[PATCH] D55163: AMDGPU: Add optimization patterns to combine fp32->fp16 conversions

[Rhys Perry via Phabricator via llvm-commits]

pendingchaos updated this revision to Diff 188516.
pendingchaos added a comment.
Herald added a subscriber: jdoerfert.

In this update:

- The 64-bit encodings are always selected and SelectCvtRtzF16F32*Mods() have been removed.
- SelectCvtRtzF16F32() is now implemented with SelectCvtRtzF16F32Lo().
- cvt_pkrtz(v, 0) and cvt_pkrtz(0, v) are handled in SelectCvtRtzF16F32LoHiImpl().
- The test explicitly tests the different encodings for the different sub targets.
- The getConstantValue() helper is used.


CHANGES SINCE LAST ACTION
  https://reviews.llvm.org/D55163/new/

https://reviews.llvm.org/D55163

Files:
  lib/Target/AMDGPU/AMDGPUISelDAGToDAG.cpp
  lib/Target/AMDGPU/SIInstrInfo.td
  lib/Target/AMDGPU/SIInstructions.td
  test/CodeGen/AMDGPU/cvt_pkrtz_f16_f32_combine.ll

-------------- next part --------------
A non-text attachment was scrubbed...
Name: D55163.188516.patch
Type: text/x-patch
Size: 12861 bytes
Desc: not available
URL: <http://lists.llvm.org/pipermail/llvm-commits/attachments/20190227/497a046a/attachment-0001.bin>