pushed now. satish On Tue, 10 Apr 2012, Barry Smith wrote:
Satish,
Please add this patch.
Thanks
Barry
Begin forwarded message:
From: Lisandro Dalcin <[email protected]> Date: April 10, 2012 12:43:39 PM CDT To: For users of the development version of PETSc <[email protected]> Subject: Re: [petsc-dev] A few remaining issues with Mat block sizes Reply-To: For users of the development version of PETSc <[email protected]>
On 10 April 2012 17:58, Barry Smith <[email protected]> wrote:
Lisandro,
Thanks, I have incorporated these and a few other blocksize fixes.
Many thanks, Barry! However, I still think you are missing this one (I found the issue after running petsc4py testsuite). Pasted below and also attached. Please review and push.
diff -r 2d1c2b03ef3a src/mat/impls/aij/seq/aij.c --- a/src/mat/impls/aij/seq/aij.c Tue Apr 10 09:57:31 2012 -0500 +++ b/src/mat/impls/aij/seq/aij.c Tue Apr 10 20:36:48 2012 +0300 @@ -3477,9 +3477,12 @@ PetscErrorCode ierr;
PetscFunctionBegin; + if (Ii[0]) SETERRQ1(PETSC_COMM_SELF,PETSC_ERR_ARG_OUTOFRANGE, "Ii[0] must be 0 it is %D", Ii[0]); + + ierr = PetscLayoutSetUp(B->rmap);CHKERRQ(ierr); + ierr = PetscLayoutSetUp(B->cmap);CHKERRQ(ierr); + ierr = MatGetSize(B, &m, &n);CHKERRQ(ierr); - - if (Ii[0]) SETERRQ1(PETSC_COMM_SELF,PETSC_ERR_ARG_OUTOFRANGE, "Ii[0] must be 0 it is %D", Ii[0]); ierr = PetscMalloc((m+1) * sizeof(PetscInt), &nnz);CHKERRQ(ierr); for(i = 0; i < m; i++) { nz = Ii[i+1]- Ii[i];
-- Lisandro Dalcin --------------- CIMEC (INTEC/CONICET-UNL) Predio CONICET-Santa Fe Colectora RN 168 Km 472, Paraje El Pozo 3000 Santa Fe, Argentina Tel: +54-342-4511594 (ext 1011) Tel/Fax: +54-342-4511169 [see attached file: matblocksize-seqaij-csr.diff]