qemu/include
Fam Zheng a77fd4bb29 block: Fix bdrv_drain in coroutine
Using the nested aio_poll() in coroutine is a bad idea. This patch
replaces the aio_poll loop in bdrv_drain with a BH, if called in
coroutine.

For example, the bdrv_drain() in mirror.c can hang when a guest issued
request is pending on it in qemu_co_mutex_lock().

Mirror coroutine in this case has just finished a request, and the block
job is about to complete. It calls bdrv_drain() which waits for the
other coroutine to complete. The other coroutine is a scsi-disk request.
The deadlock happens when the latter is in turn pending on the former to
yield/terminate, in qemu_co_mutex_lock(). The state flow is as below
(assuming a qcow2 image):

  mirror coroutine               scsi-disk coroutine
  -------------------------------------------------------------
  do last write

    qcow2:qemu_co_mutex_lock()
    ...
                                 scsi disk read

                                   tracked request begin

                                   qcow2:qemu_co_mutex_lock.enter

    qcow2:qemu_co_mutex_unlock()

  bdrv_drain
    while (has tracked request)
      aio_poll()

In the scsi-disk coroutine, the qemu_co_mutex_lock() will never return
because the mirror coroutine is blocked in the aio_poll(blocking=true).

With this patch, the added qemu_coroutine_yield() allows the scsi-disk
coroutine to make progress as expected:

  mirror coroutine               scsi-disk coroutine
  -------------------------------------------------------------
  do last write

    qcow2:qemu_co_mutex_lock()
    ...
                                 scsi disk read

                                   tracked request begin

                                   qcow2:qemu_co_mutex_lock.enter

    qcow2:qemu_co_mutex_unlock()

  bdrv_drain.enter
>   schedule BH
>   qemu_coroutine_yield()
>                                  qcow2:qemu_co_mutex_lock.return
>                                  ...
                                   tracked request end
    ...
    (resumed from BH callback)
  bdrv_drain.return
  ...

Reported-by: Laurent Vivier <lvivier@redhat.com>
Signed-off-by: Fam Zheng <famz@redhat.com>
Message-id: 1459855253-5378-2-git-send-email-famz@redhat.com
Suggested-by: Paolo Bonzini <pbonzini@redhat.com>
Signed-off-by: Fam Zheng <famz@redhat.com>
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
2016-04-11 16:59:09 +01:00
..
block block: Fix bdrv_drain in coroutine 2016-04-11 16:59:09 +01:00
crypto include/crypto: Include qapi-types.h or qemu/bswap.h instead of qemu-common.h 2016-03-22 22:20:16 +01:00
disas disas: Add print_insn to disassemble info 2015-07-09 15:20:40 +02:00
exec qemu-log: dfilter-ise exec, out_asm, op and opt_op 2016-03-22 22:20:18 +01:00
fpu include: Clean up includes 2016-02-23 12:43:05 +00:00
hw virtio: merge virtio_queue_aio_set_host_notifier_handler with virtio_queue_set_aio 2016-04-07 19:57:33 +03:00
io io: implement socket watch for win32 using WSAEventSelect+select 2016-03-10 17:19:07 +00:00
libdecnumber include: Clean up includes 2016-02-23 12:43:05 +00:00
migration migration (ordinary): move bdrv_invalidate_cache_all of of coroutine context 2016-02-26 20:39:50 +05:30
monitor Use scripts/clean-includes to drop redundant qemu/typedefs.h 2016-03-22 22:20:16 +01:00
net Use scripts/clean-includes to drop redundant qemu/typedefs.h 2016-03-22 22:20:16 +01:00
qapi Use scripts/clean-includes to drop redundant qemu/typedefs.h 2016-03-22 22:20:16 +01:00
qemu * FreeBSD build fixes (atomics, qapi/error.h) 2016-04-05 11:03:18 +01:00
qom * Log filtering from Alex and Peter 2016-03-24 21:42:40 +00:00
standard-headers update Linux headers to 4.6 2016-04-05 11:46:52 +02:00
sysemu kvm: add kvm_device_supported() helper function 2016-03-30 17:27:24 +01:00
ui ui/virtio-gpu: add and use qemu_create_displaysurface_pixman 2016-04-11 12:32:01 +02:00
elf.h include: Clean up includes 2016-02-23 12:43:05 +00:00
glib-compat.h glib: add compatibility interface for g_hash_table_add() 2015-10-30 09:01:03 +01:00
qemu-common.h util: move declarations out of qemu-common.h 2016-03-22 22:20:17 +01:00
qemu-io.h qemu-io: Use BlockBackend 2015-02-16 15:07:19 +00:00
qjson.h QJSON: Add JSON writer 2015-02-05 17:16:14 +01:00
trace-tcg.h trace: [tcg] Generate TCG tracing routines 2014-08-12 14:26:12 +01:00
trace.h trace: [tcg] Include event definitions in "trace.h" 2014-08-12 14:26:12 +01:00