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
|
// RUN: mlir-opt %s -sparsification | FileCheck %s
#SV = #sparse_tensor.encoding<{ lvlTypes = [ "compressed" ] }>
#trait = {
indexing_maps = [
affine_map<(i) -> (i)>, // A (in)
affine_map<(i) -> (i)> // X (out)
],
iterator_types = ["parallel"]
}
// CHECK-LABEL: func.func @allout_inplace(
// CHECK-SAME: %[[VAL_0:.*]]: tensor<10xi32, #{{.*}}>,
// CHECK-SAME: %[[VAL_1:.*]]: tensor<10xf32>) -> tensor<10xf32> {
// CHECK-DAG: %[[VAL_2:.*]] = arith.constant 0 : index
// CHECK-DAG: %[[VAL_3:.*]] = arith.constant 0.000000e+00 : f32
// CHECK-DAG: %[[VAL_4:.*]] = arith.constant 1 : index
// CHECK: %[[VAL_5:.*]] = sparse_tensor.positions %[[VAL_0]] {level = 0 : index} : tensor<10xi32, #{{.*}}> to memref<?xindex>
// CHECK: %[[VAL_6:.*]] = sparse_tensor.coordinates %[[VAL_0]] {level = 0 : index} : tensor<10xi32, #{{.*}}> to memref<?xindex>
// CHECK: %[[VAL_7:.*]] = sparse_tensor.values %[[VAL_0]] : tensor<10xi32, #{{.*}}> to memref<?xi32>
// CHECK: %[[VAL_8:.*]] = bufferization.to_memref %[[VAL_1]] : memref<10xf32>
// CHECK: linalg.fill ins(%[[VAL_3]] : f32) outs(%[[VAL_8]] : memref<10xf32>)
// CHECK: %[[VAL_9:.*]] = memref.load %[[VAL_5]]{{\[}}%[[VAL_2]]] : memref<?xindex>
// CHECK: %[[VAL_10:.*]] = memref.load %[[VAL_5]]{{\[}}%[[VAL_4]]] : memref<?xindex>
// CHECK: scf.for %[[VAL_11:.*]] = %[[VAL_9]] to %[[VAL_10]] step %[[VAL_4]] {
// CHECK: %[[VAL_12:.*]] = memref.load %[[VAL_6]]{{\[}}%[[VAL_11]]] : memref<?xindex>
// CHECK: %[[VAL_13:.*]] = memref.load %[[VAL_7]]{{\[}}%[[VAL_11]]] : memref<?xi32>
// CHECK: %[[VAL_14:.*]] = arith.sitofp %[[VAL_13]] : i32 to f32
// CHECK: memref.store %[[VAL_14]], %[[VAL_8]]{{\[}}%[[VAL_12]]] : memref<10xf32>
// CHECK: }
// CHECK: %[[VAL_15:.*]] = bufferization.to_tensor %[[VAL_8]] : memref<10xf32>
// CHECK: return %[[VAL_15]] : tensor<10xf32>
// CHECK: }
func.func @allout_inplace(%arga: tensor<10xi32, #SV>,
%argb: tensor<10xf32>) -> tensor<10xf32> {
%0 = linalg.generic #trait
ins(%arga: tensor<10xi32, #SV>)
outs(%argb: tensor<10xf32>) {
^bb(%a: i32, %x : f32):
%cst = arith.sitofp %a : i32 to f32
linalg.yield %cst : f32
} -> tensor<10xf32>
return %0 : tensor<10xf32>
}
// CHECK-LABEL: func.func @allout_materialize(
// CHECK-SAME: %[[VAL_0:.*]]: tensor<10xi32, #{{.*}}>) -> tensor<10xf32> {
// CHECK-DAG: %[[VAL_1:.*]] = arith.constant 0 : index
// CHECK-DAG: %[[VAL_2:.*]] = arith.constant 0.000000e+00 : f32
// CHECK-DAG: %[[VAL_3:.*]] = arith.constant 1 : index
// CHECK: %[[VAL_4:.*]] = bufferization.alloc_tensor() : tensor<10xf32>
// CHECK: %[[VAL_5:.*]] = sparse_tensor.positions %[[VAL_0]] {level = 0 : index} : tensor<10xi32, #{{.*}}> to memref<?xindex>
// CHECK: %[[VAL_6:.*]] = sparse_tensor.coordinates %[[VAL_0]] {level = 0 : index} : tensor<10xi32, #{{.*}}> to memref<?xindex>
// CHECK: %[[VAL_7:.*]] = sparse_tensor.values %[[VAL_0]] : tensor<10xi32, #{{.*}}> to memref<?xi32>
// CHECK: %[[VAL_8:.*]] = bufferization.to_memref %[[VAL_4]] : memref<10xf32>
// CHECK: linalg.fill ins(%[[VAL_2]] : f32) outs(%[[VAL_8]] : memref<10xf32>)
// CHECK: %[[VAL_9:.*]] = memref.load %[[VAL_5]]{{\[}}%[[VAL_1]]] : memref<?xindex>
// CHECK: %[[VAL_10:.*]] = memref.load %[[VAL_5]]{{\[}}%[[VAL_3]]] : memref<?xindex>
// CHECK: scf.for %[[VAL_11:.*]] = %[[VAL_9]] to %[[VAL_10]] step %[[VAL_3]] {
// CHECK: %[[VAL_12:.*]] = memref.load %[[VAL_6]]{{\[}}%[[VAL_11]]] : memref<?xindex>
// CHECK: %[[VAL_13:.*]] = memref.load %[[VAL_7]]{{\[}}%[[VAL_11]]] : memref<?xi32>
// CHECK: %[[VAL_14:.*]] = arith.sitofp %[[VAL_13]] : i32 to f32
// CHECK: memref.store %[[VAL_14]], %[[VAL_8]]{{\[}}%[[VAL_12]]] : memref<10xf32>
// CHECK: }
// CHECK: %[[VAL_15:.*]] = bufferization.to_tensor %[[VAL_8]] : memref<10xf32>
// CHECK: return %[[VAL_15]] : tensor<10xf32>
// CHECK: }
func.func @allout_materialize(%arga: tensor<10xi32, #SV>) -> tensor<10xf32> {
%m = bufferization.alloc_tensor() : tensor<10xf32>
%0 = linalg.generic #trait
ins(%arga: tensor<10xi32, #SV>)
outs(%m: tensor<10xf32>) {
^bb(%a: i32, %x : f32):
%cst = arith.sitofp %a : i32 to f32
linalg.yield %cst : f32
} -> tensor<10xf32>
return %0 : tensor<10xf32>
}
// CHECK-LABEL: func.func @update_inplace(
// CHECK-SAME: %[[VAL_0:.*]]: tensor<10xf32, #{{.*}}>,
// CHECK-SAME: %[[VAL_1:.*]]: tensor<10xf32>) -> tensor<10xf32> {
// CHECK-DAG: %[[VAL_2:.*]] = arith.constant 0 : index
// CHECK-DAG: %[[VAL_3:.*]] = arith.constant 1 : index
// CHECK: %[[VAL_4:.*]] = sparse_tensor.positions %[[VAL_0]] {level = 0 : index} : tensor<10xf32, #{{.*}}> to memref<?xindex>
// CHECK: %[[VAL_5:.*]] = sparse_tensor.coordinates %[[VAL_0]] {level = 0 : index} : tensor<10xf32, #{{.*}}> to memref<?xindex>
// CHECK: %[[VAL_6:.*]] = sparse_tensor.values %[[VAL_0]] : tensor<10xf32, #{{.*}}> to memref<?xf32>
// CHECK: %[[VAL_7:.*]] = bufferization.to_memref %[[VAL_1]] : memref<10xf32>
// CHECK: %[[VAL_8:.*]] = memref.load %[[VAL_4]]{{\[}}%[[VAL_2]]] : memref<?xindex>
// CHECK: %[[VAL_9:.*]] = memref.load %[[VAL_4]]{{\[}}%[[VAL_3]]] : memref<?xindex>
// CHECK: scf.for %[[VAL_10:.*]] = %[[VAL_8]] to %[[VAL_9]] step %[[VAL_3]] {
// CHECK: %[[VAL_11:.*]] = memref.load %[[VAL_5]]{{\[}}%[[VAL_10]]] : memref<?xindex>
// CHECK: %[[VAL_12:.*]] = memref.load %[[VAL_6]]{{\[}}%[[VAL_10]]] : memref<?xf32>
// CHECK: %[[VAL_13:.*]] = memref.load %[[VAL_7]]{{\[}}%[[VAL_11]]] : memref<10xf32>
// CHECK: %[[VAL_14:.*]] = arith.addf %[[VAL_12]], %[[VAL_13]] : f32
// CHECK: memref.store %[[VAL_14]], %[[VAL_7]]{{\[}}%[[VAL_11]]] : memref<10xf32>
// CHECK: }
// CHECK: %[[VAL_15:.*]] = bufferization.to_tensor %[[VAL_7]] : memref<10xf32>
// CHECK: return %[[VAL_15]] : tensor<10xf32>
// CHECK: }
func.func @update_inplace(%arga: tensor<10xf32, #SV>,
%argb: tensor<10xf32>) -> tensor<10xf32> {
%0 = linalg.generic #trait
ins(%arga: tensor<10xf32, #SV>)
outs(%argb: tensor<10xf32>) {
^bb(%a: f32, %x : f32):
%up = arith.addf %a, %x : f32
linalg.yield %up : f32
} -> tensor<10xf32>
return %0 : tensor<10xf32>
}
|
