Simon Pilgrim a3f50fb06d [X86] isVectorShiftByScalarCheap - vXi8 select(shift(x,splat0),shift(x,splat1)) is better than shift(x,select(splat0,splat1)) ...

Even though we don't have vXi8 vector shifts (apart from XOP), it is still better to prefer shift (or funnel-shift/rotate) by scalar where possible.

https://llvm.godbolt.org/z/6ss6ffTxv

Differential Revision: https://reviews.llvm.org/D116191

2021-12-23 14:30:02 +00:00

..

AArch64

[AArch64] Sink splat shuffles to lane index intrinsics

2021-11-22 08:11:35 +00:00

AMDGPU

…

ARM

[CodeGenPrepare] The instruction to be sunk should be inserted before its user in a block

2021-08-17 18:58:15 +08:00

Mips

…

NVPTX

…

PowerPC

…

RISCV

[CodeGenPrepare] Avoid a scalable-vector crash in ctlz/cttz

2021-10-20 16:45:55 +01:00

SPARC

…

X86

[X86] isVectorShiftByScalarCheap - vXi8 select(shift(x,splat0),shift(x,splat1)) is better than shift(x,select(splat0,splat1))

2021-12-23 14:30:02 +00:00

dom-tree.ll

Revert "[NFC] remove explicit default value for strboolattr attribute in tests"

2021-05-24 19:43:40 +02:00

sink-shift-and-trunc.ll

…

skip-merging-case-block.ll

…