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
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
|
; NOTE: Assertions have been autogenerated by utils/update_llc_test_checks.py
; RUN: llc -mtriple=amdgcn-amd-amdhsa -mcpu=gfx900 -O0 < %s | FileCheck %s
; FP is in CSR range, modified.
define hidden fastcc void @callee_has_fp() #1 {
; CHECK-LABEL: callee_has_fp:
; CHECK: ; %bb.0:
; CHECK-NEXT: s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0)
; CHECK-NEXT: s_mov_b32 s4, s33
; CHECK-NEXT: s_mov_b32 s33, s32
; CHECK-NEXT: s_add_i32 s32, s32, 0x200
; CHECK-NEXT: v_mov_b32_e32 v0, 1
; CHECK-NEXT: buffer_store_dword v0, off, s[0:3], s33
; CHECK-NEXT: s_waitcnt vmcnt(0)
; CHECK-NEXT: s_mov_b32 s32, s33
; CHECK-NEXT: s_mov_b32 s33, s4
; CHECK-NEXT: s_setpc_b64 s[30:31]
%alloca = alloca i32, addrspace(5)
store volatile i32 1, ptr addrspace(5) %alloca
ret void
}
; Has no stack objects, but introduces them due to the CSR spill. We
; see the FP modified in the callee with IPRA. We should not have
; redundant spills of s33 or assert.
define internal fastcc void @csr_vgpr_spill_fp_callee() #0 {
; CHECK-LABEL: csr_vgpr_spill_fp_callee:
; CHECK: ; %bb.0: ; %bb
; CHECK-NEXT: s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0)
; CHECK-NEXT: s_mov_b32 s18, s33
; CHECK-NEXT: s_mov_b32 s33, s32
; CHECK-NEXT: s_xor_saveexec_b64 s[16:17], -1
; CHECK-NEXT: buffer_store_dword v1, off, s[0:3], s33 offset:4 ; 4-byte Folded Spill
; CHECK-NEXT: s_mov_b64 exec, s[16:17]
; CHECK-NEXT: s_add_i32 s32, s32, 0x400
; CHECK-NEXT: buffer_store_dword v40, off, s[0:3], s33 ; 4-byte Folded Spill
; CHECK-NEXT: v_writelane_b32 v1, s30, 0
; CHECK-NEXT: v_writelane_b32 v1, s31, 1
; CHECK-NEXT: s_getpc_b64 s[16:17]
; CHECK-NEXT: s_add_u32 s16, s16, callee_has_fp@rel32@lo+4
; CHECK-NEXT: s_addc_u32 s17, s17, callee_has_fp@rel32@hi+12
; CHECK-NEXT: s_mov_b64 s[22:23], s[2:3]
; CHECK-NEXT: s_mov_b64 s[20:21], s[0:1]
; CHECK-NEXT: s_mov_b64 s[0:1], s[20:21]
; CHECK-NEXT: s_mov_b64 s[2:3], s[22:23]
; CHECK-NEXT: s_swappc_b64 s[30:31], s[16:17]
; CHECK-NEXT: ;;#ASMSTART
; CHECK-NEXT: ; clobber csr v40
; CHECK-NEXT: ;;#ASMEND
; CHECK-NEXT: v_readlane_b32 s31, v1, 1
; CHECK-NEXT: v_readlane_b32 s30, v1, 0
; CHECK-NEXT: buffer_load_dword v40, off, s[0:3], s33 ; 4-byte Folded Reload
; CHECK-NEXT: s_mov_b32 s32, s33
; CHECK-NEXT: s_xor_saveexec_b64 s[4:5], -1
; CHECK-NEXT: buffer_load_dword v1, off, s[0:3], s33 offset:4 ; 4-byte Folded Reload
; CHECK-NEXT: s_mov_b64 exec, s[4:5]
; CHECK-NEXT: s_mov_b32 s33, s18
; CHECK-NEXT: s_waitcnt vmcnt(0)
; CHECK-NEXT: s_setpc_b64 s[30:31]
bb:
call fastcc void @callee_has_fp()
call void asm sideeffect "; clobber csr v40", "~{v40}"()
ret void
}
define amdgpu_kernel void @kernel_call() {
; CHECK-LABEL: kernel_call:
; CHECK: ; %bb.0: ; %bb
; CHECK-NEXT: s_mov_b32 s32, 0
; CHECK-NEXT: s_add_u32 flat_scratch_lo, s12, s17
; CHECK-NEXT: s_addc_u32 flat_scratch_hi, s13, 0
; CHECK-NEXT: s_add_u32 s0, s0, s17
; CHECK-NEXT: s_addc_u32 s1, s1, 0
; CHECK-NEXT: ; implicit-def: $vgpr3 : SGPR spill to VGPR lane
; CHECK-NEXT: v_writelane_b32 v3, s16, 0
; CHECK-NEXT: s_mov_b32 s13, s15
; CHECK-NEXT: s_mov_b32 s12, s14
; CHECK-NEXT: v_readlane_b32 s14, v3, 0
; CHECK-NEXT: s_getpc_b64 s[16:17]
; CHECK-NEXT: s_add_u32 s16, s16, csr_vgpr_spill_fp_callee@rel32@lo+4
; CHECK-NEXT: s_addc_u32 s17, s17, csr_vgpr_spill_fp_callee@rel32@hi+12
; CHECK-NEXT: s_mov_b64 s[22:23], s[2:3]
; CHECK-NEXT: s_mov_b64 s[20:21], s[0:1]
; CHECK-NEXT: s_mov_b32 s15, 20
; CHECK-NEXT: v_lshlrev_b32_e64 v2, s15, v2
; CHECK-NEXT: s_mov_b32 s15, 10
; CHECK-NEXT: v_lshlrev_b32_e64 v1, s15, v1
; CHECK-NEXT: v_or3_b32 v31, v0, v1, v2
; CHECK-NEXT: ; implicit-def: $sgpr15
; CHECK-NEXT: s_mov_b64 s[0:1], s[20:21]
; CHECK-NEXT: s_mov_b64 s[2:3], s[22:23]
; CHECK-NEXT: s_swappc_b64 s[30:31], s[16:17]
; CHECK-NEXT: s_endpgm
bb:
tail call fastcc void @csr_vgpr_spill_fp_callee()
ret void
}
; Same, except with a tail call.
define internal fastcc void @csr_vgpr_spill_fp_tailcall_callee() #0 {
; CHECK-LABEL: csr_vgpr_spill_fp_tailcall_callee:
; CHECK: ; %bb.0: ; %bb
; CHECK-NEXT: s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0)
; CHECK-NEXT: s_xor_saveexec_b64 s[16:17], -1
; CHECK-NEXT: buffer_store_dword v1, off, s[0:3], s32 offset:4 ; 4-byte Folded Spill
; CHECK-NEXT: s_mov_b64 exec, s[16:17]
; CHECK-NEXT: buffer_store_dword v40, off, s[0:3], s32 ; 4-byte Folded Spill
; CHECK-NEXT: v_writelane_b32 v1, s33, 0
; CHECK-NEXT: ;;#ASMSTART
; CHECK-NEXT: ; clobber csr v40
; CHECK-NEXT: ;;#ASMEND
; CHECK-NEXT: s_getpc_b64 s[16:17]
; CHECK-NEXT: s_add_u32 s16, s16, callee_has_fp@rel32@lo+4
; CHECK-NEXT: s_addc_u32 s17, s17, callee_has_fp@rel32@hi+12
; CHECK-NEXT: v_readlane_b32 s33, v1, 0
; CHECK-NEXT: buffer_load_dword v40, off, s[0:3], s32 ; 4-byte Folded Reload
; CHECK-NEXT: s_xor_saveexec_b64 s[18:19], -1
; CHECK-NEXT: buffer_load_dword v1, off, s[0:3], s32 offset:4 ; 4-byte Folded Reload
; CHECK-NEXT: s_mov_b64 exec, s[18:19]
; CHECK-NEXT: s_setpc_b64 s[16:17]
bb:
call void asm sideeffect "; clobber csr v40", "~{v40}"()
tail call fastcc void @callee_has_fp()
ret void
}
define amdgpu_kernel void @kernel_tailcall() {
; CHECK-LABEL: kernel_tailcall:
; CHECK: ; %bb.0: ; %bb
; CHECK-NEXT: s_mov_b32 s32, 0
; CHECK-NEXT: s_add_u32 flat_scratch_lo, s12, s17
; CHECK-NEXT: s_addc_u32 flat_scratch_hi, s13, 0
; CHECK-NEXT: s_add_u32 s0, s0, s17
; CHECK-NEXT: s_addc_u32 s1, s1, 0
; CHECK-NEXT: ; implicit-def: $vgpr3 : SGPR spill to VGPR lane
; CHECK-NEXT: v_writelane_b32 v3, s16, 0
; CHECK-NEXT: s_mov_b32 s13, s15
; CHECK-NEXT: s_mov_b32 s12, s14
; CHECK-NEXT: v_readlane_b32 s14, v3, 0
; CHECK-NEXT: s_getpc_b64 s[16:17]
; CHECK-NEXT: s_add_u32 s16, s16, csr_vgpr_spill_fp_tailcall_callee@rel32@lo+4
; CHECK-NEXT: s_addc_u32 s17, s17, csr_vgpr_spill_fp_tailcall_callee@rel32@hi+12
; CHECK-NEXT: s_mov_b64 s[22:23], s[2:3]
; CHECK-NEXT: s_mov_b64 s[20:21], s[0:1]
; CHECK-NEXT: s_mov_b32 s15, 20
; CHECK-NEXT: v_lshlrev_b32_e64 v2, s15, v2
; CHECK-NEXT: s_mov_b32 s15, 10
; CHECK-NEXT: v_lshlrev_b32_e64 v1, s15, v1
; CHECK-NEXT: v_or3_b32 v31, v0, v1, v2
; CHECK-NEXT: ; implicit-def: $sgpr15
; CHECK-NEXT: s_mov_b64 s[0:1], s[20:21]
; CHECK-NEXT: s_mov_b64 s[2:3], s[22:23]
; CHECK-NEXT: s_swappc_b64 s[30:31], s[16:17]
; CHECK-NEXT: s_endpgm
bb:
tail call fastcc void @csr_vgpr_spill_fp_tailcall_callee()
ret void
}
define hidden i32 @tail_call() #1 {
; CHECK-LABEL: tail_call:
; CHECK: ; %bb.0: ; %entry
; CHECK-NEXT: s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0)
; CHECK-NEXT: s_mov_b32 s4, s33
; CHECK-NEXT: s_mov_b32 s33, s32
; CHECK-NEXT: v_mov_b32_e32 v0, 0
; CHECK-NEXT: s_mov_b32 s33, s4
; CHECK-NEXT: s_setpc_b64 s[30:31]
entry:
ret i32 0
}
define hidden i32 @caller_save_vgpr_spill_fp_tail_call() #0 {
; CHECK-LABEL: caller_save_vgpr_spill_fp_tail_call:
; CHECK: ; %bb.0: ; %entry
; CHECK-NEXT: s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0)
; CHECK-NEXT: s_mov_b32 s18, s33
; CHECK-NEXT: s_mov_b32 s33, s32
; CHECK-NEXT: s_xor_saveexec_b64 s[16:17], -1
; CHECK-NEXT: buffer_store_dword v1, off, s[0:3], s33 ; 4-byte Folded Spill
; CHECK-NEXT: s_mov_b64 exec, s[16:17]
; CHECK-NEXT: s_add_i32 s32, s32, 0x400
; CHECK-NEXT: v_writelane_b32 v1, s30, 0
; CHECK-NEXT: v_writelane_b32 v1, s31, 1
; CHECK-NEXT: s_getpc_b64 s[16:17]
; CHECK-NEXT: s_add_u32 s16, s16, tail_call@rel32@lo+4
; CHECK-NEXT: s_addc_u32 s17, s17, tail_call@rel32@hi+12
; CHECK-NEXT: s_mov_b64 s[22:23], s[2:3]
; CHECK-NEXT: s_mov_b64 s[20:21], s[0:1]
; CHECK-NEXT: s_mov_b64 s[0:1], s[20:21]
; CHECK-NEXT: s_mov_b64 s[2:3], s[22:23]
; CHECK-NEXT: s_swappc_b64 s[30:31], s[16:17]
; CHECK-NEXT: v_readlane_b32 s31, v1, 1
; CHECK-NEXT: v_readlane_b32 s30, v1, 0
; CHECK-NEXT: s_mov_b32 s32, s33
; CHECK-NEXT: s_xor_saveexec_b64 s[4:5], -1
; CHECK-NEXT: buffer_load_dword v1, off, s[0:3], s33 ; 4-byte Folded Reload
; CHECK-NEXT: s_mov_b64 exec, s[4:5]
; CHECK-NEXT: s_mov_b32 s33, s18
; CHECK-NEXT: s_waitcnt vmcnt(0)
; CHECK-NEXT: s_setpc_b64 s[30:31]
entry:
%call = call i32 @tail_call()
ret i32 %call
}
define hidden i32 @caller_save_vgpr_spill_fp() #0 {
; CHECK-LABEL: caller_save_vgpr_spill_fp:
; CHECK: ; %bb.0: ; %entry
; CHECK-NEXT: s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0)
; CHECK-NEXT: s_mov_b32 s19, s33
; CHECK-NEXT: s_mov_b32 s33, s32
; CHECK-NEXT: s_xor_saveexec_b64 s[16:17], -1
; CHECK-NEXT: buffer_store_dword v2, off, s[0:3], s33 ; 4-byte Folded Spill
; CHECK-NEXT: s_mov_b64 exec, s[16:17]
; CHECK-NEXT: s_add_i32 s32, s32, 0x400
; CHECK-NEXT: v_writelane_b32 v2, s30, 0
; CHECK-NEXT: v_writelane_b32 v2, s31, 1
; CHECK-NEXT: s_getpc_b64 s[16:17]
; CHECK-NEXT: s_add_u32 s16, s16, caller_save_vgpr_spill_fp_tail_call@rel32@lo+4
; CHECK-NEXT: s_addc_u32 s17, s17, caller_save_vgpr_spill_fp_tail_call@rel32@hi+12
; CHECK-NEXT: s_mov_b64 s[22:23], s[2:3]
; CHECK-NEXT: s_mov_b64 s[20:21], s[0:1]
; CHECK-NEXT: s_mov_b64 s[0:1], s[20:21]
; CHECK-NEXT: s_mov_b64 s[2:3], s[22:23]
; CHECK-NEXT: s_swappc_b64 s[30:31], s[16:17]
; CHECK-NEXT: v_readlane_b32 s31, v2, 1
; CHECK-NEXT: v_readlane_b32 s30, v2, 0
; CHECK-NEXT: s_mov_b32 s32, s33
; CHECK-NEXT: s_xor_saveexec_b64 s[4:5], -1
; CHECK-NEXT: buffer_load_dword v2, off, s[0:3], s33 ; 4-byte Folded Reload
; CHECK-NEXT: s_mov_b64 exec, s[4:5]
; CHECK-NEXT: s_mov_b32 s33, s19
; CHECK-NEXT: s_waitcnt vmcnt(0)
; CHECK-NEXT: s_setpc_b64 s[30:31]
entry:
%call = call i32 @caller_save_vgpr_spill_fp_tail_call()
ret i32 %call
}
define protected amdgpu_kernel void @kernel() {
; CHECK-LABEL: kernel:
; CHECK: ; %bb.0: ; %entry
; CHECK-NEXT: s_mov_b32 s32, 0
; CHECK-NEXT: s_add_u32 flat_scratch_lo, s12, s17
; CHECK-NEXT: s_addc_u32 flat_scratch_hi, s13, 0
; CHECK-NEXT: s_add_u32 s0, s0, s17
; CHECK-NEXT: s_addc_u32 s1, s1, 0
; CHECK-NEXT: ; implicit-def: $vgpr3 : SGPR spill to VGPR lane
; CHECK-NEXT: v_writelane_b32 v3, s16, 0
; CHECK-NEXT: s_mov_b32 s13, s15
; CHECK-NEXT: s_mov_b32 s12, s14
; CHECK-NEXT: v_readlane_b32 s14, v3, 0
; CHECK-NEXT: s_getpc_b64 s[16:17]
; CHECK-NEXT: s_add_u32 s16, s16, caller_save_vgpr_spill_fp@rel32@lo+4
; CHECK-NEXT: s_addc_u32 s17, s17, caller_save_vgpr_spill_fp@rel32@hi+12
; CHECK-NEXT: s_mov_b64 s[22:23], s[2:3]
; CHECK-NEXT: s_mov_b64 s[20:21], s[0:1]
; CHECK-NEXT: s_mov_b32 s15, 20
; CHECK-NEXT: v_lshlrev_b32_e64 v2, s15, v2
; CHECK-NEXT: s_mov_b32 s15, 10
; CHECK-NEXT: v_lshlrev_b32_e64 v1, s15, v1
; CHECK-NEXT: v_or3_b32 v31, v0, v1, v2
; CHECK-NEXT: ; implicit-def: $sgpr15
; CHECK-NEXT: s_mov_b64 s[0:1], s[20:21]
; CHECK-NEXT: s_mov_b64 s[2:3], s[22:23]
; CHECK-NEXT: s_swappc_b64 s[30:31], s[16:17]
; CHECK-NEXT: s_endpgm
entry:
%call = call i32 @caller_save_vgpr_spill_fp()
ret void
}
attributes #0 = { "frame-pointer"="none" noinline }
attributes #1 = { "frame-pointer"="all" noinline }
|
