[all-commits] [llvm/llvm-project] 1cae79: DAG: Handle load in SimplifyDemandedVectorElts

Matt Arsenault via All-commits all-commits at lists.llvm.org
Thu Feb 6 08:45:49 PST 2025 

  Branch: refs/heads/users/arsenm/dag/simplify-demanded-vector-elts-load
  Home:   https://github.com/llvm/llvm-project
  Commit: 1cae79c1c979fa9d99e595ac528f0378544eee76
      https://github.com/llvm/llvm-project/commit/1cae79c1c979fa9d99e595ac528f0378544eee76
  Author: Matt Arsenault <Matthew.Arsenault at amd.com>
  Date:   2025-02-06 (Thu, 06 Feb 2025)

  Changed paths:
    M llvm/lib/CodeGen/SelectionDAG/TargetLowering.cpp
    M llvm/test/CodeGen/AArch64/arm64-big-endian-bitconverts.ll
    M llvm/test/CodeGen/AArch64/dag-ReplaceAllUsesOfValuesWith.ll
    M llvm/test/CodeGen/AArch64/fcmp.ll
    M llvm/test/CodeGen/AArch64/fmlal-loreg.ll
    M llvm/test/CodeGen/AArch64/icmp.ll
    M llvm/test/CodeGen/AArch64/sve-fixed-length-extract-vector-elt.ll
    M llvm/test/CodeGen/AArch64/sve-fixed-length-masked-gather.ll
    M llvm/test/CodeGen/AArch64/sve-fixed-length-masked-scatter.ll
    M llvm/test/CodeGen/AArch64/sve-streaming-mode-fixed-length-extract-vector-elt.ll
    M llvm/test/CodeGen/AMDGPU/fcopysign.f32.ll
    M llvm/test/CodeGen/AMDGPU/fcopysign.f64.ll
    M llvm/test/CodeGen/AMDGPU/greedy-reverse-local-assignment.ll
    M llvm/test/CodeGen/AMDGPU/identical-subrange-spill-infloop.ll
    M llvm/test/CodeGen/AMDGPU/implicit-kernarg-backend-usage.ll
    M llvm/test/CodeGen/AMDGPU/shader-addr64-nonuniform.ll
    M llvm/test/CodeGen/AMDGPU/trunc.ll
    M llvm/test/CodeGen/AMDGPU/vector_rebroadcast.ll
    M llvm/test/CodeGen/AMDGPU/vector_shuffle.packed.ll
    M llvm/test/CodeGen/ARM/crash-on-pow2-shufflevector.ll
    M llvm/test/CodeGen/ARM/vector-promotion.ll
    M llvm/test/CodeGen/ARM/vext.ll
    M llvm/test/CodeGen/ARM/vuzp.ll
    M llvm/test/CodeGen/Mips/cconv/vector.ll
    M llvm/test/CodeGen/Mips/msa/basic_operations.ll
    M llvm/test/CodeGen/NVPTX/i128.ll
    M llvm/test/CodeGen/NVPTX/i8x4-instructions.ll
    M llvm/test/CodeGen/PowerPC/aix-vector-byval-callee.ll
    M llvm/test/CodeGen/PowerPC/canonical-merge-shuffles.ll
    M llvm/test/CodeGen/PowerPC/const-stov.ll
    M llvm/test/CodeGen/PowerPC/pr27078.ll
    M llvm/test/CodeGen/PowerPC/pre-inc-disable.ll
    M llvm/test/CodeGen/PowerPC/v16i8_scalar_to_vector_shuffle.ll
    M llvm/test/CodeGen/PowerPC/v2i64_scalar_to_vector_shuffle.ll
    M llvm/test/CodeGen/PowerPC/v8i16_scalar_to_vector_shuffle.ll
    M llvm/test/CodeGen/PowerPC/vsx_shuffle_le.ll
    M llvm/test/CodeGen/RISCV/rvv/fixed-vectors-fp-shuffles.ll
    M llvm/test/CodeGen/Thumb2/mve-extractstore.ll
    M llvm/test/CodeGen/Thumb2/mve-insertshuffleload.ll
    M llvm/test/CodeGen/X86/SwizzleShuff.ll
    M llvm/test/CodeGen/X86/avx-vbroadcast.ll
    M llvm/test/CodeGen/X86/avx.ll
    M llvm/test/CodeGen/X86/avx1-logical-load-folding.ll
    M llvm/test/CodeGen/X86/avx512-arith.ll
    M llvm/test/CodeGen/X86/avx512-intrinsics-fast-isel.ll
    M llvm/test/CodeGen/X86/avx512-shuffles/partial_permute.ll
    M llvm/test/CodeGen/X86/avx512-shuffles/shuffle-interleave.ll
    M llvm/test/CodeGen/X86/avx512-shuffles/unpack.ll
    M llvm/test/CodeGen/X86/avx512fp16-mov.ll
    M llvm/test/CodeGen/X86/bitreverse.ll
    M llvm/test/CodeGen/X86/buildvec-insertvec.ll
    M llvm/test/CodeGen/X86/combine-fabs.ll
    M llvm/test/CodeGen/X86/combine-sdiv.ll
    M llvm/test/CodeGen/X86/combine-udiv.ll
    M llvm/test/CodeGen/X86/commute-blend-avx2.ll
    M llvm/test/CodeGen/X86/commute-blend-sse41.ll
    M llvm/test/CodeGen/X86/copysign-constant-magnitude.ll
    M llvm/test/CodeGen/X86/extract-concat.ll
    M llvm/test/CodeGen/X86/extractelement-fp.ll
    M llvm/test/CodeGen/X86/extractelement-load.ll
    M llvm/test/CodeGen/X86/fabs.ll
    M llvm/test/CodeGen/X86/fast-isel-fneg.ll
    M llvm/test/CodeGen/X86/fma-signed-zero.ll
    M llvm/test/CodeGen/X86/fp-fold.ll
    M llvm/test/CodeGen/X86/fp-intrinsics-fma.ll
    M llvm/test/CodeGen/X86/fp-logic.ll
    M llvm/test/CodeGen/X86/fp-round.ll
    M llvm/test/CodeGen/X86/fp128-cast.ll
    M llvm/test/CodeGen/X86/fp16-libcalls.ll
    M llvm/test/CodeGen/X86/freeze-vector.ll
    M llvm/test/CodeGen/X86/gfni-funnel-shifts.ll
    M llvm/test/CodeGen/X86/half.ll
    M llvm/test/CodeGen/X86/insert-into-constant-vector.ll
    M llvm/test/CodeGen/X86/insertps-combine.ll
    M llvm/test/CodeGen/X86/insertps-from-constantpool.ll
    M llvm/test/CodeGen/X86/insertps-unfold-load-bug.ll
    M llvm/test/CodeGen/X86/is_fpclass.ll
    M llvm/test/CodeGen/X86/isel-blendi-gettargetconstant.ll
    M llvm/test/CodeGen/X86/load-partial.ll
    M llvm/test/CodeGen/X86/masked_load.ll
    M llvm/test/CodeGen/X86/masked_store.ll
    M llvm/test/CodeGen/X86/mmx-arith.ll
    M llvm/test/CodeGen/X86/neg_fp.ll
    M llvm/test/CodeGen/X86/negative-sin.ll
    M llvm/test/CodeGen/X86/packus.ll
    M llvm/test/CodeGen/X86/peephole-fold-movsd.ll
    M llvm/test/CodeGen/X86/pr14161.ll
    M llvm/test/CodeGen/X86/pr30511.ll
    M llvm/test/CodeGen/X86/pr31956.ll
    M llvm/test/CodeGen/X86/pr34592.ll
    M llvm/test/CodeGen/X86/pr36553.ll
    M llvm/test/CodeGen/X86/pr40811.ll
    M llvm/test/CodeGen/X86/pr63091.ll
    M llvm/test/CodeGen/X86/sar_fold64.ll
    M llvm/test/CodeGen/X86/setcc-combine.ll
    M llvm/test/CodeGen/X86/setcc-non-simple-type.ll
    M llvm/test/CodeGen/X86/shrink_vmul.ll
    M llvm/test/CodeGen/X86/shuffle-vs-trunc-512.ll
    M llvm/test/CodeGen/X86/splat-for-size.ll
    M llvm/test/CodeGen/X86/sqrt-fastmath-tune.ll
    M llvm/test/CodeGen/X86/sqrt-fastmath-tunecpu-attr.ll
    M llvm/test/CodeGen/X86/sqrt-fastmath.ll
    M llvm/test/CodeGen/X86/srem-seteq-vec-nonsplat.ll
    M llvm/test/CodeGen/X86/sse-align-12.ll
    M llvm/test/CodeGen/X86/sse2.ll
    M llvm/test/CodeGen/X86/sse3.ll
    M llvm/test/CodeGen/X86/sse41.ll
    M llvm/test/CodeGen/X86/strict-fsub-combines.ll
    M llvm/test/CodeGen/X86/subvector-broadcast.ll
    M llvm/test/CodeGen/X86/test-shrink-bug.ll
    M llvm/test/CodeGen/X86/tuning-shuffle-unpckpd-avx512.ll
    M llvm/test/CodeGen/X86/tuning-shuffle-unpckpd.ll
    M llvm/test/CodeGen/X86/urem-seteq-vec-tautological.ll
    M llvm/test/CodeGen/X86/vec_insert-5.ll
    M llvm/test/CodeGen/X86/vec_int_to_fp.ll
    M llvm/test/CodeGen/X86/vec_shift5.ll
    M llvm/test/CodeGen/X86/vector-bitreverse.ll
    M llvm/test/CodeGen/X86/vector-constrained-fp-intrinsics-flags.ll
    M llvm/test/CodeGen/X86/vector-constrained-fp-intrinsics.ll
    M llvm/test/CodeGen/X86/vector-fshl-256.ll
    M llvm/test/CodeGen/X86/vector-fshl-512.ll
    M llvm/test/CodeGen/X86/vector-fshr-256.ll
    M llvm/test/CodeGen/X86/vector-fshr-512.ll
    M llvm/test/CodeGen/X86/vector-reduce-fmax-nnan.ll
    M llvm/test/CodeGen/X86/vector-reduce-fmin.ll
    M llvm/test/CodeGen/X86/vector-rotate-128.ll
    M llvm/test/CodeGen/X86/vector-rotate-256.ll
    M llvm/test/CodeGen/X86/vector-shift-ashr-128.ll
    M llvm/test/CodeGen/X86/vector-shift-ashr-256.ll
    M llvm/test/CodeGen/X86/vector-shift-ashr-512.ll
    M llvm/test/CodeGen/X86/vector-shift-lshr-128.ll
    M llvm/test/CodeGen/X86/vector-shift-lshr-256.ll
    M llvm/test/CodeGen/X86/vector-shift-lshr-512.ll
    M llvm/test/CodeGen/X86/vector-shift-shl-128.ll
    M llvm/test/CodeGen/X86/vector-shift-shl-256.ll
    M llvm/test/CodeGen/X86/vector-shift-shl-512.ll
    M llvm/test/CodeGen/X86/vector-shuffle-128-v2.ll
    M llvm/test/CodeGen/X86/vector-shuffle-128-v4.ll
    M llvm/test/CodeGen/X86/vector-shuffle-combining-avx2.ll
    M llvm/test/CodeGen/X86/vector-shuffle-combining-ssse3.ll
    M llvm/test/CodeGen/X86/vector-shuffle-combining.ll
    M llvm/test/CodeGen/X86/vector-shuffle-v1.ll
    M llvm/test/CodeGen/X86/vector-shuffle-v192.ll
    M llvm/test/CodeGen/X86/vector-shuffle-v48.ll
    M llvm/test/CodeGen/X86/vselect.ll
    M llvm/test/CodeGen/X86/widened-broadcast.ll
    M llvm/test/CodeGen/X86/x86-interleaved-access.ll
    M llvm/test/CodeGen/X86/xop-shifts.ll
    M llvm/test/CodeGen/X86/xor.ll

  Log Message:
  -----------
  DAG: Handle load in SimplifyDemandedVectorElts

This improves some AMDGPU cases and avoids future regressions.
The combiner likes to form shuffles for cases where an extract_vector_elt
would do perfectly well, and this recovers some of the regressions from
losing load narrowing.

AMDGPU, Arch64 and RISCV test changes look broadly better. Other targets have
some improvements, but mostly regressions. In particular X86 looks much
worse. I'm guessing this is because it's shouldReduceLoadWidth is wrong.

I mostly just regenerated the checks. I assume some set of them should
switch to use volatile loads to defeat the optimization.



To unsubscribe from these emails, change your notification settings at https://github.com/llvm/llvm-project/settings/notifications

More information about the All-commits mailing list