[all-commits] [llvm/llvm-project] 8961f3: [X86] shouldReduceLoadWidth - don't split loads if...

[Simon Pilgrim via All-commits]

  Branch: refs/heads/main
  Home:   https://github.com/llvm/llvm-project
  Commit: 8961f3ee758f18f3c5c707b76406afbe6238ef7e
      https://github.com/llvm/llvm-project/commit/8961f3ee758f18f3c5c707b76406afbe6238ef7e
  Author: Simon Pilgrim <llvm-dev at redking.me.uk>
  Date:   2025-04-29 (Tue, 29 Apr 2025)

  Changed paths:
    M llvm/lib/Target/X86/X86ISelLowering.cpp
    M llvm/lib/Target/X86/X86ISelLowering.h
    M llvm/test/CodeGen/X86/oddsubvector.ll
    M llvm/test/CodeGen/X86/setcc-lowering.ll
    M llvm/test/CodeGen/X86/vector-interleaved-store-i16-stride-3.ll
    M llvm/test/CodeGen/X86/vector-interleaved-store-i16-stride-5.ll
    M llvm/test/CodeGen/X86/vector-interleaved-store-i16-stride-6.ll
    M llvm/test/CodeGen/X86/vector-interleaved-store-i16-stride-7.ll
    M llvm/test/CodeGen/X86/vector-interleaved-store-i8-stride-5.ll
    M llvm/test/CodeGen/X86/vector-interleaved-store-i8-stride-6.ll
    M llvm/test/CodeGen/X86/vector-interleaved-store-i8-stride-7.ll
    M llvm/test/CodeGen/X86/vselect-pcmp.ll

  Log Message:
  -----------
  [X86] shouldReduceLoadWidth - don't split loads if we can freely reuse full width legal binop (#129695)

Currently shouldReduceLoadWidth is very relaxed about when loads can be
split to avoid extractions from the original full width load - resulting
in many cases where the number of memory operations notably increases,
replacing the cost of a extract_subvector for additional loads.

This patch adjusts the 256/512-bit vector load splitting metric to
detect cases where ANY use of the full width load is be used directly -
in which case we will now reuse that load for smaller types, unless we'd
need to extract an upper subvector / integer element - i.e. we now
correctly treat (extract_subvector cst, 0) as free.

We retain the existing logic of never splitting loads if all uses are
extract+stores but we improve this by peeking through bitcasts while
looking for extract_subvector/store chains.

This required a number of fixes - shouldReduceLoadWidth now needs to
peek through bitcasts UP the use-chain to find final users (limited to
hasOneUse cases to reduce complexity). It also exposed an issue in
isTargetCanonicalConstantNode which assumed that a load of vector
constant data would always extract, which is no longer the case.



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