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
|
;=========================== begin_copyright_notice ============================
;
; Copyright (C) 2025 Intel Corporation
;
; SPDX-License-Identifier: MIT
;
;============================ end_copyright_notice =============================
; REQUIRES: regkeys
; RUN: igc_opt --opaque-pointers --regkey DisableCodeScheduling=0 --regkey EnableCodeSchedulingIfNoSpills=1 \
; RUN: --regkey PrintToConsole=1 --regkey DumpCodeScheduling=1 --igc-code-scheduling \
; RUN: --regkey CodeSchedulingRPThreshold=-512 \
; RUN: --regkey "CodeSchedulingConfig=10;1;0;30000;100;0;100000;1000;6000;200;10;20;500;50;0;1;1;0;1;1;8;1;32;1;200;64;0;1;20;200;200;5;16;16;34;200;1;0;128;256" \
; RUN: --regkey ForceOCLSIMDWidth=32 -S %s 2>&1 | FileCheck %s
; Checks that the register pressure is estimated correctly for the special cases related to vector shuffles.
define spir_kernel void @vector_shuffle_no_op(ptr addrspace(1) %A) {
; CHECK: Function vector_shuffle_no_op
; CHECK: Greedy MW attempt
; CHECK: {{([0-9]+,[ ]*[0-9]+[ ]*).*[ ]*}} [[BASE_ADDR:%.*]] = ptrtoint ptr addrspace(1) [[A:%.*]] to i64
; (6, 512 ) MW: Node #1, MW: 3000 %load2d = call <8 x i16> @llvm.genx.GenISA.LSC2DBlockRead.v8i16(i64 %base_addr, i32 127, i32 1023, i32 127, i32 0, i32 0, i32 16, i32 16, i32 16, i32 2, i1 false, i1 false, i32 4)
; adds 512 bytes
; CHECK: {{([0-9]+,[ ]*512[ ]*).*[ ]*}} [[LOAD2D:%.*]] = call <8 x i16> @llvm.genx.GenISA.LSC2DBlockRead.v8i16(i64 [[BASE_ADDR]], i32 127, i32 1023, i32 127, i32 0, i32 0, i32 16, i32 16, i32 16, i32 2, i1 false, i1 false, i32 4)
; the EE and IE instructions are marked as NOP and don't add regpressure
; (22, 0 ) Im: NOP Node #2, MW: 3000 %EE1.0 = extractelement <8 x i16> %load2d, i32 0
; CHECK: {{([0-9]+,[ ]*0[ ]*).*NOP.*[ ]*}} [[EE1_0:%.*]] = extractelement <8 x i16> [[LOAD2D]], i32 0
; CHECK: {{([0-9]+,[ ]*0[ ]*).*NOP.*[ ]*}} [[IE1_0:%.*]] = insertelement <4 x i16> undef, i16 [[EE1_0]], i32 0
; CHECK: {{([0-9]+,[ ]*0[ ]*).*NOP.*[ ]*}} [[EE1_1:%.*]] = extractelement <8 x i16> [[LOAD2D]], i32 1
; CHECK: {{([0-9]+,[ ]*0[ ]*).*NOP.*[ ]*}} [[IE1_1:%.*]] = insertelement <4 x i16> [[IE1_0]], i16 [[EE1_1]], i32 1
; CHECK: {{([0-9]+,[ ]*0[ ]*).*NOP.*[ ]*}} [[EE1_2:%.*]] = extractelement <8 x i16> [[LOAD2D]], i32 2
; CHECK: {{([0-9]+,[ ]*0[ ]*).*NOP.*[ ]*}} [[IE1_2:%.*]] = insertelement <4 x i16> [[IE1_1]], i16 [[EE1_2]], i32 2
; CHECK: {{([0-9]+,[ ]*0[ ]*).*NOP.*[ ]*}} [[EE1_3:%.*]] = extractelement <8 x i16> [[LOAD2D]], i32 3
; CHECK: {{([0-9]+,[ ]*0[ ]*).*NOP.*[ ]*}} [[IE1_3:%.*]] = insertelement <4 x i16> [[IE1_2]], i16 [[EE1_3]], i32 3
; CHECK: {{([0-9]+,[ ]*0[ ]*).*NOP.*[ ]*}} [[EE2_0:%.*]] = extractelement <8 x i16> [[LOAD2D]], i32 4
; CHECK: {{([0-9]+,[ ]*0[ ]*).*NOP.*[ ]*}} [[IE2_0:%.*]] = insertelement <4 x i16> undef, i16 [[EE2_0]], i32 0
; CHECK: {{([0-9]+,[ ]*0[ ]*).*NOP.*[ ]*}} [[EE2_1:%.*]] = extractelement <8 x i16> [[LOAD2D]], i32 5
; CHECK: {{([0-9]+,[ ]*0[ ]*).*NOP.*[ ]*}} [[IE2_1:%.*]] = insertelement <4 x i16> [[IE2_0]], i16 [[EE2_1]], i32 1
; CHECK: {{([0-9]+,[ ]*0[ ]*).*NOP.*[ ]*}} [[EE2_2:%.*]] = extractelement <8 x i16> [[LOAD2D]], i32 6
; CHECK: {{([0-9]+,[ ]*0[ ]*).*NOP.*[ ]*}} [[IE2_2:%.*]] = insertelement <4 x i16> [[IE2_1]], i16 [[EE2_2]], i32 2
; CHECK: {{([0-9]+,[ ]*0[ ]*).*NOP.*[ ]*}} [[EE2_3:%.*]] = extractelement <8 x i16> [[LOAD2D]], i32 7
; CHECK: {{([0-9]+,[ ]*0[ ]*).*NOP.*[ ]*}} [[IE2_3:%.*]] = insertelement <4 x i16> [[IE2_2]], i16 [[EE2_3]], i32 3
; (22, 512 ) MW: Node #34, MW: 0 %dpas1 = call <4 x float> @llvm.genx.GenISA.sub.group.dpas.v4f32.v4f32.v4i16.v4i32(<4 x float> zeroinitializer, <4 x i16> %IE1.7, <4 x i32> zeroinitializer, i32 1, i32 1, i32 1, i32 1, i1 false)
; (38, 0 ) MW: Node #35, MW: 0 %dpas2 = call <4 x float> @llvm.genx.GenISA.sub.group.dpas.v4f32.v4f32.v4i16.v4i32(<4 x float> zeroinitializer, <4 x i16> %IE2.7, <4 x i32> zeroinitializer, i32 1, i32 1, i32 1, i32 1, i1 false)
; first DPAS increases the regpressure by 512 bytes, the second one doesn't add any regpressure because the whole vector dies
; CHECK: {{([0-9]+,[ ]*512[ ]*).*[ ]*}} [[DPAS1:%.*]] = call <4 x float> @llvm.genx.GenISA.sub.group.dpas.v4f32.v4f32.v4i16.v4i32(<4 x float> zeroinitializer, <4 x i16> [[IE1_3]], <4 x i32> zeroinitializer, i32 1, i32 1, i32 1, i32 1, i1 false)
; CHECK: {{([0-9]+,[ ]*0[ ]*).*[ ]*}} [[DPAS2:%.*]] = call <4 x float> @llvm.genx.GenISA.sub.group.dpas.v4f32.v4f32.v4i16.v4i32(<4 x float> zeroinitializer, <4 x i16> [[IE2_3]], <4 x i32> zeroinitializer, i32 1, i32 1, i32 1, i32 1, i1 false)
; CHECK: ret void
;
entry:
%base_addr = ptrtoint ptr addrspace(1) %A to i64
%load2d = call <8 x i16> @llvm.genx.GenISA.LSC2DBlockRead.v8i16(i64 %base_addr, i32 127, i32 1023, i32 127, i32 0, i32 0, i32 16, i32 16, i32 16, i32 2, i1 false, i1 false, i32 4)
%EE1.0 = extractelement <8 x i16> %load2d, i32 0
%IE1.0 = insertelement <4 x i16> undef, i16 %EE1.0, i32 0
%EE1.1 = extractelement <8 x i16> %load2d, i32 1
%IE1.1 = insertelement <4 x i16> %IE1.0, i16 %EE1.1, i32 1
%EE1.2 = extractelement <8 x i16> %load2d, i32 2
%IE1.2 = insertelement <4 x i16> %IE1.1, i16 %EE1.2, i32 2
%EE1.3 = extractelement <8 x i16> %load2d, i32 3
%IE1.3 = insertelement <4 x i16> %IE1.2, i16 %EE1.3, i32 3
%EE2.0 = extractelement <8 x i16> %load2d, i32 4
%IE2.0 = insertelement <4 x i16> undef, i16 %EE2.0, i32 0
%EE2.1 = extractelement <8 x i16> %load2d, i32 5
%IE2.1 = insertelement <4 x i16> %IE2.0, i16 %EE2.1, i32 1
%EE2.2 = extractelement <8 x i16> %load2d, i32 6
%IE2.2 = insertelement <4 x i16> %IE2.1, i16 %EE2.2, i32 2
%EE2.3 = extractelement <8 x i16> %load2d, i32 7
%IE2.3 = insertelement <4 x i16> %IE2.2, i16 %EE2.3, i32 3
%dpas1 = call <4 x float> @llvm.genx.GenISA.sub.group.dpas.v4f32.v4f32.v4i16.v4i32(<4 x float> zeroinitializer, <4 x i16> %IE1.3, <4 x i32> zeroinitializer, i32 1, i32 1, i32 1, i32 1, i1 false)
%dpas2 = call <4 x float> @llvm.genx.GenISA.sub.group.dpas.v4f32.v4f32.v4i16.v4i32(<4 x float> zeroinitializer, <4 x i16> %IE2.3, <4 x i32> zeroinitializer, i32 1, i32 1, i32 1, i32 1, i1 false)
ret void
}
define spir_kernel void @vector_shuffle(ptr addrspace(1) %A) {
; CHECK: Function vector_shuffle
; CHECK: Greedy MW attempt
; CHECK: {{([0-9]+,[ ]*[0-9]+[ ]*).*[ ]*}} [[BASE_ADDR:%.*]] = ptrtoint ptr addrspace(1) [[A:%.*]] to i64
; (6, 512 ) MW: Node #1, MW: 3000 %load2d = call <8 x i16> @llvm.genx.GenISA.LSC2DBlockRead.v8i16(i64 %base_addr, i32 127, i32 1023, i32 127, i32 0, i32 0, i32 16, i32 16, i32 16, i32 2, i1 false, i1 false, i32 4)
; adds 512 bytes, but we also estimate the regpressure burst from the shuffles, so use 2x (1024 bytes) when making the decision
; CHECK: {{([0-9]+,[ ]*1024[ ]*).*[ ]*}} [[LOAD2D:%.*]] = call <8 x i16> @llvm.genx.GenISA.LSC2DBlockRead.v8i16(i64 [[BASE_ADDR]], i32 127, i32 1023, i32 127, i32 0, i32 0, i32 16, i32 16, i32 16, i32 2, i1 false, i1 false, i32 4)
; the EE and IE instructions are marked as VS. IEs add regpressure. The last IE kills the original vector
; CHECK: {{([0-9]+,[ ]*0[ ]*).*VS.*[ ]*}} [[EE1_0:%.*]] = extractelement <8 x i16> [[LOAD2D]], i32 0
; CHECK: {{([0-9]+,[ ]*256[ ]*).*VS.*[ ]*}} [[IE1_0:%.*]] = insertelement <4 x i16> undef, i16 [[EE1_0]], i32 0
; CHECK: {{([0-9]+,[ ]*0[ ]*).*VS.*[ ]*}} [[EE1_1:%.*]] = extractelement <8 x i16> [[LOAD2D]], i32 2
; CHECK: {{([0-9]+,[ ]*0[ ]*).*VS.*[ ]*}} [[IE1_1:%.*]] = insertelement <4 x i16> [[IE1_0]], i16 [[EE1_1]], i32 1
; CHECK: {{([0-9]+,[ ]*0[ ]*).*VS.*[ ]*}} [[EE1_2:%.*]] = extractelement <8 x i16> [[LOAD2D]], i32 4
; CHECK: {{([0-9]+,[ ]*0[ ]*).*VS.*[ ]*}} [[IE1_2:%.*]] = insertelement <4 x i16> [[IE1_1]], i16 [[EE1_2]], i32 2
; CHECK: {{([0-9]+,[ ]*0[ ]*).*VS.*[ ]*}} [[EE1_3:%.*]] = extractelement <8 x i16> [[LOAD2D]], i32 6
; CHECK: {{([0-9]+,[ ]*0[ ]*).*VS.*[ ]*}} [[IE1_3:%.*]] = insertelement <4 x i16> [[IE1_2]], i16 [[EE1_3]], i32 3
; CHECK: {{([0-9]+,[ ]*0[ ]*).*VS.*[ ]*}} [[EE2_0:%.*]] = extractelement <8 x i16> [[LOAD2D]], i32 1
; CHECK: {{([0-9]+,[ ]*256[ ]*).*VS.*[ ]*}} [[IE2_0:%.*]] = insertelement <4 x i16> undef, i16 [[EE2_0]], i32 0
; CHECK: {{([0-9]+,[ ]*0[ ]*).*VS.*[ ]*}} [[EE2_1:%.*]] = extractelement <8 x i16> [[LOAD2D]], i32 3
; CHECK: {{([0-9]+,[ ]*0[ ]*).*VS.*[ ]*}} [[IE2_1:%.*]] = insertelement <4 x i16> [[IE2_0]], i16 [[EE2_1]], i32 1
; CHECK: {{([0-9]+,[ ]*0[ ]*).*VS.*[ ]*}} [[EE2_2:%.*]] = extractelement <8 x i16> [[LOAD2D]], i32 5
; CHECK: {{([0-9]+,[ ]*0[ ]*).*VS.*[ ]*}} [[IE2_2:%.*]] = insertelement <4 x i16> [[IE2_1]], i16 [[EE2_2]], i32 2
; CHECK: {{([0-9]+,[ ]*0[ ]*).*VS.*[ ]*}} [[EE2_3:%.*]] = extractelement <8 x i16> [[LOAD2D]], i32 7
; CHECK: {{([0-9]+,[ ]*-512[ ]*).*VS.*[ ]*}} [[IE2_3:%.*]] = insertelement <4 x i16> [[IE2_2]], i16 [[EE2_3]], i32 3
; both DPAS increase the regpressure by 256 (sub vector of 256 dies, 512 created)
; CHECK: {{([0-9]+,[ ]*256[ ]*).*[ ]*}} [[DPAS1:%.*]] = call <4 x float> @llvm.genx.GenISA.sub.group.dpas.v4f32.v4f32.v4i16.v4i32(<4 x float> zeroinitializer, <4 x i16> [[IE1_3]], <4 x i32> zeroinitializer, i32 1, i32 1, i32 1, i32 1, i1 false)
; CHECK: {{([0-9]+,[ ]*256[ ]*).*[ ]*}} [[DPAS2:%.*]] = call <4 x float> @llvm.genx.GenISA.sub.group.dpas.v4f32.v4f32.v4i16.v4i32(<4 x float> zeroinitializer, <4 x i16> [[IE2_3]], <4 x i32> zeroinitializer, i32 1, i32 1, i32 1, i32 1, i1 false)
; CHECK: ret void
;
entry:
%base_addr = ptrtoint ptr addrspace(1) %A to i64
%load2d = call <8 x i16> @llvm.genx.GenISA.LSC2DBlockRead.v8i16(i64 %base_addr, i32 127, i32 1023, i32 127, i32 0, i32 0, i32 16, i32 16, i32 16, i32 2, i1 false, i1 false, i32 4)
%EE1.0 = extractelement <8 x i16> %load2d, i32 0
%IE1.0 = insertelement <4 x i16> undef, i16 %EE1.0, i32 0
%EE1.1 = extractelement <8 x i16> %load2d, i32 2
%IE1.1 = insertelement <4 x i16> %IE1.0, i16 %EE1.1, i32 1
%EE1.2 = extractelement <8 x i16> %load2d, i32 4
%IE1.2 = insertelement <4 x i16> %IE1.1, i16 %EE1.2, i32 2
%EE1.3 = extractelement <8 x i16> %load2d, i32 6
%IE1.3 = insertelement <4 x i16> %IE1.2, i16 %EE1.3, i32 3
%EE2.0 = extractelement <8 x i16> %load2d, i32 1
%IE2.0 = insertelement <4 x i16> undef, i16 %EE2.0, i32 0
%EE2.1 = extractelement <8 x i16> %load2d, i32 3
%IE2.1 = insertelement <4 x i16> %IE2.0, i16 %EE2.1, i32 1
%EE2.2 = extractelement <8 x i16> %load2d, i32 5
%IE2.2 = insertelement <4 x i16> %IE2.1, i16 %EE2.2, i32 2
%EE2.3 = extractelement <8 x i16> %load2d, i32 7
%IE2.3 = insertelement <4 x i16> %IE2.2, i16 %EE2.3, i32 3
%dpas1 = call <4 x float> @llvm.genx.GenISA.sub.group.dpas.v4f32.v4f32.v4i16.v4i32(<4 x float> zeroinitializer, <4 x i16> %IE1.3, <4 x i32> zeroinitializer, i32 1, i32 1, i32 1, i32 1, i1 false)
%dpas2 = call <4 x float> @llvm.genx.GenISA.sub.group.dpas.v4f32.v4f32.v4i16.v4i32(<4 x float> zeroinitializer, <4 x i16> %IE2.3, <4 x i32> zeroinitializer, i32 1, i32 1, i32 1, i32 1, i1 false)
ret void
}
define spir_kernel void @coalesced_scalars(ptr addrspace(1) %0) {
; CHECK: Function coalesced_scalars
; CHECK: Greedy MW attempt
; the IE instructions are marked as SCA. First IE adds regpressure
; then the last usage of the scalar (fadd) kills the hanging values
; CHECK: {{([0-9]+,[ ]*[0-9-]+[ ]*).*[ ]*}} [[TMP17:%.*]] = fmul fast float [[TMP9:%.*]], [[TMP1:%.*]]
; CHECK: {{([0-9]+,[ ]*512[ ]*).*SCA.*[ ]*}} [[TMP19:%.*]] = insertelement <4 x float> zeroinitializer, float [[TMP17]], i64 0
; CHECK: {{([0-9]+,[ ]*[0-9-]+[ ]*).*[ ]*}} [[TMP18:%.*]] = fmul fast float [[TMP10:%.*]], [[TMP2:%.*]]
; CHECK: {{([0-9]+,[ ]*[0-9-]+[ ]*).*SCA.*[ ]*}} [[TMP21:%.*]] = insertelement <4 x float> [[TMP19]], float [[TMP18]], i64 1
; CHECK: {{([0-9]+,[ ]*[0-9-]+[ ]*).*[ ]*}} [[TMP20:%.*]] = fmul fast float [[TMP11:%.*]], [[TMP3:%.*]]
; CHECK: {{([0-9]+,[ ]*0[ ]*).*SCA.*[ ]*}} [[TMP23:%.*]] = insertelement <4 x float> [[TMP21]], float [[TMP20]], i64 2
; CHECK: {{([0-9]+,[ ]*[0-9-]+[ ]*).*[ ]*}} [[TMP22:%.*]] = fmul fast float [[TMP12:%.*]], [[TMP4:%.*]]
; CHECK: {{([0-9]+,[ ]*0[ ]*).*SCA.*[ ]*}} [[TMP25:%.*]] = insertelement <4 x float> [[TMP23]], float [[TMP22]], i64 3
; dpas don't add any regpressure, they reuse the registers of the created vector
; CHECK: {{([0-9]+,[ ]*0[ ]*).*[ ]*}} [[TMP33:%.*]] = call <4 x float> @llvm.genx.GenISA.sub.group.dpas.v4f32.v4f32.v4i16.v4i32(<4 x float> [[TMP25]], <4 x i16> zeroinitializer, <4 x i32> zeroinitializer, i32 0, i32 0, i32 0, i32 0, i1 false)
; CHECK: {{([0-9]+,[ ]*0[ ]*).*[ ]*}} [[TMP34:%.*]] = call <4 x float> @llvm.genx.GenISA.sub.group.dpas.v4f32.v4f32.v4i16.v4i32(<4 x float> [[TMP33]], <4 x i16> zeroinitializer, <4 x i32> zeroinitializer, i32 0, i32 0, i32 0, i32 0, i1 false)
; extract the values from the vector. The extractelement instructions are marked as V2S and don't increase regpressure
; the vector hangs, so the last EE doesn't reduce regpressure. In this case the EEs are used in PHI nodes so these hanging vals are not killed
; CHECK: {{([0-9]+,[ ]*0[ ]*).*V2S.*[ ]*}} [[TMP35:%.*]] = extractelement <4 x float> [[TMP34]], i64 0
; CHECK: {{([0-9]+,[ ]*0[ ]*).*V2S.*[ ]*}} [[TMP36:%.*]] = extractelement <4 x float> [[TMP34]], i64 1
; CHECK: {{([0-9]+,[ ]*0[ ]*).*V2S.*[ ]*}} [[TMP37:%.*]] = extractelement <4 x float> [[TMP34]], i64 2
; CHECK: {{([0-9]+,[ ]*0[ ]*).*V2S.*[ ]*}} [[TMP38:%.*]] = extractelement <4 x float> [[TMP34]], i64 3
; fadd kills the hanging values from SCA and creates one float -> -384 (3 * 32 * 4)
; CHECK: {{([0-9]+,[ ]*-384[ ]*).*[ ]*}} [[TMP50:%.*]] = fadd fast float [[TMP37:%.*]], [[TMP22]]
; this fadd doesn't kill the hanging EEs (they are used in PHI nodes), so it increases regpressure by 128 (32 * 4)
; CHECK: {{([0-9]+,[ ]*128[ ]*).*[ ]*}} [[TMP51:%.*]] = fadd fast float [[TMP38:%.*]], 4.000000e+00
entry:
br label %._crit_edge
._crit_edge: ; preds = %._crit_edge, %0
%1 = phi float [ 0.000000e+00, %entry ], [ %19, %._crit_edge ]
%2 = phi float [ 0.000000e+00, %entry ], [ %20, %._crit_edge ]
%3 = phi float [ 0.000000e+00, %entry ], [ %21, %._crit_edge ]
%4 = phi float [ 0.000000e+00, %entry ], [ %22, %._crit_edge ]
%5 = call float @llvm.exp2.f32(float 0.000000e+00)
%6 = call float @llvm.exp2.f32(float 0.000000e+00)
%7 = call float @llvm.exp2.f32(float 0.000000e+00)
%8 = call float @llvm.exp2.f32(float 0.000000e+00)
%9 = fmul fast float %5, %1
%10 = fmul fast float %6, %2
%11 = fmul fast float %7, %3
%12 = fmul fast float %8, %4
%13 = insertelement <4 x float> zeroinitializer, float %9, i64 0
%14 = insertelement <4 x float> %13, float %10, i64 1
%15 = insertelement <4 x float> %14, float %11, i64 2
%16 = insertelement <4 x float> %15, float %12, i64 3
%17 = call <4 x float> @llvm.genx.GenISA.sub.group.dpas.v4f32.v4f32.v4i16.v4i32(<4 x float> %16, <4 x i16> zeroinitializer, <4 x i32> zeroinitializer, i32 0, i32 0, i32 0, i32 0, i1 false)
%18 = call <4 x float> @llvm.genx.GenISA.sub.group.dpas.v4f32.v4f32.v4i16.v4i32(<4 x float> %17, <4 x i16> zeroinitializer, <4 x i32> zeroinitializer, i32 0, i32 0, i32 0, i32 0, i1 false)
%19 = extractelement <4 x float> %18, i64 0
%20 = extractelement <4 x float> %18, i64 1
%21 = extractelement <4 x float> %18, i64 2
%22 = extractelement <4 x float> %18, i64 3
%23 = fadd fast float %21, %12
%24 = fadd fast float %22, 4.000000e+00
br label %._crit_edge
}
define spir_kernel void @vector_to_scalars_pattern(ptr addrspace(1) %A) {
; CHECK: Function vector_to_scalars_pattern
; CHECK: Greedy MW attempt
; DPAS increases regpressure
; CHECK: {{([0-9]+,[ ]*512[ ]*).*[ ]*}} [[DPAS:%.*]] = call <4 x float> @llvm.genx.GenISA.sub.group.dpas.v4f32.v4f32.v4i16.v4i32(<4 x float> zeroinitializer, <4 x i16> undef, <4 x i32> zeroinitializer, i32 1, i32 1, i32 1, i32 1, i1 false)
; EE don't increase regpressure
; CHECK: {{([0-9]+,[ ]*0[ ]*).*V2S.*[ ]*}} [[EE0:%.*]] = extractelement <4 x float> [[DPAS]], i64 0
; CHECK: {{([0-9]+,[ ]*0[ ]*).*V2S.*[ ]*}} [[EE1:%.*]] = extractelement <4 x float> [[DPAS]], i64 1
; CHECK: {{([0-9]+,[ ]*0[ ]*).*V2S.*[ ]*}} [[EE2:%.*]] = extractelement <4 x float> [[DPAS]], i64 2
; CHECK: {{([0-9]+,[ ]*0[ ]*).*V2S.*[ ]*}} [[EE3:%.*]] = extractelement <4 x float> [[DPAS]], i64 3
; The vector doesn't die on the last EE, it hangs.
; The uses of the EEs increase regpressure
; CHECK: {{([0-9]+,[ ]*128[ ]*).*[ ]*}} [[USE0:%.*]] = fadd fast float [[EE0]], 1.0
; CHECK: {{([0-9]+,[ ]*128[ ]*).*[ ]*}} [[USE1:%.*]] = fadd fast float [[EE1]], 2.0
; CHECK: {{([0-9]+,[ ]*128[ ]*).*[ ]*}} [[USE2:%.*]] = fadd fast float [[EE2]], 3.0
; The vector dies on the last EE usage
; CHECK: {{([0-9]+,[ ]*-384[ ]*).*[ ]*}} [[USE3:%.*]] = fadd fast float [[EE3]], 4.0
; CHECK: ret void
entry:
%dpas = call <4 x float> @llvm.genx.GenISA.sub.group.dpas.v4f32.v4f32.v4i16.v4i32(<4 x float> zeroinitializer, <4 x i16> undef, <4 x i32> zeroinitializer, i32 1, i32 1, i32 1, i32 1, i1 false)
%ee0 = extractelement <4 x float> %dpas, i64 0
%ee1 = extractelement <4 x float> %dpas, i64 1
%ee2 = extractelement <4 x float> %dpas, i64 2
%ee3 = extractelement <4 x float> %dpas, i64 3
%use0 = fadd fast float %ee0, 1.0
%use1 = fadd fast float %ee1, 2.0
%use2 = fadd fast float %ee2, 3.0
%use3 = fadd fast float %ee3, 4.0
ret void
}
declare <4 x float> @llvm.genx.GenISA.sub.group.dpas.v4f32.v4f32.v4i16.v4i32(
<4 x float>, <4 x i16>, <4 x i32>, i32, i32, i32, i32, i1) #1
declare float @llvm.exp2.f32(float) #2
declare <8 x i16> @llvm.genx.GenISA.LSC2DBlockRead.v8i16(i64, i32, i32, i32, i32, i32, i32, i32, i32, i32, i1, i1, i32) #1
attributes #0 = { convergent nounwind }
attributes #1 = { convergent nounwind readnone willreturn }
attributes #2 = { nofree nosync nounwind readnone speculatable willreturn }
|
