projects / platform / upstream / llvm.git / commit
? search:
summary | shortlog | log | commit | commitdiff | tree
(parent: 1f1b441) | patch
[AVX-512] Remove masking from 512-bit vpermil builtins. The backend now has versions...
authorCraig Topper <craig.topper@gmail.com>
Sun, 11 Dec 2016 01:26:52 +0000 (01:26 +0000)
committerCraig Topper <craig.topper@gmail.com>
Sun, 11 Dec 2016 01:26:52 +0000 (01:26 +0000)
commit678b07fe3ca59571f3559ab911305175d7c16b2d
tree933a88751eba5d5e5439585e9717774789db7709tree | snapshot
parent1f1b441267795db464d2200ecfc6706696c2b77fcommit | diff
[AVX-512] Remove masking from 512-bit vpermil builtins. The backend now has versions without masking so wrap it with select.

This will allow the backend to constant fold these to generic shuffle vectors like 128-bit and 256-bit without having to working about handling masking.

llvm-svn: 289351
clang/include/clang/Basic/BuiltinsX86.def diff | blob | history
clang/lib/Headers/avx512fintrin.h diff | blob | history
clang/test/CodeGen/avx512f-builtins.c diff | blob | history
Domain: System / Toolchain;