linux/block
Asias He 458f27a982 block: Avoid missed wakeup in request waitqueue
After hot-unplug a stressed disk, I found that rl->wait[] is not empty
while rl->count[] is empty and there are theads still sleeping on
get_request after the queue cleanup. With simple debug code, I found
there are exactly nr_sleep - nr_wakeup of theads in D state. So there
are missed wakeup.

  $ dmesg | grep nr_sleep
  [   52.917115] ---> nr_sleep=1046, nr_wakeup=873, delta=173
  $ vmstat 1
  1 173  0 712640  24292  96172 0 0  0  0  419  757  0  0  0 100  0

To quote Tejun:

  Ah, okay, freed_request() wakes up single waiter with the assumption
  that after the wakeup there will at least be one successful allocation
  which in turn will continue the wakeup chain until the wait list is
  empty - ie. waiter wakeup is dependent on successful request
  allocation happening after each wakeup.  With queue marked dead, any
  woken up waiter fails the allocation path, so the wakeup chaining is
  lost and we're left with hung waiters. What we need is wake_up_all()
  after drain completion.

This patch fixes the missed wakeup by waking up all the theads which
are sleeping on wait queue after queue drain.

Changes in v2: Drop waitqueue_active() optimization

Acked-by: Tejun Heo <tj@kernel.org>
Signed-off-by: Asias He <asias@redhat.com>

Fixed a bug by me, where stacked devices would oops on calling
blk_drain_queue() since ->rq.wait[] do not get initialized unless
it's a full queue setup.

Signed-off-by: Jens Axboe <axboe@kernel.dk>
2012-06-15 08:45:25 +02:00
..
partitions s390/dasd: re-prioritize partition detection message 2012-05-16 14:42:51 +02:00
blk-cgroup.c blkcg: drop local variable @q from blkg_destroy() 2012-06-06 08:35:31 +02:00
blk-cgroup.h blkcg: use radix tree to index blkgs from blkcg 2012-04-20 10:06:40 +02:00
blk-core.c block: Avoid missed wakeup in request waitqueue 2012-06-15 08:45:25 +02:00
blk-exec.c block: add missing blk_queue_dead() checks 2011-12-14 00:33:37 +01:00
blk-flush.c blk-flush: move the queue kick into 2011-10-24 16:24:31 +02:00
blk-integrity.c block: add export.h to files using EXPORT_SYMBOL/THIS_MODULE macros 2011-10-31 19:31:12 -04:00
blk-ioc.c block: avoid infinite loop in get_task_io_context() 2012-05-31 13:39:05 +02:00
blk-iopoll.c tree-wide: fix assorted typos all over the place 2009-12-04 15:39:55 +01:00
blk-lib.c block: fix patch import error in max_discard_sectors check 2011-07-23 20:34:59 +02:00
blk-map.c block: re-use existing 'reading' variable instead of checking direction again 2011-12-21 15:27:24 +01:00
blk-merge.c block: separate out blk_rq_merge_ok() and blk_try_merge() from elevator functions 2012-02-08 09:19:38 +01:00
blk-settings.c block: Introduce blk_set_stacking_limits function 2012-01-11 16:27:11 +01:00
blk-softirq.c sched, block: Unify cache detection 2012-01-27 13:28:48 +01:00
blk-sysfs.c blkcg: unify blkg's for blkcg policies 2012-03-06 21:27:23 +01:00
blk-tag.c block: fix blk_queue_end_tag() 2011-12-29 09:16:28 +01:00
blk-throttle.c blkcg: tg_stats_alloc_lock is an irq lock 2012-05-23 12:16:21 +02:00
blk-timeout.c fault-injection: add ability to export fault_attr in arbitrary directory 2011-08-03 14:25:20 -10:00
blk.h Merge branch 'for-3.5' of ../cgroup into block/for-3.5/core-merged 2012-04-01 12:55:00 -07:00
bsg-lib.c block: Change module.h -> export.h in bsg-lib.c 2011-10-31 19:31:13 -04:00
bsg.c bsg: fix sysfs link remove warning 2012-02-08 20:02:03 +01:00
cfq-iosched.c block: blkcg_policy_cfq shouldn't be used if !CONFIG_CFQ_GROUP_IOSCHED 2012-06-04 10:02:29 +02:00
compat_ioctl.c block: Add BLKROTATIONAL ioctl 2012-01-11 16:29:31 +01:00
deadline-iosched.c elevator: make elevator_init_fn() return 0/-errno 2012-03-06 21:27:21 +01:00
elevator.c blkcg: implement per-queue policy activation 2012-04-20 10:06:06 +02:00
genhd.c block: fix buffer overflow when printing partition UUIDs 2012-05-15 08:22:04 +02:00
ioctl.c Merge branch 'for-3.3/core' of git://git.kernel.dk/linux-block 2012-01-15 12:24:45 -08:00
Kconfig move fs/partitions to block/ 2012-01-03 22:54:06 -05:00
Kconfig.iosched blkcg: make CONFIG_BLK_CGROUP bool 2012-03-06 21:27:21 +01:00
Makefile separate partition format handling from generic code 2012-01-03 22:54:06 -05:00
noop-iosched.c elevator: make elevator_init_fn() return 0/-errno 2012-03-06 21:27:21 +01:00
partition-generic.c block: Fix NULL pointer dereference in sd_revalidate_disk 2012-03-02 10:38:33 +01:00
scsi_ioctl.c block: fail SCSI passthrough ioctls on partition devices 2012-01-14 15:07:24 -08:00