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
5fc2ffeabb (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>
This patch converts bio-based dm to support REQ_FLUSH/FUA instead of
now deprecated REQ_HARDBARRIER.
* -EOPNOTSUPP handling logic dropped.
* Preflush is handled as before but postflush is dropped and replaced
with passing down REQ_FUA to member request_queues. This replaces
one array wide cache flush w/ member specific FUA writes.
* __split_and_process_bio() now calls __clone_and_map_flush() directly
for flushes and guarantees all FLUSH bio's going to targets are zero
` length.
* It's now guaranteed that all FLUSH bio's which are passed onto dm
targets are zero length. bio_empty_barrier() tests are replaced
with REQ_FLUSH tests.
* Empty WRITE_BARRIERs are replaced with WRITE_FLUSHes.
* Dropped unlikely() around REQ_FLUSH tests. Flushes are not unlikely
enough to be marked with unlikely().
* Block layer now filters out REQ_FLUSH/FUA bio's if the request_queue
doesn't support cache flushing. Advertise REQ_FLUSH | REQ_FUA
capability.
* Request based dm isn't converted yet. dm_init_request_based_queue()
resets flush support to 0 for now. To avoid disturbing request
based dm code, dm->flush_error is added for bio based dm while
requested based dm continues to use dm->barrier_error.
Lightly tested linear, stripe, raid1, snap and crypt targets. Please
proceed with caution as I'm not familiar with the code base.
Signed-off-by: Tejun Heo <tj@kernel.org>
Cc: dm-devel@redhat.com
Cc: Christoph Hellwig <hch@lst.de>
Signed-off-by: Jens Axboe <jaxboe@fusionio.com>
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>
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>
Move bio completion out of dm_rh_mark_nosync in preparation for the
next patch.
Signed-off-by: Mikulas Patocka <mpatocka@redhat.com>
Reviewed-by: Takahiro Yasui <tyasui@redhat.com>
Tested-by: Takahiro Yasui <tyasui@redhat.com>
Signed-off-by: Alasdair G Kergon <agk@redhat.com>
Flush support for dm-raid1.
When it receives an empty barrier, submit it to all the devices via dm-io.
Signed-off-by: Mikulas Patocka <mpatocka@redhat.com>
Signed-off-by: Alasdair G Kergon <agk@redhat.com>
If the code can't handle allocation failures, use __GFP_NOFAIL so that
in case of memory pressure the allocator will retry indefinitely and
won't return NULL which would cause a crash in the function.
This is still not a correct fix, it may cause a classic deadlock when
memory manager waits for I/O being done and I/O waits for some free memory.
I/O code shouldn't allocate any memory. But in this case it probably
doesn't matter much in practice, people usually do not swap on RAID.
Signed-off-by: Mikulas Patocka <mpatocka@redhat.com>
Signed-off-by: Alasdair G Kergon <agk@redhat.com>
It's used by DM and MD and generally useful, so move the bio list
helpers into bio.h.
Signed-off-by: Christoph Hellwig <hch@lst.de>
Acked-by: Alasdair G Kergon <agk@redhat.com>
Signed-off-by: Jens Axboe <jens.axboe@oracle.com>
Separate the region hash code from raid1 so it can be shared by forthcoming
targets. Use BUG_ON() for failed async dm_io() calls.
Signed-off-by: Heinz Mauelshagen <hjm@redhat.com>
Signed-off-by: Alasdair G Kergon <agk@redhat.com>
Kent Overstreet
Mikulas Patocka
Lucas De Marchi
Tejun Heo
Takahiro Yasui
Christoph Hellwig
Heinz Mauelshagen