[all-commits] [llvm/llvm-project] 352a83: [InstCombine] Canonicalize non-i8 gep of mul to i8...

[David Green via All-commits]

  Branch: refs/heads/main
  Home:   https://github.com/llvm/llvm-project
  Commit: 352a836176b25abfdc26ddc5ddbd18288715a794
      https://github.com/llvm/llvm-project/commit/352a836176b25abfdc26ddc5ddbd18288715a794
  Author: David Green <david.green at arm.com>
  Date:   2024-06-26 (Wed, 26 Jun 2024)

  Changed paths:
    M llvm/lib/Transforms/InstCombine/InstructionCombining.cpp
    A llvm/test/Transforms/InstCombine/canonicalize-gep-mul.ll
    M llvm/test/Transforms/InstCombine/getelementptr.ll
    M llvm/test/Transforms/LoopVectorize/AArch64/sve-gather-scatter.ll
    M llvm/test/Transforms/LoopVectorize/AArch64/sve-interleaved-accesses.ll
    M llvm/test/Transforms/LoopVectorize/AArch64/sve-widen-phi.ll
    M llvm/test/Transforms/LoopVectorize/AArch64/uniform-args-call-variants.ll
    M llvm/test/Transforms/LoopVectorize/SystemZ/addressing.ll
    M llvm/test/Transforms/LoopVectorize/X86/interleaving.ll
    M llvm/test/Transforms/LoopVectorize/induction.ll
    M llvm/test/Transforms/LoopVectorize/interleaved-accesses.ll
    M llvm/test/Transforms/LoopVectorize/scalable-inductions.ll
    M llvm/test/Transforms/LoopVectorize/scalable-loop-unpredicated-body-scalar-tail.ll
    M llvm/test/Transforms/LowerMatrixIntrinsics/multiply-fused-loops.ll

  Log Message:
  -----------
  [InstCombine] Canonicalize non-i8 gep of mul to i8 (#96606)

This is a small canonicalization for `gep i32, p, (mul x, C)` -> `gep
i8, p, (mul x, C*4)`, so that the mul can combine both of the constant
multiplications, and we take a small step towards canonicalizing more
geps to i8.

It currently doesn't attempt to check for multiple uses on the mul, but
that should be possible if it sounds better. Let me know what you think
of the idea in general.



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