Re: [PATCH 1/1] s390x/cpumodel: add 3931 and 3932

[Thomas Huth <thuth@redhat.com>]

On 01/07/2021 09.45, David Hildenbrand wrote:
> On 30.06.21 17:56, Christian Borntraeger wrote:
>>
>>
>> On 30.06.21 17:32, Cornelia Huck wrote:
>>> On Wed, Jun 30 2021, Christian Borntraeger <borntraeger@de.ibm.com> wrote:
>>>
>>>> On 30.06.21 15:32, David Hildenbrand wrote:
>>>>> On 22.06.21 22:19, Christian Borntraeger wrote:
>>>>>> This defines 5 new facilities and the new 3931 and 3932 machines.
>>>>>> As before the name is not yet known and we do use gen16a and gen16b.
>>>>>> The new features are part of the full model.
>>>>>>
>>>>>> The default model is still empty (same as z15) and will be added
>>>>>> in a separate patch at a later point in time.
>>>>>>
>>>>>> Also add the dependencies of new facilities and as a fix for z15 add
>>>>>> a dependency from S390_FEAT_VECTOR_PACKED_DECIMAL_ENH to
>>>>>> S390_VECTOR_PACKED_DECIMAL.
>>>>>>
>>>>>> Signed-off-by: Christian Borntraeger <borntraeger@de.ibm.com>
>>>>>> ---
>>>>>>     target/s390x/cpu_features_def.h.inc |  5 +++++
>>>>>>     target/s390x/cpu_models.c           |  6 ++++++
>>>>>>     target/s390x/gen-features.c         | 14 ++++++++++++++
>>>>>>     3 files changed, 25 insertions(+)
>>>>>>
>>>>>> diff --git a/target/s390x/cpu_features_def.h.inc 
>>>>>> b/target/s390x/cpu_features_def.h.inc
>>>>>> index 7db3449e0434..c71caee74411 100644
>>>>>> --- a/target/s390x/cpu_features_def.h.inc
>>>>>> +++ b/target/s390x/cpu_features_def.h.inc
>>>>>> @@ -109,6 +109,11 @@ DEF_FEAT(VECTOR_PACKED_DECIMAL_ENH, "vxpdeh", 
>>>>>> STFL, 152, "Vector-Packed-Decimal-
>>>>>>     DEF_FEAT(MSA_EXT_9, "msa9-base", STFL, 155, 
>>>>>> "Message-security-assist-extension-9 facility (excluding subfunctions)")
>>>>>>     DEF_FEAT(ETOKEN, "etoken", STFL, 156, "Etoken facility")
>>>>>>     DEF_FEAT(UNPACK, "unpack", STFL, 161, "Unpack facility")
>>>>>> +DEF_FEAT(NNPA, "nnpa", STFL, 165, "NNPA facility")
>>>>>> +DEF_FEAT(VECTOR_PACKED_DECIMAL_ENH2, "vxpdeh2", STFL, 192, 
>>>>>> "Vector-Packed-Decimal-Enhancement facility 2")
>>>>>> +DEF_FEAT(BEAR, "bear", STFL, 193, "BEAR-enhancement facility")
>>>>>
>>>>> Usually we use "eh" for enhancement. Which would result in "beareh" or 
>>>>> alternatively "beh". But maybe the "enhancement" part is not actually 
>>>>> an enhancement, but instead this facility is more like the etoken or 
>>>>> unpack facility ...
>>>>
>>>> There was no bear facility (I think it was part of PER3).
>>>> beareh or beh would be fine with me.
>>>>
>>>>>
>>>>>> +DEF_FEAT(RDP, "rdp", STFL, 194, "Reset-DAT-protection facility")
>>>>>> +DEF_FEAT(ACTIVITY, "activity", STFL, 196, 
>>>>>> "Processor-Activity-Instrumentation facility")
>>>>>
>>>>> Would "pai" be a more appropriate feature name?
>>>>
>>>> pai would be ok for me as well.
>>>>
>>>> Conny, do you want to replace "activity" with "pai" and "bear" with 
>>>> "beareh" in your tree?
>>>
>>> I can certainly edit this to a naming everyone agrees with (no strong
>>> opinions from my side).
>>
>> Lets pick "pai" and the choose among "beareh" and "beh"
>>
> 
> I'd just go for "beareh" in case we ever have another b...enhancement 
> feature. But no strong opinion.

+1 for beareh
... the chance for confusion with other TLAs is to big otherwise.

  Thomas