Re: [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 deadloc

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

Re: [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 deadloc

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. ad

Re: [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 deadloc

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, s