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
|
//------------------------------------------------------------------------------
// GB_dense_ewise3_noaccum_template: C = A+B where all 3 matrices are dense
//------------------------------------------------------------------------------
// SuiteSparse:GraphBLAS, Timothy A. Davis, (c) 2017-2022, All Rights Reserved.
// SPDX-License-Identifier: Apache-2.0
//------------------------------------------------------------------------------
#include "GB_unused.h"
{
//--------------------------------------------------------------------------
// get A, B, and C
//--------------------------------------------------------------------------
// any matrix may be aliased to any other (C==A, C==B, and/or A==B)
GB_ATYPE *Ax = (GB_ATYPE *) A->x ;
GB_BTYPE *Bx = (GB_BTYPE *) B->x ;
GB_CTYPE *Cx = (GB_CTYPE *) C->x ;
const int64_t cnz = GB_nnz (C) ;
ASSERT (GB_as_if_full (A)) ;
ASSERT (GB_as_if_full (B)) ;
ASSERT (GB_IS_FULL (C)) ;
ASSERT (!C->iso) ;
ASSERT (!A->iso) ;
ASSERT (!B->iso) ;
int64_t p ;
//--------------------------------------------------------------------------
// C = A+B where all 3 matrices are dense
//--------------------------------------------------------------------------
#if GB_CTYPE_IS_BTYPE
if (C == B)
{
//----------------------------------------------------------------------
// C = A+C where A and C are dense
//----------------------------------------------------------------------
// C and B cannot be aliased if their types differ
#pragma omp parallel for num_threads(nthreads) schedule(static)
for (p = 0 ; p < cnz ; p++)
{
GB_GETA (aij, Ax, p, false) ; // aij = Ax [p]
GB_BINOP (GB_CX (p), aij, GB_CX (p), 0, 0) ; // Cx [p] = aij+Cx [p]
}
}
else
#endif
#if GB_CTYPE_IS_ATYPE
if (C == A)
{
//----------------------------------------------------------------------
// C = C+B where B and C are dense
//----------------------------------------------------------------------
#pragma omp parallel for num_threads(nthreads) schedule(static)
for (p = 0 ; p < cnz ; p++)
{
GB_GETB (bij, Bx, p, false) ; // bij = Bx [p]
GB_BINOP (GB_CX (p), GB_CX (p), bij, 0, 0) ; // Cx [p] += bij
}
}
else
#endif
{
//----------------------------------------------------------------------
// C = A+B where all 3 matrices are dense
//----------------------------------------------------------------------
// note that A and B may still be aliased to each other
#pragma omp parallel for num_threads(nthreads) schedule(static)
for (p = 0 ; p < cnz ; p++)
{
GB_GETA (aij, Ax, p, false) ; // aij = Ax [p]
GB_GETB (bij, Bx, p, false) ; // bij = Bx [p]
GB_BINOP (GB_CX (p), aij, bij, 0, 0) ; // Cx [p] = aij + bij
}
}
}
|
