<html>
<head>
<meta content="text/html; charset=ISO-8859-1"
http-equiv="Content-Type">
</head>
<body bgcolor="#FFFFFF" text="#000000">
<font face="Ubuntu">Hi Karli,<br>
<br>
thank you for your hint: now it works. <br>
Now I would like to speed up the solution: I was counting on
increasing the number of levels/the number of processors used, but
now I see I cannot do that.<br>
Do you have any hint to achieve better speed?<br>
Thanks!<br>
<br>
Best,<br>
Michele<br>
<br>
</font>
<div class="moz-cite-prefix">On 08/13/2013 01:33 PM, Karl Rupp
wrote:<br>
</div>
<blockquote cite="mid:520A9815.7030400@mcs.anl.gov" type="cite">Hi
Michele,
<br>
<br>
I suggest you try a different decomposition of your grid. With k
levels, you should have at least 2^{k-1} grid nodes per coordinate
direction in order to be able to correctly build a coarser mesh.
In your case, you should have at least 8 nodes (leading to coarser
levels of size 4, 2, and 1) in z direction.
<br>
<br>
Best regards,
<br>
Karli
<br>
<br>
<br>
On 08/13/2013 02:28 PM, Michele Rosso wrote:
<br>
<blockquote type="cite">Hi Barry,
<br>
<br>
I was finally able to try multigrid with a singular system and a
finer grid.
<br>
GAMG works perfectly and has no problem in handling the singular
system.
<br>
On the other hand, MG is giving me problem:
<br>
<br>
[0]PETSC ERROR: --------------------- Error Message
<br>
------------------------------------
<br>
[0]PETSC ERROR: Argument out of range!
<br>
[0]PETSC ERROR: Partition in x direction is too fine! 32 64!
<br>
[0]PETSC ERROR:
<br>
------------------------------------------------------------------------
<br>
[0]PETSC ERROR: Petsc Release Version 3.4.2, Jul, 02, 2013
<br>
[0]PETSC ERROR: See docs/changes/index.html for recent updates.
<br>
[0]PETSC ERROR: See docs/faq.html for hints about trouble
shooting.
<br>
[0]PETSC ERROR: See docs/index.html for manual pages.
<br>
[0]PETSC ERROR:
<br>
------------------------------------------------------------------------
<br>
[0]PETSC ERROR: ./hit on a arch-cray-xt5-pkgs-opt named nid01332
by
<br>
Unknown Tue Aug 13 15:06:21 2013
<br>
[0]PETSC ERROR: Libraries linked from
<br>
/nics/c/home/mrosso/LIBS/petsc-3.4.2/arch-cray-xt5-pkgs-opt/lib
<br>
[0]PETSC ERROR: Configure run at Wed Jul 31 22:48:06 2013
<br>
<br>
The input I used is:
<br>
-ksp_monitor -ksp_converged_reason -pc_type mg -pc_mg_galerkin
<br>
-pc_mg_levels 4 -options_left
<br>
<br>
I am simulating a 256^3 grid with 256 processors. Since I am
using a 2D
<br>
domain decomposition, each sub-domain contains 256x64x4 grid
points.
<br>
To be consistent with my code indexing, I had to initialize DMDA
with
<br>
reverse ordering, that is z,y,x, so when the error message says
"x is
<br>
too fine" it actually means "z is too fine".
<br>
I was wondering what is the minimum number of nodes per
direction that
<br>
would avoid this problem and how the number of levels is related
to
<br>
minimum grid size required.
<br>
Thank you!
<br>
<br>
Michele
<br>
<br>
<br>
On 08/02/2013 03:11 PM, Barry Smith wrote:
<br>
<blockquote type="cite">On Aug 2, 2013, at 4:52 PM, Michele
Rosso<a class="moz-txt-link-rfc2396E" href="mailto:mrosso@uci.edu"><mrosso@uci.edu></a> wrote:
<br>
<br>
<blockquote type="cite">Barry,
<br>
<br>
thank you very much for your help. I was trying to debug the
error with no success!
<br>
Now it works like a charm for me too!
<br>
I have still two questions for you:
<br>
<br>
1) How did you choose the number of levels to use: trial and
error?
<br>
</blockquote>
I just used 2 because it is more than one level :-). When
you use a finer mesh you can use more levels.
<br>
<br>
<blockquote type="cite">2) For a singular system (periodic),
besides the nullspace removal, should I change any
parameter?
<br>
</blockquote>
I don't know of anything.
<br>
<br>
But there is a possible problem with -pc_mg_galerkin,
PETSc does not transfer the null space information from the
fine mesh to the other meshes and technically we really want
the multigrid to remove the null space on all the levels but
usually it will work without worrying about that.
<br>
<br>
Barry
<br>
<br>
<blockquote type="cite">Again, thank you very much!
<br>
<br>
Michele
<br>
<br>
On 08/02/2013 02:38 PM, Barry Smith wrote:
<br>
<blockquote type="cite"> Finally got it. My failing
memory. I had to add the line
<br>
<br>
call KSPSetDMActive(ksp,PETSC_FALSE,ierr)
<br>
<br>
immediately after KSPSetDM() and
<br>
<br>
change
<br>
<br>
call DMCreateMatrix(da,MATMPIAIJ,A,ierr)
<br>
<br>
to
<br>
<br>
call DMCreateMatrix(da,MATAIJ,A,ierr)
<br>
<br>
so it will work in both parallel and sequential then
<br>
<br>
ksp_monitor -ksp_converged_reason -pc_type mg -ksp_view
-pc_mg_galerkin -pc_mg_levels 2
<br>
<br>
works great with 2 levels.
<br>
<br>
Barry
<br>
<br>
<br>
<br>
<br>
On Aug 1, 2013, at 6:29 PM, Michele Rosso
<br>
<a class="moz-txt-link-rfc2396E" href="mailto:mrosso@uci.edu"><mrosso@uci.edu></a>
<br>
wrote:
<br>
<br>
<br>
<blockquote type="cite">Barry,
<br>
<br>
no problem. I attached the full code in
test_poisson_solver.tar.gz.
<br>
My test code is a very reduced version of my productive
code (incompressible DNS code) thus fftw3 and the
library 2decomp&fft are needed to run it.
<br>
I attached the 2decomp&fft version I used: it is a
matter of minutes to install it, so you should not have
any problem.
<br>
Please, contact me for any question/suggestion.
<br>
I the mean time I will try to debug it.
<br>
<br>
Michele
<br>
<br>
<br>
<br>
<br>
On 08/01/2013 04:19 PM, Barry Smith wrote:
<br>
<br>
<blockquote type="cite"> Run on one process until
this is debugged. You can try the option
<br>
<br>
-start_in_debugger noxterm
<br>
<br>
and then call VecView(vec,0) in the debugger when it
gives the error below. It seems like some objects are
not getting their initial values set properly. Are you
able to email the code so we can run it and figure out
what is going on?
<br>
<br>
Barry
<br>
<br>
On Aug 1, 2013, at 5:52 PM, Michele Rosso
<br>
<br>
<a class="moz-txt-link-rfc2396E" href="mailto:mrosso@uci.edu"><mrosso@uci.edu></a>
<br>
<br>
wrote:
<br>
<br>
<br>
<br>
<blockquote type="cite">Barry,
<br>
<br>
I checked the matrix: the element (0,0) is not zero,
nor any other diagonal element is.
<br>
The matrix is symmetric positive define (i.e. the
standard Poisson matrix).
<br>
Also, -da_refine is never used (see previous
output).
<br>
I tried to run with -pc_type mg -pc_mg_galerkin
-mg_levels_pc_type jacobi -mg_levels_ksp_type
chebyshev
-mg_levels_ksp_chebyshev_estimate_eigenvalues
-ksp_view -options_left
<br>
<br>
and now the error is different:
<br>
0]PETSC ERROR: [1]PETSC ERROR: ---------------------
Error Message ------------------------------------
<br>
[1]PETSC ERROR: Floating point exception!
<br>
[1]PETSC ERROR: Vec entry at local location 0 is
not-a-number or infinite at beginning of function:
Parameter number 2!
<br>
[1]PETSC ERROR:
------------------------------------------------------------------------
<br>
[1]PETSC ERROR: Petsc Release Version 3.4.2, Jul,
02, 2013
<br>
[1]PETSC ERROR: See docs/changes/index.html for
recent updates.
<br>
[1]PETSC ERROR: See docs/faq.html for hints about
trouble shooting.
<br>
[2]PETSC ERROR: --------------------- Error Message
------------------------------------
<br>
[2]PETSC ERROR: Floating point exception!
<br>
[2]PETSC ERROR: Vec entry at local location 0 is
not-a-number or infinite at beginning of function:
Parameter number 2!
<br>
[2]PETSC ERROR:
------------------------------------------------------------------------
<br>
[2]PETSC ERROR: Petsc Release Version 3.4.2, Jul,
02, 2013
<br>
[2]PETSC ERROR: See docs/changes/index.html for
recent updates.
<br>
[2]PETSC ERROR: See docs/faq.html for hints about
trouble shooting.
<br>
[2]PETSC ERROR: [3]PETSC ERROR:
--------------------- Error Message
------------------------------------
<br>
[3]PETSC ERROR: Floating point exception!
<br>
[3]PETSC ERROR: Vec entry at local location 0 is
not-a-number or infinite at beginning of function:
Parameter number 2!
<br>
[3]PETSC ERROR:
------------------------------------------------------------------------
<br>
[3]PETSC ERROR: Petsc Release Version 3.4.2, Jul,
02, 2013
<br>
[3]PETSC ERROR: See docs/changes/index.html for
recent updates.
<br>
[3]PETSC ERROR: See docs/faq.html for hints about
trouble shooting.
<br>
[3]PETSC ERROR: See docs/index.html for manual
pages.
<br>
[3]PETSC ERROR:
------------------------------------------------------------------------
<br>
[1]PETSC ERROR: See docs/index.html for manual
pages.
<br>
[1]PETSC ERROR:
------------------------------------------------------------------------
<br>
[1]PETSC ERROR: ./test on a linux-gnu-dbg named
enterprise-A by mic Thu Aug 1 15:43:16 2013
<br>
[1]PETSC ERROR: Libraries linked from
/opt/petsc/petsc-3.4.2/linux-gnu-dbg/lib
<br>
[1]PETSC ERROR: See docs/index.html for manual
pages.
<br>
[2]PETSC ERROR:
------------------------------------------------------------------------
<br>
[2]PETSC ERROR: ./test on a linux-gnu-dbg named
enterprise-A by mic Thu Aug 1 15:43:16 2013
<br>
[2]PETSC ERROR: Libraries linked from
/opt/petsc/petsc-3.4.2/linux-gnu-dbg/lib
<br>
[2]PETSC ERROR: Configure run at Thu Aug 1 12:01:44
2013
<br>
[2]PETSC ERROR: [3]PETSC ERROR: ./test on a
linux-gnu-dbg named enterprise-A by mic Thu Aug 1
15:43:16 2013
<br>
[3]PETSC ERROR: Libraries linked from
/opt/petsc/petsc-3.4.2/linux-gnu-dbg/lib
<br>
[3]PETSC ERROR: Configure run at Thu Aug 1 12:01:44
2013
<br>
[3]PETSC ERROR: Configure options
<br>
Configure run at Thu Aug 1 12:01:44 2013
<br>
[1]PETSC ERROR: Configure options
<br>
[1]PETSC ERROR:
------------------------------------------------------------------------
<br>
[1]PETSC ERROR: VecValidValues() line 28 in
src/vec/vec/interface/rvector.c
<br>
Configure options
<br>
[2]PETSC ERROR:
------------------------------------------------------------------------
<br>
[2]PETSC ERROR: VecValidValues() line 28 in
src/vec/vec/interface/rvector.c
<br>
[3]PETSC ERROR:
------------------------------------------------------------------------
<br>
[3]PETSC ERROR: VecValidValues() line 28 in
src/vec/vec/interface/rvector.c
<br>
[3]PETSC ERROR: [1]PETSC ERROR: MatMult() line 2174
in src/mat/interface/matrix.c
<br>
[1]PETSC ERROR: [2]PETSC ERROR: MatMult() line 2174
in src/mat/interface/matrix.c
<br>
[2]PETSC ERROR: KSP_MatMult() line 204 in
src/ksp/ksp/impls/cheby//opt/petsc/petsc-3.4.2/include/petsc-private/kspimpl.h<br>
MatMult() line 2174 in src/mat/interface/matrix.c
<br>
[3]PETSC ERROR: KSP_MatMult() line 204 in
src/ksp/ksp/impls/cheby//opt/petsc/petsc-3.4.2/include/petsc-private/kspimpl.h<br>
[3]PETSC ERROR: KSP_MatMult() line 204 in
src/ksp/ksp/impls/cheby//opt/petsc/petsc-3.4.2/include/petsc-private/kspimpl.h<br>
[1]PETSC ERROR: KSPSolve_Chebyshev() line 504 in
src/ksp/ksp/impls/cheby/cheby.c
<br>
[2]PETSC ERROR: KSPSolve_Chebyshev() line 504 in
src/ksp/ksp/impls/cheby/cheby.c
<br>
[2]PETSC ERROR: KSPSolve_Chebyshev() line 504 in
src/ksp/ksp/impls/cheby/cheby.c
<br>
[3]PETSC ERROR: KSPSolve() line 441 in
src/ksp/ksp/interface/itfunc.c
<br>
[1]PETSC ERROR: KSPSolve() line 441 in
src/ksp/ksp/interface/itfunc.c
<br>
[1]PETSC ERROR: KSPSolve() line 441 in
src/ksp/ksp/interface/itfunc.c
<br>
[2]PETSC ERROR: PCMGMCycle_Private() line 19 in
src/ksp/pc/impls/mg/mg.c
<br>
[3]PETSC ERROR: PCMGMCycle_Private() line 19 in
src/ksp/pc/impls/mg/mg.c
<br>
PCMGMCycle_Private() line 19 in
src/ksp/pc/impls/mg/mg.c
<br>
[1]PETSC ERROR: PCApply_MG() line 330 in
src/ksp/pc/impls/mg/mg.c
<br>
[2]PETSC ERROR: PCApply_MG() line 330 in
src/ksp/pc/impls/mg/mg.c
<br>
[2]PETSC ERROR: PCApply() line 442 in
src/ksp/pc/interface/precon.c
<br>
[3]PETSC ERROR: PCApply_MG() line 330 in
src/ksp/pc/impls/mg/mg.c
<br>
[3]PETSC ERROR: PCApply() line 442 in
src/ksp/pc/interface/precon.c
<br>
[1]PETSC ERROR: PCApply() line 442 in
src/ksp/pc/interface/precon.c
<br>
[1]PETSC ERROR: KSP_PCApply() line 227 in
src/ksp/ksp/impls/cg//opt/petsc/petsc-3.4.2/include/petsc-private/kspimpl.h<br>
[2]PETSC ERROR: KSP_PCApply() line 227 in
src/ksp/ksp/impls/cg//opt/petsc/petsc-3.4.2/include/petsc-private/kspimpl.h<br>
[2]PETSC ERROR: [3]PETSC ERROR: KSP_PCApply() line
227 in
src/ksp/ksp/impls/cg//opt/petsc/petsc-3.4.2/include/petsc-private/kspimpl.h<br>
[3]PETSC ERROR: KSPSolve_CG() line 175 in
src/ksp/ksp/impls/cg/cg.c
<br>
[1]PETSC ERROR: KSPSolve_CG() line 175 in
src/ksp/ksp/impls/cg/cg.c
<br>
[1]PETSC ERROR: KSPSolve() line 441 in
src/ksp/ksp/interface/itfunc.c
<br>
KSPSolve_CG() line 175 in src/ksp/ksp/impls/cg/cg.c
<br>
[2]PETSC ERROR: KSPSolve() line 441 in
src/ksp/ksp/interface/itfunc.c
<br>
[3]PETSC ERROR: KSPSolve() line 441 in
src/ksp/ksp/interface/itfunc.c
<br>
--------------------- Error Message
------------------------------------
<br>
[0]PETSC ERROR: Floating point exception!
<br>
[0]PETSC ERROR: Vec entry at local location 0 is
not-a-number or infinite at beginning of function:
Parameter number 2!
<br>
[0]PETSC ERROR:
------------------------------------------------------------------------
<br>
[0]PETSC ERROR: Petsc Release Version 3.4.2, Jul,
02, 2013
<br>
[0]PETSC ERROR: See docs/changes/index.html for
recent updates.
<br>
[0]PETSC ERROR: See docs/faq.html for hints about
trouble shooting.
<br>
[0]PETSC ERROR: See docs/index.html for manual
pages.
<br>
[0]PETSC ERROR:
------------------------------------------------------------------------
<br>
[0]PETSC ERROR: ./test on a linux-gnu-dbg named
enterprise-A by mic Thu Aug 1 15:43:16 2013
<br>
[0]PETSC ERROR: Libraries linked from
/opt/petsc/petsc-3.4.2/linux-gnu-dbg/lib
<br>
[0]PETSC ERROR: Configure run at Thu Aug 1 12:01:44
2013
<br>
[0]PETSC ERROR: Configure options
<br>
[0]PETSC ERROR:
------------------------------------------------------------------------
<br>
[0]PETSC ERROR: VecValidValues() line 28 in
src/vec/vec/interface/rvector.c
<br>
[0]PETSC ERROR: MatMult() line 2174 in
src/mat/interface/matrix.c
<br>
[0]PETSC ERROR: KSP_MatMult() line 204 in
src/ksp/ksp/impls/cheby//opt/petsc/petsc-3.4.2/include/petsc-private/kspimpl.h<br>
[0]PETSC ERROR: KSPSolve_Chebyshev() line 504 in
src/ksp/ksp/impls/cheby/cheby.c
<br>
[0]PETSC ERROR: KSPSolve() line 441 in
src/ksp/ksp/interface/itfunc.c
<br>
[0]PETSC ERROR: PCMGMCycle_Private() line 19 in
src/ksp/pc/impls/mg/mg.c
<br>
[0]PETSC ERROR: PCApply_MG() line 330 in
src/ksp/pc/impls/mg/mg.c
<br>
[0]PETSC ERROR: PCApply() line 442 in
src/ksp/pc/interface/precon.c
<br>
[0]PETSC ERROR: KSP_PCApply() line 227 in
src/ksp/ksp/impls/cg//opt/petsc/petsc-3.4.2/include/petsc-private/kspimpl.h<br>
[0]PETSC ERROR: KSPSolve_CG() line 175 in
src/ksp/ksp/impls/cg/cg.c
<br>
[0]PETSC ERROR: KSPSolve() line 441 in
src/ksp/ksp/interface/itfunc.c
<br>
<br>
#PETSc Option Table entries:
<br>
-ksp_view
<br>
-mg_levels_ksp_chebyshev_estimate_eigenvalues
<br>
-mg_levels_ksp_type chebyshev
<br>
-mg_levels_pc_type jacobi
<br>
-options_left
<br>
-pc_mg_galerkin
<br>
-pc_type mg
<br>
#End of PETSc Option Table entries
<br>
There are no unused options.
<br>
<br>
Michele
<br>
<br>
<br>
On 08/01/2013 03:27 PM, Barry Smith wrote:
<br>
<br>
<br>
<blockquote type="cite"> Do a MatView() on A
before the solve (remove the -da_refine 4) so it
is small. Is the 0,0 entry 0? If the matrix has
zero on the diagonals you cannot us Gauss-Seidel
as the smoother. You can start with
-mg_levels_pc_type jacobi -mg_levels_ksp_type
chebychev
-mg_levels_ksp_chebyshev_estimate_eigenvalues
<br>
<br>
Is the matrix a Stokes-like matrix? If so then
different preconditioners are in order.
<br>
<br>
Barry
<br>
<br>
On Aug 1, 2013, at 5:21 PM, Michele Rosso
<br>
<br>
<br>
<a class="moz-txt-link-rfc2396E" href="mailto:mrosso@uci.edu"><mrosso@uci.edu></a>
<br>
<br>
<br>
wrote:
<br>
<br>
<br>
<br>
<br>
<blockquote type="cite">Barry,
<br>
<br>
here it is the fraction of code where I set the
rhs term and the matrix.
<br>
<br>
! Create matrix
<br>
call form_matrix( A , qrho, lsf, head )
<br>
call
MatAssemblyBegin(A,MAT_FINAL_ASSEMBLY,ierr)
<br>
call
MatAssemblyEnd(A,MAT_FINAL_ASSEMBLY,ierr)
<br>
<br>
! Create rhs term
<br>
call form_rhs(work, qrho, lsf, b , head)
<br>
<br>
! Solve system
<br>
call KSPSetFromOptions(ksp,ierr)
<br>
call KSPSetUp(ksp,ierr)
<br>
call KSPSolve(ksp,b,x,ierr)
<br>
call KSPGetIterationNumber(ksp, iiter
,ierr)
<br>
<br>
The subroutine form_matrix returns the Mat
object A that is filled by using
MatSetValuesStencil.
<br>
qrho, lsf and head are additional arguments that
are needed to compute the matrix value.
<br>
<br>
<br>
Michele
<br>
<br>
<br>
<br>
On 08/01/2013 03:11 PM, Barry Smith wrote:
<br>
<br>
<br>
<br>
<blockquote type="cite"> Where are you putting
the values into the matrix? It seems the
matrix has no values in it? The code is
stopping because in the Gauss-Seidel smoothing
it has detected zero diagonals.
<br>
<br>
Barry
<br>
<br>
<br>
On Aug 1, 2013, at 4:47 PM, Michele Rosso
<br>
<br>
<br>
<a class="moz-txt-link-rfc2396E" href="mailto:mrosso@uci.edu"><mrosso@uci.edu></a>
<br>
<br>
<br>
wrote:
<br>
<br>
<br>
<br>
<br>
<blockquote type="cite">Barry,
<br>
<br>
I run with : -pc_type mg -pc_mg_galerkin
-da_refine 4 -ksp_view -options_left
<br>
<br>
For the test I use a 64^3 grid and 4
processors.
<br>
<br>
The output is:
<br>
<br>
[2]PETSC ERROR: --------------------- Error
Message ------------------------------------
<br>
[2]PETSC ERROR: Arguments are incompatible!
<br>
[2]PETSC ERROR: Zero diagonal on row 0!
<br>
[2]PETSC ERROR:
------------------------------------------------------------------------
<br>
[2]PETSC ERROR: Petsc Release Version 3.4.2,
Jul, 02, 2013
<br>
[2]PETSC ERROR: See docs/changes/index.html
for recent updates.
<br>
[2]PETSC ERROR: See docs/faq.html for hints
about trouble shooting.
<br>
[2]PETSC ERROR: See docs/index.html for
manual pages.
<br>
[2]PETSC ERROR:
------------------------------------------------------------------------
<br>
[2]PETSC ERROR: ./test on a linux-gnu-dbg
named enterprise-A by mic Thu Aug 1
14:44:04 2013
<br>
[0]PETSC ERROR: [2]PETSC ERROR: Libraries
linked from
/opt/petsc/petsc-3.4.2/linux-gnu-dbg/lib
<br>
[2]PETSC ERROR: Configure run at Thu Aug 1
12:01:44 2013
<br>
[2]PETSC ERROR: Configure options
<br>
[2]PETSC ERROR:
------------------------------------------------------------------------
<br>
[2]PETSC ERROR: MatInvertDiagonal_SeqAIJ()
line 1457 in src/mat/impls/aij/seq/aij.c
<br>
[2]PETSC ERROR: MatSOR_SeqAIJ() line 1489 in
src/mat/impls/aij/seq/aij.c
<br>
--------------------- Error Message
------------------------------------
<br>
[2]PETSC ERROR: MatSOR_MPIAIJ() line 1623 in
src/mat/impls/aij/mpi/mpiaij.c
<br>
[2]PETSC ERROR: MatSOR() line 3649 in
src/mat/interface/matrix.c
<br>
[2]PETSC ERROR: [0]PETSC ERROR:
PCApply_SOR() line 35 in
src/ksp/pc/impls/sor/sor.c
<br>
[2]PETSC ERROR: PCApply() line 442 in
src/ksp/pc/interface/precon.c
<br>
[2]PETSC ERROR: KSP_PCApply() line 227 in
src/ksp/ksp/interface//opt/petsc/petsc-3.4.2/include/petsc-private/kspimpl.h<br>
Arguments are incompatible!
<br>
[2]PETSC ERROR: KSPInitialResidual() line 64
in src/ksp/ksp/interface/itres.c
<br>
[2]PETSC ERROR: KSPSolve_GMRES() line 239 in
src/ksp/ksp/impls/gmres/gmres.c
<br>
[2]PETSC ERROR: KSPSolve() line 441 in
src/ksp/ksp/interface/itfunc.c
<br>
[2]PETSC ERROR: [0]PETSC ERROR:
KSPSolve_Chebyshev() line 409 in
src/ksp/ksp/impls/cheby/cheby.c
<br>
[2]PETSC ERROR: KSPSolve() line 441 in
src/ksp/ksp/interface/itfunc.c
<br>
[2]PETSC ERROR: PCMGMCycle_Private() line 19
in src/ksp/pc/impls/mg/mg.c
<br>
Zero diagonal on row 0!
<br>
[2]PETSC ERROR: PCApply_MG() line 330 in
src/ksp/pc/impls/mg/mg.c
<br>
[2]PETSC ERROR: PCApply() line 442 in
src/ksp/pc/interface/precon.c
<br>
[0]PETSC ERROR: [2]PETSC ERROR:
KSP_PCApply() line 227 in
src/ksp/ksp/impls/cg//opt/petsc/petsc-3.4.2/include/petsc-private/kspimpl.h<br>
[2]PETSC ERROR: KSPSolve_CG() line 175 in
src/ksp/ksp/impls/cg/cg.c
<br>
------------------------------------------------------------------------
<br>
[2]PETSC ERROR: KSPSolve() line 441 in
src/ksp/ksp/interface/itfunc.c
<br>
[0]PETSC ERROR: Petsc Release Version 3.4.2,
Jul, 02, 2013
<br>
[0]PETSC ERROR: See docs/changes/index.html
for recent updates.
<br>
[3]PETSC ERROR: [0]PETSC ERROR: See
docs/faq.html for hints about trouble
shooting.
<br>
[0]PETSC ERROR: --------------------- Error
Message ------------------------------------
<br>
[3]PETSC ERROR: Arguments are incompatible!
<br>
[3]PETSC ERROR: Zero diagonal on row 0!
<br>
[3]PETSC ERROR:
------------------------------------------------------------------------
<br>
[3]PETSC ERROR: Petsc Release Version 3.4.2,
Jul, 02, 2013
<br>
[3]PETSC ERROR: See docs/changes/index.html
for recent updates.
<br>
[3]PETSC ERROR: See docs/faq.html for hints
about trouble shooting.
<br>
[3]PETSC ERROR: See docs/index.html for
manual pages.
<br>
[3]PETSC ERROR:
------------------------------------------------------------------------
<br>
See docs/index.html for manual pages.
<br>
[3]PETSC ERROR: ./test on a linux-gnu-dbg
named enterprise-A by mic Thu Aug 1
14:44:04 2013
<br>
[3]PETSC ERROR: Libraries linked from
/opt/petsc/petsc-3.4.2/linux-gnu-dbg/lib
<br>
[1]PETSC ERROR: [3]PETSC ERROR: Configure
run at Thu Aug 1 12:01:44 2013
<br>
[3]PETSC ERROR: Configure options
<br>
[3]PETSC ERROR:
------------------------------------------------------------------------
<br>
[3]PETSC ERROR: --------------------- Error
Message ------------------------------------
<br>
MatInvertDiagonal_SeqAIJ() line 1457 in
src/mat/impls/aij/seq/aij.c
<br>
[3]PETSC ERROR: MatSOR_SeqAIJ() line 1489 in
src/mat/impls/aij/seq/aij.c
<br>
[3]PETSC ERROR: [0]PETSC ERROR:
MatSOR_MPIAIJ() line 1623 in
src/mat/impls/aij/mpi/mpiaij.c
<br>
[1]PETSC ERROR: Arguments are incompatible!
<br>
[1]PETSC ERROR: Zero diagonal on row 0!
<br>
[1]PETSC ERROR:
------------------------------------------------------------------------
<br>
[1]PETSC ERROR: Petsc Release Version 3.4.2,
Jul, 02, 2013
<br>
[1]PETSC ERROR: See docs/changes/index.html
for recent updates.
<br>
[1]PETSC ERROR: See docs/faq.html for hints
about trouble shooting.
<br>
[1]PETSC ERROR: See docs/index.html for
manual pages.
<br>
[1]PETSC ERROR:
------------------------------------------------------------------------
<br>
[1]PETSC ERROR: ./test on a linux-gnu-dbg
named enterprise-A by mic Thu Aug 1
14:44:04 2013
<br>
[1]PETSC ERROR: Libraries linked from
/opt/petsc/petsc-3.4.2/linux-gnu-dbg/lib
<br>
[1]PETSC ERROR: Configure run at Thu Aug 1
12:01:44 2013
<br>
[1]PETSC ERROR: Configure options
<br>
[1]PETSC ERROR:
------------------------------------------------------------------------
<br>
[1]PETSC ERROR: MatInvertDiagonal_SeqAIJ()
line 1457 in src/mat/impls/aij/seq/aij.c
<br>
[1]PETSC ERROR: [3]PETSC ERROR: MatSOR()
line 3649 in src/mat/interface/matrix.c
<br>
[3]PETSC ERROR: PCApply_SOR() line 35 in
src/ksp/pc/impls/sor/sor.c
<br>
[3]PETSC ERROR: PCApply() line 442 in
src/ksp/pc/interface/precon.c
<br>
[3]PETSC ERROR: KSP_PCApply() line 227 in
src/ksp/ksp/interface//opt/petsc/petsc-3.4.2/include/petsc-private/kspimpl.h<br>
[3]PETSC ERROR: KSPInitialResidual() line 64
in src/ksp/ksp/interface/itres.c
<br>
[3]PETSC ERROR: KSPSolve_GMRES() line 239 in
src/ksp/ksp/impls/gmres/gmres.c
<br>
[3]PETSC ERROR: KSPSolve() line 441 in
src/ksp/ksp/interface/itfunc.c
<br>
[3]PETSC ERROR: KSPSolve_Chebyshev() line
409 in src/ksp/ksp/impls/cheby/cheby.c
<br>
[3]PETSC ERROR: KSPSolve() line 441 in
src/ksp/ksp/interface/itfunc.c
<br>
[3]PETSC ERROR: PCMGMCycle_Private() line 19
in src/ksp/pc/impls/mg/mg.c
<br>
[3]PETSC ERROR: PCApply_MG() line 330 in
src/ksp/pc/impls/mg/mg.c
<br>
[3]PETSC ERROR: PCApply() line 442 in
src/ksp/pc/interface/precon.c
<br>
[3]PETSC ERROR: KSP_PCApply() line 227 in
src/ksp/ksp/impls/cg//opt/petsc/petsc-3.4.2/include/petsc-private/kspimpl.h<br>
[3]PETSC ERROR: KSPSolve_CG() line 175 in
src/ksp/ksp/impls/cg/cg.c
<br>
[3]PETSC ERROR: KSPSolve() line 441 in
src/ksp/ksp/interface/itfunc.c
<br>
MatSOR_SeqAIJ() line 1489 in
src/mat/impls/aij/seq/aij.c
<br>
[1]PETSC ERROR: MatSOR_MPIAIJ() line 1623 in
src/mat/impls/aij/mpi/mpiaij.c
<br>
[1]PETSC ERROR: MatSOR() line 3649 in
src/mat/interface/matrix.c
<br>
[1]PETSC ERROR: PCApply_SOR() line 35 in
src/ksp/pc/impls/sor/sor.c
<br>
[1]PETSC ERROR: PCApply() line 442 in
src/ksp/pc/interface/precon.c
<br>
[1]PETSC ERROR: KSP_PCApply() line 227 in
src/ksp/ksp/interface//opt/petsc/petsc-3.4.2/include/petsc-private/kspimpl.h<br>
[1]PETSC ERROR: KSPInitialResidual() line 64
in src/ksp/ksp/interface/itres.c
<br>
[1]PETSC ERROR: KSPSolve_GMRES() line 239 in
src/ksp/ksp/impls/gmres/gmres.c
<br>
[1]PETSC ERROR: KSPSolve() line 441 in
src/ksp/ksp/interface/itfunc.c
<br>
[1]PETSC ERROR: KSPSolve_Chebyshev() line
409 in src/ksp/ksp/impls/cheby/cheby.c
<br>
[1]PETSC ERROR: KSPSolve() line 441 in
src/ksp/ksp/interface/itfunc.c
<br>
[1]PETSC ERROR: PCMGMCycle_Private() line 19
in src/ksp/pc/impls/mg/mg.c
<br>
[1]PETSC ERROR: PCApply_MG() line 330 in
src/ksp/pc/impls/mg/mg.c
<br>
[1]PETSC ERROR: PCApply() line 442 in
src/ksp/pc/interface/precon.c
<br>
[1]PETSC ERROR: KSP_PCApply() line 227 in
src/ksp/ksp/impls/cg//opt/petsc/petsc-3.4.2/include/petsc-private/kspimpl.h<br>
[1]PETSC ERROR: KSPSolve_CG() line 175 in
src/ksp/ksp/impls/cg/cg.c
<br>
[1]PETSC ERROR: KSPSolve() line 441 in
src/ksp/ksp/interface/itfunc.c
<br>
------------------------------------------------------------------------
<br>
[0]PETSC ERROR: ./test on a linux-gnu-dbg
named enterprise-A by mic Thu Aug 1
14:44:04 2013
<br>
[0]PETSC ERROR: Libraries linked from
/opt/petsc/petsc-3.4.2/linux-gnu-dbg/lib
<br>
[0]PETSC ERROR: Configure run at Thu Aug 1
12:01:44 2013
<br>
[0]PETSC ERROR: Configure options
<br>
[0]PETSC ERROR:
------------------------------------------------------------------------
<br>
[0]PETSC ERROR: MatInvertDiagonal_SeqAIJ()
line 1457 in src/mat/impls/aij/seq/aij.c
<br>
[0]PETSC ERROR: MatSOR_SeqAIJ() line 1489 in
src/mat/impls/aij/seq/aij.c
<br>
[0]PETSC ERROR: MatSOR_MPIAIJ() line 1623 in
src/mat/impls/aij/mpi/mpiaij.c
<br>
[0]PETSC ERROR: MatSOR() line 3649 in
src/mat/interface/matrix.c
<br>
[0]PETSC ERROR: PCApply_SOR() line 35 in
src/ksp/pc/impls/sor/sor.c
<br>
[0]PETSC ERROR: PCApply() line 442 in
src/ksp/pc/interface/precon.c
<br>
[0]PETSC ERROR: KSP_PCApply() line 227 in
src/ksp/ksp/interface//opt/petsc/petsc-3.4.2/include/petsc-private/kspimpl.h<br>
[0]PETSC ERROR: KSPInitialResidual() line 64
in src/ksp/ksp/interface/itres.c
<br>
[0]PETSC ERROR: KSPSolve_GMRES() line 239 in
src/ksp/ksp/impls/gmres/gmres.c
<br>
[0]PETSC ERROR: KSPSolve() line 441 in
src/ksp/ksp/interface/itfunc.c
<br>
[0]PETSC ERROR: KSPSolve_Chebyshev() line
409 in src/ksp/ksp/impls/cheby/cheby.c
<br>
[0]PETSC ERROR: KSPSolve() line 441 in
src/ksp/ksp/interface/itfunc.c
<br>
[0]PETSC ERROR: PCMGMCycle_Private() line 19
in src/ksp/pc/impls/mg/mg.c
<br>
[0]PETSC ERROR: PCApply_MG() line 330 in
src/ksp/pc/impls/mg/mg.c
<br>
[0]PETSC ERROR: PCApply() line 442 in
src/ksp/pc/interface/precon.c
<br>
[0]PETSC ERROR: KSP_PCApply() line 227 in
src/ksp/ksp/impls/cg//opt/petsc/petsc-3.4.2/include/petsc-private/kspimpl.h<br>
[0]PETSC ERROR: KSPSolve_CG() line 175 in
src/ksp/ksp/impls/cg/cg.c
<br>
[0]PETSC ERROR: KSPSolve() line 441 in
src/ksp/ksp/interface/itfunc.c
<br>
#PETSc Option Table entries:
<br>
-da_refine 4
<br>
-ksp_view
<br>
-options_left
<br>
-pc_mg_galerkin
<br>
-pc_type mg
<br>
#End of PETSc Option Table entries
<br>
There is one unused database option. It is:
<br>
Option left: name:-da_refine value: 4
<br>
<br>
<br>
Here is the code I use to setup DMDA and
KSP:
<br>
<br>
call DMDACreate3d( PETSC_COMM_WORLD
, &
<br>
&
DMDA_BOUNDARY_PERIODIC ,
DMDA_BOUNDARY_PERIODIC, &
<br>
&
DMDA_BOUNDARY_PERIODIC ,
DMDA_STENCIL_STAR, &
<br>
& N_Z , N_Y , N_X ,
N_B3 , N_B2 , 1_ip, 1_ip , 1_ip , &
<br>
& int(NNZ,ip)
,int(NNY,ip) , NNX, da , ierr)
<br>
! Create Global Vectors
<br>
call DMCreateGlobalVector(da,b,ierr)
<br>
call VecDuplicate(b,x,ierr)
<br>
! Set initial guess for first use
of the module to 0
<br>
call VecSet(x,0.0_rp,ierr)
<br>
! Create matrix
<br>
call
DMCreateMatrix(da,MATMPIAIJ,A,ierr)
<br>
! Create solver
<br>
call
KSPCreate(PETSC_COMM_WORLD,ksp,ierr)
<br>
call KSPSetDM(ksp,da,ierr)
<br>
call
KSPSetOperators(ksp,A,A,DIFFERENT_NONZERO_PATTERN,ierr)
<br>
! call
KSPSetOperators(ksp,A,A,SAME_NONZERO_PATTERN,ierr)
<br>
call KSPSetType(ksp,KSPCG,ierr)
<br>
call
KSPSetNormType(ksp,KSP_NORM_UNPRECONDITIONED,ierr)
! Real residual
<br>
call
KSPSetInitialGuessNonzero(ksp,PETSC_TRUE,ierr)
<br>
call KSPSetTolerances(ksp, tol
,PETSC_DEFAULT_DOUBLE_PRECISION,&
<br>
&
PETSC_DEFAULT_DOUBLE_PRECISION,PETSC_DEFAULT_INTEGER,ierr)
<br>
<br>
! To allow using option from command
line
<br>
call KSPSetFromOptions(ksp,ierr)
<br>
<br>
<br>
Michele
<br>
<br>
<br>
<br>
<br>
On 08/01/2013 01:04 PM, Barry Smith wrote:
<br>
<br>
<br>
<br>
<blockquote type="cite"> You can use the
option -pc_mg_galerkin and then MG will
compute the coarser matrices with a sparse
matrix matrix matrix product so you should
not need to change your code to try it
out. You still need to use the KSPSetDM()
and -da_refine n to get it working
<br>
<br>
If it doesn't work, send us all the
output.
<br>
<br>
Barry
<br>
<br>
<br>
On Aug 1, 2013, at 2:47 PM, Michele Rosso
<br>
<br>
<br>
<br>
<a class="moz-txt-link-rfc2396E" href="mailto:mrosso@uci.edu"><mrosso@uci.edu></a>
<br>
<br>
<br>
<br>
wrote:
<br>
<br>
<br>
<br>
<br>
<br>
<blockquote type="cite">Barry,
<br>
you are correct, I did not use it. I
think I get now where is the problem.
Correct me if I am wrong, but for the
<br>
geometric multigrid to work, ksp must be
provided with subroutines to compute the
matrix and the rhs at any level through
<br>
KSPSetComputeOperators and
KSPSetComputeRHS.
<br>
I do not do that, I simply build a rhs
vector and a matrix and then I solve the
system.
<br>
If you confirm what I just wrote, I will
try to modify my code accordingly and
get back to you.
<br>
Thank you,
<br>
Michele
<br>
On 08/01/2013 11:48 AM, Barry Smith
wrote:
<br>
<br>
<br>
<br>
<br>
<blockquote type="cite"> Do you use
KSPSetDM(ksp,da); ? See
src/ksp/ksp/examples/tutorials/ex19.c
<br>
<br>
Barry
<br>
<br>
On Aug 1, 2013, at 1:35 PM, Michele
Rosso
<br>
<br>
<br>
<br>
<br>
<a class="moz-txt-link-rfc2396E" href="mailto:mrosso@uci.edu"><mrosso@uci.edu></a>
<br>
<br>
<br>
<br>
<br>
wrote:
<br>
<br>
<br>
<br>
<br>
<br>
<br>
<blockquote type="cite">Barry,
<br>
<br>
I am using a finite difference
Cartesian uniform grid and DMDA and
so far it has not given me any
problem.
<br>
I am using a ksp solver (not snes).
In a previous thread, I was told an
odd number of grid points was needed
for the geometric multigrid, is that
correct?
<br>
I tried to run my case with
<br>
<br>
<br>
-pc_type mg -da_refine 4
<br>
<br>
<br>
<br>
but it does not seem to use the
-da_refine option:
<br>
<br>
mpiexec -np 4 ./test -pc_type mg
-da_refine 4 -ksp_view
-options_left
<br>
<br>
<br>
KSP Object: 4 MPI processes
<br>
type: cg
<br>
maximum iterations=10000
<br>
tolerances: relative=1e-08,
absolute=1e-50, divergence=10000
<br>
left preconditioning
<br>
using nonzero initial guess
<br>
using UNPRECONDITIONED norm type
for convergence test
<br>
PC Object: 4 MPI processes
<br>
type: mg
<br>
MG: type is MULTIPLICATIVE,
levels=1 cycles=v
<br>
Cycles per PCApply=1
<br>
Not using Galerkin computed
coarse grid matrices
<br>
Coarse grid solver -- level
-------------------------------
<br>
KSP Object:
(mg_levels_0_) 4 MPI processes
<br>
type: chebyshev
<br>
Chebyshev: eigenvalue
estimates: min = 0.134543, max =
1.47998
<br>
Chebyshev: estimated using:
[0 0.1; 0 1.1]
<br>
KSP Object:
(mg_levels_0_est_) 4 MPI
processes
<br>
type: gmres
<br>
GMRES: restart=30, using
Classical (unmodified) Gram-Schmidt
Orthogonalization with no iterative
refinement
<br>
GMRES: happy breakdown
tolerance 1e-30
<br>
maximum iterations=10,
initial guess is zero
<br>
tolerances:
relative=1e-05, absolute=1e-50,
divergence=10000
<br>
left preconditioning
<br>
using NONE norm type for
convergence test
<br>
PC Object:
(mg_levels_0_) 4 MPI
processes
<br>
type: sor
<br>
SOR: type =
local_symmetric, iterations = 1,
local iterations = 1, omega = 1
<br>
linear system matrix =
precond matrix:
<br>
Matrix Object: 4
MPI processes
<br>
type: mpiaij
<br>
rows=262144, cols=262144
<br>
total: nonzeros=1835008,
allocated nonzeros=1835008
<br>
total number of mallocs
used during MatSetValues calls =0
<br>
maximum iterations=1, initial
guess is zero
<br>
tolerances: relative=1e-05,
absolute=1e-50, divergence=10000
<br>
left preconditioning
<br>
using NONE norm type for
convergence test
<br>
PC Object: (mg_levels_0_)
4 MPI processes
<br>
type: sor
<br>
SOR: type = local_symmetric,
iterations = 1, local iterations =
1, omega = 1
<br>
linear system matrix = precond
matrix:
<br>
Matrix Object: 4 MPI
processes
<br>
type: mpiaij
<br>
rows=262144, cols=262144
<br>
total: nonzeros=1835008,
allocated nonzeros=1835008
<br>
total number of mallocs used
during MatSetValues calls =0
<br>
linear system matrix = precond
matrix:
<br>
Matrix Object: 4 MPI processes
<br>
type: mpiaij
<br>
rows=262144, cols=262144
<br>
total: nonzeros=1835008,
allocated nonzeros=1835008
<br>
total number of mallocs used
during MatSetValues calls =0
<br>
Solution = 1.53600013
sec
<br>
#PETSc Option Table entries:
<br>
-da_refine 4
<br>
-ksp_view
<br>
-options_left
<br>
-pc_type mg
<br>
#End of PETSc Option Table entries
<br>
There is one unused database option.
It is:
<br>
Option left: name:-da_refine value:
4
<br>
<br>
Michele
<br>
<br>
On 08/01/2013 11:21 AM, Barry Smith
wrote:
<br>
<br>
<br>
<br>
<br>
<br>
<blockquote type="cite"> What
kind of mesh are you using? Are
you using DMDA? If you are using
DMDA (and have written your code
to use it "correctly") then it
should be trivial to run with
geometric multigrid and geometric
multigrid should be a bit faster.
<br>
<br>
For example on
src/snes/examples/tutorials/ex19.c
I run with ./ex19 -pc_type mg
-da_refine 4 and it refines the
original DMDA 4 times and uses
geometric multigrid with 5 levels.
<br>
<br>
<br>
Barry
<br>
<br>
<br>
On Aug 1, 2013, at 1:14 PM,
Michele Rosso
<br>
<br>
<br>
<br>
<br>
<a class="moz-txt-link-rfc2396E" href="mailto:mrosso@uci.edu"><mrosso@uci.edu></a>
<br>
<br>
<br>
<br>
<br>
wrote:
<br>
<br>
<br>
<br>
<br>
<br>
<br>
<blockquote type="cite">Hi,
<br>
<br>
I am successfully using PETSc
(v3.4.2) to solve a 3D
Poisson's equation with CG +
GAMG as I was suggested to do in
a previous thread.
<br>
So far I am using GAMG with the
default settings, i.e.
<br>
<br>
-pc_type gamg
-pc_gamg_agg_nsmooths 1
<br>
<br>
The speed of the solution is
satisfactory, but I would like
to know if you have any
suggestions to further speed it
up, particularly
<br>
if there is any parameters worth
looking into to achieve an even
faster solution, for example
number of levels and so on.
<br>
So far I am using Dirichlet's
BCs for my test case, but I will
soon have periodic conditions:
in this case, does GAMG require
particular settings?
<br>
Finally, I did not try geometric
multigrid: do you think it is
worth a shot?
<br>
<br>
Here are my current settings:
<br>
<br>
I run with
<br>
<br>
-pc_type gamg
-pc_gamg_agg_nsmooths 1
-ksp_view -options_left
<br>
<br>
and the output is:
<br>
<br>
KSP Object: 4 MPI processes
<br>
type: cg
<br>
maximum iterations=10000
<br>
tolerances: relative=1e-08,
absolute=1e-50, divergence=10000
<br>
left preconditioning
<br>
using nonzero initial guess
<br>
using UNPRECONDITIONED norm
type for convergence test
<br>
PC Object: 4 MPI processes
<br>
type: gamg
<br>
MG: type is MULTIPLICATIVE,
levels=3 cycles=v
<br>
Cycles per PCApply=1
<br>
Using Galerkin computed
coarse grid matrices
<br>
Coarse grid solver -- level
-------------------------------
<br>
KSP Object:
(mg_coarse_) 4 MPI processes
<br>
type: preonly
<br>
maximum iterations=1,
initial guess is zero
<br>
tolerances:
relative=1e-05, absolute=1e-50,
divergence=10000
<br>
left preconditioning
<br>
using NONE norm type for
convergence test
<br>
PC Object:
(mg_coarse_) 4 MPI processes
<br>
type: bjacobi
<br>
block Jacobi: number of
blocks = 4
<br>
Local solve info for
each block is in the following
KSP and PC objects:
<br>
[0] number of local
blocks = 1, first local block
number = 0
<br>
[0] local block
number 0
<br>
KSP Object:
(mg_coarse_sub_) 1 MPI
processes
<br>
type: preonly
<br>
maximum iterations=1,
initial guess is zero
<br>
tolerances:
relative=1e-05, absolute=1e-50,
divergence=10000
<br>
KSP Object:
(mg_coarse_sub_) left
preconditioning
<br>
using NONE norm type
for convergence test
<br>
PC Object:
(mg_coarse_sub_) 1 MPI
processes
<br>
type: preonly
<br>
1 MPI processes
<br>
type: lu
<br>
maximum iterations=1,
initial guess is zero
<br>
tolerances:
relative=1e-05, absolute=1e-50,
divergence=10000
<br>
LU: out-of-place
factorization
<br>
left
preconditioning
<br>
using NONE norm type
for convergence test
<br>
PC Object:
(mg_coarse_sub_) 1 MPI
processes
<br>
type: lu
<br>
tolerance for zero
pivot 2.22045e-14
<br>
using diagonal
shift on blocks to prevent zero
pivot
<br>
matrix ordering: nd
<br>
LU: out-of-place
factorization
<br>
tolerance for zero
pivot 2.22045e-14
<br>
using diagonal
shift on blocks to prevent zero
pivot
<br>
matrix ordering: nd
<br>
factor fill ratio
given 5, needed 0
<br>
Factored matrix
follows:
<br>
factor fill ratio
given 5, needed 4.13207
<br>
Factored matrix
follows:
<br>
Matrix
Object: Matrix
Object: 1 MPI
processes
<br>
type: seqaij
<br>
rows=395,
cols=395
<br>
package
used to perform factorization:
petsc
<br>
total:
nonzeros=132379, allocated
nonzeros=132379
<br>
total number
of mallocs used during
MatSetValues calls =0
<br>
not
using I-node routines
<br>
1 MPI processes
<br>
type: seqaij
<br>
linear system matrix
= precond matrix:
<br>
rows=0,
cols=0
<br>
package
used to perform factorization:
petsc
<br>
total:
nonzeros=1, allocated nonzeros=1
<br>
total
number of mallocs used during
MatSetValues calls =0
<br>
not using
I-node routines
<br>
linear system
matrix = precond matrix:
<br>
Matrix Object: 1
MPI processes
<br>
type: seqaij
<br>
Matrix Object:KSP
Object: 1 MPI
processes
<br>
type: seqaij
<br>
rows=0, cols=0
<br>
total: nonzeros=0,
allocated nonzeros=0
<br>
total number of
mallocs used during MatSetValues
calls =0
<br>
not using
I-node routines
<br>
rows=395, cols=395
<br>
total:
nonzeros=32037, allocated
nonzeros=32037
<br>
total number of
mallocs used during MatSetValues
calls =0
<br>
not using I-node
routines
<br>
- - - - - - - - - - -
- - - - - - -
<br>
KSP Object:
(mg_coarse_sub_) 1 MPI
processes
<br>
type: preonly
<br>
maximum iterations=1,
initial guess is zero
<br>
tolerances:
relative=1e-05, absolute=1e-50,
divergence=10000
<br>
left preconditioning
<br>
using NONE norm type
for convergence test
<br>
PC Object:
(mg_coarse_sub_) 1 MPI
processes
<br>
type: lu
<br>
LU: out-of-place
factorization
<br>
tolerance for zero
pivot 2.22045e-14
<br>
using diagonal
shift on blocks to prevent zero
pivot
<br>
matrix ordering: nd
<br>
factor fill ratio
given 5, needed 0
<br>
Factored matrix
follows:
<br>
Matrix
Object: 1 MPI
processes
<br>
type: seqaij
<br>
rows=0,
cols=0
<br>
package used
to perform factorization: petsc
<br>
total:
nonzeros=1, allocated nonzeros=1
<br>
total number
of mallocs used during
MatSetValues calls =0
<br>
not using
I-node routines
<br>
linear system matrix
= precond matrix:
<br>
Matrix
Object: 1 MPI
processes
<br>
type: seqaij
<br>
rows=0, cols=0
<br>
total: nonzeros=0,
allocated nonzeros=0
<br>
total number of
mallocs used during MatSetValues
calls =0
<br>
not using I-node
routines
<br>
(mg_coarse_sub_) 1
MPI processes
<br>
type: preonly
<br>
maximum iterations=1,
initial guess is zero
<br>
tolerances:
relative=1e-05, absolute=1e-50,
divergence=10000
<br>
left preconditioning
<br>
using NONE norm type
for convergence test
<br>
PC Object:
(mg_coarse_sub_) 1 MPI
processes
<br>
type: lu
<br>
LU: out-of-place
factorization
<br>
tolerance for zero
pivot 2.22045e-14
<br>
using diagonal
shift on blocks to prevent zero
pivot
<br>
matrix ordering: nd
<br>
factor fill ratio
given 5, needed 0
<br>
Factored matrix
follows:
<br>
Matrix
Object: 1 MPI
processes
<br>
type: seqaij
<br>
rows=0,
cols=0
<br>
package used
to perform factorization: petsc
<br>
total:
nonzeros=1, allocated nonzeros=1
<br>
total number
of mallocs used during
MatSetValues calls =0
<br>
not using
I-node routines
<br>
linear system matrix
= precond matrix:
<br>
Matrix
Object: 1 MPI
processes
<br>
type: seqaij
<br>
rows=0, cols=0
<br>
total: nonzeros=0,
allocated nonzeros=0
<br>
total number of
mallocs used during MatSetValues
calls =0
<br>
not using I-node
routines
<br>
[1] number of local
blocks = 1, first local block
number = 1
<br>
[1] local block number
0
<br>
- - - - - - - - - - - -
- - - - - -
<br>
[2] number of local
blocks = 1, first local block
number = 2
<br>
[2] local block number
0
<br>
- - - - - - - - - - - -
- - - - - -
<br>
[3] number of local
blocks = 1, first local block
number = 3
<br>
[3] local block number
0
<br>
- - - - - - - - - - - -
- - - - - -
<br>
linear system matrix =
precond matrix:
<br>
Matrix Object: 4
MPI processes
<br>
type: mpiaij
<br>
rows=395, cols=395
<br>
total: nonzeros=32037,
allocated nonzeros=32037
<br>
total number of mallocs
used during MatSetValues calls
=0
<br>
not using I-node (on
process 0) routines
<br>
Down solver (pre-smoother) on
level 1
-------------------------------
<br>
KSP Object:
(mg_levels_1_) 4 MPI
processes
<br>
type: chebyshev
<br>
Chebyshev: eigenvalue
estimates: min = 0.0636225, max
= 1.33607
<br>
maximum iterations=2
<br>
tolerances:
relative=1e-05, absolute=1e-50,
divergence=10000
<br>
left preconditioning
<br>
using nonzero initial
guess
<br>
using NONE norm type for
convergence test
<br>
PC Object:
(mg_levels_1_) 4 MPI
processes
<br>
type: jacobi
<br>
linear system matrix =
precond matrix:
<br>
Matrix Object: 4
MPI processes
<br>
type: mpiaij
<br>
rows=23918, cols=23918
<br>
total: nonzeros=818732,
allocated nonzeros=818732
<br>
total number of mallocs
used during MatSetValues calls
=0
<br>
not using I-node (on
process 0) routines
<br>
Up solver (post-smoother)
same as down solver
(pre-smoother)
<br>
Down solver (pre-smoother) on
level 2
-------------------------------
<br>
KSP Object:
(mg_levels_2_) 4 MPI
processes
<br>
type: chebyshev
<br>
Chebyshev: eigenvalue
estimates: min = 0.0971369, max
= 2.03987
<br>
maximum iterations=2
<br>
tolerances:
relative=1e-05, absolute=1e-50,
divergence=10000
<br>
left preconditioning
<br>
using nonzero initial
guess
<br>
using NONE norm type for
convergence test
<br>
PC Object:
(mg_levels_2_) 4 MPI
processes
<br>
type: jacobi
<br>
linear system matrix =
precond matrix:
<br>
Matrix Object: 4
MPI processes
<br>
type: mpiaij
<br>
rows=262144,
cols=262144
<br>
total:
nonzeros=1835008, allocated
nonzeros=1835008
<br>
total number of mallocs
used during MatSetValues calls
=0
<br>
Up solver (post-smoother)
same as down solver
(pre-smoother)
<br>
linear system matrix =
precond matrix:
<br>
Matrix Object: 4 MPI
processes
<br>
type: mpiaij
<br>
rows=262144, cols=262144
<br>
total: nonzeros=1835008,
allocated nonzeros=1835008
<br>
total number of mallocs
used during MatSetValues calls
=0
<br>
#PETSc Option Table entries:
<br>
-ksp_view
<br>
-options_left
<br>
-pc_gamg_agg_nsmooths 1
<br>
-pc_type gamg
<br>
#End of PETSc Option Table
entries
<br>
There are no unused options.
<br>
<br>
<br>
Thank you,
<br>
Michele
<br>
<br>
<br>
<br>
<br>
<br>
</blockquote>
</blockquote>
</blockquote>
</blockquote>
</blockquote>
</blockquote>
</blockquote>
</blockquote>
</blockquote>
</blockquote>
</blockquote>
</blockquote>
<test_poisson_solver.tar.gz><2decomp_fft-1.5.847-modified.tar.gz>
<br>
<br>
</blockquote>
</blockquote>
</blockquote>
<br>
</blockquote>
<br>
</blockquote>
<br>
</blockquote>
<br>
</body>
</html>