projects / platform / upstream / llvm.git / commit
? search:
summary | shortlog | log | commit | commitdiff | tree
(parent: 96824ab) | patch
[AArch64][CodeGen] Restrict bfloat vector operations to what's actually supported
authorTies Stuij <ties.stuij@arm.com>
Fri, 28 Aug 2020 08:31:11 +0000 (09:31 +0100)
committerTies Stuij <ties.stuij@arm.com>
Fri, 28 Aug 2020 10:44:37 +0000 (11:44 +0100)
commitd678e14c55bea734524c2a294c93a7d9c0b8d99f
tree6df6c4e14e05ea280397c47caed1d25ca3c8b37etree | snapshot
parent96824abe7d80fc499032dab598968436132fcfb5commit | diff
[AArch64][CodeGen] Restrict bfloat vector operations to what's actually supported

Previously in addTypeForNeon, we would set the operations for bfloat vectors
like other generic types. But as bfloat is a storage-only type a number of
operations shouldn't be set. This patch fixes that.

Reviewed By: dmgreen

Differential Revision: https://reviews.llvm.org/D85101
llvm/lib/Target/AArch64/AArch64ISelLowering.cpp diff | blob | history
llvm/test/CodeGen/AArch64/bf16-vector-bitcast.ll diff | blob | history
llvm/test/CodeGen/AArch64/bf16-vector-shuffle.ll diff | blob | history
llvm/test/CodeGen/AArch64/bf16.ll diff | blob | history
Domain: System / Toolchain;