File: store-barrier.ll

package info (click to toggle)
llvm-toolchain-17 1%3A17.0.6-22
  • links: PTS, VCS
  • area: main
  • in suites: forky, sid, trixie
  • size: 1,799,624 kB
  • sloc: cpp: 6,428,607; ansic: 1,383,196; asm: 793,408; python: 223,504; objc: 75,364; f90: 60,502; lisp: 33,869; pascal: 15,282; sh: 9,684; perl: 7,453; ml: 4,937; awk: 3,523; makefile: 2,889; javascript: 2,149; xml: 888; fortran: 619; cs: 573
file content (43 lines) | stat: -rw-r--r-- 2,134 bytes parent folder | download | duplicates (4)
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
; RUN: llc -march=amdgcn -verify-machineinstrs -mattr=+load-store-opt -enable-misched < %s | FileCheck %s
; RUN: llc -march=amdgcn -mcpu=bonaire -verify-machineinstrs -mattr=+load-store-opt -enable-misched < %s | FileCheck %s

; This test is for a bug in the machine scheduler where stores without
; an underlying object would be moved across the barrier.  In this
; test, the <2 x i8> store will be split into two i8 stores, so they
; won't have an underlying object.

; CHECK-LABEL: {{^}}test:
; CHECK: ds_write_b8
; CHECK: ds_write_b8
; CHECK: s_barrier
; CHECK: s_endpgm
; Function Attrs: nounwind
define amdgpu_kernel void @test(ptr addrspace(3) nocapture %arg, ptr addrspace(1) nocapture readonly %arg1, ptr addrspace(1) nocapture readonly %arg2, ptr addrspace(1) nocapture %arg3, i32 %arg4, i64 %tmp9) #0 {
bb:
  %tmp10 = getelementptr inbounds i32, ptr addrspace(1) %arg2, i64 %tmp9
  %tmp13 = load i32, ptr addrspace(1) %tmp10, align 2
  %tmp14 = getelementptr inbounds <2 x i8>, ptr addrspace(3) %arg, i32 %tmp13
  %tmp15 = load <2 x i8>, ptr addrspace(3) %tmp14, align 1
  %tmp16 = add i32 %tmp13, 1
  %tmp17 = getelementptr inbounds <2 x i8>, ptr addrspace(3) %arg, i32 %tmp16
  store <2 x i8> %tmp15, ptr addrspace(3) %tmp17, align 1
  tail call void @llvm.amdgcn.s.barrier()
  %tmp25 = load i32, ptr addrspace(1) %tmp10, align 4
  %tmp26 = sext i32 %tmp25 to i64
  %tmp27 = sext i32 %arg4 to i64
  %tmp28 = getelementptr inbounds <2 x i8>, ptr addrspace(3) %arg, i32 %tmp25, i32 %arg4
  %tmp29 = load i8, ptr addrspace(3) %tmp28, align 1
  %tmp30 = getelementptr inbounds <2 x i8>, ptr addrspace(1) %arg3, i64 %tmp26, i64 %tmp27
  store i8 %tmp29, ptr addrspace(1) %tmp30, align 1
  %tmp32 = getelementptr inbounds <2 x i8>, ptr addrspace(3) %arg, i32 %tmp25, i32 0
  %tmp33 = load i8, ptr addrspace(3) %tmp32, align 1
  %tmp35 = getelementptr inbounds <2 x i8>, ptr addrspace(1) %arg3, i64 %tmp26, i64 0
  store i8 %tmp33, ptr addrspace(1) %tmp35, align 1
  ret void
}

; Function Attrs: convergent nounwind
declare void @llvm.amdgcn.s.barrier() #1

attributes #0 = { nounwind }
attributes #1 = { convergent nounwind }