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

Re: [petsc-dev] threadcomm memory leak
by Barry Smith 16 Jul '12

16 Jul '12

On Jul 16, 2012, at 11:25 AM, Satish Balay wrote: > Thats was almost 15 years ago :) from the time we used RCS - and > didn't maintain commit messages. > > I think its probably more because mpiuni doesn't support more than 1 > communicator and attributes are stashed in a global structure [not per > communicator]. > Why shouldn't they be the same? Everything should still work fine. Where in PETSc do we make the assumption MPI_COMM_SELF is not MPI_COMM_WORLD and why? Except for Shi's new stuff and the new elemental stuff everything else in PETSc has worked fine with them being the same for 15 years. Hence the problem is likely with Shri's and Hong's interactions with MPI; not with the current MPIUNI. Barry > Perhaps mpiuni needs fixing.. > > Satish > > > On Mon, 16 Jul 2012, Jed Brown wrote: > >> Satish, what was the rationale behind this commit? I can't think of an >> excuse for making them be literally the same. >> >> On Mon, Jul 16, 2012 at 11:11 AM, Satish Balay <balay(a)mcs.anl.gov> wrote: >> >>> Actually for np=1 MPI_COMM_SELF==MPI_COMM_WORLD. >>> >>> mpi.h was changed to reflect this - but mpif.h was not. >>> >>> Satish >>> >>> ---------- >>> balay@petsc:~/Petsc-historic/src/sys/src/mpiuni $ rcsdiff -r1.52 -r1.53 >>> mpi.h >>> =================================================================== >>> RCS file: RCS/mpi.h,v >>> retrieving revision 1.52 >>> retrieving revision 1.53 >>> diff -r1.52 -r1.53 >>> 1c1 >>> < /* $Id: mpi.h,v 1.52 1997/07/28 21:41:02 balay Exp $ */ >>> --- >>>> /* $Id: mpi.h,v 1.53 1997/08/21 01:06:19 balay Exp $ */ >>> 22c22 >>> < #define MPI_COMM_SELF 2 >>> --- >>>> #define MPI_COMM_SELF MPI_COMM_WORLD >>> 427c427 >>> < (MPIUNI_TMP = (void *) (comm), \ >>> --- >>>> (*(newcomm) = (comm), \ >>> 429d428 >>> < MPIUNI_TMP = (void *) (newcomm), \ >>> >> >

1 0

MatNullSpace with fortran data types
by Blaise Bourdin 16 Jul '12

16 Jul '12

Hi, I am having trouble with MatNullSPace in fortran with fortran data types enabled. I don't mind if MatNullSpace is not a f90 data type, but right now, neither the old style MatNullSpace nsp nor the more f90-ish type(MatNullSpace) nsp seem to work. galerkin:PetscF90 bourdin$ make -f makefile.MatNullSpace MatNullSpace /opt/HPC/mpich2-1.4.1p1-intel11.1/bin/mpif90 -c -fPIC -g -I/opt/HPC/petsc-3.3/include -I/opt/HPC/petsc-3.3/Darwin-intel11.1-mef90-g/include -I/opt/HPC/petsc-3.3/include/sieve -I/opt/HPC/petsc-3.3/Darwin-intel11.1-mef90-g/cbind/include -I/opt/HPC/petsc-3.3/Darwin-intel11.1-mef90-g/forbind/include -I/opt/HPC/mpich2-1.4.1p1-intel11.1/include -o MatNullSpace.o MatNullSpace.F90 MatNullSpace.F90(18): error #6457: This derived type name has not been declared. [MATNULLSPACE] type(MatNullSpace) nsp -----------^ MatNullSpace.F90(34): error #6404: This name does not have a type, and must have an explicit type. [NSP] call MatNullSpaceCreate(PETSC_COMM_WORLD,PETSC_TRUE,0,PETSC_NULL_OBJECT,nsp,ierr) ------------------------------------------------------------------------------^ compilation aborted for MatNullSpace.F90 (code 1) make: [MatNullSpace.o] Error 1 (ignored) or galerkin:PetscF90 bourdin$ make -f makefile.MatNullSpace MatNullSpace /opt/HPC/mpich2-1.4.1p1-intel11.1/bin/mpif90 -c -fPIC -g -I/opt/HPC/petsc-3.3/include -I/opt/HPC/petsc-3.3/Darwin-intel11.1-mef90-g/include -I/opt/HPC/petsc-3.3/include/sieve -I/opt/HPC/petsc-3.3/Darwin-intel11.1-mef90-g/cbind/include -I/opt/HPC/petsc-3.3/Darwin-intel11.1-mef90-g/forbind/include -I/opt/HPC/mpich2-1.4.1p1-intel11.1/include -o MatNullSpace.o MatNullSpace.F90 MatNullSpace.F90(19): error #5082: Syntax error, found IDENTIFIER 'NSP' when expecting one of: ( % : . = => MatNullSpace nsp Peeking in include/finclude, it looks like the matching fortran data type has not been declared. I added it (and that of MatFDColoring) to include/finclude/ftn-customn/petscmatdef.h90, but I still get an error when trying to compile a trivial example. What am I missing? I am attaching an simple example and the incomplete patch I though would be enough to define the fortrasn data type for MatNullSpace. Blaise -- Department of Mathematics and Center for Computation & Technology Louisiana State University, Baton Rouge, LA 70803, USA Tel. +1 (225) 578 1612, Fax +1 (225) 578 4276 http://www.math.lsu.edu/~bourdin

1 0

Re: [petsc-dev] threadcomm memory leak
by Satish Balay 16 Jul '12

16 Jul '12

Thats was almost 15 years ago :) from the time we used RCS - and didn't maintain commit messages. I think its probably more because mpiuni doesn't support more than 1 communicator and attributes are stashed in a global structure [not per communicator]. Perhaps mpiuni needs fixing.. Satish On Mon, 16 Jul 2012, Jed Brown wrote: > Satish, what was the rationale behind this commit? I can't think of an > excuse for making them be literally the same. > > On Mon, Jul 16, 2012 at 11:11 AM, Satish Balay <balay(a)mcs.anl.gov> wrote: > > > Actually for np=1 MPI_COMM_SELF==MPI_COMM_WORLD. > > > > mpi.h was changed to reflect this - but mpif.h was not. > > > > Satish > > > > ---------- > > balay@petsc:~/Petsc-historic/src/sys/src/mpiuni $ rcsdiff -r1.52 -r1.53 > > mpi.h > > =================================================================== > > RCS file: RCS/mpi.h,v > > retrieving revision 1.52 > > retrieving revision 1.53 > > diff -r1.52 -r1.53 > > 1c1 > > < /* $Id: mpi.h,v 1.52 1997/07/28 21:41:02 balay Exp $ */ > > --- > > > /* $Id: mpi.h,v 1.53 1997/08/21 01:06:19 balay Exp $ */ > > 22c22 > > < #define MPI_COMM_SELF 2 > > --- > > > #define MPI_COMM_SELF MPI_COMM_WORLD > > 427c427 > > < (MPIUNI_TMP = (void *) (comm), \ > > --- > > > (*(newcomm) = (comm), \ > > 429d428 > > < MPIUNI_TMP = (void *) (newcomm), \ > > >

1 0

Re: [petsc-dev] threadcomm memory leak
by Jed Brown 16 Jul '12

16 Jul '12

Satish, what was the rationale behind this commit? I can't think of an excuse for making them be literally the same. On Mon, Jul 16, 2012 at 11:11 AM, Satish Balay <balay(a)mcs.anl.gov> wrote: > Actually for np=1 MPI_COMM_SELF==MPI_COMM_WORLD. > > mpi.h was changed to reflect this - but mpif.h was not. > > Satish > > ---------- > balay@petsc:~/Petsc-historic/src/sys/src/mpiuni $ rcsdiff -r1.52 -r1.53 > mpi.h > =================================================================== > RCS file: RCS/mpi.h,v > retrieving revision 1.52 > retrieving revision 1.53 > diff -r1.52 -r1.53 > 1c1 > < /* $Id: mpi.h,v 1.52 1997/07/28 21:41:02 balay Exp $ */ > --- > > /* $Id: mpi.h,v 1.53 1997/08/21 01:06:19 balay Exp $ */ > 22c22 > < #define MPI_COMM_SELF 2 > --- > > #define MPI_COMM_SELF MPI_COMM_WORLD > 427c427 > < (MPIUNI_TMP = (void *) (comm), \ > --- > > (*(newcomm) = (comm), \ > 429d428 > < MPIUNI_TMP = (void *) (newcomm), \ >

1 0

Re: [petsc-dev] threadcomm memory leak
by Satish Balay 16 Jul '12

16 Jul '12

On Mon, 16 Jul 2012, Blaise Bourdin wrote: > > On Jul 16, 2012, at 10:11 AM, Jed Brown wrote: > > > On Mon, Jul 16, 2012 at 10:09 AM, Shri <abhyshr(a)mcs.anl.gov> wrote: > > Yes, i think so too. If no one objects then i can do that. > > Btw, MPI_COMM_SELF and MPI_COMM_WORLD are same in mpiuni/mpi.h > > #define MPI_COMM_WORLD 1 > > #define MPI_COMM_SELF MPI_COMM_WORLD > > > > while the fortran version declares these communicators as different. mpiuni/mpif.h > > integer MPI_COMM_WORLD > > parameter (MPI_COMM_WORLD = 1) > > integer MPI_COMM_SELF > > parameter (MPI_COMM_SELF = 2) > > > > WAT? The Fortran version did something right? > > That's because it is tested by people who do stuff with it ;) Actually for np=1 MPI_COMM_SELF==MPI_COMM_WORLD. mpi.h was changed to reflect this - but mpif.h was not. Satish ---------- balay@petsc:~/Petsc-historic/src/sys/src/mpiuni $ rcsdiff -r1.52 -r1.53 mpi.h =================================================================== RCS file: RCS/mpi.h,v retrieving revision 1.52 retrieving revision 1.53 diff -r1.52 -r1.53 1c1 < /* $Id: mpi.h,v 1.52 1997/07/28 21:41:02 balay Exp $ */ --- > /* $Id: mpi.h,v 1.53 1997/08/21 01:06:19 balay Exp $ */ 22c22 < #define MPI_COMM_SELF 2 --- > #define MPI_COMM_SELF MPI_COMM_WORLD 427c427 < (MPIUNI_TMP = (void *) (comm), \ --- > (*(newcomm) = (comm), \ 429d428 < MPIUNI_TMP = (void *) (newcomm), \ balay@petsc:~/Petsc-historic/src/sys/src/mpiuni $

1 0

Re: [petsc-dev] threadcomm memory leak
by Blaise Bourdin 16 Jul '12

16 Jul '12

On Jul 16, 2012, at 10:11 AM, Jed Brown wrote: > On Mon, Jul 16, 2012 at 10:09 AM, Shri <abhyshr(a)mcs.anl.gov> wrote: > Yes, i think so too. If no one objects then i can do that. > Btw, MPI_COMM_SELF and MPI_COMM_WORLD are same in mpiuni/mpi.h > #define MPI_COMM_WORLD 1 > #define MPI_COMM_SELF MPI_COMM_WORLD > > while the fortran version declares these communicators as different. mpiuni/mpif.h > integer MPI_COMM_WORLD > parameter (MPI_COMM_WORLD = 1) > integer MPI_COMM_SELF > parameter (MPI_COMM_SELF = 2) > > WAT? The Fortran version did something right? That's because it is tested by people who do stuff with it ;) Blaise -- Department of Mathematics and Center for Computation & Technology Louisiana State University, Baton Rouge, LA 70803, USA Tel. +1 (225) 578 1612, Fax +1 (225) 578 4276 http://www.math.lsu.edu/~bourdin

1 0

Re: [petsc-dev] threadcomm memory leak
by Jed Brown 16 Jul '12

16 Jul '12

On Mon, Jul 16, 2012 at 10:09 AM, Shri <abhyshr(a)mcs.anl.gov> wrote: > Yes, i think so too. If no one objects then i can do that. > Btw, MPI_COMM_SELF and MPI_COMM_WORLD are same in mpiuni/mpi.h > #define MPI_COMM_WORLD 1 > #define MPI_COMM_SELF MPI_COMM_WORLD > > while the fortran version declares these communicators as different. > mpiuni/mpif.h > integer MPI_COMM_WORLD > parameter (MPI_COMM_WORLD = 1) > integer MPI_COMM_SELF > parameter (MPI_COMM_SELF = 2) > WAT? The Fortran version did something right?

1 0

Re: [petsc-dev] threadcomm memory leak
by Shri 16 Jul '12

16 Jul '12

On Jul 16, 2012, at 9:46 AM, Jed Brown wrote: > On Mon, Jul 16, 2012 at 9:44 AM, Shri <abhyshr(a)mcs.anl.gov> wrote: > You need to do similar for your elemental grid to avoid memory leaks when using mpiuni. > http://petsc.cs.iit.edu/petsc/petsc-dev/rev/169c19e9ea54 > > Isn't the right solution to fix MPIUNI so that they aren't literally the same comm? Yes, i think so too. If no one objects then i can do that. Btw, MPI_COMM_SELF and MPI_COMM_WORLD are same in mpiuni/mpi.h #define MPI_COMM_WORLD 1 #define MPI_COMM_SELF MPI_COMM_WORLD while the fortran version declares these communicators as different. mpiuni/mpif.h integer MPI_COMM_WORLD parameter (MPI_COMM_WORLD = 1) integer MPI_COMM_SELF parameter (MPI_COMM_SELF = 2)

1 0

Re: [petsc-dev] threadcomm memory leak
by Jed Brown 16 Jul '12

16 Jul '12

On Mon, Jul 16, 2012 at 9:44 AM, Shri <abhyshr(a)mcs.anl.gov> wrote: > You need to do similar for your elemental grid to avoid memory leaks > when using mpiuni. > http://petsc.cs.iit.edu/petsc/petsc-dev/rev/169c19e9ea54 > Isn't the right solution to fix MPIUNI so that they aren't literally the same comm?

1 0

Re: [petsc-dev] threadcomm memory leak
by Shri 16 Jul '12

16 Jul '12

Hong, You need to do similar for your elemental grid to avoid memory leaks when using mpiuni. http://petsc.cs.iit.edu/petsc/petsc-dev/rev/169c19e9ea54 Shri On Jul 16, 2012, at 8:08 AM, Shri wrote: > > On Jul 15, 2012, at 5:26 PM, Shri wrote: > >> >> On Jul 15, 2012, at 4:57 PM, Barry Smith wrote: >> >>> >>> On Jul 15, 2012, at 4:48 PM, Shri wrote: >>> >>>> >>>> On Jul 15, 2012, at 3:27 PM, Barry Smith wrote: >>>> >>>>> >>>>> Shri, >>>>> >>>>> You need to test your code WITHOUT pthreads OR openmp configure. >>>> That is what i am doing, no openmp or pthreads. >>>> >>>>> You have the call to PetscThreadCommFinalize() protected by PTHREADCLASSES >>>> >>>> There is no PetscThreadCommFinalize(). There is a 'PetscThreadsFinalize()' call in PetscFinalize() that is protected by PTHREADCLASSES. PetscThreadsFinalize() is for finalizing the previous pthread pool model that i had added on which vec_type pthread and mat_type seqaijpthread are based. It has nothing to do with threadcomm. Once i add new vector and matrix code using threadcomm , i'll remove the old pthread code and the pthread vector and matrix class. >>> >>> Why not just remove all the old pthread code now; it will make things clearer. >> >> Don't want to hurl people who are using/testing it into darkness. >> >>> >>>>> but not PetscThreadCommInitializePackage() projected! Also why do you call PetscThreadCommPackageInitialize() at the beginning but PetscThreadCommFinalize() at the end and not PetscThreadCommFinalizePackage()? >>>> >>>> PetscThreadCommFinalizePackage is registerfinalized in PetscThreadCommInitializePackage() >>>> >>>> ierr = PetscRegisterFinalize(PetscThreadCommFinalizePackage);CHKERRQ(ierr); >>> >>> Sorry my mistake. >>>> >>>> So it is called on PetscRegisterFinalizeAll(). >>>>> >>>>> BTW: If I #if def out the PetscThreadCommPackageInitialize() then nothing works (I have built without pthread classes and without openmp). So there is some bad stuff going on. >>>> >>>> PetscThreadCommPackageInitialize() attaches the threadcomm attribute on the MPI communicator and there are several places where the threadcomm attribute is requested (sys/objects/subcomm.c, src/sys/tagm.c). Hence, the errors. >>> >>> Try a build with mpiuni. >> >> Aah, i can see the memory leak now. Thanks. It looks like the threadcomm object is not being freed. Something to do with destroying the MPIUNI attribute perhaps? Will check. > > Fixed. http://petsc.cs.iit.edu/petsc/petsc-dev/rev/169c19e9ea54 > The memory leak was due to the COMM_SELF and COMM_WORLD communicators being same for mpiuni. This was leaving an extra reference for the innercomm. > > Shri > >> >> Shri >>> Maybe the memory leak is because that isn't handling stuff possible. My build with memory leaks is with mpiuni. >>> >>> Thanks >>> >>> >>> Barry >>> >>>> >>>> Thanks, >>>> Shri >>>>> >>>>> Barry >>>>> >>>>> On Jul 15, 2012, at 11:54 AM, Shri wrote: >>>>> >>>>>> >>>>>> On Jul 14, 2012, at 7:55 PM, Satish Balay wrote: >>>>>> >>>>>>> I see this with 32bit linux build - but not 64bit linux build. >>>>>> >>>>>> I tried a fresh clone on my Mac and a linux cluster (both x86_64) but still did not get any memory leaks. Perhaps this is a 32 bit build issue as Satish mentions. What configure options do i need to have a 32 bit build on x86_64? CC="gcc -m32" FC="gfortran -m32"? >>>>>> >>>>>> Thanks, >>>>>> Shri >>>>>> >>>>>>> >>>>>>> also 32bit linux build gives errors with -fp_trap in ddot(). Haven't checked why. >>>>>>> >>>>>>> Satish >>>>>>> >>>>>>> >>>>>>> On Sat, 14 Jul 2012, Barry Smith wrote: >>>>>>> >>>>>>>> >>>>>>>> Many of them with make test >>>>>>>> >>>>>>>> Running test examples to verify correct installation >>>>>>>> Using PETSC_DIR=/Users/barrysmith/Src/petsc-dev and PETSC_ARCH=arch-float128-opt >>>>>>>> Possible error running C/C++ src/snes/examples/tutorials/ex19 with 1 MPI process >>>>>>>> See http://www.mcs.anl.gov/petsc/documentation/faq.html >>>>>>>> lid velocity = Qg, prandtl # = Qg, grashof # = Qg >>>>>>>> Number of SNES iterations = 2 >>>>>>>> [0]Total space allocated 6880 bytes >>>>>>>> [ 0]16 bytes PetscThreadCommReductionCreate() line 260 in /Users/barrysmith/Src/petsc-dev/src/sys/threadcomm/interface/threadcommred.c >>>>>>>> [ 0]16 bytes PetscThreadCommReductionCreate() line 259 in /Users/barrysmith/Src/petsc-dev/src/sys/threadcomm/interface/threadcommred.c >>>>>>>> [ 0]48 bytes PetscThreadCommReductionCreate() line 258 in /Users/barrysmith/Src/petsc-dev/src/sys/threadcomm/interface/threadcommred.c >>>>>>>> [ 0]16 bytes PetscThreadCommInitialize() line 698 in /Users/barrysmith/Src/petsc-dev/src/sys/threadcomm/interface/threadcomm.c >>>>>>>> [ 0]176 bytes PetscThreadCommInitialize() line 694 in /Users/barrysmith/Src/petsc-dev/src/sys/threadcomm/interface/threadcomm.c >>>>>>>> [ 0]16 bytes PetscThreadCommInitialize() line 698 in /Users/barrysmith/Src/petsc-dev/src/sys/threadcomm/interface/threadcomm.c >>>>>>>> [ 0]176 bytes PetscThreadCommInitialize() line 694 in /Users/barrysmith/Src/petsc-dev/src/sys/threadcomm/interface/threadcomm.c >>>>>>>> [ 0]16 bytes PetscThreadCommInitialize() line 698 in /Users/barrysmith/Src/petsc-dev/src/sys/threadcomm/interface/threadcomm.c >>>>>>>> [ 0]176 bytes PetscThreadCommInitialize() line 694 in /Users/barrysmith/Src/petsc-dev/src/sys/threadcomm/interface/threadcomm.c >>>>>>>> [ 0]16 bytes PetscThreadCommInitialize() line 698 in /Users/barrysmith/Src/petsc-dev/src/sys/threadcomm/interface/threadcomm.c >>>>>>>> [ 0]176 bytes PetscThreadCommInitialize() line 694 in /Users/barrysmith/Src/petsc-dev/src/sys/threadcomm/interface/threadcomm.c >>>>>>>> [ 0]16 bytes PetscThreadCommInitialize() line 698 in /Users/barrysmith/Src/petsc-dev/src/sys/threadcomm/interface/threadcomm.c >>>>>>>> [ 0]176 bytes PetscThreadCommInitialize() line 694 in /Users/barrysmith/Src/petsc-dev/src/sys/threadcomm/interface/threadcomm.c >>>>>>>> [ 0]16 bytes PetscThreadCommInitialize() line 698 in /Users/barrysmith/Src/petsc-dev/src/sys/threadcomm/interface/threadcomm.c >>>>>>>> [ 0]176 bytes PetscThreadCommInitialize() line 694 in /Users/barrysmith/Src/petsc-dev/src/sys/threadcomm/interface/threadcomm.c >>>>>>>> [ 0]16 bytes PetscThreadCommInitialize() line 698 in /Users/barrysmith/Src/petsc-dev/src/sys/threadcomm/interface/threadcomm.c >>>>>>>> [ 0]176 bytes PetscThreadCommInitialize() line 694 in /Users/barrysmith/Src/petsc-dev/src/sys/threadcomm/interface/threadcomm.c >>>>>>>> [ 0]16 bytes PetscThreadCommInitialize() line 698 in /Users/barrysmith/Src/petsc-dev/src/sys/threadcomm/interface/threadcomm.c >>>>>>>> [ 0]176 bytes PetscThreadCommInitialize() line 694 in /Users/barrysmith/Src/petsc-dev/src/sys/threadcomm/interface/threadcomm.c >>>>>>>> [ 0]16 bytes PetscThreadCommInitialize() line 698 in /Users/barrysmith/Src/petsc-dev/src/sys/threadcomm/interface/threadcomm.c >>>>>>>> [ 0]176 bytes PetscThreadCommInitialize() line 694 in /Users/barrysmith/Src/petsc-dev/src/sys/threadcomm/interface/threadcomm.c >>>>>>>> [ 0]16 bytes PetscThreadCommInitialize() line 698 in /Users/barrysmith/Src/petsc-dev/src/sys/threadcomm/interface/threadcomm.c >>>>>>>> [ 0]176 bytes PetscThreadCommInitialize() line 694 in /Users/barrysmith/Src/petsc-dev/src/sys/threadcomm/interface/threadcomm.c >>>>>>>> [ 0]16 bytes PetscThreadCommInitialize() line 698 in /Users/barrysmith/Src/petsc-dev/src/sys/threadcomm/interface/threadcomm.c >>>>>>>> [ 0]176 bytes PetscThreadCommInitialize() line 694 in /Users/barrysmith/Src/petsc-dev/src/sys/threadcomm/interface/threadcomm.c >>>>>>>> [ 0]16 bytes PetscThreadCommInitialize() line 698 in /Users/barrysmith/Src/petsc-dev/src/sys/threadcomm/interface/threadcomm.c >>>>>>>> [ 0]176 bytes PetscThreadCommInitialize() line 694 in /Users/barrysmith/Src/petsc-dev/src/sys/threadcomm/interface/threadcomm.c >>>>>>>> [ 0]16 bytes PetscThreadCommInitialize() line 698 in /Users/barrysmith/Src/petsc-dev/src/sys/threadcomm/interface/threadcomm.c >>>>>>>> [ 0]176 bytes PetscThreadCommInitialize() line 694 in /Users/barrysmith/Src/petsc-dev/src/sys/threadcomm/interface/threadcomm.c >>>>>>>> [ 0]16 bytes PetscThreadCommInitialize() line 698 in /Users/barrysmith/Src/petsc-dev/src/sys/threadcomm/interface/threadcomm.c >>>>>>>> [ 0]176 bytes PetscThreadCommInitialize() line 694 in /Users/barrysmith/Src/petsc-dev/src/sys/threadcomm/interface/threadcomm.c >>>>>>>> [ 0]16 bytes PetscThreadCommInitialize() line 698 in /Users/barrysmith/Src/petsc-dev/src/sys/threadcomm/interface/threadcomm.c >>>>>>>> [ 0]176 bytes PetscThreadCommInitialize() line 694 in /Users/barrysmith/Src/petsc-dev/src/sys/threadcomm/interface/threadcomm.c >>>>>>>> [ 0]16 bytes PetscThreadCommInitialize() line 698 in /Users/barrysmith/Src/petsc-dev/src/sys/threadcomm/interface/threadcomm.c >>>>>>>> [ 0]176 bytes PetscThreadCommInitialize() line 694 in /Users/barrysmith/Src/petsc-dev/src/sys/threadcomm/interface/threadcomm.c >>>>>>>> [ 0]16 bytes PetscThreadCommInitialize() line 698 in /Users/barrysmith/Src/petsc-dev/src/sys/threadcomm/interface/threadcomm.c >>>>>>>> [ 0]176 bytes PetscThreadCommInitialize() line 694 in /Users/barrysmith/Src/petsc-dev/src/sys/threadcomm/interface/threadcomm.c >>>>>>>> [ 0]16 bytes PetscThreadCommInitialize() line 698 in /Users/barrysmith/Src/petsc-dev/src/sys/threadcomm/interface/threadcomm.c >>>>>>>> [ 0]176 bytes PetscThreadCommInitialize() line 694 in /Users/barrysmith/Src/petsc-dev/src/sys/threadcomm/interface/threadcomm.c >>>>>>>> [ 0]16 bytes PetscThreadCommInitialize() line 698 in /Users/barrysmith/Src/petsc-dev/src/sys/threadcomm/interface/threadcomm.c >>>>>>>> [ 0]176 bytes PetscThreadCommInitialize() line 694 in /Users/barrysmith/Src/petsc-dev/src/sys/threadcomm/interface/threadcomm.c >>>>>>>> [ 0]16 bytes PetscThreadCommInitialize() line 698 in /Users/barrysmith/Src/petsc-dev/src/sys/threadcomm/interface/threadcomm.c >>>>>>>> [ 0]176 bytes PetscThreadCommInitialize() line 694 in /Users/barrysmith/Src/petsc-dev/src/sys/threadcomm/interface/threadcomm.c >>>>>>>> [ 0]16 bytes PetscThreadCommInitialize() line 698 in /Users/barrysmith/Src/petsc-dev/src/sys/threadcomm/interface/threadcomm.c >>>>>>>> [ 0]176 bytes PetscThreadCommInitialize() line 694 in /Users/barrysmith/Src/petsc-dev/src/sys/threadcomm/interface/threadcomm.c >>>>>>>> [ 0]16 bytes PetscThreadCommInitialize() line 698 in /Users/barrysmith/Src/petsc-dev/src/sys/threadcomm/interface/threadcomm.c >>>>>>>> [ 0]176 bytes PetscThreadCommInitialize() line 694 in /Users/barrysmith/Src/petsc-dev/src/sys/threadcomm/interface/threadcomm.c >>>>>>>> [ 0]16 bytes PetscThreadCommInitialize() line 698 in /Users/barrysmith/Src/petsc-dev/src/sys/threadcomm/interface/threadcomm.c >>>>>>>> [ 0]176 bytes PetscThreadCommInitialize() line 694 in /Users/barrysmith/Src/petsc-dev/src/sys/threadcomm/interface/threadcomm.c >>>>>>>> [ 0]16 bytes PetscThreadCommInitialize() line 698 in /Users/barrysmith/Src/petsc-dev/src/sys/threadcomm/interface/threadcomm.c >>>>>>>> [ 0]176 bytes PetscThreadCommInitialize() line 694 in /Users/barrysmith/Src/petsc-dev/src/sys/threadcomm/interface/threadcomm.c >>>>>>>> [ 0]16 bytes PetscThreadCommInitialize() line 698 in /Users/barrysmith/Src/petsc-dev/src/sys/threadcomm/interface/threadcomm.c >>>>>>>> [ 0]176 bytes PetscThreadCommInitialize() line 694 in /Users/barrysmith/Src/petsc-dev/src/sys/threadcomm/interface/threadcomm.c >>>>>>>> [ 0]16 bytes PetscThreadCommInitialize() line 698 in /Users/barrysmith/Src/petsc-dev/src/sys/threadcomm/interface/threadcomm.c >>>>>>>> [ 0]176 bytes PetscThreadCommInitialize() line 694 in /Users/barrysmith/Src/petsc-dev/src/sys/threadcomm/interface/threadcomm.c >>>>>>>> [ 0]16 bytes PetscThreadCommInitialize() line 698 in /Users/barrysmith/Src/petsc-dev/src/sys/threadcomm/interface/threadcomm.c >>>>>>>> [ 0]176 bytes PetscThreadCommInitialize() line 694 in /Users/barrysmith/Src/petsc-dev/src/sys/threadcomm/interface/threadcomm.c >>>>>>>> [ 0]16 bytes PetscThreadCommInitialize() line 698 in /Users/barrysmith/Src/petsc-dev/src/sys/threadcomm/interface/threadcomm.c >>>>>>>> [ 0]176 bytes PetscThreadCommInitialize() line 694 in /Users/barrysmith/Src/petsc-dev/src/sys/threadcomm/interface/threadcomm.c >>>>>>>> [ 0]16 bytes PetscThreadCommInitialize() line 698 in /Users/barrysmith/Src/petsc-dev/src/sys/threadcomm/interface/threadcomm.c >>>>>>>> [ 0]176 bytes PetscThreadCommInitialize() line 694 in /Users/barrysmith/Src/petsc-dev/src/sys/threadcomm/interface/threadcomm.c >>>>>>>> [ 0]16 bytes PetscThreadCommInitialize() line 698 in /Users/barrysmith/Src/petsc-dev/src/sys/threadcomm/interface/threadcomm.c >>>>>>>> [ 0]176 bytes PetscThreadCommInitialize() line 694 in /Users/barrysmith/Src/petsc-dev/src/sys/threadcomm/interface/threadcomm.c >>>>>>>> [ 0]16 bytes PetscThreadCommInitialize() line 698 in /Users/barrysmith/Src/petsc-dev/src/sys/threadcomm/interface/threadcomm.c >>>>>>>> [ 0]176 bytes PetscThreadCommInitialize() line 694 in /Users/barrysmith/Src/petsc-dev/src/sys/threadcomm/interface/threadcomm.c >>>>>>>> [ 0]16 bytes PetscThreadCommInitialize() line 698 in /Users/barrysmith/Src/petsc-dev/src/sys/threadcomm/interface/threadcomm.c >>>>>>>> [ 0]176 bytes PetscThreadCommInitialize() line 694 in /Users/barrysmith/Src/petsc-dev/src/sys/threadcomm/interface/threadcomm.c >>>>>>>> [ 0]272 bytes PetscThreadCommInitialize() line 692 in /Users/barrysmith/Src/petsc-dev/src/sys/threadcomm/interface/threadcomm.c >>>>>>>> [ 0]16 bytes PetscThreadCommSetAffinities() line 331 in /Users/barrysmith/Src/petsc-dev/src/sys/threadcomm/interface/threadcomm.c >>>>>>>> [ 0]48 bytes PetscThreadCommCreate() line 122 in /Users/barrysmith/Src/petsc-dev/src/sys/threadcomm/interface/threadcomm.c >>>>>>>> [ 0]320 bytes PetscThreadCommCreate() line 118 in /Users/barrysmith/Src/petsc-dev/src/sys/threadcomm/interface/threadcomm.c >>>>>>>> Completed test examples >>>>>>>> barry-smiths-macbook-pro:petsc-dev barrysmith$ >>>>>>>> >>>>>>>> >>>>>>>> On Jul 14, 2012, at 6:36 PM, Shri wrote: >>>>>>>> >>>>>>>>> Which example are you running? I don't see any memory leak for the examples i've tested (threadcomm examples sys/threadcomm/examples/tutorials/ and snes/examples/tutorials/ex19.c,ex5.c) >>>>>>>>> On Jul 14, 2012, at 4:22 PM, Barry Smith wrote: >>>>>>>>> >>>>>>>>>> >>>>>>>>>> Please always test for memory leaks before pushing >>>>>>>>>> >>>>>>>>>> [ 0]48 bytes PetscThreadCommCreate() line 122 in /Users/barrysmith/Src/petsc-dev/src/sys/threadcomm/interface/threadcomm.c >>>>>>>>>> [0] PetscThreadCommInitialize() line 685 in /Users/barrysmith/Src/petsc-dev/src/sys/threadcomm/interface/threadcomm.c >>>>>>>>>> [0] PetscThreadCommInitializePackage() line 53 in /Users/barrysmith/Src/petsc-dev/src/sys/threadcomm/interface/dlregisthreadcomm.c >>>>>>>>>> [ 0]320 bytes PetscThreadCommCreate() line 118 in /Users/barrysmith/Src/petsc-dev/src/sys/threadcomm/interface/threadcomm.c >>>>>>>>>> [0] PetscThreadCommInitialize() line 685 in /Users/barrysmith/Src/petsc-dev/src/sys/threadcomm/interface/threadcomm.c >>>>>>>>>> [0] PetscThreadCommInitializePackage() line 53 in /Users/barrysmith/Src/petsc-dev/src/sys/threadcomm/interface/dlregisthreadcomm.c >>>>>>>>>> >>>>>>>>>> >>>>>>>>>> >>>>>>>>> >>>>>>>> >>>>>>>> >>>>>>> >>>>>> >>>>> >>>> >>> >> >

1 0