| 1 |
|
| 2 |
/******************************************************* |
| 3 |
* |
| 4 |
* Copyright (c) 2003-2010 by University of Queensland |
| 5 |
* Earth Systems Science Computational Center (ESSCC) |
| 6 |
* http://www.uq.edu.au/esscc |
| 7 |
* |
| 8 |
* Primary Business: Queensland, Australia |
| 9 |
* Licensed under the Open Software License version 3.0 |
| 10 |
* http://www.opensource.org/licenses/osl-3.0.php |
| 11 |
* |
| 12 |
*******************************************************/ |
| 13 |
|
| 14 |
|
| 15 |
#ifndef INC_PASO_BLOCKOPS |
| 16 |
#define INC_PASO_BLOCKOPS |
| 17 |
|
| 18 |
#include "Common.h" |
| 19 |
#include "Paso.h" |
| 20 |
|
| 21 |
void Paso_BlockOps_allMV(dim_t n_block,dim_t n,double* D,index_t* pivot,double* x); |
| 22 |
|
| 23 |
#define Paso_BlockOps_Cpy_1(R, V) \ |
| 24 |
{\ |
| 25 |
*(R+0)=*(V+0);\ |
| 26 |
}\ |
| 27 |
|
| 28 |
#define Paso_BlockOps_Cpy_2(R, V) \ |
| 29 |
{\ |
| 30 |
*(R+0)=*(V+0);\ |
| 31 |
*(R+1)=*(V+1);\ |
| 32 |
}\ |
| 33 |
|
| 34 |
#define Paso_BlockOps_Cpy_3(R, V) \ |
| 35 |
{\ |
| 36 |
*(R+0)=*(V+0);\ |
| 37 |
*(R+1)=*(V+1);\ |
| 38 |
*(R+2)=*(V+2);\ |
| 39 |
}\ |
| 40 |
|
| 41 |
#define Paso_BlockOps_Cpy_N(N,R,V) \ |
| 42 |
{\ |
| 43 |
memcpy((void*)R, (void*)V, ( (size_t) N ) * sizeof(double) );\ |
| 44 |
}\ |
| 45 |
|
| 46 |
#define Paso_BlockOps_MV_1( R, MAT, V) \ |
| 47 |
{ \ |
| 48 |
register double S1=*V;\ |
| 49 |
register double A11=*MAT;\ |
| 50 |
*R=A11 * S1;\ |
| 51 |
}\ |
| 52 |
|
| 53 |
|
| 54 |
#define Paso_BlockOps_MV_2(R, MAT, V) \ |
| 55 |
{ \ |
| 56 |
register double S1=*(V+0);\ |
| 57 |
register double S2=*(V+1);\ |
| 58 |
register double A11=*(MAT+0);\ |
| 59 |
register double A12=*(MAT+2);\ |
| 60 |
register double A21=*(MAT+1);\ |
| 61 |
register double A22=*(MAT+3);\ |
| 62 |
*(R+0) = A11 * S1 + A12 * S2;\ |
| 63 |
*(R+1) = A21 * S1 + A22 * S2;\ |
| 64 |
}\ |
| 65 |
|
| 66 |
|
| 67 |
#define Paso_BlockOps_MV_3(R, MAT, V) \ |
| 68 |
{ \ |
| 69 |
register double S1=*(V+0);\ |
| 70 |
register double S2=*(V+1);\ |
| 71 |
register double S3=*(V+2);\ |
| 72 |
register double A11=*(MAT+0);\ |
| 73 |
register double A21=*(MAT+1);\ |
| 74 |
register double A31=*(MAT+2);\ |
| 75 |
register double A12=*(MAT+3);\ |
| 76 |
register double A22=*(MAT+4);\ |
| 77 |
register double A32=*(MAT+5);\ |
| 78 |
register double A13=*(MAT+6);\ |
| 79 |
register double A23=*(MAT+7);\ |
| 80 |
register double A33=*(MAT+8);\ |
| 81 |
*(R+0)=A11 * S1 + A12 * S2 + A13 * S3; \ |
| 82 |
*(R+1)=A21 * S1 + A22 * S2 + A23 * S3;\ |
| 83 |
*(R+2)=A31 * S1 + A32 * S2 + A33 * S3;\ |
| 84 |
}\ |
| 85 |
|
| 86 |
/* this does not work in place */ |
| 87 |
#define Paso_BlockOps_MV_N(N, R, MAT, V) \ |
| 88 |
{ \ |
| 89 |
dim_t p,q; \ |
| 90 |
for (p=0; p<N; p++ ) { \ |
| 91 |
register double S= *(MAT+p) * (*(V+0)); \ |
| 92 |
for ( q=1; q<N; q++ ) S+=*(MAT+p+N*q) * (*(V+q)); \ |
| 93 |
*(R+p)=S; \ |
| 94 |
} \ |
| 95 |
} \ |
| 96 |
|
| 97 |
#define Paso_BlockOps_Solve_N(N, R, MAT, pivot, V) \ |
| 98 |
{ \ |
| 99 |
*(pivot+0)=0;\ |
| 100 |
Esys_setError(TYPE_ERROR, "Paso_BlockOps_Solve_N: Right now there is support block size less than 4 only"); \ |
| 101 |
} \ |
| 102 |
|
| 103 |
#define Paso_BlockOps_SMV_1(R, MAT, V) \ |
| 104 |
{\ |
| 105 |
register double S1=*(V+0); \ |
| 106 |
register double A11=*(MAT+0);\ |
| 107 |
*(R+0)-=A11 * S1;\ |
| 108 |
}\ |
| 109 |
|
| 110 |
#define Paso_BlockOps_SMV_2(R,MAT, V) \ |
| 111 |
{\ |
| 112 |
register double S1=*(V+0); \ |
| 113 |
register double S2=*(V+1);\ |
| 114 |
register double A11=*(MAT+0);\ |
| 115 |
register double A12=*(MAT+2);\ |
| 116 |
register double A21=*(MAT+1);\ |
| 117 |
register double A22=*(MAT+3);\ |
| 118 |
*(R+0)-=A11 * S1 + A12 * S2;\ |
| 119 |
*(R+1)-=A21 * S1 + A22 * S2;\ |
| 120 |
}\ |
| 121 |
|
| 122 |
#define Paso_BlockOps_SMV_3(R, MAT, V) \ |
| 123 |
{\ |
| 124 |
register double S1=*(V+0);\ |
| 125 |
register double S2=*(V+1);\ |
| 126 |
register double S3=*(V+2);\ |
| 127 |
register double A11=*(MAT+0);\ |
| 128 |
register double A21=*(MAT+1);\ |
| 129 |
register double A31=*(MAT+2);\ |
| 130 |
register double A12=*(MAT+3);\ |
| 131 |
register double A22=*(MAT+4);\ |
| 132 |
register double A32=*(MAT+5);\ |
| 133 |
register double A13=*(MAT+6);\ |
| 134 |
register double A23=*(MAT+7);\ |
| 135 |
register double A33=*(MAT+8);\ |
| 136 |
*(R+0)-=A11 * S1 + A12 * S2 + A13 * S3; \ |
| 137 |
*(R+1)-=A21 * S1 + A22 * S2 + A23 * S3;\ |
| 138 |
*(R+2)-=A31 * S1 + A32 * S2 + A33 * S3;\ |
| 139 |
}\ |
| 140 |
|
| 141 |
/* this does not work in place */ |
| 142 |
#define Paso_BlockOps_SMV_N(N, R, MAT, V) \ |
| 143 |
{\ |
| 144 |
dim_t p,q; \ |
| 145 |
for (p=0; p<N; p++ ) { \ |
| 146 |
register double S= *(MAT+p) * *(V+0); \ |
| 147 |
for ( q=1; q<N; q++ ) S+=*(MAT+p+N*q) * (*(V+q)); \ |
| 148 |
*(R+p)-=S; \ |
| 149 |
}\ |
| 150 |
} \ |
| 151 |
|
| 152 |
#endif /* #INC_Paso_BlockOpsOPS */ |
Parent Directory
| 
Revision Log