xref: /petsc/src/mat/impls/aij/seq/superlu/superlu.c (revision 7adad9577d7f34aa90d02da3975546c0571d04a7)

#define PETSCMAT_DLL

/*  --------------------------------------------------------------------

     This file implements a subclass of the SeqAIJ matrix class that uses
     the SuperLU 3.0 sparse solver. You can use this as a starting point for
     implementing your own subclass of a PETSc matrix class.

     This demonstrates a way to make an implementation inheritence of a PETSc
     matrix type. This means constructing a new matrix type (SuperLU) by changing some
     of the methods of the previous type (SeqAIJ), adding additional data, and possibly
     additional method. (See any book on object oriented programming).
*/

/*
     Defines the data structure for the base matrix type (SeqAIJ)
*/
#include "src/mat/impls/aij/seq/aij.h"

/*
     SuperLU include files
*/
EXTERN_C_BEGIN
#if defined(PETSC_USE_COMPLEX)
#include "slu_zdefs.h"
#else
#include "slu_ddefs.h"
#endif
#include "slu_util.h"
EXTERN_C_END

/*
     This is the data we are "ADDING" to the SeqAIJ matrix type to get the SuperLU matrix type
*/
typedef struct {
  SuperMatrix       A,L,U,B,X;
  superlu_options_t options;
  PetscInt          *perm_c; /* column permutation vector */
  PetscInt          *perm_r; /* row permutations from partial pivoting */
  PetscInt          *etree;
  PetscReal         *R, *C;
  char              equed[1];
  PetscInt          lwork;
  void              *work;
  PetscReal         rpg, rcond;
  mem_usage_t       mem_usage;
  MatStructure      flg;

  /*
     This is where the methods for the superclass (SeqAIJ) are kept while we
     reset the pointers in the function table to the new (SuperLU) versions
  */
  PetscErrorCode (*MatDuplicate)(Mat,MatDuplicateOption,Mat*);
  PetscErrorCode (*MatView)(Mat,PetscViewer);
  PetscErrorCode (*MatAssemblyEnd)(Mat,MatAssemblyType);
  PetscErrorCode (*MatLUFactorSymbolic)(Mat,IS,IS,MatFactorInfo*,Mat*);
  PetscErrorCode (*MatDestroy)(Mat);

  /* Flag to clean up (non-global) SuperLU objects during Destroy */
  PetscTruth CleanUpSuperLU;
} Mat_SuperLU;

extern PetscErrorCode MatFactorInfo_SuperLU(Mat,PetscViewer);
extern PetscErrorCode MatLUFactorNumeric_SuperLU(Mat,MatFactorInfo *,Mat *);
extern PetscErrorCode MatDestroy_SuperLU(Mat);
extern PetscErrorCode MatView_SuperLU(Mat,PetscViewer);
extern PetscErrorCode MatAssemblyEnd_SuperLU(Mat,MatAssemblyType);
extern PetscErrorCode MatSolve_SuperLU(Mat,Vec,Vec);
extern PetscErrorCode MatSolveTranspose_SuperLU(Mat,Vec,Vec);
extern PetscErrorCode MatLUFactorSymbolic_SuperLU(Mat,IS,IS,MatFactorInfo *,Mat *);
extern PetscErrorCode MatDuplicate_SuperLU(Mat, MatDuplicateOption, Mat *);

/*
    Takes a SuperLU matrix (that is a SeqAIJ matrix with the additional SuperLU data-structures
   and methods) and converts it back to a regular SeqAIJ matrix.
*/
EXTERN_C_BEGIN
#undef __FUNCT__
#define __FUNCT__ "MatConvert_SuperLU_SeqAIJ"
PetscErrorCode PETSCMAT_DLLEXPORT MatConvert_SuperLU_SeqAIJ(Mat A,MatType type,MatReuse reuse,Mat *newmat)
{
  PetscErrorCode ierr;
  Mat            B=*newmat;
  Mat_SuperLU    *lu=(Mat_SuperLU *)A->spptr;

  PetscFunctionBegin;
  if (reuse == MAT_INITIAL_MATRIX) {
    ierr = MatDuplicate(A,MAT_COPY_VALUES,&B);CHKERRQ(ierr);
  }
  /* Reset the original SeqAIJ function pointers */
  B->ops->duplicate        = lu->MatDuplicate;
  B->ops->view             = lu->MatView;
  B->ops->assemblyend      = lu->MatAssemblyEnd;
  B->ops->lufactorsymbolic = lu->MatLUFactorSymbolic;
  B->ops->destroy          = lu->MatDestroy;
  ierr     = PetscFree(lu);CHKERRQ(ierr);
  A->spptr = PETSC_NULL;

  ierr = PetscObjectComposeFunction((PetscObject)B,"MatConvert_seqaij_superlu_C","",PETSC_NULL);CHKERRQ(ierr);
  ierr = PetscObjectComposeFunction((PetscObject)B,"MatConvert_superlu_seqaij_C","",PETSC_NULL);CHKERRQ(ierr);

  /* change the type name back to its original value */
  ierr = PetscObjectChangeTypeName((PetscObject)B,MATSEQAIJ);CHKERRQ(ierr);
  *newmat = B;
  PetscFunctionReturn(0);
}
EXTERN_C_END

EXTERN_C_BEGIN
#undef __FUNCT__
#define __FUNCT__ "MatConvert_SeqAIJ_SuperLU"
PetscErrorCode PETSCMAT_DLLEXPORT MatConvert_SeqAIJ_SuperLU(Mat A,MatType type,MatReuse reuse,Mat *newmat)
{
  PetscErrorCode ierr;
  Mat            B=*newmat;
  Mat_SuperLU    *lu;

  PetscFunctionBegin;
  if (reuse == MAT_INITIAL_MATRIX){
    ierr = MatDuplicate(A,MAT_COPY_VALUES,&B);CHKERRQ(ierr);
  }

  ierr = PetscNewLog(B,Mat_SuperLU,&lu);CHKERRQ(ierr);
  /* save the original SeqAIJ methods that we are changing */
  lu->MatDuplicate         = A->ops->duplicate;
  lu->MatView              = A->ops->view;
  lu->MatAssemblyEnd       = A->ops->assemblyend;
  lu->MatLUFactorSymbolic  = A->ops->lufactorsymbolic;
  lu->MatDestroy           = A->ops->destroy;
  lu->CleanUpSuperLU       = PETSC_FALSE;

  /* add to the matrix the location for all the SuperLU data is to be stored */
  B->spptr                 = (void*)lu; /* attach Mat_SuperLU to B->spptr is a bad design! */

  /* set the methods in the function table to the SuperLU versions */
  B->ops->duplicate        = MatDuplicate_SuperLU;
  B->ops->view             = MatView_SuperLU;
  B->ops->assemblyend      = MatAssemblyEnd_SuperLU;
  B->ops->lufactorsymbolic = MatLUFactorSymbolic_SuperLU;
  B->ops->choleskyfactorsymbolic = 0;
  B->ops->destroy          = MatDestroy_SuperLU;

  ierr = PetscObjectComposeFunctionDynamic((PetscObject)B,"MatConvert_seqaij_superlu_C",
                                           "MatConvert_SeqAIJ_SuperLU",MatConvert_SeqAIJ_SuperLU);CHKERRQ(ierr);
  ierr = PetscObjectComposeFunctionDynamic((PetscObject)B,"MatConvert_superlu_seqaij_C",
                                           "MatConvert_SuperLU_SeqAIJ",MatConvert_SuperLU_SeqAIJ);CHKERRQ(ierr);
  ierr = PetscInfo(0,"Using SuperLU for SeqAIJ LU factorization and solves.\n");CHKERRQ(ierr);
  ierr = PetscObjectChangeTypeName((PetscObject)B,MATSUPERLU);CHKERRQ(ierr);
  *newmat = B;
  PetscFunctionReturn(0);
}
EXTERN_C_END

/*
    Utility function
*/
#undef __FUNCT__
#define __FUNCT__ "MatFactorInfo_SuperLU"
PetscErrorCode MatFactorInfo_SuperLU(Mat A,PetscViewer viewer)
{
  Mat_SuperLU       *lu= (Mat_SuperLU*)A->spptr;
  PetscErrorCode    ierr;
  superlu_options_t options;

  PetscFunctionBegin;
  /* check if matrix is superlu_dist type */
  if (A->ops->solve != MatSolve_SuperLU) PetscFunctionReturn(0);

  options = lu->options;
  ierr = PetscViewerASCIIPrintf(viewer,"SuperLU run parameters:\n");CHKERRQ(ierr);
  ierr = PetscViewerASCIIPrintf(viewer,"  Equil: %s\n",(options.Equil != NO) ? "YES": "NO");CHKERRQ(ierr);
  ierr = PetscViewerASCIIPrintf(viewer,"  ColPerm: %D\n",options.ColPerm);CHKERRQ(ierr);
  ierr = PetscViewerASCIIPrintf(viewer,"  IterRefine: %D\n",options.IterRefine);CHKERRQ(ierr);
  ierr = PetscViewerASCIIPrintf(viewer,"  SymmetricMode: %s\n",(options.SymmetricMode != NO) ? "YES": "NO");CHKERRQ(ierr);
  ierr = PetscViewerASCIIPrintf(viewer,"  DiagPivotThresh: %g\n",options.DiagPivotThresh);CHKERRQ(ierr);
  ierr = PetscViewerASCIIPrintf(viewer,"  PivotGrowth: %s\n",(options.PivotGrowth != NO) ? "YES": "NO");CHKERRQ(ierr);
  ierr = PetscViewerASCIIPrintf(viewer,"  ConditionNumber: %s\n",(options.ConditionNumber != NO) ? "YES": "NO");CHKERRQ(ierr);
  ierr = PetscViewerASCIIPrintf(viewer,"  RowPerm: %D\n",options.RowPerm);CHKERRQ(ierr);
  ierr = PetscViewerASCIIPrintf(viewer,"  ReplaceTinyPivot: %s\n",(options.ReplaceTinyPivot != NO) ? "YES": "NO");CHKERRQ(ierr);
  ierr = PetscViewerASCIIPrintf(viewer,"  PrintStat: %s\n",(options.PrintStat != NO) ? "YES": "NO");CHKERRQ(ierr);
  ierr = PetscViewerASCIIPrintf(viewer,"  lwork: %D\n",lu->lwork);CHKERRQ(ierr);

  PetscFunctionReturn(0);
}

/*
    These are the methods provided to REPLACE the corresponding methods of the
   SeqAIJ matrix class. Other methods of SeqAIJ are not replaced
*/
#undef __FUNCT__
#define __FUNCT__ "MatLUFactorNumeric_SuperLU"
PetscErrorCode MatLUFactorNumeric_SuperLU(Mat A,MatFactorInfo *info,Mat *F)
{
  Mat_SeqAIJ     *aa = (Mat_SeqAIJ*)(A)->data;
  Mat_SuperLU    *lu = (Mat_SuperLU*)(*F)->spptr;
  PetscErrorCode ierr;
  PetscInt       sinfo;
  SuperLUStat_t  stat;
  PetscReal      ferr, berr;
  NCformat       *Ustore;
  SCformat       *Lstore;

  PetscFunctionBegin;
  if (lu->flg == SAME_NONZERO_PATTERN){ /* successing numerical factorization */
    lu->options.Fact = SamePattern;
    /* Ref: ~SuperLU_3.0/EXAMPLE/dlinsolx2.c */
    Destroy_SuperMatrix_Store(&lu->A);
    if ( lu->lwork >= 0 ) {
      Destroy_SuperNode_Matrix(&lu->L);
      Destroy_CompCol_Matrix(&lu->U);
      lu->options.Fact = SamePattern;
    }
  }

  /* Create the SuperMatrix for lu->A=A^T:
       Since SuperLU likes column-oriented matrices,we pass it the transpose,
       and then solve A^T X = B in MatSolve(). */
#if defined(PETSC_USE_COMPLEX)
  zCreate_CompCol_Matrix(&lu->A,A->cmap.n,A->rmap.n,aa->nz,(doublecomplex*)aa->a,aa->j,aa->i,
                           SLU_NC,SLU_Z,SLU_GE);
#else
  dCreate_CompCol_Matrix(&lu->A,A->cmap.n,A->rmap.n,aa->nz,aa->a,aa->j,aa->i,
                           SLU_NC,SLU_D,SLU_GE);
#endif

  /* Initialize the statistics variables. */
  StatInit(&stat);

  /* Numerical factorization */
  lu->B.ncol = 0;  /* Indicate not to solve the system */
#if defined(PETSC_USE_COMPLEX)
   zgssvx(&lu->options, &lu->A, lu->perm_c, lu->perm_r, lu->etree, lu->equed, lu->R, lu->C,
           &lu->L, &lu->U, lu->work, lu->lwork, &lu->B, &lu->X, &lu->rpg, &lu->rcond, &ferr, &berr,
           &lu->mem_usage, &stat, &sinfo);
#else
  dgssvx(&lu->options, &lu->A, lu->perm_c, lu->perm_r, lu->etree, lu->equed, lu->R, lu->C,
           &lu->L, &lu->U, lu->work, lu->lwork, &lu->B, &lu->X, &lu->rpg, &lu->rcond, &ferr, &berr,
           &lu->mem_usage, &stat, &sinfo);
#endif
  if ( !sinfo || sinfo == lu->A.ncol+1 ) {
    if ( lu->options.PivotGrowth )
      ierr = PetscPrintf(PETSC_COMM_SELF,"  Recip. pivot growth = %e\n", lu->rpg);
    if ( lu->options.ConditionNumber )
      ierr = PetscPrintf(PETSC_COMM_SELF,"  Recip. condition number = %e\n", lu->rcond);
  } else if ( sinfo > 0 ){
    if ( lu->lwork == -1 ) {
      ierr = PetscPrintf(PETSC_COMM_SELF,"  ** Estimated memory: %D bytes\n", sinfo - lu->A.ncol);
    } else {
      ierr = PetscPrintf(PETSC_COMM_SELF,"  Warning: gssvx() returns info %D\n",sinfo);
    }
  } else { /* sinfo < 0 */
    SETERRQ2(PETSC_ERR_LIB, "info = %D, the %D-th argument in gssvx() had an illegal value", sinfo,-sinfo);
  }

  if ( lu->options.PrintStat ) {
    ierr = PetscPrintf(PETSC_COMM_SELF,"MatLUFactorNumeric_SuperLU():\n");
    StatPrint(&stat);
    Lstore = (SCformat *) lu->L.Store;
    Ustore = (NCformat *) lu->U.Store;
    ierr = PetscPrintf(PETSC_COMM_SELF,"  No of nonzeros in factor L = %D\n", Lstore->nnz);
    ierr = PetscPrintf(PETSC_COMM_SELF,"  No of nonzeros in factor U = %D\n", Ustore->nnz);
    ierr = PetscPrintf(PETSC_COMM_SELF,"  No of nonzeros in L+U = %D\n", Lstore->nnz + Ustore->nnz - lu->A.ncol);
    ierr = PetscPrintf(PETSC_COMM_SELF,"  L\\U MB %.3f\ttotal MB needed %.3f\texpansions %D\n",
	       lu->mem_usage.for_lu/1e6, lu->mem_usage.total_needed/1e6,
	       lu->mem_usage.expansions);
  }
  StatFree(&stat);

  lu->flg = SAME_NONZERO_PATTERN;
  PetscFunctionReturn(0);
}

#undef __FUNCT__
#define __FUNCT__ "MatDestroy_SuperLU"
PetscErrorCode MatDestroy_SuperLU(Mat A)
{
  PetscErrorCode ierr;
  Mat_SuperLU    *lu=(Mat_SuperLU*)A->spptr;

  PetscFunctionBegin;
  if (lu->CleanUpSuperLU) { /* Free the SuperLU datastructures */
    Destroy_SuperMatrix_Store(&lu->A);
    Destroy_SuperMatrix_Store(&lu->B);
    Destroy_SuperMatrix_Store(&lu->X);

    ierr = PetscFree(lu->etree);CHKERRQ(ierr);
    ierr = PetscFree(lu->perm_r);CHKERRQ(ierr);
    ierr = PetscFree(lu->perm_c);CHKERRQ(ierr);
    ierr = PetscFree(lu->R);CHKERRQ(ierr);
    ierr = PetscFree(lu->C);CHKERRQ(ierr);
    if ( lu->lwork >= 0 ) {
      Destroy_SuperNode_Matrix(&lu->L);
      Destroy_CompCol_Matrix(&lu->U);
    }
  }
  ierr = MatConvert_SuperLU_SeqAIJ(A,MATSEQAIJ,MAT_REUSE_MATRIX,&A);CHKERRQ(ierr);
  ierr = (*A->ops->destroy)(A);CHKERRQ(ierr);
  PetscFunctionReturn(0);
}

#undef __FUNCT__
#define __FUNCT__ "MatView_SuperLU"
PetscErrorCode MatView_SuperLU(Mat A,PetscViewer viewer)
{
  PetscErrorCode    ierr;
  PetscTruth        iascii;
  PetscViewerFormat format;
  Mat_SuperLU       *lu=(Mat_SuperLU*)(A->spptr);

  PetscFunctionBegin;
  ierr = (*lu->MatView)(A,viewer);CHKERRQ(ierr);

  ierr = PetscTypeCompare((PetscObject)viewer,PETSC_VIEWER_ASCII,&iascii);CHKERRQ(ierr);
  if (iascii) {
    ierr = PetscViewerGetFormat(viewer,&format);CHKERRQ(ierr);
    if (format == PETSC_VIEWER_ASCII_INFO) {
      ierr = MatFactorInfo_SuperLU(A,viewer);CHKERRQ(ierr);
    }
  }
  PetscFunctionReturn(0);
}

#undef __FUNCT__
#define __FUNCT__ "MatAssemblyEnd_SuperLU"
PetscErrorCode MatAssemblyEnd_SuperLU(Mat A,MatAssemblyType mode) {
  PetscErrorCode ierr;
  Mat_SuperLU    *lu=(Mat_SuperLU*)(A->spptr);

  PetscFunctionBegin;
  ierr = (*lu->MatAssemblyEnd)(A,mode);CHKERRQ(ierr);
  lu->MatLUFactorSymbolic  = A->ops->lufactorsymbolic;
  A->ops->lufactorsymbolic = MatLUFactorSymbolic_SuperLU;
  PetscFunctionReturn(0);
}


#undef __FUNCT__
#define __FUNCT__ "MatSolve_SuperLU_Private"
PetscErrorCode MatSolve_SuperLU_Private(Mat A,Vec b,Vec x)
{
  Mat_SuperLU    *lu = (Mat_SuperLU*)A->spptr;
  PetscScalar    *barray,*xarray;
  PetscErrorCode ierr;
  PetscInt       info,i;
  SuperLUStat_t  stat;
  PetscReal      ferr,berr;

  PetscFunctionBegin;
  if ( lu->lwork == -1 ) {
    PetscFunctionReturn(0);
  }
  lu->B.ncol = 1;   /* Set the number of right-hand side */
  ierr = VecGetArray(b,&barray);CHKERRQ(ierr);
  ierr = VecGetArray(x,&xarray);CHKERRQ(ierr);

#if defined(PETSC_USE_COMPLEX)
  ((DNformat*)lu->B.Store)->nzval = (doublecomplex*)barray;
  ((DNformat*)lu->X.Store)->nzval = (doublecomplex*)xarray;
#else
  ((DNformat*)lu->B.Store)->nzval = barray;
  ((DNformat*)lu->X.Store)->nzval = xarray;
#endif

  /* Initialize the statistics variables. */
  StatInit(&stat);

  lu->options.Fact  = FACTORED; /* Indicate the factored form of A is supplied. */
#if defined(PETSC_USE_COMPLEX)
  zgssvx(&lu->options, &lu->A, lu->perm_c, lu->perm_r, lu->etree, lu->equed, lu->R, lu->C,
           &lu->L, &lu->U, lu->work, lu->lwork, &lu->B, &lu->X, &lu->rpg, &lu->rcond, &ferr, &berr,
           &lu->mem_usage, &stat, &info);
#else
  dgssvx(&lu->options, &lu->A, lu->perm_c, lu->perm_r, lu->etree, lu->equed, lu->R, lu->C,
           &lu->L, &lu->U, lu->work, lu->lwork, &lu->B, &lu->X, &lu->rpg, &lu->rcond, &ferr, &berr,
           &lu->mem_usage, &stat, &info);
#endif
  ierr = VecRestoreArray(b,&barray);CHKERRQ(ierr);
  ierr = VecRestoreArray(x,&xarray);CHKERRQ(ierr);

  if ( !info || info == lu->A.ncol+1 ) {
    if ( lu->options.IterRefine ) {
      ierr = PetscPrintf(PETSC_COMM_SELF,"Iterative Refinement:\n");
      ierr = PetscPrintf(PETSC_COMM_SELF,"  %8s%8s%16s%16s\n", "rhs", "Steps", "FERR", "BERR");
      for (i = 0; i < 1; ++i)
        ierr = PetscPrintf(PETSC_COMM_SELF,"  %8d%8d%16e%16e\n", i+1, stat.RefineSteps, ferr, berr);
    }
  } else if ( info > 0 ){
    if ( lu->lwork == -1 ) {
      ierr = PetscPrintf(PETSC_COMM_SELF,"  ** Estimated memory: %D bytes\n", info - lu->A.ncol);
    } else {
      ierr = PetscPrintf(PETSC_COMM_SELF,"  Warning: gssvx() returns info %D\n",info);
    }
  } else if (info < 0){
    SETERRQ2(PETSC_ERR_LIB, "info = %D, the %D-th argument in gssvx() had an illegal value", info,-info);
  }

  if ( lu->options.PrintStat ) {
    ierr = PetscPrintf(PETSC_COMM_SELF,"MatSolve__SuperLU():\n");
    StatPrint(&stat);
  }
  StatFree(&stat);
  PetscFunctionReturn(0);
}

#undef __FUNCT__
#define __FUNCT__ "MatSolve_SuperLU"
PetscErrorCode MatSolve_SuperLU(Mat A,Vec b,Vec x)
{
  Mat_SuperLU    *lu = (Mat_SuperLU*)A->spptr;
  PetscErrorCode ierr;

  PetscFunctionBegin;
  lu->options.Trans = TRANS;
  ierr = MatSolve_SuperLU_Private(A,b,x);CHKERRQ(ierr);
  PetscFunctionReturn(0);
}

#undef __FUNCT__
#define __FUNCT__ "MatSolveTranspose_SuperLU"
PetscErrorCode MatSolveTranspose_SuperLU(Mat A,Vec b,Vec x)
{
  Mat_SuperLU    *lu = (Mat_SuperLU*)A->spptr;
  PetscErrorCode ierr;

  PetscFunctionBegin;
  lu->options.Trans = NOTRANS;
  ierr = MatSolve_SuperLU_Private(A,b,x);CHKERRQ(ierr);
  PetscFunctionReturn(0);
}


/*
   Note the r permutation is ignored
*/
#undef __FUNCT__
#define __FUNCT__ "MatLUFactorSymbolic_SuperLU"
PetscErrorCode MatLUFactorSymbolic_SuperLU(Mat A,IS r,IS c,MatFactorInfo *info,Mat *F)
{
  Mat            B;
  Mat_SuperLU    *lu;
  PetscErrorCode ierr;
  PetscInt       m=A->rmap.n,n=A->cmap.n,indx;
  PetscTruth     flg;
  const char   *colperm[]={"NATURAL","MMD_ATA","MMD_AT_PLUS_A","COLAMD"}; /* MY_PERMC - not supported by the petsc interface yet */
  const char   *iterrefine[]={"NOREFINE", "SINGLE", "DOUBLE", "EXTRA"};
  const char   *rowperm[]={"NOROWPERM", "LargeDiag"}; /* MY_PERMC - not supported by the petsc interface yet */

  PetscFunctionBegin;
  ierr = MatCreate(((PetscObject)A)->comm,&B);CHKERRQ(ierr);
  ierr = MatSetSizes(B,A->rmap.n,A->cmap.n,PETSC_DETERMINE,PETSC_DETERMINE);CHKERRQ(ierr);
  ierr = MatSetType(B,((PetscObject)A)->type_name);CHKERRQ(ierr);
  ierr = MatSeqAIJSetPreallocation(B,0,PETSC_NULL);CHKERRQ(ierr);

  B->ops->lufactornumeric = MatLUFactorNumeric_SuperLU;
  B->ops->solve           = MatSolve_SuperLU;
  B->ops->solvetranspose  = MatSolveTranspose_SuperLU;
  B->factor               = FACTOR_LU;
  B->assembled            = PETSC_TRUE;  /* required by -ksp_view */

  lu = (Mat_SuperLU*)(B->spptr);

  /* Set SuperLU options */
    /* the default values for options argument:
	options.Fact = DOFACT;
        options.Equil = YES;
    	options.ColPerm = COLAMD;
	options.DiagPivotThresh = 1.0;
    	options.Trans = NOTRANS;
    	options.IterRefine = NOREFINE;
    	options.SymmetricMode = NO;
    	options.PivotGrowth = NO;
    	options.ConditionNumber = NO;
    	options.PrintStat = YES;
    */
  set_default_options(&lu->options);
  /* equilibration causes error in solve(), thus not supported here. See dgssvx.c for possible reason. */
  lu->options.Equil = NO;
  lu->options.PrintStat = NO;
  lu->lwork = 0;   /* allocate space internally by system malloc */

  ierr = PetscOptionsBegin(((PetscObject)A)->comm,((PetscObject)A)->prefix,"SuperLU Options","Mat");CHKERRQ(ierr);
  ierr = PetscOptionsEList("-mat_superlu_colperm","ColPerm","None",colperm,4,colperm[3],&indx,&flg);CHKERRQ(ierr);
  if (flg) {lu->options.ColPerm = (colperm_t)indx;}
  ierr = PetscOptionsEList("-mat_superlu_iterrefine","IterRefine","None",iterrefine,4,iterrefine[0],&indx,&flg);CHKERRQ(ierr);
  if (flg) { lu->options.IterRefine = (IterRefine_t)indx;}
  ierr = PetscOptionsTruth("-mat_superlu_symmetricmode","SymmetricMode","None",PETSC_FALSE,&flg,0);CHKERRQ(ierr);
  if (flg) lu->options.SymmetricMode = YES;
  ierr = PetscOptionsReal("-mat_superlu_diagpivotthresh","DiagPivotThresh","None",lu->options.DiagPivotThresh,&lu->options.DiagPivotThresh,PETSC_NULL);CHKERRQ(ierr);
  ierr = PetscOptionsTruth("-mat_superlu_pivotgrowth","PivotGrowth","None",PETSC_FALSE,&flg,0);CHKERRQ(ierr);
  if (flg) lu->options.PivotGrowth = YES;
  ierr = PetscOptionsTruth("-mat_superlu_conditionnumber","ConditionNumber","None",PETSC_FALSE,&flg,0);CHKERRQ(ierr);
  if (flg) lu->options.ConditionNumber = YES;
  ierr = PetscOptionsEList("-mat_superlu_rowperm","rowperm","None",rowperm,2,rowperm[0],&indx,&flg);CHKERRQ(ierr);
  if (flg) {lu->options.RowPerm = (rowperm_t)indx;}
  ierr = PetscOptionsTruth("-mat_superlu_replacetinypivot","ReplaceTinyPivot","None",PETSC_FALSE,&flg,0);CHKERRQ(ierr);
  if (flg) lu->options.ReplaceTinyPivot = YES;
  ierr = PetscOptionsTruth("-mat_superlu_printstat","PrintStat","None",PETSC_FALSE,&flg,0);CHKERRQ(ierr);
  if (flg) lu->options.PrintStat = YES;
  ierr = PetscOptionsInt("-mat_superlu_lwork","size of work array in bytes used by factorization","None",lu->lwork,&lu->lwork,PETSC_NULL);CHKERRQ(ierr);
  if (lu->lwork > 0 ){
    ierr = PetscMalloc(lu->lwork,&lu->work);CHKERRQ(ierr);
  } else if (lu->lwork != 0 && lu->lwork != -1){
    ierr = PetscPrintf(PETSC_COMM_SELF,"   Warning: lwork %D is not supported by SUPERLU. The default lwork=0 is used.\n",lu->lwork);
    lu->lwork = 0;
  }
  PetscOptionsEnd();

#ifdef SUPERLU2
  ierr = PetscObjectComposeFunctionDynamic((PetscObject)B,"MatCreateNull","MatCreateNull_SuperLU",
                                    (void(*)(void))MatCreateNull_SuperLU);CHKERRQ(ierr);
#endif

  /* Allocate spaces (notice sizes are for the transpose) */
  ierr = PetscMalloc(m*sizeof(PetscInt),&lu->etree);CHKERRQ(ierr);
  ierr = PetscMalloc(n*sizeof(PetscInt),&lu->perm_r);CHKERRQ(ierr);
  ierr = PetscMalloc(m*sizeof(PetscInt),&lu->perm_c);CHKERRQ(ierr);
  ierr = PetscMalloc(n*sizeof(PetscInt),&lu->R);CHKERRQ(ierr);
  ierr = PetscMalloc(m*sizeof(PetscInt),&lu->C);CHKERRQ(ierr);

  /* create rhs and solution x without allocate space for .Store */
#if defined(PETSC_USE_COMPLEX)
  zCreate_Dense_Matrix(&lu->B, m, 1, PETSC_NULL, m, SLU_DN, SLU_Z, SLU_GE);
  zCreate_Dense_Matrix(&lu->X, m, 1, PETSC_NULL, m, SLU_DN, SLU_Z, SLU_GE);
#else
  dCreate_Dense_Matrix(&lu->B, m, 1, PETSC_NULL, m, SLU_DN, SLU_D, SLU_GE);
  dCreate_Dense_Matrix(&lu->X, m, 1, PETSC_NULL, m, SLU_DN, SLU_D, SLU_GE);
#endif

  lu->flg            = DIFFERENT_NONZERO_PATTERN;
  lu->CleanUpSuperLU = PETSC_TRUE;

  *F = B;
  ierr = PetscLogObjectMemory(B,(A->rmap.n+A->cmap.n)*sizeof(PetscInt));CHKERRQ(ierr);
  PetscFunctionReturn(0);
}


#undef __FUNCT__
#define __FUNCT__ "MatDuplicate_SuperLU"
PetscErrorCode MatDuplicate_SuperLU(Mat A, MatDuplicateOption op, Mat *M) {
  PetscErrorCode ierr;
  Mat_SuperLU    *lu=(Mat_SuperLU *)A->spptr;

  PetscFunctionBegin;
  ierr = (*lu->MatDuplicate)(A,op,M);CHKERRQ(ierr);
  ierr = PetscMemcpy((*M)->spptr,lu,sizeof(Mat_SuperLU));CHKERRQ(ierr);
  PetscFunctionReturn(0);
}


/*MC
  MATSUPERLU - MATSUPERLU = "superlu" - A matrix type providing direct solvers (LU) for sequential matrices
  via the external package SuperLU.

  If SuperLU is installed (see the manual for
  instructions on how to declare the existence of external packages),
  a matrix type can be constructed which invokes SuperLU solvers.
  After calling MatCreate(...,A), simply call MatSetType(A,MATSUPERLU).

  This matrix inherits from MATSEQAIJ.  As a result, MatSeqAIJSetPreallocation() is
  supported for this matrix type.  One can also call MatConvert for an inplace conversion to or from
  the MATSEQAIJ type without data copy.

  Options Database Keys:
+ -mat_type superlu - sets the matrix type to "superlu" during a call to MatSetFromOptions()
. -mat_superlu_ordering <0,1,2,3> - 0: natural ordering,
                                    1: MMD applied to A'*A,
                                    2: MMD applied to A'+A,
                                    3: COLAMD, approximate minimum degree column ordering
. -mat_superlu_iterrefine - have SuperLU do iterative refinement after the triangular solve
                          choices: NOREFINE, SINGLE, DOUBLE, EXTRA; default is NOREFINE
- -mat_superlu_printstat - print SuperLU statistics about the factorization

   Level: beginner

.seealso: PCLU
M*/

/*
    Constructor for the new derived matrix class. It simply creates the base
   matrix class and then adds the additional information/methods needed by SuperLU.
*/
EXTERN_C_BEGIN
#undef __FUNCT__
#define __FUNCT__ "MatCreate_SuperLU"
PetscErrorCode PETSCMAT_DLLEXPORT MatCreate_SuperLU(Mat A)
{
  PetscErrorCode ierr;

  PetscFunctionBegin;
  ierr = MatSetType(A,MATSEQAIJ);CHKERRQ(ierr);
  ierr = MatConvert_SeqAIJ_SuperLU(A,MATSUPERLU,MAT_REUSE_MATRIX,&A);CHKERRQ(ierr);
  PetscFunctionReturn(0);
}
EXTERN_C_END