April 2011 - petsc-dev - mailman3.cels.anl.gov

Re: [petsc-dev] troff man pages
by Barry Smith 22 Apr '11

22 Apr '11

On Apr 22, 2011, at 12:26 PM, Jed Brown wrote: > > > On Fri, Apr 22, 2011 at 18:53, Barry Smith <bsmith(a)mcs.anl.gov> wrote: > Rather than use an outdated tool that uses an outdated display model ("real man pages") can you not find an "our editors" that "could automatically display the formatted" HTML "page when we are typing slowly in the body of a function". > > Yes, we can do this too, but it involves another user configuration variable for where to find the relevant pages where as man pages already have an established index. Also, man pages are somewhat more structured so it's easier to display only the most relevant parts. And man pages work across different libraries instead of needing to look in a different index when typing slow within the body of calls to each of PETSc, MPI, Lapack, pthread, etc. In other words the older technology is better than the newer technology, not an uncommon occurrence. I'm fine with you hooking up traditional man pages, but note that we shove a bunch of stuff into the .html manual page that sowing doesn't directly handle. So likely the traditional man pages would be different. Barry > > Surely emacs is not so outdated that they cannot use our html manual pages in this role? > > Emacs may be totally outdated, but it can certainly do this.

1 0

Re: [petsc-dev] troff man pages
by Jed Brown 22 Apr '11

22 Apr '11

On Fri, Apr 22, 2011 at 18:53, Barry Smith <bsmith(a)mcs.anl.gov> wrote: > Rather than use an outdated tool that uses an outdated display model ("real > man pages") can you not find an "our editors" that "could automatically > display the formatted" HTML "page when we are typing slowly in the body of a > function". Yes, we can do this too, but it involves another user configuration variable for where to find the relevant pages where as man pages already have an established index. Also, man pages are somewhat more structured so it's easier to display only the most relevant parts. And man pages work across different libraries instead of needing to look in a different index when typing slow within the body of calls to each of PETSc, MPI, Lapack, pthread, etc. Surely emacs is not so outdated that they cannot use our html manual pages > in this role? Emacs may be totally outdated, but it can certainly do this.

1 0

Re: [petsc-dev] troff man pages
by Barry Smith 22 Apr '11

22 Apr '11

On Apr 22, 2011, at 11:42 AM, Jed Brown wrote: > MPICH2 uses the same documentation software and has real man pages, why doesn't PETSc? We originally did "real man pages", then did both, I think, for a while and then decided to just use the "modern" approach of html. I think it was a little bit of a pain in the butt to have both plus the disk space was a bigger deal back then. > > Part of why I ask is that our editors could automatically display the formatted man page when we are typing slowly in the body of a function call. I typically split my screen and jump to the definition when I'm unsure of things, but an automatically displayed man page would be better in some ways. Rather than use an outdated tool that uses an outdated display model ("real man pages") can you not find an "our editors" that "could automatically display the formatted" HTML "page when we are typing slowly in the body of a function". Surely emacs is not so outdated that they cannot use our html manual pages in this role? I say this partially in jest, but only partially. Barry

1 0

Re: [petsc-dev] Floating point exception in ex47cu.cu
by Barry Smith 22 Apr '11

22 Apr '11

On Apr 22, 2011, at 11:22 AM, Евгений Козлов wrote: > Hello, > > I am interesting in using PETSc for iterative solving sparse linear > system on multi-GPU systems. > > > RedHat 5.5, Cuda 3.2 > > Question1: Is it my problem or a bug of the algorithm? Don't know why that happens. > > Question2: Where can I find simple doc or example, which describe how > to solve sparse linear systems on multi-GPU systems using PETSc? Many of the examples will run in parallel with GPU for example: src/snes/examples/tutorials ex19 -pc_type jacobi -da_vec_type cusp -da_mat_type aijcusp Barry > > > -- > Best regards, > Eugene

1 0

troff man pages
by Jed Brown 22 Apr '11

22 Apr '11

MPICH2 uses the same documentation software and has real man pages, why doesn't PETSc? Part of why I ask is that our editors could automatically display the formatted man page when we are typing slowly in the body of a function call. I typically split my screen and jump to the definition when I'm unsure of things, but an automatically displayed man page would be better in some ways.

1 0

Floating point exception in ex47cu.cu
by Евгений Козлов 22 Apr '11

22 Apr '11

Hello, I am interesting in using PETSc for iterative solving sparse linear system on multi-GPU systems. First of all, I compiled PETSc-dev and tried to run some examples, which will run on GPU. I found src/snes/examples/tutorials/ex47cu.cu. It was compiled and executed. Out of original program src/snes/examples/tutorials/ex47cu.cu: [0]PETSC ERROR: --------------------- Error Message ------------------------------------ [0]PETSC ERROR: Floating point exception! [0]PETSC ERROR: User provided compute function generated a Not-a-Number! [0]PETSC ERROR: ------------------------------------------------------------------------ [0]PETSC ERROR: Petsc Development HG revision: d3e10315d68b1dd5481adb2889c7d354880da362 HG Date: Wed Apr 20 21:03:56 2011 -0500 [0]PETSC ERROR: See docs/changes/index.html for recent updates. [0]PETSC ERROR: See docs/faq.html for hints about trouble shooting. [0]PETSC ERROR: See docs/index.html for manual pages. [0]PETSC ERROR: ------------------------------------------------------------------------ [0]PETSC ERROR: ex47cu on a arch-linu named cn03 by kukushkinav Fri Apr 22 18:33:32 2011 [0]PETSC ERROR: Libraries linked from /home/kukushkinav/lib [0]PETSC ERROR: Configure run at Thu Apr 21 19:18:22 2011 [0]PETSC ERROR: Configure options --prefix=/home/kukushkinav --with-blas-lapack-dir=/opt/intel/composerxe-2011.0.084/mkl --with-mpi-dir=/opt/intel/impi/4.0.1.007/intel64/bin --with-cuda=1 --with-cusp=1 --with-thrust=1 --with-thrust-dir=/home/kukushkinav/include --with-cusp-dir=/home/kukushkinav/include [0]PETSC ERROR: ------------------------------------------------------------------------ [0]PETSC ERROR: SNESSolve_LS() line 167 in src/snes/impls/ls/ls.c [0]PETSC ERROR: SNESSolve() line 2407 in src/snes/interface/snes.c [0]PETSC ERROR: main() line 38 in src/snes/examples/tutorials/ex47cu.cu application called MPI_Abort(MPI_COMM_WORLD, 72) - process 0 Then I tried to find the place in source with the problem, I changed the function in struct ApplyStencil to void operator()(Tuple t) { thrust::get<0>(t) = 1; } Result: [0]PETSC ERROR: --------------------- Error Message ------------------------------------ [0]PETSC ERROR: Floating point exception! [0]PETSC ERROR: Infinite or not-a-number generated in mdot, entry 0! [0]PETSC ERROR: ------------------------------------------------------------------------ [0]PETSC ERROR: Petsc Development HG revision: d3e10315d68b1dd5481adb2889c7d354880da362 HG Date: Wed Apr 20 21:03:56 2011 -0500 [0]PETSC ERROR: See docs/changes/index.html for recent updates. [0]PETSC ERROR: See docs/faq.html for hints about trouble shooting. [0]PETSC ERROR: See docs/index.html for manual pages. [0]PETSC ERROR: ------------------------------------------------------------------------ [0]PETSC ERROR: ex47cu on a arch-linu named cn11 by kukushkinav Fri Apr 22 18:58:04 2011 [0]PETSC ERROR: Libraries linked from /home/kukushkinav/lib [0]PETSC ERROR: Configure run at Thu Apr 21 19:18:22 2011 [0]PETSC ERROR: Configure options --prefix=/home/kukushkinav --with-blas-lapack-dir=/opt/intel/composerxe-2011.0.084/mkl --with-mpi-dir=/opt/intel/impi/4.0.1.007/intel64/bin --with-cuda=1 --with-cusp=1 --with-thrust=1 --with-thrust-dir=/home/kukushkinav/include --with-cusp-dir=/home/kukushkinav/include [0]PETSC ERROR: ------------------------------------------------------------------------ [0]PETSC ERROR: VecMDot() line 1146 in src/vec/vec/interface/rvector.c [0]PETSC ERROR: KSPGMRESClassicalGramSchmidtOrthogonalization() line 66 in src/ksp/ksp/impls/gmres/borthog2.c [0]PETSC ERROR: GMREScycle() line 161 in src/ksp/ksp/impls/gmres/gmres.c [0]PETSC ERROR: KSPSolve_GMRES() line 244 in src/ksp/ksp/impls/gmres/gmres.c [0]PETSC ERROR: KSPSolve() line 426 in src/ksp/ksp/interface/itfunc.c [0]PETSC ERROR: SNES_KSPSolve() line 3107 in src/snes/interface/snes.c [0]PETSC ERROR: SNESSolve_LS() line 190 in src/snes/impls/ls/ls.c [0]PETSC ERROR: SNESSolve() line 2407 in src/snes/interface/snes.cС уважением, Евгений [0]PETSC ERROR: main() line 38 in src/snes/examples/tutorials/ex47cu.cu RedHat 5.5, Cuda 3.2 Question1: Is it my problem or a bug of the algorithm? Question2: Where can I find simple doc or example, which describe how to solve sparse linear systems on multi-GPU systems using PETSc? -- Best regards, Eugene

1 0

Re: [petsc-dev] bug in BuildSystem check
by Matthew Knepley 22 Apr '11

22 Apr '11

On Fri, Apr 22, 2011 at 8:28 AM, Jed Brown <jed(a)59a2.org> wrote: > On Fri, Apr 22, 2011 at 15:20, Barry Smith <bsmith(a)mcs.anl.gov> wrote: > >> > Are you sure? >> >> Absolutely! I do not care if it links ok. We cannot use something >> that doesn't have a proper prototype because (1) it spews crap all over a >> PETSc build that worries people (rightly so) and (2) how can you trust the >> link when you don't know how the compiler interpreted the arguments in the >> compile. > > > There is also the problem that if the prototype is not available, the > argument will be implicitly converted from float to double which may change > it's value. The standard actually requires that "isinf" and relatives be a > macro, the type of the argument is checked and a more specific function > (which may or may not be called isinf, and might be a compiler builtin) is > called. > Okay, pushed something. Not yet tested. Matt -- 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

1 0

Re: [petsc-dev] bug in BuildSystem check
by Jed Brown 22 Apr '11

22 Apr '11

On Fri, Apr 22, 2011 at 15:20, Barry Smith <bsmith(a)mcs.anl.gov> wrote: > > Are you sure? > > Absolutely! I do not care if it links ok. We cannot use something > that doesn't have a proper prototype because (1) it spews crap all over a > PETSc build that worries people (rightly so) and (2) how can you trust the > link when you don't know how the compiler interpreted the arguments in the > compile. There is also the problem that if the prototype is not available, the argument will be implicitly converted from float to double which may change it's value. The standard actually requires that "isinf" and relatives be a macro, the type of the argument is checked and a more specific function (which may or may not be called isinf, and might be a compiler builtin) is called.

1 0

Re: [petsc-dev] bug in BuildSystem check
by Barry Smith 22 Apr '11

22 Apr '11

On Apr 22, 2011, at 8:16 AM, Matthew Knepley wrote: > On Fri, Apr 22, 2011 at 8:12 AM, Barry Smith <bsmith(a)mcs.anl.gov> wrote: > > Matt, > > In the test for isinf() the compile warns about lack of prototype. BuildSystem should stop at this point and NOT set HAVE_ISINF. Could you please fix this? > > Are you sure? Absolutely! I do not care if it links ok. We cannot use something that doesn't have a proper prototype because (1) it spews crap all over a PETSc build that worries people (rightly so) and (2) how can you trust the link when you don't know how the compiler interpreted the arguments in the compile. Please fix it. Barry > It linked just fine. If you still want to signal an error, we can do it by looking for that string. > > Matt > > Thanks > > Barry > > > > sh: gcc -c -o conftest.o -Wall -Wwrite-strings -Wno-strict-aliasing -g3 conftest.c > Executing: gcc -c -o conftest.o -Wall -Wwrite-strings -Wno-strict-aliasing -g3 conftest.c > sh: > Possible ERROR while running compiler: error message = {conftest.c: In function 'main': > conftest.c:6: warning: implicit declaration of function 'isinf' > conftest.c:6: warning: unused variable 'a' > } > Source: > #include "confdefs.h" > #include "conffix.h" > #include <math.h> > > int main() { > double b = 2.0; int a = isinf(b); > ; > return 0; > } > Pushing language C > Popping language C > sh: gcc -o conftest -Wall -Wwrite-strings -Wno-strict-aliasing -g3 conftest.o -L/opt/local/lib/gcc/sparc-sun-solaris2.10/4.4.1 -L/opt/local/lib/gcc -L/usr/ccs/lib -L/opt/local/lib -ldl -lgcc_eh -ldl -lsocket -lnsl > Executing: gcc -o conftest -Wall -Wwrite-strings -Wno-strict-aliasing -g3 conftest.o -L/opt/local/lib/gcc/sparc-sun-solaris2.10/4.4.1 -L/opt/local/lib/gcc -L/usr/ccs/lib -L/opt/local/lib -ldl -lgcc_eh -ldl -lsocket -lnsl > sh: > Defined "HAVE_ISINF" to "1" > > > > > -- > 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

1 0

Re: [petsc-dev] bug in BuildSystem check
by Matthew Knepley 22 Apr '11

22 Apr '11

On Fri, Apr 22, 2011 at 8:12 AM, Barry Smith <bsmith(a)mcs.anl.gov> wrote: > > Matt, > > In the test for isinf() the compile warns about lack of prototype. > BuildSystem should stop at this point and NOT set HAVE_ISINF. Could you > please fix this? > Are you sure? It linked just fine. If you still want to signal an error, we can do it by looking for that string. Matt > Thanks > > Barry > > > > sh: gcc -c -o conftest.o -Wall -Wwrite-strings -Wno-strict-aliasing -g3 > conftest.c > Executing: gcc -c -o conftest.o -Wall -Wwrite-strings -Wno-strict-aliasing > -g3 conftest.c > sh: > Possible ERROR while running compiler: error message = {conftest.c: In > function 'main': > conftest.c:6: warning: implicit declaration of function 'isinf' > conftest.c:6: warning: unused variable 'a' > } > Source: > #include "confdefs.h" > #include "conffix.h" > #include <math.h> > > int main() { > double b = 2.0; int a = isinf(b); > ; > return 0; > } > Pushing language C > Popping language C > sh: gcc -o conftest -Wall -Wwrite-strings -Wno-strict-aliasing -g3 > conftest.o -L/opt/local/lib/gcc/sparc-sun-solaris2.10/4.4.1 > -L/opt/local/lib/gcc -L/usr/ccs/lib -L/opt/local/lib -ldl -lgcc_eh -ldl > -lsocket -lnsl > Executing: gcc -o conftest -Wall -Wwrite-strings -Wno-strict-aliasing > -g3 conftest.o -L/opt/local/lib/gcc/sparc-sun-solaris2.10/4.4.1 > -L/opt/local/lib/gcc -L/usr/ccs/lib -L/opt/local/lib -ldl -lgcc_eh -ldl > -lsocket -lnsl > sh: > Defined "HAVE_ISINF" to "1" > > -- 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

1 0