Jay Foad 78281fd12c Revert "[AMDGPU] InstCombine llvm.amdgcn.ds.bpermute with uniform arguments (#129895)" ...

This reverts commit be5149a315.

It caused build failures in the openmp-offload-amdgpu-runtime buildbot
and others.

2025-03-06 15:05:19 +00:00

..

amdgcn-demanded-vector-elts-inseltpoison.ll

…

amdgcn-demanded-vector-elts.ll

…

amdgcn-intrinsics-gfx8.ll

…

amdgcn-intrinsics.ll

Revert "[AMDGPU] InstCombine llvm.amdgcn.ds.bpermute with uniform arguments (#129895)"

2025-03-06 15:05:19 +00:00

amdgcn-simplify-image-buffer-stores.ll

…

bitcast-fold-lane-ops.ll

AMDGPU: Fold bitcasts into readfirstlane, readlane, and permlane64 (#128494)

2025-02-27 20:59:11 +07:00

fma_legacy.ll

…

fmed3.ll

…

fmul_legacy.ll

…

image-d16.ll

…

issue68120.ll

…

lane-index-simplify-demanded-bits.ll

AMDGPU: Simplify demanded bits on readlane/writeline index arguments (#117963)

2024-12-06 10:31:14 -05:00

lit.local.cfg

…

llvm.amdgcn.wavefrontsize.ll

[IR] Convert from nocapture to captures(none) (#123181)

2025-01-29 16:56:47 +01:00

mbcnt.ll

…

memcpy-from-constant.ll

Revert "[InstCombine] Infer nuw for gep inbounds from base of object" (#120460)

2024-12-18 19:06:34 +01:00

mfma-scale.ll

AMDGPU: Shrink used number of registers for mfma scale based on format (#117047)

2024-11-21 09:08:05 -08:00

permlane64.ll

AMDGPU: Fold bitcasts into readfirstlane, readlane, and permlane64 (#128494)

2025-02-27 20:59:11 +07:00

rcp-contract-rsq.ll

Regenerate checks in test/Transforms/InstCombine/AMDGPU/. NFC.

2025-03-06 14:29:34 +00:00

select-from-load.ll

…

simplify-demanded-vector-elts-lane-intrinsics.ll

[IR] Return correct memory effects for convergencectrl (#129874)

2025-03-05 22:14:47 +08:00

tan.ll

…