On 12/09/16 10:39, Vladimir Murzin wrote:
> On 09/09/16 17:59, Marc Zyngier wrote:
>> On 08/09/16 17:06, Vladimir Murzin wrote:
>>> vgic-v3 driver queries CPU affinity level up to Aff3, which is valid
>>> for arm64. However, for arm up to Aff2 levels are supported, so
>>> querying for third level ends with upper bits of MPIDR are treated as
>>> valid affinity level which is not true. Make sure we report zero for
>>> any affinity level above two.
>>>
>>> Signed-off-by: Vladimir Murzin <vladimir.murzin@xxxxxxx>
>>> ---
>>> arch/arm/include/asm/cputype.h | 3 ++-
>>> 1 file changed, 2 insertions(+), 1 deletion(-)
>>>
>>> diff --git a/arch/arm/include/asm/cputype.h b/arch/arm/include/asm/cputype.h
>>> index 1ee94c7..f08fac4 100644
>>> --- a/arch/arm/include/asm/cputype.h
>>> +++ b/arch/arm/include/asm/cputype.h
>>> @@ -55,9 +55,10 @@
>>>
>>> #define MPIDR_LEVEL_BITS 8
>>> #define MPIDR_LEVEL_MASK ((1 << MPIDR_LEVEL_BITS) - 1)
>>> +#define MPIDR_LEVEL_SHIFT(level) (MPIDR_LEVEL_BITS * level)
>>>
>>> #define MPIDR_AFFINITY_LEVEL(mpidr, level) \
>>> - ((mpidr >> (MPIDR_LEVEL_BITS * level)) & MPIDR_LEVEL_MASK)
>>> + (((mpidr & MPIDR_HWID_BITMASK) >> MPIDR_LEVEL_SHIFT(level)) & MPIDR_LEVEL_MASK)
>>>
>>> #define ARM_CPU_IMP_ARM 0x41
>>> #define ARM_CPU_IMP_INTEL 0x69
>>>
>>
>> There is something I don't quite get. Is this patch really necessary?
>> Are there cases where we construct a MPIDR that can have Aff3 set on 32bit?
>
> I've just checked and it seems that I've been keeping the change for
> MPIDR_AFFINITY_LEVEL() since vgic-old era where it was used with
>
> static u32 compress_mpidr(unsigned long mpidr)
> {
> u32 ret;
>
> ret = MPIDR_AFFINITY_LEVEL(mpidr, 0);
> ret |= MPIDR_AFFINITY_LEVEL(mpidr, 1) << 8;
> ret |= MPIDR_AFFINITY_LEVEL(mpidr, 2) << 16;
> ret |= MPIDR_AFFINITY_LEVEL(mpidr, 3) << 24;
>
> return ret;
> }
>
> and now that code gone and nobody passes level 3 into this macro, so I
> can drop this change.
>
> However, I do need MPIDR_LEVEL_SHIFT() macro since it is used in vgic-v3.
>
> Should I leave patch as is or you have something in mind?
I think that for the sake of keeping the change minimal, it'd be better
to just add the MPIDR_LEVEL_SHIFT macro, and drop the other changes.
Thanks,
M.
--
Jazz is not dead. It just smells funny...
_______________________________________________
kvmarm mailing list
kvmarm@xxxxxxxxxxxxxxxxxxxxx
https://lists.cs.columbia.edu/mailman/listinfo/kvmarm
