March 2012 - petsc-dev - mailman3.cels.anl.gov

Re: [petsc-dev] buildsystem: cusp location
by Jed Brown 25 Mar '12

25 Mar '12

On Sun, Mar 25, 2012 at 11:59, Satish Balay <balay(a)mcs.anl.gov> wrote: > When configure is doing --download-cusp - it can determine how cusp is > organized. [I think we try to place in PETSC_DIR/PETSC_ARCH/include. > But since cusp decides how it installs itself [say user installs and > installs manually] - we try to support that mode using > --with-cusp-dir= option. There is no 'include/' dir in that mode of > install. > Well isn't that what --with-cusp-include= is for? I even wouldn't object to being lenient, but I don't think it makes sense to abuse the meaning of --with-cusp-dir=. > > Yes - currently the situation is not ideal. The ideal situation is > folks install cusp thrust - the recommended way - i.e along with cuda > in cuda includes. One has to do this with thrust anyway - otherwise > nvcc will always pick up thrust bundled with cuda - and now what you > specify with -I/path-to-thrust. > > The other problem currently with cusp/thrust is - we've been using dev > snapshot of cusp/thrust for a while. I don't know if there is a > release snapshot of these packages that we can switch to now [or not]. >

1 0

Re: [petsc-dev] buildsystem: cusp location
by Satish Balay 25 Mar '12

25 Mar '12

On Sun, 25 Mar 2012, Sean Farley wrote: > So, I wanted to try out cuda (since I heard you can get 400x speedup) but I > ran into a problem with cusp. I specified where I installed cusp: > --with-cusp-dir=$HOME/local but configure gave the error: > > ******************************************************************************* > UNABLE to CONFIGURE with GIVEN OPTIONS (see configure.log for > details): > ------------------------------------------------------------------------------- > --with-cusp-dir=/opt/local did not work > ******************************************************************************* > > Looking at configure.log, I see that flag passed to nvcc is -I$HOME/local > instead of -I$HOME/local/include. Investigating even more, cusp.py has the > following line: > > self.includedir = 'include' # default and --download have different > includedirs > > I consider this a flaw. Just because a library doesn't build a .so / > .dylib, we shove the header files anywhere? From what I can see, we could > either: > > 1) [backwards compatible] search both --with-cusp-dir=$FOO and $FOO/include > > 2) set self.includedir = 'include' permanently; not just in Install() > > I was never any good at Jenga so (2) is what I prefer :-) Thoughts? > When configure is doing --download-cusp - it can determine how cusp is organized. [I think we try to place in PETSC_DIR/PETSC_ARCH/include. But since cusp decides how it installs itself [say user installs and installs manually] - we try to support that mode using --with-cusp-dir= option. There is no 'include/' dir in that mode of install. Yes - currently the situation is not ideal. The ideal situation is folks install cusp thrust - the recommended way - i.e along with cuda in cuda includes. One has to do this with thrust anyway - otherwise nvcc will always pick up thrust bundled with cuda - and now what you specify with -I/path-to-thrust. The other problem currently with cusp/thrust is - we've been using dev snapshot of cusp/thrust for a while. I don't know if there is a release snapshot of these packages that we can switch to now [or not]. Satish

1 0

buildsystem: cusp location
by Sean Farley 25 Mar '12

25 Mar '12

So, I wanted to try out cuda (since I heard you can get 400x speedup) but I ran into a problem with cusp. I specified where I installed cusp: --with-cusp-dir=$HOME/local but configure gave the error: ******************************************************************************* UNABLE to CONFIGURE with GIVEN OPTIONS (see configure.log for details): ------------------------------------------------------------------------------- --with-cusp-dir=/opt/local did not work ******************************************************************************* Looking at configure.log, I see that flag passed to nvcc is -I$HOME/local instead of -I$HOME/local/include. Investigating even more, cusp.py has the following line: self.includedir = 'include' # default and --download have different includedirs I consider this a flaw. Just because a library doesn't build a .so / .dylib, we shove the header files anywhere? From what I can see, we could either: 1) [backwards compatible] search both --with-cusp-dir=$FOO and $FOO/include 2) set self.includedir = 'include' permanently; not just in Install() I was never any good at Jenga so (2) is what I prefer :-) Thoughts?

1 0

Re: [petsc-dev] Petsc functions must ALWAYS return error codes
by Mark F. Adams 25 Mar '12

25 Mar '12

Barry, The iterators are a problem. THe code now looks like: for(pos=PetscCDGetHeadPos(agg_llists,kk) ; pos ; pos=PetscCDGetNextPos(agg_llists,kk,pos)){ .... } I could use: ierr = PetscCDGetHeadPos(agg_llists,kk, &pos); while( pos ){ ... ierr = PetscCDGetNextPos(agg_llists, kk, &pos ); } Is this what you would like or do you have another API suggestion? Mark On Mar 24, 2012, at 11:21 PM, Barry Smith wrote: > > There are a bunch of these functions that don't return error codes and need to be fixed. > > Barry > > PetscInt LLNGetID( const PetscCDIntNd *a_this ); > PetscErrorCode PetscCDAppendID( PetscCoarsenData *ail, PetscInt a_idx, PetscInt a_gid ); > PetscErrorCode PetscCDAppendRemove( PetscCoarsenData *ail, PetscInt a_destidx, PetscInt a_srcidx ); > PetscErrorCode PetscCDAppendNode( PetscCoarsenData *ail, PetscInt a_idx, PetscCDIntNd *a_n ); > PetscErrorCode PetscCDRemoveNextNode( PetscCoarsenData *ail, PetscInt a_idx, PetscCDIntNd *a_last ); > PetscErrorCode PetscCDRemoveAllAt( PetscCoarsenData *ail, PetscInt a_idx ); > PetscInt PetscCDSizeAt( const PetscCoarsenData *ail, PetscInt a_idx ); > PetscErrorCode PetscCDSetChuckSize( PetscCoarsenData *ail, PetscInt a_sz ); > PetscErrorCode PetscCDPrint( const PetscCoarsenData *ail, MPI_Comm comm ); > PetscErrorCode PetscCDGetMIS( PetscCoarsenData *ail, IS * ); > PetscErrorCode PetscCDGetMat( const PetscCoarsenData *ail, Mat * ); > PetscErrorCode PetscCDSetMat( PetscCoarsenData *ail, Mat ); > typedef PetscCDIntNd* PetscCDPos; > const PetscCDPos PetscCDGetHeadPos( const PetscCoarsenData *ail, PetscInt idx ); > const PetscCDPos PetscCDGetNextPos( const PetscCoarsenData *ail, PetscInt idx, PetscCDPos cpos); > > > >

1 0

Petsc functions must ALWAYS return error codes
by Barry Smith 24 Mar '12

24 Mar '12

There are a bunch of these functions that don't return error codes and need to be fixed. Barry PetscInt LLNGetID( const PetscCDIntNd *a_this ); PetscErrorCode PetscCDAppendID( PetscCoarsenData *ail, PetscInt a_idx, PetscInt a_gid ); PetscErrorCode PetscCDAppendRemove( PetscCoarsenData *ail, PetscInt a_destidx, PetscInt a_srcidx ); PetscErrorCode PetscCDAppendNode( PetscCoarsenData *ail, PetscInt a_idx, PetscCDIntNd *a_n ); PetscErrorCode PetscCDRemoveNextNode( PetscCoarsenData *ail, PetscInt a_idx, PetscCDIntNd *a_last ); PetscErrorCode PetscCDRemoveAllAt( PetscCoarsenData *ail, PetscInt a_idx ); PetscInt PetscCDSizeAt( const PetscCoarsenData *ail, PetscInt a_idx ); PetscErrorCode PetscCDSetChuckSize( PetscCoarsenData *ail, PetscInt a_sz ); PetscErrorCode PetscCDPrint( const PetscCoarsenData *ail, MPI_Comm comm ); PetscErrorCode PetscCDGetMIS( PetscCoarsenData *ail, IS * ); PetscErrorCode PetscCDGetMat( const PetscCoarsenData *ail, Mat * ); PetscErrorCode PetscCDSetMat( PetscCoarsenData *ail, Mat ); typedef PetscCDIntNd* PetscCDPos; const PetscCDPos PetscCDGetHeadPos( const PetscCoarsenData *ail, PetscInt idx ); const PetscCDPos PetscCDGetNextPos( const PetscCoarsenData *ail, PetscInt idx, PetscCDPos cpos);

1 0

Re: [petsc-dev] Reminder: *every* symbol with extern linkage *must* be namespaced
by Lisandro Dalcin 24 Mar '12

24 Mar '12

On 24 March 2012 04:28, Jed Brown <jedbrown(a)mcs.anl.gov> wrote: > On Sat, Mar 17, 2012 at 15:32, Mark F. Adams <mark.adams(a)columbia.edu> > wrote: >> >> done, but it won't show up until I check in my current repo, which could >> be a week. > > > http://petsc.cs.iit.edu/petsc/petsc-dev/rev/597bc483f35d > > Mark, this patch adds types and symbols that are not namespaced. > Also, I'm getting GCC a few GCC warnings: /Users/dalcinl/Devel/petsc-dev/include/petsc-private/matimpl.h:391: warning: type qualifiers ignored on function return type /Users/dalcinl/Devel/petsc-dev/include/petsc-private/matimpl.h:392: warning: type qualifiers ignored on function return type -- 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

1 0

Re: [petsc-dev] PCFieldSplitSetFields() API change
by Barry Smith 23 Mar '12

23 Mar '12

On Mar 23, 2012, at 6:22 PM, Jed Brown wrote: > But what does using B as 00 block actually mean? It seems to me that you pivot, producing P' P J where P J has B in the 00 location. I am not arguing if it is pivoting or not. I'm arguing what the user interface should be. Say I want D to be the 00 block then I just set that with PCFieldSplitSetFields() as the zeroth field. My point is that I am proposing a SINGLE interface for setting the blocks locations in all circumstances, while you are proposing having an additional interface for certain 00 configurations (pivoting) thus with your model you first call the PCFieldSplitSetFields() and then something else. What is your proposed API? Barry > > On Mar 23, 2012 6:13 PM, "Barry Smith" <bsmith(a)mcs.anl.gov> wrote: > > On Mar 23, 2012, at 6:00 PM, Jed Brown wrote: > > > On Fri, Mar 23, 2012 at 17:48, Barry Smith <bsmith(a)mcs.anl.gov> wrote: > > I'm confused. I am NOT injecting it into the original definition, I am injecting it into the PCFieldSplit in the same way that your "permutation" would inject it just into the PCFieldSplit, it is just (in my opinion) a more user friendly way of the user providing this permutation. Note that this "two fields" business is NOT going in the DM (and hence into the original definition of the problem) and would normally only be used when the DM definition was "bad" (due to zero block for example) just like your permutation would only be used when things were bad. > > > > 1. The PCFieldSplitSetIS() interface does not have this. > > > > 2. The reason for pivoting is not that the DM ordering was "bad", it's that a symmetric ordering without the 0 block does not exist (factorization does not work with _any_ symmetric ordering). So you pivot for factorization and restore. > > I don't agree. Consider [A B ; C D] where all four matrices are square and non-singular. It may be that using B as the 00 block is just better for convergence etc > > > > > 3. You now have one name for a set of rows and columns that may not "go together". > > That name is only for setting fieldsplit options etc. It is not fieldnames in the way that DM fieldnames are names of fields. > > > > > Somehow I think this is just more confusing and less run-time-friendly, but maybe that's my fault. In any case, the change has to be consistent (PCFieldSplitSetIS) and we have to decide how the DM can specify when pivoting is needed. > > You may be right. Spec out your alternative Functions I would call and command line options I would use to get the same effect as the affect Jungho and I were trying to achieve. I was bothered by the separation in your original proposal between the > selection of the blocks the "pivoting" of your model. I don't really see using a different block in 00 as necessarily a "pivoting" that is why you proposal turned me off. > > > Barry > >

1 0

communicators
by Jed Brown 23 Mar '12

23 Mar '12

Pretty much every use of COMM_WORLD outside of src/sys/ is a bug. $ find src -name '*.c' | grep -v '$examples\|src/sys\|src/contrib\|src/benchmarks$' | xargs grep COMM_WORLD I just fixed a few obvious cases before running this. If there is COMM_WORLD in your code, please consider fixing it. src/dm/impls/adda/adda.c: ierr = MPI_Comm_size(PETSC_COMM_WORLD,&comm_size);CHKERRQ(ierr); src/dm/impls/da/usfft/matusfft.c: ierr = PetscPrintf(PETSC_COMM_WORLD, "direction = %d, usfft->ndim = %d\n", direction, usfft->ndim);CHKERRQ(ierr); src/dm/impls/da/usfft/matusfft.c: ierr = PetscPrintf(PETSC_COMM_WORLD, "usfft->outdim[%d] = %d\n", ii, usfft->outdim[ii]);CHKERRQ(ierr); src/dm/impls/iga/iga.c: SETERRQ(PETSC_COMM_WORLD, PETSC_ERR_ARG_OUTOFRANGE, "Discretization inconsistent: polynomial order must be greater than degree of continuity"); src/dm/impls/iga/iga.c: ierr = DMDACreate(PETSC_COMM_WORLD,&iga->da_dof); CHKERRQ(ierr); src/dm/impls/iga/iga.c: SETERRQ(PETSC_COMM_WORLD, PETSC_ERR_FILE_OPEN, "Cannot find geometry file"); src/dm/impls/iga/iga.c: SETERRQ(PETSC_COMM_WORLD, PETSC_ERR_ARG_OUTOFRANGE, "Geometry dimension != problem dimension"); src/dm/impls/iga/iga.c: ierr = DMDACreate(PETSC_COMM_WORLD,&iga->da_dof); CHKERRQ(ierr); src/dm/impls/iga/iga.c: ierr = DMDACreate(PETSC_COMM_WORLD,&iga->da_geometry); CHKERRQ(ierr); src/dm/impls/iga/iga.c: SETERRQ(PETSC_COMM_WORLD, PETSC_ERR_ARG_OUTOFRANGE, "Discretization inconsistent: polynomial order must be greater than degree of continuity"); src/dm/impls/iga/iga.c: SETERRQ(PETSC_COMM_WORLD, PETSC_ERR_ARG_OUTOFRANGE, "Initialization routine for tapered meshes does not yet support periodicity"); src/dm/impls/iga/iga.c: SETERRQ(PETSC_COMM_WORLD, PETSC_ERR_ARG_OUTOFRANGE, "Initialization routine for tapered meshes does not yet support periodicity"); src/dm/impls/iga/iga.c: ierr = DMDACreate(PETSC_COMM_WORLD,&iga->da_dof); CHKERRQ(ierr); src/dm/impls/iga/iga.c: SETERRQ(PETSC_COMM_WORLD, PETSC_ERR_ARG_OUTOFRANGE, "Discretization inconsistent: polynomial order must be greater than degree of continuity"); src/dm/impls/iga/iga.c: ierr = DMDACreate(PETSC_COMM_WORLD,&iga->da_dof); CHKERRQ(ierr); src/dm/impls/iga/iga.c: SETERRQ(PETSC_COMM_WORLD, PETSC_ERR_ARG_OUTOFRANGE, "Discretization inconsistent: polynomial order must be greater than degree of continuity"); src/dm/impls/iga/iga.c: ierr = DMDACreate(PETSC_COMM_WORLD,&iga->da_dof); CHKERRQ(ierr); src/dm/impls/iga/iga.c: SETERRQ1(PETSC_COMM_WORLD, PETSC_ERR_ARG_OUTOFRANGE, "Unimplemented number of gauss points %d!", n); src/dm/impls/iga/iga.c: SETERRQ(PETSC_COMM_WORLD, PETSC_ERR_ARG_OUTOFRANGE, "Inserted knots beyond original knot vector limits"); src/dm/impls/mesh/ftn-custom/zmesh.c: *ierr = PetscPrintf(PETSC_COMM_WORLD, "%s %d allocations %d bytes\n", pN, logger.getNumAllocations(pN), logger.getAllocationTotal(pN)); src/dm/impls/mesh/ftn-custom/zmesh.c: *ierr = PetscPrintf(PETSC_COMM_WORLD, "%s %d deallocations %d bytes\n", pN, logger.getNumDeallocations(pN), logger.getDeallocationTotal(pN)); src/dm/impls/mesh/mesh.c: SETERRQ(PETSC_COMM_WORLD, PETSC_ERR_SUP, "Peter needs to incorporate his code."); src/dm/impls/mesh/mesh.c: SETERRQ(PETSC_COMM_WORLD, PETSC_ERR_SUP, "Peter needs to incorporate his code."); src/dm/impls/mesh/meshlagrit.c: //Obj<PETSC_MESH_TYPE> m = ALE::LaGriT::Builder::readMesh(PETSC_COMM_WORLD, 3, options->baseFilename, options->interpolate, options->debug);' src/ksp/pc/impls/asa/asa.c: ierr = PetscPrintf(PETSC_COMM_WORLD, src/ksp/pc/impls/gamg/agg.c:/* PetscPrintf(PETSC_COMM_WORLD," **** [%d]%s %d total done, %d nodes (%d local done), min agg. size = %d\n",mype,__FUNCT__,ii,kk/bs,ndone,jj); */ src/ksp/pc/impls/gamg/agg.c: PetscPrintf(PETSC_COMM_WORLD,"\t\t[%d]%s New grid %d nodes\n",mype,__FUNCT__,ii/col_bs); src/ksp/pc/impls/gamg/geo.c: PetscPrintf(PETSC_COMM_WORLD,"[%d]%s done\n",0,__FUNCT__); src/ksp/pc/impls/gamg/geo.c: PetscPrintf(PETSC_COMM_WORLD,"[%d]%s ERROE: no selected points on coarse grid\n",mype,__FUNCT__); src/ksp/pc/impls/gamg/geo.c: PetscPrintf(PETSC_COMM_WORLD,"[%d]%s failed metric for coarse grid %e\n",mype,__FUNCT__,tm); src/ksp/pc/impls/gamg/geo.c: PetscPrintf(PETSC_COMM_WORLD,"[%d]%s worst metric for coarse grid = %e\n",mype,__FUNCT__,metric); src/ksp/pc/impls/is/nn/nn.c: ierr = PetscViewerASCIIOpen(PETSC_COMM_WORLD,"coarse.m",&viewer);CHKERRQ(ierr); src/ksp/pc/impls/tfs/comm.c: MPI_Comm_size(MPI_COMM_WORLD,&PCTFS_num_nodes); src/ksp/pc/impls/tfs/comm.c: MPI_Comm_rank(MPI_COMM_WORLD,&PCTFS_my_id); src/ksp/pc/impls/tfs/comm.c: {ierr = MPI_Send(vals,n,MPIU_INT,edge_not_pow_2,MSGTAG0+PCTFS_my_id,MPI_COMM_WORLD);CHKERRQ(ierr);} src/ksp/pc/impls/tfs/comm.c: ierr = MPI_Recv(work,n,MPIU_INT,MPI_ANY_SOURCE,MSGTAG0+edge_not_pow_2, MPI_COMM_WORLD,&status);CHKERRQ(ierr); src/ksp/pc/impls/tfs/comm.c: {ierr = MPI_Send(vals,n,MPIU_INT,dest,MSGTAG2+PCTFS_my_id,MPI_COMM_WORLD);CHKERRQ(ierr);} src/ksp/pc/impls/tfs/comm.c: ierr = MPI_Recv(work,n,MPIU_INT,MPI_ANY_SOURCE,MSGTAG2+dest,MPI_COMM_WORLD, &status);CHKERRQ(ierr); src/ksp/pc/impls/tfs/comm.c: {ierr = MPI_Send(vals,n,MPIU_INT,dest,MSGTAG4+PCTFS_my_id,MPI_COMM_WORLD);CHKERRQ(ierr);} src/ksp/pc/impls/tfs/comm.c: ierr = MPI_Recv(vals,n,MPIU_INT,MPI_ANY_SOURCE,MSGTAG4+dest,MPI_COMM_WORLD, &status);CHKERRQ(ierr); src/ksp/pc/impls/tfs/comm.c: ierr = MPI_Recv(vals,n,MPIU_INT,MPI_ANY_SOURCE,MSGTAG5+edge_not_pow_2,MPI_COMM_WORLD,&status);CHKERRQ(ierr); src/ksp/pc/impls/tfs/comm.c: {ierr = MPI_Send(vals,n,MPIU_INT,edge_not_pow_2,MSGTAG5+PCTFS_my_id,MPI_COMM_WORLD);CHKERRQ(ierr);} src/ksp/pc/impls/tfs/comm.c: {ierr = MPI_Send(vals,n,MPIU_SCALAR,edge_not_pow_2,MSGTAG0+PCTFS_my_id,MPI_COMM_WORLD);CHKERRQ(ierr);} src/ksp/pc/impls/tfs/comm.c: ierr = MPI_Recv(work,n,MPIU_SCALAR,MPI_ANY_SOURCE,MSGTAG0+edge_not_pow_2,MPI_COMM_WORLD,&status);CHKERRQ(ierr); src/ksp/pc/impls/tfs/comm.c: {ierr = MPI_Send(vals,n,MPIU_SCALAR,dest,MSGTAG2+PCTFS_my_id,MPI_COMM_WORLD);CHKERRQ(ierr);} src/ksp/pc/impls/tfs/comm.c: ierr = MPI_Recv(work,n,MPIU_SCALAR,MPI_ANY_SOURCE,MSGTAG2+dest,MPI_COMM_WORLD, &status);CHKERRQ(ierr); src/ksp/pc/impls/tfs/comm.c: {ierr = MPI_Send(vals,n,MPIU_SCALAR,dest,MSGTAG4+PCTFS_my_id,MPI_COMM_WORLD);CHKERRQ(ierr);} src/ksp/pc/impls/tfs/comm.c: ierr = MPI_Recv(vals,n,MPIU_SCALAR,MPI_ANY_SOURCE,MSGTAG4+dest,MPI_COMM_WORLD, &status);CHKERRQ(ierr); src/ksp/pc/impls/tfs/comm.c: ierr = MPI_Recv(vals,n,MPIU_SCALAR,MPI_ANY_SOURCE,MSGTAG5+edge_not_pow_2, MPI_COMM_WORLD,&status);CHKERRQ(ierr); src/ksp/pc/impls/tfs/comm.c: {ierr = MPI_Send(vals,n,MPIU_SCALAR,edge_not_pow_2,MSGTAG5+PCTFS_my_id,MPI_COMM_WORLD);CHKERRQ(ierr);} src/ksp/pc/impls/tfs/comm.c: {ierr = MPI_Send(vals,n,MPIU_SCALAR,dest,MSGTAG2+PCTFS_my_id,MPI_COMM_WORLD);CHKERRQ(ierr);} src/ksp/pc/impls/tfs/comm.c: ierr = MPI_Recv(work,n,MPIU_SCALAR,MPI_ANY_SOURCE,MSGTAG2+dest,MPI_COMM_WORLD,&status);CHKERRQ(ierr); src/ksp/pc/impls/tfs/comm.c: {ierr = MPI_Send(vals,n,MPIU_SCALAR,dest,MSGTAG4+PCTFS_my_id,MPI_COMM_WORLD);CHKERRQ(ierr);} src/ksp/pc/impls/tfs/comm.c: ierr = MPI_Recv(vals,n,MPIU_SCALAR,MPI_ANY_SOURCE,MSGTAG4+dest,MPI_COMM_WORLD,&status);CHKERRQ(ierr); src/ksp/pc/impls/tfs/comm.c: {ierr = MPI_Send(vals+segs[edge],stage_n,MPIU_SCALAR,dest,type, MPI_COMM_WORLD);CHKERRQ(ierr);} src/ksp/pc/impls/tfs/comm.c: ierr = MPI_Recv(work,stage_n,MPIU_SCALAR,MPI_ANY_SOURCE,type,MPI_COMM_WORLD,&status);CHKERRQ(ierr); src/ksp/pc/impls/tfs/comm.c: {ierr = MPI_Send(vals+segs[level-1-edge],stage_n,MPIU_SCALAR,dest,type,MPI_COMM_WORLD);CHKERRQ(ierr);} src/ksp/pc/impls/tfs/comm.c: ierr = MPI_Recv(vals+segs[level-1-edge],stage_n,MPIU_SCALAR, MPI_ANY_SOURCE,type,MPI_COMM_WORLD,&status);CHKERRQ(ierr); src/ksp/pc/impls/tfs/comm.c: {ierr = MPI_Send(vals,n,MPIU_INT,dest,MSGTAG2+PCTFS_my_id,MPI_COMM_WORLD);CHKERRQ(ierr);} src/ksp/pc/impls/tfs/comm.c: ierr = MPI_Recv(work,n,MPIU_INT,MPI_ANY_SOURCE,MSGTAG2+dest,MPI_COMM_WORLD, &status);CHKERRQ(ierr); src/ksp/pc/impls/tfs/comm.c: {ierr = MPI_Send(vals,n,MPIU_INT,dest,MSGTAG4+PCTFS_my_id,MPI_COMM_WORLD);CHKERRQ(ierr);} src/ksp/pc/impls/tfs/comm.c: ierr = MPI_Recv(vals,n,MPIU_INT,MPI_ANY_SOURCE,MSGTAG4+dest,MPI_COMM_WORLD,&status);CHKERRQ(ierr); src/ksp/pc/impls/tfs/gs.c: ierr = gsi_via_bit_mask(gs);CHKERRABORT(PETSC_COMM_WORLD,ierr); src/ksp/pc/impls/tfs/gs.c: ierr = MPI_Comm_group(MPI_COMM_WORLD,&PCTFS_gs_group);CHKERRABORT(PETSC_COMM_WORLD,ierr); src/ksp/pc/impls/tfs/gs.c: ierr = MPI_Comm_create(MPI_COMM_WORLD,PCTFS_gs_group,&PCTFS_gs_comm);CHKERRABORT(PETSC_COMM_WORLD,ierr); src/ksp/pc/impls/tfs/gs.c: ierr = PetscMemzero(gs,sizeof(PCTFS_gs_id));CHKERRABORT(PETSC_COMM_WORLD,ierr); src/ksp/pc/impls/tfs/gs.c: if (!in_elms) SETERRABORT(PETSC_COMM_WORLD,PETSC_ERR_PLIB,"elms point to nothing!!!\n"); src/ksp/pc/impls/tfs/gs.c: if (nel<0) SETERRABORT(PETSC_COMM_WORLD,PETSC_ERR_PLIB,"can't have fewer than 0 elms!!!\n"); src/ksp/pc/impls/tfs/gs.c: {ierr = PetscInfo(0,"I don't have any elements!!!\n");CHKERRABORT(PETSC_COMM_WORLD,ierr);} src/ksp/pc/impls/tfs/gs.c: if (j!=nel) SETERRABORT(PETSC_COMM_WORLD,PETSC_ERR_PLIB,"nel j mismatch!\n"); src/ksp/pc/impls/tfs/gs.c: ierr = PetscInfo(0,"gsi_check_args() :: elm list *not* sorted!\n");CHKERRABORT(PETSC_COMM_WORLD,ierr); src/ksp/pc/impls/tfs/gs.c: ierr = PCTFS_SMI_sort((void*)elms, (void*)companion, nel, SORT_INTEGER);CHKERRABORT(PETSC_COMM_WORLD,ierr); src/ksp/pc/impls/tfs/gs.c: {ierr = PetscInfo(0,"gsi_check_args() :: elm list sorted!\n");CHKERRABORT(PETSC_COMM_WORLD,ierr);} src/ksp/pc/impls/tfs/gs.c: ierr = PCTFS_giop(vals,work,sizeof(oprs)/sizeof(oprs[0])-1,oprs);CHKERRABORT(PETSC_COMM_WORLD,ierr); src/ksp/pc/impls/tfs/gs.c: if (vals[3]<0) SETERRABORT(PETSC_COMM_WORLD,PETSC_ERR_PLIB,"gsi_check_args() :: system not semi-pos def \n"); src/ksp/pc/impls/tfs/gs.c: if (vals[4]==INT_MAX) SETERRABORT(PETSC_COMM_WORLD,PETSC_ERR_PLIB,"gsi_check_args() :: system ub too large !\n"); src/ksp/pc/impls/tfs/gs.c: if (gs->negl<=0) SETERRABORT(PETSC_COMM_WORLD,PETSC_ERR_PLIB,"gsi_check_args() :: system empty or neg :: %d\n"); src/ksp/pc/impls/tfs/xxt.c: if ((1<<(xxt_handle->level=PCTFS_i_log2_num_nodes))!=PCTFS_num_nodes) SETERRQ2(PETSC_COMM_SELF,PETSC_ERR_PLIB,"only 2^k for now and MPI_COMM_WORLD!!! %D != %D\n",1<<PCTFS_i_log2_num_nodes,PCTFS_num_nodes); src/ksp/pc/impls/tfs/xxt.c: if (!PCTFS_my_id) {ierr = PetscPrintf(PETSC_COMM_WORLD,"XXT_stats() :: no stats available!\n");CHKERRQ(ierr);} src/ksp/pc/impls/tfs/xxt.c: ierr = PetscPrintf(PETSC_COMM_WORLD,"%D :: min xxt_nnz=%D\n",PCTFS_my_id,vals[0]);CHKERRQ(ierr); src/ksp/pc/impls/tfs/xxt.c: ierr = PetscPrintf(PETSC_COMM_WORLD,"%D :: max xxt_nnz=%D\n",PCTFS_my_id,vals[1]);CHKERRQ(ierr); src/ksp/pc/impls/tfs/xxt.c: ierr = PetscPrintf(PETSC_COMM_WORLD,"%D :: avg xxt_nnz=%g\n",PCTFS_my_id,1.0*vals[2]/PCTFS_num_nodes);CHKERRQ(ierr); src/ksp/pc/impls/tfs/xxt.c: ierr = PetscPrintf(PETSC_COMM_WORLD,"%D :: tot xxt_nnz=%D\n",PCTFS_my_id,vals[2]);CHKERRQ(ierr); src/ksp/pc/impls/tfs/xxt.c: ierr = PetscPrintf(PETSC_COMM_WORLD,"%D :: xxt C(2d) =%g\n",PCTFS_my_id,vals[2]/(pow(1.0*vals[5],1.5)));CHKERRQ(ierr); src/ksp/pc/impls/tfs/xxt.c: ierr = PetscPrintf(PETSC_COMM_WORLD,"%D :: xxt C(3d) =%g\n",PCTFS_my_id,vals[2]/(pow(1.0*vals[5],1.6667)));CHKERRQ(ierr); src/ksp/pc/impls/tfs/xxt.c: ierr = PetscPrintf(PETSC_COMM_WORLD,"%D :: min xxt_n =%D\n",PCTFS_my_id,vals[3]);CHKERRQ(ierr); src/ksp/pc/impls/tfs/xxt.c: ierr = PetscPrintf(PETSC_COMM_WORLD,"%D :: max xxt_n =%D\n",PCTFS_my_id,vals[4]);CHKERRQ(ierr); src/ksp/pc/impls/tfs/xxt.c: ierr = PetscPrintf(PETSC_COMM_WORLD,"%D :: avg xxt_n =%g\n",PCTFS_my_id,1.0*vals[5]/PCTFS_num_nodes);CHKERRQ(ierr); src/ksp/pc/impls/tfs/xxt.c: ierr = PetscPrintf(PETSC_COMM_WORLD,"%D :: tot xxt_n =%D\n",PCTFS_my_id,vals[5]);CHKERRQ(ierr); src/ksp/pc/impls/tfs/xxt.c: ierr = PetscPrintf(PETSC_COMM_WORLD,"%D :: min xxt_buf=%D\n",PCTFS_my_id,vals[6]);CHKERRQ(ierr); src/ksp/pc/impls/tfs/xxt.c: ierr = PetscPrintf(PETSC_COMM_WORLD,"%D :: max xxt_buf=%D\n",PCTFS_my_id,vals[7]);CHKERRQ(ierr); src/ksp/pc/impls/tfs/xxt.c: ierr = PetscPrintf(PETSC_COMM_WORLD,"%D :: avg xxt_buf=%g\n",PCTFS_my_id,1.0*vals[8]/PCTFS_num_nodes);CHKERRQ(ierr); src/ksp/pc/impls/tfs/xxt.c: ierr = PetscPrintf(PETSC_COMM_WORLD,"%D :: min xxt_slv=%g\n",PCTFS_my_id,fvals[0]);CHKERRQ(ierr); src/ksp/pc/impls/tfs/xxt.c: ierr = PetscPrintf(PETSC_COMM_WORLD,"%D :: max xxt_slv=%g\n",PCTFS_my_id,fvals[1]);CHKERRQ(ierr); src/ksp/pc/impls/tfs/xxt.c: ierr = PetscPrintf(PETSC_COMM_WORLD,"%D :: avg xxt_slv=%g\n",PCTFS_my_id,fvals[2]/PCTFS_num_nodes);CHKERRQ(ierr); src/ksp/pc/impls/tfs/xxt.c: {ierr = PetscPrintf(PETSC_COMM_WORLD,"xxt_generate() :: null space exists %D %D %D\n",m,j,xxt_handle->ns);} src/ksp/pc/impls/tfs/xyt.c: if ((1<<(xyt_handle->level=PCTFS_i_log2_num_nodes))!=PCTFS_num_nodes) SETERRQ2(PETSC_COMM_SELF,PETSC_ERR_PLIB,"only 2^k for now and MPI_COMM_WORLD!!! %D != %D\n",1<<PCTFS_i_log2_num_nodes,PCTFS_num_nodes); src/ksp/pc/impls/tfs/xyt.c: if (!PCTFS_my_id) PetscPrintf(PETSC_COMM_WORLD,"XYT_stats() :: no stats available!\n"); src/ksp/pc/impls/tfs/xyt.c: PetscPrintf(PETSC_COMM_WORLD,"%D :: min xyt_nnz=%D\n",PCTFS_my_id,vals[0]); src/ksp/pc/impls/tfs/xyt.c: PetscPrintf(PETSC_COMM_WORLD,"%D :: max xyt_nnz=%D\n",PCTFS_my_id,vals[1]); src/ksp/pc/impls/tfs/xyt.c: PetscPrintf(PETSC_COMM_WORLD,"%D :: avg xyt_nnz=%g\n",PCTFS_my_id,1.0*vals[2]/PCTFS_num_nodes); src/ksp/pc/impls/tfs/xyt.c: PetscPrintf(PETSC_COMM_WORLD,"%D :: tot xyt_nnz=%D\n",PCTFS_my_id,vals[2]); src/ksp/pc/impls/tfs/xyt.c: PetscPrintf(PETSC_COMM_WORLD,"%D :: xyt C(2d) =%g\n",PCTFS_my_id,vals[2]/(pow(1.0*vals[5],1.5))); src/ksp/pc/impls/tfs/xyt.c: PetscPrintf(PETSC_COMM_WORLD,"%D :: xyt C(3d) =%g\n",PCTFS_my_id,vals[2]/(pow(1.0*vals[5],1.6667))); src/ksp/pc/impls/tfs/xyt.c: PetscPrintf(PETSC_COMM_WORLD,"%D :: min xyt_n =%D\n",PCTFS_my_id,vals[3]); src/ksp/pc/impls/tfs/xyt.c: PetscPrintf(PETSC_COMM_WORLD,"%D :: max xyt_n =%D\n",PCTFS_my_id,vals[4]); src/ksp/pc/impls/tfs/xyt.c: PetscPrintf(PETSC_COMM_WORLD,"%D :: avg xyt_n =%g\n",PCTFS_my_id,1.0*vals[5]/PCTFS_num_nodes); src/ksp/pc/impls/tfs/xyt.c: PetscPrintf(PETSC_COMM_WORLD,"%D :: tot xyt_n =%D\n",PCTFS_my_id,vals[5]); src/ksp/pc/impls/tfs/xyt.c: PetscPrintf(PETSC_COMM_WORLD,"%D :: min xyt_buf=%D\n",PCTFS_my_id,vals[6]); src/ksp/pc/impls/tfs/xyt.c: PetscPrintf(PETSC_COMM_WORLD,"%D :: max xyt_buf=%D\n",PCTFS_my_id,vals[7]); src/ksp/pc/impls/tfs/xyt.c: PetscPrintf(PETSC_COMM_WORLD,"%D :: avg xyt_buf=%g\n",PCTFS_my_id,1.0*vals[8]/PCTFS_num_nodes); src/ksp/pc/impls/tfs/xyt.c: PetscPrintf(PETSC_COMM_WORLD,"%D :: min xyt_slv=%g\n",PCTFS_my_id,fvals[0]); src/ksp/pc/impls/tfs/xyt.c: PetscPrintf(PETSC_COMM_WORLD,"%D :: max xyt_slv=%g\n",PCTFS_my_id,fvals[1]); src/ksp/pc/impls/tfs/xyt.c: PetscPrintf(PETSC_COMM_WORLD,"%D :: avg xyt_slv=%g\n",PCTFS_my_id,fvals[2]/PCTFS_num_nodes); src/ksp/pc/impls/tfs/xyt.c: {ierr = PetscPrintf(PETSC_COMM_WORLD,"xyt_generate() :: null space exists %D %D %D\n",m,j,xyt_handle->ns);} src/mat/coarsen/impls/hem/hem.c: ierr = MPI_Comm_rank( PETSC_COMM_WORLD, &mype ); CHKERRQ(ierr); src/mat/coarsen/impls/hem/hem.c: /* PetscPrintf(PETSC_COMM_WORLD,"\t[%d]%s %d) found singleton %d (n=%d)\n",mype,__FUNCT__,iter,lid+my0,nn); */ src/mat/coarsen/impls/hem/hem.c: if(verbose>2) PetscPrintf(PETSC_COMM_WORLD,"\t[%d]%s %d.%d: %d active edges.\n", src/mat/impls/aij/mpi/pastix/pastix.c: PASTIX_CHECKMATRIX(MPI_COMM_WORLD,verb,((isSym != 0) ? API_SYM_YES : API_SYM_NO),API_YES,*n,&tmpcolptr,&tmprows,(PastixScalar**)&tmpvalues,NULL,1); src/mat/impls/aij/mpi/mpiaij.c:#define CHKERRQ(ierr) CHKERRABORT(PETSC_COMM_WORLD,ierr) src/mat/impls/dense/mpi/mpidense.c: PLAPACK is initialized (that is the initial matrices created) are on subcommunicators of MPI_COMM_WORLD, these subcommunicators src/mat/impls/fft/fftw/fftw.c: SETERRQ(PETSC_COMM_WORLD,PETSC_ERR_SUP,"not support for real numbers yet"); src/mat/impls/fft/fftw/fftw.c: SETERRQ(PETSC_COMM_WORLD,PETSC_ERR_SUP,"not support for real numbers yet"); src/mat/impls/fft/fftw/fftw.c: ierr = MPI_Comm_size(PETSC_COMM_WORLD, &size);CHKERRQ(ierr); src/mat/impls/fft/fftw/fftw.c: ierr = MPI_Comm_rank(PETSC_COMM_WORLD, &rank);CHKERRQ(ierr); src/mat/impls/fft/fftw/fftw.c: SETERRQ(PETSC_COMM_WORLD,PETSC_ERR_SUP,"FFTW does not allow parallel real 1D transform"); src/mat/impls/fft/fftw/fftw.c: ierr = VecCreateMPIWithArray(PETSC_COMM_WORLD,(PetscInt)n1,N1,(PetscScalar*)data_finr,fin);CHKERRQ(ierr); src/mat/impls/fft/fftw/fftw.c: ierr = VecCreateMPIWithArray(PETSC_COMM_WORLD,(PetscInt)n1,N1,(PetscScalar*)data_boutr,bout);CHKERRQ(ierr); src/mat/impls/fft/fftw/fftw.c: ierr = VecCreateMPIWithArray(PETSC_COMM_WORLD,(PetscInt)n1,N1,(PetscScalar*)data_fout,fout);CHKERRQ(ierr); src/mat/impls/fft/fftw/fftw.c: ierr = VecCreateMPIWithArray(PETSC_COMM_WORLD,(PetscInt)n1,N1,(PetscScalar*)data_finr,fin);CHKERRQ(ierr); src/mat/impls/fft/fftw/fftw.c: ierr = VecCreateMPIWithArray(PETSC_COMM_WORLD,(PetscInt)n1,N1,(PetscScalar*)data_boutr,bout);CHKERRQ(ierr); src/mat/impls/fft/fftw/fftw.c: ierr = VecCreateMPIWithArray(PETSC_COMM_WORLD,n1,N1,(PetscScalar*)data_fout,fout);CHKERRQ(ierr); src/mat/impls/fft/fftw/fftw.c: ierr = VecCreateMPIWithArray(PETSC_COMM_WORLD,(PetscInt)n,N1,(PetscScalar*)data_finr,fin);CHKERRQ(ierr); src/mat/impls/fft/fftw/fftw.c: ierr = VecCreateMPIWithArray(PETSC_COMM_WORLD,(PetscInt)n,N1,(PetscScalar*)data_boutr,bout);CHKERRQ(ierr); src/mat/impls/fft/fftw/fftw.c: ierr = VecCreateMPIWithArray(PETSC_COMM_WORLD,n,N1,(PetscScalar*)data_fout,fout);CHKERRQ(ierr); src/mat/impls/fft/fftw/fftw.c: ierr = ISCreateStride(PETSC_COMM_WORLD,local_n0,local_0_start,1,&list1); src/mat/impls/fft/fftw/fftw.c: ierr = ISCreateStride(PETSC_COMM_WORLD,local_n0,low,1,&list2); src/mat/impls/fft/fftw/fftw.c: ierr = ISCreateStride(PETSC_COMM_WORLD,local_n0*dim[1],local_0_start*dim[1],1,&list1); src/mat/impls/fft/fftw/fftw.c: ierr = ISCreateStride(PETSC_COMM_WORLD,local_n0*dim[1],low,1,&list2); src/mat/impls/fft/fftw/fftw.c: ierr = ISCreateStride(PETSC_COMM_WORLD,local_n0*dim[1]*dim[2],local_0_start*dim[1]*dim[2],1,&list1); src/mat/impls/fft/fftw/fftw.c: ierr = ISCreateStride(PETSC_COMM_WORLD,local_n0*dim[1]*dim[2],low,1,&list2); src/mat/impls/fft/fftw/fftw.c: ierr = ISCreateStride(PETSC_COMM_WORLD,local_n0*(fftw->partial_dim),local_0_start*(fftw->partial_dim),1,&list1); src/mat/impls/fft/fftw/fftw.c: ierr = ISCreateStride(PETSC_COMM_WORLD,local_n0*(fftw->partial_dim),low,1,&list2); src/mat/impls/fft/fftw/fftw.c: ierr = ISCreateStride(PETSC_COMM_WORLD,local_n1,local_1_start,1,&list1); src/mat/impls/fft/fftw/fftw.c: ierr = ISCreateStride(PETSC_COMM_WORLD,local_n1,low,1,&list2); src/mat/impls/fft/fftw/fftw.c: ierr = ISCreateStride(PETSC_COMM_WORLD,local_n0*dim[1],local_0_start*dim[1],1,&list1); src/mat/impls/fft/fftw/fftw.c: ierr = ISCreateStride(PETSC_COMM_WORLD,local_n0*dim[1],low,1,&list2); src/mat/impls/fft/fftw/fftw.c: ierr = ISCreateStride(PETSC_COMM_WORLD,local_n0*dim[1]*dim[2],local_0_start*dim[1]*dim[2],1,&list1); src/mat/impls/fft/fftw/fftw.c: ierr = ISCreateStride(PETSC_COMM_WORLD,local_n0*dim[1]*dim[2],low,1,&list2); src/mat/impls/fft/fftw/fftw.c: ierr = ISCreateStride(PETSC_COMM_WORLD,local_n0*(fftw->partial_dim),local_0_start*(fftw->partial_dim),1,&list1); src/mat/impls/fft/fftw/fftw.c: ierr = ISCreateStride(PETSC_COMM_WORLD,local_n0*(fftw->partial_dim),low,1,&list2); src/mat/impls/fft/fftw/fftw.c: alloc_local = fftw_mpi_local_size_2d_transposed(dim[0],dim[1]/2+1,PETSC_COMM_WORLD,&local_n0,&local_0_start,&local_n1,&local_1_start); src/mat/impls/fft/fftw/fftw.c: alloc_local = fftw_mpi_local_size_3d_transposed(dim[0],dim[1],dim[2]/2+1,PETSC_COMM_WORLD,&local_n0,&local_0_start,&local_n1,&local_1_start); src/mat/impls/fft/fftw/fftw.c: alloc_local = fftw_mpi_local_size_transposed(ndim,pdim,PETSC_COMM_WORLD,&local_n0,&local_0_start,&local_n1,&local_1_start); src/snes/impls/vi/rs/virs.c: ierr = ISCreateGeneral(PETSC_COMM_WORLD,cnt,coarseindex,PETSC_OWN_POINTER,&inactive);CHKERRQ(ierr); src/snes/impls/vi/rs/virs.c: ierr = ISCreateGeneral(PETSC_COMM_WORLD,cnt,nrows,PETSC_OWN_POINTER,&IS_inact);CHKERRQ(ierr); src/snes/impls/vi/rs/virs.c: ierr = VecView(vi->xu,PETSC_VIEWER_BINARY_(PETSC_COMM_WORLD));CHKERRQ(ierr); src/snes/impls/vi/rs/virs.c: ierr = VecView(vi->xl,PETSC_VIEWER_BINARY_(PETSC_COMM_WORLD));CHKERRQ(ierr); src/snes/impls/vi/rs/virs.c: ierr = VecView(X,PETSC_VIEWER_BINARY_(PETSC_COMM_WORLD));CHKERRQ(ierr); src/snes/impls/vi/rs/virs.c: ierr = VecView(F,PETSC_VIEWER_BINARY_(PETSC_COMM_WORLD));CHKERRQ(ierr); src/snes/impls/vi/rs/virs.c: ierr = ISView(IS_inact,PETSC_VIEWER_BINARY_(PETSC_COMM_WORLD));CHKERRQ(ierr); src/snes/impls/vi/rsaug/virsaug.c: ierr = ISCreateGeneral(PETSC_COMM_WORLD,cnt,coarseindex,PETSC_OWN_POINTER,&inactive);CHKERRQ(ierr); src/snes/impls/vi/rsaug/virsaug.c: ierr = ISCreateGeneral(PETSC_COMM_WORLD,cnt,nrows,PETSC_OWN_POINTER,&IS_inact);CHKERRQ(ierr); src/snes/impls/vi/rsaug/virsaug.c: ierr = VecView(vi->xu,PETSC_VIEWER_BINARY_(PETSC_COMM_WORLD));CHKERRQ(ierr); src/snes/impls/vi/rsaug/virsaug.c: ierr = VecView(vi->xl,PETSC_VIEWER_BINARY_(PETSC_COMM_WORLD));CHKERRQ(ierr); src/snes/impls/vi/rsaug/virsaug.c: ierr = VecView(X,PETSC_VIEWER_BINARY_(PETSC_COMM_WORLD));CHKERRQ(ierr); src/snes/impls/vi/rsaug/virsaug.c: ierr = VecView(F,PETSC_VIEWER_BINARY_(PETSC_COMM_WORLD));CHKERRQ(ierr); src/snes/impls/vi/rsaug/virsaug.c: ierr = ISView(IS_inact,PETSC_VIEWER_BINARY_(PETSC_COMM_WORLD));CHKERRQ(ierr); src/snes/interface/snes.c: ierr = PetscPrintf(PETSC_COMM_WORLD,"Testing FormFunction() for local min\n"); src/snes/interface/snes.c: ierr = PetscPrintf(PETSC_COMM_WORLD,"i = %D\n",i);CHKERRQ(ierr); src/snes/interface/snes.c: ierr = PetscPrintf(PETSC_COMM_WORLD," j norm %D %18.16e\n",j,norm);CHKERRQ(ierr); src/snes/utils/adutils/damgsnesad.c: ierr = PetscPrintf(PETSC_COMM_WORLD,"I am here");CHKERRQ(ierr); src/snes/utils/adutils/damgsnesad.c: } ierr = PetscPrintf(PETSC_COMM_WORLD,"I am here");CHKERRQ(ierr); src/snes/utils/adutils/damgsnesad.c: ierr = PetscPrintf(PETSC_COMM_WORLD,"iter=%d",i+1);CHKERRQ(ierr); src/snes/utils/adutils/damgsnesad.c: ierr = PetscPrintf(PETSC_COMM_WORLD,"step=%g,oldnorm=%g,norm=%g ",ynorm,fnorm,gnorm);CHKERRQ(ierr); src/snes/utils/adutils/damgsnesad.c: ierr = PetscPrintf(PETSC_COMM_WORLD,"betaHS=%g,betaDY=%g\n",betaHS,betaDY);CHKERRQ(ierr); src/snes/utils/adutils/damgsnesad.c: ierr = PetscPrintf(PETSC_COMM_WORLD,"beta=%g,c_2=%g\n",beta,fabs(gnorm/ynorm));CHKERRQ(ierr); src/snes/utils/adutils/damgsnesad.c: ierr = PetscPrintf(PETSC_COMM_WORLD,"size=%g\n",fabs(gnorm/(beta*ynorm)));CHKERRQ(ierr); src/snes/utils/adutils/damgsnesad.c: ierr = PetscPrintf(PETSC_COMM_WORLD,"reason=%d\n",snes->reason);CHKERRQ(ierr); src/snes/utils/adutils/damgsnesad.c: ierr = PetscPrintf(PETSC_COMM_WORLD,"\n k=%d,iter=%d fmin=%g ",k,iter++,sqrt(fminnorm));CHKERRQ(ierr); src/snes/utils/adutils/damgsnesad.c: //ierr = PetscPrintf(PETSC_COMM_WORLD,"fmin=%g",sqrt(fminnorm));CHKERRQ(ierr); src/snes/utils/adutils/damgsnesad.c: //ierr = PetscPrintf(PETSC_COMM_WORLD,"Crite A\n");CHKERRQ(ierr); src/snes/utils/adutils/damgsnesad.c: // ierr = PetscPrintf(PETSC_COMM_WORLD,"Select A\n");CHKERRQ(ierr); src/snes/utils/adutils/damgsnesad.c: ierr = PetscPrintf(PETSC_COMM_WORLD,"restart for C ");CHKERRQ(ierr); src/snes/utils/adutils/damgsnesad.c: ierr = PetscPrintf(PETSC_COMM_WORLD,"restart for D ");CHKERRQ(ierr); src/snes/utils/adutils/damgsnesad.c: ierr = PetscPrintf(PETSC_COMM_WORLD,"k=%d, fnorm=%g\n", src/ts/impls/explicit/rk/rk.c: /* ierr = PetscPrintf(PETSC_COMM_WORLD,"norm err: %f maxerror: %f dt: %f",norm,rk->maxerror,ts->time_step);CHKERRQ(ierr); */ src/ts/impls/explicit/rk/rk.c: /*ierr = PetscPrintf(PETSC_COMM_WORLD,"changing fac %f\n",fac);*/ src/ts/impls/explicit/rk/rk.c: ierr = PetscPrintf(PETSC_COMM_WORLD,"Very small steps: %f\n",ts->time_step);CHKERRQ(ierr); src/vec/is/utils/iscoloring.c: communicator (e.g., PETSC_COMM_WORLD). If the index sets were created src/vec/vec/impls/mpi/pvec2.c: MPI_Abort(MPI_COMM_WORLD,1); src/vec/vec/impls/mpi/pvec2.c: MPI_Abort(MPI_COMM_WORLD,1); src/vec/vec/interface/vector.c: PetscRandomCreate(PETSC_COMM_WORLD,&rctx); src/vec/vec/utils/comb.c: MPI_Abort(MPI_COMM_WORLD,1); src/vec/vec/utils/comb.c: MPI_Abort(MPI_COMM_WORLD,1);

1 0

Re: [petsc-dev] petsc release soon
by Jed Brown 23 Mar '12

23 Mar '12

On Mon, Mar 12, 2012 at 14:18, Barry Smith <bsmith(a)mcs.anl.gov> wrote: > Hence I like the petsc-private. Jed gets what he wants and I get what I > want. Plus it doesn't break the current PETSc application code. > Pushed: http://petsc.cs.iit.edu/petsc/petsc-dev/rev/9a1730720b0d Hg tracks the rename, so merges should go smoothly.

1 0

Re: [petsc-dev] Reminder: *every* symbol with extern linkage *must* be namespaced
by Jed Brown 23 Mar '12

23 Mar '12

On Sat, Mar 17, 2012 at 15:32, Mark F. Adams <mark.adams(a)columbia.edu>wrote: > done, but it won't show up until I check in my current repo, which could > be a week. > http://petsc.cs.iit.edu/petsc/petsc-dev/rev/597bc483f35d Mark, this patch adds types and symbols that are not namespaced. Please pull before fixing, I just renamed all the private headers to include/petsc-private/.

1 0