projects / platform / upstream / llvm.git / commit
? search:
summary | shortlog | log | commit | commitdiff | tree
(parent: 8fab9f8) | patch
[X86] Pass to transform tdpbf16ps intrinsics to scalar operation.
authorBing1 Yu <bing1.yu@intel.com>
Mon, 22 Mar 2021 01:48:59 +0000 (09:48 +0800)
committerBing1 Yu <bing1.yu@intel.com>
Mon, 22 Mar 2021 05:00:40 +0000 (13:00 +0800)
commit113f077f808feb00c109dc99dcedcf067c01bcef
tree5fda6a8cfa4d8a98b166bf8485239e9098ecb472tree | snapshot
parent8fab9f824fcaa38edebbd0f6b86f8612a4212cddcommit | diff
[X86] Pass to transform tdpbf16ps intrinsics to scalar operation.

In previous patch https://reviews.llvm.org/D93594, we only scalarize tilezero, tileload, tilestore and tiledpbssd. In this patch we scalarize tdpbf16ps intrinsic.

Reviewed By: pengfei

Differential Revision: https://reviews.llvm.org/D96110
llvm/lib/Target/X86/X86LowerAMXIntrinsics.cpp diff | blob | history
llvm/test/CodeGen/X86/AMX/amx-low-intrinsics.ll diff | blob | history
Domain: System / Toolchain;