From mboxrd@z Thu Jan 1 00:00:00 1970 Received: from eggs.gnu.org ([2001:4830:134:3::10]:53539) by lists.gnu.org with esmtp (Exim 4.71) (envelope-from ) id 1fyJQn-0006Pm-77 for qemu-devel@nongnu.org; Fri, 07 Sep 2018 12:16:32 -0400 Received: from Debian-exim by eggs.gnu.org with spam-scanned (Exim 4.71) (envelope-from ) id 1fyJQg-0004Un-HP for qemu-devel@nongnu.org; Fri, 07 Sep 2018 12:16:19 -0400 From: Kevin Wolf Date: Fri, 7 Sep 2018 18:15:20 +0200 Message-Id: <20180907161520.26349-15-kwolf@redhat.com> In-Reply-To: <20180907161520.26349-1-kwolf@redhat.com> References: <20180907161520.26349-1-kwolf@redhat.com> Subject: [Qemu-devel] [PATCH 14/14] test-bdrv-drain: Test nested poll in bdrv_drain_poll_top_level() List-Id: List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , To: qemu-block@nongnu.org Cc: kwolf@redhat.com, mreitz@redhat.com, famz@redhat.com, pbonzini@redhat.com, slp@redhat.com, qemu-devel@nongnu.org This is a regression test for a deadlock that could occur in callbacks called from the aio_poll() in bdrv_drain_poll_top_level(). The AioContext lock wasn't released and therefore would be taken a second time in the callback. This would cause a possible AIO_WAIT_WHILE() in the callback to hang. Signed-off-by: Kevin Wolf --- tests/test-bdrv-drain.c | 13 +++++++++++++ 1 file changed, 13 insertions(+) diff --git a/tests/test-bdrv-drain.c b/tests/test-bdrv-drain.c index 9641a20dd8..07ce5bfddc 100644 --- a/tests/test-bdrv-drain.c +++ b/tests/test-bdrv-drain.c @@ -636,6 +636,17 @@ static void test_iothread_aio_cb(void *opaque, int ret) qemu_event_set(&done_event); } +static void test_iothread_main_thread_bh(void *opaque) +{ + struct test_iothread_data *data = opaque; + + /* Test that the AioContext is not yet locked in a random BH that is + * executed during drain, otherwise this would deadlock. */ + aio_context_acquire(bdrv_get_aio_context(data->bs)); + bdrv_flush(data->bs); + aio_context_release(bdrv_get_aio_context(data->bs)); +} + /* * Starts an AIO request on a BDS that runs in the AioContext of iothread 1. * The request involves a BH on iothread 2 before it can complete. @@ -705,6 +716,8 @@ static void test_iothread_common(enum drain_type drain_type, int drain_thread) aio_context_acquire(ctx_a); } + aio_bh_schedule_oneshot(ctx_a, test_iothread_main_thread_bh, &data); + /* The request is running on the IOThread a. Draining its block device * will make sure that it has completed as far as the BDS is concerned, * but the drain in this thread can continue immediately after -- 2.13.6