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
|
// RUN: mlir-opt %s -async-parallel-for \
// RUN: -async-to-async-runtime \
// RUN: -async-runtime-ref-counting \
// RUN: -async-runtime-ref-counting-opt \
// RUN: -convert-async-to-llvm \
// RUN: -convert-scf-to-cf \
// RUN: -finalize-memref-to-llvm \
// RUN: -arith-expand \
// RUN: -memref-expand \
// RUN: -convert-func-to-llvm \
// RUN: -reconcile-unrealized-casts \
// RUN: | mlir-cpu-runner \
// RUN: -e entry -entry-point-result=void -O0 \
// RUN: -shared-libs=%mlir_runner_utils \
// RUN: -shared-libs=%mlir_async_runtime\
// RUN: | FileCheck %s --dump-input=always
// RUN: mlir-opt %s -async-parallel-for \
// RUN: -async-to-async-runtime \
// RUN: -async-runtime-policy-based-ref-counting \
// RUN: -convert-async-to-llvm \
// RUN: -convert-scf-to-cf \
// RUN: -finalize-memref-to-llvm \
// RUN: -arith-expand \
// RUN: -memref-expand \
// RUN: -convert-func-to-llvm \
// RUN: -reconcile-unrealized-casts \
// RUN: | mlir-cpu-runner \
// RUN: -e entry -entry-point-result=void -O0 \
// RUN: -shared-libs=%mlir_runner_utils \
// RUN: -shared-libs=%mlir_async_runtime\
// RUN: | FileCheck %s --dump-input=always
// RUN: mlir-opt %s -async-parallel-for="async-dispatch=false \
// RUN: num-workers=20 \
// RUN: min-task-size=1" \
// RUN: -async-to-async-runtime \
// RUN: -async-runtime-ref-counting \
// RUN: -async-runtime-ref-counting-opt \
// RUN: -convert-async-to-llvm \
// RUN: -convert-scf-to-cf \
// RUN: -finalize-memref-to-llvm \
// RUN: -arith-expand \
// RUN: -memref-expand \
// RUN: -convert-func-to-llvm \
// RUN: -reconcile-unrealized-casts \
// RUN: | mlir-cpu-runner \
// RUN: -e entry -entry-point-result=void -O0 \
// RUN: -shared-libs=%mlir_runner_utils \
// RUN: -shared-libs=%mlir_async_runtime\
// RUN: | FileCheck %s --dump-input=always
// Suppress constant folding by introducing "dynamic" zero value at runtime.
func.func private @zero() -> index {
%0 = arith.constant 0 : index
return %0 : index
}
func.func @entry() {
%c0 = arith.constant 0.0 : f32
%c1 = arith.constant 1 : index
%c2 = arith.constant 2 : index
%c3 = arith.constant 3 : index
%lb = arith.constant 0 : index
%ub = arith.constant 9 : index
%A = memref.alloc() : memref<9xf32>
%U = memref.cast %A : memref<9xf32> to memref<*xf32>
// Initialize memref with zeros because we do load and store to in every test
// to verify that we process each element of the iteration space once.
scf.parallel (%i) = (%lb) to (%ub) step (%c1) {
memref.store %c0, %A[%i] : memref<9xf32>
}
// 1. %i = (0) to (9) step (1)
scf.parallel (%i) = (%lb) to (%ub) step (%c1) {
%0 = arith.index_cast %i : index to i32
%1 = arith.sitofp %0 : i32 to f32
%2 = memref.load %A[%i] : memref<9xf32>
%3 = arith.addf %1, %2 : f32
memref.store %3, %A[%i] : memref<9xf32>
}
// CHECK: [0, 1, 2, 3, 4, 5, 6, 7, 8]
call @printMemrefF32(%U): (memref<*xf32>) -> ()
scf.parallel (%i) = (%lb) to (%ub) step (%c1) {
memref.store %c0, %A[%i] : memref<9xf32>
}
// 2. %i = (0) to (9) step (2)
scf.parallel (%i) = (%lb) to (%ub) step (%c2) {
%0 = arith.index_cast %i : index to i32
%1 = arith.sitofp %0 : i32 to f32
%2 = memref.load %A[%i] : memref<9xf32>
%3 = arith.addf %1, %2 : f32
memref.store %3, %A[%i] : memref<9xf32>
}
// CHECK: [0, 0, 2, 0, 4, 0, 6, 0, 8]
call @printMemrefF32(%U): (memref<*xf32>) -> ()
scf.parallel (%i) = (%lb) to (%ub) step (%c1) {
memref.store %c0, %A[%i] : memref<9xf32>
}
// 3. %i = (-20) to (-11) step (3)
%lb0 = arith.constant -20 : index
%ub0 = arith.constant -11 : index
scf.parallel (%i) = (%lb0) to (%ub0) step (%c3) {
%0 = arith.index_cast %i : index to i32
%1 = arith.sitofp %0 : i32 to f32
%2 = arith.constant 20 : index
%3 = arith.addi %i, %2 : index
%4 = memref.load %A[%3] : memref<9xf32>
%5 = arith.addf %1, %4 : f32
memref.store %5, %A[%3] : memref<9xf32>
}
// CHECK: [-20, 0, 0, -17, 0, 0, -14, 0, 0]
call @printMemrefF32(%U): (memref<*xf32>) -> ()
// 4. Check that loop with zero iterations doesn't crash at runtime.
%lb1 = call @zero(): () -> (index)
%ub1 = call @zero(): () -> (index)
scf.parallel (%i) = (%lb1) to (%ub1) step (%c1) {
%false = arith.constant 0 : i1
cf.assert %false, "should never be executed"
}
memref.dealloc %A : memref<9xf32>
return
}
func.func private @printMemrefF32(memref<*xf32>) attributes { llvm.emit_c_interface }
|
