[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
[Qemu-block] [PULL 06/61] migration: avoid recursive AioContext locking
From: |
Kevin Wolf |
Subject: |
[Qemu-block] [PULL 06/61] migration: avoid recursive AioContext locking in save_vmstate() |
Date: |
Fri, 23 Jun 2017 18:21:04 +0200 |
From: Stefan Hajnoczi <address@hidden>
AioContext was designed to allow nested acquire/release calls. It uses
a recursive mutex so callers don't need to worry about nesting...or so
we thought.
BDRV_POLL_WHILE() is used to wait for block I/O requests. It releases
the AioContext temporarily around aio_poll(). This gives IOThreads a
chance to acquire the AioContext to process I/O completions.
It turns out that recursive locking and BDRV_POLL_WHILE() don't mix.
BDRV_POLL_WHILE() only releases the AioContext once, so the IOThread
will not be able to acquire the AioContext if it was acquired
multiple times.
Instead of trying to release AioContext n times in BDRV_POLL_WHILE(),
this patch simply avoids nested locking in save_vmstate(). It's the
simplest fix and we should step back to consider the big picture with
all the recent changes to block layer threading.
This patch is the final fix to solve 'savevm' hanging with -object
iothread.
Signed-off-by: Stefan Hajnoczi <address@hidden>
Reviewed-by: Eric Blake <address@hidden>
Reviewed-by: Paolo Bonzini <address@hidden>
Signed-off-by: Kevin Wolf <address@hidden>
---
migration/savevm.c | 12 +++++++++++-
1 file changed, 11 insertions(+), 1 deletion(-)
diff --git a/migration/savevm.c b/migration/savevm.c
index 6bfd489..5846d9c 100644
--- a/migration/savevm.c
+++ b/migration/savevm.c
@@ -2144,6 +2144,14 @@ int save_snapshot(const char *name, Error **errp)
goto the_end;
}
+ /* The bdrv_all_create_snapshot() call that follows acquires the AioContext
+ * for itself. BDRV_POLL_WHILE() does not support nested locking because
+ * it only releases the lock once. Therefore synchronous I/O will deadlock
+ * unless we release the AioContext before bdrv_all_create_snapshot().
+ */
+ aio_context_release(aio_context);
+ aio_context = NULL;
+
ret = bdrv_all_create_snapshot(sn, bs, vm_state_size, &bs);
if (ret < 0) {
error_setg(errp, "Error while creating snapshot on '%s'",
@@ -2154,7 +2162,9 @@ int save_snapshot(const char *name, Error **errp)
ret = 0;
the_end:
- aio_context_release(aio_context);
+ if (aio_context) {
+ aio_context_release(aio_context);
+ }
if (saved_vm_running) {
vm_start();
}
--
1.8.3.1
- [Qemu-block] [PULL 00/61] Block layer patches, Kevin Wolf, 2017/06/23
- [Qemu-block] [PULL 01/61] commit: Fix completion with extra reference, Kevin Wolf, 2017/06/23
- [Qemu-block] [PULL 02/61] qemu-iotests: Allow starting new qemu after cleanup, Kevin Wolf, 2017/06/23
- [Qemu-block] [PULL 04/61] block: count bdrv_co_rw_vmstate() requests, Kevin Wolf, 2017/06/23
- [Qemu-block] [PULL 03/61] qemu-iotests: Test exiting qemu with running job, Kevin Wolf, 2017/06/23
- [Qemu-block] [PULL 05/61] block: use BDRV_POLL_WHILE() in bdrv_rw_vmstate(), Kevin Wolf, 2017/06/23
- [Qemu-block] [PULL 06/61] migration: avoid recursive AioContext locking in save_vmstate(),
Kevin Wolf <=
- [Qemu-block] [PULL 07/61] migration: use bdrv_drain_all_begin/end() instead bdrv_drain_all(), Kevin Wolf, 2017/06/23
- [Qemu-block] [PULL 10/61] throttle: Update throttle-groups.c documentation, Kevin Wolf, 2017/06/23
- [Qemu-block] [PULL 09/61] doc: Document driver-specific -blockdev options, Kevin Wolf, 2017/06/23
- [Qemu-block] [PULL 08/61] doc: Document generic -blockdev options, Kevin Wolf, 2017/06/23
- [Qemu-block] [PULL 12/61] migration: hold AioContext lock for loadvm qemu_fclose(), Kevin Wolf, 2017/06/23
- [Qemu-block] [PULL 13/61] qemu-iotests: 068: extract _qemu() function, Kevin Wolf, 2017/06/23
- [Qemu-block] [PULL 14/61] qemu-iotests: 068: use -drive/-device instead of -hda, Kevin Wolf, 2017/06/23
- [Qemu-block] [PULL 11/61] virtio-pci: use ioeventfd even when KVM is disabled, Kevin Wolf, 2017/06/23