Am 26.06.2012 um 17:39 schrieb Cornelia Huck <cornelia.h...@de.ibm.com>:
> On Tue, 26 Jun 2012 16:56:08 +0200
> Alexander Graf <ag...@suse.de> wrote:
>
>>
>> On 26.06.2012, at 16:06, Cornelia Huck wrote:
>>
>>> From: Heiko Carstens <heiko.carst...@de.ibm.com>
>>>
>>> If an invalid parameter is passed or the addressed cpu is in an
>>> incorrect state sigp set prefix will store a status.
>>> This status must only have bits set as defined by the architecture.
>>> The current kvm implementation missed to clear bits and also did
>>> not set the intended status bit ("and" instead of "or" operation).
>>>
>>> Signed-off-by: Heiko Carstens <heiko.carst...@de.ibm.com>
>>> Signed-off-by: Cornelia Huck <cornelia.h...@de.ibm.com>
>>
>> What was the net effect of this for a guest? Any problems rising from it?
>
> The guest might see some unexpected status bits if it did call sigp set
> prefix in an incorrect way. I'm not aware that anybody has actually
> seen that.
Yeah, we only need the set prefix on vm init and here it should always succeed,
so this one is not all that urgent :).
Alex
>
> Cornelia
>
--
To unsubscribe from this list: send the line "unsubscribe kvm" in
the body of a message to majord...@vger.kernel.org
More majordomo info at http://vger.kernel.org/majordomo-info.html
