Fixes: https://bugzilla.redhat.com/show_bug.cgi?id=1447551
If one tries to issue a block_resize while a guest is busy
accessing the disk, it is possible that qemu may deadlock
when invoking aio_poll from both the main loop and the iothread.
Replace another instance of bdrv_drain_all that doesn't
quite belong.
Suggested-by: Paolo Bonzini <pbonzini@redhat.com>
Signed-off-by: John Snow <jsnow@redhat.com>
---
blockdev.c | 5 ++---
1 file changed, 2 insertions(+), 3 deletions(-)
diff --git a/blockdev.c b/blockdev.c
index 0b38c3d..9ff06f3 100644
--- a/blockdev.c
+++ b/blockdev.c
@@ -2923,10 +2923,9 @@ void qmp_block_resize(bool has_device, const char *device,
goto out;
}
- /* complete all in-flight operations before resizing the device */
- bdrv_drain_all();
-
+ bdrv_drained_begin(bs);
ret = blk_truncate(blk, size, errp);
+ bdrv_drained_end(bs);
out:
blk_unref(blk);
--
2.9.3
On 05/10/2017 12:39 PM, John Snow wrote: > Fixes: https://bugzilla.redhat.com/show_bug.cgi?id=1447551 > > If one tries to issue a block_resize while a guest is busy > accessing the disk, it is possible that qemu may deadlock > when invoking aio_poll from both the main loop and the iothread. > > Replace another instance of bdrv_drain_all that doesn't > quite belong. > > Suggested-by: Paolo Bonzini <pbonzini@redhat.com> > Signed-off-by: John Snow <jsnow@redhat.com> > --- > blockdev.c | 5 ++--- > 1 file changed, 2 insertions(+), 3 deletions(-) > Reviewed-by: Eric Blake <eblake@redhat.com> > diff --git a/blockdev.c b/blockdev.c > index 0b38c3d..9ff06f3 100644 > --- a/blockdev.c > +++ b/blockdev.c > @@ -2923,10 +2923,9 @@ void qmp_block_resize(bool has_device, const char *device, > goto out; > } > > - /* complete all in-flight operations before resizing the device */ > - bdrv_drain_all(); > - > + bdrv_drained_begin(bs); > ret = blk_truncate(blk, size, errp); > + bdrv_drained_end(bs); > > out: > blk_unref(blk); > -- Eric Blake, Principal Software Engineer Red Hat, Inc. +1-919-301-3266 Virtualization: qemu.org | libvirt.org
On 10/05/2017 19:39, John Snow wrote: > Fixes: https://bugzilla.redhat.com/show_bug.cgi?id=1447551 > > If one tries to issue a block_resize while a guest is busy > accessing the disk, it is possible that qemu may deadlock > when invoking aio_poll from both the main loop and the iothread. > > Replace another instance of bdrv_drain_all that doesn't > quite belong. > > Suggested-by: Paolo Bonzini <pbonzini@redhat.com> > Signed-off-by: John Snow <jsnow@redhat.com> > --- > blockdev.c | 5 ++--- > 1 file changed, 2 insertions(+), 3 deletions(-) > > diff --git a/blockdev.c b/blockdev.c > index 0b38c3d..9ff06f3 100644 > --- a/blockdev.c > +++ b/blockdev.c > @@ -2923,10 +2923,9 @@ void qmp_block_resize(bool has_device, const char *device, > goto out; > } > > - /* complete all in-flight operations before resizing the device */ > - bdrv_drain_all(); > - > + bdrv_drained_begin(bs); > ret = blk_truncate(blk, size, errp); > + bdrv_drained_end(bs); > > out: > blk_unref(blk); > Reviewed-by: Paolo Bonzini <pbonzini@redhat.com>
Am 10.05.2017 um 19:39 hat John Snow geschrieben: > Fixes: https://bugzilla.redhat.com/show_bug.cgi?id=1447551 > > If one tries to issue a block_resize while a guest is busy > accessing the disk, it is possible that qemu may deadlock > when invoking aio_poll from both the main loop and the iothread. > > Replace another instance of bdrv_drain_all that doesn't > quite belong. > > Suggested-by: Paolo Bonzini <pbonzini@redhat.com> > Signed-off-by: John Snow <jsnow@redhat.com> Thanks, applied to the block branch. Kevin
© 2016 - 2024 Red Hat, Inc.