[all-commits] [llvm/llvm-project] 08b20d: [RISCV] Generaize reduction tree matching to fp su...

[Philip Reames via All-commits]

  Branch: refs/heads/main
  Home:   https://github.com/llvm/llvm-project
  Commit: 08b20d838549e8ff3939c6793b5067b0d9b36337
      https://github.com/llvm/llvm-project/commit/08b20d838549e8ff3939c6793b5067b0d9b36337
  Author: Philip Reames <preames at rivosinc.com>
  Date:   2023-10-09 (Mon, 09 Oct 2023)

  Changed paths:
    M llvm/lib/Target/RISCV/RISCVISelLowering.cpp
    M llvm/test/CodeGen/RISCV/rvv/fixed-vectors-reduction-formation.ll

  Log Message:
  -----------
  [RISCV] Generaize reduction tree matching to fp sum reductions (#68599)

This builds on the transform introduced in f0505c3, and generalizes to
all integer operations in 45a334d. This change adds support for floating
point sumation.

A couple of notes:
* I chose to leave fmaxnum and fminnum unhandled for the moment. They
have a slightly different set of legality rules.
* We could form strictly sequenced FADD reductions for FADDs without
fast math flags. As the ordered reductions are more expensive, I left
thinking about this as a future exercise.
* This can't yet match the full vector reduce + start value idiom. That
will be an upcoming set of changes.