diff --git a/amgprec/amg_base_prec_type.F90 b/amgprec/amg_base_prec_type.F90 index dbe5fe5f..4ad8a4a0 100644 --- a/amgprec/amg_base_prec_type.F90 +++ b/amgprec/amg_base_prec_type.F90 @@ -1,15 +1,15 @@ -! -! +! +! ! AMG4PSBLAS version 1.0 ! Algebraic Multigrid Package ! based on PSBLAS (Parallel Sparse BLAS version 3.7) -! -! (C) Copyright 2020 -! -! Salvatore Filippone -! Pasqua D'Ambra -! Fabio Durastante -! +! +! (C) Copyright 2020 +! +! Salvatore Filippone +! Pasqua D'Ambra +! Fabio Durastante +! ! Redistribution and use in source and binary forms, with or without ! modification, are permitted provided that the following conditions ! are met: @@ -21,7 +21,7 @@ ! 3. The name of the AMG4PSBLAS group or the names of its contributors may ! not be used to endorse or promote products derived from this ! software without specific written permission. -! +! ! THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS ! ``AS IS'' AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED ! TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR @@ -33,8 +33,8 @@ ! CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ! ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE ! POSSIBILITY OF SUCH DAMAGE. -! -! +! +! ! File: amg_base_prec_type.F90 ! ! Module: amg_base_prec_type @@ -50,16 +50,16 @@ ! ! It contains routines for ! - converting character constants defining the preconditioner into integer -! constants; +! constants; ! - checking if the preconditioner is correctly defined; ! - printing a description of the preconditioner; -! - deallocating the preconditioner data structure. +! - deallocating the preconditioner data structure. ! module amg_base_prec_type ! - ! This reduces the size of .mod file. Without the ONLY clause compilation + ! This reduces the size of .mod file. Without the ONLY clause compilation ! blows up on some systems. ! use psb_const_mod @@ -78,7 +78,7 @@ module amg_base_prec_type & psb_err_from_subroutine_, psb_err_missing_override_method_, & & psb_error_handler, psb_out_unit, psb_err_unit - ! + ! ! Version numbers ! character(len=*), parameter :: amg_version_string_ = "1.0.0" @@ -120,7 +120,7 @@ module amg_base_prec_type procedure, pass(pm) :: printout => d_ml_parms_printout end type amg_dml_parms - + type amg_iaggr_data ! @@ -134,32 +134,32 @@ module amg_base_prec_type integer(psb_ipk_) :: min_coarse_size = -ione integer(psb_ipk_) :: min_coarse_size_per_process = -ione integer(psb_lpk_) :: target_coarse_size - ! 2. maximum number of levels. Defaults to 20 + ! 2. maximum number of levels. Defaults to 20 integer(psb_ipk_) :: max_levs = 20_psb_ipk_ end type amg_iaggr_data - + type, extends(amg_iaggr_data) :: amg_saggr_data - ! 3. min_cr_ratio = 1.5 + ! 3. min_cr_ratio = 1.5 real(psb_spk_) :: min_cr_ratio = 1.5_psb_spk_ real(psb_spk_) :: op_complexity = szero real(psb_spk_) :: avg_cr = szero end type amg_saggr_data type, extends(amg_iaggr_data) :: amg_daggr_data - ! 3. min_cr_ratio = 1.5 + ! 3. min_cr_ratio = 1.5 real(psb_dpk_) :: min_cr_ratio = 1.5_psb_dpk_ real(psb_dpk_) :: op_complexity = dzero real(psb_dpk_) :: avg_cr = dzero end type amg_daggr_data - + ! ! Entries in iprcparm ! ! These are in baseprec - ! - integer(psb_ipk_), parameter :: amg_smoother_type_ = 1 + ! + integer(psb_ipk_), parameter :: amg_smoother_type_ = 1 integer(psb_ipk_), parameter :: amg_sub_solve_ = 2 integer(psb_ipk_), parameter :: amg_sub_restr_ = 3 integer(psb_ipk_), parameter :: amg_sub_prol_ = 4 @@ -169,7 +169,7 @@ module amg_base_prec_type ! ! These are in onelev - ! + ! integer(psb_ipk_), parameter :: amg_ml_cycle_ = 20 integer(psb_ipk_), parameter :: amg_smoother_sweeps_pre_ = 21 integer(psb_ipk_), parameter :: amg_smoother_sweeps_post_ = 22 @@ -181,7 +181,7 @@ module amg_base_prec_type integer(psb_ipk_), parameter :: amg_aggr_eig_ = 28 integer(psb_ipk_), parameter :: amg_aggr_filter_ = 29 integer(psb_ipk_), parameter :: amg_coarse_mat_ = 30 - integer(psb_ipk_), parameter :: amg_coarse_solve_ = 31 + integer(psb_ipk_), parameter :: amg_coarse_solve_ = 31 integer(psb_ipk_), parameter :: amg_coarse_sweeps_ = 32 integer(psb_ipk_), parameter :: amg_coarse_fillin_ = 33 integer(psb_ipk_), parameter :: amg_coarse_subsolve_ = 34 @@ -196,7 +196,7 @@ module amg_base_prec_type ! ! Legal values for entry: amg_smoother_type_ - ! + ! integer(psb_ipk_), parameter :: amg_min_prec_ = 0 integer(psb_ipk_), parameter :: amg_noprec_ = 0 integer(psb_ipk_), parameter :: amg_base_smooth_ = 0 @@ -243,7 +243,7 @@ module amg_base_prec_type integer(psb_ipk_), parameter :: amg_ilu_scale_none_ = 0 integer(psb_ipk_), parameter :: amg_ilu_scale_maxval_ = 1 integer(psb_ipk_), parameter :: amg_ilu_scale_diag_ = 2 - integer(psb_ipk_), parameter :: amg_ilu_scale_arwsum_ = 3 + integer(psb_ipk_), parameter :: amg_ilu_scale_arwsum_ = 3 integer(psb_ipk_), parameter :: amg_ilu_scale_aclsum_ = 4 integer(psb_ipk_), parameter :: amg_ilu_scale_arcsum_ = 5 ! For the time being enable only maxval scale @@ -261,19 +261,21 @@ module amg_base_prec_type integer(psb_ipk_), parameter :: amg_new_ml_prec_ = 7 integer(psb_ipk_), parameter :: amg_mult_dev_ml_ = 7 integer(psb_ipk_), parameter :: amg_max_ml_cycle_ = 8 - ! + ! ! Legal values for entry: amg_par_aggr_alg_ ! integer(psb_ipk_), parameter :: amg_dec_aggr_ = 0 integer(psb_ipk_), parameter :: amg_sym_dec_aggr_ = 1 - integer(psb_ipk_), parameter :: amg_ext_aggr_ = 2 - integer(psb_ipk_), parameter :: amg_max_par_aggr_alg_ = amg_ext_aggr_ + integer(psb_ipk_), parameter :: amg_ext_aggr_ = 2 + integer(psb_ipk_), parameter :: amg_coupled_aggr_ = 3 + integer(psb_ipk_), parameter :: amg_max_par_aggr_alg_ = amg_coupled_aggr_ ! ! Legal values for entry: amg_aggr_type_ ! integer(psb_ipk_), parameter :: amg_noalg_ = 0 integer(psb_ipk_), parameter :: amg_soc1_ = 1 integer(psb_ipk_), parameter :: amg_soc2_ = 2 + integer(psb_ipk_), parameter :: amg_matchboxp_ = 3 ! ! Legal values for entry: amg_aggr_prol_ ! @@ -288,7 +290,7 @@ module amg_base_prec_type integer(psb_ipk_), parameter :: amg_no_filter_mat_ = 0 integer(psb_ipk_), parameter :: amg_filter_mat_ = 1 integer(psb_ipk_), parameter :: amg_max_filter_mat_ = amg_filter_mat_ - ! + ! ! Legal values for entry: amg_aggr_ord_ ! integer(psb_ipk_), parameter :: amg_aggr_ord_nat_ = 0 @@ -308,7 +310,7 @@ module amg_base_prec_type ! integer(psb_ipk_), parameter :: amg_distr_mat_ = 0 integer(psb_ipk_), parameter :: amg_repl_mat_ = 1 - integer(psb_ipk_), parameter :: amg_max_coarse_mat_ = amg_repl_mat_ + integer(psb_ipk_), parameter :: amg_max_coarse_mat_ = amg_repl_mat_ ! ! Legal values for entry: amg_prec_status_ ! @@ -338,7 +340,7 @@ module amg_base_prec_type ! ! Fields for sparse matrices ensembles stored in av() - ! + ! integer(psb_ipk_), parameter :: amg_l_pr_ = 1 integer(psb_ipk_), parameter :: amg_u_pr_ = 2 integer(psb_ipk_), parameter :: amg_bp_ilu_avsz_ = 2 @@ -347,7 +349,7 @@ module amg_base_prec_type integer(psb_ipk_), parameter :: amg_sm_pr_t_ = 5 integer(psb_ipk_), parameter :: amg_sm_pr_ = 6 integer(psb_ipk_), parameter :: amg_smth_avsz_ = 6 - integer(psb_ipk_), parameter :: amg_max_avsz_ = amg_smth_avsz_ + integer(psb_ipk_), parameter :: amg_max_avsz_ = amg_smth_avsz_ ! ! Character constants used by amg_file_prec_descr @@ -362,12 +364,13 @@ module amg_base_prec_type character(len=15), parameter, private :: & & matrix_names(0:1)=(/'distributed ','replicated '/) character(len=18), parameter, private :: & - & aggr_type_names(0:2)=(/'None ',& - & 'SOC measure 1 ', 'SOC Measure 2 '/) + & aggr_type_names(0:3)=(/'None ',& + & 'SOC measure 1 ', 'SOC Measure 2 ',& + & 'Parallel Matching '/) character(len=18), parameter, private :: & - & par_aggr_alg_names(0:2)=(/& + & par_aggr_alg_names(0:3)=(/& & 'decoupled aggr. ', 'sym. dec. aggr. ',& - & 'user defined aggr.'/) + & 'user defined aggr.', 'coupled aggr. '/) character(len=18), parameter, private :: & & ord_names(0:1)=(/'Natural ordering ','Desc. degree ord. '/) character(len=6), parameter, private :: & @@ -395,7 +398,7 @@ module amg_base_prec_type module procedure amg_icheck_def, amg_scheck_def, amg_dcheck_def end interface - interface psb_bcast + interface psb_bcast module procedure amg_ml_bcast, amg_sml_bcast, amg_dml_bcast end interface psb_bcast @@ -408,9 +411,9 @@ module amg_base_prec_type ! Will need a more sophisticated strategy. ! logical, private, save :: do_remap=.false. - + contains - + function amg_get_do_remap() result(res) implicit none logical :: res @@ -424,7 +427,7 @@ contains do_remap = val end subroutine amg_set_do_remap - + ! ! Function: amg_stringval ! @@ -439,10 +442,10 @@ contains ! function amg_stringval(string) result(val) use psb_prec_const_mod - implicit none + implicit none ! Arguments character(len=*), intent(in) :: string - integer(psb_ipk_) :: val + integer(psb_ipk_) :: val character(len=*), parameter :: name='amg_stringval' ! Local variable integer :: index_tab @@ -450,14 +453,14 @@ contains index_tab=index(string,char(9)) if (index_tab.NE.0) then string2=string(1:index_tab-1) - else + else string2=string endif select case(psb_toupper(trim(string2))) case('NONE') val = 0 case('HALO') - val = psb_halo_ + val = psb_halo_ case('SUM') val = psb_sum_ case('AVG') @@ -553,56 +556,56 @@ contains case('OUTER_SWEEPS') val = amg_outer_sweeps_ case('LOCAL_SOLVER') - val = amg_local_solver_ + val = amg_local_solver_ case('GLOBAL_SOLVER') - val = amg_global_solver_ + val = amg_global_solver_ case default val = -1 end select end function amg_stringval subroutine ml_parms_get_coarse(pm,pmin) - implicit none + implicit none class(amg_ml_parms), intent(inout) :: pm class(amg_ml_parms), intent(in) :: pmin pm%coarse_mat = pmin%coarse_mat pm%coarse_solve = pmin%coarse_solve end subroutine ml_parms_get_coarse - - - + + + subroutine ml_parms_printout(pm,iout) - implicit none + implicit none class(amg_ml_parms), intent(in) :: pm integer(psb_ipk_), intent(in) :: iout - + write(iout,*) 'ML : ',pm%ml_cycle write(iout,*) 'Sweeps: ',pm%sweeps_pre,pm%sweeps_post write(iout,*) 'AGGR : ',pm%par_aggr_alg,pm%aggr_prol, pm%aggr_ord write(iout,*) ' : ',pm%aggr_omega_alg,pm%aggr_eig,pm%aggr_filter write(iout,*) 'COARSE: ',pm%coarse_mat,pm%coarse_solve end subroutine ml_parms_printout - - + + subroutine s_ml_parms_printout(pm,iout) - implicit none + implicit none class(amg_sml_parms), intent(in) :: pm integer(psb_ipk_), intent(in) :: iout - + call pm%amg_ml_parms%printout(iout) write(iout,*) 'REAL : ',pm%aggr_omega_val,pm%aggr_thresh end subroutine s_ml_parms_printout - - + + subroutine d_ml_parms_printout(pm,iout) - implicit none + implicit none class(amg_dml_parms), intent(in) :: pm integer(psb_ipk_), intent(in) :: iout - + call pm%amg_ml_parms%printout(iout) write(iout,*) 'REAL : ',pm%aggr_omega_val,pm%aggr_thresh end subroutine d_ml_parms_printout - + ! ! Routines printing out a description of the preconditioner @@ -618,7 +621,7 @@ contains info = psb_success_ if ((pm%ml_cycle>=amg_no_ml_).and.(pm%ml_cycle<=amg_max_ml_cycle_)) then - + write(iout,*) ' Multilevel cycle: ',& & ml_names(pm%ml_cycle) select case (pm%ml_cycle) @@ -644,7 +647,7 @@ contains info = psb_success_ if ((pm%ml_cycle>=amg_no_ml_).and.(pm%ml_cycle<=amg_max_ml_cycle_)) then - + write(iout,*) ' Parallel aggregation algorithm: ',& & par_aggr_alg_names(pm%par_aggr_alg) if (pm%aggr_type>0) write(iout,*) ' Aggregation type: ',& @@ -656,23 +659,23 @@ contains write(iout,*) ' Aggregation prolongator: ', & & aggr_prols(pm%aggr_prol) if (pm%aggr_prol /= amg_no_smooth_) then - write(iout,*) ' with: ', aggr_filters(pm%aggr_filter) - if (pm%aggr_omega_alg == amg_eig_est_) then + write(iout,*) ' with: ', aggr_filters(pm%aggr_filter) + if (pm%aggr_omega_alg == amg_eig_est_) then write(iout,*) ' Damping omega computation: spectral radius estimate' write(iout,*) ' Spectral radius estimate: ', & & eigen_estimates(pm%aggr_eig) - else if (pm%aggr_omega_alg == amg_user_choice_) then + else if (pm%aggr_omega_alg == amg_user_choice_) then write(iout,*) ' Damping omega computation: user defined value.' - else + else write(iout,*) ' Damping omega computation: unknown value in iprcparm!!' end if end if !end if else write(iout,*) ' Multilevel type: Unkonwn value. Something is amiss....',& - & pm%ml_cycle + & pm%ml_cycle end if - + return end subroutine ml_parms_mldescr @@ -689,13 +692,13 @@ contains logical :: coarse_ info = psb_success_ - if (present(coarse)) then + if (present(coarse)) then coarse_ = coarse else coarse_ = .false. end if - if (coarse_) then + if (coarse_) then call pm%coarsedescr(iout,info) end if @@ -718,12 +721,12 @@ contains write(iout,*) ' Coarse matrix: ',& & matrix_names(pm%coarse_mat) select case(pm%coarse_solve) - case (amg_bjac_,amg_as_) + case (amg_bjac_,amg_as_) write(iout,*) ' Number of sweeps : ',& & pm%sweeps_pre write(iout,*) ' Coarse solver: ',& & 'Block Jacobi' - case (amg_l1_bjac_) + case (amg_l1_bjac_) write(iout,*) ' Number of sweeps : ',& & pm%sweeps_pre write(iout,*) ' Coarse solver: ',& @@ -790,7 +793,7 @@ contains ! function is_legal_base_prec(ip) - implicit none + implicit none integer(psb_ipk_), intent(in) :: ip logical :: is_legal_base_prec @@ -798,60 +801,68 @@ contains return end function is_legal_base_prec function is_int_non_negative(ip) - implicit none + implicit none integer(psb_ipk_), intent(in) :: ip logical :: is_int_non_negative - is_int_non_negative = (ip >= 0) + is_int_non_negative = (ip >= 0) return end function is_int_non_negative function is_legal_ilu_scale(ip) - implicit none + implicit none integer(psb_ipk_), intent(in) :: ip logical :: is_legal_ilu_scale is_legal_ilu_scale = ((ip >= amg_ilu_scale_none_).and.(ip <= amg_max_ilu_scale_)) return end function is_legal_ilu_scale function is_int_positive(ip) - implicit none + implicit none integer(psb_ipk_), intent(in) :: ip logical :: is_int_positive - is_int_positive = (ip >= 1) + is_int_positive = (ip >= 1) return end function is_int_positive function is_legal_prolong(ip) - implicit none + implicit none integer(psb_ipk_), intent(in) :: ip logical :: is_legal_prolong is_legal_prolong = ((ip>=psb_none_).and.(ip<=psb_square_root_)) return end function is_legal_prolong function is_legal_restrict(ip) - implicit none + implicit none integer(psb_ipk_), intent(in) :: ip logical :: is_legal_restrict is_legal_restrict = ((ip == psb_nohalo_).or.(ip==psb_halo_)) return end function is_legal_restrict function is_legal_ml_cycle(ip) - implicit none + implicit none integer(psb_ipk_), intent(in) :: ip logical :: is_legal_ml_cycle is_legal_ml_cycle = ((ip>=amg_no_ml_).and.(ip<=amg_max_ml_cycle_)) return end function is_legal_ml_cycle - function is_legal_ml_par_aggr_alg(ip) - implicit none + function is_legal_coupled_par_aggr_alg(ip) + implicit none integer(psb_ipk_), intent(in) :: ip - logical :: is_legal_ml_par_aggr_alg + logical :: is_legal_coupled_par_aggr_alg - is_legal_ml_par_aggr_alg = ((ip>=amg_dec_aggr_).and.(ip<=amg_max_par_aggr_alg_)) + is_legal_coupled_par_aggr_alg = (ip == amg_coupled_aggr_) return - end function is_legal_ml_par_aggr_alg + end function is_legal_coupled_par_aggr_alg + function is_legal_decoupled_par_aggr_alg(ip) + implicit none + integer(psb_ipk_), intent(in) :: ip + logical :: is_legal_decoupled_par_aggr_alg + + is_legal_decoupled_par_aggr_alg = ((ip>=amg_dec_aggr_).and.(ip<=amg_max_par_aggr_alg_)) + return + end function is_legal_decoupled_par_aggr_alg function is_legal_ml_aggr_type(ip) - implicit none + implicit none integer(psb_ipk_), intent(in) :: ip logical :: is_legal_ml_aggr_type @@ -859,7 +870,7 @@ contains return end function is_legal_ml_aggr_type function is_legal_ml_aggr_ord(ip) - implicit none + implicit none integer(psb_ipk_), intent(in) :: ip logical :: is_legal_ml_aggr_ord @@ -867,7 +878,7 @@ contains return end function is_legal_ml_aggr_ord function is_legal_ml_aggr_omega_alg(ip) - implicit none + implicit none integer(psb_ipk_), intent(in) :: ip logical :: is_legal_ml_aggr_omega_alg @@ -875,7 +886,7 @@ contains return end function is_legal_ml_aggr_omega_alg function is_legal_ml_aggr_eig(ip) - implicit none + implicit none integer(psb_ipk_), intent(in) :: ip logical :: is_legal_ml_aggr_eig @@ -883,7 +894,7 @@ contains return end function is_legal_ml_aggr_eig function is_legal_ml_aggr_prol(ip) - implicit none + implicit none integer(psb_ipk_), intent(in) :: ip logical :: is_legal_ml_aggr_prol @@ -891,7 +902,7 @@ contains return end function is_legal_ml_aggr_prol function is_legal_ml_coarse_mat(ip) - implicit none + implicit none integer(psb_ipk_), intent(in) :: ip logical :: is_legal_ml_coarse_mat @@ -899,7 +910,7 @@ contains return end function is_legal_ml_coarse_mat function is_legal_aggr_filter(ip) - implicit none + implicit none integer(psb_ipk_), intent(in) :: ip logical :: is_legal_aggr_filter @@ -907,7 +918,7 @@ contains return end function is_legal_aggr_filter function is_distr_ml_coarse_mat(ip) - implicit none + implicit none integer(psb_ipk_), intent(in) :: ip logical :: is_distr_ml_coarse_mat @@ -915,7 +926,7 @@ contains return end function is_distr_ml_coarse_mat function is_legal_ml_fact(ip) - implicit none + implicit none integer(psb_ipk_), intent(in) :: ip logical :: is_legal_ml_fact ! Here the minimum is really 1, amg_fact_none_ is not acceptable. @@ -925,7 +936,7 @@ contains end function is_legal_ml_fact function is_legal_ilu_fact(ip) use psb_prec_const_mod - implicit none + implicit none integer(psb_ipk_), intent(in) :: ip logical :: is_legal_ilu_fact @@ -934,14 +945,14 @@ contains return end function is_legal_ilu_fact function is_legal_d_omega(ip) - implicit none + implicit none real(psb_dpk_), intent(in) :: ip logical :: is_legal_d_omega is_legal_d_omega = ((ip>=0.0d0).and.(ip<=2.0d0)) return end function is_legal_d_omega function is_legal_d_fact_thrs(ip) - implicit none + implicit none real(psb_dpk_), intent(in) :: ip logical :: is_legal_d_fact_thrs @@ -949,7 +960,7 @@ contains return end function is_legal_d_fact_thrs function is_legal_d_aggr_thrs(ip) - implicit none + implicit none real(psb_dpk_), intent(in) :: ip logical :: is_legal_d_aggr_thrs @@ -958,14 +969,14 @@ contains end function is_legal_d_aggr_thrs function is_legal_s_omega(ip) - implicit none + implicit none real(psb_spk_), intent(in) :: ip logical :: is_legal_s_omega is_legal_s_omega = ((ip>=0.0).and.(ip<=2.0)) return end function is_legal_s_omega function is_legal_s_fact_thrs(ip) - implicit none + implicit none real(psb_spk_), intent(in) :: ip logical :: is_legal_s_fact_thrs @@ -973,7 +984,7 @@ contains return end function is_legal_s_fact_thrs function is_legal_s_aggr_thrs(ip) - implicit none + implicit none real(psb_spk_), intent(in) :: ip logical :: is_legal_s_aggr_thrs @@ -983,11 +994,11 @@ contains subroutine amg_icheck_def(ip,name,id,is_legal) - implicit none + implicit none integer(psb_ipk_), intent(inout) :: ip integer(psb_ipk_), intent(in) :: id character(len=*), intent(in) :: name - interface + interface function is_legal(i) import :: psb_ipk_ integer(psb_ipk_), intent(in) :: i @@ -996,7 +1007,7 @@ contains end interface character(len=20), parameter :: rname='amg_check_def' - if (.not.is_legal(ip)) then + if (.not.is_legal(ip)) then write(0,*)trim(rname),': Error: Illegal value for ',& & name,' :',ip, '. defaulting to ',id ip = id @@ -1004,11 +1015,11 @@ contains end subroutine amg_icheck_def subroutine amg_scheck_def(ip,name,id,is_legal) - implicit none + implicit none real(psb_spk_), intent(inout) :: ip real(psb_spk_), intent(in) :: id character(len=*), intent(in) :: name - interface + interface function is_legal(i) use psb_base_mod, only : psb_spk_ real(psb_spk_), intent(in) :: i @@ -1017,7 +1028,7 @@ contains end interface character(len=20), parameter :: rname='amg_check_def' - if (.not.is_legal(ip)) then + if (.not.is_legal(ip)) then write(0,*)trim(rname),': Error: Illegal value for ',& & name,' :',ip, '. defaulting to ',id ip = id @@ -1025,11 +1036,11 @@ contains end subroutine amg_scheck_def subroutine amg_dcheck_def(ip,name,id,is_legal) - implicit none + implicit none real(psb_dpk_), intent(inout) :: ip real(psb_dpk_), intent(in) :: id character(len=*), intent(in) :: name - interface + interface function is_legal(i) use psb_base_mod, only : psb_dpk_ real(psb_dpk_), intent(in) :: i @@ -1038,7 +1049,7 @@ contains end interface character(len=20), parameter :: rname='amg_check_def' - if (.not.is_legal(ip)) then + if (.not.is_legal(ip)) then write(0,*)trim(rname),': Error: Illegal value for ',& & name,' :',ip, '. defaulting to ',id ip = id @@ -1047,7 +1058,7 @@ contains function pr_to_str(iprec) - implicit none + implicit none integer(psb_ipk_), intent(in) :: iprec character(len=10) :: pr_to_str @@ -1055,11 +1066,11 @@ contains select case(iprec) case(amg_noprec_) pr_to_str='NOPREC' - case(amg_jac_) + case(amg_jac_) pr_to_str='JAC' - case(amg_bjac_) + case(amg_bjac_) pr_to_str='BJAC' - case(amg_as_) + case(amg_as_) pr_to_str='AS' end select @@ -1067,7 +1078,7 @@ contains subroutine amg_ml_bcast(ctxt,dat,root) - implicit none + implicit none type(psb_ctxt_type), intent(in) :: ctxt type(amg_ml_parms), intent(inout) :: dat integer(psb_ipk_), intent(in), optional :: root @@ -1089,7 +1100,7 @@ contains subroutine amg_sml_bcast(ctxt,dat,root) - implicit none + implicit none type(psb_ctxt_type), intent(in) :: ctxt type(amg_sml_parms), intent(inout) :: dat integer(psb_ipk_), intent(in), optional :: root @@ -1100,7 +1111,7 @@ contains end subroutine amg_sml_bcast subroutine amg_dml_bcast(ctxt,dat,root) - implicit none + implicit none type(psb_ctxt_type), intent(in) :: ctxt type(amg_dml_parms), intent(inout) :: dat integer(psb_ipk_), intent(in), optional :: root @@ -1112,7 +1123,7 @@ contains subroutine ml_parms_clone(pm,pmout,info) - implicit none + implicit none class(amg_ml_parms), intent(inout) :: pm class(amg_ml_parms), intent(out) :: pmout integer(psb_ipk_), intent(out) :: info @@ -1132,19 +1143,19 @@ contains pmout%coarse_solve = pm%coarse_solve end subroutine ml_parms_clone - + subroutine s_ml_parms_clone(pm,pmout,info) - implicit none + implicit none class(amg_sml_parms), intent(inout) :: pm class(amg_ml_parms), intent(out) :: pmout integer(psb_ipk_), intent(out) :: info - + integer(psb_ipk_) :: err_act integer(psb_ipk_) :: ierr(5) character(len=20) :: name='clone' - + info = 0 select type(pout => pmout) class is (amg_sml_parms) @@ -1159,21 +1170,21 @@ contains call psb_get_erraction(err_act) call psb_error_handler(err_act) end select - + end subroutine s_ml_parms_clone subroutine d_ml_parms_clone(pm,pmout,info) - implicit none + implicit none class(amg_dml_parms), intent(inout) :: pm class(amg_ml_parms), intent(out) :: pmout integer(psb_ipk_), intent(out) :: info - + integer(psb_ipk_) :: err_act integer(psb_ipk_) :: ierr(5) character(len=20) :: name='clone' - + info = 0 select type(pout => pmout) class is (amg_dml_parms) @@ -1189,13 +1200,13 @@ contains call psb_error_handler(err_act) return end select - + end subroutine d_ml_parms_clone function amg_s_equal_aggregation(parms1, parms2) result(val) type(amg_sml_parms), intent(in) :: parms1, parms2 logical :: val - + val = (parms1%par_aggr_alg == parms2%par_aggr_alg ) .and. & & (parms1%aggr_type == parms2%aggr_type ) .and. & & (parms1%aggr_ord == parms2%aggr_ord ) .and. & @@ -1210,7 +1221,7 @@ contains function amg_d_equal_aggregation(parms1, parms2) result(val) type(amg_dml_parms), intent(in) :: parms1, parms2 logical :: val - + val = (parms1%par_aggr_alg == parms2%par_aggr_alg ) .and. & & (parms1%aggr_type == parms2%aggr_type ) .and. & & (parms1%aggr_ord == parms2%aggr_ord ) .and. & @@ -1221,5 +1232,5 @@ contains & (parms1%aggr_omega_val == parms2%aggr_omega_val ) .and. & & (parms1%aggr_thresh == parms2%aggr_thresh ) end function amg_d_equal_aggregation - + end module amg_base_prec_type diff --git a/amgprec/amg_d_matchboxp_mod.f90 b/amgprec/amg_d_matchboxp_mod.f90 index 2578be73..3be08ca7 100644 --- a/amgprec/amg_d_matchboxp_mod.f90 +++ b/amgprec/amg_d_matchboxp_mod.f90 @@ -80,7 +80,7 @@ module dmatchboxp_mod subroutine dMatchBoxPC(nlver,nledge,verlocptr,verlocind,edgelocweight,& & verdistance, mate, myrank, numprocs, icomm,& & msgindsent,msgactualsent,msgpercent,& - & ph0_time, ph1_time, ph2_time, ph1_card, ph2_card) bind(c,name='MatchBoxPC') + & ph0_time, ph1_time, ph2_time, ph1_card, ph2_card) bind(c,name='dMatchBoxPC') use iso_c_binding import :: psb_c_ipk_, psb_c_lpk_, psb_c_mpk_, psb_c_epk_ implicit none diff --git a/amgprec/amg_s_matchboxp_mod.f90 b/amgprec/amg_s_matchboxp_mod.f90 index 123061f1..4f8d8181 100644 --- a/amgprec/amg_s_matchboxp_mod.f90 +++ b/amgprec/amg_s_matchboxp_mod.f90 @@ -80,7 +80,7 @@ module smatchboxp_mod subroutine sMatchBoxPC(nlver,nledge,verlocptr,verlocind,edgelocweight,& & verdistance, mate, myrank, numprocs, icomm,& & msgindsent,msgactualsent,msgpercent,& - & ph0_time, ph1_time, ph2_time, ph1_card, ph2_card) bind(c,name='MatchBoxPC') + & ph0_time, ph1_time, ph2_time, ph1_card, ph2_card) bind(c,name='sMatchBoxPC') use iso_c_binding import :: psb_c_ipk_, psb_c_lpk_, psb_c_mpk_, psb_c_epk_ implicit none diff --git a/amgprec/impl/aggregator/MatchBoxPC.h b/amgprec/impl/aggregator/MatchBoxPC.h index a353a486..090e0265 100644 --- a/amgprec/impl/aggregator/MatchBoxPC.h +++ b/amgprec/impl/aggregator/MatchBoxPC.h @@ -172,6 +172,24 @@ MilanLongInt* msgIndSent, MilanLongInt* msgActualSent, MilanReal* msgPercent, MilanReal* ph0_time, MilanReal* ph1_time, MilanReal* ph2_time, MilanLongInt* ph1_card, MilanLongInt* ph2_card ); +void dMatchBoxPC(MilanLongInt NLVer, MilanLongInt NLEdge, + MilanLongInt* verLocPtr, MilanLongInt* verLocInd, MilanReal* edgeLocWeight, + MilanLongInt* verDistance, + MilanLongInt* Mate, + MilanInt myRank, MilanInt numProcs, MilanInt icomm, + MilanLongInt* msgIndSent, MilanLongInt* msgActualSent, MilanReal* msgPercent, + MilanReal* ph0_time, MilanReal* ph1_time, MilanReal* ph2_time, + MilanLongInt* ph1_card, MilanLongInt* ph2_card ); + +void sMatchBoxPC(MilanLongInt NLVer, MilanLongInt NLEdge, + MilanLongInt* verLocPtr, MilanLongInt* verLocInd, MilanFloat* edgeLocWeight, + MilanLongInt* verDistance, + MilanLongInt* Mate, + MilanInt myRank, MilanInt numProcs, MilanInt icomm, + MilanLongInt* msgIndSent, MilanLongInt* msgActualSent, MilanReal* msgPercent, + MilanReal* ph0_time, MilanReal* ph1_time, MilanReal* ph2_time, + MilanLongInt* ph1_card, MilanLongInt* ph2_card ); + #ifdef __cplusplus } #endif diff --git a/amgprec/impl/aggregator/amg_c_dec_aggregator_tprol.f90 b/amgprec/impl/aggregator/amg_c_dec_aggregator_tprol.f90 index 229a8e57..cb8fb6a7 100644 --- a/amgprec/impl/aggregator/amg_c_dec_aggregator_tprol.f90 +++ b/amgprec/impl/aggregator/amg_c_dec_aggregator_tprol.f90 @@ -1,15 +1,15 @@ -! -! +! +! ! AMG4PSBLAS version 1.0 ! Algebraic Multigrid Package ! based on PSBLAS (Parallel Sparse BLAS version 3.7) -! -! (C) Copyright 2021 -! -! Salvatore Filippone -! Pasqua D'Ambra -! Fabio Durastante -! +! +! (C) Copyright 2021 +! +! Salvatore Filippone +! Pasqua D'Ambra +! Fabio Durastante +! ! Redistribution and use in source and binary forms, with or without ! modification, are permitted provided that the following conditions ! are met: @@ -21,7 +21,7 @@ ! 3. The name of the AMG4PSBLAS group or the names of its contributors may ! not be used to endorse or promote products derived from this ! software without specific written permission. -! +! ! THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS ! ``AS IS'' AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED ! TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR @@ -33,26 +33,26 @@ ! CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ! ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE ! POSSIBILITY OF SUCH DAMAGE. -! -! +! +! ! File: amg_c_dec_aggregator_tprol.f90 ! ! Subroutine: amg_c_dec_aggregator_tprol ! Version: complex ! -! This routine is mainly an interface to soc_map_bld where the real work is performed. +! This routine is mainly an interface to soc_map_bld where the real work is performed. ! It takes care of some consistency checking, and calls map_to_tprol, which is ! refactored and shared among all the aggregation methods that produce a simple ! integer mapping. ! -! +! ! Arguments: ! ag - type(amg_c_dec_aggregator_type), input/output. ! The aggregator object, carrying with itself the mapping algorithm. ! parms - The auxiliary parameters object ! ag_data - Auxiliary global aggregation parameters object -! -! +! +! ! a - type(psb_cspmat_type). ! The sparse matrix structure containing the local part of the ! fine-level matrix. @@ -70,10 +70,10 @@ ! nlaggr(i) contains the aggregates held by process i. ! t_prol - type(psb_cspmat_type), output ! The tentative prolongator, based on ilaggr. -! +! ! info - integer, output. -! Error code. -! +! Error code. +! subroutine amg_c_dec_aggregator_build_tprol(ag,parms,ag_data,& & a,desc_a,ilaggr,nlaggr,t_prol,info) use psb_base_mod @@ -81,7 +81,7 @@ subroutine amg_c_dec_aggregator_build_tprol(ag,parms,ag_data,& use amg_c_inner_mod implicit none class(amg_c_dec_aggregator_type), target, intent(inout) :: ag - type(amg_sml_parms), intent(inout) :: parms + type(amg_sml_parms), intent(inout) :: parms type(amg_saggr_data), intent(in) :: ag_data type(psb_cspmat_type), intent(in) :: a type(psb_desc_type), intent(in) :: desc_a @@ -112,7 +112,7 @@ subroutine amg_c_dec_aggregator_build_tprol(ag,parms,ag_data,& call amg_check_def(parms%ml_cycle,'Multilevel cycle',& & amg_mult_ml_,is_legal_ml_cycle) call amg_check_def(parms%par_aggr_alg,'Aggregation',& - & amg_dec_aggr_,is_legal_ml_par_aggr_alg) + & amg_dec_aggr_,is_legal_decoupled_par_aggr_alg) call amg_check_def(parms%aggr_ord,'Ordering',& & amg_aggr_ord_nat_,is_legal_ml_aggr_ord) call amg_check_def(parms%aggr_thresh,'Aggr_Thresh',szero,is_legal_s_aggr_thrs) @@ -130,11 +130,11 @@ subroutine amg_c_dec_aggregator_build_tprol(ag,parms,ag_data,& call psb_errpush(info,name,a_err='soc_map_bld/map_to_tprol') goto 9999 endif - + call psb_erractionrestore(err_act) return 9999 call psb_error_handler(err_act) return - + end subroutine amg_c_dec_aggregator_build_tprol diff --git a/amgprec/impl/aggregator/amg_c_symdec_aggregator_tprol.f90 b/amgprec/impl/aggregator/amg_c_symdec_aggregator_tprol.f90 index 2b32cb57..9974a233 100644 --- a/amgprec/impl/aggregator/amg_c_symdec_aggregator_tprol.f90 +++ b/amgprec/impl/aggregator/amg_c_symdec_aggregator_tprol.f90 @@ -1,15 +1,15 @@ -! -! +! is_legal_decoupled_par_aggr_alg +! ! AMG4PSBLAS version 1.0 ! Algebraic Multigrid Package ! based on PSBLAS (Parallel Sparse BLAS version 3.7) -! -! (C) Copyright 2021 -! -! Salvatore Filippone -! Pasqua D'Ambra -! Fabio Durastante -! +! +! (C) Copyright 2021 +! +! Salvatore Filippone +! Pasqua D'Ambra +! Fabio Durastante +! ! Redistribution and use in source and binary forms, with or without ! modification, are permitted provided that the following conditions ! are met: @@ -21,7 +21,7 @@ ! 3. The name of the AMG4PSBLAS group or the names of its contributors may ! not be used to endorse or promote products derived from this ! software without specific written permission. -! +! ! THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS ! ``AS IS'' AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED ! TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR @@ -33,28 +33,28 @@ ! CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ! ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE ! POSSIBILITY OF SUCH DAMAGE. -! -! +! +! ! File: amg_c_symdec_aggregator_tprol.f90 ! ! Subroutine: amg_c_symdec_aggregator_tprol ! Version: complex ! ! -! This routine is mainly an interface to soc_map_bld where the real work is performed. +! This routine is mainly an interface to soc_map_bld where the real work is performed. ! It takes care of some consistency checking, and calls map_to_tprol, which is ! refactored and shared among all the aggregation methods that produce a simple -! integer mapping. It also symmetrizes the pattern of the local matrix A. +! integer mapping. It also symmetrizes the pattern of the local matrix A. +! ! ! -! ! Arguments: ! Arguments: ! ag - type(amg_c_dec_aggregator_type), input/output. ! The aggregator object, carrying with itself the mapping algorithm. ! parms - The auxiliary parameters object ! ag_data - Auxiliary global aggregation parameters object -! +! ! a - type(psb_cspmat_type). ! The sparse matrix structure containing the local part of the ! fine-level matrix. @@ -72,10 +72,10 @@ ! nlaggr(i) contains the aggregates held by process i. ! op_prol - type(psb_cspmat_type), output ! The tentative prolongator, based on ilaggr. -! +! ! info - integer, output. -! Error code. -! +! Error code. +! subroutine amg_c_symdec_aggregator_build_tprol(ag,parms,ag_data,& & a,desc_a,ilaggr,nlaggr,op_prol,info) use psb_base_mod @@ -84,7 +84,7 @@ subroutine amg_c_symdec_aggregator_build_tprol(ag,parms,ag_data,& use amg_c_inner_mod implicit none class(amg_c_symdec_aggregator_type), target, intent(inout) :: ag - type(amg_sml_parms), intent(inout) :: parms + type(amg_sml_parms), intent(inout) :: parms type(amg_saggr_data), intent(in) :: ag_data type(psb_cspmat_type), intent(in) :: a type(psb_desc_type), intent(in) :: desc_a @@ -117,7 +117,7 @@ subroutine amg_c_symdec_aggregator_build_tprol(ag,parms,ag_data,& call amg_check_def(parms%ml_cycle,'Multilevel cycle',& & amg_mult_ml_,is_legal_ml_cycle) call amg_check_def(parms%par_aggr_alg,'Aggregation',& - & amg_dec_aggr_,is_legal_ml_par_aggr_alg) + & amg_dec_aggr_,is_legal_decoupled_par_aggr_alg) call amg_check_def(parms%aggr_ord,'Ordering',& & amg_aggr_ord_nat_,is_legal_ml_aggr_ord) call amg_check_def(parms%aggr_thresh,'Aggr_Thresh',szero,is_legal_s_aggr_thrs) @@ -129,7 +129,7 @@ subroutine amg_c_symdec_aggregator_build_tprol(ag,parms,ag_data,& call atmp%set_ncols(nr) if (info == psb_success_) call atmp%transp(atrans) if (info == psb_success_) call atrans%cscnv(info,type='COO') - if (info == psb_success_) call psb_rwextd(nr,atmp,info,b=atrans,rowscale=.false.) + if (info == psb_success_) call psb_rwextd(nr,atmp,info,b=atrans,rowscale=.false.) call atmp%set_nrows(nr) call atmp%set_ncols(nr) if (info == psb_success_) call atrans%free() @@ -145,7 +145,7 @@ subroutine amg_c_symdec_aggregator_build_tprol(ag,parms,ag_data,& & desc_a,nlaggr,ilaggr,info) if (info == psb_success_) call atmp%free() - if (info == psb_success_) call amg_map_to_tprol(desc_a,ilaggr,nlaggr,op_prol,info) + if (info == psb_success_) call amg_map_to_tprol(desc_a,ilaggr,nlaggr,op_prol,info) if (info /= psb_success_) then info=psb_err_from_subroutine_ call psb_errpush(info,name,a_err='soc_map_bld/map_to_tprol') diff --git a/amgprec/impl/aggregator/amg_d_dec_aggregator_tprol.f90 b/amgprec/impl/aggregator/amg_d_dec_aggregator_tprol.f90 index 8ef77da8..e3d1e73c 100644 --- a/amgprec/impl/aggregator/amg_d_dec_aggregator_tprol.f90 +++ b/amgprec/impl/aggregator/amg_d_dec_aggregator_tprol.f90 @@ -1,15 +1,15 @@ -! -! +! +! ! AMG4PSBLAS version 1.0 ! Algebraic Multigrid Package ! based on PSBLAS (Parallel Sparse BLAS version 3.7) -! -! (C) Copyright 2021 -! -! Salvatore Filippone -! Pasqua D'Ambra -! Fabio Durastante -! +! +! (C) Copyright 2021 +! +! Salvatore Filippone +! Pasqua D'Ambra +! Fabio Durastante +! ! Redistribution and use in source and binary forms, with or without ! modification, are permitted provided that the following conditions ! are met: @@ -21,7 +21,7 @@ ! 3. The name of the AMG4PSBLAS group or the names of its contributors may ! not be used to endorse or promote products derived from this ! software without specific written permission. -! +! ! THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS ! ``AS IS'' AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED ! TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR @@ -33,26 +33,26 @@ ! CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ! ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE ! POSSIBILITY OF SUCH DAMAGE. -! -! +! +! ! File: amg_d_dec_aggregator_tprol.f90 ! ! Subroutine: amg_d_dec_aggregator_tprol ! Version: real ! -! This routine is mainly an interface to soc_map_bld where the real work is performed. +! This routine is mainly an interface to soc_map_bld where the real work is performed. ! It takes care of some consistency checking, and calls map_to_tprol, which is ! refactored and shared among all the aggregation methods that produce a simple ! integer mapping. ! -! +! ! Arguments: ! ag - type(amg_d_dec_aggregator_type), input/output. ! The aggregator object, carrying with itself the mapping algorithm. ! parms - The auxiliary parameters object ! ag_data - Auxiliary global aggregation parameters object -! -! +! +! ! a - type(psb_dspmat_type). ! The sparse matrix structure containing the local part of the ! fine-level matrix. @@ -70,10 +70,10 @@ ! nlaggr(i) contains the aggregates held by process i. ! t_prol - type(psb_dspmat_type), output ! The tentative prolongator, based on ilaggr. -! +! ! info - integer, output. -! Error code. -! +! Error code. +! subroutine amg_d_dec_aggregator_build_tprol(ag,parms,ag_data,& & a,desc_a,ilaggr,nlaggr,t_prol,info) use psb_base_mod @@ -81,7 +81,7 @@ subroutine amg_d_dec_aggregator_build_tprol(ag,parms,ag_data,& use amg_d_inner_mod implicit none class(amg_d_dec_aggregator_type), target, intent(inout) :: ag - type(amg_dml_parms), intent(inout) :: parms + type(amg_dml_parms), intent(inout) :: parms type(amg_daggr_data), intent(in) :: ag_data type(psb_dspmat_type), intent(in) :: a type(psb_desc_type), intent(in) :: desc_a @@ -112,7 +112,7 @@ subroutine amg_d_dec_aggregator_build_tprol(ag,parms,ag_data,& call amg_check_def(parms%ml_cycle,'Multilevel cycle',& & amg_mult_ml_,is_legal_ml_cycle) call amg_check_def(parms%par_aggr_alg,'Aggregation',& - & amg_dec_aggr_,is_legal_ml_par_aggr_alg) + & amg_dec_aggr_,is_legal_decoupled_par_aggr_alg) call amg_check_def(parms%aggr_ord,'Ordering',& & amg_aggr_ord_nat_,is_legal_ml_aggr_ord) call amg_check_def(parms%aggr_thresh,'Aggr_Thresh',dzero,is_legal_d_aggr_thrs) @@ -130,11 +130,11 @@ subroutine amg_d_dec_aggregator_build_tprol(ag,parms,ag_data,& call psb_errpush(info,name,a_err='soc_map_bld/map_to_tprol') goto 9999 endif - + call psb_erractionrestore(err_act) return 9999 call psb_error_handler(err_act) return - + end subroutine amg_d_dec_aggregator_build_tprol diff --git a/amgprec/impl/aggregator/amg_d_parmatch_aggregator_tprol.f90 b/amgprec/impl/aggregator/amg_d_parmatch_aggregator_tprol.f90 index b004bb3a..0b9c3167 100644 --- a/amgprec/impl/aggregator/amg_d_parmatch_aggregator_tprol.f90 +++ b/amgprec/impl/aggregator/amg_d_parmatch_aggregator_tprol.f90 @@ -177,7 +177,7 @@ subroutine amg_d_parmatch_aggregator_build_tprol(ag,parms,ag_data,& call amg_check_def(parms%ml_cycle,'Multilevel cycle',& & amg_mult_ml_,is_legal_ml_cycle) call amg_check_def(parms%par_aggr_alg,'Aggregation',& - & amg_dec_aggr_,is_legal_ml_par_aggr_alg) + & amg_coupled_aggr_,is_legal_coupled_par_aggr_alg) call amg_check_def(parms%aggr_ord,'Ordering',& & amg_aggr_ord_nat_,is_legal_ml_aggr_ord) call amg_check_def(parms%aggr_thresh,'Aggr_Thresh',dzero,is_legal_d_aggr_thrs) diff --git a/amgprec/impl/aggregator/amg_d_symdec_aggregator_tprol.f90 b/amgprec/impl/aggregator/amg_d_symdec_aggregator_tprol.f90 index b33c632d..35efe63d 100644 --- a/amgprec/impl/aggregator/amg_d_symdec_aggregator_tprol.f90 +++ b/amgprec/impl/aggregator/amg_d_symdec_aggregator_tprol.f90 @@ -1,15 +1,15 @@ -! -! +! is_legal_decoupled_par_aggr_alg +! ! AMG4PSBLAS version 1.0 ! Algebraic Multigrid Package ! based on PSBLAS (Parallel Sparse BLAS version 3.7) -! -! (C) Copyright 2021 -! -! Salvatore Filippone -! Pasqua D'Ambra -! Fabio Durastante -! +! +! (C) Copyright 2021 +! +! Salvatore Filippone +! Pasqua D'Ambra +! Fabio Durastante +! ! Redistribution and use in source and binary forms, with or without ! modification, are permitted provided that the following conditions ! are met: @@ -21,7 +21,7 @@ ! 3. The name of the AMG4PSBLAS group or the names of its contributors may ! not be used to endorse or promote products derived from this ! software without specific written permission. -! +! ! THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS ! ``AS IS'' AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED ! TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR @@ -33,28 +33,28 @@ ! CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ! ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE ! POSSIBILITY OF SUCH DAMAGE. -! -! +! +! ! File: amg_d_symdec_aggregator_tprol.f90 ! ! Subroutine: amg_d_symdec_aggregator_tprol ! Version: real ! ! -! This routine is mainly an interface to soc_map_bld where the real work is performed. +! This routine is mainly an interface to soc_map_bld where the real work is performed. ! It takes care of some consistency checking, and calls map_to_tprol, which is ! refactored and shared among all the aggregation methods that produce a simple -! integer mapping. It also symmetrizes the pattern of the local matrix A. +! integer mapping. It also symmetrizes the pattern of the local matrix A. +! ! ! -! ! Arguments: ! Arguments: ! ag - type(amg_d_dec_aggregator_type), input/output. ! The aggregator object, carrying with itself the mapping algorithm. ! parms - The auxiliary parameters object ! ag_data - Auxiliary global aggregation parameters object -! +! ! a - type(psb_dspmat_type). ! The sparse matrix structure containing the local part of the ! fine-level matrix. @@ -72,10 +72,10 @@ ! nlaggr(i) contains the aggregates held by process i. ! op_prol - type(psb_dspmat_type), output ! The tentative prolongator, based on ilaggr. -! +! ! info - integer, output. -! Error code. -! +! Error code. +! subroutine amg_d_symdec_aggregator_build_tprol(ag,parms,ag_data,& & a,desc_a,ilaggr,nlaggr,op_prol,info) use psb_base_mod @@ -84,7 +84,7 @@ subroutine amg_d_symdec_aggregator_build_tprol(ag,parms,ag_data,& use amg_d_inner_mod implicit none class(amg_d_symdec_aggregator_type), target, intent(inout) :: ag - type(amg_dml_parms), intent(inout) :: parms + type(amg_dml_parms), intent(inout) :: parms type(amg_daggr_data), intent(in) :: ag_data type(psb_dspmat_type), intent(in) :: a type(psb_desc_type), intent(in) :: desc_a @@ -117,7 +117,7 @@ subroutine amg_d_symdec_aggregator_build_tprol(ag,parms,ag_data,& call amg_check_def(parms%ml_cycle,'Multilevel cycle',& & amg_mult_ml_,is_legal_ml_cycle) call amg_check_def(parms%par_aggr_alg,'Aggregation',& - & amg_dec_aggr_,is_legal_ml_par_aggr_alg) + & amg_dec_aggr_,is_legal_decoupled_par_aggr_alg) call amg_check_def(parms%aggr_ord,'Ordering',& & amg_aggr_ord_nat_,is_legal_ml_aggr_ord) call amg_check_def(parms%aggr_thresh,'Aggr_Thresh',dzero,is_legal_d_aggr_thrs) @@ -129,7 +129,7 @@ subroutine amg_d_symdec_aggregator_build_tprol(ag,parms,ag_data,& call atmp%set_ncols(nr) if (info == psb_success_) call atmp%transp(atrans) if (info == psb_success_) call atrans%cscnv(info,type='COO') - if (info == psb_success_) call psb_rwextd(nr,atmp,info,b=atrans,rowscale=.false.) + if (info == psb_success_) call psb_rwextd(nr,atmp,info,b=atrans,rowscale=.false.) call atmp%set_nrows(nr) call atmp%set_ncols(nr) if (info == psb_success_) call atrans%free() @@ -145,7 +145,7 @@ subroutine amg_d_symdec_aggregator_build_tprol(ag,parms,ag_data,& & desc_a,nlaggr,ilaggr,info) if (info == psb_success_) call atmp%free() - if (info == psb_success_) call amg_map_to_tprol(desc_a,ilaggr,nlaggr,op_prol,info) + if (info == psb_success_) call amg_map_to_tprol(desc_a,ilaggr,nlaggr,op_prol,info) if (info /= psb_success_) then info=psb_err_from_subroutine_ call psb_errpush(info,name,a_err='soc_map_bld/map_to_tprol') diff --git a/amgprec/impl/aggregator/amg_s_dec_aggregator_tprol.f90 b/amgprec/impl/aggregator/amg_s_dec_aggregator_tprol.f90 index 71fc4fa1..0ab5274e 100644 --- a/amgprec/impl/aggregator/amg_s_dec_aggregator_tprol.f90 +++ b/amgprec/impl/aggregator/amg_s_dec_aggregator_tprol.f90 @@ -1,15 +1,15 @@ -! -! +! +! ! AMG4PSBLAS version 1.0 ! Algebraic Multigrid Package ! based on PSBLAS (Parallel Sparse BLAS version 3.7) -! -! (C) Copyright 2021 -! -! Salvatore Filippone -! Pasqua D'Ambra -! Fabio Durastante -! +! +! (C) Copyright 2021 +! +! Salvatore Filippone +! Pasqua D'Ambra +! Fabio Durastante +! ! Redistribution and use in source and binary forms, with or without ! modification, are permitted provided that the following conditions ! are met: @@ -21,7 +21,7 @@ ! 3. The name of the AMG4PSBLAS group or the names of its contributors may ! not be used to endorse or promote products derived from this ! software without specific written permission. -! +! ! THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS ! ``AS IS'' AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED ! TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR @@ -33,26 +33,26 @@ ! CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ! ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE ! POSSIBILITY OF SUCH DAMAGE. -! -! +! +! ! File: amg_s_dec_aggregator_tprol.f90 ! ! Subroutine: amg_s_dec_aggregator_tprol ! Version: real ! -! This routine is mainly an interface to soc_map_bld where the real work is performed. +! This routine is mainly an interface to soc_map_bld where the real work is performed. ! It takes care of some consistency checking, and calls map_to_tprol, which is ! refactored and shared among all the aggregation methods that produce a simple ! integer mapping. ! -! +! ! Arguments: ! ag - type(amg_s_dec_aggregator_type), input/output. ! The aggregator object, carrying with itself the mapping algorithm. ! parms - The auxiliary parameters object ! ag_data - Auxiliary global aggregation parameters object -! -! +! +! ! a - type(psb_sspmat_type). ! The sparse matrix structure containing the local part of the ! fine-level matrix. @@ -70,10 +70,10 @@ ! nlaggr(i) contains the aggregates held by process i. ! t_prol - type(psb_sspmat_type), output ! The tentative prolongator, based on ilaggr. -! +! ! info - integer, output. -! Error code. -! +! Error code. +! subroutine amg_s_dec_aggregator_build_tprol(ag,parms,ag_data,& & a,desc_a,ilaggr,nlaggr,t_prol,info) use psb_base_mod @@ -81,7 +81,7 @@ subroutine amg_s_dec_aggregator_build_tprol(ag,parms,ag_data,& use amg_s_inner_mod implicit none class(amg_s_dec_aggregator_type), target, intent(inout) :: ag - type(amg_sml_parms), intent(inout) :: parms + type(amg_sml_parms), intent(inout) :: parms type(amg_saggr_data), intent(in) :: ag_data type(psb_sspmat_type), intent(in) :: a type(psb_desc_type), intent(in) :: desc_a @@ -112,7 +112,7 @@ subroutine amg_s_dec_aggregator_build_tprol(ag,parms,ag_data,& call amg_check_def(parms%ml_cycle,'Multilevel cycle',& & amg_mult_ml_,is_legal_ml_cycle) call amg_check_def(parms%par_aggr_alg,'Aggregation',& - & amg_dec_aggr_,is_legal_ml_par_aggr_alg) + & amg_dec_aggr_,is_legal_decoupled_par_aggr_alg) call amg_check_def(parms%aggr_ord,'Ordering',& & amg_aggr_ord_nat_,is_legal_ml_aggr_ord) call amg_check_def(parms%aggr_thresh,'Aggr_Thresh',szero,is_legal_s_aggr_thrs) @@ -130,11 +130,11 @@ subroutine amg_s_dec_aggregator_build_tprol(ag,parms,ag_data,& call psb_errpush(info,name,a_err='soc_map_bld/map_to_tprol') goto 9999 endif - + call psb_erractionrestore(err_act) return 9999 call psb_error_handler(err_act) return - + end subroutine amg_s_dec_aggregator_build_tprol diff --git a/amgprec/impl/aggregator/amg_s_parmatch_aggregator_tprol.f90 b/amgprec/impl/aggregator/amg_s_parmatch_aggregator_tprol.f90 index e8d8fecc..f239de95 100644 --- a/amgprec/impl/aggregator/amg_s_parmatch_aggregator_tprol.f90 +++ b/amgprec/impl/aggregator/amg_s_parmatch_aggregator_tprol.f90 @@ -177,7 +177,7 @@ subroutine amg_s_parmatch_aggregator_build_tprol(ag,parms,ag_data,& call amg_check_def(parms%ml_cycle,'Multilevel cycle',& & amg_mult_ml_,is_legal_ml_cycle) call amg_check_def(parms%par_aggr_alg,'Aggregation',& - & amg_dec_aggr_,is_legal_ml_par_aggr_alg) + & amg_coupled_aggr_,is_legal_coupled_par_aggr_alg) call amg_check_def(parms%aggr_ord,'Ordering',& & amg_aggr_ord_nat_,is_legal_ml_aggr_ord) call amg_check_def(parms%aggr_thresh,'Aggr_Thresh',szero,is_legal_s_aggr_thrs) diff --git a/amgprec/impl/aggregator/amg_s_symdec_aggregator_tprol.f90 b/amgprec/impl/aggregator/amg_s_symdec_aggregator_tprol.f90 index 3b8462c1..5a9548eb 100644 --- a/amgprec/impl/aggregator/amg_s_symdec_aggregator_tprol.f90 +++ b/amgprec/impl/aggregator/amg_s_symdec_aggregator_tprol.f90 @@ -1,15 +1,15 @@ -! -! +! is_legal_decoupled_par_aggr_alg +! ! AMG4PSBLAS version 1.0 ! Algebraic Multigrid Package ! based on PSBLAS (Parallel Sparse BLAS version 3.7) -! -! (C) Copyright 2021 -! -! Salvatore Filippone -! Pasqua D'Ambra -! Fabio Durastante -! +! +! (C) Copyright 2021 +! +! Salvatore Filippone +! Pasqua D'Ambra +! Fabio Durastante +! ! Redistribution and use in source and binary forms, with or without ! modification, are permitted provided that the following conditions ! are met: @@ -21,7 +21,7 @@ ! 3. The name of the AMG4PSBLAS group or the names of its contributors may ! not be used to endorse or promote products derived from this ! software without specific written permission. -! +! ! THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS ! ``AS IS'' AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED ! TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR @@ -33,28 +33,28 @@ ! CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ! ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE ! POSSIBILITY OF SUCH DAMAGE. -! -! +! +! ! File: amg_s_symdec_aggregator_tprol.f90 ! ! Subroutine: amg_s_symdec_aggregator_tprol ! Version: real ! ! -! This routine is mainly an interface to soc_map_bld where the real work is performed. +! This routine is mainly an interface to soc_map_bld where the real work is performed. ! It takes care of some consistency checking, and calls map_to_tprol, which is ! refactored and shared among all the aggregation methods that produce a simple -! integer mapping. It also symmetrizes the pattern of the local matrix A. +! integer mapping. It also symmetrizes the pattern of the local matrix A. +! ! ! -! ! Arguments: ! Arguments: ! ag - type(amg_s_dec_aggregator_type), input/output. ! The aggregator object, carrying with itself the mapping algorithm. ! parms - The auxiliary parameters object ! ag_data - Auxiliary global aggregation parameters object -! +! ! a - type(psb_sspmat_type). ! The sparse matrix structure containing the local part of the ! fine-level matrix. @@ -72,10 +72,10 @@ ! nlaggr(i) contains the aggregates held by process i. ! op_prol - type(psb_sspmat_type), output ! The tentative prolongator, based on ilaggr. -! +! ! info - integer, output. -! Error code. -! +! Error code. +! subroutine amg_s_symdec_aggregator_build_tprol(ag,parms,ag_data,& & a,desc_a,ilaggr,nlaggr,op_prol,info) use psb_base_mod @@ -84,7 +84,7 @@ subroutine amg_s_symdec_aggregator_build_tprol(ag,parms,ag_data,& use amg_s_inner_mod implicit none class(amg_s_symdec_aggregator_type), target, intent(inout) :: ag - type(amg_sml_parms), intent(inout) :: parms + type(amg_sml_parms), intent(inout) :: parms type(amg_saggr_data), intent(in) :: ag_data type(psb_sspmat_type), intent(in) :: a type(psb_desc_type), intent(in) :: desc_a @@ -117,7 +117,7 @@ subroutine amg_s_symdec_aggregator_build_tprol(ag,parms,ag_data,& call amg_check_def(parms%ml_cycle,'Multilevel cycle',& & amg_mult_ml_,is_legal_ml_cycle) call amg_check_def(parms%par_aggr_alg,'Aggregation',& - & amg_dec_aggr_,is_legal_ml_par_aggr_alg) + & amg_dec_aggr_,is_legal_decoupled_par_aggr_alg) call amg_check_def(parms%aggr_ord,'Ordering',& & amg_aggr_ord_nat_,is_legal_ml_aggr_ord) call amg_check_def(parms%aggr_thresh,'Aggr_Thresh',szero,is_legal_s_aggr_thrs) @@ -129,7 +129,7 @@ subroutine amg_s_symdec_aggregator_build_tprol(ag,parms,ag_data,& call atmp%set_ncols(nr) if (info == psb_success_) call atmp%transp(atrans) if (info == psb_success_) call atrans%cscnv(info,type='COO') - if (info == psb_success_) call psb_rwextd(nr,atmp,info,b=atrans,rowscale=.false.) + if (info == psb_success_) call psb_rwextd(nr,atmp,info,b=atrans,rowscale=.false.) call atmp%set_nrows(nr) call atmp%set_ncols(nr) if (info == psb_success_) call atrans%free() @@ -145,7 +145,7 @@ subroutine amg_s_symdec_aggregator_build_tprol(ag,parms,ag_data,& & desc_a,nlaggr,ilaggr,info) if (info == psb_success_) call atmp%free() - if (info == psb_success_) call amg_map_to_tprol(desc_a,ilaggr,nlaggr,op_prol,info) + if (info == psb_success_) call amg_map_to_tprol(desc_a,ilaggr,nlaggr,op_prol,info) if (info /= psb_success_) then info=psb_err_from_subroutine_ call psb_errpush(info,name,a_err='soc_map_bld/map_to_tprol') diff --git a/amgprec/impl/aggregator/amg_z_dec_aggregator_tprol.f90 b/amgprec/impl/aggregator/amg_z_dec_aggregator_tprol.f90 index abe8bf05..5a90fcd5 100644 --- a/amgprec/impl/aggregator/amg_z_dec_aggregator_tprol.f90 +++ b/amgprec/impl/aggregator/amg_z_dec_aggregator_tprol.f90 @@ -1,15 +1,15 @@ -! -! +! +! ! AMG4PSBLAS version 1.0 ! Algebraic Multigrid Package ! based on PSBLAS (Parallel Sparse BLAS version 3.7) -! -! (C) Copyright 2021 -! -! Salvatore Filippone -! Pasqua D'Ambra -! Fabio Durastante -! +! +! (C) Copyright 2021 +! +! Salvatore Filippone +! Pasqua D'Ambra +! Fabio Durastante +! ! Redistribution and use in source and binary forms, with or without ! modification, are permitted provided that the following conditions ! are met: @@ -21,7 +21,7 @@ ! 3. The name of the AMG4PSBLAS group or the names of its contributors may ! not be used to endorse or promote products derived from this ! software without specific written permission. -! +! ! THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS ! ``AS IS'' AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED ! TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR @@ -33,26 +33,26 @@ ! CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ! ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE ! POSSIBILITY OF SUCH DAMAGE. -! -! +! +! ! File: amg_z_dec_aggregator_tprol.f90 ! ! Subroutine: amg_z_dec_aggregator_tprol ! Version: complex ! -! This routine is mainly an interface to soc_map_bld where the real work is performed. +! This routine is mainly an interface to soc_map_bld where the real work is performed. ! It takes care of some consistency checking, and calls map_to_tprol, which is ! refactored and shared among all the aggregation methods that produce a simple ! integer mapping. ! -! +! ! Arguments: ! ag - type(amg_z_dec_aggregator_type), input/output. ! The aggregator object, carrying with itself the mapping algorithm. ! parms - The auxiliary parameters object ! ag_data - Auxiliary global aggregation parameters object -! -! +! +! ! a - type(psb_zspmat_type). ! The sparse matrix structure containing the local part of the ! fine-level matrix. @@ -70,10 +70,10 @@ ! nlaggr(i) contains the aggregates held by process i. ! t_prol - type(psb_zspmat_type), output ! The tentative prolongator, based on ilaggr. -! +! ! info - integer, output. -! Error code. -! +! Error code. +! subroutine amg_z_dec_aggregator_build_tprol(ag,parms,ag_data,& & a,desc_a,ilaggr,nlaggr,t_prol,info) use psb_base_mod @@ -81,7 +81,7 @@ subroutine amg_z_dec_aggregator_build_tprol(ag,parms,ag_data,& use amg_z_inner_mod implicit none class(amg_z_dec_aggregator_type), target, intent(inout) :: ag - type(amg_dml_parms), intent(inout) :: parms + type(amg_dml_parms), intent(inout) :: parms type(amg_daggr_data), intent(in) :: ag_data type(psb_zspmat_type), intent(in) :: a type(psb_desc_type), intent(in) :: desc_a @@ -112,7 +112,7 @@ subroutine amg_z_dec_aggregator_build_tprol(ag,parms,ag_data,& call amg_check_def(parms%ml_cycle,'Multilevel cycle',& & amg_mult_ml_,is_legal_ml_cycle) call amg_check_def(parms%par_aggr_alg,'Aggregation',& - & amg_dec_aggr_,is_legal_ml_par_aggr_alg) + & amg_dec_aggr_,is_legal_decoupled_par_aggr_alg) call amg_check_def(parms%aggr_ord,'Ordering',& & amg_aggr_ord_nat_,is_legal_ml_aggr_ord) call amg_check_def(parms%aggr_thresh,'Aggr_Thresh',dzero,is_legal_d_aggr_thrs) @@ -130,11 +130,11 @@ subroutine amg_z_dec_aggregator_build_tprol(ag,parms,ag_data,& call psb_errpush(info,name,a_err='soc_map_bld/map_to_tprol') goto 9999 endif - + call psb_erractionrestore(err_act) return 9999 call psb_error_handler(err_act) return - + end subroutine amg_z_dec_aggregator_build_tprol diff --git a/amgprec/impl/aggregator/amg_z_symdec_aggregator_tprol.f90 b/amgprec/impl/aggregator/amg_z_symdec_aggregator_tprol.f90 index e76add39..84de6849 100644 --- a/amgprec/impl/aggregator/amg_z_symdec_aggregator_tprol.f90 +++ b/amgprec/impl/aggregator/amg_z_symdec_aggregator_tprol.f90 @@ -1,15 +1,15 @@ -! -! +! is_legal_decoupled_par_aggr_alg +! ! AMG4PSBLAS version 1.0 ! Algebraic Multigrid Package ! based on PSBLAS (Parallel Sparse BLAS version 3.7) -! -! (C) Copyright 2021 -! -! Salvatore Filippone -! Pasqua D'Ambra -! Fabio Durastante -! +! +! (C) Copyright 2021 +! +! Salvatore Filippone +! Pasqua D'Ambra +! Fabio Durastante +! ! Redistribution and use in source and binary forms, with or without ! modification, are permitted provided that the following conditions ! are met: @@ -21,7 +21,7 @@ ! 3. The name of the AMG4PSBLAS group or the names of its contributors may ! not be used to endorse or promote products derived from this ! software without specific written permission. -! +! ! THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS ! ``AS IS'' AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED ! TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR @@ -33,28 +33,28 @@ ! CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ! ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE ! POSSIBILITY OF SUCH DAMAGE. -! -! +! +! ! File: amg_z_symdec_aggregator_tprol.f90 ! ! Subroutine: amg_z_symdec_aggregator_tprol ! Version: complex ! ! -! This routine is mainly an interface to soc_map_bld where the real work is performed. +! This routine is mainly an interface to soc_map_bld where the real work is performed. ! It takes care of some consistency checking, and calls map_to_tprol, which is ! refactored and shared among all the aggregation methods that produce a simple -! integer mapping. It also symmetrizes the pattern of the local matrix A. +! integer mapping. It also symmetrizes the pattern of the local matrix A. +! ! ! -! ! Arguments: ! Arguments: ! ag - type(amg_z_dec_aggregator_type), input/output. ! The aggregator object, carrying with itself the mapping algorithm. ! parms - The auxiliary parameters object ! ag_data - Auxiliary global aggregation parameters object -! +! ! a - type(psb_zspmat_type). ! The sparse matrix structure containing the local part of the ! fine-level matrix. @@ -72,10 +72,10 @@ ! nlaggr(i) contains the aggregates held by process i. ! op_prol - type(psb_zspmat_type), output ! The tentative prolongator, based on ilaggr. -! +! ! info - integer, output. -! Error code. -! +! Error code. +! subroutine amg_z_symdec_aggregator_build_tprol(ag,parms,ag_data,& & a,desc_a,ilaggr,nlaggr,op_prol,info) use psb_base_mod @@ -84,7 +84,7 @@ subroutine amg_z_symdec_aggregator_build_tprol(ag,parms,ag_data,& use amg_z_inner_mod implicit none class(amg_z_symdec_aggregator_type), target, intent(inout) :: ag - type(amg_dml_parms), intent(inout) :: parms + type(amg_dml_parms), intent(inout) :: parms type(amg_daggr_data), intent(in) :: ag_data type(psb_zspmat_type), intent(in) :: a type(psb_desc_type), intent(in) :: desc_a @@ -117,7 +117,7 @@ subroutine amg_z_symdec_aggregator_build_tprol(ag,parms,ag_data,& call amg_check_def(parms%ml_cycle,'Multilevel cycle',& & amg_mult_ml_,is_legal_ml_cycle) call amg_check_def(parms%par_aggr_alg,'Aggregation',& - & amg_dec_aggr_,is_legal_ml_par_aggr_alg) + & amg_dec_aggr_,is_legal_decoupled_par_aggr_alg) call amg_check_def(parms%aggr_ord,'Ordering',& & amg_aggr_ord_nat_,is_legal_ml_aggr_ord) call amg_check_def(parms%aggr_thresh,'Aggr_Thresh',dzero,is_legal_d_aggr_thrs) @@ -129,7 +129,7 @@ subroutine amg_z_symdec_aggregator_build_tprol(ag,parms,ag_data,& call atmp%set_ncols(nr) if (info == psb_success_) call atmp%transp(atrans) if (info == psb_success_) call atrans%cscnv(info,type='COO') - if (info == psb_success_) call psb_rwextd(nr,atmp,info,b=atrans,rowscale=.false.) + if (info == psb_success_) call psb_rwextd(nr,atmp,info,b=atrans,rowscale=.false.) call atmp%set_nrows(nr) call atmp%set_ncols(nr) if (info == psb_success_) call atrans%free() @@ -145,7 +145,7 @@ subroutine amg_z_symdec_aggregator_build_tprol(ag,parms,ag_data,& & desc_a,nlaggr,ilaggr,info) if (info == psb_success_) call atmp%free() - if (info == psb_success_) call amg_map_to_tprol(desc_a,ilaggr,nlaggr,op_prol,info) + if (info == psb_success_) call amg_map_to_tprol(desc_a,ilaggr,nlaggr,op_prol,info) if (info /= psb_success_) then info=psb_err_from_subroutine_ call psb_errpush(info,name,a_err='soc_map_bld/map_to_tprol') diff --git a/amgprec/impl/level/amg_c_base_onelev_csetc.F90 b/amgprec/impl/level/amg_c_base_onelev_csetc.F90 index 3b75263f..732f334c 100644 --- a/amgprec/impl/level/amg_c_base_onelev_csetc.F90 +++ b/amgprec/impl/level/amg_c_base_onelev_csetc.F90 @@ -242,10 +242,10 @@ subroutine amg_c_base_onelev_csetc(lv,what,val,info,pos,idx) end if end if - select case(ival) - case(amg_dec_aggr_) + select case(val) + case('DEC') allocate(amg_c_dec_aggregator_type :: lv%aggr, stat=info) - case(amg_sym_dec_aggr_) + case('SYMDEC') allocate(amg_c_symdec_aggregator_type :: lv%aggr, stat=info) case default info = psb_err_internal_error_ diff --git a/amgprec/impl/level/amg_c_base_onelev_cseti.F90 b/amgprec/impl/level/amg_c_base_onelev_cseti.F90 index b1135622..deba9001 100644 --- a/amgprec/impl/level/amg_c_base_onelev_cseti.F90 +++ b/amgprec/impl/level/amg_c_base_onelev_cseti.F90 @@ -1,15 +1,15 @@ -! -! +! +! ! AMG4PSBLAS version 1.0 ! Algebraic Multigrid Package ! based on PSBLAS (Parallel Sparse BLAS version 3.7) -! -! (C) Copyright 2021 -! -! Salvatore Filippone -! Pasqua D'Ambra -! Fabio Durastante -! +! +! (C) Copyright 2021 +! +! Salvatore Filippone +! Pasqua D'Ambra +! Fabio Durastante +! ! Redistribution and use in source and binary forms, with or without ! modification, are permitted provided that the following conditions ! are met: @@ -21,7 +21,7 @@ ! 3. The name of the AMG4PSBLAS group or the names of its contributors may ! not be used to endorse or promote products derived from this ! software without specific written permission. -! +! ! THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS ! ``AS IS'' AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED ! TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR @@ -33,10 +33,10 @@ ! CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ! ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE ! POSSIBILITY OF SUCH DAMAGE. -! -! +! +! subroutine amg_c_base_onelev_cseti(lv,what,val,info,pos,idx) - + use psb_base_mod use amg_c_onelev_mod, amg_protect_name => amg_c_base_onelev_cseti use amg_c_base_aggregator_mod @@ -59,13 +59,13 @@ subroutine amg_c_base_onelev_cseti(lv,what,val,info,pos,idx) Implicit None ! Arguments - class(amg_c_onelev_type), intent(inout) :: lv - character(len=*), intent(in) :: what + class(amg_c_onelev_type), intent(inout) :: lv + character(len=*), intent(in) :: what integer(psb_ipk_), intent(in) :: val integer(psb_ipk_), intent(out) :: info character(len=*), optional, intent(in) :: pos integer(psb_ipk_), intent(in), optional :: idx - ! Local + ! Local integer(psb_ipk_) :: ipos_, err_act character(len=20) :: name='c_base_onelev_cseti' type(amg_c_base_smoother_type) :: amg_c_base_smoother_mold @@ -84,7 +84,7 @@ subroutine amg_c_base_onelev_cseti(lv,what,val,info,pos,idx) #if defined(HAVE_MUMPS_) type(amg_c_mumps_solver_type) :: amg_c_mumps_solver_mold #endif - + call psb_erractionsave(err_act) info = psb_success_ @@ -100,14 +100,14 @@ subroutine amg_c_base_onelev_cseti(lv,what,val,info,pos,idx) else ipos_ = amg_smooth_both_ end if - + select case (psb_toupper(what)) case ('SMOOTHER_TYPE') - select case (val) + select case (val) case (amg_noprec_) call lv%set(amg_c_base_smoother_mold,info,pos=pos) if (info == 0) call lv%set(amg_c_id_solver_mold,info,pos=pos) - + case (amg_jac_) call lv%set(amg_c_jac_smoother_mold,info,pos=pos) if (info == 0) call lv%set(amg_c_diag_solver_mold,info,pos=pos) @@ -115,11 +115,11 @@ subroutine amg_c_base_onelev_cseti(lv,what,val,info,pos,idx) case (amg_l1_jac_) call lv%set(amg_c_jac_smoother_mold,info,pos=pos) if (info == 0) call lv%set(amg_c_l1_diag_solver_mold,info,pos=pos) - + case (amg_bjac_) call lv%set(amg_c_jac_smoother_mold,info,pos=pos) if (info == 0) call lv%set(amg_c_ilu_solver_mold,info,pos=pos) - + case (amg_l1_bjac_) call lv%set(amg_c_l1_jac_smoother_mold,info,pos=pos) if (info == 0) call lv%set(amg_c_ilu_solver_mold,info,pos=pos) @@ -133,61 +133,61 @@ subroutine amg_c_base_onelev_cseti(lv,what,val,info,pos,idx) if (info == 0) call lv%set(amg_c_gs_solver_mold,info,pos='pre') call lv%set(amg_c_jac_smoother_mold,info,pos='post') if (info == 0) call lv%set(amg_c_bwgs_solver_mold,info,pos='post') - + case default ! - ! Do nothing and hope for the best :) + ! Do nothing and hope for the best :) ! end select - if ((ipos_==amg_smooth_pre_).or.(ipos_==amg_smooth_both_)) then + if ((ipos_==amg_smooth_pre_).or.(ipos_==amg_smooth_both_)) then if (allocated(lv%sm)) call lv%sm%default() end if if ((ipos_==amg_smooth_post_).or.(ipos_==amg_smooth_both_)) then if (allocated(lv%sm2a)) call lv%sm2a%default() end if - + case('SUB_SOLVE') - select case (val) + select case (val) case (amg_f_none_) call lv%set(amg_c_id_solver_mold,info,pos=pos) - + case (amg_diag_scale_) call lv%set(amg_c_diag_solver_mold,info,pos=pos) - + case (amg_l1_diag_scale_) call lv%set(amg_c_l1_diag_solver_mold,info,pos=pos) - + case (amg_gs_) call lv%set(amg_c_gs_solver_mold,info,pos=pos) - + case (amg_bwgs_) call lv%set(amg_c_bwgs_solver_mold,info,pos=pos) - + case (psb_ilu_n_,psb_milu_n_,psb_ilu_t_) call lv%set(amg_c_ilu_solver_mold,info,pos=pos) if (info == 0) then - if ((ipos_==amg_smooth_pre_) .or.(ipos_==amg_smooth_both_)) then + if ((ipos_==amg_smooth_pre_) .or.(ipos_==amg_smooth_both_)) then call lv%sm%sv%set('SUB_SOLVE',val,info) end if - if ((ipos_==amg_smooth_post_).or.(ipos_==amg_smooth_both_))then + if ((ipos_==amg_smooth_post_).or.(ipos_==amg_smooth_both_))then if (allocated(lv%sm2a)) call lv%sm2a%sv%set('SUB_SOLVE',val,info) end if end if #ifdef HAVE_SLU_ - case (amg_slu_) + case (amg_slu_) call lv%set(amg_c_slu_solver_mold,info,pos=pos) #endif #ifdef HAVE_MUMPS_ - case (amg_mumps_) + case (amg_mumps_) call lv%set(amg_c_mumps_solver_mold,info,pos=pos) #endif case default ! - ! Do nothing and hope for the best :) + ! Do nothing and hope for the best :) ! end select - + case ('SMOOTHER_SWEEPS') if ((ipos_==amg_smooth_pre_) .or.(ipos_==amg_smooth_both_)) & @@ -208,7 +208,7 @@ subroutine amg_c_base_onelev_cseti(lv,what,val,info,pos,idx) return end if end if - + select case(val) case(amg_dec_aggr_) allocate(amg_c_dec_aggregator_type :: lv%aggr, stat=info) @@ -218,7 +218,7 @@ subroutine amg_c_base_onelev_cseti(lv,what,val,info,pos,idx) info = psb_err_internal_error_ end select if (info == psb_success_) call lv%aggr%default() - + case ('AGGR_ORD') lv%parms%aggr_ord = val @@ -245,13 +245,13 @@ subroutine amg_c_base_onelev_cseti(lv,what,val,info,pos,idx) lv%parms%coarse_solve = val case default - if ((ipos_==amg_smooth_pre_) .or.(ipos_==amg_smooth_both_)) then - if (allocated(lv%sm)) then + if ((ipos_==amg_smooth_pre_) .or.(ipos_==amg_smooth_both_)) then + if (allocated(lv%sm)) then call lv%sm%set(what,val,info,idx=idx) end if end if - if ((ipos_==amg_smooth_post_).or.(ipos_==amg_smooth_both_))then - if (allocated(lv%sm2a)) then + if ((ipos_==amg_smooth_post_).or.(ipos_==amg_smooth_both_))then + if (allocated(lv%sm2a)) then call lv%sm2a%set(what,val,info,idx=idx) end if end if diff --git a/amgprec/impl/level/amg_d_base_onelev_csetc.F90 b/amgprec/impl/level/amg_d_base_onelev_csetc.F90 index 1cc62a3d..eaacabbc 100644 --- a/amgprec/impl/level/amg_d_base_onelev_csetc.F90 +++ b/amgprec/impl/level/amg_d_base_onelev_csetc.F90 @@ -42,6 +42,7 @@ subroutine amg_d_base_onelev_csetc(lv,what,val,info,pos,idx) use amg_d_base_aggregator_mod use amg_d_dec_aggregator_mod use amg_d_symdec_aggregator_mod + use amg_d_parmatch_aggregator_mod use amg_d_jac_smoother use amg_d_as_smoother use amg_d_diag_solver @@ -262,11 +263,13 @@ subroutine amg_d_base_onelev_csetc(lv,what,val,info,pos,idx) end if end if - select case(ival) - case(amg_dec_aggr_) + select case(val) + case('DEC') allocate(amg_d_dec_aggregator_type :: lv%aggr, stat=info) - case(amg_sym_dec_aggr_) + case('SYMDEC') allocate(amg_d_symdec_aggregator_type :: lv%aggr, stat=info) + case('COUP','COUPLED') + allocate(amg_d_parmatch_aggregator_type :: lv%aggr, stat=info) case default info = psb_err_internal_error_ end select diff --git a/amgprec/impl/level/amg_d_base_onelev_cseti.F90 b/amgprec/impl/level/amg_d_base_onelev_cseti.F90 index 39b5fd43..b5ca549b 100644 --- a/amgprec/impl/level/amg_d_base_onelev_cseti.F90 +++ b/amgprec/impl/level/amg_d_base_onelev_cseti.F90 @@ -1,15 +1,15 @@ -! -! +! +! ! AMG4PSBLAS version 1.0 ! Algebraic Multigrid Package ! based on PSBLAS (Parallel Sparse BLAS version 3.7) -! -! (C) Copyright 2021 -! -! Salvatore Filippone -! Pasqua D'Ambra -! Fabio Durastante -! +! +! (C) Copyright 2021 +! +! Salvatore Filippone +! Pasqua D'Ambra +! Fabio Durastante +! ! Redistribution and use in source and binary forms, with or without ! modification, are permitted provided that the following conditions ! are met: @@ -21,7 +21,7 @@ ! 3. The name of the AMG4PSBLAS group or the names of its contributors may ! not be used to endorse or promote products derived from this ! software without specific written permission. -! +! ! THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS ! ``AS IS'' AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED ! TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR @@ -33,15 +33,16 @@ ! CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ! ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE ! POSSIBILITY OF SUCH DAMAGE. -! -! +! +! subroutine amg_d_base_onelev_cseti(lv,what,val,info,pos,idx) - + use psb_base_mod use amg_d_onelev_mod, amg_protect_name => amg_d_base_onelev_cseti use amg_d_base_aggregator_mod use amg_d_dec_aggregator_mod use amg_d_symdec_aggregator_mod + use amg_d_parmatch_aggregator_mod use amg_d_jac_smoother use amg_d_as_smoother use amg_d_diag_solver @@ -65,13 +66,13 @@ subroutine amg_d_base_onelev_cseti(lv,what,val,info,pos,idx) Implicit None ! Arguments - class(amg_d_onelev_type), intent(inout) :: lv - character(len=*), intent(in) :: what + class(amg_d_onelev_type), intent(inout) :: lv + character(len=*), intent(in) :: what integer(psb_ipk_), intent(in) :: val integer(psb_ipk_), intent(out) :: info character(len=*), optional, intent(in) :: pos integer(psb_ipk_), intent(in), optional :: idx - ! Local + ! Local integer(psb_ipk_) :: ipos_, err_act character(len=20) :: name='d_base_onelev_cseti' type(amg_d_base_smoother_type) :: amg_d_base_smoother_mold @@ -96,7 +97,7 @@ subroutine amg_d_base_onelev_cseti(lv,what,val,info,pos,idx) #if defined(HAVE_MUMPS_) type(amg_d_mumps_solver_type) :: amg_d_mumps_solver_mold #endif - + call psb_erractionsave(err_act) info = psb_success_ @@ -112,14 +113,14 @@ subroutine amg_d_base_onelev_cseti(lv,what,val,info,pos,idx) else ipos_ = amg_smooth_both_ end if - + select case (psb_toupper(what)) case ('SMOOTHER_TYPE') - select case (val) + select case (val) case (amg_noprec_) call lv%set(amg_d_base_smoother_mold,info,pos=pos) if (info == 0) call lv%set(amg_d_id_solver_mold,info,pos=pos) - + case (amg_jac_) call lv%set(amg_d_jac_smoother_mold,info,pos=pos) if (info == 0) call lv%set(amg_d_diag_solver_mold,info,pos=pos) @@ -127,11 +128,11 @@ subroutine amg_d_base_onelev_cseti(lv,what,val,info,pos,idx) case (amg_l1_jac_) call lv%set(amg_d_jac_smoother_mold,info,pos=pos) if (info == 0) call lv%set(amg_d_l1_diag_solver_mold,info,pos=pos) - + case (amg_bjac_) call lv%set(amg_d_jac_smoother_mold,info,pos=pos) if (info == 0) call lv%set(amg_d_ilu_solver_mold,info,pos=pos) - + case (amg_l1_bjac_) call lv%set(amg_d_l1_jac_smoother_mold,info,pos=pos) if (info == 0) call lv%set(amg_d_ilu_solver_mold,info,pos=pos) @@ -145,53 +146,53 @@ subroutine amg_d_base_onelev_cseti(lv,what,val,info,pos,idx) if (info == 0) call lv%set(amg_d_gs_solver_mold,info,pos='pre') call lv%set(amg_d_jac_smoother_mold,info,pos='post') if (info == 0) call lv%set(amg_d_bwgs_solver_mold,info,pos='post') - + case default ! - ! Do nothing and hope for the best :) + ! Do nothing and hope for the best :) ! end select - if ((ipos_==amg_smooth_pre_).or.(ipos_==amg_smooth_both_)) then + if ((ipos_==amg_smooth_pre_).or.(ipos_==amg_smooth_both_)) then if (allocated(lv%sm)) call lv%sm%default() end if if ((ipos_==amg_smooth_post_).or.(ipos_==amg_smooth_both_)) then if (allocated(lv%sm2a)) call lv%sm2a%default() end if - + case('SUB_SOLVE') - select case (val) + select case (val) case (amg_f_none_) call lv%set(amg_d_id_solver_mold,info,pos=pos) - + case (amg_diag_scale_) call lv%set(amg_d_diag_solver_mold,info,pos=pos) - + case (amg_l1_diag_scale_) call lv%set(amg_d_l1_diag_solver_mold,info,pos=pos) - + case (amg_gs_) call lv%set(amg_d_gs_solver_mold,info,pos=pos) - + case (amg_bwgs_) call lv%set(amg_d_bwgs_solver_mold,info,pos=pos) - + case (psb_ilu_n_,psb_milu_n_,psb_ilu_t_) call lv%set(amg_d_ilu_solver_mold,info,pos=pos) if (info == 0) then - if ((ipos_==amg_smooth_pre_) .or.(ipos_==amg_smooth_both_)) then + if ((ipos_==amg_smooth_pre_) .or.(ipos_==amg_smooth_both_)) then call lv%sm%sv%set('SUB_SOLVE',val,info) end if - if ((ipos_==amg_smooth_post_).or.(ipos_==amg_smooth_both_))then + if ((ipos_==amg_smooth_post_).or.(ipos_==amg_smooth_both_))then if (allocated(lv%sm2a)) call lv%sm2a%sv%set('SUB_SOLVE',val,info) end if end if #ifdef HAVE_SLU_ - case (amg_slu_) + case (amg_slu_) call lv%set(amg_d_slu_solver_mold,info,pos=pos) #endif #ifdef HAVE_MUMPS_ - case (amg_mumps_) + case (amg_mumps_) call lv%set(amg_d_mumps_solver_mold,info,pos=pos) #endif #ifdef HAVE_SLUDIST_ @@ -204,10 +205,10 @@ subroutine amg_d_base_onelev_cseti(lv,what,val,info,pos,idx) #endif case default ! - ! Do nothing and hope for the best :) + ! Do nothing and hope for the best :) ! end select - + case ('SMOOTHER_SWEEPS') if ((ipos_==amg_smooth_pre_) .or.(ipos_==amg_smooth_both_)) & @@ -228,7 +229,7 @@ subroutine amg_d_base_onelev_cseti(lv,what,val,info,pos,idx) return end if end if - + select case(val) case(amg_dec_aggr_) allocate(amg_d_dec_aggregator_type :: lv%aggr, stat=info) @@ -238,7 +239,7 @@ subroutine amg_d_base_onelev_cseti(lv,what,val,info,pos,idx) info = psb_err_internal_error_ end select if (info == psb_success_) call lv%aggr%default() - + case ('AGGR_ORD') lv%parms%aggr_ord = val @@ -265,13 +266,13 @@ subroutine amg_d_base_onelev_cseti(lv,what,val,info,pos,idx) lv%parms%coarse_solve = val case default - if ((ipos_==amg_smooth_pre_) .or.(ipos_==amg_smooth_both_)) then - if (allocated(lv%sm)) then + if ((ipos_==amg_smooth_pre_) .or.(ipos_==amg_smooth_both_)) then + if (allocated(lv%sm)) then call lv%sm%set(what,val,info,idx=idx) end if end if - if ((ipos_==amg_smooth_post_).or.(ipos_==amg_smooth_both_))then - if (allocated(lv%sm2a)) then + if ((ipos_==amg_smooth_post_).or.(ipos_==amg_smooth_both_))then + if (allocated(lv%sm2a)) then call lv%sm2a%set(what,val,info,idx=idx) end if end if diff --git a/amgprec/impl/level/amg_s_base_onelev_csetc.F90 b/amgprec/impl/level/amg_s_base_onelev_csetc.F90 index 7f8a82c8..f5f944bf 100644 --- a/amgprec/impl/level/amg_s_base_onelev_csetc.F90 +++ b/amgprec/impl/level/amg_s_base_onelev_csetc.F90 @@ -42,6 +42,7 @@ subroutine amg_s_base_onelev_csetc(lv,what,val,info,pos,idx) use amg_s_base_aggregator_mod use amg_s_dec_aggregator_mod use amg_s_symdec_aggregator_mod + use amg_s_parmatch_aggregator_mod use amg_s_jac_smoother use amg_s_as_smoother use amg_s_diag_solver @@ -242,11 +243,13 @@ subroutine amg_s_base_onelev_csetc(lv,what,val,info,pos,idx) end if end if - select case(ival) - case(amg_dec_aggr_) + select case(val) + case('DEC') allocate(amg_s_dec_aggregator_type :: lv%aggr, stat=info) - case(amg_sym_dec_aggr_) + case('SYMDEC') allocate(amg_s_symdec_aggregator_type :: lv%aggr, stat=info) + case('COUP','COUPLED') + allocate(amg_s_parmatch_aggregator_type :: lv%aggr, stat=info) case default info = psb_err_internal_error_ end select diff --git a/amgprec/impl/level/amg_s_base_onelev_cseti.F90 b/amgprec/impl/level/amg_s_base_onelev_cseti.F90 index 7101117e..1211a662 100644 --- a/amgprec/impl/level/amg_s_base_onelev_cseti.F90 +++ b/amgprec/impl/level/amg_s_base_onelev_cseti.F90 @@ -1,15 +1,15 @@ -! -! +! +! ! AMG4PSBLAS version 1.0 ! Algebraic Multigrid Package ! based on PSBLAS (Parallel Sparse BLAS version 3.7) -! -! (C) Copyright 2021 -! -! Salvatore Filippone -! Pasqua D'Ambra -! Fabio Durastante -! +! +! (C) Copyright 2021 +! +! Salvatore Filippone +! Pasqua D'Ambra +! Fabio Durastante +! ! Redistribution and use in source and binary forms, with or without ! modification, are permitted provided that the following conditions ! are met: @@ -21,7 +21,7 @@ ! 3. The name of the AMG4PSBLAS group or the names of its contributors may ! not be used to endorse or promote products derived from this ! software without specific written permission. -! +! ! THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS ! ``AS IS'' AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED ! TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR @@ -33,15 +33,16 @@ ! CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ! ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE ! POSSIBILITY OF SUCH DAMAGE. -! -! +! +! subroutine amg_s_base_onelev_cseti(lv,what,val,info,pos,idx) - + use psb_base_mod use amg_s_onelev_mod, amg_protect_name => amg_s_base_onelev_cseti use amg_s_base_aggregator_mod use amg_s_dec_aggregator_mod use amg_s_symdec_aggregator_mod + use amg_s_parmatch_aggregator_mod use amg_s_jac_smoother use amg_s_as_smoother use amg_s_diag_solver @@ -59,13 +60,13 @@ subroutine amg_s_base_onelev_cseti(lv,what,val,info,pos,idx) Implicit None ! Arguments - class(amg_s_onelev_type), intent(inout) :: lv - character(len=*), intent(in) :: what + class(amg_s_onelev_type), intent(inout) :: lv + character(len=*), intent(in) :: what integer(psb_ipk_), intent(in) :: val integer(psb_ipk_), intent(out) :: info character(len=*), optional, intent(in) :: pos integer(psb_ipk_), intent(in), optional :: idx - ! Local + ! Local integer(psb_ipk_) :: ipos_, err_act character(len=20) :: name='s_base_onelev_cseti' type(amg_s_base_smoother_type) :: amg_s_base_smoother_mold @@ -84,7 +85,7 @@ subroutine amg_s_base_onelev_cseti(lv,what,val,info,pos,idx) #if defined(HAVE_MUMPS_) type(amg_s_mumps_solver_type) :: amg_s_mumps_solver_mold #endif - + call psb_erractionsave(err_act) info = psb_success_ @@ -100,14 +101,14 @@ subroutine amg_s_base_onelev_cseti(lv,what,val,info,pos,idx) else ipos_ = amg_smooth_both_ end if - + select case (psb_toupper(what)) case ('SMOOTHER_TYPE') - select case (val) + select case (val) case (amg_noprec_) call lv%set(amg_s_base_smoother_mold,info,pos=pos) if (info == 0) call lv%set(amg_s_id_solver_mold,info,pos=pos) - + case (amg_jac_) call lv%set(amg_s_jac_smoother_mold,info,pos=pos) if (info == 0) call lv%set(amg_s_diag_solver_mold,info,pos=pos) @@ -115,11 +116,11 @@ subroutine amg_s_base_onelev_cseti(lv,what,val,info,pos,idx) case (amg_l1_jac_) call lv%set(amg_s_jac_smoother_mold,info,pos=pos) if (info == 0) call lv%set(amg_s_l1_diag_solver_mold,info,pos=pos) - + case (amg_bjac_) call lv%set(amg_s_jac_smoother_mold,info,pos=pos) if (info == 0) call lv%set(amg_s_ilu_solver_mold,info,pos=pos) - + case (amg_l1_bjac_) call lv%set(amg_s_l1_jac_smoother_mold,info,pos=pos) if (info == 0) call lv%set(amg_s_ilu_solver_mold,info,pos=pos) @@ -133,61 +134,61 @@ subroutine amg_s_base_onelev_cseti(lv,what,val,info,pos,idx) if (info == 0) call lv%set(amg_s_gs_solver_mold,info,pos='pre') call lv%set(amg_s_jac_smoother_mold,info,pos='post') if (info == 0) call lv%set(amg_s_bwgs_solver_mold,info,pos='post') - + case default ! - ! Do nothing and hope for the best :) + ! Do nothing and hope for the best :) ! end select - if ((ipos_==amg_smooth_pre_).or.(ipos_==amg_smooth_both_)) then + if ((ipos_==amg_smooth_pre_).or.(ipos_==amg_smooth_both_)) then if (allocated(lv%sm)) call lv%sm%default() end if if ((ipos_==amg_smooth_post_).or.(ipos_==amg_smooth_both_)) then if (allocated(lv%sm2a)) call lv%sm2a%default() end if - + case('SUB_SOLVE') - select case (val) + select case (val) case (amg_f_none_) call lv%set(amg_s_id_solver_mold,info,pos=pos) - + case (amg_diag_scale_) call lv%set(amg_s_diag_solver_mold,info,pos=pos) - + case (amg_l1_diag_scale_) call lv%set(amg_s_l1_diag_solver_mold,info,pos=pos) - + case (amg_gs_) call lv%set(amg_s_gs_solver_mold,info,pos=pos) - + case (amg_bwgs_) call lv%set(amg_s_bwgs_solver_mold,info,pos=pos) - + case (psb_ilu_n_,psb_milu_n_,psb_ilu_t_) call lv%set(amg_s_ilu_solver_mold,info,pos=pos) if (info == 0) then - if ((ipos_==amg_smooth_pre_) .or.(ipos_==amg_smooth_both_)) then + if ((ipos_==amg_smooth_pre_) .or.(ipos_==amg_smooth_both_)) then call lv%sm%sv%set('SUB_SOLVE',val,info) end if - if ((ipos_==amg_smooth_post_).or.(ipos_==amg_smooth_both_))then + if ((ipos_==amg_smooth_post_).or.(ipos_==amg_smooth_both_))then if (allocated(lv%sm2a)) call lv%sm2a%sv%set('SUB_SOLVE',val,info) end if end if #ifdef HAVE_SLU_ - case (amg_slu_) + case (amg_slu_) call lv%set(amg_s_slu_solver_mold,info,pos=pos) #endif #ifdef HAVE_MUMPS_ - case (amg_mumps_) + case (amg_mumps_) call lv%set(amg_s_mumps_solver_mold,info,pos=pos) #endif case default ! - ! Do nothing and hope for the best :) + ! Do nothing and hope for the best :) ! end select - + case ('SMOOTHER_SWEEPS') if ((ipos_==amg_smooth_pre_) .or.(ipos_==amg_smooth_both_)) & @@ -208,7 +209,7 @@ subroutine amg_s_base_onelev_cseti(lv,what,val,info,pos,idx) return end if end if - + select case(val) case(amg_dec_aggr_) allocate(amg_s_dec_aggregator_type :: lv%aggr, stat=info) @@ -218,7 +219,7 @@ subroutine amg_s_base_onelev_cseti(lv,what,val,info,pos,idx) info = psb_err_internal_error_ end select if (info == psb_success_) call lv%aggr%default() - + case ('AGGR_ORD') lv%parms%aggr_ord = val @@ -245,13 +246,13 @@ subroutine amg_s_base_onelev_cseti(lv,what,val,info,pos,idx) lv%parms%coarse_solve = val case default - if ((ipos_==amg_smooth_pre_) .or.(ipos_==amg_smooth_both_)) then - if (allocated(lv%sm)) then + if ((ipos_==amg_smooth_pre_) .or.(ipos_==amg_smooth_both_)) then + if (allocated(lv%sm)) then call lv%sm%set(what,val,info,idx=idx) end if end if - if ((ipos_==amg_smooth_post_).or.(ipos_==amg_smooth_both_))then - if (allocated(lv%sm2a)) then + if ((ipos_==amg_smooth_post_).or.(ipos_==amg_smooth_both_))then + if (allocated(lv%sm2a)) then call lv%sm2a%set(what,val,info,idx=idx) end if end if diff --git a/amgprec/impl/level/amg_z_base_onelev_csetc.F90 b/amgprec/impl/level/amg_z_base_onelev_csetc.F90 index 3a49c635..47b76a24 100644 --- a/amgprec/impl/level/amg_z_base_onelev_csetc.F90 +++ b/amgprec/impl/level/amg_z_base_onelev_csetc.F90 @@ -262,10 +262,10 @@ subroutine amg_z_base_onelev_csetc(lv,what,val,info,pos,idx) end if end if - select case(ival) - case(amg_dec_aggr_) + select case(val) + case('DEC') allocate(amg_z_dec_aggregator_type :: lv%aggr, stat=info) - case(amg_sym_dec_aggr_) + case('SYMDEC') allocate(amg_z_symdec_aggregator_type :: lv%aggr, stat=info) case default info = psb_err_internal_error_ diff --git a/amgprec/impl/level/amg_z_base_onelev_cseti.F90 b/amgprec/impl/level/amg_z_base_onelev_cseti.F90 index a602dd66..b6a447a4 100644 --- a/amgprec/impl/level/amg_z_base_onelev_cseti.F90 +++ b/amgprec/impl/level/amg_z_base_onelev_cseti.F90 @@ -1,15 +1,15 @@ -! -! +! +! ! AMG4PSBLAS version 1.0 ! Algebraic Multigrid Package ! based on PSBLAS (Parallel Sparse BLAS version 3.7) -! -! (C) Copyright 2021 -! -! Salvatore Filippone -! Pasqua D'Ambra -! Fabio Durastante -! +! +! (C) Copyright 2021 +! +! Salvatore Filippone +! Pasqua D'Ambra +! Fabio Durastante +! ! Redistribution and use in source and binary forms, with or without ! modification, are permitted provided that the following conditions ! are met: @@ -21,7 +21,7 @@ ! 3. The name of the AMG4PSBLAS group or the names of its contributors may ! not be used to endorse or promote products derived from this ! software without specific written permission. -! +! ! THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS ! ``AS IS'' AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED ! TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR @@ -33,10 +33,10 @@ ! CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ! ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE ! POSSIBILITY OF SUCH DAMAGE. -! -! +! +! subroutine amg_z_base_onelev_cseti(lv,what,val,info,pos,idx) - + use psb_base_mod use amg_z_onelev_mod, amg_protect_name => amg_z_base_onelev_cseti use amg_z_base_aggregator_mod @@ -65,13 +65,13 @@ subroutine amg_z_base_onelev_cseti(lv,what,val,info,pos,idx) Implicit None ! Arguments - class(amg_z_onelev_type), intent(inout) :: lv - character(len=*), intent(in) :: what + class(amg_z_onelev_type), intent(inout) :: lv + character(len=*), intent(in) :: what integer(psb_ipk_), intent(in) :: val integer(psb_ipk_), intent(out) :: info character(len=*), optional, intent(in) :: pos integer(psb_ipk_), intent(in), optional :: idx - ! Local + ! Local integer(psb_ipk_) :: ipos_, err_act character(len=20) :: name='z_base_onelev_cseti' type(amg_z_base_smoother_type) :: amg_z_base_smoother_mold @@ -96,7 +96,7 @@ subroutine amg_z_base_onelev_cseti(lv,what,val,info,pos,idx) #if defined(HAVE_MUMPS_) type(amg_z_mumps_solver_type) :: amg_z_mumps_solver_mold #endif - + call psb_erractionsave(err_act) info = psb_success_ @@ -112,14 +112,14 @@ subroutine amg_z_base_onelev_cseti(lv,what,val,info,pos,idx) else ipos_ = amg_smooth_both_ end if - + select case (psb_toupper(what)) case ('SMOOTHER_TYPE') - select case (val) + select case (val) case (amg_noprec_) call lv%set(amg_z_base_smoother_mold,info,pos=pos) if (info == 0) call lv%set(amg_z_id_solver_mold,info,pos=pos) - + case (amg_jac_) call lv%set(amg_z_jac_smoother_mold,info,pos=pos) if (info == 0) call lv%set(amg_z_diag_solver_mold,info,pos=pos) @@ -127,11 +127,11 @@ subroutine amg_z_base_onelev_cseti(lv,what,val,info,pos,idx) case (amg_l1_jac_) call lv%set(amg_z_jac_smoother_mold,info,pos=pos) if (info == 0) call lv%set(amg_z_l1_diag_solver_mold,info,pos=pos) - + case (amg_bjac_) call lv%set(amg_z_jac_smoother_mold,info,pos=pos) if (info == 0) call lv%set(amg_z_ilu_solver_mold,info,pos=pos) - + case (amg_l1_bjac_) call lv%set(amg_z_l1_jac_smoother_mold,info,pos=pos) if (info == 0) call lv%set(amg_z_ilu_solver_mold,info,pos=pos) @@ -145,53 +145,53 @@ subroutine amg_z_base_onelev_cseti(lv,what,val,info,pos,idx) if (info == 0) call lv%set(amg_z_gs_solver_mold,info,pos='pre') call lv%set(amg_z_jac_smoother_mold,info,pos='post') if (info == 0) call lv%set(amg_z_bwgs_solver_mold,info,pos='post') - + case default ! - ! Do nothing and hope for the best :) + ! Do nothing and hope for the best :) ! end select - if ((ipos_==amg_smooth_pre_).or.(ipos_==amg_smooth_both_)) then + if ((ipos_==amg_smooth_pre_).or.(ipos_==amg_smooth_both_)) then if (allocated(lv%sm)) call lv%sm%default() end if if ((ipos_==amg_smooth_post_).or.(ipos_==amg_smooth_both_)) then if (allocated(lv%sm2a)) call lv%sm2a%default() end if - + case('SUB_SOLVE') - select case (val) + select case (val) case (amg_f_none_) call lv%set(amg_z_id_solver_mold,info,pos=pos) - + case (amg_diag_scale_) call lv%set(amg_z_diag_solver_mold,info,pos=pos) - + case (amg_l1_diag_scale_) call lv%set(amg_z_l1_diag_solver_mold,info,pos=pos) - + case (amg_gs_) call lv%set(amg_z_gs_solver_mold,info,pos=pos) - + case (amg_bwgs_) call lv%set(amg_z_bwgs_solver_mold,info,pos=pos) - + case (psb_ilu_n_,psb_milu_n_,psb_ilu_t_) call lv%set(amg_z_ilu_solver_mold,info,pos=pos) if (info == 0) then - if ((ipos_==amg_smooth_pre_) .or.(ipos_==amg_smooth_both_)) then + if ((ipos_==amg_smooth_pre_) .or.(ipos_==amg_smooth_both_)) then call lv%sm%sv%set('SUB_SOLVE',val,info) end if - if ((ipos_==amg_smooth_post_).or.(ipos_==amg_smooth_both_))then + if ((ipos_==amg_smooth_post_).or.(ipos_==amg_smooth_both_))then if (allocated(lv%sm2a)) call lv%sm2a%sv%set('SUB_SOLVE',val,info) end if end if #ifdef HAVE_SLU_ - case (amg_slu_) + case (amg_slu_) call lv%set(amg_z_slu_solver_mold,info,pos=pos) #endif #ifdef HAVE_MUMPS_ - case (amg_mumps_) + case (amg_mumps_) call lv%set(amg_z_mumps_solver_mold,info,pos=pos) #endif #ifdef HAVE_SLUDIST_ @@ -204,10 +204,10 @@ subroutine amg_z_base_onelev_cseti(lv,what,val,info,pos,idx) #endif case default ! - ! Do nothing and hope for the best :) + ! Do nothing and hope for the best :) ! end select - + case ('SMOOTHER_SWEEPS') if ((ipos_==amg_smooth_pre_) .or.(ipos_==amg_smooth_both_)) & @@ -228,7 +228,7 @@ subroutine amg_z_base_onelev_cseti(lv,what,val,info,pos,idx) return end if end if - + select case(val) case(amg_dec_aggr_) allocate(amg_z_dec_aggregator_type :: lv%aggr, stat=info) @@ -238,7 +238,7 @@ subroutine amg_z_base_onelev_cseti(lv,what,val,info,pos,idx) info = psb_err_internal_error_ end select if (info == psb_success_) call lv%aggr%default() - + case ('AGGR_ORD') lv%parms%aggr_ord = val @@ -265,13 +265,13 @@ subroutine amg_z_base_onelev_cseti(lv,what,val,info,pos,idx) lv%parms%coarse_solve = val case default - if ((ipos_==amg_smooth_pre_) .or.(ipos_==amg_smooth_both_)) then - if (allocated(lv%sm)) then + if ((ipos_==amg_smooth_pre_) .or.(ipos_==amg_smooth_both_)) then + if (allocated(lv%sm)) then call lv%sm%set(what,val,info,idx=idx) end if end if - if ((ipos_==amg_smooth_post_).or.(ipos_==amg_smooth_both_))then - if (allocated(lv%sm2a)) then + if ((ipos_==amg_smooth_post_).or.(ipos_==amg_smooth_both_))then + if (allocated(lv%sm2a)) then call lv%sm2a%set(what,val,info,idx=idx) end if end if diff --git a/docs/html/index.html b/docs/html/index.html index 2d1ac7b7..18ae1e4d 100644 --- a/docs/html/index.html +++ b/docs/html/index.html @@ -105,75 +105,80 @@ class="cmr-12"> 4.1 Examples +
 4.2 GPU example
5 User Interface
 5.1 Method init
 5.2 Method set
 5.3 Method hierarchy_build
 5.4 Method smoothers_build
 5.5 Method build
 5.6 Method apply
 5.7 Method free
 5.8 Method descr
 5.9 Auxiliary Methods
6 Adding new smoother and solver objects to AMG4PSBLAS
7 Error Handling
A License
B Contributor Covenant Code of Conduct
References diff --git a/docs/html/userhtml.html b/docs/html/userhtml.html index 2d1ac7b7..18ae1e4d 100644 --- a/docs/html/userhtml.html +++ b/docs/html/userhtml.html @@ -105,75 +105,80 @@ class="cmr-12"> 4.1 Examples +
 4.2 GPU example
5 User Interface
 5.1 Method init
 5.2 Method set
 5.3 Method hierarchy_build
 5.4 Method smoothers_build
 5.5 Method build
 5.6 Method apply
 5.7 Method free
 5.8 Method descr
 5.9 Auxiliary Methods
6 Adding new smoother and solver objects to AMG4PSBLAS
7 Error Handling
A License
B Contributor Covenant Code of Conduct
References diff --git a/docs/html/userhtmlli1.html b/docs/html/userhtmlli1.html index fc40c5d3..dcc83f43 100644 --- a/docs/html/userhtmlli1.html +++ b/docs/html/userhtmlli1.html @@ -77,11 +77,11 @@ class="cmr-12">PSCToolkit (Parallel Sparse Computation Toolkit) software framewo class="cmr-12">of a software development project started in 2007, named MLD2P4, which originally implemented a multilevel version of some domain decomposition preconditioners of -additive-Schwarz type, and was based on a parallel decoupled version of the well known +additive-Schwarz type, and was based on a parallel decoupled version of the well known smoothed aggregation method to generate the multilevel hierarchy of coarser  4.1 Examples +
 4.2 GPU example
5 User Interface
 5.1 Method init
 5.2 Method set
 5.3 Method hierarchy_build
 5.4 Method smoothers_build
 5.5 Method build
 5.6 Method apply
 5.7 Method free
 5.8 Method descr
 5.9 Auxiliary Methods
  5.9.1 Method: dump
  5.9.2 Method: clone
  5.9.3 Method: sizeof
  5.9.4 Method: allocate_wrk
  5.9.5 Method: free_wrk
6 Adding new smoother and solver objects to AMG4PSBLAS
7 Error Handling
A License
B Contributor Covenant Code of Conduct diff --git a/docs/html/userhtmlli5.html b/docs/html/userhtmlli5.html index b9362d17..d712d8cf 100644 --- a/docs/html/userhtmlli5.html +++ b/docs/html/userhtmlli5.html @@ -25,7 +25,7 @@ href="userhtml.html#userhtmlli5.html" >up]

References

diff --git a/docs/html/userhtmlse1.html b/docs/html/userhtmlse1.html index f7af8224..f0b32b47 100644 --- a/docs/html/userhtmlse1.html +++ b/docs/html/userhtmlse1.html @@ -299,7 +299,7 @@ class="cmr-12">smoothers and solvers for building new versions of the preconditi Section 6).

, where the Fortran codes of a few sample programs are also shown A reference guide for the user interface routines is provided in Section 5. Information on the extension of the package through the addition smoothers and solvers is reported in Section 6. The error handling mechanism used by the package is briefly described in Section 7. The copyright terms concerning the distribution and modification of AMG4PSBLAS are re Appendix A. diff --git a/docs/html/userhtmlse2.html b/docs/html/userhtmlse2.html index 8c0e20d1..03862797 100644 --- a/docs/html/userhtmlse2.html +++ b/docs/html/userhtmlse2.html @@ -42,7 +42,7 @@ class="cmr-12">where contact points for further information can be also found.

The software is available under a modified BSD license, as specified in Appendix A; list of all the preconditioner parameters and their allowed and d is provided in Section 5, Tables 2-8. @@ -208,7 +208,7 @@ class="cmr-12">All the previous routines are available as methods of the precond detailed description of them is given in Section 5. Examples showing the basic use of 2-8 for further details of +href="userhtmlsu9.html#x19-18009r2">2-8 for further details of the preconditioner.

 4.1 Examples +
 4.2 GPU example
+ diff --git a/docs/html/userhtmlse5.html b/docs/html/userhtmlse5.html index 8f6ba4e7..a7e1fb0e 100644 --- a/docs/html/userhtmlse5.html +++ b/docs/html/userhtmlse5.html @@ -29,7 +29,7 @@ class="cmr-12">up]

5 User Interface

The basic user interface of AMG4PBLAS consists of eight methods. The six methods @@ -151,7 +151,7 @@ class="cmr-12">the precision of the sparse matrix and preconditioner data struct Section 5.2).

A description of each method is given in the remainder of this se  5.1 Method init
 5.2 Method set
 5.3 Method hierarchy_build
 5.4 Method smoothers_build
 5.5 Method build
 5.6 Method apply
 5.7 Method free
 5.8 Method descr
 5.9 Auxiliary Methods
  5.9.1 Method: dump
  5.9.2 Method: clone
  5.9.3 Method: sizeof
  5.9.4 Method: allocate_wrk
  5.9.5 Method: free_wrk diff --git a/docs/html/userhtmlse6.html b/docs/html/userhtmlse6.html index 5cbf8ab8..6282c36f 100644 --- a/docs/html/userhtmlse6.html +++ b/docs/html/userhtmlse6.html @@ -29,7 +29,7 @@ class="cmr-12">up]

6 Adding new smoother and solver objects to AMG4PSBLAS

Developers can add completely new smoother and/or solver classes derived from the @@ -37,7 +37,7 @@ class="cmr-12">Developers can add completely new smoother and/or solver classes class="cmr-12">base objects in the library (see Remark 2 in Section 5.2), without recompiling the The user-defined new solver to be employed in the preconditioner.

The other arguments are defined in the way described in Sec. 5.2. As an example, in the pass it as follows: +<<<<<<< HEAD 

+=======
+   
+>>>>>>> development
   ! sparse matrix and preconditioner
   type(psb_dspmat_type) :: a
   type(amg_dprec_type)  :: prec
diff --git a/docs/html/userhtmlse7.html b/docs/html/userhtmlse7.html
index 20a74c45..df4dc649 100644
--- a/docs/html/userhtmlse7.html
+++ b/docs/html/userhtmlse7.html
@@ -29,7 +29,7 @@ class="cmr-12">up] 

    
7    Error Handling

 
The error handling in AMG4PSBLAS is based on the PSBLAS error handling. Error
diff --git a/docs/html/userhtmlse8.html b/docs/html/userhtmlse8.html
index c719a48e..e322b719 100644
--- a/docs/html/userhtmlse8.html
+++ b/docs/html/userhtmlse8.html
@@ -29,14 +29,18 @@ class="cmr-12">up] 

    
A    License

 
AMG4PSBLAS is freely distributable under the following copyright terms:
                                                                                
 
                                                                                
+<<<<<<< HEAD
    
+=======
+   
+>>>>>>> development
                            AMG4PSBLAS  version 1.0
               Algebraic MultiGrid Preconditioners Package
              based on PSBLAS (Parallel Sparse BLAS version 3.7)
@@ -78,7 +82,11 @@ class="cmr-12">abide by its terms:
                                                                                
 
                                                                                
+<<<<<<< HEAD
    
+=======
+   
+>>>>>>> development
                            MLD2P4  version 2.2
   MultiLevel Domain Decomposition Parallel Preconditioners Package
              based on PSBLAS (Parallel Sparse BLAS version 3.5)
@@ -125,6 +133,7 @@ class="cmr-12">. The relevant license is reproduced here:
                                                                                
 
                                                                                
+<<<<<<< HEAD
    
 // ***********************************************************************
 //
@@ -134,6 +143,17 @@ class="cmr-12">. The relevant license is reproduced here:
 //
 // ***********************************************************************
 //
+=======
+   
+// ***********************************************************************
+//
+//        MatchboxP: A C++ library for approximate weighted matching
+//               Mahantesh Halappanavar (hala@pnnl.gov)
+//               Pacific Northwest National Laboratory
+//
+// ***********************************************************************
+//
+>>>>>>> development
 //       Copyright (2021) Battelle Memorial Institute
 //                      All rights reserved.
 //
diff --git a/docs/html/userhtmlse9.html b/docs/html/userhtmlse9.html
index 5a1864d1..3db740b8 100644
--- a/docs/html/userhtmlse9.html
+++ b/docs/html/userhtmlse9.html
@@ -29,12 +29,12 @@ class="cmr-12">up] 

    
B    Contributor Covenant Code of Conduct

    

    

 
Our Pledge
    We as members, contributors, and leaders pledge to make participation in
@@ -100,7 +100,7 @@ class="cmr-12">Other  conduct  which  could  reasonably  be  considered  inappro
      professional setting
 
Enforcement Responsibilities
    Community leaders are responsible for clarifying and enforcing our standards of
@@ -138,7 +138,7 @@ class="cmr-12">leaders are obligated to respect the privacy and security of the
 incident.
 
Enforcement Guidelines
    Community leaders will follow these Community Impact Guidelines in
@@ -148,7 +148,7 @@ class="cmr-12">determining the consequences for any action they deem in violatio
 class="cmr-12">Conduct:
      
    
      
  1. Correction
      
    Community Impact: clarity around the nature of the violation and an explanation of
 class="cmr-12">behavior was inappropriate. A public apology may be requested.
      
    2. 
      
  2. Warning
      
    Community Impact: channels like social media. Violating these terms may lead to a t
 class="cmr-12">or permanent ban.
      
    3. 
      
  3. Temporary Ban
      
    Community Impact: interaction with those enforcing the Code of Conduct, is allowed
 class="cmr-12">period. Violating these terms may lead to a permanent ban.
      
    4. 
      
  4. Permanent Ban
      
    Community Impact:  A permanent ban from any sort of public interaction within
      the community.

 
Attribution
    This Code of Conduct is adapted from the Contributor Covenant, version 2.0,
diff --git a/docs/html/userhtmlsu1.html b/docs/html/userhtmlsu1.html
index d3d37856..6946e556 100644
--- a/docs/html/userhtmlsu1.html
+++ b/docs/html/userhtmlsu1.html
@@ -124,10 +124,39 @@ class="cmr-12">Please note that the four previous libraries must have Fortran in
 class="cmr-12">AMG4PSBLAS; usually this means that they should all be built with the same
 compiler being used for AMG4PSBLAS.
+
   If you want to use the PSBLAS support for NVIDIA GPUs, you will also
+need:
+     

+PSBLAS-EXT 
Parallel  Sparse  BLAS  (PSBLAS)  Extensions,  available  from
+     psctoolkit.github.io/products/psblasext/; version 1.3.0 (or later).
+     

+SPGPU 
Sparse CUDA kernels for NVIDIA GPUs; available from GitHub, see
+     also psctoolkit.github.io/products/psblasext/.

+
See also Sec 4.2.
                                                                                
 
                                                                                
-   
[next] [up] 

-
   
     
  
diff --git a/docs/html/userhtmlsu10.html b/docs/html/userhtmlsu10.html
index caba46d1..ced70cf9 100644
--- a/docs/html/userhtmlsu10.html
+++ b/docs/html/userhtmlsu10.html
@@ -1,7 +1,7 @@
   
  
-Method smoothers_build 
+Method hierarchy_build 
  
  
  
@@ -10,7 +10,7 @@
  
 
-   
[nextup] 

    
5.4    Method smoothers5.3    Method hierarchy_build

 

-

-

+
call p%smoothers_buildhierarchy_build(a,desc_a,p,info[,amold,vmold,imold])

+class="cmtt-12">)

 

-
This method builds the smoothers and the coarsest-level solvers for the multilevel
+
This method builds the hierarchy of matrices and restriction/prolongation operators for
 preconditioner the multilevel preconditioner p, according to the requirements made by the user through the methods
-, according to the requirements made by the user
+through the methods init and set, and based on the aggregation hierarchy produced by a previous call to
-hierarchy_build (see Section 5.3).
-
.
+
Arguments
-   
 
 
a     
a     
>>>>>> development
 class="cmtt-12">type(psb_in).                                  

 
      
      
>>>>>> development
 class="cmr-12">The sparse matrix structure containing the local part of the matrix
 to be preconditioned. Note that 20].            

 
desc_a
desc_a
>>>>>> development
 class="cmtt-12">type(psb_desc_typein).                                     

 
      
      
>>>>>> development
 class="cmr-12">The communication descriptor of a. See the PSBLAS User’s Guide
@@ -149,10 +164,17 @@ class="cmr-12">20].                                                                    

 
info  
info  
>>>>>> development
 class="cmtt-12">integer, out).                                                       

 
      
      
>>>>>> development
 class="cmr-12">Error code. If no error, 0 is returned. See Section 7 for details.     

 
amold 
The desired dynamic type for internal integer vector components;<
 class="cmr-12">this allows e.g. running on GPUs.                                           
 
>>>>>> development
 class="td11">       

                                                                                
 
@@ -262,7 +294,7 @@ class="td11">       

                                                                                
 
                                                                                
-   
[next] [up] 

-
   
     
  
diff --git a/docs/html/userhtmlsu11.html b/docs/html/userhtmlsu11.html
index 7a182b68..c9fb0871 100644
--- a/docs/html/userhtmlsu11.html
+++ b/docs/html/userhtmlsu11.html
@@ -1,7 +1,7 @@
   
  
-Method build 
+Method smoothers_build 
  
  
  
@@ -10,7 +10,7 @@
  
 
-   
[nextup] 

    
5.5    Method build

+class="cmr-12">5.4    Method smoothers_build
 

-

-

+
call p%buildsmoothers_build(a,desc_a,p,info[,amold,imold])

 

-
This method builds the preconditioner p according to the requirements made by the
+
This method builds the smoothers and the coarsest-level solvers for the multilevel
 user through the methods preconditioner p, according to the requirements made by the user through the methods
+init and set (see Sections 5.3 and 5.4 for multilevel
-preconditioners). It is mostly provided for backward compatibility; indeed, it is
-internally implemented by invoking the two previous methods hierarchy_build and
+class="cmtt-12">set, and based on the aggregation hierarchy produced by a previous call to
 smoothers_build, whose nomenclature would however be somewhat unnatural when
-dealing with simple one-level preconditioners.
-
hierarchy_build (see Section 5.3).
+
Arguments
-   
 
 
a     
a     
>>>>>> development
 class="cmtt-12">type(psb_in).                                  

 
      
      
>>>>>> development
 class="cmr-12">The sparse matrix structure containing the local part of the matrix
 to be preconditioned. Note that 20].            

 
desc_a
desc_a
>>>>>> development
 class="cmtt-12">type(psb_desc_typein).                                     

 
      
      
>>>>>> development
 class="cmr-12">The communication descriptor of a. See the PSBLAS User’s Guide
@@ -157,10 +178,17 @@ class="cmr-12">20].                                                                    

 
info  
info  
>>>>>> development
 class="cmtt-12">integer, out).                                                       

 
      
      
>>>>>> development
 class="cmr-12">Error code. If no error, 0 is returned. See Section 7 for details.     

 
amold 
amold 
>>>>>> development
 class="cmtt-12">class(psb_ optional.       

 
      
      
>>>>>> development
 class="cmr-12">The  desired  dynamic  type  for  internal  matrix  components;  this
 allows  e.g.  running  on  GPUs;  it  needs  not  be  the  same  on  all
@@ -212,10 +259,17 @@ class="cmr-12">20].             

 
vmold 
vmold 
>>>>>> development
 class="cmtt-12">class(psb_ optional.         

 
      
      
>>>>>> development
 class="cmr-12">The  desired  dynamic  type  for  internal  vector  components;  this
 allows e.g. running on GPUs.                                                 

 
imold 
imold 
>>>>>> development
 class="cmtt-12">class(psb_i_base_vect_type optional.         

 
      
      
>>>>>> development
 class="cmr-12">The desired dynamic type for internal integer vector components;
 this allows e.g. running on GPUs.                                           

 
>>>>>> development
 class="td11">       

-
The method can be used to build multilevel preconditioners too.
                                                                                
 
                                                                                
                                                                                
 
                                                                                
-   
[next] [up] 

-
   
     
  
diff --git a/docs/html/userhtmlsu12.html b/docs/html/userhtmlsu12.html
index cae2bdd9..cdb8bf0d 100644
--- a/docs/html/userhtmlsu12.html
+++ b/docs/html/userhtmlsu12.html
@@ -1,7 +1,7 @@
   
  
-Method apply 
+Method build 
  
  
  
@@ -10,7 +10,7 @@
  
 
-   
[nextup] 

    
5.6    Method apply

+class="cmr-12">5.5    Method build
 

-

-

+
call p%applybuild(x,ya,desc_a,info [,transamold,vmold,workimold])

 

-
This method computes y = op(B-1) x, where B is a previously built preconditioner,
+
This method builds the preconditioner p according to the requirements made by the
 stored into p, and op denotes the preconditioner itself or its transpose, according to the
+class="cmr-12">user through the methods init and set (see Sections 5.3 and 5.4 for multilevel
 value of trans. Note that, when AMG4PSBLAS is used with a Krylov solver from
+class="cmr-12">preconditioners). It is mostly provided for backward compatibility; indeed, it is
 PSBLAS, p%apply is called within the PSBLAS method psb_krylov and hence it is
+class="cmr-12">internally implemented by invoking the two previous methods hierarchy_build and
+smoothers_build, whose nomenclature would however be somewhat unnatural when
 completely transparent to the user.
-
dealing with simple one-level preconditioners.
+
Arguments
-   
 
 
y     

+class="cmr-12">.                                  
 
+class="cmr-12">under use. See the PSBLAS User’s Guide for details [20].            
 
+class="cmr-12">for details [20].                                                                    
 
+class="cmtt-12">), optional.       
 
+class="cmr-12">processes. See the PSBLAS User’s Guide for details [20].             
 
+
+
+
+class="cmr-12">this allows e.g. running on GPUs.                                           
 
type
a     
type(kind_parameterpsb_xspmat_type), intent>>>>>> development
+class="cmtt-12">(in), dimension(:), intent(out)—.                  
      
type and kind_parameter
+=======
+ style="vertical-align:baseline;" id="TBL-13-2-">
      
The sparse matrix structure containing the local part of the matrix
+>>>>>>> development
 must  be  chosen  according  to  the  to be preconditioned. Note that x must be chosen according to the
+real/complex,  single/double
+class="cmr-12">, single/double precision version of AMG4PSBLAS
 precision version of AMG4PSBLAS under use.                          
desc_a
desc_a
>>>>>> development
 class="cmtt-12">type(psb_desc_typein).                                     

 
      
The  communication  descriptor  associated  to  the  matrix  to  be
+=======
+ style="vertical-align:baseline;" id="TBL-13-4-">
      
The communication descriptor of a. See the PSBLAS User’s Guide
+>>>>>>> development
 preconditioned.                                                                   
info  
info  
>>>>>> development
 class="cmtt-12">integer, out).                                                       

 
      
      
>>>>>> development
 class="cmr-12">Error code. If no error, 0 is returned. See Section 7 for details.     

 
trans 
character
amold 
class>>>>>> development
 class="cmtt-12">(len=1), optional,psb_x_base_sparse_mat), intent(in).                
      
,’t’
+=======
+ style="vertical-align:baseline;" id="TBL-13-8-">
      
The  desired  dynamic  type  for  internal  matrix  components;  this
+>>>>>>> development
 then op(B-1) = B-T  (transpose of B-1); if trans = ’C’,’c’ then
+class="cmr-12">allows  e.g.  running  on  GPUs;  it  needs  not  be  the  same  on  all
 op(B-1) = B-C (conjugate transpose of B-1).                           
work  
Workspace.  Its  size  should  be  at  least  4 *
vmold 
class(psb_x_base_vect_type),>>>>>> development
 class="cmtt-12"> psb_cd_get_local_
-colsintent(desc_a) (see the PSBLAS User’s Guide). Note that type and
+class="cmtt-12">in), optional.         
      
The  desired  dynamic  type  for  internal  vector  components;  this
 kind_parameter  must be chosen according to the real/complex,
+class="cmr-12">allows e.g. running on GPUs.                                                 
imold 
class(psb_i_base_vect_type), intent(in), optional.         
      
The desired dynamic type for internal integer vector components;
 single/double precision version of AMG4PSBLAS under use.        
>>>>>> development
 class="td11">       

+
The method can be used to build multilevel preconditioners too.
                                                                                
 
                                                                                
                                                                                
 
                                                                                
-                                                                               
-
-                                                                               
-   
[next] [up] 

-
   
     
  
diff --git a/docs/html/userhtmlsu13.html b/docs/html/userhtmlsu13.html
index 9ab70af9..215abe6e 100644
--- a/docs/html/userhtmlsu13.html
+++ b/docs/html/userhtmlsu13.html
@@ -1,7 +1,7 @@
   
  
-Method free 
+Method apply 
  
  
  
@@ -10,7 +10,7 @@
  
 
-   
[nextup] 

    
5.7    Method free

+class="cmr-12">5.6    Method apply
 

-

-

+
call p%freeapply(px,y,desc_a,info)

+class="cmtt-12"> [,trans,work])

 

-
This method deallocates the preconditioner data structure 
This method computes y = op(B-1) x, where B is a previously built preconditioner,
+stored into p.
-
, and op denotes the preconditioner itself or its transpose, according to the
+value of trans. Note that, when AMG4PSBLAS is used with a Krylov solver from
+PSBLAS, p%apply is called within the PSBLAS method psb_krylov and hence it is
+completely transparent to the user.
+
Arguments
-   
 
 
info  

+
+
+
+
+
+
+class="cmr-12">.
+
+
+
+
+class="cmti-12">kind_parameter  must be chosen according to the real/complex,
+single/double precision version of AMG4PSBLAS under use.
x     
type(kind_parameter), dimension(:), intent(in)—.                    
      
The local part of the vector x. Note that type and kind_parameter
+must  be  chosen  according  to  the  real/complex,  single/double
+precision version of AMG4PSBLAS under use.                          
y     
type(kind_parameter), dimension(:), intent(out)—.                  
      
The local part of the vector y. Note that type and kind_parameter
+must  be  chosen  according  to  the  real/complex,  single/double
+precision version of AMG4PSBLAS under use.                          
desc_a
type(psb_desc_type), intent(in).                                     
      
The  communication  descriptor  associated  to  the  matrix  to  be
+preconditioned.                                                                   
info  
>>>>>> development
 class="cmtt-12">integer, intent(out).                                            
                                                       
 
      
      
>>>>>> development
 class="cmr-12">Error code. If no error, 0 is returned. See Section 7 for
+class="cmr-12">for details.     
trans 
character(len=1), optional, intent(in).                
      
If trans = ’N’,’n’ then op(B-1)  =  B-1; if trans = ’T’,’t’
+then op(B-1) = B-T  (transpose of B-1); if trans = ’C’,’c’ then
+op(B-1) = B-C (conjugate transpose of B-1).                           
work  
type(kind_parameter), dimension(:), optional, target—.            
      
Workspace.  Its  size  should  be  at  least  4 * psb_cd_get_local_
+cols(desc_a) (see the PSBLAS User’s Guide). Note that type and
 details.                                                                   
        
 
>>>>>> development
 class="td11">       

                                                                                
 
@@ -89,7 +356,10 @@ class="td11">       

                                                                                
 
                                                                                
-   
[next] [up] 

-
   
     
  
diff --git a/docs/html/userhtmlsu14.html b/docs/html/userhtmlsu14.html
index 4c399a06..0204408d 100644
--- a/docs/html/userhtmlsu14.html
+++ b/docs/html/userhtmlsu14.html
@@ -1,7 +1,7 @@
   
  
-Method descr 
+Method free 
  
  
  
@@ -10,7 +10,7 @@
  
 
-   
[nextup] 

    
5.8    Method descr

+class="cmr-12">5.7    Method free
 

-

-

+
call p%descrfree(info, [iout, rootp, verbosity])

+class="cmtt-12">info)

 

-
This method prints a description of the preconditioner p to the standard output or to a
-file. It must be called after hierachy_build and smoothers_build, or build, have
-been called.
-
This method deallocates the preconditioner data structure p.
+
Arguments
-   
 
 
info      

+class="cmr-12">.
+class="cmr-12">details.
info  
>>>>>> development
 class="cmtt-12">integer, intent(out).                                                
                                            
 
           
      
>>>>>> development
 class="cmr-12">Error code. If no error, 0 is returned. See Section 7 for details.

 
The verbosity level of the description. Default value is 0. For
 values higher than 0, it prints out further information, e.g., for
+=======
+class="cmr-12">for
+>>>>>>> development
 a distributed multilevel preconditioner the size of the coarse
-matrices on every process.                                              
                                                                   
 
            

+=======
+ style="vertical-align:baseline;" id="TBL-15-3-">       

+>>>>>>> development
                                                                                
 
                                                                                
-   
[next] [up] 

-
   
     
  
diff --git a/docs/html/userhtmlsu15.html b/docs/html/userhtmlsu15.html
index 9179e431..fd563dcb 100644
--- a/docs/html/userhtmlsu15.html
+++ b/docs/html/userhtmlsu15.html
@@ -1,7 +1,7 @@
   
  
-Auxiliary Methods 
+Method descr 
  
  
  
@@ -10,8 +10,11 @@
  
 
-   
[next] [prev] [up] 

    
5.9    Auxiliary Methods

-
Various functionalities are implemented as additional methods of the preconditioner
-object.
-

-   
5.9.1    Method: dump

+class="cmr-12">5.8    Method descr
 

-

-

+
call p%dumpdescr(info[,istart,iend,prefix,head,ac,rp,smoother [iout,solver root,global_num verbosity])

 

-
Dump on file.
-
This method prints a description of the preconditioner p to the standard output or to a
+file. It must be called after hierachy_build and smoothers_build, or build, have
+been called.
+
Arguments
-   
 
 
info  

+class="cmr-12">.
+class="cmr-12">for details.
+class="cmr-12">.
+class="cmr-12">printed; the default is the standard output.
info      
>>>>>> development
 class="cmtt-12">integer, intent(out).                                                       
                                                
 
      
           
>>>>>> development
 class="cmr-12">Error code. If no error, 0 is returned. See Section 7 for details.     

 
amold 
psb_x_base_sparse_mat),
iout       
integer,>>>>>> development
 class="cmtt-12"> intent(in), optional.       
                                 
 
      
The  desired  dynamic  type  for  internal  matrix  components;  this
+=======
+ style="vertical-align:baseline;" id="TBL-16-4-">
           
The id of the file where the preconditioner description will be
+>>>>>>> development
 allows  e.g.  running  on  GPUs;  it  needs  not  be  the  same  on  all
-processes. See the PSBLAS User’s Guide for details [20].             
                       
 
       

 

@@ -202,11 +226,18 @@ class="cmr-12">version of AMG4PSBLAS under use.
 class="td11">
info  
root       
>>>>>> development
 class="cmtt-12">integer, intent(out).                                                       
@@ -253,21 +284,31 @@ class="cmtt-12">global  
logical,in),>>>>>> development
 class="cmtt-12"> optional.                                                            
+class="cmr-12">.                                 
 
      
Whether the global or local preconditioner memory occupation is
+=======
+ style="vertical-align:baseline;" id="TBL-16-6-">
           
The id of the process where the preconditioner description
+>>>>>>> development
 desired. Default: .false..                                                     
+class="cmr-12">will be printed; the default is psb_root_.                          
        
 
 
info  
verbosity  
>>>>>> development
 class="cmtt-12">integer, intent(out).                                                       
@@ -368,12 +416,15 @@ class="cmtt-12">), intent(>>>>>> development
 class="cmtt-12">in), optional.         
+class="cmr-12">.                                 
 
      
for details.     
        

+=======
+ style="vertical-align:baseline;" id="TBL-16-8-">
           
The verbosity level of the description. Default value is 0. For
+values higher than 0, it prints out further information, e.g., for
+a distributed multilevel preconditioner the size of the coarse
+matrices on every process.                                              
+            

+>>>>>>> development
                                                                                
 
                                                                                
-                                                                               
-
-                                                                               
-                                                                               
-
-                                                                               
-   
[next] [prev] [] [up] 

-
      
+
     
  
diff --git a/docs/html/userhtmlsu16.html b/docs/html/userhtmlsu16.html
index 97965c2b..99eaf060 100644
--- a/docs/html/userhtmlsu16.html
+++ b/docs/html/userhtmlsu16.html
@@ -1,20 +1,17 @@
   
  
-Method free 
+Auxiliary Methods 
  
- 
- 
+ 
+ 
  
  
  
 
-   
[next] [prev] [] [tail] [up] 

    
6.7    Method free

+class="cmr-12">5.9    Auxiliary Methods
+
Various functionalities are implemented as additional methods of the preconditioner
+object.
+

+   
5.9.1    Method: dump

 

-

-

+
call p%free(p,info)

+class="cmtt-12"> p%dump(info[,istart,iend,prefix,head,ac,rp,smoother,solver,global_num])

 

-
This method deallocates the preconditioner data structure p.
-
Dump on file.
+
Arguments
-   
 
 

-
+
+
+
+class="cmr-12">allows  e.g.  running  on  GPUs;  it  needs  not  be  the  same  on  all
+processes. See the PSBLAS User’s Guide for details [20].
info  
integer, intent(out).                                            
      
info  
integer, intent(out).                                                       
      
Error code. If no error, 0 is returned. See Section 8 for
+href="userhtmlse7.html#x28-320007">7 for details.     
amold 
class(psb_x_base_sparse_mat), intent(in), optional.       
      
The  desired  dynamic  type  for  internal  matrix  components;  this
 details.                                                                   
             
 
       

+

+   
5.9.2    Method: clone

+

+

+
call p%clone(pout,info)

+

                                                                                
 
                                                                                
+
Create a (deep) copy of the preconditioner object.
+
Arguments
+   
 
+
+
+
+
pout  
type(amg_xprec_type), intent(out).                                  
      
The copy of the preconditioner data structure. Note that x must
+be chosen according to the real/complex, single/double precision
+version of AMG4PSBLAS under use.                                       
info  
integer, intent(out).                                                       
      
Error code. If no error, 0 is returned. See Section 7 for details.     
       

+

+   
5.9.3    Method: sizeof

+

+

+
sz = p%sizeof([global])

+

+   
 
+
+
global 
 
logical, optional.                                                            
      
Whether the global or local preconditioner memory occupation is
+desired. Default: .false..                                                     
       

+
 Return memory footprint in bytes.
+

+   
5.9.4    Method: allocate_wrk

+

+

+
call p%allocate_wrk(info[, vmold])

+

+
Allocate internal work vectors. Each application of the preconditioner uses a number of
+work vectors which are allocated internally as necessary; therefore allocation and
+deallocation of memory occurs multiple times during the execution of a Krylov method.
+In most cases this strategy is perfectly acceptable, but on some platforms, most
+notably GPUs, memory allocation is a slow operation, and the default behaviour would
+lead to a slowdown. This method allows to trade space for time by preallocating
+the internal workspace outside of the invocation of a Krylov method. When
                                                                                
 
                                                                                
-   
using GPUs or other specialized devices, the vmold argument is also necessary
+to ensure the internal work vectors are of the appropriate dynamic type to
+exploit the accelerator hardware; when allocation occurs internally this is
+taken care of based on the dynamic type of the x argument to the apply
+method.
+
Arguments
+   
 
+
+
+
+
info  
integer, intent(out).                                                       
      
Error code. If no error, 0 is returned. See Section 7 for details.     
vmold 
class(psb_x_base_vect_type), intent(in), optional.         
      
The  desired  dynamic  type  for  internal  vector  components;  this
+allows e.g. running on GPUs.                                                 
       

+

+   
5.9.5    Method: free_wrk

+

+

+
call p%free_wrk(info)

+

+
Deallocate internal work vectors.
+
Arguments
+   
 
+
+
info  
integer, intent(out).                                                       
      
Error code. If no error, 0 is returned. See Section 7 for details.     
       

+                                                                               
+
+                                                                               
+                                                                               
+
+                                                                               
+                                                                               
+
+                                                                               
+   
[next] [prev] [] [front] [up] 

-
     
+
      
  
diff --git a/docs/html/userhtmlsu2.html b/docs/html/userhtmlsu2.html
index 64e76ece..e73a0092 100644
--- a/docs/html/userhtmlsu2.html
+++ b/docs/html/userhtmlsu2.html
@@ -10,7 +10,7 @@
  
 
-   
[next] 

 class="cmr-12">3.2    Optional third party libraries
-
We provide interfaces to the following third-party software libraries; note that these are
 optional, but if you enable them some defaults for multilevel preconditioners may
 change to reflect their presence.
-

+

      

 UMFPACK 
.

                                                                                
 
                                                                                
-   
[next] [up] 

-
   
     
  
diff --git a/docs/html/userhtmlsu3.html b/docs/html/userhtmlsu3.html
index 4f2dd495..36b6cf45 100644
--- a/docs/html/userhtmlsu3.html
+++ b/docs/html/userhtmlsu3.html
@@ -10,7 +10,7 @@
  
 
-   
[next] 

 class="cmr-12">3.3    Configuration options
-
In order to build AMG4PSBLAS, the first step is to use the configure script in the
 main directory to generate the necessary makefile.
-
   
   As a minimal example consider the following:
                                                                                
 
@@ -45,7 +45,11 @@ class="cmr-12">As a minimal example consider the following:
    
 ./configure --with-psblas=PSB-INSTALL-DIR
 

+<<<<<<< HEAD
 
 
 >>>>>> development
 class="cmr-12">which assumes that the various MPI compilers and support libraries are available in
 the standard directories on the system, and specifies only the PSBLAS install directory
@@ -59,7 +63,7 @@ class="cmtt-12">./configure --help, which
 produces: 

+class="cmr-12">produces: 

 
    issues>.
    
    

-
For instance, if a user has built and installed PSBLAS 3.7 under the /opt directory and is
@@ -4251,7 +4255,11 @@ class="cmr-12">might be configured with:
 ./configure --with-psblas=/opt/psblas-3.7/ \
 --with-umfpackincdir=/usr/include/suitesparse/
 

+<<<<<<< HEAD
 
 
 >>>>>> development
 class="cmr-12">Once the configure script has completed execution, it will have generated the file
 Make.inc which will then be used by all Makefiles in the directory tree; t
 class="cmr-12">copied in the install directory under the name Make.inc.AMG4PSBLAS.
-
   
   To use the MUMPS solver package, the user has to add the appropriate options to
 the configure script; by default we are looking for the libraries --with-extra-libs configure
 option.
-
   
   To build the library the user will now enter
                                                                                
 
@@ -4288,7 +4296,11 @@ class="cmr-12">To build the library the user will now enter
    
 make
 

+<<<<<<< HEAD
 
 
 >>>>>> development
 class="cmr-12">followed (optionally) by
                                                                                
 
@@ -4296,11 +4308,15 @@ class="cmr-12">followed (optionally) by
    
 make install
 

+<<<<<<< HEAD
 

+=======
+

+>>>>>>> development
                                                                                
 
                                                                                
-   
[next] [up] 

-
   
     
  
diff --git a/docs/html/userhtmlsu4.html b/docs/html/userhtmlsu4.html
index 92856733..3d45cd22 100644
--- a/docs/html/userhtmlsu4.html
+++ b/docs/html/userhtmlsu4.html
@@ -10,7 +10,7 @@
  
 
-   
[next] 

 class="cmr-12">3.4    Bug reporting
-
If you find any bugs in our codes, please report them through our issues page
 on
 https://github.com/psctoolkit/amg4psblas/issues

-
   
   To enable us to track the bug, please provide a log from the failing application, the
 test conditions, and ideally a self-contained test program reproducing the
@@ -48,7 +48,7 @@ class="cmr-12">issue.
                                                                                
 
                                                                                
-   
[next] [up] 

-
   
     
  
diff --git a/docs/html/userhtmlsu5.html b/docs/html/userhtmlsu5.html
index d3e22707..bbbed858 100644
--- a/docs/html/userhtmlsu5.html
+++ b/docs/html/userhtmlsu5.html
@@ -10,7 +10,7 @@
  
 
-   
[prev] 

 class="cmr-12">3.5    Example and test programs
-
The package contains the examples and runs subdirectories, to experiment with the full range
      of preconditioners implemented in the package.
-
The fileread directories contain sample programs that read sparse matrices from files,
diff --git a/docs/html/userhtmlsu6.html b/docs/html/userhtmlsu6.html
index 921424cf..c0eb1e37 100644
--- a/docs/html/userhtmlsu6.html
+++ b/docs/html/userhtmlsu6.html
@@ -12,6 +12,9 @@
 >
    
[next] [tail] [].
 
   The setup and application of the default multilevel preconditioner for the real single
-precision and the complex, single and double precision, versions are obtained
                                                                                
 
                                                                                
+precision and the complex, single and double precision, versions are obtained
 with straightforward modifications of the previous example (see Section 5 for
 setup of a one-level Schwarz preconditioner.
[next] [front] [up] 

-
   
     
  
diff --git a/docs/html/userhtmlsu7.html b/docs/html/userhtmlsu7.html
index d7800f58..cc5567aa 100644
--- a/docs/html/userhtmlsu7.html
+++ b/docs/html/userhtmlsu7.html
@@ -1,7 +1,7 @@
   
  
-Method init 
+GPU example 
  
  
  
@@ -10,23 +10,44 @@
  
 
-   
[nextprev] [prev-tail] [tail] [up] 

    
5.1    Method init

+class="cmr-12">4.2    GPU example
+
The code reported in Figure 5 shows how to set up a program exploiting the combined
+GPU capabilities of PSBLAS and AMG4PSBLAS.
+                                                                               
+
+                                                                               
+
   

+                                                                               
+
+                                                                               
 

+<<<<<<< HEAD
 

 
callfor details.     
        

+=======
+

+
+! build a one-level RAS with overlap 2 and ILU(0) on the local blocks.
+call P%init(’AS’,info)
+call P%set(’SUB_OVR’,2,info)
+call P%build(A,desc_A,info)
+... ...
+! solve Ax=b with preconditioned BiCGSTAB
+  call psb_krylov(’BICGSTAB’,A,P,b,x,tol,desc_A,info)
+

+

+
 
Listing 5: setup of a one-level Schwarz preconditioner.

+>>>>>>> development
                                                                                
 
                                                                                
+   

                                                                                
 
                                                                                
-   
[nextprev] [prev-tail] [front] [up] 

-
   
     
  
diff --git a/docs/html/userhtmlsu8.html b/docs/html/userhtmlsu8.html
index 3afae7dd..891ee601 100644
--- a/docs/html/userhtmlsu8.html
+++ b/docs/html/userhtmlsu8.html
@@ -1,7 +1,7 @@
   
  
-Method set 
+Method init 
  
  
  
@@ -10,17 +10,11 @@
  
 
-   
[next] [prev] [prev-tail] [tail] [up] 

    
5.2    Method set

+class="cmr-12">5.1    Method init
 

-

-

+
call p%setinit(whatcontxt,valptype,info [,ilev, ilmax, pos, idx])

-
This method sets the parameters defining the preconditioner )

+
This method allocates and initializes the preconditioner p. More precisely, the
+class="cmr-12">, according to the
 parameter identified by what is assigned the value contained in val.
-
preconditioner type chosen by the user.
+
Arguments
-                                                                               
-
-                                                                               
-   
 
 
, optional,
+is ignored.
 class="td11">
idx   

+
+
+
+class="cmr-12">.
contxt 
 
type(psb_ctxt_type),>>>>>> development
 class="cmtt-12"> intent(in).                                         

 
one, i.e., level 1 is the finest level.
 class="td11">
pos   
.                                     
      
The communication context.                                                  
ptype  
 
>>>>>> development
 class="cmtt-12">character(len=*), optional, intent(in).                          

 
                               
) .
      
The type of preconditioner. Its values are specified in Table 1.     
      
Note that strings are case insensitive.                                      
info  
>>>>>> development
 class="cmtt-12">integer, optional, intent(inout).                                         
                                                       
 
      
                    

 
 
Table 5: Parameters defining the solver at the coarsest level (continued).

+=======
+ style="vertical-align:baseline;" id="TBL-2-7-">
      
Error code. If no error, 0 is returned. See Section 7 for details.     
+       

+>>>>>>> development
                                                                                
 
                                                                                
-   

-   

-   

                                                                                
 
                                                                                
+<<<<<<< HEAD
 
   

@@ -2565,22 +2597,19 @@ class="content">Parameters defining the smoother or the details of the one-level
 
                                                                                
    
>>>>>> development
 class="cmr-12">[next] [prev] [prev-tail] [front] [up] 

-
   
     
  
diff --git a/docs/html/userhtmlsu9.html b/docs/html/userhtmlsu9.html
index 5353f18b..1113134e 100644
--- a/docs/html/userhtmlsu9.html
+++ b/docs/html/userhtmlsu9.html
@@ -1,7 +1,7 @@
   
  
-Method hierarchy_build 
+Method set 
  
  
  
@@ -10,7 +10,7 @@
  
 
-   
[nextup] 

    
5.3    Method hierarchy_build

+class="cmr-12">5.2    Method set
 

-

-

+
call p%hierarchy_buildset(awhat,desc_aval,info)

-

-
This method builds the hierarchy of matrices and restriction/prolongation operators for
-the multilevel preconditioner  [,ilev, ilmax, pos, idx])

+
This method sets the parameters defining the preconditioner p, according to the requirements made by the user
+class="cmr-12">. More precisely, the
 through the methods init and setparameter identified by what is assigned the value contained in val.
-
Arguments
-   
 
 
a     

+
+
+
+
+
+
+class="cmr-12">.                                         
 
+
+
+
+class="cmr-12">.                          
 
+
+class="cmr-12">.                                         
  
+
type
what  
character(len=*).                                                             
      
The parameter to be set. It can be specified through its name; the
+string is case-insensitive. See Tables 2-8.                                  
val   
integer   or    character(len=*)   or    real(psb_xspmat_type),psb_spk_)   or
+real>>>>>> development
+class="cmtt-12">(psb_dpk_), intent(in).                                               
      
The value of the parameter to be set. The list of allowed values and
+the corresponding data types is given in Tables 2-8. When the value
+is of type character(len=*), it is also treated as case insensitive.
info  
integer, intent(out).                                                       
      
Error code. If no error, 0 is returned. See Section 7 for details.     
ilev  
integer, optional, intent(in).                                  
      
The sparse matrix structure containing the local part of the matrix
+=======
+ style="vertical-align:baseline;" id="TBL-3-8-">
      
For   the   multilevel   preconditioner,   the   level   at   which   the
 to be preconditioned. Note that x must be chosen according to the
+class="cmr-12">preconditioner parameter has to be set. The levels are numbered
+>>>>>>> development
+in increasing order starting from the finest one, i.e., level 1 is the
+finest level. If ilev is not present, the parameter identified by what
+is set at all levels that are appropriate (see Tables 2-8).              
ilmax 
integer, optional, intent(in).                                         
      
For the multilevel preconditioner, when both ilev and ilmax are
+present, the settings are applied at all levels ilev:ilmax. When
 real/complex, single/double precision version of AMG4PSBLAS
+class="cmtt-12">ilev is present but ilmax is not, then the default is ilmax=ilev.
 under use. See the PSBLAS User’s Guide for details [
desc_a
typeThe levels are numbered in increasing order starting from the finest
+one, i.e., level 1 is the finest level.                                           
pos   
character>>>>>> development
 class="cmtt-12">(psb_desc_type),len=*), optional, intent(in).                                     
      
.
 class="td11">
info  
      
Whether  the  other  arguments  apply  only  to  the  pre-smoother
+(’PRE’) or to the post-smoother (’POST’). If pos is not present,
+the  other  arguments  are  applied  to  both  smoothers.  If  the
+preconditioner is one-level or the parameter identified by what does
+not concern the smoothers, pos is ignored.                               
idx   
>>>>>> development
 class="cmtt-12">integer, optional, intent(outin).                                                       
      
Error code. If no error, 0 is returned. See Section
      
An auxiliary input argument that can be passed to the underlying
+objects.                                                                             

+

+
   A variety of preconditioners can be obtained by setting the appropriate
+preconditioner parameters. These parameters can be logically divided into four groups,
+i.e., parameters defining
+     
    
+     
  1. the type of multilevel cycle and how many cycles must be applied;
+     
    2. 
+     
  2. the coarsening algorithm;
+     
    3. 
+     
  3. the solver at the coarsest level (for multilevel preconditioners only);
+     
    4. 
+     
  4. the   smoother   of   the   multilevel   preconditioners,   or   the   one-level
+                                                                               
+
+                                                                               
+     preconditioner.

+
A list of the parameters that can be set, along with their allowed and default values, is
+given in Tables>>>>>> development
 class="cmr-12"> 7 for details.     
+href="#x19-18009r2">2-8.

+
   Remark 2. A smoother is usually obtained by combining two objects: a
+smoother (’SMOOTHER_TYPE’) and a local solver (’SUB_SOLVE’), as specified in
+Tables 7-8. For example, the block-Jacobi smoother using ILU(0) on the blocks is
+obtained by combining the block-Jacobi smoother object with the ILU(0) solver
+object. Similarly, the hybrid Gauss-Seidel smoother (see Note in Table 7) is
+obtained by combining the block-Jacobi smoother object with a single sweep of
+the Gauss-Seidel solver object, while the point-Jacobi smoother is the result
+of combining the block-Jacobi smoother object with a single sweep of the
+point-Jacobi solver object. In the same way are obtained the ℓ1-versions of the
+smoothers. However, for simplicity, shortcuts are provided to set all versions of
+point-Jacobi, hybrid (forward) Gauss-Seidel, and hybrid backward Gauss-Seidel, i.e.,
+the previous smoothers can be defined just by setting ’SMOOTHER_TYPE’ to
+certain specific values (see Tables 7), without the need to set ’SUB_SOLVE’ as
+well.
+
   The smoother and solver objects are arranged in a hierarchical manner. When
+specifying a smoother object, its parameters, including the local solver, are set to
+their default values, and when a solver object is specified, its defaults are also
+set, overriding in both cases any previous settings even if explicitly specified.
+Therefore if the user sets a smoother, and wishes to use a solver different from
+the default one, the call to set the solver must come after the call to set the
+smoother.
+
   Similar considerations apply to the point-Jacobi, Gauss-Seidel and block-Jacobi
+coarsest-level solvers, and shortcuts are available in this case too (see Table 5).
+

+
   Remark 3. Many of the coarsest-level solvers cannot be used with both replicated
+and distributed coarsest-matrix layouts; therefore, setting the solver after the layout
+may change the layout. Similarly, setting the layout after the solver may change the
+solver.
+
   More precisely, UMFPACK and SuperLU require the coarsest-level matrix to be
+replicated, while SuperLU_Dist and KRM require it to be distributed. In these cases,
+setting the coarsest-level solver implies that the layout is redefined according to the
+solver, ovverriding any previous settings. MUMPS, point-Jacobi, hybrid Gauss-Seidel
+and block-Jacobi can be applied to replicated and distributed matrices, thus their
+choice does not modify any previously specified layout. It is worth noting that, when
+the matrix is replicated, the point-Jacobi, hybrid Gauss-Seidel and block-Jacobi solvers
+and their ℓ1- versions reduce to the corresponding local solver objects (see Remark 2).
+For the point-Jacobi and Gauss-Seidel solvers, these objects correspond to a single
+                                                                               
+
+                                                                               
+point-Jacobi sweep and a single Gauss-Seidel sweep, respectively, which are very poor
+solvers.
+
   On the other hand, the distributed layout can be used with any solver but
+UMFPACK and SuperLU; therefore, if any of these two solvers has already been
+selected, the coarsest-level solver is changed to block-Jacobi, with the previously chosen
+solver applied to the local blocks. Likewise, the replicated layout can be used with any
+solver but SuperLu_Dist and KRM; therefore, if SuperLu_Dist or KRM have been
+previously set, the coarsest-level solver is changed to the default sequential
+solver.
+
   Remark 4. The argument idx can be used to allow finer control for those solvers;
+for instance, by specifying the keyword ’MUMPS_IPAR_ENTRY’ and an appropriate value
+for idx, it is possible to set any entry in the MUMPS integer control array. See also
+Sec. 6.
+   

+                                                                               
+
+                                                                               
+
   

+                                                                               
+
+                                                                               
+

+

+
 
+
+
what              
data type         
val         
default      
comments                                    
’ML_CYCLE’ 
character(len=*)   
’VCYCLE’
+
’WCYCLE’
+
’KCYCLE’
+
’ADD’       
’VCYCLE’    
Multilevel cycle: V-cycle, W-cycle, K-cycle,
+and additive composition.                      
’CYCLE_SWEEPS’    
integer                                                                  
Any integer
+
number ≥ 1   
1                 
Number of multilevel cycles.                   
                  

+
 
Table 2: Parameters defining the multilevel cycle and the number of cycles to be
+applied. 

+                                                                               
+
+                                                                               
+   

+   

+   

+                                                                               
+
+                                                                               
+
   

+                                                                               
+
+                                                                               
+

+

+                                                                               
+
+                                                                               
+
 
+
+
+
+
+
+
+
+
Note. The aggregation algorithm stops when at least one of the following criteria is met: the coarse size threshold, 
what                        
data type        
val        
default       
comments                                  
’MIN_COARSE_SIZE_PER_PROCESS’ 
integer   
Any number
+
> 0             
200               
Coarse  size  threshold  per  process.  The
+aggregation stops if the global number of
+variables of the computed coarsest matrix
+is lower than or equal to this threshold
+multiplied  by  the  number  of  processes
+(see Note).                                        
’MIN_COARSE_SIZE’            
integer                                                                  
Any number
+
> 0             
-1                 
Coarse  size  threshold.  The  aggregation
+stops  if  the  global  number  of  variables
+of the computed coarsest matrix is lower
+than                   or                   equal
+to this threshold (see Note). If negative,
+it is ignored in favour of the default for
+’MIN_COARSE_SIZE_PER_PROCESS’.       
’MIN_CR_RATIO’               
real                                                                     
Any number
+
> 1             
1.5                
Minimum                           coarsening
+ratio. The aggregation stops if the ratio
+between the global matrix dimensions at
+two  consecutive  levels  is  lower  than  or
+equal to this threshold (see Note).         
’MAX_LEVS’                   
integer                                                                  
Any integer
+
number > 1  
20                
Maximum    number    of    levels.    The
+aggregation stops if the number of levels
+reaches this value (see Note).               
’PAR_AGGR_ALG’               
character(len=*)                                                         
’DEC’,
+’SYMDEC’,
+’COUPLED’  
’DEC’       
Parallel aggregation algorithm.
+
the  SYMDEC  option  applies  decoupled
+aggregation  to  the  sparsity  pattern  of
+A + AT .                                           
’AGGR_TYPE’                  
character(len=*)                                                         
’SOC1’,
+’SOC2’,
+’MATCHBOXP’
’SOC1’      
Type of aggregation algorithm: currently,
+for                                              the
+decoupled aggregation we implement two
+measures of strength of connection, the
+one by Vaněk, Mandel and Brezina [33],
+and the one by Gratton et al [23]. The
+coupled aggregation is based on a parallel
+version of the half-approximate matching
+implemented in the MatchBox-P software
+package [9].                                      
’AGGR_SIZE’                  
integer                                                                  
Any integer
+
power   of   2,
+with
+aggr_size ≥
+2                
4                  
Maximum  size  of  aggregates  when  the
+coupled  aggregation  based  on  matching
+is   applied.   For   aggressive   coarsening
+with  size  of  aggregate  larger  than  8
+we  recommend  the  use  of  smoothed
+prolongators. Used only with ’COUPLED’
+and ’MATCHBOXP’                   
’AGGR_PROL’                  
character(len=*)                                                         
’SMOOTHED’,
+’UNSMOOTHED’ 
’SMOOTHED’  
Prolongator  used  by  the  aggregation
+algorithm: smoothed or unsmoothed (i.e.,
+tentative prolongator).                        

 
       

+=======
+ style="vertical-align:baseline;" id="TBL-5-11-">
the minimum coarsening ratio, or the maximum number of levels is reached.

+
Therefore, the actual number of levels may be smaller than the specified maximum number of levels. 

+
                             
+                                                                               
+
+                                                                               
+

+
 
Table 3: Parameters defining the aggregation algorithm. 

+

+>>>>>>> development
+                                                                               
+
+                                                                               
+   

+   

+   

+                                                                               
+
+                                                                               
+
   

+                                                                               
+
+                                                                               
+

+

+
 
+
+
+
Note. Different thresholds at different levels, such as those used in [33, Section 5.1], can be easily set by invoking the rou-

+
tine set with the parameter ilev.

+
what               
data type               
val         
default      
comments                                
’AGGR_ORD’ 
character(len=*)   
’NATURAL’
+
’DEGREE’    
’NATURAL’  
Initial    ordering    of    indices    for
+the  decoupled  aggregation  algorithm:
+either  natural  ordering  or  sorted  by
+descending degrees of the nodes in the
+matrix graph.                                 
’AGGR_THRESH’      
real(kind_parameter)                                                     
Any real
+
number ∈
+[0,1]             
0.01             
The  threshold  θ  in  the  strength  of
+connection algorithm. See also the note
+at the bottom of this table.               
’AGGR_FILTER’      
character(len=*)                                                         
’FILTER’
+
’NOFILTER’  
’NOFILTER’ 
Matrix used in computing the smoothed
+prolongator: filtered or unfiltered.       
                   

+
 
Table 4: Parameters defining the aggregation algorithm (continued). 

+                                                                               
+
+                                                                               
+   

+   

+   

+                                                                               
+
+                                                                               
+
   

+                                                                               
+
+                                                                               
+

+

+                                                                               
+
+                                                                               
+
 
+
+
+
Note. Defaults for COARSE_SOLVE and COARSE_SUBSOLVE are chosen in the following order:

+
single precision version – MUMPS if installed, then SLU if installed, ILU otherwise;

+
double precision version – UMF if installed, then MUMPS if installed, then SLU if installed, ILU otherwise.

+
+
+
+
+
Note. Further options for coarse solvers are contained in Table 6.

+
For a first use it is suggested to use the default options obtained by simply selecting the solver type.

+
what               
data type               
val      
default 
comments                                              
’COARSE_MAT’ 
character(len=*)   
’DIST’
+
’REPL’   
’REPL’  
Coarsest  matrix  layout:  distributed  among  the
+processes or replicated on each of them.               
’COARSE_SOLVE’      
character(len=*)                                                         
’MUMPS’
+
’UMF’
+
’SLU’
+
’SLUDIST’
+
’JACOBI’
+
’GS’
+
’BJAC’
+
’KRM’
+’L1-JACOBI’
+
’L1-BJAC’
+
’L1-FBGS’
See Note.
Solver   used   at   the   coarsest   level:   sequential
+LU   from   MUMPS,   UMFPACK,   or   SuperLU
+(plus   triangular   solve);   distributed   LU   from
+MUMPS or SuperLU_Dist (plus triangular solve);
+point-Jacobi, hybrid Gauss-Seidel or block-Jacobi
+and  related  ℓ1-versions;  Krylov  Method  (flexible
+Conjugate Gradient) coupled with the block-Jacobi
+preconditioner  with  ILU(0)  on  the  blocks.  Note
+that  UMF and  SLU require  the  coarsest  matrix  to
+be replicated, SLUDIST, JACOBI, GS, BJAC and KRM
+require it to be distributed, MUMPS can be used with
+either a replicated or a distributed matrix. When
+any of the previous solvers is specified, the matrix
+layout is set to a default value which allows the use
+of the solver (see Remark 3, p. 24). Note also that
+UMFPACK and SuperLU_Dist are available only in
+double precision.                                             
’COARSE_SUBSOLVE’   
character(len=*)                                                         
’ILU’
+
’ILUT’
+
’MILU’
+
’MUMPS’
+
’SLU’
+
’UMF’
+
’INVT’
+
’INVK’
+
’AINV’   
See Note.
Solver  for  the  diagonal  blocks  of  the  coarsest
+matrix, in case the block Jacobi solver is chosen
+as coarsest-level solver: ILU(p), ILU(p,t), MILU(p),
+LU   from   MUMPS,   SuperLU   or   UMFPACK
+(plus   triangular   solve),   Approximate   Inverses
+INVK(p,q), INVT(p1,p2,t1,t2) and AINV(t); note
+that  approximate  inverses  are  specifically  suited
+for  GPUs  since  they  do  not  employ  triangular
+system solve kernels, see [3]. Note that UMFPACK
+and  SuperLU_Dist  are  available  only  in  double
+precision.                                                       
what                
data type               
val      
default 
comments                                              
’COARSE_SWEEPS’     
integer                                                                  
Any
+integer
+
number   >
+0             
10          
Number  of  sweeps  when  JACOBI,  GS or  BJAC is
+chosen as coarsest-level solver.                           
’COARSE_FILLIN’     
integer                                                                  
Any
+integer
+
number   ≥
+0             
0            
Fill-in  level  p of  the  ILU  factorizations  and  first
+fill-in for the approximate inverses.                     
’COARSE_ILUTHRS’    
real(kind_parameter)                                                     
Any real
+
number   ≥
+0             
0            
Drop tolerance t in the ILU(p,t) factorization and
+first drop-tolerance for the approximate inverses.   
                    

+                                                                               
+
+                                                                               
+

+
 
Table 5: Parameters defining the solver at the coarsest level (continued).

+                                                                               
+
+                                                                               
+   

+   

+   

+                                                                               
+
+                                                                               
+
   

+                                                                               
+
+                                                                               
+

+

+                                                                               
+
+                                                                               
+
 
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
what               
data type               
val       
default 
comments                                        
’BJAC_STOP’ 
character(len=*)   
’FALSE’
+
’TRUE’    
’FALSE’ 
Select whether to use a stopping criterion for
+the  Block-Jacobi  method  used  as  a  coarse
+solver.                                                    
’BJAC_TRACE’        
character(len=*)                                                         
’FALSE’
+
’TRUE’    
’FALSE’ 
Select  whether  to  print  a  trace  for  the
+calculated   residual   for   the   Block-Jacobi
+method used as a coarse solver.                   
’BJAC_ITRACE’       
integer                                                                  
Any integer
+
> 0           
-1           
Number of iterations after which a trace is to
+be printed.                                              
’BJAC_RESCHECK’     
integer                                                                  
Any integer
+
> 0           
-1           
Number of iterations after which a residual is
+to be calculated.                                      
’BJAC_STOPTOL’      
real(kind_parameter)                                                     
Any real
+
< 1           
0            
Tolerance  for  the  stopping  criterion  on  the
+residual.                                                 
’KRM_METHOD’        
character(len=*)                                                         
’CG’
+
’FCG’
+
’CGS’
+
’CGR’
+
’BICG’
+
’BICGSTAB’
+
’BICGSTABL’
+
’RGMRES’  
’FCG’   
A string that defines the iterative method to
+be  used  when  employing  a  Krylov  method
+’KRM’ as  a  coarse  solver.  CG the  Conjugate
+Gradient method; FCG the Flexible Conjugate
+Gradient method; CGS the Conjugate Gradient
+Stabilized
+method;   GCR  the   Generalized   Conjugate
+Residual method; FCG the Flexible Conjugate
+Gradient  method;  BICG  the  Bi-Conjugate
+Gradient method; BICGSTAB the Bi-Conjugate
+Gradient  Stabilized  method;  BICGSTABL the
+Bi-Conjugate   Gradient   Stabilized   method
+with   restarting;   RGMRES  the   Generalized
+Minimal  Residual  method  with  restarting.
+Refer to the PSBLAS guide [20] for further
+information.                                            
’KRM_KPREC’         
character(len=*)                                                         
Table 1 
’BJAC’ 
The     one-level
+preconditioners from the Table 1 can be used
+for the coarse Krylov solver.                       
’KRM_SUB_SOLVE’     
character(len=*)                                                         
Table 5 
’ILU’ 
Solver for the diagonal blocks of the coarsest
+matrix preconditioner, in case the block Jacobi
+solver                     is                     chosen
+as ’KRM_KPREC’: ILU(p), ILU(p,t), MILU(p),
+LU  from  MUMPS,  SuperLU  or  UMFPACK
+(plus triangular solve), Approximate Inverses
+INVK(p,q), INVT(p1,p2,t1,t2) and AINV(t);
+The same caveat from Table 5 applies here.  
’KRM_GLOBAL’        
character(len=*)                                                         
’TRUE’,
+’FALSE’   
’FALSE’ 
Choose  between  a  global  Krylov  solver,  all
+unknowns on a single node, or a distributed
+one.  The  default  choice  is  the  distributed
+solver.                                                    
’KRM_EPS’           
real(kind_parameter)                                                     
Real < 1 
10-6  
The stopping tolerance.                
’KRM_IRST’          
integer                                                                  
Integer
+
≥ 1           
30          
An integer specifying the restart parameter.
+This is employed for the BiCGSTABL or RGMRES
+methods, otherwise it is ignored.                 
’KRM_ISTOPC’        
integer                                                                  
Integers
+1,2,3         
2            
If  1 then  the  method  uses  the  normwise
+backward error in the infinity norm; if 2, the
+it uses the relative residual in the 2-norm; if 3
+the relative residual reduction in the 2-norm is
+used instead; refer to the PSBLAS [20] guide
+for the details.                                         
’KRM_ITMAX’         
integer                                                                  
Integer
+
≥ 1           
40          
The   maximum   number   of   iterations   to
+perform.                                                 
’KRM_ITRACE’        
integer                                                                  
Integer
+
≥ 0           
-1           
If         >             0         print         out
+an informational message about convergence
+every ’KRM_ITRACE’ iterations. If = 0 print a
+message in case of convergence failure.          
’KRM_FILLIN’        
integer                                                                  
Integer
+
≥ 0           
0            
Fill-in level p of the ILU factorizations and
+first fill-in for the approximate inverses.        
                    
                                                       
+
+                                                                               
+

+
 
Table 6: Additional parameters defining the solver at the coarsest level.

+                                                                               
+
+                                                                               
+   

+   

+   

+                                                                               
+
+                                                                               
+
   

+                                                                               
+
+                                                                               
+

+

+
 
+
+
+
+
what                
data type         
val       
default                 
comments                                    
’SMOOTHER_TYPE’ 
character(len=*)   
’JACOBI’
+
’GS’
+
’BGS’
+
’BJAC’
+
’AS’
+
’L1-JACOBI’
+
’L1-BJAC’
+
’L1-FBGS’ 
’FBGS’              
Type of smoother used in the multilevel
+preconditioner:     point-Jacobi,     hybrid
+(forward)  Gauss-Seidel,  hybrid  backward
+Gauss-Seidel,    block-Jacobi,    ℓ1-Jacobi,
+ℓ1–hybrid       (forward)       Gauss-Seidel,
+ℓ1-point-Jacobi and Additive Schwarz.
+
It is ignored by one-level preconditioners. 
’SUB_SOLVE’         
character(len=*)                                                                
’JACOBI’
+’GS’
+
’BGS’
+
’ILU’
+
’ILUT’
+
’MILU’
+
’MUMPS’
+
’SLU’
+
’UMF’
+
’INVT’
+
’INVK’
+
’AINV’    
GS and  BGS for  pre-
+and  post-smoothers  of
+multilevel
+preconditioners,
+respectively
+
ILU  for   block-Jacobi
+and  Additive  Schwarz
+one-level
+preconditioners           
The  local  solver  to  be  used  with  the
+smoother or one-level preconditioner (see
+Remark 2, page 24): point-Jacobi, hybrid
+(forward)  Gauss-Seidel,  hybrid  backward
+Gauss-Seidel, ILU(p), ILU(p,t), MILU(p),
+LU                from                MUMPS,
+SuperLU  or  UMFPACK  (plus  triangular
+solve), Approximate Inverses INVK(p,q),
+INVT(p1,p2,t1,t2)   and   AINV(t);   note
+that approximate inverses are specifically
+suited for GPUs since they do not employ
+triangular system solve kernels, see [3]. See
+Note for details on hybrid Gauss-Seidel.   
’SMOOTHER_SWEEPS’   
integer                                                                         
Any integer
+
number ≥ 0
1                             
Number  of  sweeps  of  the  smoother  or
+one-level preconditioner. In the multilevel
+case,  no  pre-smother  or  post-smoother
+is  used  if  this  parameter  is  set  to  0
+together with pos=’PRE’ or pos=’POST’,
+respectively.                                        
’SUB_OVR’           
integer                                                                         
Any integer
+
number ≥ 0
1                             
Number  of  overlap  layers,  for  Additive
+Schwarz only.                                      
                    

+
  
Table 7:   Parameters   defining   the   smoother   or   the   details   of   the   one-level
+preconditioner. 

                                                                                
 
                                                                                
+   

+   

+   

                                                                                
 
                                                                                
-   
   

+                                                                               
+
+                                                                               
+

+

+
 
+
+
+
+
+
+
+
what             
data type               
val           
default         
comments                                     
’SUB_RESTR’ 
character(len=*)   
’HALO’
+
’NONE’        
’HALO’        
Type  of  restriction  operator,  for  Additive
+Schwarz only: HALO for taking into account
+the overlap, ’NONE’ for neglecting it.
+
Note  that  HALO must  be  chosen  for  the
+classical  Addditive  Schwarz  smoother  and
+its RAS variant.                                    
’SUB_PROL’        
character(len=*)                                                                
’SUM’
+
’NONE’        
’NONE’        
Type of prolongation operator, for Additive
+Schwarz   only:   ’SUM’  for   adding   the
+contributions from the overlap, ’NONE’ for
+neglecting them.
+
Note  that  ’SUM’ must  be  chosen  for  the
+classical  Additive  Schwarz  smoother,  and
+’NONE’ for its RAS variant.                     
’SUB_FILLIN’      
integer                                                                         
Any integer
+
number ≥ 0      
0                     
Fill-in   level   p  of   the   incomplete   LU
+factorizations.                                       
’SUB_ILUTHRS’     
real(kind_parameter)                                                            
Any real
+number ≥ 0      
0                     
Drop    tolerance    t   in    the    ILU(p,t)
+factorization.                                        
’MUMPS_LOC_GLOB’  
character(len=*)                                                                
’LOCAL_SOLVER’
+
’GLOBAL_SOLVER’ 
’GLOBAL_SOLVER’ 
Whether  MUMPS  should  be  used  as  a
+distributed solver, or as a serial solver acting
+only on the part of the matrix local to each
+process.                                               
’MUMPS_IPAR_ENTRY’ 
integer                                                                         
Any integer
+number            
0                     
Set an entry in the MUMPS integer control
+array,  as  chosen  via  the  idx  optional
+argument.                                            
’MUMPS_RPAR_ENTRY’ 
real                                                                            
Any real number 
0 
Set an entry in the MUMPS  real  control
+array,  as  chosen  via  the  idx  optional
+argument.                                            
                  

+
 
Table 8: Parameters defining the smoother or the details of the one-level preconditioner
+(continued).

+                                                                               
+
+                                                                               
+   

+   

+                                                                               
+
+                                                                               
+                                                                               
+
+                                                                               
+   
[next] [up] 

-
   
     
  
diff --git a/docs/src/building.tex b/docs/src/building.tex
index 06b0b836..cc54f95c 100644
--- a/docs/src/building.tex
+++ b/docs/src/building.tex
@@ -48,7 +48,7 @@ work in the expected way. Our advice is first to download the LAPACK tarfile fro
 \href{http://www.netlib.org/lapack}{www.netlib.org/lapack} and install it independently of ATLAS. In this case,
 you need to modify the OPTS and NOOPT definitions for including -fPIC compilation option
 in the make.inc file of the LAPACK library.
-\item[MPI] \cite{MPI2,MPI1} A version of MPI is available on most
+ \item[MPI] \cite{MPI2,MPI1} A version of MPI is available on most
   high-performance computing systems.
  \item[PSBLAS] \cite{PSBLASGUIDE,psblas_00} Parallel Sparse BLAS (PSBLAS) is
   available from
@@ -61,6 +61,18 @@ Please note that the four previous libraries must have Fortran
 interfaces compatible with AMG4PSBLAS; usually this means that they
 should all be built with the same compiler being used for  AMG4PSBLAS.
 
+If you want to use the PSBLAS support for NVIDIA GPUs, you will also
+need:
+\begin{description}
+ \item[PSBLAS-EXT] Parallel Sparse BLAS (PSBLAS) Extensions, 
+  available from
+  \href{https://psctoolkit.github.io/products/psblasext/}{psctoolkit.github.io/products/psblasext/}; version   1.3.0  (or later). 
+ \item[SPGPU] Sparse CUDA kernels for NVIDIA GPUs; available from
+   GitHub, see also
+   \href{https://psctoolkit.github.io/products/psblasext/}{psctoolkit.github.io/products/psblasext/}.
+ \end{description}
+ See also Sec~\ref{sec:gpu-example}.
+ 
 \subsection{Optional third party libraries\label{sec:third-party}}
 
 We provide interfaces to the following third-party software libraries;
diff --git a/docs/src/gettingstarted.tex b/docs/src/gettingstarted.tex
index 42399ee6..e4322ce0 100644
--- a/docs/src/gettingstarted.tex
+++ b/docs/src/gettingstarted.tex
@@ -416,6 +416,45 @@ call P%build(A,desc_A,info)
 \end{listing}
 
 
+
+\subsection{GPU example\label{sec:gpu-example}}
+
+The code reported in Figure~\ref{fig:gpu-ex1} shows how to set  up a
+program exploiting the combined GPU capabilities of PSBLAS and
+AMG4PSBLAS.
+\begin{listing}[h!]
+\ifpdf
+\begin{minted}[breaklines=true,bgcolor=bg,fontsize=\small]{fortran}
+! build a one-level RAS with overlap 2 and ILU(0) on the local blocks.
+call P%init('AS',info)
+call P%set('SUB_OVR',2,info)
+call P%build(A,desc_A,info)
+... ...
+! solve Ax=b with preconditioned BiCGSTAB
+  call psb_krylov('BICGSTAB',A,P,b,x,tol,desc_A,info)
+\end{minted}
+\else
+\begin{center}
+\begin{minipage}{.90\textwidth}
+{\small
+\begin{verbatim}
+! build a one-level RAS with overlap 2 and ILU(0) on the local blocks.
+call P%init('AS',info)
+call P%set('SUB_OVR',2,info)
+call P%build(A,desc_A,info)
+... ...
+! solve Ax=b with preconditioned BiCGSTAB
+  call psb_krylov('BICGSTAB',A,P,b,x,tol,desc_A,info)
+\end{verbatim}
+}
+\end{minipage}
+\end{center}
+\fi
+\caption{setup of a one-level Schwarz preconditioner.\label{fig:gpu-ex1}}
+\end{listing}
+
+
+
 %%% Local Variables:
 %%% mode: latex
 %%% TeX-master: "userguide"