[PATCH] D22064: [X86] Make some cast costs more precise
Simon Pilgrim via llvm-commits
llvm-commits at lists.llvm.org
Fri Jul 8 12:44:04 PDT 2016
RKSimon added a comment.
A couple of minors but otherwise this looks good to me. The AVX512 people should probably give the final OK though.
================
Comment at: lib/Target/X86/X86TargetTransformInfo.cpp:540
@@ -539,3 +539,3 @@
{ ISD::UINT_TO_FP, MVT::v8f32, MVT::v8i64, 1 },
{ ISD::UINT_TO_FP, MVT::v8f64, MVT::v8i64, 1 },
----------------
mkuper wrote:
> delena wrote:
> > mkuper wrote:
> > > RKSimon wrote:
> > > > Depending on how thorough we need to be shouldn't there be AVX512DQ+AVX512VL UINT_TO_FP cases for 128/256 bit vectors?
> > > Probably.
> > > I'd rather leave that to the Intel folks, they can probably get more precise numbers for SKX.
> > In this case, even if you have only DQ without VL, the conversion is in ZMM instead of YMM, but the cost is the same.
> We don't do this right now, see below.
OK - please add a TODO comment to the table for the AVX512DQ 128/256 entries.
================
Comment at: lib/Target/X86/X86TargetTransformInfo.cpp:717
@@ -708,1 +716,3 @@
+ { ISD::FP_EXTEND, MVT::v4f64, MVT::v4f32, 1 },
+ { ISD::FP_ROUND, MVT::v4f32, MVT::v4f64, 1 },
};
----------------
Test?
http://reviews.llvm.org/D22064
More information about the llvm-commits
mailing list