[llvm-commits] [llvm] r53146 - in /llvm/trunk/lib/Target/Mips: Mips.td MipsAsmPrinter.cpp MipsCallingConv.td MipsISelDAGToDAG.cpp MipsISelLowering.cpp MipsISelLowering.h MipsInstrFPU.td MipsInstrFormats.td MipsInstrInfo.cpp MipsInstrInfo.h MipsIn

[Bruno Cardoso Lopes]

Thanks you both! Just commited a properly (i think) version. :)

On Tue, Jul 8, 2008 at 3:16 AM, Chris Lattner <clattner at apple.com> wrote:
>
> On Jul 7, 2008, at 8:10 PM, Eric Christopher wrote:
>
>> Couple of questions:
>>
>> +  if (!Subtarget->isAllegrex()) {
>> +    setOperationAction(ISD::SIGN_EXTEND_INREG, MVT::i8, Expand);
>> +    setOperationAction(ISD::SIGN_EXTEND_INREG, MVT::i16, Expand);
>> +  }
>>
>> +  bool hasAllegrexVFPU() const { return HasAllegrexVFPU; };
>> +  bool isAllegrex() const { return IsAllegrex; };
>>
>> For these it's probably best to return depending on CPU feature not
>> necessarily cpu type. There's a lot of commonality between mips chips.
>
> Great point, it would be best to define these as a subtarget feature
> and then have the cpu table indicate which ones support the feature
> directly.  For an example, please grep for 'FeatureFSqrt' in PPC.td
> for example.
>
> -Chris
> _______________________________________________
> llvm-commits mailing list
> llvm-commits at cs.uiuc.edu
> http://lists.cs.uiuc.edu/mailman/listinfo/llvm-commits
>



-- 
Bruno Cardoso Lopes
http://www.brunocardoso.cc
"When faced with untenable alternatives, you
 should consider your imperative."