config/pac.m4
configure.ac
configure
mlprec/Makefile
tests/fileread/Makefile
tests/pdegen/Makefile
tests/pdegen/runs/ppde.inp
Now accepts the install path of PSBLAS.
mlprec/mld_c_jac_smoother.f03
mlprec/mld_d_jac_smoother.f03
mlprec/mld_dprecset.F90
mlprec/mld_s_jac_smoother.f03
mlprec/mld_sprecset.F90
mlprec/mld_z_jac_smoother.f03
Check in Jacobi smoother for empty non-diagonal part, and force 1
sweep in that case.
mlprec/mld_dmlprec_aply.f90
mlprec/mld_dmlprec_bld.f90
mlprec/mld_dprecset.f90
mlprec/mld_move_alloc.f90
mlprec/mld_move_alloc_mod.f90
tests/pdegen/ppde.f90
tests/pdegen/runs/ppde.inp
First working version of multilevel. Further testing to be performed
yet.
mlprec/mld_c_prec_type.f90
mlprec/mld_d_as_smoother.f03
mlprec/mld_d_diag_solver.f03
mlprec/mld_d_ilu_solver.f03
mlprec/mld_d_jac_smoother.f03
mlprec/mld_d_prec_type.f03
mlprec/mld_dprecaply.f90
mlprec/mld_dprecbld.f90
mlprec/mld_dprecinit.F90
mlprec/mld_s_prec_type.f90
mlprec/mld_z_prec_type.f90
tests/pdegen/runs/ppde.inp
Used new Smoother/Solver object hierarchy, for 1 level.
The precinit/precset is still the old one.
Makefile
mld_base_prec_type.f90
mld_c_prec_type.f90
mld_d_prec_type.f90
mld_d_prec_type.mod
mld_prec_type.f90
mld_s_prec_type.f90
mld_z_prec_type.f90
Brought in separation of prec_type from trunk.
mlprec/mld_daggrmat_smth_asb.F90
mlprec/mld_dcoarse_bld.f90
mlprec/mld_dilu_bld.f90
mlprec/mld_dsub_solve.f90
tests/pdegen/runs/ppde.inp
Found and fixed bugs in building. Some preconditioners work, testing
not complete yet.
Makefile
config/pac.m4
configure.ac
configure
mlprec/Makefile
mlprec/mld_daggrmap_bld.f90
mlprec/mld_daggrmat_asb.f90
mlprec/mld_daggrmat_minnrg_asb.F90
mlprec/mld_daggrmat_nosmth_asb.F90
mlprec/mld_daggrmat_smth_asb.F90
mlprec/mld_das_aply.f90
mlprec/mld_das_bld.f90
mlprec/mld_dbaseprec_bld.f90
mlprec/mld_dcoarse_bld.f90
mlprec/mld_ddiag_bld.f90
mlprec/mld_dfact_bld.f90
mlprec/mld_dilu0_fact.f90
mlprec/mld_dilu_bld.f90
mlprec/mld_diluk_fact.f90
mlprec/mld_dilut_fact.f90
mlprec/mld_dmlprec_bld.f90
mlprec/mld_dprecaply.f90
mlprec/mld_dprecbld.f90
mlprec/mld_dslu_bld.f90
mlprec/mld_dslud_bld.f90
mlprec/mld_dsp_renum.f90
mlprec/mld_dsub_solve.f90
mlprec/mld_dumf_bld.f90
mlprec/mld_inner_mod.f90
mlprec/mld_prec_mod.f90
mlprec/mld_prec_type.f90
tests/pdegen/Makefile
tests/pdegen/ppde.f90
MLD2P4: first compilable version for D, still a lot to do to make it
RUN properly.
mlprec/mld_daggrmat_asb.f90
mlprec/mld_daggrmat_minnrg_asb.F90
mlprec/mld_dmlprec_aply.f90
mlprec/mld_dprecinit.F90
mlprec/mld_dprecset.f90
mlprec/mld_inner_mod.f90
mlprec/mld_prec_type.f90
mld2p4:
First attempt at implementing MIN ENERGY smoother. Only for double
real so far.
mlprec/mld_caggrmat_smth_asb.F90
mlprec/mld_daggrmat_smth_asb.F90
mlprec/mld_saggrmat_smth_asb.F90
mlprec/mld_zaggrmat_smth_asb.F90
mld2p4:
Switched order of arguments in sp_scal; now it is more logical
(D,A) <-> A = DA
examples/fileread/mld_cexample_1lev.f90
examples/fileread/mld_cexample_ml.f90
examples/fileread/mld_dexample_1lev.f90
examples/fileread/mld_dexample_ml.f90
examples/fileread/mld_sexample_1lev.f90
examples/fileread/mld_sexample_ml.f90
examples/fileread/mld_zexample_1lev.f90
examples/fileread/mld_zexample_ml.f90
examples/pdegen/mld_dexample_1lev.f90
examples/pdegen/mld_dexample_ml.f90
examples/pdegen/mld_sexample_1lev.f90
examples/pdegen/mld_sexample_ml.f90
mlprec/mld_cdiag_bld.f90
mlprec/mld_cmlprec_bld.f90
mlprec/mld_cprecaply.f90
mlprec/mld_cprecbld.f90
mlprec/mld_cumf_bld.f90
mlprec/mld_ddiag_bld.f90
mlprec/mld_dmlprec_bld.f90
mlprec/mld_dprecaply.f90
mlprec/mld_dprecbld.f90
mlprec/mld_inner_mod.f90
mlprec/mld_sdiag_bld.f90
mlprec/mld_smlprec_bld.f90
mlprec/mld_sprecaply.f90
mlprec/mld_sprecbld.f90
mlprec/mld_zdiag_bld.f90
mlprec/mld_zmlprec_bld.f90
mlprec/mld_zprecaply.f90
mlprec/mld_zprecbld.f90
mlprec/mld_zumf_bld.f90
tests/fileread/cf_sample.f90
tests/fileread/df_sample.f90
tests/fileread/sf_sample.f90
tests/fileread/zf_sample.f90
tests/pdegen/ppde.f90
tests/pdegen/spde.f90
Changed interface in matdist. It is unspecified in the Fortran
standard whether a subroutine can be used to disambiguate a specific
interface with respect to an integer vector.
Fixes for various small inconsistencies in interfaces revealed by NAG
compiler.
mld_caggrmat_smth_asb.F90
mld_daggrmat_smth_asb.F90
mld_saggrmat_smth_asb.F90
mld_zaggrmat_smth_asb.F90
Fixed filtering into smoothed aggregation to prpoerly use the
strong-coupling definition.
Salvatore Filippone