On 26/11/18 22:00, Eduardo Habkost wrote:
> On Mon, Nov 26, 2018 at 06:13:49PM +0100, Paolo Bonzini wrote:
>> On 26/11/18 16:28, Roman Kagan wrote:
>>> When started in compat configuration of SynIC, e.g.
>>>
>>> qemu-system-x86_64 -machine pc-i440fx-2.10,accel=kvm \
>>> -cpu host,-vmx,hv-relaxed,hv_spinlocks=0x1fff,hv-vpindex,hv-synic
>>>
>>> or explicitly
>>>
>>> qemu-system-x86_64 -enable-kvm -cpu host,hv-synic,x-hv-synic-kvm-only=on
>>>
>>> QEMU crashes in hyperv_synic_reset() trying to access the non-present
>>> qobject for SynIC.
>>>
>>> Add the missing check for NULL.
>>>
>>> Reported-by: Vitaly Kuznetsov <vkuzn...@redhat.com>
>>> Reported-by: Igor Mammedov <imamm...@redhat.com>
>>> Fixes: 9b4cf107b09d18ac30f46fd1c4de8585ccba030c
>>> Fixes: 4a93722f9c279184e95b1e1ad775c01deec05065
>>> Signed-off-by: Roman Kagan <rka...@virtuozzo.com>
>>> ---
>>> hw/hyperv/hyperv.c | 6 +++++-
>>> 1 file changed, 5 insertions(+), 1 deletion(-)
>>>
>>> diff --git a/hw/hyperv/hyperv.c b/hw/hyperv/hyperv.c
>>> index a28e7249d8..8758635227 100644
>>> --- a/hw/hyperv/hyperv.c
>>> +++ b/hw/hyperv/hyperv.c
>>> @@ -136,7 +136,11 @@ void hyperv_synic_add(CPUState *cs)
>>>
>>> void hyperv_synic_reset(CPUState *cs)
>>> {
>>> - device_reset(DEVICE(get_synic(cs)));
>>> + SynICState *synic = get_synic(cs);
>>> +
>>> + if (synic) {
>>> + device_reset(DEVICE(synic));
>>> + }
>>> }
>>>
>>> static const TypeInfo synic_type_info = {
>>>
>>
>> Queued, thanks.
>
> Oops, I had queued it earlier today and just submitted a pull
> request.
No big deal, it will be included twice. :)
Paolo
