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
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
|
// RUN: mlir-opt %s -split-input-file -affine-data-copy-generate="generate-dma=false fast-mem-space=0 skip-non-unit-stride-loops" | FileCheck %s
// Small buffer size to trigger fine copies.
// RUN: mlir-opt %s -split-input-file -affine-data-copy-generate="generate-dma=false fast-mem-space=0 fast-mem-capacity=1" | FileCheck --check-prefix=CHECK-SMALL %s
// Test affine data copy with a memref filter. We use a test pass that invokes
// affine data copy utility on the input loop nest.
// '-test-affine-data-copy-memref-filter' passes the first memref found in an
// affine.load op in the innermost loop as a filter.
// RUN: mlir-opt %s -split-input-file -test-affine-data-copy='memref-filter' | FileCheck %s --check-prefix=FILTER
// RUN: mlir-opt %s -split-input-file -test-affine-data-copy='for-memref-region' | FileCheck %s --check-prefix=MEMREF_REGION
// -copy-skip-non-stride-loops forces the copies to be placed right inside the
// tile space loops, avoiding the sensitivity of copy placement depth to memory
// footprint -- so that one could write a definite test case and not have to
// update it each time something related to the cost functions change.
#id = affine_map<(d0) -> (d0)>
#ub = affine_map<(d0) -> (d0 + 128)>
// Map used to index the buffer while computing.
// CHECK-DAG: [[$MAP_IDENTITY:map[0-9a-zA-Z_]*]] = affine_map<(d0) -> (d0)>
// CHECK-DAG: [[$MAP_PLUS_128:map[0-9a-zA-Z_]*]] = affine_map<(d0) -> (d0 + 128)>
// CHECK-LABEL: func @matmul
// FILTER-LABEL: func @matmul
func.func @matmul(%A: memref<4096x4096xf32>, %B: memref<4096x4096xf32>, %C: memref<4096x4096xf32>) -> memref<4096x4096xf32> {
affine.for %i = 0 to 4096 step 128 {
affine.for %j = 0 to 4096 step 128 {
affine.for %k = 0 to 4096 step 128 {
affine.for %ii = #id(%i) to #ub(%i) {
affine.for %jj = #id(%j) to #ub(%j) {
affine.for %kk = #id(%k) to #ub(%k) {
%5 = affine.load %A[%ii, %kk] : memref<4096x4096xf32>
%6 = affine.load %B[%kk, %jj] : memref<4096x4096xf32>
%7 = affine.load %C[%ii, %jj] : memref<4096x4096xf32>
%8 = arith.mulf %5, %6 : f32
%9 = arith.addf %7, %8 : f32
affine.store %9, %C[%ii, %jj] : memref<4096x4096xf32>
}
}
}
}
}
}
return %C : memref<4096x4096xf32>
}
// Buffers of size 128x128 get created here for all three matrices.
// CHECK: affine.for %[[I:.*]] = 0 to 4096 step 128 {
// CHECK: affine.for %[[J:.*]] = 0 to 4096 step 128 {
// CHECK: [[BUFC:%[0-9a-zA-Z_]+]] = memref.alloc() : memref<128x128xf32>
// The result matrix's copy gets hoisted out.
// Result matrix copy-in.
// CHECK: affine.for %[[II:.*]] = #[[$MAP_IDENTITY]](%{{.*}}) to #[[$MAP_PLUS_128]](%{{.*}}) {
// CHECK: affine.for %[[JJ:.*]] = #[[$MAP_IDENTITY]](%{{.*}}) to #[[$MAP_PLUS_128]](%{{.*}}) {
// CHECK: affine.load %{{.*}}[%{{.*}}, %{{.*}}] : memref<4096x4096xf32>
// CHECK: affine.store %{{.*}}, [[BUFC]][%[[II]] - %[[I]], %[[JJ]] - %[[J]]] : memref<128x128xf32>
// CHECK: }
// CHECK: }
// LHS matrix copy-in.
// CHECK: affine.for %[[K:.*]] = 0 to 4096 step 128 {
// CHECK: [[BUFA:%[0-9a-zA-Z_]+]] = memref.alloc() : memref<128x128xf32>
// CHECK: affine.for %[[II:.*]] = #[[$MAP_IDENTITY]](%{{.*}}) to #[[$MAP_PLUS_128]](%{{.*}}) {
// CHECK: affine.for %[[KK:.*]] = #[[$MAP_IDENTITY]](%{{.*}}) to #[[$MAP_PLUS_128]](%{{.*}}) {
// CHECK: affine.load %{{.*}}[%{{.*}}, %{{.*}}] : memref<4096x4096xf32>
// CHECK: affine.store %{{.*}}, [[BUFA]][%[[II]] - %[[I]], %[[KK]] - %[[K]]] : memref<128x128xf32>
// CHECK: }
// CHECK: }
// RHS matrix copy-in.
// CHECK: [[BUFB:%[0-9a-zA-Z_]+]] = memref.alloc() : memref<128x128xf32>
// CHECK: affine.for %[[KK:.*]] = #[[$MAP_IDENTITY]](%{{.*}}) to #[[$MAP_PLUS_128]](%{{.*}}) {
// CHECK: affine.for %[[JJ:.*]] = #[[$MAP_IDENTITY]](%{{.*}}) to #[[$MAP_PLUS_128]](%{{.*}}) {
// CHECK: affine.load %{{.*}}[%{{.*}}, %{{.*}}] : memref<4096x4096xf32>
// CHECK: affine.store %{{.*}}, [[BUFB]][%[[KK]] - %[[K]], %[[JJ]] - %[[J]]] : memref<128x128xf32>
// CHECK: }
// CHECK: }
// Computation on the fast buffers.
// CHECK: affine.for %{{.*}} = #[[$MAP_IDENTITY]](%{{.*}}) to #[[$MAP_PLUS_128]](%{{.*}}) {
// CHECK: affine.for %{{.*}} = #[[$MAP_IDENTITY]](%{{.*}}) to #[[$MAP_PLUS_128]](%{{.*}}) {
// CHECK: affine.for %{{.*}} = #[[$MAP_IDENTITY]](%{{.*}}) to #[[$MAP_PLUS_128]](%{{.*}}) {
// CHECK: affine.load [[BUFA]][-%{{.*}} + %{{.*}}, -%{{.*}} + %{{.*}}] : memref<128x128xf32>
// CHECK: affine.load [[BUFB]][-%{{.*}} + %{{.*}}, -%{{.*}} + %{{.*}}] : memref<128x128xf32>
// CHECK: affine.load [[BUFC]][-%{{.*}} + %{{.*}}, -%{{.*}} + %{{.*}}] : memref<128x128xf32>
// CHECK: arith.mulf %{{.*}}, %{{.*}} : f32
// CHECK: arith.addf %{{.*}}, %{{.*}} : f32
// CHECK: affine.store %{{.*}}, [[BUFC]][-%{{.*}} + %{{.*}}, -%{{.*}} + %{{.*}}] : memref<128x128xf32>
// CHECK: }
// CHECK: }
// CHECK: }
// CHECK: memref.dealloc [[BUFB]] : memref<128x128xf32>
// CHECK: memref.dealloc [[BUFA]] : memref<128x128xf32>
// CHECK: }
// Result matrix copy out.
// CHECK: affine.for %{{.*}} = #[[$MAP_IDENTITY]](%{{.*}}) to #[[$MAP_PLUS_128]](%{{.*}}) {
// CHECK: affine.for %{{.*}} = #[[$MAP_IDENTITY]](%{{.*}}) to #[[$MAP_PLUS_128]](%{{.*}}) {
// CHECK: affine.load [[BUFC]][%{{.*}} - %{{.*}}, %{{.*}} - %{{.*}}] : memref<128x128xf32>
// CHECK: store %{{.*}}, %{{.*}}[%{{.*}}, %{{.*}}] : memref<4096x4096xf32>
// CHECK: }
// CHECK: }
// CHECK: memref.dealloc [[BUFC]] : memref<128x128xf32>
// CHECK: }
// CHECK: }
// Check that only one memref is copied when memref filter is used.
// FILTER: affine.for %{{.*}} = 0 to 4096 step 128 {
// FILTER: memref.alloc() : memref<128x4096xf32>
// FILTER-NOT: memref.alloc()
// FILTER: affine.for
// FILTER: affine.for %{{.*}} = 0 to 4096 {
// FILTER: affine.for %{{.*}} = 0 to 4096 step 128 {
// FILTER-NEXT: affine.for %{{.*}} = 0 to 4096 step 128 {
// FILTER-NEXT: affine.for %{{.*}} = #map{{.*}}(%{{.*}}) to #map{{.*}}(%{{.*}}) {
// FILTER-NEXT: affine.for %{{.*}} = #map{{.*}}(%{{.*}}) to #map{{.*}}(%{{.*}}) {
// FILTER-NEXT: affine.for %{{.*}} = #map{{.*}}(%{{.*}}) to #map{{.*}}(%{{.*}}) {
// FILTER: memref.dealloc %{{.*}} : memref<128x4096xf32>
// FILTER-NOT: memref.dealloc %{{.*}} : memref<128x4096xf32>
// -----
//
// This test case will lead to single element buffers. These are eventually
// expected to be turned into registers via alloca and mem2reg.
//
// CHECK-SMALL-LABEL: func @single_elt_buffers
// FILTER-LABEL: func @single_elt_buffers
// MEMREF_REGION-LABEL: func @single_elt_buffers
func.func @single_elt_buffers(%arg0: memref<1024x1024xf32>, %arg1: memref<1024x1024xf32>, %arg2: memref<1024x1024xf32>) -> memref<1024x1024xf32> {
affine.for %i = 0 to 1024 {
affine.for %j = 0 to 1024 {
affine.for %k = 0 to 1024 {
%6 = affine.load %arg1[%k, %j] : memref<1024x1024xf32>
%7 = affine.load %arg2[%i, %j] : memref<1024x1024xf32>
%9 = arith.addf %6, %7 : f32
affine.store %9, %arg2[%i, %j] : memref<1024x1024xf32>
}
}
}
return %arg2 : memref<1024x1024xf32>
}
// CHECK-SMALL: affine.for %arg{{.*}} = 0 to 1024 {
// CHECK-SMALL: affine.for %arg{{.*}} = 0 to 1024 {
// CHECK-SMALL: memref.alloc() : memref<1x1xf32>
// CHECK-SMALL: affine.load %arg{{.*}}[%{{.*}}, %{{.*}}] : memref<1024x1024xf32>
// CHECK-SMALL: affine.store %{{.*}}, %{{.*}}[0, 0] : memref<1x1xf32>
// CHECK-SMALL: affine.for %arg{{.*}} = 0 to 1024 {
// CHECK-SMALL: memref.alloc() : memref<1x1xf32>
// CHECK-SMALL: affine.load %arg{{.*}}[%{{.*}}, %{{.*}}] : memref<1024x1024xf32>
// CHECK-SMALL: affine.store %{{.*}}, %{{.*}}[0, 0] : memref<1x1xf32>
// CHECK-SMALL: affine.load %{{.*}}[0, 0] : memref<1x1xf32>
// CHECK-SMALL: affine.load %{{.*}}[0, 0] : memref<1x1xf32>
// CHECK-SMALL: arith.addf %{{.*}}, %{{.*}} : f32
// CHECK-SMALL: affine.store %{{.*}}, %{{.*}}[0, 0] : memref<1x1xf32>
// CHECK-SMALL: memref.dealloc %{{.*}} : memref<1x1xf32>
// CHECK-SMALL: }
// CHECK-SMALL: affine.load %{{.*}}[0, 0] : memref<1x1xf32>
// CHECK-SMALL: affine.store %{{.*}}, %arg{{.*}}[%{{.*}}, %{{.*}}] : memref<1024x1024xf32>
// CHECK-SMALL: memref.dealloc %{{.*}} : memref<1x1xf32>
// CHECK-SMALL: }
// CHECK-SMALL: }
// CHECK-SMALL: return
// Check that only one memref is copied when memref filter is used.
// FILTER: memref.alloc() : memref<1024x1024xf32>
// FILTER-NOT: memref.alloc()
// FILTER: affine.for %{{.*}} = 0 to 1024 {
// FILTER: affine.for %{{.*}} = 0 to 1024 {
// FILTER: affine.for %{{.*}} = 0 to 1024 {
// FILTER-NEXT: affine.for %{{.*}} = 0 to 1024 {
// FILTER-NEXT: affine.for %{{.*}} = 0 to 1024 {
// FILTER: memref.dealloc %{{.*}} : memref<1024x1024xf32>
// FILTER-NOT: memref.dealloc
// FILTER: return
// CHeck that only one memref is copied, because for-memref-region is enabled
// (and the first ever encountered load is analyzed).
// MEMREF_REGION: memref.alloc() : memref<1024x1024xf32>
// MEMREF_REGION-NOT: memref.alloc()
// MEMREF_REGION: affine.for %{{.*}} = 0 to 1024 {
// MEMREF_REGION: affine.for %{{.*}} = 0 to 1024 {
// MEMREF_REGION: }
// MEMREF_REGION: }
// MEMREF_REGION-NEXT: affine.for %{{.*}} = 0 to 1024 {
// MEMREF_REGION-NEXT: affine.for %{{.*}} = 0 to 1024 {
// MEMREF_REGION-NEXT: affine.for %{{.*}} = 0 to 1024 {
// MEMREF_REGION: memref.dealloc %{{.*}} : memref<1024x1024xf32>
// MEMREF_REGION-NOT: memref.dealloc
// MEMREF_REGION-NEXT: return
// -----
// This pattern typically appears with tiling with tile sizes that don't divide
// the loop trip counts.
#map_ub = affine_map<(d0) -> (4096, d0 + 100)>
// CHECK-DAG: [[$MAP_IDENTITY:map[0-9a-zA-Z_]*]] = affine_map<(d0) -> (d0)>
// CHECK-DAG: [[$MAP_MIN_UB1:map[0-9a-zA-Z_]*]] = affine_map<(d0) -> (d0 + 100, 4096)>
// CHECK-DAG: [[$MAP_MIN_UB2:map[0-9a-zA-Z_]*]] = affine_map<(d0) -> (4096, d0 + 100)>
// CHECK-LABEL: func @min_upper_bound
func.func @min_upper_bound(%A: memref<4096xf32>) -> memref<4096xf32> {
affine.for %i = 0 to 4096 step 100 {
affine.for %ii = affine_map<(d0) -> (d0)>(%i) to min #map_ub(%i) {
%5 = affine.load %A[%ii] : memref<4096xf32>
%6 = arith.mulf %5, %5 : f32
affine.store %6, %A[%ii] : memref<4096xf32>
}
}
return %A : memref<4096xf32>
}
// CHECK: affine.for %[[IV1:.*]] = 0 to 4096 step 100
// CHECK: %[[BUF:.*]] = memref.alloc() : memref<100xf32>
// CHECK-NEXT: affine.for %[[IV2:.*]] = #[[$MAP_IDENTITY]](%[[IV1]]) to min #[[$MAP_MIN_UB1]](%[[IV1]]) {
// CHECK-NEXT: affine.load %{{.*}}[%[[IV2]]] : memref<4096xf32>
// CHECK-NEXT: affine.store %{{.*}}, %[[BUF]][%[[IV2]] - %[[IV1]]] : memref<100xf32>
// CHECK-NEXT: }
// CHECK-NEXT: affine.for %[[IV2:.*]] = #[[$MAP_IDENTITY]](%[[IV1]]) to min #[[$MAP_MIN_UB2]](%[[IV1]]) {
// CHECK-NEXT: affine.load %[[BUF]][-%[[IV1]] + %[[IV2]]] : memref<100xf32>
// CHECK-NEXT: arith.mulf
// CHECK-NEXT: affine.store %{{.*}}, %[[BUF]][-%[[IV1]] + %[[IV2]]] : memref<100xf32>
// CHECK-NEXT: }
// CHECK: affine.for %[[IV2:.*]] = #[[$MAP_IDENTITY]](%[[IV1]]) to min #[[$MAP_MIN_UB1]](%[[IV1]]) {
// CHECK-NEXT: affine.load %[[BUF]][%[[IV2]] - %[[IV1]]] : memref<100xf32>
// CHECK-NEXT: affine.store %{{.*}}, %{{.*}}[%[[IV2]]] : memref<4096xf32>
// CHECK-NEXT: }
// CHECK-NEXT: memref.dealloc %[[BUF]] : memref<100xf32>
// CHECK-NEXT: }
// -----
// Lower bound is a max; upper bound is a min. This pattern typically appears
// with multi-level tiling when the tile sizes used don't divide loop trip
// counts.
#lb = affine_map<()[s0, s1] -> (s0 * 512, s1 * 6)>
#ub = affine_map<()[s0, s1] -> (s0 * 512 + 512, s1 * 6 + 6)>
// CHECK-DAG: #[[$LB:.*]] = affine_map<()[s0, s1] -> (s0 * 512, s1 * 6)>
// CHECK-DAG: #[[$UB:.*]] = affine_map<()[s0, s1] -> (s0 * 512 + 512, s1 * 6 + 6)>
// CHECK-LABEL: max_lower_bound(%{{.*}}: memref<2048x516xf64>,
// CHECK-SAME: [[i:arg[0-9a-zA-Z_]+]]
// CHECK-SAME: [[j:arg[0-9a-zA-Z_]+]]
func.func @max_lower_bound(%M: memref<2048x516xf64>, %i : index, %j : index) {
affine.for %ii = 0 to 2048 {
affine.for %jj = max #lb()[%i, %j] to min #ub()[%i, %j] {
affine.load %M[%ii, %jj] : memref<2048x516xf64>
}
}
return
}
// CHECK: %[[BUF:.*]] = memref.alloc() : memref<2048x6xf64>
// CHECK-NEXT: affine.for %[[ii:.*]] = 0 to 2048 {
// CHECK-NEXT: affine.for %[[jj:.*]] = max #[[$LB]]()[%[[i]], %[[j]]] to min #[[$UB]]()[%[[i]], %[[j]]] {
// CHECK-NEXT: affine.load %{{.*}}[%[[ii]], %[[jj]]] : memref<2048x516xf64>
// CHECK-NEXT: affine.store %{{.*}}, %[[BUF]][%[[ii]], %[[jj]] - symbol(%[[j]]) * 6] : memref<2048x6xf64>
// CHECK-NEXT: }
// CHECK-NEXT: }
// CHECK-NEXT: affine.for %[[ii_:.*]] = 0 to 2048 {
// CHECK-NEXT: affine.for %[[jj_:.*]] = max #[[$LB]]()[%{{.*}}, %{{.*}}] to min #[[$UB]]()[%{{.*}}, %{{.*}}] {
// CHECK-NEXT: affine.load %[[BUF]][%[[ii_]], %[[jj_]] - symbol(%[[j]]) * 6] : memref<2048x6xf64>
// CHECK-NEXT: }
// CHECK-NEXT: }
// CHECK-NEXT: memref.dealloc %[[BUF]] : memref<2048x6xf64>
// -----
// CHECK-LABEL: func @empty_loops
func.func @empty_loops(%arg0: memref<1024x1024xf64>) {
// Empty loops - so no copy generation happens.
affine.for %i = 0 to 0 {
affine.load %arg0[0, %i] : memref<1024x1024xf64>
}
affine.for %i = 0 to -16 {
affine.load %arg0[0, %i] : memref<1024x1024xf64>
}
return
// CHECK-NOT: memref.alloc
// CHECK: return
}
#map16 = affine_map<(d0, d1, d2) -> (d0 * 40 + d1 * 8 + d2 * 2)>
#map17 = affine_map<(d0, d1, d2) -> (d0 * 40 + d1 * 8 + d2 * 2 + 2)>
// CHECK-LABEL: func @affine_parallel
func.func @affine_parallel(%85:memref<2x5x4x2xi64>) {
affine.for %arg0 = 0 to 2 {
affine.parallel (%arg1) = (0) to (5) {
affine.parallel (%arg2) = (0) to (4) {
affine.for %arg3 = #map16(%arg0, %arg1, %arg2) to #map17(%arg0, %arg1, %arg2) {
%105 = affine.load %85[((%arg3 floordiv 2) floordiv 4) floordiv 5, ((%arg3 floordiv 2) floordiv 4) mod 5, (%arg3 floordiv 2) mod 4, %arg3 mod 2] : memref<2x5x4x2xi64>
}
}
}
}
// CHECK: affine.for
// CHECK-NEXT: affine.for %{{.*}} = 0 to 5
// CHECK-NEXT: affine.for %{{.*}} = 0 to 4
// CHECK-NEXT: affine.for %{{.*}} = 0 to 2
// CHECK: affine.for
// CHECK-NEXT: affine.parallel
// CHECK-NEXT: affine.parallel
return
}
// CHECK-LABEL: func @index_elt_type
func.func @index_elt_type(%arg0: memref<1x2x4x8xindex>) {
affine.for %arg1 = 0 to 1 {
affine.for %arg2 = 0 to 2 {
affine.for %arg3 = 0 to 4 {
affine.for %arg4 = 0 to 8 {
affine.store %arg4, %arg0[%arg1, %arg2, %arg3, %arg4] : memref<1x2x4x8xindex>
}
}
}
}
// CHECK: affine.for %{{.*}} = 0 to 1
// CHECK-NEXT: affine.for %{{.*}} = 0 to 2
// CHECK-NEXT: affine.for %{{.*}} = 0 to 4
// CHECK-NEXT: affine.for %{{.*}} = 0 to 8
// CHECK: affine.for %{{.*}} = 0 to 2
// CHECK-NEXT: affine.for %{{.*}} = 0 to 4
// CHECK-NEXT: affine.for %{{.*}} = 0 to 8
return
}
|
