[PATCH] D67048: [AMDGPU] Set default flat work group size to (1, 256) for HIP

[Yaxun Liu via Phabricator via cfe-commits]

This revision was automatically updated to reflect the committed changes.
Closed by commit rL370808: [AMDGPU] Set default flat work group size to (1,256) for HIP (authored by yaxunl, committed by ).
Herald added a project: LLVM.
Herald added a subscriber: llvm-commits.

Changed prior to commit:
  https://reviews.llvm.org/D67048?vs=218261&id=218508#toc

Repository:
  rL LLVM

CHANGES SINCE LAST ACTION
  https://reviews.llvm.org/D67048/new/

https://reviews.llvm.org/D67048

Files:
  cfe/trunk/lib/CodeGen/TargetInfo.cpp
  cfe/trunk/test/CodeGenCUDA/kernel-amdgcn.cu


Index: cfe/trunk/lib/CodeGen/TargetInfo.cpp
===================================================================
--- cfe/trunk/lib/CodeGen/TargetInfo.cpp
+++ cfe/trunk/lib/CodeGen/TargetInfo.cpp
@@ -7915,8 +7915,9 @@
 
   const bool IsOpenCLKernel = M.getLangOpts().OpenCL &&
                               FD->hasAttr<OpenCLKernelAttr>();
-  if ((IsOpenCLKernel ||
-       (M.getLangOpts().HIP && FD->hasAttr<CUDAGlobalAttr>())) &&
+  const bool IsHIPKernel = M.getLangOpts().HIP &&
+                           FD->hasAttr<CUDAGlobalAttr>();
+  if ((IsOpenCLKernel || IsHIPKernel) &&
       (M.getTriple().getOS() == llvm::Triple::AMDHSA))
     F->addFnAttr("amdgpu-implicitarg-num-bytes", "56");
 
@@ -7942,7 +7943,7 @@
       F->addFnAttr("amdgpu-flat-work-group-size", AttrVal);
     } else
       assert(Max == 0 && "Max must be zero");
-  } else if (IsOpenCLKernel) {
+  } else if (IsOpenCLKernel || IsHIPKernel) {
     // By default, restrict the maximum size to 256.
     F->addFnAttr("amdgpu-flat-work-group-size", "1,256");
   }
Index: cfe/trunk/test/CodeGenCUDA/kernel-amdgcn.cu
===================================================================
--- cfe/trunk/test/CodeGenCUDA/kernel-amdgcn.cu
+++ cfe/trunk/test/CodeGenCUDA/kernel-amdgcn.cu
@@ -1,4 +1,4 @@
-// RUN: %clang_cc1 -triple amdgcn -fcuda-is-device -emit-llvm %s -o - | FileCheck %s
+// RUN: %clang_cc1 -triple amdgcn -fcuda-is-device -emit-llvm -x hip %s -o - | FileCheck %s
 #include "Inputs/cuda.h"
 
 // CHECK: define amdgpu_kernel void @_ZN1A6kernelEv
@@ -25,7 +25,7 @@
   EmptyKernelPtr Empty() { return EmptyKernel<void>; } 
 };
 
-// CHECK: define amdgpu_kernel void @_Z15template_kernelI1AEvT_
+// CHECK: define amdgpu_kernel void @_Z15template_kernelI1AEvT_{{.*}} #[[ATTR:[0-9][0-9]*]]
 template<class T>
 __global__ void template_kernel(T x) {}
 
@@ -39,3 +39,4 @@
   launch((void*)D.Empty());
   return 0;
 }
+// CHECK: attributes #[[ATTR]] = {{.*}}"amdgpu-flat-work-group-size"="1,256"


-------------- next part --------------
A non-text attachment was scrubbed...
Name: D67048.218508.patch
Type: text/x-patch
Size: 1971 bytes
Desc: not available
URL: <http://lists.llvm.org/pipermail/cfe-commits/attachments/20190903/b1dd2d40/attachment.bin>