2008-10-21 18:45:06 +02:00
|
|
|
/*
|
|
|
|
* Copyright (C) 2003 Sistina Software Limited.
|
|
|
|
* Copyright (C) 2004-2008 Red Hat, Inc. All rights reserved.
|
|
|
|
*
|
|
|
|
* This file is released under the GPL.
|
|
|
|
*/
|
|
|
|
|
|
|
|
#include <linux/dm-dirty-log.h>
|
|
|
|
#include <linux/dm-region-hash.h>
|
|
|
|
|
|
|
|
#include <linux/ctype.h>
|
|
|
|
#include <linux/init.h>
|
|
|
|
#include <linux/module.h>
|
include cleanup: Update gfp.h and slab.h includes to prepare for breaking implicit slab.h inclusion from percpu.h
percpu.h is included by sched.h and module.h and thus ends up being
included when building most .c files. percpu.h includes slab.h which
in turn includes gfp.h making everything defined by the two files
universally available and complicating inclusion dependencies.
percpu.h -> slab.h dependency is about to be removed. Prepare for
this change by updating users of gfp and slab facilities include those
headers directly instead of assuming availability. As this conversion
needs to touch large number of source files, the following script is
used as the basis of conversion.
http://userweb.kernel.org/~tj/misc/slabh-sweep.py
The script does the followings.
* Scan files for gfp and slab usages and update includes such that
only the necessary includes are there. ie. if only gfp is used,
gfp.h, if slab is used, slab.h.
* When the script inserts a new include, it looks at the include
blocks and try to put the new include such that its order conforms
to its surrounding. It's put in the include block which contains
core kernel includes, in the same order that the rest are ordered -
alphabetical, Christmas tree, rev-Xmas-tree or at the end if there
doesn't seem to be any matching order.
* If the script can't find a place to put a new include (mostly
because the file doesn't have fitting include block), it prints out
an error message indicating which .h file needs to be added to the
file.
The conversion was done in the following steps.
1. The initial automatic conversion of all .c files updated slightly
over 4000 files, deleting around 700 includes and adding ~480 gfp.h
and ~3000 slab.h inclusions. The script emitted errors for ~400
files.
2. Each error was manually checked. Some didn't need the inclusion,
some needed manual addition while adding it to implementation .h or
embedding .c file was more appropriate for others. This step added
inclusions to around 150 files.
3. The script was run again and the output was compared to the edits
from #2 to make sure no file was left behind.
4. Several build tests were done and a couple of problems were fixed.
e.g. lib/decompress_*.c used malloc/free() wrappers around slab
APIs requiring slab.h to be added manually.
5. The script was run on all .h files but without automatically
editing them as sprinkling gfp.h and slab.h inclusions around .h
files could easily lead to inclusion dependency hell. Most gfp.h
inclusion directives were ignored as stuff from gfp.h was usually
wildly available and often used in preprocessor macros. Each
slab.h inclusion directive was examined and added manually as
necessary.
6. percpu.h was updated not to include slab.h.
7. Build test were done on the following configurations and failures
were fixed. CONFIG_GCOV_KERNEL was turned off for all tests (as my
distributed build env didn't work with gcov compiles) and a few
more options had to be turned off depending on archs to make things
build (like ipr on powerpc/64 which failed due to missing writeq).
* x86 and x86_64 UP and SMP allmodconfig and a custom test config.
* powerpc and powerpc64 SMP allmodconfig
* sparc and sparc64 SMP allmodconfig
* ia64 SMP allmodconfig
* s390 SMP allmodconfig
* alpha SMP allmodconfig
* um on x86_64 SMP allmodconfig
8. percpu.h modifications were reverted so that it could be applied as
a separate patch and serve as bisection point.
Given the fact that I had only a couple of failures from tests on step
6, I'm fairly confident about the coverage of this conversion patch.
If there is a breakage, it's likely to be something in one of the arch
headers which should be easily discoverable easily on most builds of
the specific arch.
Signed-off-by: Tejun Heo <tj@kernel.org>
Guess-its-ok-by: Christoph Lameter <cl@linux-foundation.org>
Cc: Ingo Molnar <mingo@redhat.com>
Cc: Lee Schermerhorn <Lee.Schermerhorn@hp.com>
2010-03-24 09:04:11 +01:00
|
|
|
#include <linux/slab.h>
|
2008-10-21 18:45:06 +02:00
|
|
|
#include <linux/vmalloc.h>
|
|
|
|
|
|
|
|
#include "dm.h"
|
|
|
|
|
|
|
|
#define DM_MSG_PREFIX "region hash"
|
|
|
|
|
|
|
|
/*-----------------------------------------------------------------
|
|
|
|
* Region hash
|
|
|
|
*
|
|
|
|
* The mirror splits itself up into discrete regions. Each
|
|
|
|
* region can be in one of three states: clean, dirty,
|
|
|
|
* nosync. There is no need to put clean regions in the hash.
|
|
|
|
*
|
|
|
|
* In addition to being present in the hash table a region _may_
|
|
|
|
* be present on one of three lists.
|
|
|
|
*
|
|
|
|
* clean_regions: Regions on this list have no io pending to
|
|
|
|
* them, they are in sync, we are no longer interested in them,
|
|
|
|
* they are dull. dm_rh_update_states() will remove them from the
|
|
|
|
* hash table.
|
|
|
|
*
|
|
|
|
* quiesced_regions: These regions have been spun down, ready
|
|
|
|
* for recovery. rh_recovery_start() will remove regions from
|
|
|
|
* this list and hand them to kmirrord, which will schedule the
|
|
|
|
* recovery io with kcopyd.
|
|
|
|
*
|
|
|
|
* recovered_regions: Regions that kcopyd has successfully
|
|
|
|
* recovered. dm_rh_update_states() will now schedule any delayed
|
|
|
|
* io, up the recovery_count, and remove the region from the
|
|
|
|
* hash.
|
|
|
|
*
|
|
|
|
* There are 2 locks:
|
|
|
|
* A rw spin lock 'hash_lock' protects just the hash table,
|
|
|
|
* this is never held in write mode from interrupt context,
|
|
|
|
* which I believe means that we only have to disable irqs when
|
|
|
|
* doing a write lock.
|
|
|
|
*
|
|
|
|
* An ordinary spin lock 'region_lock' that protects the three
|
|
|
|
* lists in the region_hash, with the 'state', 'list' and
|
|
|
|
* 'delayed_bios' fields of the regions. This is used from irq
|
|
|
|
* context, so all other uses will have to suspend local irqs.
|
|
|
|
*---------------------------------------------------------------*/
|
|
|
|
struct dm_region_hash {
|
|
|
|
uint32_t region_size;
|
|
|
|
unsigned region_shift;
|
|
|
|
|
|
|
|
/* holds persistent region state */
|
|
|
|
struct dm_dirty_log *log;
|
|
|
|
|
|
|
|
/* hash table */
|
|
|
|
rwlock_t hash_lock;
|
|
|
|
mempool_t *region_pool;
|
|
|
|
unsigned mask;
|
|
|
|
unsigned nr_buckets;
|
|
|
|
unsigned prime;
|
|
|
|
unsigned shift;
|
|
|
|
struct list_head *buckets;
|
|
|
|
|
|
|
|
unsigned max_recovery; /* Max # of regions to recover in parallel */
|
|
|
|
|
|
|
|
spinlock_t region_lock;
|
|
|
|
atomic_t recovery_in_flight;
|
|
|
|
struct semaphore recovery_count;
|
|
|
|
struct list_head clean_regions;
|
|
|
|
struct list_head quiesced_regions;
|
|
|
|
struct list_head recovered_regions;
|
|
|
|
struct list_head failed_recovered_regions;
|
|
|
|
|
2009-12-11 00:51:59 +01:00
|
|
|
/*
|
2010-09-03 11:56:19 +02:00
|
|
|
* If there was a flush failure no regions can be marked clean.
|
2009-12-11 00:51:59 +01:00
|
|
|
*/
|
2010-09-03 11:56:19 +02:00
|
|
|
int flush_failure;
|
2009-12-11 00:51:59 +01:00
|
|
|
|
2008-10-21 18:45:06 +02:00
|
|
|
void *context;
|
|
|
|
sector_t target_begin;
|
|
|
|
|
|
|
|
/* Callback function to schedule bios writes */
|
|
|
|
void (*dispatch_bios)(void *context, struct bio_list *bios);
|
|
|
|
|
|
|
|
/* Callback function to wakeup callers worker thread. */
|
|
|
|
void (*wakeup_workers)(void *context);
|
|
|
|
|
|
|
|
/* Callback function to wakeup callers recovery waiters. */
|
|
|
|
void (*wakeup_all_recovery_waiters)(void *context);
|
|
|
|
};
|
|
|
|
|
|
|
|
struct dm_region {
|
|
|
|
struct dm_region_hash *rh; /* FIXME: can we get rid of this ? */
|
|
|
|
region_t key;
|
|
|
|
int state;
|
|
|
|
|
|
|
|
struct list_head hash_list;
|
|
|
|
struct list_head list;
|
|
|
|
|
|
|
|
atomic_t pending;
|
|
|
|
struct bio_list delayed_bios;
|
|
|
|
};
|
|
|
|
|
|
|
|
/*
|
|
|
|
* Conversion fns
|
|
|
|
*/
|
|
|
|
static region_t dm_rh_sector_to_region(struct dm_region_hash *rh, sector_t sector)
|
|
|
|
{
|
|
|
|
return sector >> rh->region_shift;
|
|
|
|
}
|
|
|
|
|
|
|
|
sector_t dm_rh_region_to_sector(struct dm_region_hash *rh, region_t region)
|
|
|
|
{
|
|
|
|
return region << rh->region_shift;
|
|
|
|
}
|
|
|
|
EXPORT_SYMBOL_GPL(dm_rh_region_to_sector);
|
|
|
|
|
|
|
|
region_t dm_rh_bio_to_region(struct dm_region_hash *rh, struct bio *bio)
|
|
|
|
{
|
2013-10-12 00:44:27 +02:00
|
|
|
return dm_rh_sector_to_region(rh, bio->bi_iter.bi_sector -
|
|
|
|
rh->target_begin);
|
2008-10-21 18:45:06 +02:00
|
|
|
}
|
|
|
|
EXPORT_SYMBOL_GPL(dm_rh_bio_to_region);
|
|
|
|
|
|
|
|
void *dm_rh_region_context(struct dm_region *reg)
|
|
|
|
{
|
|
|
|
return reg->rh->context;
|
|
|
|
}
|
|
|
|
EXPORT_SYMBOL_GPL(dm_rh_region_context);
|
|
|
|
|
|
|
|
region_t dm_rh_get_region_key(struct dm_region *reg)
|
|
|
|
{
|
|
|
|
return reg->key;
|
|
|
|
}
|
|
|
|
EXPORT_SYMBOL_GPL(dm_rh_get_region_key);
|
|
|
|
|
|
|
|
sector_t dm_rh_get_region_size(struct dm_region_hash *rh)
|
|
|
|
{
|
|
|
|
return rh->region_size;
|
|
|
|
}
|
|
|
|
EXPORT_SYMBOL_GPL(dm_rh_get_region_size);
|
|
|
|
|
|
|
|
/*
|
|
|
|
* FIXME: shall we pass in a structure instead of all these args to
|
|
|
|
* dm_region_hash_create()????
|
|
|
|
*/
|
|
|
|
#define RH_HASH_MULT 2654435387U
|
|
|
|
#define RH_HASH_SHIFT 12
|
|
|
|
|
|
|
|
#define MIN_REGIONS 64
|
|
|
|
struct dm_region_hash *dm_region_hash_create(
|
|
|
|
void *context, void (*dispatch_bios)(void *context,
|
|
|
|
struct bio_list *bios),
|
|
|
|
void (*wakeup_workers)(void *context),
|
|
|
|
void (*wakeup_all_recovery_waiters)(void *context),
|
|
|
|
sector_t target_begin, unsigned max_recovery,
|
|
|
|
struct dm_dirty_log *log, uint32_t region_size,
|
|
|
|
region_t nr_regions)
|
|
|
|
{
|
|
|
|
struct dm_region_hash *rh;
|
|
|
|
unsigned nr_buckets, max_buckets;
|
|
|
|
size_t i;
|
|
|
|
|
|
|
|
/*
|
|
|
|
* Calculate a suitable number of buckets for our hash
|
|
|
|
* table.
|
|
|
|
*/
|
|
|
|
max_buckets = nr_regions >> 6;
|
|
|
|
for (nr_buckets = 128u; nr_buckets < max_buckets; nr_buckets <<= 1)
|
|
|
|
;
|
|
|
|
nr_buckets >>= 1;
|
|
|
|
|
|
|
|
rh = kmalloc(sizeof(*rh), GFP_KERNEL);
|
|
|
|
if (!rh) {
|
|
|
|
DMERR("unable to allocate region hash memory");
|
|
|
|
return ERR_PTR(-ENOMEM);
|
|
|
|
}
|
|
|
|
|
|
|
|
rh->context = context;
|
|
|
|
rh->dispatch_bios = dispatch_bios;
|
|
|
|
rh->wakeup_workers = wakeup_workers;
|
|
|
|
rh->wakeup_all_recovery_waiters = wakeup_all_recovery_waiters;
|
|
|
|
rh->target_begin = target_begin;
|
|
|
|
rh->max_recovery = max_recovery;
|
|
|
|
rh->log = log;
|
|
|
|
rh->region_size = region_size;
|
|
|
|
rh->region_shift = ffs(region_size) - 1;
|
|
|
|
rwlock_init(&rh->hash_lock);
|
|
|
|
rh->mask = nr_buckets - 1;
|
|
|
|
rh->nr_buckets = nr_buckets;
|
|
|
|
|
|
|
|
rh->shift = RH_HASH_SHIFT;
|
|
|
|
rh->prime = RH_HASH_MULT;
|
|
|
|
|
|
|
|
rh->buckets = vmalloc(nr_buckets * sizeof(*rh->buckets));
|
|
|
|
if (!rh->buckets) {
|
|
|
|
DMERR("unable to allocate region hash bucket memory");
|
|
|
|
kfree(rh);
|
|
|
|
return ERR_PTR(-ENOMEM);
|
|
|
|
}
|
|
|
|
|
|
|
|
for (i = 0; i < nr_buckets; i++)
|
|
|
|
INIT_LIST_HEAD(rh->buckets + i);
|
|
|
|
|
|
|
|
spin_lock_init(&rh->region_lock);
|
|
|
|
sema_init(&rh->recovery_count, 0);
|
|
|
|
atomic_set(&rh->recovery_in_flight, 0);
|
|
|
|
INIT_LIST_HEAD(&rh->clean_regions);
|
|
|
|
INIT_LIST_HEAD(&rh->quiesced_regions);
|
|
|
|
INIT_LIST_HEAD(&rh->recovered_regions);
|
|
|
|
INIT_LIST_HEAD(&rh->failed_recovered_regions);
|
2010-09-03 11:56:19 +02:00
|
|
|
rh->flush_failure = 0;
|
2008-10-21 18:45:06 +02:00
|
|
|
|
|
|
|
rh->region_pool = mempool_create_kmalloc_pool(MIN_REGIONS,
|
|
|
|
sizeof(struct dm_region));
|
|
|
|
if (!rh->region_pool) {
|
|
|
|
vfree(rh->buckets);
|
|
|
|
kfree(rh);
|
|
|
|
rh = ERR_PTR(-ENOMEM);
|
|
|
|
}
|
|
|
|
|
|
|
|
return rh;
|
|
|
|
}
|
|
|
|
EXPORT_SYMBOL_GPL(dm_region_hash_create);
|
|
|
|
|
|
|
|
void dm_region_hash_destroy(struct dm_region_hash *rh)
|
|
|
|
{
|
|
|
|
unsigned h;
|
|
|
|
struct dm_region *reg, *nreg;
|
|
|
|
|
|
|
|
BUG_ON(!list_empty(&rh->quiesced_regions));
|
|
|
|
for (h = 0; h < rh->nr_buckets; h++) {
|
|
|
|
list_for_each_entry_safe(reg, nreg, rh->buckets + h,
|
|
|
|
hash_list) {
|
|
|
|
BUG_ON(atomic_read(®->pending));
|
|
|
|
mempool_free(reg, rh->region_pool);
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
|
|
|
if (rh->log)
|
|
|
|
dm_dirty_log_destroy(rh->log);
|
|
|
|
|
2015-09-13 14:15:05 +02:00
|
|
|
mempool_destroy(rh->region_pool);
|
2008-10-21 18:45:06 +02:00
|
|
|
vfree(rh->buckets);
|
|
|
|
kfree(rh);
|
|
|
|
}
|
|
|
|
EXPORT_SYMBOL_GPL(dm_region_hash_destroy);
|
|
|
|
|
|
|
|
struct dm_dirty_log *dm_rh_dirty_log(struct dm_region_hash *rh)
|
|
|
|
{
|
|
|
|
return rh->log;
|
|
|
|
}
|
|
|
|
EXPORT_SYMBOL_GPL(dm_rh_dirty_log);
|
|
|
|
|
|
|
|
static unsigned rh_hash(struct dm_region_hash *rh, region_t region)
|
|
|
|
{
|
|
|
|
return (unsigned) ((region * rh->prime) >> rh->shift) & rh->mask;
|
|
|
|
}
|
|
|
|
|
|
|
|
static struct dm_region *__rh_lookup(struct dm_region_hash *rh, region_t region)
|
|
|
|
{
|
|
|
|
struct dm_region *reg;
|
|
|
|
struct list_head *bucket = rh->buckets + rh_hash(rh, region);
|
|
|
|
|
|
|
|
list_for_each_entry(reg, bucket, hash_list)
|
|
|
|
if (reg->key == region)
|
|
|
|
return reg;
|
|
|
|
|
|
|
|
return NULL;
|
|
|
|
}
|
|
|
|
|
|
|
|
static void __rh_insert(struct dm_region_hash *rh, struct dm_region *reg)
|
|
|
|
{
|
|
|
|
list_add(®->hash_list, rh->buckets + rh_hash(rh, reg->key));
|
|
|
|
}
|
|
|
|
|
|
|
|
static struct dm_region *__rh_alloc(struct dm_region_hash *rh, region_t region)
|
|
|
|
{
|
|
|
|
struct dm_region *reg, *nreg;
|
|
|
|
|
|
|
|
nreg = mempool_alloc(rh->region_pool, GFP_ATOMIC);
|
|
|
|
if (unlikely(!nreg))
|
2009-06-22 11:12:13 +02:00
|
|
|
nreg = kmalloc(sizeof(*nreg), GFP_NOIO | __GFP_NOFAIL);
|
2008-10-21 18:45:06 +02:00
|
|
|
|
|
|
|
nreg->state = rh->log->type->in_sync(rh->log, region, 1) ?
|
|
|
|
DM_RH_CLEAN : DM_RH_NOSYNC;
|
|
|
|
nreg->rh = rh;
|
|
|
|
nreg->key = region;
|
|
|
|
INIT_LIST_HEAD(&nreg->list);
|
|
|
|
atomic_set(&nreg->pending, 0);
|
|
|
|
bio_list_init(&nreg->delayed_bios);
|
|
|
|
|
|
|
|
write_lock_irq(&rh->hash_lock);
|
|
|
|
reg = __rh_lookup(rh, region);
|
|
|
|
if (reg)
|
|
|
|
/* We lost the race. */
|
|
|
|
mempool_free(nreg, rh->region_pool);
|
|
|
|
else {
|
|
|
|
__rh_insert(rh, nreg);
|
|
|
|
if (nreg->state == DM_RH_CLEAN) {
|
|
|
|
spin_lock(&rh->region_lock);
|
|
|
|
list_add(&nreg->list, &rh->clean_regions);
|
|
|
|
spin_unlock(&rh->region_lock);
|
|
|
|
}
|
|
|
|
|
|
|
|
reg = nreg;
|
|
|
|
}
|
|
|
|
write_unlock_irq(&rh->hash_lock);
|
|
|
|
|
|
|
|
return reg;
|
|
|
|
}
|
|
|
|
|
|
|
|
static struct dm_region *__rh_find(struct dm_region_hash *rh, region_t region)
|
|
|
|
{
|
|
|
|
struct dm_region *reg;
|
|
|
|
|
|
|
|
reg = __rh_lookup(rh, region);
|
|
|
|
if (!reg) {
|
|
|
|
read_unlock(&rh->hash_lock);
|
|
|
|
reg = __rh_alloc(rh, region);
|
|
|
|
read_lock(&rh->hash_lock);
|
|
|
|
}
|
|
|
|
|
|
|
|
return reg;
|
|
|
|
}
|
|
|
|
|
|
|
|
int dm_rh_get_state(struct dm_region_hash *rh, region_t region, int may_block)
|
|
|
|
{
|
|
|
|
int r;
|
|
|
|
struct dm_region *reg;
|
|
|
|
|
|
|
|
read_lock(&rh->hash_lock);
|
|
|
|
reg = __rh_lookup(rh, region);
|
|
|
|
read_unlock(&rh->hash_lock);
|
|
|
|
|
|
|
|
if (reg)
|
|
|
|
return reg->state;
|
|
|
|
|
|
|
|
/*
|
|
|
|
* The region wasn't in the hash, so we fall back to the
|
|
|
|
* dirty log.
|
|
|
|
*/
|
|
|
|
r = rh->log->type->in_sync(rh->log, region, may_block);
|
|
|
|
|
|
|
|
/*
|
|
|
|
* Any error from the dirty log (eg. -EWOULDBLOCK) gets
|
|
|
|
* taken as a DM_RH_NOSYNC
|
|
|
|
*/
|
|
|
|
return r == 1 ? DM_RH_CLEAN : DM_RH_NOSYNC;
|
|
|
|
}
|
|
|
|
EXPORT_SYMBOL_GPL(dm_rh_get_state);
|
|
|
|
|
|
|
|
static void complete_resync_work(struct dm_region *reg, int success)
|
|
|
|
{
|
|
|
|
struct dm_region_hash *rh = reg->rh;
|
|
|
|
|
|
|
|
rh->log->type->set_region_sync(rh->log, reg->key, success);
|
|
|
|
|
|
|
|
/*
|
|
|
|
* Dispatch the bios before we call 'wake_up_all'.
|
|
|
|
* This is important because if we are suspending,
|
|
|
|
* we want to know that recovery is complete and
|
|
|
|
* the work queue is flushed. If we wake_up_all
|
|
|
|
* before we dispatch_bios (queue bios and call wake()),
|
|
|
|
* then we risk suspending before the work queue
|
|
|
|
* has been properly flushed.
|
|
|
|
*/
|
|
|
|
rh->dispatch_bios(rh->context, ®->delayed_bios);
|
|
|
|
if (atomic_dec_and_test(&rh->recovery_in_flight))
|
|
|
|
rh->wakeup_all_recovery_waiters(rh->context);
|
|
|
|
up(&rh->recovery_count);
|
|
|
|
}
|
|
|
|
|
|
|
|
/* dm_rh_mark_nosync
|
|
|
|
* @ms
|
|
|
|
* @bio
|
|
|
|
*
|
|
|
|
* The bio was written on some mirror(s) but failed on other mirror(s).
|
|
|
|
* We can successfully endio the bio but should avoid the region being
|
|
|
|
* marked clean by setting the state DM_RH_NOSYNC.
|
|
|
|
*
|
|
|
|
* This function is _not_ safe in interrupt context!
|
|
|
|
*/
|
2009-12-11 00:52:05 +01:00
|
|
|
void dm_rh_mark_nosync(struct dm_region_hash *rh, struct bio *bio)
|
2008-10-21 18:45:06 +02:00
|
|
|
{
|
|
|
|
unsigned long flags;
|
|
|
|
struct dm_dirty_log *log = rh->log;
|
|
|
|
struct dm_region *reg;
|
|
|
|
region_t region = dm_rh_bio_to_region(rh, bio);
|
|
|
|
int recovering = 0;
|
|
|
|
|
2010-09-03 11:56:19 +02:00
|
|
|
if (bio->bi_rw & REQ_FLUSH) {
|
|
|
|
rh->flush_failure = 1;
|
2009-12-11 00:51:59 +01:00
|
|
|
return;
|
|
|
|
}
|
|
|
|
|
dm raid1: fix crash with mirror recovery and discard
This patch fixes a crash when a discard request is sent during mirror
recovery.
Firstly, some background. Generally, the following sequence happens during
mirror synchronization:
- function do_recovery is called
- do_recovery calls dm_rh_recovery_prepare
- dm_rh_recovery_prepare uses a semaphore to limit the number
simultaneously recovered regions (by default the semaphore value is 1,
so only one region at a time is recovered)
- dm_rh_recovery_prepare calls __rh_recovery_prepare,
__rh_recovery_prepare asks the log driver for the next region to
recover. Then, it sets the region state to DM_RH_RECOVERING. If there
are no pending I/Os on this region, the region is added to
quiesced_regions list. If there are pending I/Os, the region is not
added to any list. It is added to the quiesced_regions list later (by
dm_rh_dec function) when all I/Os finish.
- when the region is on quiesced_regions list, there are no I/Os in
flight on this region. The region is popped from the list in
dm_rh_recovery_start function. Then, a kcopyd job is started in the
recover function.
- when the kcopyd job finishes, recovery_complete is called. It calls
dm_rh_recovery_end. dm_rh_recovery_end adds the region to
recovered_regions or failed_recovered_regions list (depending on
whether the copy operation was successful or not).
The above mechanism assumes that if the region is in DM_RH_RECOVERING
state, no new I/Os are started on this region. When I/O is started,
dm_rh_inc_pending is called, which increases reg->pending count. When
I/O is finished, dm_rh_dec is called. It decreases reg->pending count.
If the count is zero and the region was in DM_RH_RECOVERING state,
dm_rh_dec adds it to the quiesced_regions list.
Consequently, if we call dm_rh_inc_pending/dm_rh_dec while the region is
in DM_RH_RECOVERING state, it could be added to quiesced_regions list
multiple times or it could be added to this list when kcopyd is copying
data (it is assumed that the region is not on any list while kcopyd does
its jobs). This results in memory corruption and crash.
There already exist bypasses for REQ_FLUSH requests: REQ_FLUSH requests
do not belong to any region, so they are always added to the sync list
in do_writes. dm_rh_inc_pending does not increase count for REQ_FLUSH
requests. In mirror_end_io, dm_rh_dec is never called for REQ_FLUSH
requests. These bypasses avoid the crash possibility described above.
These bypasses were improperly implemented for REQ_DISCARD when
the mirror target gained discard support in commit
5fc2ffeabb9ee0fc0e71ff16b49f34f0ed3d05b4 (dm raid1: support discard).
In do_writes, REQ_DISCARD requests is always added to the sync queue and
immediately dispatched (even if the region is in DM_RH_RECOVERING). However,
dm_rh_inc and dm_rh_dec is called for REQ_DISCARD resusts. So it violates the
rule that no I/Os are started on DM_RH_RECOVERING regions, and causes the list
corruption described above.
This patch changes it so that REQ_DISCARD requests follow the same path
as REQ_FLUSH. This avoids the crash.
Reference: https://bugzilla.redhat.com/837607
Signed-off-by: Mikulas Patocka <mpatocka@redhat.com>
Cc: stable@kernel.org
Signed-off-by: Alasdair G Kergon <agk@redhat.com>
2012-07-20 15:25:03 +02:00
|
|
|
if (bio->bi_rw & REQ_DISCARD)
|
|
|
|
return;
|
|
|
|
|
2008-10-21 18:45:06 +02:00
|
|
|
/* We must inform the log that the sync count has changed. */
|
|
|
|
log->type->set_region_sync(log, region, 0);
|
|
|
|
|
|
|
|
read_lock(&rh->hash_lock);
|
|
|
|
reg = __rh_find(rh, region);
|
|
|
|
read_unlock(&rh->hash_lock);
|
|
|
|
|
|
|
|
/* region hash entry should exist because write was in-flight */
|
|
|
|
BUG_ON(!reg);
|
|
|
|
BUG_ON(!list_empty(®->list));
|
|
|
|
|
|
|
|
spin_lock_irqsave(&rh->region_lock, flags);
|
|
|
|
/*
|
|
|
|
* Possible cases:
|
|
|
|
* 1) DM_RH_DIRTY
|
2011-03-31 03:57:33 +02:00
|
|
|
* 2) DM_RH_NOSYNC: was dirty, other preceding writes failed
|
2008-10-21 18:45:06 +02:00
|
|
|
* 3) DM_RH_RECOVERING: flushing pending writes
|
|
|
|
* Either case, the region should have not been connected to list.
|
|
|
|
*/
|
|
|
|
recovering = (reg->state == DM_RH_RECOVERING);
|
|
|
|
reg->state = DM_RH_NOSYNC;
|
|
|
|
BUG_ON(!list_empty(®->list));
|
|
|
|
spin_unlock_irqrestore(&rh->region_lock, flags);
|
|
|
|
|
|
|
|
if (recovering)
|
|
|
|
complete_resync_work(reg, 0);
|
|
|
|
}
|
|
|
|
EXPORT_SYMBOL_GPL(dm_rh_mark_nosync);
|
|
|
|
|
|
|
|
void dm_rh_update_states(struct dm_region_hash *rh, int errors_handled)
|
|
|
|
{
|
|
|
|
struct dm_region *reg, *next;
|
|
|
|
|
|
|
|
LIST_HEAD(clean);
|
|
|
|
LIST_HEAD(recovered);
|
|
|
|
LIST_HEAD(failed_recovered);
|
|
|
|
|
|
|
|
/*
|
|
|
|
* Quickly grab the lists.
|
|
|
|
*/
|
|
|
|
write_lock_irq(&rh->hash_lock);
|
|
|
|
spin_lock(&rh->region_lock);
|
|
|
|
if (!list_empty(&rh->clean_regions)) {
|
|
|
|
list_splice_init(&rh->clean_regions, &clean);
|
|
|
|
|
|
|
|
list_for_each_entry(reg, &clean, list)
|
|
|
|
list_del(®->hash_list);
|
|
|
|
}
|
|
|
|
|
|
|
|
if (!list_empty(&rh->recovered_regions)) {
|
|
|
|
list_splice_init(&rh->recovered_regions, &recovered);
|
|
|
|
|
|
|
|
list_for_each_entry(reg, &recovered, list)
|
|
|
|
list_del(®->hash_list);
|
|
|
|
}
|
|
|
|
|
|
|
|
if (!list_empty(&rh->failed_recovered_regions)) {
|
|
|
|
list_splice_init(&rh->failed_recovered_regions,
|
|
|
|
&failed_recovered);
|
|
|
|
|
|
|
|
list_for_each_entry(reg, &failed_recovered, list)
|
|
|
|
list_del(®->hash_list);
|
|
|
|
}
|
|
|
|
|
|
|
|
spin_unlock(&rh->region_lock);
|
|
|
|
write_unlock_irq(&rh->hash_lock);
|
|
|
|
|
|
|
|
/*
|
|
|
|
* All the regions on the recovered and clean lists have
|
|
|
|
* now been pulled out of the system, so no need to do
|
|
|
|
* any more locking.
|
|
|
|
*/
|
|
|
|
list_for_each_entry_safe(reg, next, &recovered, list) {
|
|
|
|
rh->log->type->clear_region(rh->log, reg->key);
|
|
|
|
complete_resync_work(reg, 1);
|
|
|
|
mempool_free(reg, rh->region_pool);
|
|
|
|
}
|
|
|
|
|
|
|
|
list_for_each_entry_safe(reg, next, &failed_recovered, list) {
|
|
|
|
complete_resync_work(reg, errors_handled ? 0 : 1);
|
|
|
|
mempool_free(reg, rh->region_pool);
|
|
|
|
}
|
|
|
|
|
|
|
|
list_for_each_entry_safe(reg, next, &clean, list) {
|
|
|
|
rh->log->type->clear_region(rh->log, reg->key);
|
|
|
|
mempool_free(reg, rh->region_pool);
|
|
|
|
}
|
|
|
|
|
|
|
|
rh->log->type->flush(rh->log);
|
|
|
|
}
|
|
|
|
EXPORT_SYMBOL_GPL(dm_rh_update_states);
|
|
|
|
|
|
|
|
static void rh_inc(struct dm_region_hash *rh, region_t region)
|
|
|
|
{
|
|
|
|
struct dm_region *reg;
|
|
|
|
|
|
|
|
read_lock(&rh->hash_lock);
|
|
|
|
reg = __rh_find(rh, region);
|
|
|
|
|
|
|
|
spin_lock_irq(&rh->region_lock);
|
|
|
|
atomic_inc(®->pending);
|
|
|
|
|
|
|
|
if (reg->state == DM_RH_CLEAN) {
|
|
|
|
reg->state = DM_RH_DIRTY;
|
|
|
|
list_del_init(®->list); /* take off the clean list */
|
|
|
|
spin_unlock_irq(&rh->region_lock);
|
|
|
|
|
|
|
|
rh->log->type->mark_region(rh->log, reg->key);
|
|
|
|
} else
|
|
|
|
spin_unlock_irq(&rh->region_lock);
|
|
|
|
|
|
|
|
|
|
|
|
read_unlock(&rh->hash_lock);
|
|
|
|
}
|
|
|
|
|
|
|
|
void dm_rh_inc_pending(struct dm_region_hash *rh, struct bio_list *bios)
|
|
|
|
{
|
|
|
|
struct bio *bio;
|
|
|
|
|
2009-12-11 00:51:59 +01:00
|
|
|
for (bio = bios->head; bio; bio = bio->bi_next) {
|
dm raid1: fix crash with mirror recovery and discard
This patch fixes a crash when a discard request is sent during mirror
recovery.
Firstly, some background. Generally, the following sequence happens during
mirror synchronization:
- function do_recovery is called
- do_recovery calls dm_rh_recovery_prepare
- dm_rh_recovery_prepare uses a semaphore to limit the number
simultaneously recovered regions (by default the semaphore value is 1,
so only one region at a time is recovered)
- dm_rh_recovery_prepare calls __rh_recovery_prepare,
__rh_recovery_prepare asks the log driver for the next region to
recover. Then, it sets the region state to DM_RH_RECOVERING. If there
are no pending I/Os on this region, the region is added to
quiesced_regions list. If there are pending I/Os, the region is not
added to any list. It is added to the quiesced_regions list later (by
dm_rh_dec function) when all I/Os finish.
- when the region is on quiesced_regions list, there are no I/Os in
flight on this region. The region is popped from the list in
dm_rh_recovery_start function. Then, a kcopyd job is started in the
recover function.
- when the kcopyd job finishes, recovery_complete is called. It calls
dm_rh_recovery_end. dm_rh_recovery_end adds the region to
recovered_regions or failed_recovered_regions list (depending on
whether the copy operation was successful or not).
The above mechanism assumes that if the region is in DM_RH_RECOVERING
state, no new I/Os are started on this region. When I/O is started,
dm_rh_inc_pending is called, which increases reg->pending count. When
I/O is finished, dm_rh_dec is called. It decreases reg->pending count.
If the count is zero and the region was in DM_RH_RECOVERING state,
dm_rh_dec adds it to the quiesced_regions list.
Consequently, if we call dm_rh_inc_pending/dm_rh_dec while the region is
in DM_RH_RECOVERING state, it could be added to quiesced_regions list
multiple times or it could be added to this list when kcopyd is copying
data (it is assumed that the region is not on any list while kcopyd does
its jobs). This results in memory corruption and crash.
There already exist bypasses for REQ_FLUSH requests: REQ_FLUSH requests
do not belong to any region, so they are always added to the sync list
in do_writes. dm_rh_inc_pending does not increase count for REQ_FLUSH
requests. In mirror_end_io, dm_rh_dec is never called for REQ_FLUSH
requests. These bypasses avoid the crash possibility described above.
These bypasses were improperly implemented for REQ_DISCARD when
the mirror target gained discard support in commit
5fc2ffeabb9ee0fc0e71ff16b49f34f0ed3d05b4 (dm raid1: support discard).
In do_writes, REQ_DISCARD requests is always added to the sync queue and
immediately dispatched (even if the region is in DM_RH_RECOVERING). However,
dm_rh_inc and dm_rh_dec is called for REQ_DISCARD resusts. So it violates the
rule that no I/Os are started on DM_RH_RECOVERING regions, and causes the list
corruption described above.
This patch changes it so that REQ_DISCARD requests follow the same path
as REQ_FLUSH. This avoids the crash.
Reference: https://bugzilla.redhat.com/837607
Signed-off-by: Mikulas Patocka <mpatocka@redhat.com>
Cc: stable@kernel.org
Signed-off-by: Alasdair G Kergon <agk@redhat.com>
2012-07-20 15:25:03 +02:00
|
|
|
if (bio->bi_rw & (REQ_FLUSH | REQ_DISCARD))
|
2009-12-11 00:51:59 +01:00
|
|
|
continue;
|
2008-10-21 18:45:06 +02:00
|
|
|
rh_inc(rh, dm_rh_bio_to_region(rh, bio));
|
2009-12-11 00:51:59 +01:00
|
|
|
}
|
2008-10-21 18:45:06 +02:00
|
|
|
}
|
|
|
|
EXPORT_SYMBOL_GPL(dm_rh_inc_pending);
|
|
|
|
|
|
|
|
void dm_rh_dec(struct dm_region_hash *rh, region_t region)
|
|
|
|
{
|
|
|
|
unsigned long flags;
|
|
|
|
struct dm_region *reg;
|
|
|
|
int should_wake = 0;
|
|
|
|
|
|
|
|
read_lock(&rh->hash_lock);
|
|
|
|
reg = __rh_lookup(rh, region);
|
|
|
|
read_unlock(&rh->hash_lock);
|
|
|
|
|
|
|
|
spin_lock_irqsave(&rh->region_lock, flags);
|
|
|
|
if (atomic_dec_and_test(®->pending)) {
|
|
|
|
/*
|
|
|
|
* There is no pending I/O for this region.
|
|
|
|
* We can move the region to corresponding list for next action.
|
|
|
|
* At this point, the region is not yet connected to any list.
|
|
|
|
*
|
|
|
|
* If the state is DM_RH_NOSYNC, the region should be kept off
|
|
|
|
* from clean list.
|
|
|
|
* The hash entry for DM_RH_NOSYNC will remain in memory
|
|
|
|
* until the region is recovered or the map is reloaded.
|
|
|
|
*/
|
|
|
|
|
|
|
|
/* do nothing for DM_RH_NOSYNC */
|
2010-09-03 11:56:19 +02:00
|
|
|
if (unlikely(rh->flush_failure)) {
|
2009-12-11 00:51:59 +01:00
|
|
|
/*
|
2010-09-03 11:56:19 +02:00
|
|
|
* If a write flush failed some time ago, we
|
2009-12-11 00:51:59 +01:00
|
|
|
* don't know whether or not this write made it
|
|
|
|
* to the disk, so we must resync the device.
|
|
|
|
*/
|
|
|
|
reg->state = DM_RH_NOSYNC;
|
|
|
|
} else if (reg->state == DM_RH_RECOVERING) {
|
2008-10-21 18:45:06 +02:00
|
|
|
list_add_tail(®->list, &rh->quiesced_regions);
|
|
|
|
} else if (reg->state == DM_RH_DIRTY) {
|
|
|
|
reg->state = DM_RH_CLEAN;
|
|
|
|
list_add(®->list, &rh->clean_regions);
|
|
|
|
}
|
|
|
|
should_wake = 1;
|
|
|
|
}
|
|
|
|
spin_unlock_irqrestore(&rh->region_lock, flags);
|
|
|
|
|
|
|
|
if (should_wake)
|
|
|
|
rh->wakeup_workers(rh->context);
|
|
|
|
}
|
|
|
|
EXPORT_SYMBOL_GPL(dm_rh_dec);
|
|
|
|
|
|
|
|
/*
|
|
|
|
* Starts quiescing a region in preparation for recovery.
|
|
|
|
*/
|
|
|
|
static int __rh_recovery_prepare(struct dm_region_hash *rh)
|
|
|
|
{
|
|
|
|
int r;
|
|
|
|
region_t region;
|
|
|
|
struct dm_region *reg;
|
|
|
|
|
|
|
|
/*
|
|
|
|
* Ask the dirty log what's next.
|
|
|
|
*/
|
|
|
|
r = rh->log->type->get_resync_work(rh->log, ®ion);
|
|
|
|
if (r <= 0)
|
|
|
|
return r;
|
|
|
|
|
|
|
|
/*
|
|
|
|
* Get this region, and start it quiescing by setting the
|
|
|
|
* recovering flag.
|
|
|
|
*/
|
|
|
|
read_lock(&rh->hash_lock);
|
|
|
|
reg = __rh_find(rh, region);
|
|
|
|
read_unlock(&rh->hash_lock);
|
|
|
|
|
|
|
|
spin_lock_irq(&rh->region_lock);
|
|
|
|
reg->state = DM_RH_RECOVERING;
|
|
|
|
|
|
|
|
/* Already quiesced ? */
|
|
|
|
if (atomic_read(®->pending))
|
|
|
|
list_del_init(®->list);
|
|
|
|
else
|
|
|
|
list_move(®->list, &rh->quiesced_regions);
|
|
|
|
|
|
|
|
spin_unlock_irq(&rh->region_lock);
|
|
|
|
|
|
|
|
return 1;
|
|
|
|
}
|
|
|
|
|
|
|
|
void dm_rh_recovery_prepare(struct dm_region_hash *rh)
|
|
|
|
{
|
|
|
|
/* Extra reference to avoid race with dm_rh_stop_recovery */
|
|
|
|
atomic_inc(&rh->recovery_in_flight);
|
|
|
|
|
|
|
|
while (!down_trylock(&rh->recovery_count)) {
|
|
|
|
atomic_inc(&rh->recovery_in_flight);
|
|
|
|
if (__rh_recovery_prepare(rh) <= 0) {
|
|
|
|
atomic_dec(&rh->recovery_in_flight);
|
|
|
|
up(&rh->recovery_count);
|
|
|
|
break;
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
|
|
|
/* Drop the extra reference */
|
|
|
|
if (atomic_dec_and_test(&rh->recovery_in_flight))
|
|
|
|
rh->wakeup_all_recovery_waiters(rh->context);
|
|
|
|
}
|
|
|
|
EXPORT_SYMBOL_GPL(dm_rh_recovery_prepare);
|
|
|
|
|
|
|
|
/*
|
|
|
|
* Returns any quiesced regions.
|
|
|
|
*/
|
|
|
|
struct dm_region *dm_rh_recovery_start(struct dm_region_hash *rh)
|
|
|
|
{
|
|
|
|
struct dm_region *reg = NULL;
|
|
|
|
|
|
|
|
spin_lock_irq(&rh->region_lock);
|
|
|
|
if (!list_empty(&rh->quiesced_regions)) {
|
|
|
|
reg = list_entry(rh->quiesced_regions.next,
|
|
|
|
struct dm_region, list);
|
|
|
|
list_del_init(®->list); /* remove from the quiesced list */
|
|
|
|
}
|
|
|
|
spin_unlock_irq(&rh->region_lock);
|
|
|
|
|
|
|
|
return reg;
|
|
|
|
}
|
|
|
|
EXPORT_SYMBOL_GPL(dm_rh_recovery_start);
|
|
|
|
|
|
|
|
void dm_rh_recovery_end(struct dm_region *reg, int success)
|
|
|
|
{
|
|
|
|
struct dm_region_hash *rh = reg->rh;
|
|
|
|
|
|
|
|
spin_lock_irq(&rh->region_lock);
|
|
|
|
if (success)
|
|
|
|
list_add(®->list, ®->rh->recovered_regions);
|
dm raid1: fix null pointer dereference in suspend
When suspending a failed mirror, bios are completed by mirror_end_io() and
__rh_lookup() in dm_rh_dec() returns NULL where a non-NULL return value is
required by design. Fix this by not changing the state of the recovery failed
region from DM_RH_RECOVERING to DM_RH_NOSYNC in dm_rh_recovery_end().
Issue
On 2.6.33-rc1 kernel, I hit the bug when I suspended the failed
mirror by dmsetup command.
BUG: unable to handle kernel NULL pointer dereference at 00000020
IP: [<f94f38e2>] dm_rh_dec+0x35/0xa1 [dm_region_hash]
...
EIP: 0060:[<f94f38e2>] EFLAGS: 00010046 CPU: 0
EIP is at dm_rh_dec+0x35/0xa1 [dm_region_hash]
EAX: 00000286 EBX: 00000000 ECX: 00000286 EDX: 00000000
ESI: eff79eac EDI: eff79e80 EBP: f6915cd4 ESP: f6915cc4
DS: 007b ES: 007b FS: 00d8 GS: 0033 SS: 0068
Process dmsetup (pid: 2849, ti=f6914000 task=eff03e80 task.ti=f6914000)
...
Call Trace:
[<f9530af6>] ? mirror_end_io+0x53/0x1b1 [dm_mirror]
[<f9413104>] ? clone_endio+0x4d/0xa2 [dm_mod]
[<f9530aa3>] ? mirror_end_io+0x0/0x1b1 [dm_mirror]
[<f94130b7>] ? clone_endio+0x0/0xa2 [dm_mod]
[<c02d6bcb>] ? bio_endio+0x28/0x2b
[<f952f303>] ? hold_bio+0x2d/0x62 [dm_mirror]
[<f952f942>] ? mirror_presuspend+0xeb/0xf7 [dm_mirror]
[<c02aa3e2>] ? vmap_page_range+0xb/0xd
[<f9414c8d>] ? suspend_targets+0x2d/0x3b [dm_mod]
[<f9414ca9>] ? dm_table_presuspend_targets+0xe/0x10 [dm_mod]
[<f941456f>] ? dm_suspend+0x4d/0x150 [dm_mod]
[<f941767d>] ? dev_suspend+0x55/0x18a [dm_mod]
[<c0343762>] ? _copy_from_user+0x42/0x56
[<f9417fb0>] ? dm_ctl_ioctl+0x22c/0x281 [dm_mod]
[<f9417628>] ? dev_suspend+0x0/0x18a [dm_mod]
[<f9417d84>] ? dm_ctl_ioctl+0x0/0x281 [dm_mod]
[<c02c3c4b>] ? vfs_ioctl+0x22/0x85
[<c02c422c>] ? do_vfs_ioctl+0x4cb/0x516
[<c02c42b7>] ? sys_ioctl+0x40/0x5a
[<c0202858>] ? sysenter_do_call+0x12/0x28
Analysis
When recovery process of a region failed, dm_rh_recovery_end() function
changes the state of the region from RM_RH_RECOVERING to DM_RH_NOSYNC.
When recovery_complete() is executed between dm_rh_update_states() and
dm_writes() in do_mirror(), bios are processed with the region state,
DM_RH_NOSYNC. However, the region data is freed without checking its
pending count when dm_rh_update_states() is called next time.
When bios are finished by mirror_end_io(), __rh_lookup() in dm_rh_dec()
returns NULL even though a valid return value are expected.
Solution
Remove the state change of the recovery failed region from DM_RH_RECOVERING
to DM_RH_NOSYNC in dm_rh_recovery_end(). We can remove the state change
because:
- If the region data has been released by dm_rh_update_states(),
a new region data is created with the state of DM_RH_NOSYNC, and
bios are processed according to the DM_RH_NOSYNC state.
- If the region data has not been released by dm_rh_update_states(),
a state of the region is DM_RH_RECOVERING and bios are put in the
delayed_bio list.
The flag change from DM_RH_RECOVERING to DM_RH_NOSYNC in dm_rh_recovery_end()
was added in the following commit:
dm raid1: handle resync failures
author Jonathan Brassow <jbrassow@redhat.com>
Thu, 12 Jul 2007 16:29:04 +0000 (17:29 +0100)
http://git.kernel.org/linus/f44db678edcc6f4c2779ac43f63f0b9dfa28b724
Signed-off-by: Takahiro Yasui <tyasui@redhat.com>
Reviewed-by: Mikulas Patocka <mpatocka@redhat.com>
Signed-off-by: Alasdair G Kergon <agk@redhat.com>
2010-02-16 19:42:58 +01:00
|
|
|
else
|
2008-10-21 18:45:06 +02:00
|
|
|
list_add(®->list, ®->rh->failed_recovered_regions);
|
dm raid1: fix null pointer dereference in suspend
When suspending a failed mirror, bios are completed by mirror_end_io() and
__rh_lookup() in dm_rh_dec() returns NULL where a non-NULL return value is
required by design. Fix this by not changing the state of the recovery failed
region from DM_RH_RECOVERING to DM_RH_NOSYNC in dm_rh_recovery_end().
Issue
On 2.6.33-rc1 kernel, I hit the bug when I suspended the failed
mirror by dmsetup command.
BUG: unable to handle kernel NULL pointer dereference at 00000020
IP: [<f94f38e2>] dm_rh_dec+0x35/0xa1 [dm_region_hash]
...
EIP: 0060:[<f94f38e2>] EFLAGS: 00010046 CPU: 0
EIP is at dm_rh_dec+0x35/0xa1 [dm_region_hash]
EAX: 00000286 EBX: 00000000 ECX: 00000286 EDX: 00000000
ESI: eff79eac EDI: eff79e80 EBP: f6915cd4 ESP: f6915cc4
DS: 007b ES: 007b FS: 00d8 GS: 0033 SS: 0068
Process dmsetup (pid: 2849, ti=f6914000 task=eff03e80 task.ti=f6914000)
...
Call Trace:
[<f9530af6>] ? mirror_end_io+0x53/0x1b1 [dm_mirror]
[<f9413104>] ? clone_endio+0x4d/0xa2 [dm_mod]
[<f9530aa3>] ? mirror_end_io+0x0/0x1b1 [dm_mirror]
[<f94130b7>] ? clone_endio+0x0/0xa2 [dm_mod]
[<c02d6bcb>] ? bio_endio+0x28/0x2b
[<f952f303>] ? hold_bio+0x2d/0x62 [dm_mirror]
[<f952f942>] ? mirror_presuspend+0xeb/0xf7 [dm_mirror]
[<c02aa3e2>] ? vmap_page_range+0xb/0xd
[<f9414c8d>] ? suspend_targets+0x2d/0x3b [dm_mod]
[<f9414ca9>] ? dm_table_presuspend_targets+0xe/0x10 [dm_mod]
[<f941456f>] ? dm_suspend+0x4d/0x150 [dm_mod]
[<f941767d>] ? dev_suspend+0x55/0x18a [dm_mod]
[<c0343762>] ? _copy_from_user+0x42/0x56
[<f9417fb0>] ? dm_ctl_ioctl+0x22c/0x281 [dm_mod]
[<f9417628>] ? dev_suspend+0x0/0x18a [dm_mod]
[<f9417d84>] ? dm_ctl_ioctl+0x0/0x281 [dm_mod]
[<c02c3c4b>] ? vfs_ioctl+0x22/0x85
[<c02c422c>] ? do_vfs_ioctl+0x4cb/0x516
[<c02c42b7>] ? sys_ioctl+0x40/0x5a
[<c0202858>] ? sysenter_do_call+0x12/0x28
Analysis
When recovery process of a region failed, dm_rh_recovery_end() function
changes the state of the region from RM_RH_RECOVERING to DM_RH_NOSYNC.
When recovery_complete() is executed between dm_rh_update_states() and
dm_writes() in do_mirror(), bios are processed with the region state,
DM_RH_NOSYNC. However, the region data is freed without checking its
pending count when dm_rh_update_states() is called next time.
When bios are finished by mirror_end_io(), __rh_lookup() in dm_rh_dec()
returns NULL even though a valid return value are expected.
Solution
Remove the state change of the recovery failed region from DM_RH_RECOVERING
to DM_RH_NOSYNC in dm_rh_recovery_end(). We can remove the state change
because:
- If the region data has been released by dm_rh_update_states(),
a new region data is created with the state of DM_RH_NOSYNC, and
bios are processed according to the DM_RH_NOSYNC state.
- If the region data has not been released by dm_rh_update_states(),
a state of the region is DM_RH_RECOVERING and bios are put in the
delayed_bio list.
The flag change from DM_RH_RECOVERING to DM_RH_NOSYNC in dm_rh_recovery_end()
was added in the following commit:
dm raid1: handle resync failures
author Jonathan Brassow <jbrassow@redhat.com>
Thu, 12 Jul 2007 16:29:04 +0000 (17:29 +0100)
http://git.kernel.org/linus/f44db678edcc6f4c2779ac43f63f0b9dfa28b724
Signed-off-by: Takahiro Yasui <tyasui@redhat.com>
Reviewed-by: Mikulas Patocka <mpatocka@redhat.com>
Signed-off-by: Alasdair G Kergon <agk@redhat.com>
2010-02-16 19:42:58 +01:00
|
|
|
|
2008-10-21 18:45:06 +02:00
|
|
|
spin_unlock_irq(&rh->region_lock);
|
|
|
|
|
|
|
|
rh->wakeup_workers(rh->context);
|
|
|
|
}
|
|
|
|
EXPORT_SYMBOL_GPL(dm_rh_recovery_end);
|
|
|
|
|
|
|
|
/* Return recovery in flight count. */
|
|
|
|
int dm_rh_recovery_in_flight(struct dm_region_hash *rh)
|
|
|
|
{
|
|
|
|
return atomic_read(&rh->recovery_in_flight);
|
|
|
|
}
|
|
|
|
EXPORT_SYMBOL_GPL(dm_rh_recovery_in_flight);
|
|
|
|
|
|
|
|
int dm_rh_flush(struct dm_region_hash *rh)
|
|
|
|
{
|
|
|
|
return rh->log->type->flush(rh->log);
|
|
|
|
}
|
|
|
|
EXPORT_SYMBOL_GPL(dm_rh_flush);
|
|
|
|
|
|
|
|
void dm_rh_delay(struct dm_region_hash *rh, struct bio *bio)
|
|
|
|
{
|
|
|
|
struct dm_region *reg;
|
|
|
|
|
|
|
|
read_lock(&rh->hash_lock);
|
|
|
|
reg = __rh_find(rh, dm_rh_bio_to_region(rh, bio));
|
|
|
|
bio_list_add(®->delayed_bios, bio);
|
|
|
|
read_unlock(&rh->hash_lock);
|
|
|
|
}
|
|
|
|
EXPORT_SYMBOL_GPL(dm_rh_delay);
|
|
|
|
|
|
|
|
void dm_rh_stop_recovery(struct dm_region_hash *rh)
|
|
|
|
{
|
|
|
|
int i;
|
|
|
|
|
|
|
|
/* wait for any recovering regions */
|
|
|
|
for (i = 0; i < rh->max_recovery; i++)
|
|
|
|
down(&rh->recovery_count);
|
|
|
|
}
|
|
|
|
EXPORT_SYMBOL_GPL(dm_rh_stop_recovery);
|
|
|
|
|
|
|
|
void dm_rh_start_recovery(struct dm_region_hash *rh)
|
|
|
|
{
|
|
|
|
int i;
|
|
|
|
|
|
|
|
for (i = 0; i < rh->max_recovery; i++)
|
|
|
|
up(&rh->recovery_count);
|
|
|
|
|
|
|
|
rh->wakeup_workers(rh->context);
|
|
|
|
}
|
|
|
|
EXPORT_SYMBOL_GPL(dm_rh_start_recovery);
|
|
|
|
|
|
|
|
MODULE_DESCRIPTION(DM_NAME " region hash");
|
|
|
|
MODULE_AUTHOR("Joe Thornber/Heinz Mauelshagen <dm-devel@redhat.com>");
|
|
|
|
MODULE_LICENSE("GPL");
|