[llvm-commits] patch: x86 vector udiv

[Peter Cooper]

Hi Duncan

Yeah, Eli suggested instcombine which having looked at it makes much more sense.  I'll reimplement it there and send round another patch.

Thanks,
Pete

On Nov 2, 2011, at 1:48 AM, Duncan Sands wrote:

> Hi Peter,
> 
>> Please review this patch for pr8429 to turn vector udiv into vector shrl instructions when dividing by powers of 2.  This was already done on scalars but not on vectors.
> 
> shouldn't the IR level optimizers be doing this?
> 
> Ciao, Duncan.
> 
>> 
>> I also found that vector shifts were always scalarised on x86 even with SSE on so i turned these into the equivalent intrinsics.
>> 
>> Thanks,
>> Pete
>> 
>> 
>> 
>> _______________________________________________
>> llvm-commits mailing list
>> llvm-commits at cs.uiuc.edu
>> http://lists.cs.uiuc.edu/mailman/listinfo/llvm-commits
> 
> _______________________________________________
> llvm-commits mailing list
> llvm-commits at cs.uiuc.edu
> http://lists.cs.uiuc.edu/mailman/listinfo/llvm-commits