fc0653f31c
Recent upstream trends have moved away from explicitly using `-verify-machineinstrs`, as it's already covered by the expensive checks. This PR removes almost all `-verify-machineinstrs` from tests in `llvm/test/CodeGen/AMDGPU/*.ll`, leaving only those tests where its removal currently causes failures.
28 lines
727 B
LLVM
28 lines
727 B
LLVM
; RUN: llc -mtriple=amdgcn -mcpu=gfx908 < %s | FileCheck -check-prefixes=GCN,GFX908 %s
|
|
; RUN: llc -mtriple=amdgcn -mcpu=gfx900 < %s | FileCheck -check-prefixes=GCN,GFX900 %s
|
|
|
|
define void @foo() {
|
|
bb:
|
|
ret void
|
|
}
|
|
|
|
; FIXME: We spill v40 into AGPR, but still save and restore FP
|
|
; which is not needed in this case.
|
|
|
|
; GCN-LABEL: {{^}}caller:
|
|
|
|
; GCN: s_mov_b32 [[TMP_SGPR:s[0-9]+]], s33
|
|
; GCN: s_mov_b32 s33, s32
|
|
; GFX900: buffer_store_dword
|
|
; GFX908-DAG: v_accvgpr_write_b32
|
|
; GCN: s_swappc_b64
|
|
; GFX900: buffer_load_dword
|
|
; GFX908: v_accvgpr_read_b32
|
|
; GCN: s_mov_b32 s33, [[TMP_SGPR]]
|
|
define i64 @caller() {
|
|
bb:
|
|
call void asm sideeffect "", "~{v40}" ()
|
|
tail call void @foo()
|
|
ret i64 0
|
|
}
|