[petsc-dev] [petsc-maint #135017] Re: petsc-dev/buildsystem is broken on OS X.6+openmpi-nofortran
Aron Ahmadia
aron at ahmadia.net
Mon Oct 8 02:31:39 CDT 2012
Dang, that's almost a perfect match to my gcc:
i686-apple-darwin10-gcc-4.2.1 (GCC) 4.2.1 (Apple Inc. build 5666) (dot 3)
We can definitely reproduce this on two 10.6 systems out here at KAUST.
Ketch and I are both using "brew" on these boxen, but this is fairly
non-invasive with regards to this stack.
Matt, Satish, can you guys get your config and make logs to me somehow?
When I get some time I will try to bisect back to the commit in -dev that
is causing this.
A
On Sun, Oct 7, 2012 at 10:16 PM, Matthew Knepley <petsc-maint at mcs.anl.gov>wrote:
> On Sun, Oct 7, 2012 at 2:36 PM, Aron Ahmadia <aron.ahmadia at gmail.com>wrote:
>
>> Please report your gcc compiler versions.
>>
>
> knepley:/PETSc3/petsc/petsc-dev$ mpicc --show
> mpicc --show
> gcc -D_REENTRANT -Wl,-u,_munmap -Wl,-multiply_defined,suppress -lmpi
> -lopen-rte -lopen-pal
> knepley:/PETSc3/petsc/petsc-dev$ gcc --version
> gcc --version
> i686-apple-darwin10-gcc-4.2.1 (GCC) 4.2.1 (Apple Inc. build 5664)
> Copyright (C) 2007 Free Software Foundation, Inc.
> This is free software; see the source for copying conditions. There is NO
> warranty; not even for MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.
>
> Matt
>
>
>> A
>>
>> Sent from my iPhone
>>
>> On Oct 7, 2012, at 7:49 PM, Matthew Knepley <petsc-maint at mcs.anl.gov>
>> wrote:
>>
>> > On Sun, Oct 7, 2012 at 12:48 PM, Satish Balay <petsc-maint at mcs.anl.gov>
>> wrote:
>> > My test ran fine again.
>> >
>> > Perhaps you can run the example manually in a debugger to get a better
>> > trace.
>> >
>> > Yes, I can't reproduce this either.
>> >
>> > Matt
>> >
>> >
>> > Satish
>> >
>> > On Sun, 7 Oct 2012, Satish Balay wrote:
>> >
>> > > ok - then the issue is unrelated to fortran. I'll redo the 10.6 build
>> > > and see if its reproduceable.
>> > >
>> > > and secondly there is some wierd bug where fortran interface is not
>> > > getting built [you could try 'make all-legacy' and see if it gets
>> > > built correctly]
>> > >
>> > > Also gfortran-4.2* is buggy [wrt some of the f90 interface
>> > > functionality]. gfortran >= 4.3 is prefered for that. [but if not
>> > > using f90 interfaces - it doesn't matter]
>> > >
>> > > Satish
>> > >
>> > > On Sun, 7 Oct 2012, Aron Ahmadia wrote:
>> > >
>> > > > reproduced with fortran enabled, additionally:
>> > > >
>> > > > /Users/aron/sandbox/petsc/arch-darwin-c-debug/bin/mpif90 -c -Wall
>> > > > -Wno-unused-variable -g -I/Users/aron/sandbox/petsc/include
>> > > > -I/Users/aron/sandbox/petsc/arch-darwin-c-debug/include -o
>> ex5f.o ex5f.F
>> > > > Warning: Nonconforming tab character in column 1 of line 2757
>> > > > Warning: Nonconforming tab character in column 1 of line 4937
>> > > > Warning: Nonconforming tab character in column 1 of line 6914
>> > > > Warning: Nonconforming tab character in column 1 of line 8887
>> > > > Warning: Nonconforming tab character in column 1 of line 10901
>> > > > /Users/aron/sandbox/petsc/arch-darwin-c-debug/bin/mpif90
>> > > > -Wl,-multiply_defined,suppress -Wl,-multiply_defined -Wl,suppress
>> > > > -Wl,-commons,use_dylibs -Wl,-search_paths_first
>> > > > -Wl,-multiply_defined,suppress -Wl,-multiply_defined -Wl,suppress
>> > > > -Wl,-commons,use_dylibs -Wl,-search_paths_first -Wall
>> > > > -Wno-unused-variable -g -o ex5f ex5f.o
>> > > > -L/Users/aron/sandbox/petsc/arch-darwin-c-debug/lib -lpetsc
>> > > > -L/usr/X11R6/lib -lX11 -lpthread -llapack -lblas
>> > > > -L/usr/lib/gcc/i686-apple-darwin10/4.2.1/x86_64
>> > > > -L/usr/lib/i686-apple-darwin10/4.2.1
>> > > > -L/usr/lib/gcc/i686-apple-darwin10/4.2.1 -lmpichf90 -lgfortran
>> > > >
>> -L/usr/local/Cellar/gfortran/4.2.4-5664/lib/gcc/i686-apple-darwin10/4.2.1/x86_64
>> > > >
>> -L/usr/local/Cellar/gfortran/4.2.4-5664/lib/gcc/i686-apple-darwin10/4.2.1
>> > > > -L/usr/local/Cellar/gfortran/4.2.4-5664/lib/gcc
>> > > > -L/usr/local/Cellar/gfortran/4.2.4-5664/lib -ldl -lpmpich -lmpich
>> -lopa
>> > > > -lmpl -lpthread -lSystem -ldl
>> > > > Undefined symbols:
>> > > > "_dmlocaltoglobalbegin_", referenced from:
>> > > > _forminitialguess_ in ex5f.o
>> > > > "_dmgetlocalvector_", referenced from:
>> > > > _forminitialguess_ in ex5f.o
>> > > > "_matassemblybegin_", referenced from:
>> > > > _formjacobianlocal_ in ex5f.o
>> > > > "_dmdestroy_", referenced from:
>> > > > _MAIN__ in ex5f.o
>> > > > "_snessolve_", referenced from:
>> > > > _MAIN__ in ex5f.o
>> > > > "_vecrestorearray_", referenced from:
>> > > > _forminitialguess_ in ex5f.o
>> > > > "_vecdestroy_", referenced from:
>> > > > _MAIN__ in ex5f.o
>> > > > _MAIN__ in ex5f.o
>> > > > "_matsetvalueslocal_", referenced from:
>> > > > _formjacobianlocal_ in ex5f.o
>> > > > _formjacobianlocal_ in ex5f.o
>> > > > "_matassemblyend_", referenced from:
>> > > > _formjacobianlocal_ in ex5f.o
>> > > > "_dmdagetcorners_", referenced from:
>> > > > _MAIN__ in ex5f.o
>> > > > "_dmdasetlocaljacobian_", referenced from:
>> > > > _MAIN__ in ex5f.o
>> > > > "_petscinitialize_", referenced from:
>> > > > _MAIN__ in ex5f.o
>> > > > "_dmcreateglobalvector_", referenced from:
>> > > > _MAIN__ in ex5f.o
>> > > > "_petscfinalize_", referenced from:
>> > > > _MAIN__ in ex5f.o
>> > > > "_snessetdm_", referenced from:
>> > > > _MAIN__ in ex5f.o
>> > > > "_snessetfromoptions_", referenced from:
>> > > > _MAIN__ in ex5f.o
>> > > > "_snescreate_", referenced from:
>> > > > _MAIN__ in ex5f.o
>> > > > "_snesdestroy_", referenced from:
>> > > > _MAIN__ in ex5f.o
>> > > > "_dmdacreate2d_", referenced from:
>> > > > _MAIN__ in ex5f.o
>> > > > "_vecduplicate_", referenced from:
>> > > > _MAIN__ in ex5f.o
>> > > > "_petsclogflops_", referenced from:
>> > > > _formfunctionlocal_ in ex5f.o
>> > > > "_dmdagetghostcorners_", referenced from:
>> > > > _MAIN__ in ex5f.o
>> > > > "_dmrestorelocalvector_", referenced from:
>> > > > _forminitialguess_ in ex5f.o
>> > > > "_snesgetiterationnumber_", referenced from:
>> > > > _MAIN__ in ex5f.o
>> > > > "_dmlocaltoglobalend_", referenced from:
>> > > > _forminitialguess_ in ex5f.o
>> > > > "_dmdasetlocalfunction_", referenced from:
>> > > > _MAIN__ in ex5f.o
>> > > > "_dmdagetinfo_", referenced from:
>> > > > _MAIN__ in ex5f.o
>> > > > "_petscoptionsgetreal_", referenced from:
>> > > > _MAIN__ in ex5f.o
>> > > > "_vecgetarray_", referenced from:
>> > > > _forminitialguess_ in ex5f.o
>> > > > ld: symbol(s) not found
>> > > > collect2: ld returned 1 exit status
>> > > > make[3]: [ex5f] Error 1 (ignored)
>> > > >
>> > > > On Sun, Oct 7, 2012 at 7:29 PM, Satish Balay <
>> petsc-maint at mcs.anl.gov>wrote:
>> > > >
>> > > > > Hm - nightlybuilds use --download-mpich=1 - but I don't see this
>> error
>> > > > > on OSX build.
>> > > > >
>> > > > > wrt binsection - the following worked - the last time I tried.
>> > > > >
>> > > > > http://www.mcs.anl.gov/petsc/documentation/faq.html#bisect
>> > > > >
>> > > > > Satish
>> > > > >
>> > > > > On Sun, 7 Oct 2012, Aron Ahmadia wrote:
>> > > > >
>> > > > > > reproduced with --download-mpich=1
>> > > > > >
>> > > > > > how the heck do you guys bisect with two independent repos?
>> > > > > >
>> > > > > > A
>> > > > > >
>> > > > > > On Sun, Oct 7, 2012 at 6:28 PM, Aron Ahmadia <aron at ahmadia.net>
>> wrote:
>> > > > > >
>> > > > > > > I'm able to reproduce with shared libraries enabled and
>> disabled. Your
>> > > > > > > nightly logs aren't indicating failures as far as I can tell.
>> > > > > > >
>> > > > > > > changeset: 24541:729963bc657c
>> > > > > > > tag: tip
>> > > > > > > user: Stefano Zampini <stefano.zampini at gmail.com>
>> > > > > > > date: Sat Oct 06 13:39:31 2012 +0200
>> > > > > > > summary: Rearranged some deallocations when building
>> FETIDP mat.
>> > > > > > > (same failure with 2 processes), I'm looking into whether any
>> of the
>> > > > > > > specific configure options (python/shared libraries) triggers
>> the
>> > > > > failure.
>> > > > > > >
>> > > > > > > ~/s/petsc ❯❯❯ make PETSC_DIR=/Users/aron/sandbox/petsc
>> > > > > > > PETSC_ARCH=arch-darwin-c-debug test
>> > > > > > > Running test examples to verify correct installation
>> > > > > > > Using PETSC_DIR=/Users/aron/sandbox/petsc and
>> > > > > > > PETSC_ARCH=arch-darwin-c-debug
>> > > > > > > Possible error running C/C++ src/snes/examples/tutorials/ex19
>> with 1
>> > > > > MPI
>> > > > > > > process
>> > > > > > > See http://www.mcs.anl.gov/petsc/documentation/faq.html
>> > > > > > > [peregrine:31992] *** Process received signal ***
>> > > > > > > [peregrine:31992] Signal: Segmentation fault (11)
>> > > > > > > [peregrine:31992] Signal code: Address not mapped (1)
>> > > > > > > [peregrine:31992] Failing at address: 0x440000b0
>> > > > > > > [peregrine:31992] [ 0] 2 libSystem.B.dylib
>> > > > > > > 0x00007fff846391ba _sigtramp + 26
>> > > > > > > [peregrine:31992] [ 1] 3 ???
>> > > > > > > 0x0000000101434fd0 0x0 + 4316155856
>> > > > > > > [peregrine:31992] [ 2] 4 libpetsc.dylib
>> > > > > > > 0x0000000100118456 PetscInitialize + 1424
>> > > > > > > [peregrine:31992] [ 3] 5 ex19
>> > > > > > > 0x000000010000110f main + 43
>> > > > > > > [peregrine:31992] [ 4] 6 ex19
>> > > > > > > 0x00000001000010dc start + 52
>> > > > > > > [peregrine:31992] [ 5] 7 ???
>> > > > > > > 0x0000000000000007 0x0 + 7
>> > > > > > > [peregrine:31992] *** End of error message ***
>> > > > > > > mpiexec noticed that job rank 0 with PID 31992 on node
>> peregrine.local
>> > > > > > > exited on signal 11 (Segmentation fault).
>> > > > > > >
>> > > > > >
>> > > > > >
>> > > > >
>> > > >
>> > > >
>> > >
>> > >
>> >
>> >
>> >
>> >
>> > --
>> > 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
>>
>>
>
>
> --
> 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
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.mcs.anl.gov/pipermail/petsc-dev/attachments/20121008/a21aa3ea/attachment.html>
More information about the petsc-dev
mailing list