Re: [Qemu-devel] [PATCH target-arm v3 04/15] arm: xlnx-zynqmp: Add GIC

[Peter Crosthwaite]

On Wed, Mar 18, 2015 at 10:39 PM, Peter Crosthwaite
<address@hidden> wrote:
> On Wed, Mar 18, 2015 at 7:06 PM, Paolo Bonzini <address@hidden> wrote:
>>
>>
>> On 18/03/2015 14:11, Peter Crosthwaite wrote:
>>> So I actually bisected this as a recent regression on:
>>>
>>> commit a464982499b2f637f6699e3d03e0a9d2e0b5288b (refs/bisect/bad)
>>> Author: Paolo Bonzini <address@hidden>
>>> Date:   Wed Feb 11 17:15:18 2015 +0100
>>>     rcu: run RCU callbacks under the BQL
>>>     This needs to go away sooner or later, but one complication is the
>>>     complex VFIO data structures that are modified in instance_finalize.
>>>     Take a shortcut for now.
>>>
>>>     Reviewed-by: Michael Roth <address@hidden>
>>>     Tested-by: Michael Roth <address@hidden>
>>>     Signed-off-by: Paolo Bonzini <address@hidden>
>>>
>>> Segfault backtrace:
>>
>> Please rebase to 2.3.0-rc0 to weed out the two problems related to early
>> qemu_mutex_lock_iothread() (this one, and the breakage of -daemonize).
>>
>
> So I am fully up to date with my baseline but the bug still replicates
> intermittently. I did however notice during the bisect that the
> failure probability did improve later in the history probably due to
> the patches you refer to.
>

I have debugged this further. New patch on list that builds on top of
of your fix to catch the case where the CPU is inited but not yet
realized.

Regards,
Peter

> I have a replicator handy if needed?
>
> Regards,
> Peter
>
>
>> Paolo
>>