[llvm-commits] [llvm] r108735 - in /llvm/trunk: include/llvm/Target/TargetLowering.h lib/CodeGen/SelectionDAG/TargetLowering.cpp
Eric Christopher
echristo at apple.com
Mon Jul 19 15:53:30 PDT 2010
On Jul 19, 2010, at 3:52 PM, Jakob Stoklund Olesen wrote:
>
> On Jul 19, 2010, at 3:41 PM, Eric Christopher wrote:
>
>>
>> On Jul 19, 2010, at 3:38 PM, Jakob Stoklund Olesen wrote:
>>
>>> QPR has 16 registers, DPR has 32.
>>>
>>> If each DPR counts as a QPR, the scheduler is going to behave as if there were only 16 DPR registers available. That is almost as bad as scheduling purely for register pressure.
>>>
>>
>> True, but you could maybe count each QPR as two dprs and if you run into an occasion where that matters it'd suck, but I bet it'd be infrequent.
>
> Right. That's what I meant by scaling. I think it is necessary.
>
Ah. Yes.
> Evan is probably right that QPRs and DPRs won't be live at the same time.
>
> But we should at least get the number of registers right.
>
> Overriding findRepresentativeClass for ARM does get the register counts right, though.
Right. Violent agreement. :)
-eric
More information about the llvm-commits
mailing list