From mboxrd@z Thu Jan 1 00:00:00 1970 Received: from eggs.gnu.org ([2001:4830:134:3::10]:35295) by lists.gnu.org with esmtp (Exim 4.71) (envelope-from ) id 1emG3Z-0001zn-D6 for qemu-devel@nongnu.org; Thu, 15 Feb 2018 04:42:25 -0500 Received: from Debian-exim by eggs.gnu.org with spam-scanned (Exim 4.71) (envelope-from ) id 1emG3Y-00007L-El for qemu-devel@nongnu.org; Thu, 15 Feb 2018 04:42:21 -0500 From: Stefan Hajnoczi Date: Thu, 15 Feb 2018 09:42:08 +0000 Message-Id: <20180215094210.27770-2-stefanha@redhat.com> In-Reply-To: <20180215094210.27770-1-stefanha@redhat.com> References: <20180215094210.27770-1-stefanha@redhat.com> Subject: [Qemu-devel] [PULL v2 1/3] vl: pause vcpus before stopping iothreads List-Id: List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , To: qemu-devel@nongnu.org Cc: qemu-block@nongnu.org, Peter Maydell , Stefan Hajnoczi Commit dce8921b2baaf95974af8176406881872067adfa ("iothread: Stop threads before main() quits") introduced iothread_stop_all() to avoid the following virtio-scsi assertion failure: assert(blk_get_aio_context(d->conf.blk) == s->ctx); Back then the assertion failed because when bdrv_close_all() made d->conf.blk NULL, blk_get_aio_context() returned the global AioContext instead of s->ctx. The same assertion can still fail today when vcpus submit new I/O requests after iothread_stop_all() has moved the BDS to the global AioContext. This patch hardens the iothread_stop_all() approach by pausing vcpus before calling iothread_stop_all(). Note that the assertion failure is a race condition. It is not possible to reproduce it reliably. Signed-off-by: Stefan Hajnoczi Message-id: 20180201110708.8080-1-stefanha@redhat.com Signed-off-by: Stefan Hajnoczi --- vl.c | 12 ++++++++++-- 1 file changed, 10 insertions(+), 2 deletions(-) diff --git a/vl.c b/vl.c index 21878496ec..7a5554bc41 100644 --- a/vl.c +++ b/vl.c @@ -4767,10 +4767,18 @@ int main(int argc, char **argv, char **envp) main_loop(); replay_disable_events(); + + /* The ordering of the following is delicate. Stop vcpus to prevent new + * I/O requests being queued by the guest. Then stop IOThreads (this + * includes a drain operation and completes all request processing). At + * this point emulated devices are still associated with their IOThreads + * (if any) but no longer have any work to do. Only then can we close + * block devices safely because we know there is no more I/O coming. + */ + pause_all_vcpus(); iothread_stop_all(); - - pause_all_vcpus(); bdrv_close_all(); + res_free(); /* vhost-user must be cleaned up before chardevs. */ -- 2.14.3