|
IFPACK Development
|
00001 /*@HEADER 00002 // *********************************************************************** 00003 // 00004 // Ifpack: Object-Oriented Algebraic Preconditioner Package 00005 // Copyright (2009) Sandia Corporation 00006 // 00007 // Under terms of Contract DE-AC04-94AL85000, there is a non-exclusive 00008 // license for use of this work by or on behalf of the U.S. Government. 00009 // 00010 // This library is free software; you can redistribute it and/or modify 00011 // it under the terms of the GNU Lesser General Public License as 00012 // published by the Free Software Foundation; either version 2.1 of the 00013 // License, or (at your option) any later version. 00014 // 00015 // This library is distributed in the hope that it will be useful, but 00016 // WITHOUT ANY WARRANTY; without even the implied warranty of 00017 // MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU 00018 // Lesser General Public License for more details. 00019 // 00020 // You should have received a copy of the GNU Lesser General Public 00021 // License along with this library; if not, write to the Free Software 00022 // Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 00023 // USA 00024 // Questions? Contact Michael A. Heroux (maherou@sandia.gov) 00025 // 00026 // *********************************************************************** 00027 //@HEADER 00028 */ 00029 00030 #include "Euclid_dh.h" 00031 #include "Mat_dh.h" 00032 #include "Factor_dh.h" 00033 #include "Parser_dh.h" 00034 #include "TimeLog_dh.h" 00035 #include "SubdomainGraph_dh.h" 00036 00037 00038 static void scale_rhs_private (Euclid_dh ctx, double *rhs); 00039 static void permute_vec_n2o_private (Euclid_dh ctx, double *xIN, 00040 double *xOUT); 00041 static void permute_vec_o2n_private (Euclid_dh ctx, double *xIN, 00042 double *xOUT); 00043 00044 #undef __FUNC__ 00045 #define __FUNC__ "Euclid_dhApply" 00046 void 00047 Euclid_dhApply (Euclid_dh ctx, double *rhs, double *lhs) 00048 { 00049 START_FUNC_DH double *rhs_, *lhs_; 00050 double t1, t2; 00051 00052 t1 = MPI_Wtime (); 00053 00054 /* default settings; for everything except PILU */ 00055 ctx->from = 0; 00056 ctx->to = ctx->m; 00057 00058 /* case 1: no preconditioning */ 00059 if (!strcmp (ctx->algo_ilu, "none") || !strcmp (ctx->algo_par, "none")) 00060 { 00061 int i, m = ctx->m; 00062 for (i = 0; i < m; ++i) 00063 lhs[i] = rhs[i]; 00064 goto END_OF_FUNCTION; 00065 } 00066 00067 /*---------------------------------------------------------------- 00068 * permute and scale rhs vector 00069 *----------------------------------------------------------------*/ 00070 /* permute rhs vector */ 00071 if (ctx->sg != NULL) 00072 { 00073 00074 permute_vec_n2o_private (ctx, rhs, lhs); 00075 CHECK_V_ERROR; 00076 rhs_ = lhs; 00077 lhs_ = ctx->work2; 00078 } 00079 else 00080 { 00081 rhs_ = rhs; 00082 lhs_ = lhs; 00083 } 00084 00085 /* scale rhs vector */ 00086 if (ctx->isScaled) 00087 { 00088 00089 scale_rhs_private (ctx, rhs_); 00090 CHECK_V_ERROR; 00091 } 00092 00093 /* note: rhs_ is permuted, scaled; the input, "rhs" vector has 00094 not been disturbed. 00095 */ 00096 00097 /*---------------------------------------------------------------- 00098 * big switch to choose the appropriate triangular solve 00099 *----------------------------------------------------------------*/ 00100 00101 /* sequential and mpi block jacobi cases */ 00102 if (np_dh == 1 || !strcmp (ctx->algo_par, "bj")) 00103 { 00104 Factor_dhSolveSeq (rhs_, lhs_, ctx); 00105 CHECK_V_ERROR; 00106 } 00107 00108 00109 /* pilu case */ 00110 else 00111 { 00112 Factor_dhSolve (rhs_, lhs_, ctx); 00113 CHECK_V_ERROR; 00114 } 00115 00116 /*---------------------------------------------------------------- 00117 * unpermute lhs vector 00118 * (note: don't need to unscale, because we were clever) 00119 *----------------------------------------------------------------*/ 00120 if (ctx->sg != NULL) 00121 { 00122 permute_vec_o2n_private (ctx, lhs_, lhs); 00123 CHECK_V_ERROR; 00124 } 00125 00126 END_OF_FUNCTION:; 00127 00128 t2 = MPI_Wtime (); 00129 /* collective timing for triangular solves */ 00130 ctx->timing[TRI_SOLVE_T] += (t2 - t1); 00131 00132 /* collective timing for setup+krylov+triSolves 00133 (intent is to time linear solve, but this is 00134 at best probelematical!) 00135 */ 00136 ctx->timing[TOTAL_SOLVE_TEMP_T] = t2 - ctx->timing[SOLVE_START_T]; 00137 00138 /* total triangular solve count */ 00139 ctx->its += 1; 00140 ctx->itsTotal += 1; 00141 00142 END_FUNC_DH} 00143 00144 00145 #undef __FUNC__ 00146 #define __FUNC__ "scale_rhs_private" 00147 void 00148 scale_rhs_private (Euclid_dh ctx, double *rhs) 00149 { 00150 START_FUNC_DH int i, m = ctx->m; 00151 REAL_DH *scale = ctx->scale; 00152 00153 /* if matrix was scaled, must scale the rhs */ 00154 if (scale != NULL) 00155 { 00156 for (i = 0; i < m; ++i) 00157 { 00158 rhs[i] *= scale[i]; 00159 } 00160 } 00161 END_FUNC_DH} 00162 00163 00164 #undef __FUNC__ 00165 #define __FUNC__ "permute_vec_o2n_private" 00166 void 00167 permute_vec_o2n_private (Euclid_dh ctx, double *xIN, double *xOUT) 00168 { 00169 START_FUNC_DH int i, m = ctx->m; 00170 int *o2n = ctx->sg->o2n_col; 00171 for (i = 0; i < m; ++i) 00172 xOUT[i] = xIN[o2n[i]]; 00173 END_FUNC_DH} 00174 00175 00176 #undef __FUNC__ 00177 #define __FUNC__ "permute_vec_n2o_private" 00178 void 00179 permute_vec_n2o_private (Euclid_dh ctx, double *xIN, double *xOUT) 00180 { 00181 START_FUNC_DH int i, m = ctx->m; 00182 int *n2o = ctx->sg->n2o_row; 00183 for (i = 0; i < m; ++i) 00184 xOUT[i] = xIN[n2o[i]]; 00185 END_FUNC_DH}
1.7.4