Files
archived-llvm/test/CodeGen/AMDGPU/private-access-no-objects.ll
Matt Arsenault c31d80dbf5 AMDGPU: Assume spilling will occur at -O0
Because everything live is spilled at the end of a
block by fast regalloc, assume this will happen and
avoid the copies of the resource descriptor.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@284119 91177308-0d34-0410-b5e6-96231b3b80d8
2016-10-13 13:10:00 +00:00

34 lines
936 B
LLVM

; RUN: llc -O0 -mtriple=amdgcn--amdhsa -mcpu=fiji -verify-machineinstrs < %s | FileCheck -check-prefix=GCN -check-prefix=VI -check-prefix=OPTNONE %s
; GCN-LABEL: {{^}}store_to_undef:
; -O0 should assume spilling, so the input scratch resource descriptor
; -should be used directly without any copies.
; OPTNONE-NOT: s_mov_b32
; OPTNONE: buffer_store_dword v{{[0-9]+}}, v{{[0-9]+}}, s[0:3], s7 offen{{$}}
define void @store_to_undef() #0 {
store volatile i32 0, i32* undef
ret void
}
; GCN-LABEL: {{^}}store_to_inttoptr:
define void @store_to_inttoptr() #0 {
store volatile i32 0, i32* inttoptr (i32 123 to i32*)
ret void
}
; GCN-LABEL: {{^}}load_from_undef:
define void @load_from_undef() #0 {
%ld = load volatile i32, i32* undef
ret void
}
; GCN-LABEL: {{^}}load_from_inttoptr:
define void @load_from_inttoptr() #0 {
%ld = load volatile i32, i32* inttoptr (i32 123 to i32*)
ret void
}
attributes #0 = { nounwind }