[Qemu-devel] [PATCH 1/1] virtio-blk: fix race on guest notifiers

[Halil Pasic]

The commits 03de2f527 "virtio-blk: do not use vring in dataplane"  and
9ffe337c08 "virtio-blk: always use dataplane path if ioeventfd is active"
changed how notifications are done for virtio-blk substantially. Due to a
race condition, interrupts are lost when irqfd behind the guest notifier
is torn down after notify_guest_bh was scheduled but before it actually
runs.

Let's fix this by forcing guest notifications before cleaning up the
irqfd's. Let's also add some explanatory comments.

Cc: address@hidden
Signed-off-by: Halil Pasic <address@hidden>
Reported-by: Michael A. Tebolt <address@hidden>
Suggested-by: Paolo Bonzini <address@hidden>
---

This patch supersedes my previous attempt to fix the same issue which was
running under the title "virtio: fallback from irqfd to non-irqfd
notify".

Tested it briefly without iothread and  over the night with iothread, but
more testing would not hurt.

@Paolo: Could you please check the comments? They reflect my
understanding which may be very wrong.
---
 hw/block/dataplane/virtio-blk.c | 9 ++++++++-
 1 file changed, 8 insertions(+), 1 deletion(-)

diff --git a/hw/block/dataplane/virtio-blk.c b/hw/block/dataplane/virtio-blk.c
index 5556f0e..13dd14d 100644
--- a/hw/block/dataplane/virtio-blk.c
+++ b/hw/block/dataplane/virtio-blk.c
@@ -258,9 +258,16 @@ void virtio_blk_data_plane_stop(VirtIODevice *vdev)
         virtio_queue_aio_set_host_notifier_handler(vq, s->ctx, NULL);
     }
 
-    /* Drain and switch bs back to the QEMU main loop */
+    /* Drain and switch bs back to the QEMU main loop. After drain, the
+     * device will not submit (nor comple) any requests until dataplane
+     * starts again.
+     */
     blk_set_aio_context(s->conf->conf.blk, qemu_get_aio_context());
 
+    /* Notify guest before the guest notifiers get cleaned up */
+    qemu_bh_cancel(s->bh);
+    notify_guest_bh(s);
+
     aio_context_release(s->ctx);
 
     for (i = 0; i < nvqs; i++) {
-- 
2.8.4