9e9907f1cf
Similar to 806761a762.
For IR files without a target triple, -mtriple= specifies the full
target triple while -march= merely sets the architecture part of the
default target triple, leaving a target triple which may not make sense,
e.g. amdgpu-apple-darwin.
Therefore, -march= is error-prone and not recommended for tests without
a target triple. The issue has been benign as we recognize
$unknown-apple-darwin as ELF instead of rejecting it outrightly.
This patch changes AMDGPU tests to not rely on the default
OS/environment components. Tests that need fixes are not changed:
```
LLVM :: CodeGen/AMDGPU/fabs.f64.ll
LLVM :: CodeGen/AMDGPU/fabs.ll
LLVM :: CodeGen/AMDGPU/floor.ll
LLVM :: CodeGen/AMDGPU/fneg-fabs.f64.ll
LLVM :: CodeGen/AMDGPU/fneg-fabs.ll
LLVM :: CodeGen/AMDGPU/r600-infinite-loop-bug-while-reorganizing-vector.ll
LLVM :: CodeGen/AMDGPU/schedule-if-2.ll
```
32 lines
871 B
LLVM
32 lines
871 B
LLVM
; RUN: llc -mtriple=amdgcn < %s | FileCheck %s
|
|
; RUN: llc -mtriple=amdgcn < %s -global-isel | FileCheck %s
|
|
|
|
; CHECK-LABEL: {{^}}unknown_wgs:
|
|
; CHECK: s_barrier
|
|
define amdgpu_kernel void @unknown_wgs() {
|
|
tail call void @llvm.amdgcn.s.barrier() #0
|
|
ret void
|
|
}
|
|
|
|
; CHECK-LABEL: {{^}}flat_wgs_attr_32_128:
|
|
; CHECK: s_barrier
|
|
define amdgpu_kernel void @flat_wgs_attr_32_128() #1 {
|
|
tail call void @llvm.amdgcn.s.barrier() #0
|
|
ret void
|
|
}
|
|
|
|
; CHECK-LABEL: {{^}}flat_wgs_attr_32_64:
|
|
; CHECK: :
|
|
; CHECK-NEXT: ; wave barrier
|
|
; CHECK-NEXT: s_endpgm
|
|
define amdgpu_kernel void @flat_wgs_attr_32_64() #2 {
|
|
tail call void @llvm.amdgcn.s.barrier() #0
|
|
ret void
|
|
}
|
|
|
|
declare void @llvm.amdgcn.s.barrier() #0
|
|
|
|
attributes #0 = { convergent nounwind }
|
|
attributes #1 = { nounwind "amdgpu-flat-work-group-size"="32,128" }
|
|
attributes #2 = { nounwind "amdgpu-flat-work-group-size"="32,64" }
|