projects / platform / upstream / llvm.git / commit
? search:
summary | shortlog | log | commit | commitdiff | tree
(parent: db69ea4) | patch
[AArch64][SVE] Optimize bitcasts between unpacked half/i16 vectors.
authorSander de Smalen <sander.desmalen@arm.com>
Mon, 19 Jul 2021 06:13:14 +0000 (07:13 +0100)
committerSander de Smalen <sander.desmalen@arm.com>
Mon, 19 Jul 2021 07:29:28 +0000 (08:29 +0100)
commit0ed0573527404ee37d9ac700c9beab642e38657b
treed882ca72d7d142f6330fb5b66eafba8a8528956etree | snapshot
parentdb69ea40a91a683a4b1a2d581c5b5a2b46154677commit | diff
[AArch64][SVE] Optimize bitcasts between unpacked half/i16 vectors.

The case for nxv2f32/nxv2i32 was already covered by D104573.
This patch builds on top of that by making the mechanism work for
nxv2[b]f16/nxv2i16, nxv4[b]f16/nxv4i16 as well.

Reviewed By: efriedma

Differential Revision: https://reviews.llvm.org/D106138
llvm/lib/Target/AArch64/AArch64ISelLowering.cpp diff | blob | history
llvm/test/CodeGen/AArch64/sve-bitcast.ll diff | blob | history
Domain: System / Toolchain;