Re: [Qemu-devel] [PATCH v10 00/23] MTTCG Base enabling patches with ARM enablement

[Alex Bennée]

Pranith Kumar <address@hidden> writes:

> Hi Alex,
>
> On Mon, Feb 6, 2017 at 10:30 AM, Alex Bennée <address@hidden> wrote:
>> Hi Richard/Peter,
>>
>> This is a mostly ARM focused update to last weeks v9.
>>
>> It has necessitated one change to the cputlb API. It was pointed out
>> that translators often have to special case a bunch of things if you
>> longjmp() out of a helper. As a result the cputlb _synched() calls are
>> no longer QEMU_NORETURN but do document the importance of the guest to
>> exit the block as soon as synchronisation is required. In ARM's case
>> this is already done as the TLB flushes are treated as CP write
>> operation which ends the block by default. This means I was able to
>> drop the two patches that dealt with ARM_CP_PC, simplifying the code.
>>
>> The other moderate change was fixing up target-arm/powerctl to
>> properly model the ON_PENDING powerstate which is key to vCPUs
>> handling otherwise race-prone start-up sequences. The power off and
>> reset methods where also updated to update the CPUState structures in
>> the targets context.
>>
>> Otherwise there is the usual array of review tags and a few minor
>> fixes documented as normal bellow the --- line.
>>
>> A version of the tree can be found at:
>>
>>   https://github.com/stsquad/qemu/tree/mttcg/base-patches-v10
>>
>
> I've sent my Reviewed-by/tested-by for v9 but those seem to be lost.
> I've tested v10 too.
>
> So please add my tested and reviewed tags:
>
> Tested-and-Reviewed-by: Pranith Kumar <address@hidden>

Sorry I missed that as it was applied to the cover letter (everyone
keeps finding corner cases for my tooling ;-). I'll apply it to patch 23
where we default on.

>
> Thanks,


--
Alex Bennée