On May 18, 2007, at 11:59 AM, Dan Gohman wrote:
> Hello all,
>
> While doing some experiments with vector ValueTypes, I noticed that
> the PowerPC target uses MVT::LAST_VECTOR_VALUETYPE inconsistently with
> the way it is declared. I don't have the resources to test PowerPC
> changes, so I'm just sending this patch to the list in case anyone
> who does is interested in it.
The patch looks safe to me, but how is it semantically different?
-Chris
> Dan
>
> Index: lib/Target/PowerPC/PPCISelLowering.cpp
> ===================================================================
> RCS file: /var/cvs/llvm/llvm/lib/Target/PowerPC/PPCISelLowering.cpp,v
> retrieving revision 1.270
> diff -u -r1.270 PPCISelLowering.cpp
> --- lib/Target/PowerPC/PPCISelLowering.cpp
> +++ lib/Target/PowerPC/PPCISelLowering.cpp
> @@ -222,7 +222,7 @@
> // First set operation action for all vector types to expand.
> Then we
> // will selectively turn on ones that can be effectively
> codegen'd.
> for (unsigned VT = (unsigned)MVT::FIRST_VECTOR_VALUETYPE;
> - VT != (unsigned)MVT::LAST_VECTOR_VALUETYPE; ++VT) {
> + VT <= (unsigned)MVT::LAST_VECTOR_VALUETYPE; ++VT) {
> // add/sub are legal for all supported vector VT's.
> setOperationAction(ISD::ADD , (MVT::ValueType)VT, Legal);
> setOperationAction(ISD::SUB , (MVT::ValueType)VT, Legal);
> _______________________________________________
> llvm-commits mailing list
> llvm-commits@cs.uiuc.edu
> http://lists.cs.uiuc.edu/mailman/listinfo/llvm-commits
_______________________________________________
llvm-commits mailing list
llvm-commits@cs.uiuc.edu
http://lists.cs.uiuc.edu/mailman/listinfo/llvm-commits
