[libvirt] [PATCH] fix bug libvirt daemon deadlock when another force console break down an existed console client when this deadlock hanppened, libvirtd backtrace as follow, a typical ABBA deadlock circumstance:
(gdb) thread 23 [Switching to thread 23 (Thread 0x7fbb54810700 (LWP 296966))] (gdb) bt at rpc/virnetserverclient.c:1503 client=client@entry=0x5641d20e20a0, msg=0x7fbb08003ab0, rerr=rerr@entry=0x7fbb5480f7b0, procedure=201, type=type@entry=3, serial=9) at rpc/virnetserverprogram.c:173 msg=<optimized out>, rerr=rerr@entry=0x7fbb5480f7b0, procedure=<optimized out>, serial=<optimized out>) at rpc/virnetserverprogram.c:222 opaque=opaque@entry=0x5641d20e20a0) at stream.c:246 force=force@entry=true) at conf/virchrdev.c:386 at qemu/qemu_driver.c:16188 st=st@entry=0x7fbb08008fc0, flags=1) at libvirt-domain.c:9363 rerr=0x7fbb5480fc10, msg=<optimized out>, client=0x5641d20e4e80) at remote_dispatch.h:8540 rerr=0x7fbb5480fc10, args=0x7fbb08005eb0, ret=<optimized out>) at remote_dispatch.h:8505 server=0x5641d20be420, prog=0x5641d20da580) at rpc/virnetserverprogram.c:437 msg=0x5641d20e9a20) at rpc/virnetserverprogram.c:307 srv=0x5641d20be420) at rpc/virnetserver.c:148 ---Type <return> to continue, or q <return> to quit--- (gdb) thread 26 [Switching to thread 26 (Thread 0x7fbb87a378c0 (LWP 295636))] (gdb) bt at stream.c:286 at rpc/virnetserverclient.c:1247 at rpc/virnetserverclient.c:1457 at util/vireventpoll.c:508 Signed-off-by: LanceLiu <liu.lance.89@gmail.com> --- src/remote/remote_daemon_stream.c | 4 +++- 1 file changed, 3 insertions(+), 1 deletion(-) diff --git a/src/remote/remote_daemon_stream.c b/src/remote/remote_daemon_stream.c index 1f6e783..b27348f 100644 --- a/src/remote/remote_daemon_stream.c +++ b/src/remote/remote_daemon_stream.c @@ -284,14 +284,16 @@ daemonStreamEvent(virStreamPtr st, int events, void *opaque) * -1 on fatal client error */ static int -daemonStreamFilter(virNetServerClientPtr client ATTRIBUTE_UNUSED, +daemonStreamFilter(virNetServerClientPtr client, virNetMessagePtr msg, void *opaque) { daemonClientStream *stream = opaque; int ret = 0; + virObjectUnlock(client); virMutexLock(&stream->priv->lock); + virObjectLock(client); if (msg->header.type != VIR_NET_STREAM && msg->header.type != VIR_NET_STREAM_HOLE) -- 1.8.3.1
On 11/19/19 12:39 PM, LanceLiu wrote:
(gdb) thread 23 [Switching to thread 23 (Thread 0x7fbb54810700 (LWP 296966))] (gdb) bt at rpc/virnetserverclient.c:1503 client=client@entry=0x5641d20e20a0, msg=0x7fbb08003ab0, rerr=rerr@entry=0x7fbb5480f7b0, procedure=201, type=type@entry=3, serial=9) at rpc/virnetserverprogram.c:173 msg=<optimized out>, rerr=rerr@entry=0x7fbb5480f7b0, procedure=<optimized out>, serial=<optimized out>) at rpc/virnetserverprogram.c:222 opaque=opaque@entry=0x5641d20e20a0) at stream.c:246 force=force@entry=true) at conf/virchrdev.c:386 at qemu/qemu_driver.c:16188 st=st@entry=0x7fbb08008fc0, flags=1) at libvirt-domain.c:9363 rerr=0x7fbb5480fc10, msg=<optimized out>, client=0x5641d20e4e80) at remote_dispatch.h:8540 rerr=0x7fbb5480fc10, args=0x7fbb08005eb0, ret=<optimized out>) at remote_dispatch.h:8505 server=0x5641d20be420, prog=0x5641d20da580) at rpc/virnetserverprogram.c:437 msg=0x5641d20e9a20) at rpc/virnetserverprogram.c:307 srv=0x5641d20be420) at rpc/virnetserver.c:148 ---Type <return> to continue, or q <return> to quit---
(gdb) thread 26 [Switching to thread 26 (Thread 0x7fbb87a378c0 (LWP 295636))] (gdb) bt at stream.c:286 at rpc/virnetserverclient.c:1247 at rpc/virnetserverclient.c:1457 at util/vireventpoll.c:508
Signed-off-by: LanceLiu <liu.lance.89@gmail.com> --- src/remote/remote_daemon_stream.c | 4 +++- 1 file changed, 3 insertions(+), 1 deletion(-)
diff --git a/src/remote/remote_daemon_stream.c b/src/remote/remote_daemon_stream.c index 1f6e783..b27348f 100644 --- a/src/remote/remote_daemon_stream.c +++ b/src/remote/remote_daemon_stream.c @@ -284,14 +284,16 @@ daemonStreamEvent(virStreamPtr st, int events, void *opaque) * -1 on fatal client error */ static int -daemonStreamFilter(virNetServerClientPtr client ATTRIBUTE_UNUSED, +daemonStreamFilter(virNetServerClientPtr client,
This does not look rebased onto current git HEAD ;-) "Luckily", the problem you're fixing is still there.
virNetMessagePtr msg, void *opaque) { daemonClientStream *stream = opaque; int ret = 0;
+ virObjectUnlock(client); virMutexLock(&stream->priv->lock); + virObjectLock(client);
if (msg->header.type != VIR_NET_STREAM && msg->header.type != VIR_NET_STREAM_HOLE)
I've reworded the commit message, put a comment to virNetServerClientFilterFunc() typedef, ACKed and pushed. Congratulations on your first libvirt contribution! Michal
Thank you! But it looks like libvirtd daemon stream module still has bug lead to process segfault if applied this patch(if not, it will lead libvirtd deadlock).I have got the cause, but it seems difficult to fix it with present codes。 Bug reproduce: 1. add sleep(3) in daemonStreamFilter() so it is easily to get the point, as follows static int daemonStreamFilter(virNetServerClientPtr client, virNetMessagePtr msg, void *opaque) { daemonClientStream *stream = opaque; int ret = 0; static int first = 1; // if (first) { // notify_force_console(); // VIR_ERROR("notify force console to break this client down!client: %p", client); // sleep(30); // VIR_ERROR("sleep 10 finished"); // } VIR_ERROR("stream: %p, filter_id: %d", stream, stream->filterID); virObjectUnlock(client); sleep(3); virMutexLock(&stream->priv->lock); virObjectLock(client); 2. build the libvirtd with new code, then replace the libvirtd 3. use virsh console to open a vm's console tty, then just input many chars in the console 4. open new terminal, then use virsh console --force to break the previous client, then you'll got libvirtd segfault it looks because when daemonStreamEvent() with para events as VIR_STREAM_EVENT_ERROR is called, it will remove stream and free stream, but daemonStreamFilter still waiting stream->priv->lock mutex, thus cause invalid memory access To fix this, I think the most effective and clean way is to drop client->priv->lock elements, just use client object lock。But for current code, it's not easy to do that。 also， a workaroud scheme is to add a big lock, but is ugly and low efficient。 Any good ideas? Michal Privoznik <mprivozn@redhat.com> 于2019年11月19日周二 下午11:41写道：
(gdb) thread 23 [Switching to thread 23 (Thread 0x7fbb54810700 (LWP 296966))] (gdb) bt at rpc/virnetserverclient.c:1503 client=client@entry=0x5641d20e20a0, msg=0x7fbb08003ab0, rerr=rerr@entry=0x7fbb5480f7b0, procedure=201, type=type@entry=3, serial=9) at rpc/virnetserverprogram.c:173 msg=<optimized out>, rerr=rerr@entry=0x7fbb5480f7b0,
On 11/19/19 12:39 PM, LanceLiu wrote: procedure=<optimized out>, serial=<optimized out>)
at rpc/virnetserverprogram.c:222 opaque=opaque@entry=0x5641d20e20a0) at stream.c:246 force=force@entry=true) at conf/virchrdev.c:386 at qemu/qemu_driver.c:16188 st=st@entry=0x7fbb08008fc0, flags=1) at libvirt-domain.c:9363 rerr=0x7fbb5480fc10, msg=<optimized out>, client=0x5641d20e4e80) at
remote_dispatch.h:8540
rerr=0x7fbb5480fc10, args=0x7fbb08005eb0, ret=<optimized out>) at
remote_dispatch.h:8505
server=0x5641d20be420, prog=0x5641d20da580) at
rpc/virnetserverprogram.c:437
msg=0x5641d20e9a20) at rpc/virnetserverprogram.c:307 srv=0x5641d20be420) at rpc/virnetserver.c:148 ---Type <return> to continue, or q <return> to quit---
(gdb) thread 26 [Switching to thread 26 (Thread 0x7fbb87a378c0 (LWP 295636))] (gdb) bt at stream.c:286 at rpc/virnetserverclient.c:1247 at rpc/virnetserverclient.c:1457 at util/vireventpoll.c:508
Signed-off-by: LanceLiu <liu.lance.89@gmail.com> --- src/remote/remote_daemon_stream.c | 4 +++- 1 file changed, 3 insertions(+), 1 deletion(-)
diff --git a/src/remote/remote_daemon_stream.c
b/src/remote/remote_daemon_stream.c
index 1f6e783..b27348f 100644 --- a/src/remote/remote_daemon_stream.c +++ b/src/remote/remote_daemon_stream.c @@ -284,14 +284,16 @@ daemonStreamEvent(virStreamPtr st, int events, void *opaque) * -1 on fatal client error */ static int -daemonStreamFilter(virNetServerClientPtr client ATTRIBUTE_UNUSED, +daemonStreamFilter(virNetServerClientPtr client,
This does not look rebased onto current git HEAD ;-) "Luckily", the problem you're fixing is still there.
virNetMessagePtr msg, void *opaque) { daemonClientStream *stream = opaque; int ret = 0;
+ virObjectUnlock(client); virMutexLock(&stream->priv->lock); + virObjectLock(client);
if (msg->header.type != VIR_NET_STREAM && msg->header.type != VIR_NET_STREAM_HOLE)
I've reworded the commit message, put a comment to virNetServerClientFilterFunc() typedef, ACKed and pushed.
Congratulations on your first libvirt contribution!
Michal
On 11/22/19 8:00 AM, Lance Liu wrote:
Thank you! But it looks like libvirtd daemon stream module still has bug lead to process segfault if applied this patch(if not, it will lead libvirtd deadlock).I have got the cause, but it seems difficult to fix it with present codes。 Bug reproduce: 1. add sleep(3) in daemonStreamFilter() so it is easily to get the point, as follows static int daemonStreamFilter(virNetServerClientPtr client, virNetMessagePtr msg, void *opaque) { daemonClientStream *stream = opaque; int ret = 0; static int first = 1;
// if (first) { // notify_force_console(); // VIR_ERROR("notify force console to break this client down!client: %p", client); // sleep(30); // VIR_ERROR("sleep 10 finished"); // } VIR_ERROR("stream: %p, filter_id: %d", stream, stream->filterID); virObjectUnlock(client); sleep(3); virMutexLock(&stream->priv->lock); virObjectLock(client); 2. build the libvirtd with new code, then replace the libvirtd 3. use virsh console to open a vm's console tty, then just input many chars in the console 4. open new terminal, then use virsh console --force to break the previous client, then you'll got libvirtd segfault
it looks because when daemonStreamEvent() with para events as VIR_STREAM_EVENT_ERROR is called, it will remove stream and free stream, but daemonStreamFilter still waiting stream->priv->lock mutex, thus cause invalid memory access
Yes, this is exactly what is happening.
To fix this, I think the most effective and clean way is to drop client->priv->lock elements, just use client object lock。But for current code, it's not easy to do that。 also， a workaroud scheme is to add a big lock, but is ugly and low efficient。
Any good ideas?
I think we need to get back to drawing board. In the original problem you observed ABBA, but also was patching older libvirt. I convinced myself that the problem still occurs even with (at that point) current master: 1) virNetServerClientDispatchEvent() locks @client, virNetServerClientDispatchRead() daemonStreamFilter() which locks stream->priv->lock; on the other hand: 2) daemonStreamEvent() locks stream->priv->lock, virNetServerClientClose() which locks @client; /* there are other functions called over @client which also lock @client, I've just picked one */ So there is ABBA pattern. However, there might be another lock that is breaking the pattern (although I don't see it now): CAB and another thread then does CBA. Do you have a reproducer for your original problem? Does it reproduce on the current master (after you revert the patch I've merged of course). Michal
participants (3)
-
Lance Liu -
LanceLiu
-
Michal Privoznik