Actual source code: lgmresimpl.h

petsc-3.10.5 2019-03-28
Report Typos and Errors
  1: /* A. Baker */
  2: /*
  3:    Private data structure used by the LGMRES method.
  4: */


  9: #define KSPGMRES_NO_MACROS
 10:  #include <../src/ksp/ksp/impls/gmres/gmresimpl.h>

 12: typedef struct {
 13:   KSPGMRESHEADER

 15:   /* LGMRES_MOD - make these for the z vectors - new storage for lgmres */
 16:   Vec *augvecs;                      /* holds the error approximation vectors for lgmres. */
 17:   Vec **augvecs_user_work;           /* same purpose as user_work above, but this one is
 18:                                          for our error approx vectors */
 19:   /* currently only augvecs_user_work[0] is used, not sure if this will be */
 20:   /* extended in the future to use more, or if this is a design bug */
 21:   PetscInt aug_vv_allocated;           /* aug_vv_allocated is the number of allocated lgmres
 22:                                           augmentation vectors */
 23:   PetscInt aug_vecs_allocated;         /* aug_vecs_allocated is the total number of augmentation vecs
 24:                                           available - used to simplify the dynamic
 25:                                        allocation of vectors */
 26:   PetscScalar *hwork;                /* work array to hold Hessenberg product */

 28:   PetscInt augwork_alloc;            /*size of chunk allocated for augmentation vectors */

 30:   PetscInt aug_dim;                  /* max number of augmented directions to add */

 32:   PetscInt aug_ct;                   /* number of aug. vectors available */

 34:   PetscInt *aug_order;               /*keeps track of order to use aug. vectors*/

 36:   PetscBool approx_constant;        /* = 1 then the approx space at each restart will
 37:                                   be  size max_k .  Therefore, more than (max_k - aug_dim)
 38:                                   krylov vectors may be used if less than aug_dim error
 39:                                   approximations are available (in the first few restarts,
 40:                                   for example) to keep the space a constant size. */

 42:   PetscInt matvecs;                 /*keep track of matvecs */
 43: } KSP_LGMRES;


 46: #define HH(a,b)  (lgmres->hh_origin + (b)*(lgmres->max_k+2)+(a))
 47: /* HH will be size (max_k+2)*(max_k+1)  -  think of HH as
 48:    being stored columnwise (inc. zeros) for access purposes. */
 49: #define HES(a,b) (lgmres->hes_origin + (b)*(lgmres->max_k+1)+(a))
 50: /* HES will be size (max_k + 1) * (max_k + 1) -
 51:    again, think of HES as being stored columnwise */
 52: #define CC(a)    (lgmres->cc_origin + (a)) /* CC will be length (max_k+1) - cosines */
 53: #define SS(a)    (lgmres->ss_origin + (a)) /* SS will be length (max_k+1) - sines */
 54: #define GRS(a)    (lgmres->rs_origin + (a)) /* GRS will be length (max_k+2) - rt side */

 56: /* vector names */
 57: #define VEC_OFFSET     2
 58: #define VEC_TEMP       lgmres->vecs[0]               /* work space */
 59: #define VEC_TEMP_MATOP lgmres->vecs[1]               /* work space */
 60: #define VEC_VV(i)      lgmres->vecs[VEC_OFFSET+i]    /* use to access
 61:                                                         othog basis vectors */
 62: /*LGMRES_MOD */
 63: #define AUG_OFFSET     1
 64: #define AUGVEC(i)      lgmres->augvecs[AUG_OFFSET+i]   /*error approx vecors */
 65: #define AUG_ORDER(i)   lgmres->aug_order[i]            /*order in which to augment */
 66: #define A_AUGVEC(i)    lgmres->augvecs[AUG_OFFSET+i+lgmres->aug_dim] /*A times error vector */
 67: #define AUG_TEMP       lgmres->augvecs[0]              /* work vector */
 68: #endif