gluster: Abort on AIO completion failure
authorBharata B Rao <bharata@linux.vnet.ibm.com>
Tue, 27 Aug 2013 08:15:41 +0000 (13:45 +0530)
committerKevin Wolf <kwolf@redhat.com>
Fri, 30 Aug 2013 13:28:52 +0000 (15:28 +0200)
commit9faa574f7d07109e2256c0b4b63e8711d650f2d8
tree77ae2cd916d5c3cbfe00a21783419fac47cd5bb7
parente5b1d99f5528315dc77aab369ae060d7cbad1e2a
gluster: Abort on AIO completion failure

Currently if gluster AIO callback thread fails to notify the QEMU thread about
AIO completion, we try graceful recovery by marking the disk drive as
inaccessible. This error recovery code is race-prone as found by Asias and
Stefan. However as found out by Paolo, this kind of error is impossible and
hence simplify the code that handles this error recovery.

Signed-off-by: Bharata B Rao <bharata@linux.vnet.ibm.com>
Reviewed-by: Stefan Hajnoczi <stefanha@redhat.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
block/gluster.c