Jed, Thanks. However: [ 19%] Building C object CMakeFiles/petsc.dir/src/sys/ftn-custom/zstartf.c.o [ 19%] Building C object CMakeFiles/petsc.dir/src/sys/ftn-custom/zsys.c.o Building C object CMakeFiles/petsc.dir/src/sys/ftn-custom/zutils.c.o /panfs/panfs.gfz-hpcc.cluster/home/mt/agrayver/lib/petsc-dev1/src/sys/ftn-custom/zstart.c(117): error: expected a type specifier extern MPI_Op MPIU_SUM; ^ /panfs/panfs.gfz-hpcc.cluster/home/mt/agrayver/lib/petsc-dev1/src/sys/ftn-custom/zstart.c(117): error: "MPI_Op" has already been declared in the current scope extern MPI_Op MPIU_SUM; ^ /panfs/panfs.gfz-hpcc.cluster/home/mt/agrayver/lib/petsc-dev1/src/sys/ftn-custom/zstart.c(128): error: function "MPI_Op" is not a type name extern MPI_Op PetscMaxSum_Op; ^ /panfs/panfs.gfz-hpcc.cluster/home/mt/agrayver/lib/petsc-dev1/src/sys/ftn-custom/zstart.c(220): error: function "MPI_Op" is not a type name extern MPI_Op PetscADMax_Op; ^ /panfs/panfs.gfz-hpcc.cluster/home/mt/agrayver/lib/petsc-dev1/src/sys/ftn-custom/zstart.c(221): error: function "MPI_Op" is not a type name extern MPI_Op PetscADMin_Op; ^ /panfs/panfs.gfz-hpcc.cluster/home/mt/agrayver/lib/petsc-dev1/src/sys/ftn-custom/zstart.c(343): error: expression must be an lvalue or a function designator *ierr = MPI_Op_create(PetscSum_Local,1,&MPIU_SUM); On 23.04.2012 03:23, Jed Brown wrote:
Thanks for the test case. Fixed here:
http://petsc.cs.iit.edu/petsc/petsc-dev/rev/7ac7040f19d7
On Sun, Apr 22, 2012 at 08:11, Alexander Grayver <[email protected] <mailto:[email protected]>> wrote:
Matt,
Here are matrices (double complex, 3 MB each): http://dl.dropbox.com/u/60982984/X.dat http://dl.dropbox.com/u/60982984/Y.dat
Code:
static char help[] = ""; int main(int argc,char **args) { Mat A,B; PetscErrorCode ierr; PetscViewer viewer; PetscScalar cone;
PetscInitialize(&argc,&args,(char *)0,help);
ierr = PetscViewerBinaryOpen(PETSC_COMM_WORLD,"Y.dat",FILE_MODE_READ,&viewer);CHKERRQ(ierr); ierr = MatCreate(PETSC_COMM_WORLD,&A);CHKERRQ(ierr); ierr = MatSetFromOptions(A);CHKERRQ(ierr); ierr = MatLoad(A,viewer);CHKERRQ(ierr); ierr = PetscViewerDestroy(&viewer);CHKERRQ(ierr);
ierr = PetscViewerBinaryOpen(PETSC_COMM_WORLD,"X.dat",FILE_MODE_READ,&viewer);CHKERRQ(ierr); ierr = MatCreate(PETSC_COMM_WORLD,&B);CHKERRQ(ierr); ierr = MatSetFromOptions(B);CHKERRQ(ierr); ierr = MatLoad(B,viewer);CHKERRQ(ierr); ierr = PetscViewerDestroy(&viewer);CHKERRQ(ierr);
cone = 1.0; ierr = MatAYPX(A,cone,B,DIFFERENT_NONZERO_PATTERN); ierr = PetscFinalize(); return 0; }
It works fine if you run it with: mpirun -n 10 solveTest -mat_type aij
and crashes if n > 10
Thanks.
On 22.04.2012 14:51, Matthew Knepley wrote:
On Sun, Apr 22, 2012 at 8:41 AM, Alexander Grayver <[email protected] <mailto:[email protected]>> wrote:
Sorry, I don't quite get. Is this something I can avoid by doing preallocation properly or petsc issue that will be fixed?
This looks like a PETSc bug to me, but it would be nice to get input that triggers it.
Thanks,
Matt
On 20.04.2012 14:23, Jed Brown wrote:
On Fri, Apr 20, 2012 at 05:18, Matthew Knepley <[email protected] <mailto:[email protected]>> wrote:
Did someone intend to fix this, and then stop? MatAXPY_BasicWithPreallocation() is the same as MatAXPY_Basic().
Look how it's called. I don't know what's going on here.
ierr = MatAXPYGetPreallocation_SeqAIJ(yy->A,xx->A,nnz_d);CHKERRQ(ierr); ierr = MatAXPYGetPreallocation_SeqAIJ(yy->B,xx->B,nnz_o);CHKERRQ(ierr); ierr = MatMPIAIJSetPreallocation(B,0,nnz_d,0,nnz_o);CHKERRQ(ierr); ierr = MatAXPY_BasicWithPreallocation(B,Y,a,X,str);CHKERRQ(ierr);
-- Regards, Alexander
-- What most experimenters take for granted before they begin their experiments is infinitely more interesting than any results to which their experiments lead. -- Norbert Wiener
-- Regards, Alexander
-- Regards, Alexander