<div dir="ltr"><div class="gmail_extra"><div class="gmail_quote">On Tue, Feb 9, 2016 at 9:31 AM, Hom Nath Gharti <span dir="ltr"><<a href="mailto:hng.email@gmail.com" target="_blank">hng.email@gmail.com</a>></span> wrote:<br><blockquote class="gmail_quote" style="margin:0px 0px 0px 0.8ex;border-left-width:1px;border-left-color:rgb(204,204,204);border-left-style:solid;padding-left:1ex">Thanks a lot Matt!<br>
<br>
Were you referring to<br>
<a href="http://www.mcs.anl.gov/petsc/petsc-current/src/snes/examples/tutorials/ex62.c.html" rel="noreferrer" target="_blank">http://www.mcs.anl.gov/petsc/petsc-current/src/snes/examples/tutorials/ex62.c.html</a><br>
?<br>
<br>
I do not see any statements related to PCFieldSplit there. Am I<br>
missing something here?<br></blockquote><div><br></div><div>Yes. The whole point is not to put anything specific to FIELDSPLIT in the code. It is all in options. For example,</div><div><br></div><div> ./ex62 -run_type full -refinement_limit 0.00625 -bc_type dirichlet -interpolate 1 -vel_petscspace_order 2 -pres_petscspace_order 1 -ksp_type fgmres -ksp_gmres_restart 100 -ksp_rtol 1.0e-9 -pc_type fieldsplit -pc_fieldsplit_type schur -pc_fieldsplit_schur_factorization_type upper -fieldsplit_pressure_ksp_rtol 1e-10 -fieldsplit_velocity_ksp_type gmres -fieldsplit_velocity_pc_type lu -fieldsplit_pressure_pc_type jacobi -snes_monitor_short -ksp_monitor_short -snes_converged_reason -ksp_converged_reason -snes_view -show_solution 0</div><div><br></div><div>The reason that it works is that the DM calls PCFIeldSplitSetIS() for each field in the DM.</div><div><br></div><div> Thanks,</div><div><br></div><div> Matt</div><div> </div><blockquote class="gmail_quote" style="margin:0px 0px 0px 0.8ex;border-left-width:1px;border-left-color:rgb(204,204,204);border-left-style:solid;padding-left:1ex">
Thanks,<br>
Hom Nath<br>
<div class=""><div class="h5"><br>
On Tue, Feb 9, 2016 at 10:19 AM, Matthew Knepley <<a href="mailto:knepley@gmail.com">knepley@gmail.com</a>> wrote:<br>
> On Tue, Feb 9, 2016 at 9:10 AM, Hom Nath Gharti <<a href="mailto:hng.email@gmail.com">hng.email@gmail.com</a>> wrote:<br>
>><br>
>> Thank you so much Barry!<br>
>><br>
>> For my small test case, with -pc_fieldsplit_block_size 4, the program<br>
>> runs, although the answer was not correct. At least now I get<br>
>> something to look upon. I am using PCFieldSplitSetIS to set the<br>
>> fields. Do I still need to use -pc_fieldsplit_block_size?<br>
><br>
><br>
> No, this is only for co-located discretizations.<br>
><br>
>><br>
>> In my case each grid point may have different variable sets. For<br>
>> example, the grid point in the solid region has displacement and<br>
>> gravity potential: ux, uy, uz, and \phi. Whereas the grid point in the<br>
>> fluid region has scalar potential, pressure and gravity potential:<br>
>> \chi, p, and \phi. And the solid-fluid interface has all of them. Do<br>
>> you think I can still use PCFIELDSPLIT in this situation?<br>
><br>
><br>
> We have examples, like SNES ex62, which do exactly this.<br>
><br>
> Thanks,<br>
><br>
> Matt<br>
><br>
>><br>
>> Best,<br>
>> Hom Nath<br>
>><br>
>><br>
>><br>
>> On Mon, Feb 8, 2016 at 2:27 PM, Barry Smith <<a href="mailto:bsmith@mcs.anl.gov">bsmith@mcs.anl.gov</a>> wrote:<br>
>> ><br>
>> ><br>
>> > In this case you need to provide two pieces of information to the<br>
>> > PCFIELDSPLIT. What we call the "block size" or bs which is the number of<br>
>> > "basic fields" in the problem. For example if at each grid point you have<br>
>> > x-velocity, y-velocity, and pressure the block size is 3. The second is you<br>
>> > need to tell PCFIELDSPLIT what "basic fields" are in each split you want to<br>
>> > use.<br>
>> ><br>
>> > In your case you can do this with -pc_fieldsplit_block_size 4.<br>
>> > (Note if you use a DM to define your problem the PCFIELDSPLIT will<br>
>> > automatically get the bs from the DM so you do not need to provide it.<br>
>> > Similarly if you set a block size on your Mat the PCFIELDSPLIT will use that<br>
>> > so you don't need to set it).<br>
>> ><br>
>> > Then use -pc_fieldsplit_0_fields 0,1 to indicate your first split is<br>
>> > the 0 and 1 basic fields and -pc_fieldsplit_1_fields 2,3 to indicate the<br>
>> > second split is the 2 and 3 basic fields.<br>
>> > (By default if you do not provide this then PCFIELDSPLIT will use bs<br>
>> > splits (4 in your case) one for each basic field.<br>
>> ><br>
>> > Barry<br>
>> ><br>
>> >> On Feb 8, 2016, at 11:20 AM, Hom Nath Gharti <<a href="mailto:hng.email@gmail.com">hng.email@gmail.com</a>><br>
>> >> wrote:<br>
>> >><br>
>> >> Hi Matt, Hi all,<br>
>> >><br>
>> >> I am trying to have some feel for PCFIELDSPLIT testing on a very small<br>
>> >> matrix (attached below). I have 4 degrees of freedoms. I use 4<br>
>> >> processors. I want to split 4 dofs into two fields each having two<br>
>> >> dofs. I don't know whether this my be a problem for petsc. When I use<br>
>> >> the command:<br>
>> >> srun -n 4 ./xtestfs -ksp_monitor -ksp_view<br>
>> >><br>
>> >> It runs fine.<br>
>> >><br>
>> >> But when I want to use field split options using the command:<br>
>> >> srun -n 4 ./xtestfs -ksp_monitor -ksp_type fgmres -pc_type fieldsplit<br>
>> >> -fieldsplit_0_ksp_type gmres -fieldsplit_0_pc_type bjacobi<br>
>> >> -fieldsplit_1_pc_type jacobi<br>
>> >><br>
>> >> I get the following error message:<br>
>> >> [0]PETSC ERROR: Petsc has generated inconsistent data<br>
>> >> [0]PETSC ERROR: Unhandled case, must have at least two fields, not 1<br>
>> >> [0]PETSC ERROR: See<br>
>> >> <a href="http://www.mcs.anl.gov/petsc/documentation/faq.html" rel="noreferrer" target="_blank">http://www.mcs.anl.gov/petsc/documentation/faq.html</a> for trouble<br>
>> >> shooting.<br>
>> >> [0]PETSC ERROR: Petsc Release Version 3.6.3, Dec, 03, 2015<br>
>> >> [0]PETSC ERROR: /tigress/hgharti/test/testfs/./xtestfs on a<br>
>> >> arch-linux2-c-debug named tiger-r11n1 by hgharti Mon Feb 8 11:40:03<br>
>> >> 2016<br>
>> >> [0]PETSC ERROR: Configure options<br>
>> >><br>
>> >> --with-blas-lapack-dir=/opt/intel/composer_xe_2015.2.164/mkl/lib/intel64/<br>
>> >> --with-cc=mpicc --with-cxx=mpicxx --with-fc=mpif90<br>
>> >> --with-mpiexec=mpiexec --with-debugging=1 --download-scalapack<br>
>> >> --download-mumps --download-pastix --download-superlu<br>
>> >> --download-superlu_dist --download-metis --download-parmetis<br>
>> >> --download-ptscotch --download-hypre<br>
>> >> [0]PETSC ERROR: #1 PCFieldSplitSetDefaults() line 469 in<br>
>> >><br>
>> >> /home/hgharti/lsoft/petsc-3.6.3-intel16-mpi5/src/ksp/pc/impls/fieldsplit/fieldsplit.c<br>
>> >> [0]PETSC ERROR: #2 PCSetUp_FieldSplit() line 486 in<br>
>> >><br>
>> >> /home/hgharti/lsoft/petsc-3.6.3-intel16-mpi5/src/ksp/pc/impls/fieldsplit/fieldsplit.c<br>
>> >> [0]PETSC ERROR: #3 PCSetUp() line 983 in<br>
>> >><br>
>> >> /home/hgharti/lsoft/petsc-3.6.3-intel16-mpi5/src/ksp/pc/interface/precon.c<br>
>> >> [0]PETSC ERROR: #4 KSPSetUp() line 332 in<br>
>> >><br>
>> >> /home/hgharti/lsoft/petsc-3.6.3-intel16-mpi5/src/ksp/ksp/interface/itfunc.c<br>
>> >> [0]PETSC ERROR: #5 KSPSolve() line 546 in<br>
>> >><br>
>> >> /home/hgharti/lsoft/petsc-3.6.3-intel16-mpi5/src/ksp/ksp/interface/itfunc.c<br>
>> >> WARNING! There are options you set that were not used!<br>
>> >> WARNING! could be spelling mistake, etc!<br>
>> >> Option left: name:-fieldsplit_0_ksp_type value: gmres<br>
>> >> Option left: name:-fieldsplit_0_pc_type value: bjacobi<br>
>> >> Option left: name:-fieldsplit_1_pc_type value: jacobi<br>
>> >> forrtl: error (76): Abort trap signal<br>
>> >><br>
>> >> I tried several trials but could not fix the problem. Is it the<br>
>> >> FORTRAN problem or am I doing something wrong? Any suggestions would<br>
>> >> be greatly appreciated.<br>
>> >> For your information I use:<br>
>> >> PETSc-3.6.3<br>
>> >> intel-16.0 compiler<br>
>> >> intel-mpi-5.1.1<br>
>> >><br>
>> >> Program<br>
>> >><br>
>> >><br>
>> >> Best,<br>
>> >> Hom Nath<br>
>> >><br>
>> >> ! simple test case for PCFIELDSPLIT<br>
>> >> !<br>
>> >> -----------------------------------------------------------------------<br>
>> >> program testfs<br>
>> >> implicit none<br>
>> >> #include "petsc/finclude/petscsys.h"<br>
>> >> #include "petsc/finclude/petscvec.h"<br>
>> >> #include "petsc/finclude/petscvec.h90"<br>
>> >> #include "petsc/finclude/petscmat.h"<br>
>> >> #include "petsc/finclude/petscksp.h"<br>
>> >> #include "petsc/finclude/petscpc.h"<br>
>> >> #include "petsc/finclude/petscviewer.h"<br>
>> >> #include "petsc/finclude/petscviewer.h90"<br>
>> >><br>
>> >> Vec x,b,u<br>
>> >> Mat A<br>
>> >> KSP ksp<br>
>> >> PC pc<br>
>> >> PetscErrorCode ierr<br>
>> >> PetscBool flg<br>
>> >> PetscInt errcode,its,maxitr,myrank,nproc<br>
>> >> PetscInt nedof,nelmt,ndof,nenod,ngdof,fixdof,nsparse,neq<br>
>> >> PetscInt,allocatable :: krow_sparse(:),ggdof_elmt(:,:), &<br>
>> >> inode_elmt(:,:)<br>
>> >><br>
>> >> PetscReal,allocatable :: storef(:,:),storekmat(:,:,:)<br>
>> >> PetscInt gdof0(2),gdof1(2)<br>
>> >><br>
>> >> ! initialize MPI<br>
>> >><br>
>> >> call MPI_INIT(errcode)<br>
>> >> if(errcode /= 0)stop 'ERROR: cannot initialize MPI!'<br>
>> >> call MPI_COMM_RANK(MPI_COMM_WORLD,myrank,errcode)<br>
>> >> if(errcode /= 0)stop 'ERROR: cannot find processor ID'<br>
>> >> call MPI_COMM_SIZE(MPI_COMM_WORLD,nproc,errcode)<br>
>> >> if(errcode /= 0)stop 'ERROR: cannot find number of processors!'<br>
>> >><br>
>> >> ! define some parameters<br>
>> >> ! 1D model with 4 elements. Each elements consits of 2 nodes. Node 0 is<br>
>> >> fixed.<br>
>> >> ! 1-----------2-----------3-----------4-----------5<br>
>> >> nelmt=1 ! per processor<br>
>> >> nenod=2 ! number of nodes per element<br>
>> >> nedof=2 ! number of DOFs per element<br>
>> >> ndof=2 ! number of DOFs<br>
>> >> ngdof=4 ! total number of global DOFs<br>
>> >> fixdof=1<br>
>> >><br>
>> >> if(myrank==0)then<br>
>> >> neq=1 ! number of equations<br>
>> >> nsparse=1<br>
>> >> else<br>
>> >> neq=2 ! number of equations<br>
>> >> nsparse=4<br>
>> >> endif<br>
>> >><br>
>> >> allocate(storef(nedof,nelmt),storekmat(nedof,nedof,nelmt),<br>
>> >> &<br>
>> >> ggdof_elmt(nedof,nelmt),inode_elmt(nenod,nelmt),krow_sparse(nsparse))<br>
>> >><br>
>> >> storef=0.0_8 ! local RHS vector<br>
>> >> storekmat=0.0_8 ! local stiffness matrix<br>
>> >> if(myrank==0)then<br>
>> >> krow_sparse(:)=1<br>
>> >> storef(:,1)=(/12.5000_8, 12.5000_8/)<br>
>> >> storekmat(:,:,1) = reshape((/1.2333333332_8, 0.0166666667_8, &<br>
>> >> 0.0166666667_8,<br>
>> >> 1.2333333332_8/),(/nedof,nedof/))<br>
>> >> inode_elmt(:,1) = (/1,2/)<br>
>> >> ggdof_elmt(:,1) = (/0,1/) ! 0 because left node is fixed<br>
>> >> elseif(myrank==1)then<br>
>> >> krow_sparse(:)=(/1,1,2,2/)<br>
>> >> storef(:,1)=(/12.5000_8, 12.5000_8/)<br>
>> >> storekmat(:,:,1) = reshape((/1.2333333332_8, 0.0166666667_8, &<br>
>> >> 0.0166666667_8,<br>
>> >> 1.2333333332_8/),(/nedof,nedof/))<br>
>> >> inode_elmt(:,1) = (/1,2/)<br>
>> >> ggdof_elmt(:,1) = (/1,2/)<br>
>> >> elseif(myrank==2)then<br>
>> >> krow_sparse(:)=(/1,1,2,2/)<br>
>> >> storef(:,1)=(/12.5000_8, 12.5000_8/)<br>
>> >> storekmat(:,:,1) = reshape((/1.2333333332_8, 0.0166666667_8, &<br>
>> >> 0.0166666667_8,<br>
>> >> 1.2333333332_8/),(/nedof,nedof/))<br>
>> >> inode_elmt(:,1) = (/1,2/)<br>
>> >> ggdof_elmt(:,1) = (/2,3/)<br>
>> >> elseif(myrank==3)then<br>
>> >> krow_sparse(:)=(/1,1,2,2/)<br>
>> >> storef(:,1)=(/12.5000_8, 22.5000_8/)<br>
>> >> storekmat(:,:,1) = reshape((/1.2333333332_8, 0.0166666667_8, &<br>
>> >> 0.0166666667_8,<br>
>> >> 1.2333333332_8/),(/nedof,nedof/))<br>
>> >> inode_elmt(:,1) = (/1,2/)<br>
>> >> ggdof_elmt(:,1) = (/3,4/)<br>
>> >> endif<br>
>> >><br>
>> >> call petsc_initialize<br>
>> >> call petsc_set_matrix<br>
>> >> call petsc_set_vector<br>
>> >> call petsc_solve<br>
>> >> call petsc_finalize<br>
>> >><br>
>> >> ! deallocate<br>
>> >> deallocate(storef,storekmat,ggdof_elmt,inode_elmt)<br>
>> >> deallocate(krow_sparse)<br>
>> >> contains<br>
>> >><br>
>> >> !-------------------------------------------------------------------------------<br>
>> >> subroutine petsc_initialize()<br>
>> >> implicit none<br>
>> >> PetscInt :: istart,iend<br>
>> >> PetscInt :: nzeros_max,nzeros_min<br>
>> >> PetscReal :: atol,dtol,rtol<br>
>> >> PetscInt,allocatable :: nzeros(:)<br>
>> >> IS gdof0_is,gdof1_is<br>
>> >><br>
>> >> call PetscInitialize(PETSC_NULL_CHARACTER,ierr)<br>
>> >> call PetscOptionsGetInt(PETSC_NULL_CHARACTER,'-n',ngdof,flg,ierr)<br>
>> >><br>
>> >> ! count number of nonzeros per row<br>
>> >> allocate(nzeros(neq))<br>
>> >> nzeros=0<br>
>> >> nzeros(krow_sparse)=nzeros(krow_sparse)+1<br>
>> >> nzeros_max=maxval(nzeros)<br>
>> >> nzeros_min=minval(nzeros)<br>
>> >><br>
>> >> ! create matrix<br>
>> >> call MatCreate(PETSC_COMM_WORLD,A,ierr)<br>
>> >> call MatSetType(A,MATMPIAIJ,ierr)<br>
>> >> CHKERRQ(ierr)<br>
>> >> call MatSetSizes(A,PETSC_DECIDE,PETSC_DECIDE,ngdof,ngdof,ierr)<br>
>> >> CHKERRQ(ierr)<br>
>> >> ! preallocation<br>
>> >> call<br>
>> >> MatMPIAIJSetPreallocation(A,nzeros_max,PETSC_NULL_INTEGER,nzeros_max, &<br>
>> >> PETSC_NULL_INTEGER,ierr)<br>
>> >> CHKERRQ(ierr)<br>
>> >><br>
>> >> call MatGetOwnershipRange(A,istart,iend,ierr)<br>
>> >> CHKERRQ(ierr)<br>
>> >> print*,'ownership:',myrank,istart,iend<br>
>> >> ! create vector<br>
>> >> call VecCreate(PETSC_COMM_WORLD,x,ierr)<br>
>> >> call VecSetSizes(x,PETSC_DECIDE,ngdof,ierr)<br>
>> >> call VecSetType(x,VECMPI,ierr)<br>
>> >> call VecDuplicate(x,b,ierr)<br>
>> >> call VecDuplicate(x,u,ierr)<br>
>> >><br>
>> >><br>
>> >> !*******************************************************************************<br>
>> >> ! get IS for split fields<br>
>> >> gdof0=(/0,1/)<br>
>> >> gdof1=(/2,3/)<br>
>> >> call<br>
>> >> ISCreateGeneral(PETSC_COMM_WORLD,2,gdof0,PETSC_COPY_VALUES,gdof0_is,ierr)<br>
>> >> CHKERRQ(ierr)<br>
>> >> call<br>
>> >> ISCreateGeneral(PETSC_COMM_WORLD,2,gdof1,PETSC_COPY_VALUES,gdof1_is,ierr)<br>
>> >> CHKERRQ(ierr)<br>
>> >><br>
>> >> !*******************************************************************************<br>
>> >><br>
>> >> ! Create linear solver context<br>
>> >> call KSPCreate(PETSC_COMM_WORLD,ksp,ierr)<br>
>> >> call KSPSetOperators(ksp,A,A,ierr)<br>
>> >><br>
>> >> call KSPGetPC(ksp,pc,ierr)<br>
>> >> CHKERRQ(ierr)<br>
>> >><br>
>> >><br>
>> >> !*******************************************************************************<br>
>> >> ! split PC<br>
>> >> call PCFieldSplitSetIS(pc,"0",gdof0_is,ierr);<br>
>> >> CHKERRQ(ierr)<br>
>> >> call ISDestroy(gdof0_is,ierr)<br>
>> >> CHKERRQ(ierr)<br>
>> >> call PCFieldSplitSetIS(pc,"1",gdof1_is,ierr);<br>
>> >> CHKERRQ(ierr)<br>
>> >> call ISDestroy(gdof1_is,ierr)<br>
>> >> CHKERRQ(ierr)<br>
>> >><br>
>> >> !*******************************************************************************<br>
>> >><br>
>> >> rtol = 1.0d-6<br>
>> >> atol = 1.0d-6<br>
>> >> dtol = 1.0d10<br>
>> >> maxitr = 1000<br>
>> >> call KSPSetTolerances(ksp,rtol,atol,dtol,maxitr,ierr)<br>
>> >> CHKERRQ(ierr)<br>
>> >> call KSPSetFromOptions(ksp,ierr)<br>
>> >> end subroutine petsc_initialize<br>
>> >><br>
>> >> !-------------------------------------------------------------------------------<br>
>> >><br>
>> >> subroutine petsc_set_matrix()<br>
>> >><br>
>> >> implicit none<br>
>> >> integer :: i,i_elmt,j,ncount<br>
>> >> integer :: ggdof(NEDOF),idof(NEDOF),igdof(NEDOF)<br>
>> >><br>
>> >> ! Set and assemble matrix.<br>
>> >> call MatZeroEntries(A,ierr)<br>
>> >> CHKERRQ(ierr)<br>
>> >> do i_elmt=1,nelmt<br>
>> >> ggdof(:)=ggdof_elmt(:,i_elmt)<br>
>> >> ggdof(:)=ggdof(:)-1 ! petsc index starts from 0<br>
>> >> ncount=0; idof=-1; igdof=-1<br>
>> >> do i=1,NEDOF<br>
>> >> do j=1,NEDOF<br>
>> >> if(ggdof(i).ge.0.and.ggdof(j).ge.0)then<br>
>> >> call MatSetValues(A,1,ggdof(i),1,ggdof(j),storekmat(i,j,i_elmt),<br>
>> >> &<br>
>> >> ADD_VALUES,ierr)<br>
>> >> CHKERRQ(ierr)<br>
>> >> endif<br>
>> >> enddo<br>
>> >> enddo<br>
>> >> enddo<br>
>> >> call MatAssemblyBegin(A,MAT_FINAL_ASSEMBLY,ierr)<br>
>> >> call MatAssemblyEnd(A,MAT_FINAL_ASSEMBLY,ierr)<br>
>> >> if(myrank==0)print*,'matrix setting & assembly complete!'<br>
>> >><br>
>> >> end subroutine petsc_set_matrix<br>
>> >><br>
>> >> !-------------------------------------------------------------------------------<br>
>> >><br>
>> >> subroutine petsc_set_vector()<br>
>> >> implicit none<br>
>> >> PetscScalar zero<br>
>> >> integer :: i,i_elmt,ncount<br>
>> >> integer :: ggdof(NEDOF),idof(NEDOF),igdof(NEDOF)<br>
>> >><br>
>> >> ! set vector<br>
>> >> zero=0.0_8<br>
>> >> call VecSet(b,zero,ierr)<br>
>> >> do i_elmt=1,nelmt<br>
>> >> ggdof(:)=ggdof_elmt(:,i_elmt)<br>
>> >> ggdof(:)=ggdof(:)-1 ! petsc index starts from 0<br>
>> >> ncount=0; idof=-1; igdof=-1<br>
>> >> do i=1,NEDOF<br>
>> >> if(ggdof(i).ge.0)then<br>
>> >> call VecSetValues(b,1,ggdof(i),storef(i,i_elmt),ADD_VALUES,ierr);<br>
>> >> CHKERRQ(ierr)<br>
>> >> endif<br>
>> >> enddo<br>
>> >> enddo<br>
>> >><br>
>> >> ! assemble vector<br>
>> >> call VecAssemblyBegin(b,ierr)<br>
>> >> call VecAssemblyEnd(b,ierr)<br>
>> >> if(myrank==0)print*,'vector setting & assembly complete!'<br>
>> >><br>
>> >> end subroutine petsc_set_vector<br>
>> >><br>
>> >> !-------------------------------------------------------------------------------<br>
>> >><br>
>> >> subroutine petsc_solve()<br>
>> >> implicit none<br>
>> >> PetscInt ireason<br>
>> >> PetscScalar a_v(1)<br>
>> >> PetscOffset a_i<br>
>> >> PetscInt n<br>
>> >> PetscReal,allocatable :: sdata(:)<br>
>> >><br>
>> >> call VecGetSize(b,n,ierr)<br>
>> >> CHKERRQ(ierr)<br>
>> >> allocate(sdata(n))<br>
>> >> sdata=0.0_8<br>
>> >> call VecGetArray(b,a_v,a_i,ierr)<br>
>> >> CHKERRQ(ierr)<br>
>> >> sdata(1:n)=a_v(a_i+1:a_i+n)<br>
>> >> call VecRestoreArray(b,a_v,a_i,ierr)<br>
>> >> CHKERRQ(ierr)<br>
>> >> print*,'rhs:',sdata<br>
>> >><br>
>> >> call KSPSolve(ksp,b,x,ierr)<br>
>> >> sdata=0.0_8<br>
>> >> call VecGetArray(x,a_v,a_i,ierr)<br>
>> >> CHKERRQ(ierr)<br>
>> >> sdata(1:n)=a_v(a_i+1:a_i+n)<br>
>> >> call VecRestoreArray(b,a_v,a_i,ierr)<br>
>> >> CHKERRQ(ierr)<br>
>> >> print*,'solution:',sdata<br>
>> >><br>
>> >> ! Check convergence<br>
>> >> call KSPGetConvergedReason(ksp,ireason,ierr)<br>
>> >> if(myrank==0)print*,'converges reason',myrank,ireason<br>
>> >> call KSPGetIterationNumber(ksp,its,ierr)<br>
>> >> if(myrank==0)print*,'Iterations:',its<br>
>> >> deallocate(sdata)<br>
>> >> end subroutine petsc_solve<br>
>> >><br>
>> >> !-------------------------------------------------------------------------------<br>
>> >><br>
>> >> subroutine petsc_finalize()<br>
>> >> implicit none<br>
>> >><br>
>> >> ! Free work space.<br>
>> >> call VecDestroy(x,ierr)<br>
>> >> call VecDestroy(u,ierr)<br>
>> >> call VecDestroy(b,ierr)<br>
>> >> call MatDestroy(A,ierr)<br>
>> >> call KSPDestroy(ksp,ierr)<br>
>> >> call PetscFinalize(ierr)<br>
>> >> CHKERRQ(ierr)<br>
>> >><br>
>> >> end subroutine petsc_finalize<br>
>> >><br>
>> >> !-------------------------------------------------------------------------------<br>
>> >><br>
>> >> end program testfs<br>
>> >><br>
>> >> On Tue, Feb 2, 2016 at 4:54 PM, Matthew Knepley <<a href="mailto:knepley@gmail.com">knepley@gmail.com</a>><br>
>> >> wrote:<br>
>> >>> On Tue, Feb 2, 2016 at 3:11 PM, Hom Nath Gharti <<a href="mailto:hng.email@gmail.com">hng.email@gmail.com</a>><br>
>> >>> wrote:<br>
>> >>>><br>
>> >>>> Thanks a lot. I will try.<br>
>> >>><br>
>> >>><br>
>> >>> Also, if you send a small test case, we can run it and tell you the<br>
>> >>> problem.<br>
>> >>><br>
>> >>> Matt<br>
>> >>><br>
>> >>>><br>
>> >>>> Hom Nath<br>
>> >>>><br>
>> >>>> On Tue, Feb 2, 2016 at 4:02 PM, Matthew Knepley <<a href="mailto:knepley@gmail.com">knepley@gmail.com</a>><br>
>> >>>> wrote:<br>
>> >>>>> On Tue, Feb 2, 2016 at 3:01 PM, Hom Nath Gharti<br>
>> >>>>> <<a href="mailto:hng.email@gmail.com">hng.email@gmail.com</a>><br>
>> >>>>> wrote:<br>
>> >>>>>><br>
>> >>>>>> Thanks again Matt. Unfortunately got the same errors with '0'. I<br>
>> >>>>>> think<br>
>> >>>>>> both are valid in Fortran.<br>
>> >>>>><br>
>> >>>>><br>
>> >>>>> Then you will have to go in the debugger and see why its not<br>
>> >>>>> creating 4<br>
>> >>>>> splits, since this is exactly<br>
>> >>>>> what it does in our tests. This is how all the SNES tests that I use<br>
>> >>>>> work. I<br>
>> >>>>> am sure its a Fortran<br>
>> >>>>> problem.<br>
>> >>>>><br>
>> >>>>> Thanks,<br>
>> >>>>><br>
>> >>>>> Matt<br>
>> >>>>><br>
>> >>>>>><br>
>> >>>>>> Hom Nath<br>
>> >>>>>><br>
>> >>>>>> On Tue, Feb 2, 2016 at 3:42 PM, Matthew Knepley <<a href="mailto:knepley@gmail.com">knepley@gmail.com</a>><br>
>> >>>>>> wrote:<br>
>> >>>>>>> On Tue, Feb 2, 2016 at 2:35 PM, Hom Nath Gharti<br>
>> >>>>>>> <<a href="mailto:hng.email@gmail.com">hng.email@gmail.com</a>><br>
>> >>>>>>> wrote:<br>
>> >>>>>>>><br>
>> >>>>>>>> Thank you so much Matt.<br>
>> >>>>>>>><br>
>> >>>>>>>> I now tried the following:<br>
>> >>>>>>>><br>
>> >>>>>>>> ======================================================<br>
>> >>>>>>>> call KSPCreate(PETSC_COMM_WORLD,ksp,ierr)<br>
>> >>>>>>>> call KSPGetPC(ksp,pc,ierr)<br>
>> >>>>>>>><br>
>> >>>>>>>> call PCFieldSplitSetIS(pc,"0",gdofu_is,ierr);<br>
>> >>>>>>>> call ISDestroy(gdofu_is,ierr)<br>
>> >>>>>>>> call PCFieldSplitSetIS(pc,"1",gdofchi_is,ierr);<br>
>> >>>>>>>> call ISDestroy(gdofchi_is,ierr)<br>
>> >>>>>>>> call PCFieldSplitSetIS(pc,"2",gdofp_is,ierr);<br>
>> >>>>>>>> call ISDestroy(gdofp_is,ierr)<br>
>> >>>>>>>> call PCFieldSplitSetIS(pc,"3",gdofphi_is,ierr);<br>
>> >>>>>>>> call ISDestroy(gdofphi_is,ierr)<br>
>> >>>>>>>><br>
>> >>>>>>>> ! Amat changes in each time steps<br>
>> >>>>>>>> call KSPSetOperators(ksp,Amat,Amat,ierr) !version >= 3.5.0<br>
>> >>>>>>>><br>
>> >>>>>>>> call KSPSolve(ksp,bvec,xvec,ierr)<br>
>> >>>>>>>> ======================================================<br>
>> >>>>>>><br>
>> >>>>>>><br>
>> >>>>>>> I am guessing that "0" is not a valid string for your Fortran<br>
>> >>>>>>> compiler.<br>
>> >>>>>>> Are<br>
>> >>>>>>> you sure<br>
>> >>>>>>> its not single quotes '0'?<br>
>> >>>>>>><br>
>> >>>>>>> Matt<br>
>> >>>>>>><br>
>> >>>>>>>><br>
>> >>>>>>>> But I get the following error:<br>
>> >>>>>>>><br>
>> >>>>>>>> [0]PETSC ERROR: --------------------- Error Message<br>
>> >>>>>>>> --------------------------------------------------------------<br>
>> >>>>>>>> [0]PETSC ERROR: Petsc has generated inconsistent data<br>
>> >>>>>>>> [0]PETSC ERROR: Unhandled case, must have at least two fields,<br>
>> >>>>>>>> not 1<br>
>> >>>>>>>> [0]PETSC ERROR: See<br>
>> >>>>>>>> <a href="http://www.mcs.anl.gov/petsc/documentation/faq.html" rel="noreferrer" target="_blank">http://www.mcs.anl.gov/petsc/documentation/faq.html</a> for trouble<br>
>> >>>>>>>> shooting.<br>
>> >>>>>>>> [0]PETSC ERROR: Petsc Release Version 3.6.3, Dec, 03, 2015<br>
>> >>>>>>>> [0]PETSC ERROR:<br>
>> >>>>>>>> /tigress/hgharti/gitwork/SPECFEM3D_GLOBEVSPP/./bin/xspecfem3D on<br>
>> >>>>>>>> a<br>
>> >>>>>>>> arch-linux2-c-debug named tiger-r3c1n7 by hgharti Tue Feb 2 15:<br>
>> >>>>>>>> 29:30 2016<br>
>> >>>>>>>> [0]PETSC ERROR: Configure options<br>
>> >>>>>>>><br>
>> >>>>>>>><br>
>> >>>>>>>><br>
>> >>>>>>>> --with-blas-lapack-dir=/opt/intel/composer_xe_2015.2.164/mkl/lib/intel64/<br>
>> >>>>>>>> --with-cc=mpicc --with-cxx=mpicxx --wit<br>
>> >>>>>>>> h-fc=mpif90 --with-mpiexec=mpiexec --with-debugging=1<br>
>> >>>>>>>> --download-scalapack --download-mumps --download-pastix<br>
>> >>>>>>>> --download-superlu --download-superlu_dist --download-metis<br>
>> >>>>>>>> --download-parmetis --download-ptscotch --download-hypre<br>
>> >>>>>>>> [0]PETSC ERROR: #1 PCFieldSplitSetDefaults() line 469 in<br>
>> >>>>>>>><br>
>> >>>>>>>><br>
>> >>>>>>>><br>
>> >>>>>>>><br>
>> >>>>>>>> /home/hgharti/lsoft/petsc-3.6.3-intel16-mpi5/src/ksp/pc/impls/fieldsplit/fieldsplit.c<br>
>> >>>>>>>> [0]PETSC ERROR: #2 PCSetUp_FieldSplit() line 486 in<br>
>> >>>>>>>><br>
>> >>>>>>>><br>
>> >>>>>>>><br>
>> >>>>>>>><br>
>> >>>>>>>> /home/hgharti/lsoft/petsc-3.6.3-intel16-mpi5/src/ksp/pc/impls/fieldsplit/fieldsplit.c<br>
>> >>>>>>>> [0]PETSC ERROR: #3 PCSetUp() line 983 in<br>
>> >>>>>>>><br>
>> >>>>>>>><br>
>> >>>>>>>><br>
>> >>>>>>>> /home/hgharti/lsoft/petsc-3.6.3-intel16-mpi5/src/ksp/pc/interface/precon.c<br>
>> >>>>>>>> [0]PETSC ERROR: #4 KSPSetUp() line 332 in<br>
>> >>>>>>>><br>
>> >>>>>>>><br>
>> >>>>>>>><br>
>> >>>>>>>><br>
>> >>>>>>>> /home/hgharti/lsoft/petsc-3.6.3-intel16-mpi5/src/ksp/ksp/interface/itfunc.c<br>
>> >>>>>>>> [0]PETSC ERROR: #5 KSPSolve() line 546 in<br>
>> >>>>>>>><br>
>> >>>>>>>><br>
>> >>>>>>>><br>
>> >>>>>>>><br>
>> >>>>>>>> /home/hgharti/lsoft/petsc-3.6.3-intel16-mpi5/src/ksp/ksp/interface/itfunc.c<br>
>> >>>>>>>> forrtl: error (76): Abort trap signal<br>
>> >>>>>>>><br>
>> >>>>>>>> Am I missing something?<br>
>> >>>>>>>><br>
>> >>>>>>>> Thanks,<br>
>> >>>>>>>> Hom Nath<br>
>> >>>>>>>><br>
>> >>>>>>>> On Tue, Feb 2, 2016 at 3:24 PM, Matthew Knepley<br>
>> >>>>>>>> <<a href="mailto:knepley@gmail.com">knepley@gmail.com</a>><br>
>> >>>>>>>> wrote:<br>
>> >>>>>>>>> On Tue, Feb 2, 2016 at 2:20 PM, Hom Nath Gharti<br>
>> >>>>>>>>> <<a href="mailto:hng.email@gmail.com">hng.email@gmail.com</a>><br>
>> >>>>>>>>> wrote:<br>
>> >>>>>>>>>><br>
>> >>>>>>>>>> Hi Matt, hi all,<br>
>> >>>>>>>>>><br>
>> >>>>>>>>>> I am trying to use PCFIELDSPLIT for my solver which consists of<br>
>> >>>>>>>>>> 4<br>
>> >>>>>>>>>> different variables, namely, u (displacement vector), \chi<br>
>> >>>>>>>>>> (displacement potential), p(pressure), and \phi (gravity<br>
>> >>>>>>>>>> potential).<br>
>> >>>>>>>>>><br>
>> >>>>>>>>>> My code segment looks like the following:<br>
>> >>>>>>>>>> ======================================================<br>
>> >>>>>>>>>> call KSPCreate(PETSC_COMM_WORLD,ksp,ierr)<br>
>> >>>>>>>>>> call KSPGetPC(ksp,pc,ierr)<br>
>> >>>>>>>>>> nsplit=4<br>
>> >>>>>>>>>> call PCFieldSplitSetBlockSize(pc,nsplit,ierr);<br>
>> >>>>>>>>><br>
>> >>>>>>>>><br>
>> >>>>>>>>> You do not need the statement above.<br>
>> >>>>>>>>><br>
>> >>>>>>>>>><br>
>> >>>>>>>>>> call PCFieldSplitSetIS(pc,"0",gdofu_is,ierr);<br>
>> >>>>>>>>>> call ISDestroy(gdofu_is,ierr)<br>
>> >>>>>>>>>> call PCFieldSplitSetIS(pc,"1",gdofchi_is,ierr);<br>
>> >>>>>>>>>> call ISDestroy(gdofchi_is,ierr)<br>
>> >>>>>>>>>> call PCFieldSplitSetIS(pc,"2",gdofp_is,ierr);<br>
>> >>>>>>>>>> call ISDestroy(gdofp_is,ierr)<br>
>> >>>>>>>>>> call PCFieldSplitSetIS(pc,"3",gdofphi_is,ierr);<br>
>> >>>>>>>>>> call ISDestroy(gdofphi_is,ierr)<br>
>> >>>>>>>>>><br>
>> >>>>>>>>>> call PCFieldSplitGetSubKSP(pc,nsplit,subksp,ierr);<br>
>> >>>>>>>>><br>
>> >>>>>>>>><br>
>> >>>>>>>>> These SetOperators() calls are wrong. I am not sure why you want<br>
>> >>>>>>>>> them<br>
>> >>>>>>>>> here.<br>
>> >>>>>>>>> Also, that means you do not need the call above.<br>
>> >>>>>>>>><br>
>> >>>>>>>>> Thanks,<br>
>> >>>>>>>>><br>
>> >>>>>>>>> Matt<br>
>> >>>>>>>>><br>
>> >>>>>>>>>><br>
>> >>>>>>>>>> call KSPSetOperators(subksp(1),Amat,Amat,ierr);<br>
>> >>>>>>>>>> call KSPSetOperators(subksp(2),Amat,Amat,ierr);<br>
>> >>>>>>>>>> call KSPSetOperators(subksp(3),Amat,Amat,ierr);<br>
>> >>>>>>>>>> call KSPSetOperators(subksp(4),Amat,Amat,ierr);<br>
>> >>>>>>>>>><br>
>> >>>>>>>>>> call KSPSolve(ksp,bvec,xvec,ierr)<br>
>> >>>>>>>>>> ======================================================<br>
>> >>>>>>>>>><br>
>> >>>>>>>>>> But I am getting the following error:<br>
>> >>>>>>>>>> [79]PETSC ERROR: Null argument, when expecting valid pointer<br>
>> >>>>>>>>>> [79]PETSC ERROR: Null Object: Parameter # 1<br>
>> >>>>>>>>>> [79]PETSC ERROR: #1 KSPSetOperators() line 536 in<br>
>> >>>>>>>>>> /home/hgharti/lsoft/petsc-3.6.3-intel16-mpi5/src/ksp/ksp/interf<br>
>> >>>>>>>>>><br>
>> >>>>>>>>>> It looks like I am doing something wrong in "call<br>
>> >>>>>>>>>> KSPSetOperators"<br>
>> >>>>>>>>>> but<br>
>> >>>>>>>>>> I could not figure that out.<br>
>> >>>>>>>>>><br>
>> >>>>>>>>>> Could anybody help me to fix this problem? I looked into almost<br>
>> >>>>>>>>>> all<br>
>> >>>>>>>>>> related examples but I could not really figure out the correct<br>
>> >>>>>>>>>> steps<br>
>> >>>>>>>>>> after "call PCFieldSplitSetIS".<br>
>> >>>>>>>>>><br>
>> >>>>>>>>>> Any help will be greatly appreciated.<br>
>> >>>>>>>>>><br>
>> >>>>>>>>>> Best,<br>
>> >>>>>>>>>> Hom nath<br>
>> >>>>>>>>>><br>
>> >>>>>>>>>> On Sun, Jan 24, 2016 at 7:14 PM, Hom Nath Gharti<br>
>> >>>>>>>>>> <<a href="mailto:hng.email@gmail.com">hng.email@gmail.com</a>><br>
>> >>>>>>>>>> wrote:<br>
>> >>>>>>>>>>> Thank you so much Matt! I will try.<br>
>> >>>>>>>>>>><br>
>> >>>>>>>>>>> Hom Nath<br>
>> >>>>>>>>>>><br>
>> >>>>>>>>>>> On Sun, Jan 24, 2016 at 6:26 AM, Matthew Knepley<br>
>> >>>>>>>>>>> <<a href="mailto:knepley@gmail.com">knepley@gmail.com</a>><br>
>> >>>>>>>>>>> wrote:<br>
>> >>>>>>>>>>>> On Fri, Jan 22, 2016 at 2:19 PM, Hom Nath Gharti<br>
>> >>>>>>>>>>>> <<a href="mailto:hng.email@gmail.com">hng.email@gmail.com</a>><br>
>> >>>>>>>>>>>> wrote:<br>
>> >>>>>>>>>>>>><br>
>> >>>>>>>>>>>>> Dear all,<br>
>> >>>>>>>>>>>>><br>
>> >>>>>>>>>>>>> I am new to PcFieldSplit.<br>
>> >>>>>>>>>>>>><br>
>> >>>>>>>>>>>>> I have a matrix formed using MATMPIAIJ. Is it possible to<br>
>> >>>>>>>>>>>>> use<br>
>> >>>>>>>>>>>>> PCFIELDSPLIT operations in this type of matrix? Or does it<br>
>> >>>>>>>>>>>>> have<br>
>> >>>>>>>>>>>>> to<br>
>> >>>>>>>>>>>>> be<br>
>> >>>>>>>>>>>>> MATMPIBIJ or MATNEST format?<br>
>> >>>>>>>>>>>><br>
>> >>>>>>>>>>>><br>
>> >>>>>>>>>>>> Yes, you can split AIJ.<br>
>> >>>>>>>>>>>><br>
>> >>>>>>>>>>>>><br>
>> >>>>>>>>>>>>> If possible for MATMPIAIJ, could anybody provide me a simple<br>
>> >>>>>>>>>>>>> example<br>
>> >>>>>>>>>>>>> or few steps? Variables in the equations are displacement<br>
>> >>>>>>>>>>>>> vector,<br>
>> >>>>>>>>>>>>> scalar potential and pressure.<br>
>> >>>>>>>>>>>><br>
>> >>>>>>>>>>>><br>
>> >>>>>>>>>>>> If you do not have a collocated discretization, then you have<br>
>> >>>>>>>>>>>> to<br>
>> >>>>>>>>>>>> use<br>
>> >>>>>>>>>>>><br>
>> >>>>>>>>>>>><br>
>> >>>>>>>>>>>><br>
>> >>>>>>>>>>>><br>
>> >>>>>>>>>>>><br>
>> >>>>>>>>>>>><br>
>> >>>>>>>>>>>><br>
>> >>>>>>>>>>>> <a href="http://www.mcs.anl.gov/petsc/petsc-current/docs/manualpages/PC/PCFieldSplitSetIS.html" rel="noreferrer" target="_blank">http://www.mcs.anl.gov/petsc/petsc-current/docs/manualpages/PC/PCFieldSplitSetIS.html</a><br>
>> >>>>>>>>>>>><br>
>> >>>>>>>>>>>> Thanks,<br>
>> >>>>>>>>>>>><br>
>> >>>>>>>>>>>> Matt<br>
>> >>>>>>>>>>>><br>
>> >>>>>>>>>>>>><br>
>> >>>>>>>>>>>>> Thanks for help.<br>
>> >>>>>>>>>>>>><br>
>> >>>>>>>>>>>>> Hom Nath<br>
>> >>>>>>>>>>>><br>
>> >>>>>>>>>>>><br>
>> >>>>>>>>>>>><br>
>> >>>>>>>>>>>><br>
>> >>>>>>>>>>>> --<br>
>> >>>>>>>>>>>> What most experimenters take for granted before they begin<br>
>> >>>>>>>>>>>> their<br>
>> >>>>>>>>>>>> experiments<br>
>> >>>>>>>>>>>> is infinitely more interesting than any results to which<br>
>> >>>>>>>>>>>> their<br>
>> >>>>>>>>>>>> experiments<br>
>> >>>>>>>>>>>> lead.<br>
>> >>>>>>>>>>>> -- Norbert Wiener<br>
>> >>>>>>>>><br>
>> >>>>>>>>><br>
>> >>>>>>>>><br>
>> >>>>>>>>><br>
>> >>>>>>>>> --<br>
>> >>>>>>>>> What most experimenters take for granted before they begin their<br>
>> >>>>>>>>> experiments<br>
>> >>>>>>>>> is infinitely more interesting than any results to which their<br>
>> >>>>>>>>> experiments<br>
>> >>>>>>>>> lead.<br>
>> >>>>>>>>> -- Norbert Wiener<br>
>> >>>>>>><br>
>> >>>>>>><br>
>> >>>>>>><br>
>> >>>>>>><br>
>> >>>>>>> --<br>
>> >>>>>>> What most experimenters take for granted before they begin their<br>
>> >>>>>>> experiments<br>
>> >>>>>>> is infinitely more interesting than any results to which their<br>
>> >>>>>>> experiments<br>
>> >>>>>>> lead.<br>
>> >>>>>>> -- Norbert Wiener<br>
>> >>>>><br>
>> >>>>><br>
>> >>>>><br>
>> >>>>><br>
>> >>>>> --<br>
>> >>>>> What most experimenters take for granted before they begin their<br>
>> >>>>> experiments<br>
>> >>>>> is infinitely more interesting than any results to which their<br>
>> >>>>> experiments<br>
>> >>>>> lead.<br>
>> >>>>> -- Norbert Wiener<br>
>> >>><br>
>> >>><br>
>> >>><br>
>> >>><br>
>> >>> --<br>
>> >>> What most experimenters take for granted before they begin their<br>
>> >>> experiments<br>
>> >>> is infinitely more interesting than any results to which their<br>
>> >>> experiments<br>
>> >>> lead.<br>
>> >>> -- Norbert Wiener<br>
>> ><br>
><br>
><br>
><br>
><br>
> --<br>
> What most experimenters take for granted before they begin their experiments<br>
> is infinitely more interesting than any results to which their experiments<br>
> lead.<br>
> -- Norbert Wiener<br>
</div></div></blockquote></div><br><br clear="all"><div><br></div>-- <br><div class="gmail_signature">What most experimenters take for granted before they begin their experiments is infinitely more interesting than any results to which their experiments lead.<br>-- Norbert Wiener</div>
</div></div>