projects / platform / upstream / llvm.git / commit
? search:
summary | shortlog | log | commit | commitdiff | tree
(parent: 6a7cd4a) | patch
[X86][AMX] enable amx cast intrinsics in FE.
authorLuo, Yuanke <yuanke.luo@intel.com>
Thu, 17 Mar 2022 14:48:33 +0000 (22:48 +0800)
committerLuo, Yuanke <yuanke.luo@intel.com>
Sat, 2 Apr 2022 06:02:35 +0000 (14:02 +0800)
commit979d876bb4e93db1be50360a4941b83319d4e114
tree59b94498f51021ed21e385d2e3bea6808a2e7826tree | snapshot
parent6a7cd4a1df6c394f5d0dae93c9be26f8f777cd7fcommit | diff
[X86][AMX] enable amx cast intrinsics in FE.

We have some discission in D99152 and llvm-dev and finially come up with
a solution to add amx specific cast intrinsics. We've support the
intrinsics in llvm IR. This patch is to replace bitcast with amx cast
intrinsics in code emitting in FE.

Differential Revision: https://reviews.llvm.org/D122567
clang/lib/CodeGen/CGBuiltin.cpp diff | blob | history
clang/test/CodeGen/X86/amx_api.c diff | blob | history
Domain: System / Toolchain;