mirror of
https://github.com/RPCSX/llvm.git
synced 2024-11-23 19:59:57 +00:00
d706d030af
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). git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@298444 91177308-0d34-0410-b5e6-96231b3b80d8
47 lines
1.7 KiB
LLVM
47 lines
1.7 KiB
LLVM
; RUN: llc -march=amdgcn -mattr=-promote-alloca -amdgpu-sroa=0 -verify-machineinstrs < %s | FileCheck -check-prefix=SI -check-prefix=FUNC %s
|
|
; RUN: llc -march=amdgcn -mcpu=tonga -mattr=-promote-alloca -amdgpu-sroa=0 -verify-machineinstrs < %s | FileCheck -check-prefix=SI -check-prefix=FUNC %s
|
|
|
|
; FUNC-LABEL: {{^}}load_i8_sext_private:
|
|
; SI: buffer_load_sbyte v{{[0-9]+}}, off, s[{{[0-9]+:[0-9]+}}], s{{[0-9]+}} offset:4{{$}}
|
|
define amdgpu_kernel void @load_i8_sext_private(i32 addrspace(1)* %out) {
|
|
entry:
|
|
%tmp0 = alloca i8
|
|
%tmp1 = load i8, i8* %tmp0
|
|
%tmp2 = sext i8 %tmp1 to i32
|
|
store i32 %tmp2, i32 addrspace(1)* %out
|
|
ret void
|
|
}
|
|
|
|
; FUNC-LABEL: {{^}}load_i8_zext_private:
|
|
; SI: buffer_load_ubyte v{{[0-9]+}}, off, s[{{[0-9]+:[0-9]+}}], s{{[0-9]+}} offset:4{{$}}
|
|
define amdgpu_kernel void @load_i8_zext_private(i32 addrspace(1)* %out) {
|
|
entry:
|
|
%tmp0 = alloca i8
|
|
%tmp1 = load i8, i8* %tmp0
|
|
%tmp2 = zext i8 %tmp1 to i32
|
|
store i32 %tmp2, i32 addrspace(1)* %out
|
|
ret void
|
|
}
|
|
|
|
; FUNC-LABEL: {{^}}load_i16_sext_private:
|
|
; SI: buffer_load_sshort v{{[0-9]+}}, off, s[{{[0-9]+:[0-9]+}}], s{{[0-9]+}} offset:4{{$}}
|
|
define amdgpu_kernel void @load_i16_sext_private(i32 addrspace(1)* %out) {
|
|
entry:
|
|
%tmp0 = alloca i16
|
|
%tmp1 = load i16, i16* %tmp0
|
|
%tmp2 = sext i16 %tmp1 to i32
|
|
store i32 %tmp2, i32 addrspace(1)* %out
|
|
ret void
|
|
}
|
|
|
|
; FUNC-LABEL: {{^}}load_i16_zext_private:
|
|
; SI: buffer_load_ushort v{{[0-9]+}}, off, s[{{[0-9]+:[0-9]+}}], s{{[0-9]+}} offset:4{{$}}
|
|
define amdgpu_kernel void @load_i16_zext_private(i32 addrspace(1)* %out) {
|
|
entry:
|
|
%tmp0 = alloca i16
|
|
%tmp1 = load volatile i16, i16* %tmp0
|
|
%tmp2 = zext i16 %tmp1 to i32
|
|
store i32 %tmp2, i32 addrspace(1)* %out
|
|
ret void
|
|
}
|