[all-commits] [llvm/llvm-project] 5a924b: [CIR][CUDA] Support built-in CUDA surface type (#1...

[Orphic Dreamer via All-commits]

  Branch: refs/heads/main
  Home:   https://github.com/llvm/llvm-project
  Commit: 5a924bf3962c7f0c84e2de46172cd1d01ae3a67d
      https://github.com/llvm/llvm-project/commit/5a924bf3962c7f0c84e2de46172cd1d01ae3a67d
  Author: Orphic Dreamer <160723353+Siya-05 at users.noreply.github.com>
  Date:   2026-07-08 (Wed, 08 Jul 2026)

  Changed paths:
    M clang/include/clang/CIR/MissingFeatures.h
    M clang/lib/CIR/CodeGen/CIRGenCUDANV.cpp
    M clang/lib/CIR/CodeGen/CIRGenTypes.cpp
    M clang/lib/CIR/CodeGen/CMakeLists.txt
    M clang/lib/CIR/CodeGen/TargetInfo.cpp
    M clang/lib/CIR/CodeGen/TargetInfo.h
    A clang/lib/CIR/CodeGen/Targets/NVPTX.cpp
    A clang/test/CIR/CodeGenCUDA/surface.cu

  Log Message:
  -----------
  [CIR][CUDA] Support built-in CUDA surface type (#196079)

Related: #179278

This patch adds initial support for CUDA built-in surface types in CIR
for device-side compilation.

CUDA surface references are lowered to the NVPTX device-handle
representation (`i64`), matching existing Clang CodeGen behavior.

Changes
* Add `getCUDADeviceBuiltinSurfaceDeviceType()` target hook to
`TargetCIRGenInfo`
* Implement NVPTX surface lowering in `NVPTXTargetCIRGenInfo`
* Handle CUDA built-in surface types in `CIRGenTypes::convertType`
* Add CIR CUDA test coverage for device-side surface lowering

Notes

* This patch only implements device-side surface type lowering support
* Texture types remain unsupported
* TBAA handling for surface/texture types is left for follow-up work

---------

Co-authored-by: mencotton <mencotton0410 at gmail.com>
Co-authored-by: David Rivera <davidriverg at gmail.com>
Co-authored-by: Andy Kaylor <akaylor at nvidia.com>



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