00:40:35 +08:00
up_read(&_origins_lock);
if (origin_md) {
if (must_restart_merging)
start_merge(snap_merging);
dm_internal_resume_fast(origin_md);
dm_put(origin_md);
}
/* Now we have correct chunk size, reregister */
reregister_snapshot(s);
mutex_lock(&s->lock);
s->active = 1;
mutex_unlock(&s->lock);
}
static uint32_t get_origin_minimum_chunksize(struct block_device *bdev)
{
uint32_t min_chunksize;
down_read(&_origins_lock);
min_chunksize = __minimum_chunk_size(__lookup_origin(bdev));
up_read(&_origins_lock);
return min_chunksize;
}
static void snapshot_merge_resume(struct dm_target *ti)
{
struct dm_snapshot *s = ti->private;
/*
* Handover exceptions from existing snapshot.
*/
snapshot_resume(ti);
/*
* snapshot-merge acts as an origin, so set ti->max_io_len
*/
ti->max_io_len = get_origin_minimum_chunksize(s->origin->bdev);
start_merge(s);
}
dm: fix truncated status strings Avoid returning a truncated table or status string instead of setting the DM_BUFFER_FULL_FLAG when the last target of a table fills the buffer. When processing a table or status request, the function retrieve_status calls ti->type->status. If ti->type->status returns non-zero, retrieve_status assumes that the buffer overflowed and sets DM_BUFFER_FULL_FLAG. However, targets don't return non-zero values from their status method on overflow. Most targets returns always zero. If a buffer overflow happens in a target that is not the last in the table, it gets noticed during the next iteration of the loop in retrieve_status; but if a buffer overflow happens in the last target, it goes unnoticed and erroneously truncated data is returned. In the current code, the targets behave in the following way: * dm-crypt returns -ENOMEM if there is not enough space to store the key, but it returns 0 on all other overflows. * dm-thin returns errors from the status method if a disk error happened. This is incorrect because retrieve_status doesn't check the error code, it assumes that all non-zero values mean buffer overflow. * all the other targets always return 0. This patch changes the ti->type->status function to return void (because most targets don't use the return code). Overflow is detected in retrieve_status: if the status method fills up the remaining space completely, it is assumed that buffer overflow happened. Cc: stable@vger.kernel.org Signed-off-by: Mikulas Patocka <mpatocka@redhat.com> Signed-off-by: Alasdair G Kergon <agk@redhat.com>
2013-03-02 06:45:44 +08:00
static void snapshot_status(struct dm_target *ti, status_type_t type,
unsigned status_flags, char *result, unsigned maxlen)
{
unsigned sz = 0;
struct dm_snapshot *snap = ti->private;
switch (type) {
case STATUSTYPE_INFO:
mutex_lock(&snap->lock);
if (!snap->valid)
DMEMIT("Invalid");
else if (snap->merge_failed)
DMEMIT("Merge failed");
else if (snap->snapshot_overflowed)
DMEMIT("Overflow");
else {
if (snap->store->type->usage) {
sector_t total_sectors, sectors_allocated,
metadata_sectors;
snap->store->type->usage(snap->store,
&total_sectors,
&sectors_allocated,
&metadata_sectors);
DMEMIT("%llu/%llu %llu",
(unsigned long long)sectors_allocated,
(unsigned long long)total_sectors,
(unsigned long long)metadata_sectors);
}
else
DMEMIT("Unknown");
}
mutex_unlock(&snap->lock);
break;
case STATUSTYPE_TABLE:
/*
* kdevname returns a static pointer so we need
* to make private copies if the output is to
* make sense.
*/
DMEMIT("%s %s", snap->origin->name, snap->cow->name);
snap->store->type->status(snap->store, type, result + sz,
maxlen - sz);
break;
}
}
static int snapshot_iterate_devices(struct dm_target *ti,
iterate_devices_callout_fn fn, void *data)
{
struct dm_snapshot *snap = ti->private;
int r;
r = fn(ti, snap->origin, 0, ti->len, data);
if (!r)
r = fn(ti, snap->cow, 0, get_dev_size(snap->cow->bdev), data);
return r;
}
/*-----------------------------------------------------------------
* Origin methods
*---------------------------------------------------------------*/
/*
* If no exceptions need creating, DM_MAPIO_REMAPPED is returned and any
* supplied bio was ignored. The caller may submit it immediately.
* (No remapping actually occurs as the origin is always a direct linear
* map.)
*
* If further exceptions are required, DM_MAPIO_SUBMITTED is returned
* and any supplied bio is added to a list to be submitted once all
* the necessary exceptions exist.
*/
static int __origin_write(struct list_head *snapshots, sector_t sector,
struct bio *bio)
{
int r = DM_MAPIO_REMAPPED;
struct dm_snapshot *snap;
struct dm_exception *e;
struct dm_snap_pending_exception *pe;
struct dm_snap_pending_exception *pe_to_start_now = NULL;
struct dm_snap_pending_exception *pe_to_start_last = NULL;
chunk_t chunk;
/* Do all the snapshots on this origin */
list_for_each_entry (snap, snapshots, list) {
/*
* Don't make new exceptions in a merging snapshot
* because it has effectively been deleted
*/
if (dm_target_is_snapshot_merge(snap->ti))
continue;
mutex_lock(&snap->lock);
/* Only deal with valid and active snapshots */
if (!snap->valid || !snap->active)
goto next_snapshot;
/* Nothing to do if writing beyond end of snapshot */
if (sector >= dm_table_get_size(snap->ti->table))
goto next_snapshot;
/*
* Remember, different snapshots can have
* different chunk sizes.
*/
chunk = sector_to_chunk(snap->store, sector);
/*
* Check exception table to see if block
* is already remapped in this snapshot
* and trigger an exception if not.
*/
e = dm_lookup_exception(&snap->complete, chunk);
if (e)
goto next_snapshot;
pe = __lookup_pending_exception(snap, chunk);
if (!pe) {
mutex_unlock(&snap->lock);
pe = alloc_pending_exception(snap);
mutex_lock(&snap->lock);
if (!snap->valid) {
free_pending_exception(pe);
goto next_snapshot;
}
e = dm_lookup_exception(&snap->complete, chunk);
if (e) {
free_pending_exception(pe);
goto next_snapshot;
}
pe = __find_pending_exception(snap, pe, chunk);
if (!pe) {
__invalidate_snapshot(snap, -ENOMEM);
goto next_snapshot;
}
}
r = DM_MAPIO_SUBMITTED;
/*
* If an origin bio was supplied, queue it to wait for the
* completion of this exception, and start this one last,
* at the end of the function.
*/
if (bio) {
bio_list_add(&pe->origin_bios, bio);
bio = NULL;
if (!pe->started) {
pe->started = 1;
pe_to_start_last = pe;
}
}
if (!pe->started) {
pe->started = 1;
pe_to_start_now = pe;
}
next_snapshot:
mutex_unlock(&snap->lock);
if (pe_to_start_now) {
start_copy(pe_to_start_now);
pe_to_start_now = NULL;
}
}
/*
* Submit the exception against which the bio is queued last,
* to give the other exceptions a head start.
*/
if (pe_to_start_last)
start_copy(pe_to_start_last);
return r;
}
/*
* Called on a write from the origin driver.
*/
static int do_origin(struct dm_dev *origin, struct bio *bio)
{
struct origin *o;
int r = DM_MAPIO_REMAPPED;
down_read(&_origins_lock);
o = __lookup_origin(origin->bdev);
if (o)
block: Abstract out bvec iterator Immutable biovecs are going to require an explicit iterator. To implement immutable bvecs, a later patch is going to add a bi_bvec_done member to this struct; for now, this patch effectively just renames things. Signed-off-by: Kent Overstreet <kmo@daterainc.com> Cc: Jens Axboe <axboe@kernel.dk> Cc: Geert Uytterhoeven <geert@linux-m68k.org> Cc: Benjamin Herrenschmidt <benh@kernel.crashing.org> Cc: Paul Mackerras <paulus@samba.org> Cc: "Ed L. Cashin" <ecashin@coraid.com> Cc: Nick Piggin <npiggin@kernel.dk> Cc: Lars Ellenberg <drbd-dev@lists.linbit.com> Cc: Jiri Kosina <jkosina@suse.cz> Cc: Matthew Wilcox <willy@linux.intel.com> Cc: Geoff Levand <geoff@infradead.org> Cc: Yehuda Sadeh <yehuda@inktank.com> Cc: Sage Weil <sage@inktank.com> Cc: Alex Elder <elder@inktank.com> Cc: ceph-devel@vger.kernel.org Cc: Joshua Morris <josh.h.morris@us.ibm.com> Cc: Philip Kelleher <pjk1939@linux.vnet.ibm.com> Cc: Rusty Russell <rusty@rustcorp.com.au> Cc: "Michael S. Tsirkin" <mst@redhat.com> Cc: Konrad Rzeszutek Wilk <konrad.wilk@oracle.com> Cc: Jeremy Fitzhardinge <jeremy@goop.org> Cc: Neil Brown <neilb@suse.de> Cc: Alasdair Kergon <agk@redhat.com> Cc: Mike Snitzer <snitzer@redhat.com> Cc: dm-devel@redhat.com Cc: Martin Schwidefsky <schwidefsky@de.ibm.com> Cc: Heiko Carstens <heiko.carstens@de.ibm.com> Cc: linux390@de.ibm.com Cc: Boaz Harrosh <bharrosh@panasas.com> Cc: Benny Halevy <bhalevy@tonian.com> Cc: "James E.J. Bottomley" <JBottomley@parallels.com> Cc: Greg Kroah-Hartman <gregkh@linuxfoundation.org> Cc: "Nicholas A. Bellinger" <nab@linux-iscsi.org> Cc: Alexander Viro <viro@zeniv.linux.org.uk> Cc: Chris Mason <chris.mason@fusionio.com> Cc: "Theodore Ts'o" <tytso@mit.edu> Cc: Andreas Dilger <adilger.kernel@dilger.ca> Cc: Jaegeuk Kim <jaegeuk.kim@samsung.com> Cc: Steven Whitehouse <swhiteho@redhat.com> Cc: Dave Kleikamp <shaggy@kernel.org> Cc: Joern Engel <joern@logfs.org> Cc: Prasad Joshi <prasadjoshi.linux@gmail.com> Cc: Trond Myklebust <Trond.Myklebust@netapp.com> Cc: KONISHI Ryusuke <konishi.ryusuke@lab.ntt.co.jp> Cc: Mark Fasheh <mfasheh@suse.com> Cc: Joel Becker <jlbec@evilplan.org> Cc: Ben Myers <bpm@sgi.com> Cc: xfs@oss.sgi.com Cc: Steven Rostedt <rostedt@goodmis.org> Cc: Frederic Weisbecker <fweisbec@gmail.com> Cc: Ingo Molnar <mingo@redhat.com> Cc: Len Brown <len.brown@intel.com> Cc: Pavel Machek <pavel@ucw.cz> Cc: "Rafael J. Wysocki" <rjw@sisk.pl> Cc: Herton Ronaldo Krzesinski <herton.krzesinski@canonical.com> Cc: Ben Hutchings <ben@decadent.org.uk> Cc: Andrew Morton <akpm@linux-foundation.org> Cc: Guo Chao <yan@linux.vnet.ibm.com> Cc: Tejun Heo <tj@kernel.org> Cc: Asai Thambi S P <asamymuthupa@micron.com> Cc: Selvan Mani <smani@micron.com> Cc: Sam Bradshaw <sbradshaw@micron.com> Cc: Wei Yongjun <yongjun_wei@trendmicro.com.cn> Cc: "Roger Pau Monné" <roger.pau@citrix.com> Cc: Jan Beulich <jbeulich@suse.com> Cc: Stefano Stabellini <stefano.stabellini@eu.citrix.com> Cc: Ian Campbell <Ian.Campbell@citrix.com> Cc: Sebastian Ott <sebott@linux.vnet.ibm.com> Cc: Christian Borntraeger <borntraeger@de.ibm.com> Cc: Minchan Kim <minchan@kernel.org> Cc: Jiang Liu <jiang.liu@huawei.com> Cc: Nitin Gupta <ngupta@vflare.org> Cc: Jerome Marchand <jmarchand@redhat.com> Cc: Joe Perches <joe@perches.com> Cc: Peng Tao <tao.peng@emc.com> Cc: Andy Adamson <andros@netapp.com> Cc: fanchaoting <fanchaoting@cn.fujitsu.com> Cc: Jie Liu <jeff.liu@oracle.com> Cc: Sunil Mushran <sunil.mushran@gmail.com> Cc: "Martin K. Petersen" <martin.petersen@oracle.com> Cc: Namjae Jeon <namjae.jeon@samsung.com> Cc: Pankaj Kumar <pankaj.km@samsung.com> Cc: Dan Magenheimer <dan.magenheimer@oracle.com> Cc: Mel Gorman <mgorman@suse.de>6
2013-10-12 06:44:27 +08:00
r = __origin_write(&o->snapshots, bio->bi_iter.bi_sector, bio);
up_read(&_origins_lock);
return r;
}
/*
* Trigger exceptions in all non-merging snapshots.
*
* The chunk size of the merging snapshot may be larger than the chunk
* size of some other snapshot so we may need to reallocate multiple
* chunks in other snapshots.
*
* We scan all the overlapping exceptions in the other snapshots.
* Returns 1 if anything was reallocated and must be waited for,
* otherwise returns 0.
*
* size must be a multiple of merging_snap's chunk_size.
*/
static int origin_write_extent(struct dm_snapshot *merging_snap,
sector_t sector, unsigned size)
{
int must_wait = 0;
sector_t n;
struct origin *o;
/*
* The origin's __minimum_chunk_size() got stored in max_io_len
* by snapshot_merge_resume().
*/
down_read(&_origins_lock);
o = __lookup_origin(merging_snap->origin->bdev);
for (n = 0; n < size; n += merging_snap->ti->max_io_len)
if (__origin_write(&o->snapshots, sector + n, NULL) ==
DM_MAPIO_SUBMITTED)
must_wait = 1;
up_read(&_origins_lock);
return must_wait;
}
/*
* Origin: maps a linear range of a device, with hooks for snapshotting.
*/
/*
* Construct an origin mapping: <dev_path>
* The context for an origin is merely a 'struct dm_dev *'
* pointing to the real device.
*/
static int origin_ctr(struct dm_target *ti, unsigned int argc, char **argv)
{
int r;
struct dm_origin *o;
if (argc != 1) {
ti->error = "origin: incorrect number of arguments";
return -EINVAL;
}
o = kmalloc(sizeof(struct dm_origin), GFP_KERNEL);
if (!o) {
ti->error = "Cannot allocate private origin structure";
r = -ENOMEM;
goto bad_alloc;
}
r = dm_get_device(ti, argv[0], dm_table_get_mode(ti->table), &o->dev);
if (r) {
ti->error = "Cannot get target device";
goto bad_open;
}
dm snapshot: suspend origin when doing exception handover In the function snapshot_resume we perform exception store handover. If there is another active snapshot target, the exception store is moved from this target to the target that is being resumed. The problem is that if there is some pending exception, it will point to an incorrect exception store after that handover, causing a crash due to dm-snap-persistent.c:get_exception()'s BUG_ON. This bug can be triggered by repeatedly changing snapshot permissions with "lvchange -p r" and "lvchange -p rw" while there are writes on the associated origin device. To fix this bug, we must suspend the origin device when doing the exception store handover to make sure that there are no pending exceptions: - introduce _origin_hash that keeps track of dm_origin structures. - introduce functions __lookup_dm_origin, __insert_dm_origin and __remove_dm_origin that manipulate the origin hash. - modify snapshot_resume so that it calls dm_internal_suspend_fast() and dm_internal_resume_fast() on the origin device. NOTE to stable@ people: When backporting to kernels 3.12-3.18, use dm_internal_suspend and dm_internal_resume instead of dm_internal_suspend_fast and dm_internal_resume_fast. When backporting to kernels older than 3.12, you need to pick functions dm_internal_suspend and dm_internal_resume from the commit fd2ed4d252701d3bbed4cd3e3d267ad469bb832a. Signed-off-by: Mikulas Patocka <mpatocka@redhat.com> Signed-off-by: Mike Snitzer <snitzer@redhat.com> Cc: stable@vger.kernel.org
2015-02-27 00:40:35 +08:00
o->ti = ti;
ti->private = o;
ti->num_flush_bios = 1;
return 0;
bad_open:
kfree(o);
bad_alloc:
return r;
}
static void origin_dtr(struct dm_target *ti)
{
struct dm_origin *o = ti->private;
dm snapshot: suspend origin when doing exception handover In the function snapshot_resume we perform exception store handover. If there is another active snapshot target, the exception store is moved from this target to the target that is being resumed. The problem is that if there is some pending exception, it will point to an incorrect exception store after that handover, causing a crash due to dm-snap-persistent.c:get_exception()'s BUG_ON. This bug can be triggered by repeatedly changing snapshot permissions with "lvchange -p r" and "lvchange -p rw" while there are writes on the associated origin device. To fix this bug, we must suspend the origin device when doing the exception store handover to make sure that there are no pending exceptions: - introduce _origin_hash that keeps track of dm_origin structures. - introduce functions __lookup_dm_origin, __insert_dm_origin and __remove_dm_origin that manipulate the origin hash. - modify snapshot_resume so that it calls dm_internal_suspend_fast() and dm_internal_resume_fast() on the origin device. NOTE to stable@ people: When backporting to kernels 3.12-3.18, use dm_internal_suspend and dm_internal_resume instead of dm_internal_suspend_fast and dm_internal_resume_fast. When backporting to kernels older than 3.12, you need to pick functions dm_internal_suspend and dm_internal_resume from the commit fd2ed4d252701d3bbed4cd3e3d267ad469bb832a. Signed-off-by: Mikulas Patocka <mpatocka@redhat.com> Signed-off-by: Mike Snitzer <snitzer@redhat.com> Cc: stable@vger.kernel.org
2015-02-27 00:40:35 +08:00
dm_put_device(ti, o->dev);
kfree(o);
}
static int origin_map(struct dm_target *ti, struct bio *bio)
{
struct dm_origin *o = ti->private;
unsigned available_sectors;
bio_set_dev(bio, o->dev->bdev);
if (unlikely(bio->bi_opf & REQ_PREFLUSH))
return DM_MAPIO_REMAPPED;
if (bio_data_dir(bio) != WRITE)
return DM_MAPIO_REMAPPED;
available_sectors = o->split_boundary -
((unsigned)bio->bi_iter.bi_sector & (o->split_boundary - 1));
if (bio_sectors(bio) > available_sectors)
dm_accept_partial_bio(bio, available_sectors);
/* Only tell snapshots if this is a write */
return do_origin(o->dev, bio);
}
static long origin_dax_direct_access(struct dm_target *ti, pgoff_t pgoff,
long nr_pages, void **kaddr, pfn_t *pfn)
{
DMWARN("device does not support dax.");
return -EIO;
}
/*
* Set the target "max_io_len" field to the minimum of all the snapshots'
* chunk sizes.
*/
static void origin_resume(struct dm_target *ti)
{
struct dm_origin *o = ti->private;
o->split_boundary = get_origin_minimum_chunksize(o->dev->bdev);
dm snapshot: suspend origin when doing exception handover In the function snapshot_resume we perform exception store handover. If there is another active snapshot target, the exception store is moved from this target to the target that is being resumed. The problem is that if there is some pending exception, it will point to an incorrect exception store after that handover, causing a crash due to dm-snap-persistent.c:get_exception()'s BUG_ON. This bug can be triggered by repeatedly changing snapshot permissions with "lvchange -p r" and "lvchange -p rw" while there are writes on the associated origin device. To fix this bug, we must suspend the origin device when doing the exception store handover to make sure that there are no pending exceptions: - introduce _origin_hash that keeps track of dm_origin structures. - introduce functions __lookup_dm_origin, __insert_dm_origin and __remove_dm_origin that manipulate the origin hash. - modify snapshot_resume so that it calls dm_internal_suspend_fast() and dm_internal_resume_fast() on the origin device. NOTE to stable@ people: When backporting to kernels 3.12-3.18, use dm_internal_suspend and dm_internal_resume instead of dm_internal_suspend_fast and dm_internal_resume_fast. When backporting to kernels older than 3.12, you need to pick functions dm_internal_suspend and dm_internal_resume from the commit fd2ed4d252701d3bbed4cd3e3d267ad469bb832a. Signed-off-by: Mikulas Patocka <mpatocka@redhat.com> Signed-off-by: Mike Snitzer <snitzer@redhat.com> Cc: stable@vger.kernel.org
2015-02-27 00:40:35 +08:00
down_write(&_origins_lock);
__insert_dm_origin(o);
up_write(&_origins_lock);
}
static void origin_postsuspend(struct dm_target *ti)
{
struct dm_origin *o = ti->private;
down_write(&_origins_lock);
__remove_dm_origin(o);
up_write(&_origins_lock);
}
dm: fix truncated status strings Avoid returning a truncated table or status string instead of setting the DM_BUFFER_FULL_FLAG when the last target of a table fills the buffer. When processing a table or status request, the function retrieve_status calls ti->type->status. If ti->type->status returns non-zero, retrieve_status assumes that the buffer overflowed and sets DM_BUFFER_FULL_FLAG. However, targets don't return non-zero values from their status method on overflow. Most targets returns always zero. If a buffer overflow happens in a target that is not the last in the table, it gets noticed during the next iteration of the loop in retrieve_status; but if a buffer overflow happens in the last target, it goes unnoticed and erroneously truncated data is returned. In the current code, the targets behave in the following way: * dm-crypt returns -ENOMEM if there is not enough space to store the key, but it returns 0 on all other overflows. * dm-thin returns errors from the status method if a disk error happened. This is incorrect because retrieve_status doesn't check the error code, it assumes that all non-zero values mean buffer overflow. * all the other targets always return 0. This patch changes the ti->type->status function to return void (because most targets don't use the return code). Overflow is detected in retrieve_status: if the status method fills up the remaining space completely, it is assumed that buffer overflow happened. Cc: stable@vger.kernel.org Signed-off-by: Mikulas Patocka <mpatocka@redhat.com> Signed-off-by: Alasdair G Kergon <agk@redhat.com>
2013-03-02 06:45:44 +08:00
static void origin_status(struct dm_target *ti, status_type_t type,
unsigned status_flags, char *result, unsigned maxlen)
{
struct dm_origin *o = ti->private;
switch (type) {
case STATUSTYPE_INFO:
result[0] = '\0';
break;
case STATUSTYPE_TABLE:
snprintf(result, maxlen, "%s", o->dev->name);
break;
}
}
static int origin_iterate_devices(struct dm_target *ti,
iterate_devices_callout_fn fn, void *data)
{
struct dm_origin *o = ti->private;
return fn(ti, o->dev, 0, ti->len, data);
}
static struct target_type origin_target = {
.name = "snapshot-origin",
dm snapshot: suspend origin when doing exception handover In the function snapshot_resume we perform exception store handover. If there is another active snapshot target, the exception store is moved from this target to the target that is being resumed. The problem is that if there is some pending exception, it will point to an incorrect exception store after that handover, causing a crash due to dm-snap-persistent.c:get_exception()'s BUG_ON. This bug can be triggered by repeatedly changing snapshot permissions with "lvchange -p r" and "lvchange -p rw" while there are writes on the associated origin device. To fix this bug, we must suspend the origin device when doing the exception store handover to make sure that there are no pending exceptions: - introduce _origin_hash that keeps track of dm_origin structures. - introduce functions __lookup_dm_origin, __insert_dm_origin and __remove_dm_origin that manipulate the origin hash. - modify snapshot_resume so that it calls dm_internal_suspend_fast() and dm_internal_resume_fast() on the origin device. NOTE to stable@ people: When backporting to kernels 3.12-3.18, use dm_internal_suspend and dm_internal_resume instead of dm_internal_suspend_fast and dm_internal_resume_fast. When backporting to kernels older than 3.12, you need to pick functions dm_internal_suspend and dm_internal_resume from the commit fd2ed4d252701d3bbed4cd3e3d267ad469bb832a. Signed-off-by: Mikulas Patocka <mpatocka@redhat.com> Signed-off-by: Mike Snitzer <snitzer@redhat.com> Cc: stable@vger.kernel.org
2015-02-27 00:40:35 +08:00
.version = {1, 9, 0},
.module = THIS_MODULE,
.ctr = origin_ctr,
.dtr = origin_dtr,
.map = origin_map,
.resume = origin_resume,
dm snapshot: suspend origin when doing exception handover In the function snapshot_resume we perform exception store handover. If there is another active snapshot target, the exception store is moved from this target to the target that is being resumed. The problem is that if there is some pending exception, it will point to an incorrect exception store after that handover, causing a crash due to dm-snap-persistent.c:get_exception()'s BUG_ON. This bug can be triggered by repeatedly changing snapshot permissions with "lvchange -p r" and "lvchange -p rw" while there are writes on the associated origin device. To fix this bug, we must suspend the origin device when doing the exception store handover to make sure that there are no pending exceptions: - introduce _origin_hash that keeps track of dm_origin structures. - introduce functions __lookup_dm_origin, __insert_dm_origin and __remove_dm_origin that manipulate the origin hash. - modify snapshot_resume so that it calls dm_internal_suspend_fast() and dm_internal_resume_fast() on the origin device. NOTE to stable@ people: When backporting to kernels 3.12-3.18, use dm_internal_suspend and dm_internal_resume instead of dm_internal_suspend_fast and dm_internal_resume_fast. When backporting to kernels older than 3.12, you need to pick functions dm_internal_suspend and dm_internal_resume from the commit fd2ed4d252701d3bbed4cd3e3d267ad469bb832a. Signed-off-by: Mikulas Patocka <mpatocka@redhat.com> Signed-off-by: Mike Snitzer <snitzer@redhat.com> Cc: stable@vger.kernel.org
2015-02-27 00:40:35 +08:00
.postsuspend = origin_postsuspend,
.status = origin_status,
.iterate_devices = origin_iterate_devices,
.direct_access = origin_dax_direct_access,
};
static struct target_type snapshot_target = {
.name = "snapshot",
.version = {1, 15, 0},
.module = THIS_MODULE,
.ctr = snapshot_ctr,
.dtr = snapshot_dtr,
.map = snapshot_map,
.end_io = snapshot_end_io,
.preresume = snapshot_preresume,
.resume = snapshot_resume,
.status = snapshot_status,
.iterate_devices = snapshot_iterate_devices,
};
static struct target_type merge_target = {
.name = dm_snapshot_merge_target_name,
.version = {1, 4, 0},
.module = THIS_MODULE,
.ctr = snapshot_ctr,
.dtr = snapshot_dtr,
.map = snapshot_merge_map,
.end_io = snapshot_end_io,
.presuspend = snapshot_merge_presuspend,
.preresume = snapshot_preresume,
.resume = snapshot_merge_resume,
.status = snapshot_status,
.iterate_devices = snapshot_iterate_devices,
};
static int __init dm_snapshot_init(void)
{
int r;
r = dm_exception_store_init();
if (r) {
DMERR("Failed to initialize exception stores");
return r;
}
r = init_origin_hash();
if (r) {
DMERR("init_origin_hash failed.");
goto bad_origin_hash;
}
exception_cache = KMEM_CACHE(dm_exception, 0);
if (!exception_cache) {
DMERR("Couldn't create exception cache.");
r = -ENOMEM;
goto bad_exception_cache;
}
pending_cache = KMEM_CACHE(dm_snap_pending_exception, 0);
if (!pending_cache) {
DMERR("Couldn't create pending cache.");
r = -ENOMEM;
goto bad_pending_cache;
}
dm: fix various targets to dm_register_target after module __init resources created A NULL pointer is seen if two concurrent "vgchange -ay -K <vg name>" processes race to load the dm-thin-pool module: PID: 25992 TASK: ffff883cd7d23500 CPU: 4 COMMAND: "vgchange" #0 [ffff883cd743d600] machine_kexec at ffffffff81038fa9 0000001 [ffff883cd743d660] crash_kexec at ffffffff810c5992 0000002 [ffff883cd743d730] oops_end at ffffffff81515c90 0000003 [ffff883cd743d760] no_context at ffffffff81049f1b 0000004 [ffff883cd743d7b0] __bad_area_nosemaphore at ffffffff8104a1a5 0000005 [ffff883cd743d800] bad_area at ffffffff8104a2ce 0000006 [ffff883cd743d830] __do_page_fault at ffffffff8104aa6f 0000007 [ffff883cd743d950] do_page_fault at ffffffff81517bae 0000008 [ffff883cd743d980] page_fault at ffffffff81514f95 [exception RIP: kmem_cache_alloc+108] RIP: ffffffff8116ef3c RSP: ffff883cd743da38 RFLAGS: 00010046 RAX: 0000000000000004 RBX: ffffffff81121b90 RCX: ffff881bf1e78cc0 RDX: 0000000000000000 RSI: 00000000000000d0 RDI: 0000000000000000 RBP: ffff883cd743da68 R8: ffff881bf1a4eb00 R9: 0000000080042000 R10: 0000000000002000 R11: 0000000000000000 R12: 00000000000000d0 R13: 0000000000000000 R14: 00000000000000d0 R15: 0000000000000246 ORIG_RAX: ffffffffffffffff CS: 0010 SS: 0018 0000009 [ffff883cd743da70] mempool_alloc_slab at ffffffff81121ba5 0000010 [ffff883cd743da80] mempool_create_node at ffffffff81122083 0000011 [ffff883cd743dad0] mempool_create at ffffffff811220f4 0000012 [ffff883cd743dae0] pool_ctr at ffffffffa08de049 [dm_thin_pool] 0000013 [ffff883cd743dbd0] dm_table_add_target at ffffffffa0005f2f [dm_mod] 0000014 [ffff883cd743dc30] table_load at ffffffffa0008ba9 [dm_mod] 0000015 [ffff883cd743dc90] ctl_ioctl at ffffffffa0009dc4 [dm_mod] The race results in a NULL pointer because: Process A (vgchange -ay -K): a. send DM_LIST_VERSIONS_CMD ioctl; b. pool_target not registered; c. modprobe dm_thin_pool and wait until end. Process B (vgchange -ay -K): a. send DM_LIST_VERSIONS_CMD ioctl; b. pool_target registered; c. table_load->dm_table_add_target->pool_ctr; d. _new_mapping_cache is NULL and panic. Note: 1. process A and process B are two concurrent processes. 2. pool_target can be detected by process B but _new_mapping_cache initialization has not ended. To fix dm-thin-pool, and other targets (cache, multipath, and snapshot) with the same problem, simply dm_register_target() after all resources created during module init (as labelled with __init) are finished. Cc: stable@vger.kernel.org Signed-off-by: monty <monty_pavel@sina.com> Signed-off-by: Mike Snitzer <snitzer@redhat.com>
2017-11-25 01:43:50 +08:00
r = dm_register_target(&snapshot_target);
if (r < 0) {
DMERR("snapshot target register failed %d", r);
goto bad_register_snapshot_target;
}
r = dm_register_target(&origin_target);
if (r < 0) {
DMERR("Origin target register failed %d", r);
goto bad_register_origin_target;
}
r = dm_register_target(&merge_target);
if (r < 0) {
DMERR("Merge target register failed %d", r);
goto bad_register_merge_target;
}
return 0;
bad_register_merge_target:
dm_unregister_target(&origin_target);
bad_register_origin_target:
dm_unregister_target(&snapshot_target);
bad_register_snapshot_target:
dm: fix various targets to dm_register_target after module __init resources created A NULL pointer is seen if two concurrent "vgchange -ay -K <vg name>" processes race to load the dm-thin-pool module: PID: 25992 TASK: ffff883cd7d23500 CPU: 4 COMMAND: "vgchange" #0 [ffff883cd743d600] machine_kexec at ffffffff81038fa9 0000001 [ffff883cd743d660] crash_kexec at ffffffff810c5992 0000002 [ffff883cd743d730] oops_end at ffffffff81515c90 0000003 [ffff883cd743d760] no_context at ffffffff81049f1b 0000004 [ffff883cd743d7b0] __bad_area_nosemaphore at ffffffff8104a1a5 0000005 [ffff883cd743d800] bad_area at ffffffff8104a2ce 0000006 [ffff883cd743d830] __do_page_fault at ffffffff8104aa6f 0000007 [ffff883cd743d950] do_page_fault at ffffffff81517bae 0000008 [ffff883cd743d980] page_fault at ffffffff81514f95 [exception RIP: kmem_cache_alloc+108] RIP: ffffffff8116ef3c RSP: ffff883cd743da38 RFLAGS: 00010046 RAX: 0000000000000004 RBX: ffffffff81121b90 RCX: ffff881bf1e78cc0 RDX: 0000000000000000 RSI: 00000000000000d0 RDI: 0000000000000000 RBP: ffff883cd743da68 R8: ffff881bf1a4eb00 R9: 0000000080042000 R10: 0000000000002000 R11: 0000000000000000 R12: 00000000000000d0 R13: 0000000000000000 R14: 00000000000000d0 R15: 0000000000000246 ORIG_RAX: ffffffffffffffff CS: 0010 SS: 0018 0000009 [ffff883cd743da70] mempool_alloc_slab at ffffffff81121ba5 0000010 [ffff883cd743da80] mempool_create_node at ffffffff81122083 0000011 [ffff883cd743dad0] mempool_create at ffffffff811220f4 0000012 [ffff883cd743dae0] pool_ctr at ffffffffa08de049 [dm_thin_pool] 0000013 [ffff883cd743dbd0] dm_table_add_target at ffffffffa0005f2f [dm_mod] 0000014 [ffff883cd743dc30] table_load at ffffffffa0008ba9 [dm_mod] 0000015 [ffff883cd743dc90] ctl_ioctl at ffffffffa0009dc4 [dm_mod] The race results in a NULL pointer because: Process A (vgchange -ay -K): a. send DM_LIST_VERSIONS_CMD ioctl; b. pool_target not registered; c. modprobe dm_thin_pool and wait until end. Process B (vgchange -ay -K): a. send DM_LIST_VERSIONS_CMD ioctl; b. pool_target registered; c. table_load->dm_table_add_target->pool_ctr; d. _new_mapping_cache is NULL and panic. Note: 1. process A and process B are two concurrent processes. 2. pool_target can be detected by process B but _new_mapping_cache initialization has not ended. To fix dm-thin-pool, and other targets (cache, multipath, and snapshot) with the same problem, simply dm_register_target() after all resources created during module init (as labelled with __init) are finished. Cc: stable@vger.kernel.org Signed-off-by: monty <monty_pavel@sina.com> Signed-off-by: Mike Snitzer <snitzer@redhat.com>
2017-11-25 01:43:50 +08:00
kmem_cache_destroy(pending_cache);
bad_pending_cache:
kmem_cache_destroy(exception_cache);
bad_exception_cache:
exit_origin_hash();
bad_origin_hash:
dm_exception_store_exit();
return r;
}
static void __exit dm_snapshot_exit(void)
{
dm_unregister_target(&snapshot_target);
dm_unregister_target(&origin_target);
dm_unregister_target(&merge_target);
exit_origin_hash();
kmem_cache_destroy(pending_cache);
kmem_cache_destroy(exception_cache);
dm_exception_store_exit();
}
/* Module hooks */
module_init(dm_snapshot_init);
module_exit(dm_snapshot_exit);
MODULE_DESCRIPTION(DM_NAME " snapshot target");
MODULE_AUTHOR("Joe Thornber");
MODULE_LICENSE("GPL");
MODULE_ALIAS("dm-snapshot-origin");
MODULE_ALIAS("dm-snapshot-merge");