From mboxrd@z Thu Jan 1 00:00:00 1970 Received: from eggs.gnu.org ([209.51.188.92]:47685) by lists.gnu.org with esmtp (Exim 4.71) (envelope-from ) id 1hIlp4-00065p-Qd for qemu-devel@nongnu.org; Mon, 22 Apr 2019 23:10:20 -0400 Received: from Debian-exim by eggs.gnu.org with spam-scanned (Exim 4.71) (envelope-from ) id 1hIleJ-0000tK-Eg for qemu-devel@nongnu.org; Mon, 22 Apr 2019 22:59:12 -0400 References: <20190416184624.15397-1-dan.streetman@canonical.com> <20190416184624.15397-2-dan.streetman@canonical.com> <9f3dac2a-15fe-8463-6aee-f6916b8d5e1c@redhat.com> From: Jason Wang Message-ID: Date: Tue, 23 Apr 2019 10:58:57 +0800 MIME-Version: 1.0 In-Reply-To: Content-Type: text/plain; charset=utf-8; format=flowed Content-Language: en-US Content-Transfer-Encoding: quoted-printable Subject: Re: [Qemu-devel] [PATCH 1/2] add VirtIONet vhost_stopped flag to prevent multiple stops List-Id: List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , To: Dan Streetman Cc: "Michael S. Tsirkin" , qemu-devel@nongnu.org, qemu-stable@nongnu.org, =?UTF-8?Q?marcandre=2elureau=40redhat=2ecom_=3e=3e_Marc-Andr=c3=a9_Lureau?= On 2019/4/23 =E4=B8=8A=E5=8D=884:14, Dan Streetman wrote: > On Sun, Apr 21, 2019 at 10:50 PM Jason Wang wrote= : >> >> On 2019/4/17 =E4=B8=8A=E5=8D=882:46, Dan Streetman wrote: >>> From: Dan Streetman >>> >>> Buglink: https://launchpad.net/bugs/1823458 >>> >>> There is a race condition when using the vhost-user driver, between a= guest >>> shutdown and the vhost-user interface being closed. This is explaine= d in >>> more detail at the bug link above; the short explanation is the vhost= -user >>> device can be closed while the main thread is in the middle of stoppi= ng >>> the vhost_net. In this case, the main thread handling shutdown will >>> enter virtio_net_vhost_status() and move into the n->vhost_started (e= lse) >>> block, and call vhost_net_stop(); while it is running that function, >>> another thread is notified that the vhost-user device has been closed= , >>> and (indirectly) calls into virtio_net_vhost_status() also. >> >> I think we need figure out why there are multiple vhost_net_stop() cal= ls >> simultaneously. E.g vhost-user register fd handlers like: >> >> qemu_chr_fe_set_handlers(&s->chr, NULL, NULL, >> net_vhost_user_event, NULL, nc0->na= me, >> NULL, >> true); >> >> which uses default main context, so it should only be called only in >> main thread. > net_vhost_user_event() schedules chr_closed_bh() to do its bottom half > work; does aio_bh_schedule_oneshot() execute its events from the main > thread? I think so if net_vhost_user_event() was called in main thread (it calls=20 qemu_get_current_aio_context()). > > For reference, the call chain is: > > chr_closed_bh() > qmp_set_link() > nc->info->link_status_changed() -> virtio_net_set_link_status() > virtio_net_set_status() > virtio_net_vhost_status() The code was added by Marc since: commit e7c83a885f865128ae3cf1946f8cb538b63cbfba Author: Marc-Andr=C3=A9 Lureau Date:=C2=A0=C2=A0 Mon Feb 27 14:49:56 2017 +0400 =C2=A0=C2=A0=C2=A0 vhost-user: delay vhost_user_stop Cc him for more thoughts. Thanks >> Thanks >> >> >>> Since the >>> vhost_net status hasn't yet changed, the second thread also enters >>> the n->vhost_started block, and also calls vhost_net_stop(). This >>> causes problems for the second thread when it tries to stop the netwo= rk >>> that's already been stopped. >>> >>> This adds a flag to the struct that's atomically set to prevent more = than >>> one thread from calling vhost_net_stop(). The atomic_fetch_inc() is = likely >>> overkill and probably could be done with a simple check-and-set, but >>> since it's a race condition there would still be a (very, very) small >>> window without using an atomic to set it. >>> >>> Signed-off-by: Dan Streetman >>> --- >>> hw/net/virtio-net.c | 3 ++- >>> include/hw/virtio/virtio-net.h | 1 + >>> 2 files changed, 3 insertions(+), 1 deletion(-) >>> >>> diff --git a/hw/net/virtio-net.c b/hw/net/virtio-net.c >>> index ffe0872fff..d36f50d5dd 100644 >>> --- a/hw/net/virtio-net.c >>> +++ b/hw/net/virtio-net.c >>> @@ -13,6 +13,7 @@ >>> >>> #include "qemu/osdep.h" >>> #include "qemu/iov.h" >>> +#include "qemu/atomic.h" >>> #include "hw/virtio/virtio.h" >>> #include "net/net.h" >>> #include "net/checksum.h" >>> @@ -240,7 +241,7 @@ static void virtio_net_vhost_status(VirtIONet *n,= uint8_t status) >>> "falling back on userspace virtio", -r); >>> n->vhost_started =3D 0; >>> } >>> - } else { >>> + } else if (atomic_fetch_inc(&n->vhost_stopped) =3D=3D 0) { >>> vhost_net_stop(vdev, n->nic->ncs, queues); >>> n->vhost_started =3D 0; >>> } >>> diff --git a/include/hw/virtio/virtio-net.h b/include/hw/virtio/virti= o-net.h >>> index b96f0c643f..d03fd933d0 100644 >>> --- a/include/hw/virtio/virtio-net.h >>> +++ b/include/hw/virtio/virtio-net.h >>> @@ -164,6 +164,7 @@ struct VirtIONet { >>> uint8_t nouni; >>> uint8_t nobcast; >>> uint8_t vhost_started; >>> + int vhost_stopped; >>> struct { >>> uint32_t in_use; >>> uint32_t first_multi; From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: X-Spam-Checker-Version: SpamAssassin 3.4.0 (2014-02-07) on aws-us-west-2-korg-lkml-1.web.codeaurora.org X-Spam-Level: X-Spam-Status: No, score=-6.9 required=3.0 tests=HEADER_FROM_DIFFERENT_DOMAINS, INCLUDES_PATCH,MAILING_LIST_MULTI,SIGNED_OFF_BY,SPF_PASS,URIBL_BLOCKED autolearn=ham autolearn_force=no version=3.4.0 Received: from mail.kernel.org (mail.kernel.org [198.145.29.99]) by smtp.lore.kernel.org (Postfix) with ESMTP id 78A0FC10F11 for ; Tue, 23 Apr 2019 03:11:26 +0000 (UTC) Received: from lists.gnu.org (lists.gnu.org [209.51.188.17]) (using TLSv1 with cipher AES256-SHA (256/256 bits)) (No client certificate requested) by mail.kernel.org (Postfix) with ESMTPS id 39995206BA for ; Tue, 23 Apr 2019 03:11:26 +0000 (UTC) DMARC-Filter: OpenDMARC Filter v1.3.2 mail.kernel.org 39995206BA Authentication-Results: mail.kernel.org; dmarc=fail (p=none dis=none) header.from=redhat.com Authentication-Results: mail.kernel.org; spf=pass smtp.mailfrom=qemu-devel-bounces+qemu-devel=archiver.kernel.org@nongnu.org Received: from localhost ([127.0.0.1]:47291 helo=lists.gnu.org) by lists.gnu.org with esmtp (Exim 4.71) (envelope-from ) id 1hIlq9-0006tx-5U for qemu-devel@archiver.kernel.org; Mon, 22 Apr 2019 23:11:25 -0400 Received: from eggs.gnu.org ([209.51.188.92]:47685) by lists.gnu.org with esmtp (Exim 4.71) (envelope-from ) id 1hIlp4-00065p-Qd for qemu-devel@nongnu.org; Mon, 22 Apr 2019 23:10:20 -0400 Received: from Debian-exim by eggs.gnu.org with spam-scanned (Exim 4.71) (envelope-from ) id 1hIleJ-0000tK-Eg for qemu-devel@nongnu.org; Mon, 22 Apr 2019 22:59:12 -0400 Received: from mx1.redhat.com ([209.132.183.28]:50320) by eggs.gnu.org with esmtps (TLS1.0:DHE_RSA_AES_256_CBC_SHA1:32) (Exim 4.71) (envelope-from ) id 1hIleJ-0000sr-6u; Mon, 22 Apr 2019 22:59:11 -0400 Received: from smtp.corp.redhat.com (int-mx03.intmail.prod.int.phx2.redhat.com [10.5.11.13]) (using TLSv1.2 with cipher AECDH-AES256-SHA (256/256 bits)) (No client certificate requested) by mx1.redhat.com (Postfix) with ESMTPS id 817E159461; Tue, 23 Apr 2019 02:59:09 +0000 (UTC) Received: from [10.72.12.166] (ovpn-12-166.pek2.redhat.com [10.72.12.166]) by smtp.corp.redhat.com (Postfix) with ESMTP id 9CBAB183F3; Tue, 23 Apr 2019 02:58:59 +0000 (UTC) To: Dan Streetman References: <20190416184624.15397-1-dan.streetman@canonical.com> <20190416184624.15397-2-dan.streetman@canonical.com> <9f3dac2a-15fe-8463-6aee-f6916b8d5e1c@redhat.com> From: Jason Wang Message-ID: Date: Tue, 23 Apr 2019 10:58:57 +0800 User-Agent: Mozilla/5.0 (X11; Linux x86_64; rv:60.0) Gecko/20100101 Thunderbird/60.6.1 MIME-Version: 1.0 In-Reply-To: Content-Type: text/plain; charset="UTF-8"; format="flowed" Content-Language: en-US X-Scanned-By: MIMEDefang 2.79 on 10.5.11.13 X-Greylist: Sender IP whitelisted, not delayed by milter-greylist-4.5.16 (mx1.redhat.com [10.5.110.39]); Tue, 23 Apr 2019 02:59:09 +0000 (UTC) Content-Transfer-Encoding: quoted-printable X-detected-operating-system: by eggs.gnu.org: GNU/Linux 2.2.x-3.x [generic] X-Received-From: 209.132.183.28 Subject: Re: [Qemu-devel] [PATCH 1/2] add VirtIONet vhost_stopped flag to prevent multiple stops X-BeenThere: qemu-devel@nongnu.org X-Mailman-Version: 2.1.21 Precedence: list List-Id: List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Cc: =?UTF-8?Q?marcandre=2elureau=40redhat=2ecom_=3e=3e_Marc-Andr=c3=a9_Lureau?= , qemu-stable@nongnu.org, qemu-devel@nongnu.org, "Michael S. Tsirkin" Errors-To: qemu-devel-bounces+qemu-devel=archiver.kernel.org@nongnu.org Sender: "Qemu-devel" Message-ID: <20190423025857._HXqbTGpCsyKiiPA5a9_Xu2sJUZGAut2otnTd_l1sl0@z> On 2019/4/23 =E4=B8=8A=E5=8D=884:14, Dan Streetman wrote: > On Sun, Apr 21, 2019 at 10:50 PM Jason Wang wrote= : >> >> On 2019/4/17 =E4=B8=8A=E5=8D=882:46, Dan Streetman wrote: >>> From: Dan Streetman >>> >>> Buglink: https://launchpad.net/bugs/1823458 >>> >>> There is a race condition when using the vhost-user driver, between a= guest >>> shutdown and the vhost-user interface being closed. This is explaine= d in >>> more detail at the bug link above; the short explanation is the vhost= -user >>> device can be closed while the main thread is in the middle of stoppi= ng >>> the vhost_net. In this case, the main thread handling shutdown will >>> enter virtio_net_vhost_status() and move into the n->vhost_started (e= lse) >>> block, and call vhost_net_stop(); while it is running that function, >>> another thread is notified that the vhost-user device has been closed= , >>> and (indirectly) calls into virtio_net_vhost_status() also. >> >> I think we need figure out why there are multiple vhost_net_stop() cal= ls >> simultaneously. E.g vhost-user register fd handlers like: >> >> qemu_chr_fe_set_handlers(&s->chr, NULL, NULL, >> net_vhost_user_event, NULL, nc0->na= me, >> NULL, >> true); >> >> which uses default main context, so it should only be called only in >> main thread. > net_vhost_user_event() schedules chr_closed_bh() to do its bottom half > work; does aio_bh_schedule_oneshot() execute its events from the main > thread? I think so if net_vhost_user_event() was called in main thread (it calls=20 qemu_get_current_aio_context()). > > For reference, the call chain is: > > chr_closed_bh() > qmp_set_link() > nc->info->link_status_changed() -> virtio_net_set_link_status() > virtio_net_set_status() > virtio_net_vhost_status() The code was added by Marc since: commit e7c83a885f865128ae3cf1946f8cb538b63cbfba Author: Marc-Andr=C3=A9 Lureau Date:=C2=A0=C2=A0 Mon Feb 27 14:49:56 2017 +0400 =C2=A0=C2=A0=C2=A0 vhost-user: delay vhost_user_stop Cc him for more thoughts. Thanks >> Thanks >> >> >>> Since the >>> vhost_net status hasn't yet changed, the second thread also enters >>> the n->vhost_started block, and also calls vhost_net_stop(). This >>> causes problems for the second thread when it tries to stop the netwo= rk >>> that's already been stopped. >>> >>> This adds a flag to the struct that's atomically set to prevent more = than >>> one thread from calling vhost_net_stop(). The atomic_fetch_inc() is = likely >>> overkill and probably could be done with a simple check-and-set, but >>> since it's a race condition there would still be a (very, very) small >>> window without using an atomic to set it. >>> >>> Signed-off-by: Dan Streetman >>> --- >>> hw/net/virtio-net.c | 3 ++- >>> include/hw/virtio/virtio-net.h | 1 + >>> 2 files changed, 3 insertions(+), 1 deletion(-) >>> >>> diff --git a/hw/net/virtio-net.c b/hw/net/virtio-net.c >>> index ffe0872fff..d36f50d5dd 100644 >>> --- a/hw/net/virtio-net.c >>> +++ b/hw/net/virtio-net.c >>> @@ -13,6 +13,7 @@ >>> >>> #include "qemu/osdep.h" >>> #include "qemu/iov.h" >>> +#include "qemu/atomic.h" >>> #include "hw/virtio/virtio.h" >>> #include "net/net.h" >>> #include "net/checksum.h" >>> @@ -240,7 +241,7 @@ static void virtio_net_vhost_status(VirtIONet *n,= uint8_t status) >>> "falling back on userspace virtio", -r); >>> n->vhost_started =3D 0; >>> } >>> - } else { >>> + } else if (atomic_fetch_inc(&n->vhost_stopped) =3D=3D 0) { >>> vhost_net_stop(vdev, n->nic->ncs, queues); >>> n->vhost_started =3D 0; >>> } >>> diff --git a/include/hw/virtio/virtio-net.h b/include/hw/virtio/virti= o-net.h >>> index b96f0c643f..d03fd933d0 100644 >>> --- a/include/hw/virtio/virtio-net.h >>> +++ b/include/hw/virtio/virtio-net.h >>> @@ -164,6 +164,7 @@ struct VirtIONet { >>> uint8_t nouni; >>> uint8_t nobcast; >>> uint8_t vhost_started; >>> + int vhost_stopped; >>> struct { >>> uint32_t in_use; >>> uint32_t first_multi;