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
|
// Copyright Yahoo. Licensed under the terms of the Apache 2.0 license. See LICENSE in the project root.
#include <vespa/eval/eval/fast_value.h>
#include <vespa/eval/eval/operation.h>
#include <vespa/eval/eval/tensor_function.h>
#include <vespa/eval/eval/test/eval_fixture.h>
#include <vespa/eval/eval/test/gen_spec.h>
#include <vespa/eval/instruction/dense_multi_matmul_function.h>
#include <vespa/vespalib/testkit/test_kit.h>
#include <vespa/vespalib/util/stash.h>
#include <vespa/vespalib/util/stringfmt.h>
using namespace vespalib;
using namespace vespalib::eval;
using namespace vespalib::eval::test;
using namespace vespalib::eval::tensor_function;
const ValueBuilderFactory &prod_factory = FastValueBuilderFactory::get();
struct FunInfo {
using LookFor = DenseMultiMatMulFunction;
size_t lhs_size;
size_t common_size;
size_t rhs_size;
size_t matmul_cnt;
bool lhs_inner;
bool rhs_inner;
void verify(const LookFor &fun) const {
EXPECT_TRUE(fun.result_is_mutable());
EXPECT_EQUAL(fun.lhs_size(), lhs_size);
EXPECT_EQUAL(fun.common_size(), common_size);
EXPECT_EQUAL(fun.rhs_size(), rhs_size);
EXPECT_EQUAL(fun.matmul_cnt(), matmul_cnt);
EXPECT_EQUAL(fun.lhs_common_inner(), lhs_inner);
EXPECT_EQUAL(fun.rhs_common_inner(), rhs_inner);
}
};
void verify_optimized(const vespalib::string &expr, const FunInfo &details)
{
TEST_STATE(expr.c_str());
CellTypeSpace stable_types(CellTypeUtils::list_stable_types(), 2);
CellTypeSpace unstable_types(CellTypeUtils::list_unstable_types(), 2);
EvalFixture::verify<FunInfo>(expr, {details}, CellTypeSpace(stable_types).same());
EvalFixture::verify<FunInfo>(expr, {}, CellTypeSpace(std::move(stable_types)).different());
EvalFixture::verify<FunInfo>(expr, {}, std::move(unstable_types));
}
void verify_not_optimized(const vespalib::string &expr) {
TEST_STATE(expr.c_str());
CellTypeSpace just_double({CellType::DOUBLE}, 2);
EvalFixture::verify<FunInfo>(expr, {}, just_double);
}
TEST("require that multi matmul can be optimized") {
FunInfo details = { .lhs_size = 2, .common_size = 3, .rhs_size = 5,
.matmul_cnt = 6, .lhs_inner = true, .rhs_inner = true };
TEST_DO(verify_optimized("reduce(A2B1C3a2d3*A2B1C3b5d3,sum,d)", details));
TEST_DO(verify_optimized("reduce(A2B1C3a2d3*A2B1C3b5d3,sum,d)", details));
}
TEST("require that single multi matmul can be optimized") {
FunInfo details = { .lhs_size = 2, .common_size = 3, .rhs_size = 5,
.matmul_cnt = 1, .lhs_inner = true, .rhs_inner = true };
TEST_DO(verify_optimized("reduce(A1a2d3*A1b5d3,sum,d)", details));
}
TEST("require that multi matmul with lambda can be optimized") {
FunInfo details = { .lhs_size = 2, .common_size = 3, .rhs_size = 5,
.matmul_cnt = 6, .lhs_inner = true, .rhs_inner = true };
TEST_DO(verify_optimized("reduce(join(A2B1C3a2d3,A2B1C3b5d3,f(x,y)(x*y)),sum,d)", details));
}
TEST("require that expressions similar to multi matmul are not optimized") {
TEST_DO(verify_not_optimized("reduce(A2B1C3a2d3*A2B1C3b5d3,sum,a)"));
TEST_DO(verify_not_optimized("reduce(A2B1C3a2d3*A2B1C3b5d3,sum,b)"));
TEST_DO(verify_not_optimized("reduce(A2B1C3a2d3*A2B1C3b5d3,prod,d)"));
TEST_DO(verify_not_optimized("reduce(A2B1C3a2d3*A2B1C3b5d3,sum)"));
TEST_DO(verify_not_optimized("reduce(join(A2B1C3a2d3,A2B1C3b5d3,f(x,y)(y*x)),sum,d)"));
TEST_DO(verify_not_optimized("reduce(join(A2B1C3a2d3,A2B1C3b5d3,f(x,y)(x+y)),sum,d)"));
TEST_DO(verify_not_optimized("reduce(join(A2B1C3a2d3,A2B1C3b5d3,f(x,y)(x*x)),sum,d)"));
TEST_DO(verify_not_optimized("reduce(join(A2B1C3a2d3,A2B1C3b5d3,f(x,y)(y*y)),sum,d)"));
TEST_DO(verify_not_optimized("reduce(join(A2B1C3a2d3,A2B1C3b5d3,f(x,y)(x*y*1)),sum,d)"));
TEST_DO(verify_not_optimized("reduce(A2B1C3a2c3*A2B1C3b5d3,sum,d)"));
TEST_DO(verify_not_optimized("reduce(A2B1C3a2c3*A2B1C3b5d3,sum,c)"));
}
TEST("require that multi matmul must have matching dimension prefix") {
TEST_DO(verify_not_optimized("reduce(B1C3a2d3*A2B1C3b5d3,sum,d)"));
TEST_DO(verify_not_optimized("reduce(A2B1C3a2d3*B1C3b5d3,sum,d)"));
}
TEST("require that multi matmul must have inner nesting of matmul dimensions") {
TEST_DO(verify_not_optimized("reduce(A2D3a2b1c3*B5D3a2b1c3,sum,D)"));
TEST_DO(verify_not_optimized("reduce(B5D3a2b1c3*A2D3a2b1c3,sum,D)"));
}
TEST("require that multi matmul ignores trivial dimensions") {
FunInfo details = { .lhs_size = 2, .common_size = 3, .rhs_size = 5,
.matmul_cnt = 6, .lhs_inner = true, .rhs_inner = true };
TEST_DO(verify_optimized("reduce(A2B1C3D1a2c1d3e1*A2B1C3b5d3,sum,d)", details));
TEST_DO(verify_optimized("reduce(A2B1C3b5d3*A2B1C3D1a2c1d3e1,sum,d)", details));
}
TEST("require that multi matmul function can be debug dumped") {
EvalFixture fixture(prod_factory, "reduce(m1*m2,sum,d)", EvalFixture::ParamRepo()
.add("m1", GenSpec::from_desc("A2B1C3a2d3"))
.add("m2", GenSpec::from_desc("A2B1C3b5d3")), true);
auto info = fixture.find_all<DenseMultiMatMulFunction>();
ASSERT_EQUAL(info.size(), 1u);
fprintf(stderr, "%s\n", info[0]->as_string().c_str());
}
TEST("require that multi matmul inner/inner works correctly") {
FunInfo details = { .lhs_size = 2, .common_size = 3, .rhs_size = 5,
.matmul_cnt = 6, .lhs_inner = true, .rhs_inner = true };
TEST_DO(verify_optimized("reduce(A2B1C3a2d3*A2B1C3b5d3,sum,d)", details));
}
TEST("require that multi matmul inner/outer works correctly") {
FunInfo details = { .lhs_size = 2, .common_size = 5, .rhs_size = 3,
.matmul_cnt = 6, .lhs_inner = true, .rhs_inner = false };
TEST_DO(verify_optimized("reduce(A2B1C3a2b5*A2B1C3b5d3,sum,b)", details));
}
TEST("require that multi matmul outer/outer works correctly") {
FunInfo details = { .lhs_size = 2, .common_size = 5, .rhs_size = 3,
.matmul_cnt = 6, .lhs_inner = false, .rhs_inner = false };
TEST_DO(verify_optimized("reduce(A2B1C3b5c2*A2B1C3b5d3,sum,b)", details));
}
TEST_MAIN() { TEST_RUN_ALL(); }
|
