intel/llvm
1
0
Fork 0
mirror of https://github.com/intel/llvm.git synced 2026-02-04 03:26:06 +08:00
Code Issues Packages Projects Releases Wiki Activity

[CUDA] Set LLVM calling convention for CUDA kernel

Browse Source 
Some targets need special LLVM calling convention for CUDA kernel.
This patch does that through a TargetCodeGenInfo hook.

It only affects amdgcn target.

Patch by Greg Rodgers.
Revised and lit tests added by Yaxun Liu.

Differential Revision: https://reviews.llvm.org/D45223

llvm-svn: 330447
This commit is contained in:
Yaxun Liu
2018-04-20 17:01:03 +00:00
parent e268304122
commit 4306f2086f
4 changed files with 52 additions and 0 deletions
Download Patch File Download Diff File Expand all files Collapse all files

3
clang/lib/CodeGen/CodeGenModule.cpp
View File

@@ -3627,6 +3627,9 @@ void CodeGenModule::EmitGlobalFunctionDefinition(GlobalDecl GD,
MaybeHandleStaticInExternC(D, Fn);
if (D->hasAttr<CUDAGlobalAttr>())
getTargetCodeGenInfo().setCUDAKernelCallingConvention(Fn);
maybeSetTrivialComdat(*D, *Fn);
CodeGenFunction(*this).GenerateCode(D, Fn, FI);