[CIG-SHORT] PyLith configure script cannot find PETSc

Fri Feb 13 13:14:07 PST 2015

Romain,

I have pushed changes to PyLith master and next branches to fix this 
issue. I also figured out how to get the m4 submodules updating 
automatically.

I have pushed changes to the installer so that it doesn't use --depth 1 
in the git clone command. This appears to also help resolve an issue you 
had.

Assuming your PETSc is built and installed, you should be able to (1) 
pull pylith (git pull), (2) autoreconf -if, (3) configure, and (4) build.

Thanks for your patience,
Brad

On 02/13/2015 07:54 AM, Matthew Knepley wrote:
> On Fri, Feb 13, 2015 at 9:47 AM, Romain Jolivet <jolivetinsar at gmail.com>
> wrote:
>
>> The configure.log file is attached.
>>
>> In the build directory of PETSc, I could find this arch-pylith directory,
>> indeed, but not where PETSc has been installed (with all the other pylith
>> stuffs like nemesis, spatialdata etc).
>> When I build pylith, I point to the directory where petsc has been
>> installed, not built. Should I do that?
>>
>
> Ah, this is the difference between installing in a custom place (usually
> now called $PETSC_DIR)
> and using the default installation in $PETSC_DIR/$PETSC_ARCH. The
> recommended PyLith
> thing was to use the default, so our tools were not checking the other
> option.
>
> Brad, we can check for an installation in $PETSC_DIR if we look for
>
>    $PETSC_DIR/include/petscconf.h
>
> or
>
>    $PETSC_DIR/lib/petsc-conf/petscrules
>
> Otherwise we use the default form. This is a clash between two different
> styles of Linux
> installation, and it is a pain.
>
>    Thanks,
>
>       Matt
>
> R
>>
>>
>>
>> Romain Jolivet
>> Cambridge, UK
>> jolivetinsar at gmail.com
>> UK: +44 7596 703 148
>> France: +33 6 52 91 76 39
>> USA: +1 (626) 560 6356
>>
>> On 13 Feb 2015, at 15:43, Matthew Knepley <knepley at mcs.anl.gov> wrote:
>>
>> On Fri, Feb 13, 2015 at 9:34 AM, Romain Jolivet <jolivetinsar at gmail.com>
>> wrote:
>>
>>> I don’t have  $PETSC_DIR/$PETSC_ARCH/lib/petsc-conf
>>> It is the one being checked for in configure.
>>> What I have is only $PETSC_DIR/lib/petsc-conf
>>> That is both on our linux cluster and on my iMac.
>>>
>>
>> The build process creates them. If they are not there, PETSc was not
>> built. The
>> configure log has a record of created files. Can you send it for the
>> configuration
>> that you cannot find this directory for?
>>
>>    Matt
>>
>>
>>> R
>>>
>>>
>>> Romain Jolivet
>>> Cambridge, UK
>>> jolivetinsar at gmail.com
>>> UK: +44 7596 703 148
>>> France: +33 6 52 91 76 39
>>> USA: +1 (626) 560 6356
>>>
>>> On 13 Feb 2015, at 15:28, Matthew Knepley <knepley at mcs.anl.gov> wrote:
>>>
>>> On Fri, Feb 13, 2015 at 9:22 AM, Romain Jolivet <jolivetinsar at gmail.com>
>>> wrote:
>>>
>>>> I got it to work, but not with the m4 file you provided.
>>>> In your file, you test for $PETSC_DIR/$PETSC_ARCH/lib/petsc-conf
>>>>
>>> While the architecture I have is $PETSC_DIR/lib/petsc-conf
>>>> The variable PETSC_ARCH is not used in the destination directory when I
>>>> install petsc…
>>>> Is that normal?
>>>>
>>>
>>> Both of these directories exist. What is being checked for?
>>>
>>>    Thanks,
>>>
>>>       Matt
>>>
>>>
>>>> R
>>>>
>>>> Romain Jolivet
>>>> Cambridge, UK
>>>> jolivetinsar at gmail.com
>>>> UK: +44 7596 703 148
>>>> France: +33 6 52 91 76 39
>>>> USA: +1 (626) 560 6356
>>>>
>>>> On 13 Feb 2015, at 14:54, Brad Aagaard <baagaard at usgs.gov> wrote:
>>>>
>>>> Romain and Matt,
>>>>
>>>> This is already fixed. Romain needs to pull the m4 dir or reclone pylith.
>>>>
>>>> In the top level pylith dir,
>>>>
>>>> cd m4
>>>> git pull origin master
>>>>
>>>> Brad
>>>>
>>>>
>>>>
>>>> On Feb 13, 2015, at 3:54 AM, Romain Jolivet <jolivetinsar at gmail.com>
>>>> wrote:
>>>>
>>>> Oh, sorry, here it is.
>>>> R
>>>>
>>>> <cit_petsc.m4>
>>>>
>>>>
>>>> Romain Jolivet
>>>> Cambridge, UK
>>>> jolivetinsar at gmail.com
>>>> UK: +44 7596 703 148
>>>> France: +33 6 52 91 76 39
>>>> USA: +1 (626) 560 6356
>>>>
>>>> On 13 Feb 2015, at 11:52, Matthew Knepley <knepley at mcs.anl.gov> wrote:
>>>>
>>>> On Fri, Feb 13, 2015 at 5:45 AM, Romain Jolivet <jolivetinsar at gmail.com>
>>>> wrote:
>>>>
>>>>> Hello,
>>>>>
>>>>> After compiling manually petsc from the git repository, I noticed the
>>>>> architecture of the file directories that are copied into the —prefix
>>>>> argument are different from why I was used to.
>>>>> As a consequence, the configure in pylith cannot find the petsc
>>>>> libraries.
>>>>> I looked into the m4/cit_petsc.m4 file used by autoreconf, and I do not
>>>>> see the case for the architecture I have.
>>>>>
>>>>> I could modify it to adapt to my system, but, there should be a better
>>>>> way.
>>>>> Attached is the modified cit_petsc.m4 I modified to compile pylith.
>>>>> Any suggestions?
>>>>>
>>>>
>>>> There is nothing attached. The PyLith master branch tracks PETSc
>>>> knepley/pylith, so all fixes
>>>> should be there.
>>>>
>>>>    Thanks,
>>>>
>>>>      Matt
>>>>
>>>>
>>>>>
>>>>> R
>>>>>
>>>>>
>>>>> —————————————————————————————————————
>>>>> —————————————————————————————————————
>>>>> Romain Jolivet
>>>>> Postdoctoral Fellow
>>>>>
>>>>> University of Cambridge
>>>>> Department of Earth Sciences
>>>>> Bullard Labs
>>>>> Madingley Rise
>>>>> Madingley Road
>>>>> Cambridge CB3 0EZ
>>>>> United Kingdom
>>>>>
>>>>> email: rpj29 at cam.ac.uk
>>>>> Phone: +44 1223 748 938
>>>>> Mobile: +44 7596 703 148
>>>>>
>>>>> France: +33 6 52 91 76 39
>>>>> US: +1 (626) 560 6356
>>>>> —————————————————————————————————————
>>>>> —————————————————————————————————————
>>>>>
>>>>> _______________________________________________
>>>>> CIG-SHORT mailing list
>>>>> CIG-SHORT at geodynamics.org
>>>>> http://lists.geodynamics.org/cgi-bin/mailman/listinfo/cig-short
>>>>
>>>>
>>>>
>>>>
>>>> --
>>>> 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
>>>>   _______________________________________________
>>>> CIG-SHORT mailing list
>>>> CIG-SHORT at geodynamics.org
>>>> http://lists.geodynamics.org/cgi-bin/mailman/listinfo/cig-short
>>>>
>>>>
>>>> _______________________________________________
>>>> CIG-SHORT mailing list
>>>> CIG-SHORT at geodynamics.org
>>>> http://lists.geodynamics.org/cgi-bin/mailman/listinfo/cig-short
>>>>
>>>> _______________________________________________
>>>> CIG-SHORT mailing list
>>>> CIG-SHORT at geodynamics.org
>>>> http://lists.geodynamics.org/cgi-bin/mailman/listinfo/cig-short
>>>>
>>>>
>>>>
>>>> _______________________________________________
>>>> CIG-SHORT mailing list
>>>> CIG-SHORT at geodynamics.org
>>>> http://lists.geodynamics.org/cgi-bin/mailman/listinfo/cig-short
>>>>
>>>
>>>
>>>
>>> --
>>> 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
>>>   _______________________________________________
>>> CIG-SHORT mailing list
>>> CIG-SHORT at geodynamics.org
>>> http://lists.geodynamics.org/cgi-bin/mailman/listinfo/cig-short
>>>
>>>
>>>
>>> _______________________________________________
>>> CIG-SHORT mailing list
>>> CIG-SHORT at geodynamics.org
>>> http://lists.geodynamics.org/cgi-bin/mailman/listinfo/cig-short
>>>
>>
>>
>>
>> --
>> 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
>>   _______________________________________________
>> CIG-SHORT mailing list
>> CIG-SHORT at geodynamics.org
>> http://lists.geodynamics.org/cgi-bin/mailman/listinfo/cig-short
>>
>>
>>
>> _______________________________________________
>> CIG-SHORT mailing list
>> CIG-SHORT at geodynamics.org
>> http://lists.geodynamics.org/cgi-bin/mailman/listinfo/cig-short
>>
>
>
>
>
>
> _______________________________________________
> CIG-SHORT mailing list
> CIG-SHORT at geodynamics.org
> http://lists.geodynamics.org/cgi-bin/mailman/listinfo/cig-short
>