[all-commits] [llvm/llvm-project] 5e5fd0: [NVPTX] Select bfloat16 add/mul/sub as fma on SM80...

[peterbell10 via All-commits]

  Branch: refs/heads/main
  Home:   https://github.com/llvm/llvm-project
  Commit: 5e5fd0e6fc50cc1198750308c11433a5b3acfd0f
      https://github.com/llvm/llvm-project/commit/5e5fd0e6fc50cc1198750308c11433a5b3acfd0f
  Author: peterbell10 <peterbell10 at openai.com>
  Date:   2025-01-16 (Thu, 16 Jan 2025)

  Changed paths:
    M llvm/lib/Target/NVPTX/NVPTXISelDAGToDAG.cpp
    M llvm/lib/Target/NVPTX/NVPTXISelDAGToDAG.h
    M llvm/lib/Target/NVPTX/NVPTXISelLowering.cpp
    M llvm/lib/Target/NVPTX/NVPTXISelLowering.h
    M llvm/lib/Target/NVPTX/NVPTXSubtarget.cpp
    M llvm/lib/Target/NVPTX/NVPTXSubtarget.h
    M llvm/test/CodeGen/NVPTX/atomics-sm90.ll
    M llvm/test/CodeGen/NVPTX/bf16-instructions.ll
    M llvm/test/CodeGen/NVPTX/bf16x2-instructions.ll
    M llvm/test/CodeGen/NVPTX/fma-relu-contract.ll
    M llvm/test/CodeGen/NVPTX/fma-relu-fma-intrinsic.ll
    M llvm/test/CodeGen/NVPTX/fma-relu-instruction-flag.ll

  Log Message:
  -----------
  [NVPTX] Select bfloat16 add/mul/sub as fma on SM80 (#121065)

SM80 has fma for bfloat16 but not add/mul/sub. Currently these ops incur
a promotion to f32, but we can avoid this by writing them in terms of
the fma:
```
FADD(a, b) -> FMA(a, 1.0, b)
FMUL(a, b) -> FMA(a, b, -0.0)
FSUB(a, b) -> FMA(b, -1.0, a)
```

Unfortunately there is no `fma.ftz` so when ftz is enabled, we still
fall back to promotion.



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