[all-commits] [llvm/llvm-project] 747d4a: [DirectX] Implement UseNativeLowPrecision shader f...

[Deric C. via All-commits]

  Branch: refs/heads/main
  Home:   https://github.com/llvm/llvm-project
  Commit: 747d4a952bf7ed4adec72ddf3c9038aeff4fe8ee
      https://github.com/llvm/llvm-project/commit/747d4a952bf7ed4adec72ddf3c9038aeff4fe8ee
  Author: Deric C. <cheung.deric at gmail.com>
  Date:   2025-04-09 (Wed, 09 Apr 2025)

  Changed paths:
    M clang/lib/CodeGen/CGHLSLRuntime.cpp
    A clang/test/CodeGenHLSL/enable-16bit-types.hlsl
    M llvm/lib/Target/DirectX/DXILShaderFlags.cpp
    M llvm/test/CodeGen/DirectX/ShaderFlags/low-precision.ll
    A llvm/test/CodeGen/DirectX/ShaderFlags/use-native-low-precision.ll

  Log Message:
  -----------
  [DirectX] Implement UseNativeLowPrecision shader flag analysis (#134288)

Fixes #112267 

Implement the shader flag analysis to set the UseNativeLowPrecision DXIL
module flag.

The flag is only able to be set when the command-line flag
`-enable-16bit-types` is passed to clang-dxc, or equivalently
`-fnative-half-type` is passed to clang.
When the command-line flag is passed, a module metadata flag called
"dx.nativelowprec" is set to 1.
The DXILShaderFlags shader flags analysis checks that the module
metadata flag "dx.nativelowprec" is set to 1 and the DXIL Version is 1.2
or greater before setting the UseNativeLowPrecision DXIL module flag.



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