[PATCH] D44675: [ARM] Codegen FP16 vmaxnm/vminnm scalar instructions

[Sjoerd Meijer via Phabricator via llvm-commits]

SjoerdMeijer updated this revision to Diff 142123.
SjoerdMeijer added a comment.

Hi Sam, thanks for checking this and pointing this out. These instructions should be selected when 'fast math' is enabled. Tests in vminmaxnm-safe.ll are checking that this won't happen when fast math is not enabled, which I forgot to add. So I've added new file fp16-vminmaxnm-safe.ll. I've also moved  the actual codegen tests to new file fp16-vminmaxnm.ll, because some cases were missing and there are a lot of them. 
Cheers.


https://reviews.llvm.org/D44675

Files:
  lib/Target/ARM/ARMISelLowering.cpp
  lib/Target/ARM/ARMInstrVFP.td
  test/CodeGen/ARM/fp16-instructions.ll
  test/CodeGen/ARM/fp16-vminmaxnm-safe.ll
  test/CodeGen/ARM/fp16-vminmaxnm.ll

-------------- next part --------------
A non-text attachment was scrubbed...
Name: D44675.142123.patch
Type: text/x-patch
Size: 21204 bytes
Desc: not available
URL: <http://lists.llvm.org/pipermail/llvm-commits/attachments/20180412/0b7da39c/attachment.bin>