[PATCH] D42993: [AArch64] Fixes for ARMv8.2-A FP16 scalar intrinsic

Abderrazek Zaafrani via Phabricator via cfe-commits cfe-commits at lists.llvm.org
Thu Feb 8 15:40:33 PST 2018


az updated this revision to Diff 133505.
az added a comment.

> Question about the failures: I am now wondering if this means we were and still  are missing tests?

Given that this work is fixing https://reviews.llvm.org/D41792 which is mainly about adding frontend intrinsic support, then there is a test for each intrinsic and I am updating the tests for the intrinsics with code generation changes (they only test the frontend). As we previously discussed it, there is some work to be done in instruction selection in order to generate good IR for some intrinisics and the ones with problems need new tests. In this patch, I am fixing few simple ones that fall into that category (see AArch64InstrInfo.td) and I agree that I should have separated this patch into clang and llvm patches instead of just one. I am adding tests for those in this revision.


https://reviews.llvm.org/D42993

Files:
  clang/lib/CodeGen/CGBuiltin.cpp
  clang/test/CodeGen/aarch64-v8.2a-fp16-intrinsics.c
  llvm/lib/Target/AArch64/AArch64InstrInfo.td
  llvm/test/CodeGen/AArch64/f16-instructions.ll

-------------- next part --------------
A non-text attachment was scrubbed...
Name: D42993.133505.patch
Type: text/x-patch
Size: 18061 bytes
Desc: not available
URL: <http://lists.llvm.org/pipermail/cfe-commits/attachments/20180208/133c5baa/attachment-0001.bin>


More information about the cfe-commits mailing list