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
44
45
46
47
48
49
50
51
52
53
54
| # RUN: llc -mtriple=amdgcn-amd-amdhsa -verify-machineinstrs -run-pass=prologepilog %s -o - | FileCheck %s
# CHECK-LABEL: all_sgpr_used
# CHECK: V_CMP_LT_U32_e64
--- |
define amdgpu_kernel void @all_sgpr_used() #0 {
ret void
}
attributes #0 = { "amdgpu-num-sgpr"="8" "frame-pointer"="all"}
...
---
name: all_sgpr_used
tracksRegLiveness: true
liveins:
- { reg: '$vgpr0' }
- { reg: '$vgpr1' }
- { reg: '$vgpr2' }
- { reg: '$sgpr4_sgpr5' }
- { reg: '$sgpr6_sgpr7' }
- { reg: '$sgpr8' }
- { reg: '$sgpr9' }
machineFunctionInfo:
explicitKernArgSize: 84
maxKernArgAlign: 8
ldsSize: 20496
isEntryFunction: true
waveLimiter: true
scratchRSrcReg: '$sgpr96_sgpr97_sgpr98_sgpr99'
scratchWaveOffsetReg: '$sgpr101'
frameOffsetReg: '$sgpr101'
stackPtrOffsetReg: '$sgpr32'
argumentInfo:
privateSegmentBuffer: { reg: '$sgpr0_sgpr1_sgpr2_sgpr3' }
dispatchPtr: { reg: '$sgpr4_sgpr5' }
kernargSegmentPtr: { reg: '$sgpr6_sgpr7' }
workGroupIDX: { reg: '$sgpr8' }
workGroupIDY: { reg: '$sgpr9' }
privateSegmentWaveByteOffset: { reg: '$sgpr10' }
workItemIDX: { reg: '$vgpr0' }
workItemIDY: { reg: '$vgpr1' }
workItemIDZ: { reg: '$vgpr2' }
body: |
bb.0:
liveins: $sgpr8, $sgpr9, $vgpr0, $vgpr1, $vgpr2, $sgpr4_sgpr5, $sgpr6_sgpr7
$sgpr0 = S_MOV_B32 0
$sgpr1 = S_MOV_B32 0
$sgpr2 = S_MOV_B32 0
$sgpr3 = S_MOV_B32 0
$sgpr4 = S_MOV_B32 0
$sgpr5 = S_MOV_B32 0
$sgpr6 = S_MOV_B32 0
$sgpr7 = S_MOV_B32 0
$vcc = V_CMP_LT_U32_e64 $sgpr8, $vgpr1, implicit $exec
...
|