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
```
45 lines
2.2 KiB
YAML
45 lines
2.2 KiB
YAML
# NOTE: Assertions have been autogenerated by utils/update_mir_test_checks.py
|
|
# RUN: llc -mtriple=amdgcn -mcpu=gfx900 -verify-machineinstrs -run-pass=post-RA-sched %s -o - | FileCheck -check-prefix=GCN %s
|
|
|
|
# Check that we move consumer further from producer, even if one of them is in a bundle.
|
|
|
|
---
|
|
name: src_bundle_latency
|
|
tracksRegLiveness: true
|
|
body: |
|
|
bb.0:
|
|
; GCN-LABEL: name: src_bundle_latency
|
|
; GCN: $vgpr0, $vgpr1 = BUNDLE undef $vgpr3_vgpr4, implicit $exec {
|
|
; GCN-NEXT: $vgpr0 = GLOBAL_LOAD_DWORD undef $vgpr3_vgpr4, 0, 0, implicit $exec
|
|
; GCN-NEXT: $vgpr1 = GLOBAL_LOAD_DWORD undef $vgpr3_vgpr4, 4, 0, implicit $exec
|
|
; GCN-NEXT: }
|
|
; GCN-NEXT: $vgpr6 = V_ADD_F32_e32 killed $vgpr0, $vgpr0, implicit $mode, implicit $exec
|
|
; GCN-NEXT: $vgpr5 = V_ADD_F32_e32 killed $vgpr1, $vgpr1, implicit $mode, implicit $exec
|
|
$vgpr0, $vgpr1 = BUNDLE undef $vgpr3_vgpr4, implicit $exec {
|
|
$vgpr0 = GLOBAL_LOAD_DWORD undef $vgpr3_vgpr4, 0, 0, implicit $exec
|
|
$vgpr1 = GLOBAL_LOAD_DWORD undef $vgpr3_vgpr4, 4, 0, implicit $exec
|
|
}
|
|
$vgpr5 = V_ADD_F32_e32 $vgpr1, $vgpr1, implicit $mode, implicit $exec
|
|
$vgpr6 = V_ADD_F32_e32 $vgpr0, $vgpr0, implicit $mode, implicit $exec
|
|
...
|
|
|
|
---
|
|
name: dst_bundle_latency
|
|
tracksRegLiveness: true
|
|
body: |
|
|
bb.0:
|
|
; GCN-LABEL: name: dst_bundle_latency
|
|
; GCN: $vgpr1 = V_ADD_F32_e32 undef $vgpr6, undef $vgpr6, implicit $mode, implicit $exec
|
|
; GCN-NEXT: $vgpr0 = V_ADD_F32_e32 undef $vgpr5, undef $vgpr5, implicit $mode, implicit $exec
|
|
; GCN-NEXT: BUNDLE killed $vgpr0, killed $vgpr1, undef $vgpr3_vgpr4, implicit $exec {
|
|
; GCN-NEXT: GLOBAL_STORE_DWORD undef $vgpr3_vgpr4, killed $vgpr1, 0, 0, implicit $exec
|
|
; GCN-NEXT: GLOBAL_STORE_DWORD undef $vgpr3_vgpr4, killed $vgpr0, 4, 0, implicit $exec
|
|
; GCN-NEXT: }
|
|
$vgpr0 = V_ADD_F32_e32 undef $vgpr5, undef $vgpr5, implicit $mode, implicit $exec
|
|
$vgpr1 = V_ADD_F32_e32 undef $vgpr6, undef $vgpr6, implicit $mode, implicit $exec
|
|
BUNDLE $vgpr0, $vgpr1, undef $vgpr3_vgpr4, implicit $exec {
|
|
GLOBAL_STORE_DWORD undef $vgpr3_vgpr4, $vgpr1, 0, 0, implicit $exec
|
|
GLOBAL_STORE_DWORD undef $vgpr3_vgpr4, $vgpr0, 4, 0, implicit $exec
|
|
}
|
|
...
|