Blob Blame Raw
From ecfbec2139f06a75a03c510e24f265a869684edb Mon Sep 17 00:00:00 2001
From: Asias He <asias@redhat.com>
Date: Thu, 12 Sep 2013 07:39:35 +0200
Subject: [PATCH 15/29] gluster: Abort on AIO completion failure

RH-Author: Asias He <asias@redhat.com>
Message-id: <1378971575-22416-5-git-send-email-asias@redhat.com>
Patchwork-id: 54324
O-Subject: [RHEL7.0 qemu-kvm PATCH 4/4] gluster: Abort on AIO completion failure
Bugzilla: 1007226
RH-Acked-by: Stefan Hajnoczi <stefanha@redhat.com>
RH-Acked-by: Paolo Bonzini <pbonzini@redhat.com>
RH-Acked-by: Fam Zheng <famz@redhat.com>

From: Bharata B Rao <bharata@linux.vnet.ibm.com>

BZ: https://bugzilla.redhat.com/show_bug.cgi?id=1004733
Brew: https://brewweb.devel.redhat.com/taskinfo?taskID=6275752

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>
(cherry picked from commit 9faa574f7d07109e2256c0b4b63e8711d650f2d8)

Conflicts:
	block/gluster.c
---
 block/gluster.c | 17 ++---------------
 1 file changed, 2 insertions(+), 15 deletions(-)

Signed-off-by: Miroslav Rezanina <mrezanin@redhat.com>
---
 block/gluster.c |   17 ++---------------
 1 files changed, 2 insertions(+), 15 deletions(-)

diff --git a/block/gluster.c b/block/gluster.c
index 7b10e9c..90a51ca 100644
--- a/block/gluster.c
+++ b/block/gluster.c
@@ -436,22 +436,9 @@ static void gluster_finish_aiocb(struct glfs_fd *fd, ssize_t ret, void *arg)
         /*
          * Gluster AIO callback thread failed to notify the waiting
          * QEMU thread about IO completion.
-         *
-         * Complete this IO request and make the disk inaccessible for
-         * subsequent reads and writes.
          */
-        error_report("Gluster failed to notify QEMU about IO completion");
-
-        qemu_mutex_lock_iothread(); /* We are in gluster thread context */
-        acb->common.cb(acb->common.opaque, -EIO);
-        qemu_aio_release(acb);
-        s->qemu_aio_count--;
-        close(s->fds[GLUSTER_FD_READ]);
-        close(s->fds[GLUSTER_FD_WRITE]);
-        qemu_aio_set_fd_handler(s->fds[GLUSTER_FD_READ], NULL, NULL, NULL,
-            NULL);
-        bs->drv = NULL; /* Make the disk inaccessible */
-        qemu_mutex_unlock_iothread();
+        error_report("Gluster AIO completion failed: %s", strerror(errno));
+        abort();
     }
 }
 
-- 
1.7.1