aboutsummaryrefslogtreecommitdiff
path: root/llvm/test/CodeGen/AMDGPU/gep-flags-stack-offsets.ll
blob: b5f0b2ff9ef4cf429e6a41003b65ab9de9047612 (plain)
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
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
; NOTE: Assertions have been autogenerated by utils/update_llc_test_checks.py UTC_ARGS: --version 5
; gfx8 required knowing no overflow happened to fold the addressing mode
; RUN: llc -mtriple=amdgcn-amd-amdhsa -mcpu=gfx803 < %s | FileCheck -check-prefix=GFX8 %s
; RUN: llc -mtriple=amdgcn-amd-amdhsa -mcpu=gfx900 < %s | FileCheck -check-prefix=GFX9 %s

define void @gep_noflags_alloca(i32 %idx, i32 %val) #0 {
; GFX8-LABEL: gep_noflags_alloca:
; GFX8:       ; %bb.0:
; GFX8-NEXT:    s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0)
; GFX8-NEXT:    v_lshlrev_b32_e32 v0, 2, v0
; GFX8-NEXT:    v_lshrrev_b32_e64 v2, 6, s32
; GFX8-NEXT:    v_add_u32_e32 v0, vcc, v0, v2
; GFX8-NEXT:    v_add_u32_e32 v0, vcc, 16, v0
; GFX8-NEXT:    buffer_store_dword v1, v0, s[0:3], 0 offen
; GFX8-NEXT:    s_waitcnt vmcnt(0)
; GFX8-NEXT:    s_setpc_b64 s[30:31]
;
; GFX9-LABEL: gep_noflags_alloca:
; GFX9:       ; %bb.0:
; GFX9-NEXT:    s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0)
; GFX9-NEXT:    v_lshrrev_b32_e64 v2, 6, s32
; GFX9-NEXT:    v_lshl_add_u32 v0, v0, 2, v2
; GFX9-NEXT:    buffer_store_dword v1, v0, s[0:3], 0 offen offset:16
; GFX9-NEXT:    s_waitcnt vmcnt(0)
; GFX9-NEXT:    s_setpc_b64 s[30:31]
  %alloca = alloca [32 x i32], addrspace(5)
  %gep0 = getelementptr [32 x i32], ptr addrspace(5) %alloca, i32 0, i32 %idx
  %gep1 = getelementptr i32, ptr addrspace(5) %gep0, i32 4
  store volatile i32 %val, ptr addrspace(5) %gep1
  ret void
}

define void @gep_inbounds_alloca(i32 %idx, i32 %val) #0 {
; GFX8-LABEL: gep_inbounds_alloca:
; GFX8:       ; %bb.0:
; GFX8-NEXT:    s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0)
; GFX8-NEXT:    v_lshlrev_b32_e32 v0, 2, v0
; GFX8-NEXT:    v_lshrrev_b32_e64 v2, 6, s32
; GFX8-NEXT:    v_add_u32_e32 v0, vcc, v0, v2
; GFX8-NEXT:    v_add_u32_e32 v0, vcc, 16, v0
; GFX8-NEXT:    buffer_store_dword v1, v0, s[0:3], 0 offen
; GFX8-NEXT:    s_waitcnt vmcnt(0)
; GFX8-NEXT:    s_setpc_b64 s[30:31]
;
; GFX9-LABEL: gep_inbounds_alloca:
; GFX9:       ; %bb.0:
; GFX9-NEXT:    s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0)
; GFX9-NEXT:    v_lshrrev_b32_e64 v2, 6, s32
; GFX9-NEXT:    v_lshl_add_u32 v0, v0, 2, v2
; GFX9-NEXT:    buffer_store_dword v1, v0, s[0:3], 0 offen offset:16
; GFX9-NEXT:    s_waitcnt vmcnt(0)
; GFX9-NEXT:    s_setpc_b64 s[30:31]
  %alloca = alloca [32 x i32], addrspace(5)
  %gep0 = getelementptr inbounds [32 x i32], ptr addrspace(5) %alloca, i32 0, i32 %idx
  %gep1 = getelementptr i32, ptr addrspace(5) %gep0, i32 4
  store volatile i32 %val, ptr addrspace(5) %gep1
  ret void
}

define void @gep_nuw_alloca(i32 %idx, i32 %val) #0 {
; GFX8-LABEL: gep_nuw_alloca:
; GFX8:       ; %bb.0:
; GFX8-NEXT:    s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0)
; GFX8-NEXT:    v_lshlrev_b32_e32 v0, 2, v0
; GFX8-NEXT:    v_lshrrev_b32_e64 v2, 6, s32
; GFX8-NEXT:    v_add_u32_e32 v0, vcc, v0, v2
; GFX8-NEXT:    v_add_u32_e32 v0, vcc, 16, v0
; GFX8-NEXT:    buffer_store_dword v1, v0, s[0:3], 0 offen
; GFX8-NEXT:    s_waitcnt vmcnt(0)
; GFX8-NEXT:    s_setpc_b64 s[30:31]
;
; GFX9-LABEL: gep_nuw_alloca:
; GFX9:       ; %bb.0:
; GFX9-NEXT:    s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0)
; GFX9-NEXT:    v_lshrrev_b32_e64 v2, 6, s32
; GFX9-NEXT:    v_lshl_add_u32 v0, v0, 2, v2
; GFX9-NEXT:    buffer_store_dword v1, v0, s[0:3], 0 offen offset:16
; GFX9-NEXT:    s_waitcnt vmcnt(0)
; GFX9-NEXT:    s_setpc_b64 s[30:31]
  %alloca = alloca [32 x i32], addrspace(5)
  %gep0 = getelementptr nuw [32 x i32], ptr addrspace(5) %alloca, i32 0, i32 %idx
  %gep1 = getelementptr i32, ptr addrspace(5) %gep0, i32 4
  store volatile i32 %val, ptr addrspace(5) %gep1
  ret void
}

define void @gep_nusw_alloca(i32 %idx, i32 %val) #0 {
; GFX8-LABEL: gep_nusw_alloca:
; GFX8:       ; %bb.0:
; GFX8-NEXT:    s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0)
; GFX8-NEXT:    v_lshlrev_b32_e32 v0, 2, v0
; GFX8-NEXT:    v_lshrrev_b32_e64 v2, 6, s32
; GFX8-NEXT:    v_add_u32_e32 v0, vcc, v0, v2
; GFX8-NEXT:    v_add_u32_e32 v0, vcc, 16, v0
; GFX8-NEXT:    buffer_store_dword v1, v0, s[0:3], 0 offen
; GFX8-NEXT:    s_waitcnt vmcnt(0)
; GFX8-NEXT:    s_setpc_b64 s[30:31]
;
; GFX9-LABEL: gep_nusw_alloca:
; GFX9:       ; %bb.0:
; GFX9-NEXT:    s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0)
; GFX9-NEXT:    v_lshrrev_b32_e64 v2, 6, s32
; GFX9-NEXT:    v_lshl_add_u32 v0, v0, 2, v2
; GFX9-NEXT:    buffer_store_dword v1, v0, s[0:3], 0 offen offset:16
; GFX9-NEXT:    s_waitcnt vmcnt(0)
; GFX9-NEXT:    s_setpc_b64 s[30:31]
  %alloca = alloca [32 x i32], addrspace(5)
  %gep0 = getelementptr nusw [32 x i32], ptr addrspace(5) %alloca, i32 0, i32 %idx
  %gep1 = getelementptr i32, ptr addrspace(5) %gep0, i32 4
  store volatile i32 %val, ptr addrspace(5) %gep1
  ret void
}

define void @gep_inbounds_nuw_alloca(i32 %idx, i32 %val) #0 {
; GFX8-LABEL: gep_inbounds_nuw_alloca:
; GFX8:       ; %bb.0:
; GFX8-NEXT:    s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0)
; GFX8-NEXT:    v_lshlrev_b32_e32 v0, 2, v0
; GFX8-NEXT:    v_lshrrev_b32_e64 v2, 6, s32
; GFX8-NEXT:    v_add_u32_e32 v0, vcc, v0, v2
; GFX8-NEXT:    v_add_u32_e32 v0, vcc, 16, v0
; GFX8-NEXT:    buffer_store_dword v1, v0, s[0:3], 0 offen
; GFX8-NEXT:    s_waitcnt vmcnt(0)
; GFX8-NEXT:    s_setpc_b64 s[30:31]
;
; GFX9-LABEL: gep_inbounds_nuw_alloca:
; GFX9:       ; %bb.0:
; GFX9-NEXT:    s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0)
; GFX9-NEXT:    v_lshrrev_b32_e64 v2, 6, s32
; GFX9-NEXT:    v_lshl_add_u32 v0, v0, 2, v2
; GFX9-NEXT:    buffer_store_dword v1, v0, s[0:3], 0 offen offset:16
; GFX9-NEXT:    s_waitcnt vmcnt(0)
; GFX9-NEXT:    s_setpc_b64 s[30:31]
  %alloca = alloca [32 x i32], addrspace(5)
  %gep0 = getelementptr inbounds nuw [32 x i32], ptr addrspace(5) %alloca, i32 0, i32 %idx
  %gep1 = getelementptr i32, ptr addrspace(5) %gep0, i32 4
  store volatile i32 %val, ptr addrspace(5) %gep1
  ret void
}

define void @gep_nusw_nuw_alloca(i32 %idx, i32 %val) #0 {
; GFX8-LABEL: gep_nusw_nuw_alloca:
; GFX8:       ; %bb.0:
; GFX8-NEXT:    s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0)
; GFX8-NEXT:    v_lshlrev_b32_e32 v0, 2, v0
; GFX8-NEXT:    v_lshrrev_b32_e64 v2, 6, s32
; GFX8-NEXT:    v_add_u32_e32 v0, vcc, v0, v2
; GFX8-NEXT:    v_add_u32_e32 v0, vcc, 16, v0
; GFX8-NEXT:    buffer_store_dword v1, v0, s[0:3], 0 offen
; GFX8-NEXT:    s_waitcnt vmcnt(0)
; GFX8-NEXT:    s_setpc_b64 s[30:31]
;
; GFX9-LABEL: gep_nusw_nuw_alloca:
; GFX9:       ; %bb.0:
; GFX9-NEXT:    s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0)
; GFX9-NEXT:    v_lshrrev_b32_e64 v2, 6, s32
; GFX9-NEXT:    v_lshl_add_u32 v0, v0, 2, v2
; GFX9-NEXT:    buffer_store_dword v1, v0, s[0:3], 0 offen offset:16
; GFX9-NEXT:    s_waitcnt vmcnt(0)
; GFX9-NEXT:    s_setpc_b64 s[30:31]
  %alloca = alloca [32 x i32], addrspace(5)
  %gep0 = getelementptr nusw nuw [32 x i32], ptr addrspace(5) %alloca, i32 0, i32 %idx
  %gep1 = getelementptr i32, ptr addrspace(5) %gep0, i32 4
  store volatile i32 %val, ptr addrspace(5) %gep1
  ret void
}

define void @gep_inbounds_nuw_alloca_nonpow2_scale(i32 %idx, i32 %val) #0 {
; GFX8-LABEL: gep_inbounds_nuw_alloca_nonpow2_scale:
; GFX8:       ; %bb.0:
; GFX8-NEXT:    s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0)
; GFX8-NEXT:    s_movk_i32 s4, 0x84
; GFX8-NEXT:    v_mul_lo_u32 v0, v0, s4
; GFX8-NEXT:    v_lshrrev_b32_e64 v2, 6, s32
; GFX8-NEXT:    v_add_u32_e32 v0, vcc, v0, v2
; GFX8-NEXT:    v_add_u32_e32 v0, vcc, 16, v0
; GFX8-NEXT:    buffer_store_dword v1, v0, s[0:3], 0 offen
; GFX8-NEXT:    s_waitcnt vmcnt(0)
; GFX8-NEXT:    s_setpc_b64 s[30:31]
;
; GFX9-LABEL: gep_inbounds_nuw_alloca_nonpow2_scale:
; GFX9:       ; %bb.0:
; GFX9-NEXT:    s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0)
; GFX9-NEXT:    s_movk_i32 s4, 0x84
; GFX9-NEXT:    v_mul_lo_u32 v0, v0, s4
; GFX9-NEXT:    v_lshrrev_b32_e64 v2, 6, s32
; GFX9-NEXT:    v_add_u32_e32 v0, v0, v2
; GFX9-NEXT:    buffer_store_dword v1, v0, s[0:3], 0 offen offset:16
; GFX9-NEXT:    s_waitcnt vmcnt(0)
; GFX9-NEXT:    s_setpc_b64 s[30:31]
  %alloca = alloca [5 x [33 x i32]], align 4, addrspace(5)
  %gep1 = getelementptr inbounds nuw [5 x [33 x i32]], ptr addrspace(5) %alloca, i32 0, i32 %idx, i32 4
  store volatile i32 %val, ptr addrspace(5) %gep1, align 4
  ret void
}

attributes #0 = { nounwind }