mirror of
https://github.com/RPCS3/llvm-mirror.git
synced 2024-10-23 21:13:02 +02:00
dd9ab77318
Currently the default C calling convention functions are treated the same as compute kernels. Make this explicit so the default calling convention can be changed to a non-kernel. Converted with perl -pi -e 's/define void/define amdgpu_kernel void/' on the relevant test directories (and undoing in one place that actually wanted a non-kernel). llvm-svn: 298444
13 lines
596 B
LLVM
13 lines
596 B
LLVM
; RUN: llc -march=amdgcn -mcpu=SI -verify-machineinstrs< %s
|
|
; RUN: llc -march=amdgcn -mcpu=tonga -verify-machineinstrs< %s
|
|
|
|
; This tests for a bug that caused a crash in
|
|
; AMDGPUDAGToDAGISel::SelectMUBUFScratch() which is used for selecting
|
|
; scratch loads and stores.
|
|
; CHECK-LABEL: {{^}}store_vector_ptrs:
|
|
define amdgpu_kernel void @store_vector_ptrs(<4 x i32*>* %out, <4 x [1024 x i32]*> %array) nounwind {
|
|
%p = getelementptr [1024 x i32], <4 x [1024 x i32]*> %array, <4 x i16> zeroinitializer, <4 x i16> <i16 16, i16 16, i16 16, i16 16>
|
|
store <4 x i32*> %p, <4 x i32*>* %out
|
|
ret void
|
|
}
|