[llvm] r215588 - [FastISel] Let the target decide first if it wants to materialize a constant.

Eric Christopher echristo at gmail.com
Thu Aug 14 10:50:14 PDT 2014


Excellent. Let's do it.

On Thu, Aug 14, 2014 at 10:49 AM, Jim Grosbach <grosbach at apple.com> wrote:
> When there’s active and steady development in the area already is when it makes sense to go ahead and do style and conventions cleanups like this. Go for it.
>
> -Jim
>
>> On Aug 14, 2014, at 10:43 AM, Eric Christopher <echristo at gmail.com> wrote:
>>
>> Mmm.. that way may lie madness. I'm always a proponent of big changes
>> to make things consistent, but let's use a 3rd party to check.
>>
>> Jim? :)
>>
>> -eric
>>
>> On Thu, Aug 14, 2014 at 10:41 AM, Juergen Ributzka <juergen at apple.com> wrote:
>>> Should we also change tblgen to not emit capitalized function names anymore for FastEmit_*?
>>>
>>> -Juergen
>>>
>>> On Aug 14, 2014, at 10:38 AM, Eric Christopher <echristo at gmail.com> wrote:
>>>
>>>> That would be pretty nice. Generally we don't do the mass changes like
>>>> that, but I think the file is pretty awkward right now and so I'd
>>>> appreciate it for sure :)
>>>>
>>>> -eric
>>>>
>>>> On Thu, Aug 14, 2014 at 10:37 AM, Juergen Ributzka <juergen at apple.com> wrote:
>>>>> Copy paste from TargetMaterializeConstant ;) The file is a little bit of a
>>>>> mess when it comes to function names. Should I do a cleanup and update all
>>>>> the function names while I am at it?
>>>>>
>>>>> -Juergen
>>>>>
>>>>>
>>>>> On Aug 13, 2014, at 6:30 PM, Eric Christopher <echristo at gmail.com> wrote:
>>>>>
>>>>>>
>>>>>>
>>>>>> -/// materializeRegForValue - Helper for getRegForValue. This function is
>>>>>> -/// called when the value isn't already available in a register and must
>>>>>> -/// be materialized with new instructions.
>>>>>> -unsigned FastISel::materializeRegForValue(const Value *V, MVT VT) {
>>>>>> +unsigned FastISel::MaterializeConstant(const Value *V, MVT VT) {
>>>>>>  unsigned Reg = 0;
>>>>>> -
>>>>>
>>>>>
>>>>> Coding standard has functions with lower case names? :)
>>>>>
>>>>> -eric
>>>>>
>>>>>
>>>
>




More information about the llvm-commits mailing list