[CIG-SHORT] Git Version of PETSc does not compile

Matthew Knepley knepley at mcs.anl.gov
Fri Feb 13 02:23:39 PST 2015 

On Fri, Feb 13, 2015 at 4:11 AM, Romain Jolivet <jolivetinsar at gmail.com>
wrote:

> On 12 Feb 2015, at 21:22, Matthew Knepley <knepley at mcs.anl.gov> wrote:
>
> On Thu, Feb 12, 2015 at 3:18 PM, Romain Jolivet <jolivetinsar at gmail.com>
> wrote:
>
>> I have the exact same issue running the pylith-installer on our little
>> cluster (no macports installed there).
>> This issue shows up for configuring petsc and pylith.
>> On the cluster, we are running git 1.9.1
>> On my mac, I am running git 1.7.12.4 (Apple Git-37)
>>
>
> You are cloning without pulling any tags. You can fix it using
>
>   git fetch —tags
>
>
> I have tried what you suggested. Now, git describe returns something, but
> ./configure still does not understands it.
> Here is what I get:
>
> *pikaru* 10:08:33 ~/MYBIN/PyLith/pylith-build/petsc-pylith > ./configure
> --prefix=/Users/rjolivet/MYBIN/PyLith/pylith-2 --with-c2html=0 --with-x=0
> --with-clanguage=C --with-mpicompilers=1 --with-debugging=0
> --with-shared-libraries=1 --download-chaco=1 --download-ml=1
> --download-fblaslapack=1 --with-hdf5=1
> --with-hdf5-include=/opt/local/include
> --with-hdf5-lib=/opt/local/lib/libhdf5.dylib --LIBS=-lz
> CPPFLAGS="-I/Users/rjolivet/MYBIN/pylith/pylith-2/include "
> LDFLAGS="-L/Users/rjolivet/MYBIN/pylith/pylith-2/lib " CFLAGS="-g -O2"
> CXXFLAGS="-g -O2 -DMPICH_IGNORE_CXX_SEEK" FCFLAGS="-g -O2"
> PETSC_DIR=/Users/rjolivet/MYBIN/PyLith/pylith-build/petsc-pylith
> PETSC_ARCH=arch-pylith
>
> ===============================================================================
>              Configuring PETSc to compile on your system
>
>
> ===============================================================================
> TESTING: configureInstallationMethod from
> PETSc.options.petscclone(config/PETSc/options/petscclone.py:15)
>
>         fatal: No tags can describe
> '04fcc16b6fc1fef768ea84f574fa77a92547ade2'.
> Try --always, or create some tags.
>
> *******************************************************************************
>          UNABLE to CONFIGURE with GIVEN OPTIONS    (see configure.log for
> details):
>
> -------------------------------------------------------------------------------
> Your petsc source tree is broken. Use "git status" to check, or remove the
> entire directory and start all over again
>
> *******************************************************************************
>
> Because I need to move forward with my project, I will add —always in
> petscclone.py where git is called, but there might be a better way.
>

That will not fix it. Please try one more thing, and I am Cc'ing
petsc-maint since Jed is the git expert.

Clone the PETSc repository again, without the extra options you used

  git clone https://bitbucket.org:petsc/petsc.git

and try

  git checkout knepley/pylith petsc-pylith
  cd petsc-pylith
  git describe

If that works, everything should work with configuration and PyLith. I am
not sure
why

    git clone --depth 1 -b knepley/pylith
https://bitbucket.org/petsc/petsc.git petsc-pylith

is messed up, but maybe Jed can tell us.

  Thanks,

     Matt

Any suggestions?
> R
>
>
>   Thanks,
>
>     Matt
>
>
>> R
>>
>>
>> Romain Jolivet
>> Cambridge, UK
>> jolivetinsar at gmail.com
>> UK: +44 7 596 703 148
>> France: +33 6 52 91 76 39
>> US: +1 (626) 560 6356
>>
>> On Feb 12, 2015, at 21: 08, Brad Aagaard wrote:
>>
>> Romain,
>>
>> Having to remove the .git directory is very, very suspicious. This seems
>> to indicate conflicts from already installed software or inconsistent
>> environment settings.
>>
>> If you are using something like Macports, then be very careful about
>> managing what gets installed. You may have incompatibilities coming from
>> stuff installed there.
>>
>> As Matt mentioned, in order to diagnose problems like these we almost
>> always need detailed explanations of what you are doing and all the output.
>>
>> Brad
>>
>> On 02/12/2015 12:32 PM, Romain Jolivet wrote:
>>
>> I will check if there is an older version of petsc in the directory where
>> I put the executable.
>>
>>
>> In the mean time, ./configure crashes when I have the .git repository. I
>> would be glad not to remove this guy, but, if I don't, I cannot compile.
>>
>> What would be the reason for "git describe" to not send back any
>> meaningful informations?
>>
>> One other way is to add --always in the configure script, where git is
>> called. Should I do that?
>>
>>
>> R
>>
>>
>>
>> Romain Jolivet
>>
>> Cambridge, UK
>>
>> jolivetinsar at gmail.com
>>
>> UK: +44 7 596 703 148
>>
>> France: +33 6 52 91 76 39
>>
>> US: +1 (626) 560 6356
>>
>>
>> On Feb 12, 2015, at 20: 00, cig-short-request at geodynamics.org wrote:
>>
>>
>> Send CIG-SHORT mailing list submissions to
>>
>> cig-short at geodynamics.org
>>
>>
>> To subscribe or unsubscribe via the World Wide Web, visit
>>
>> http://lists.geodynamics.org/cgi-bin/mailman/listinfo/cig-short
>>
>> or, via email, send a message with subject or body 'help' to
>>
>> cig-short-request at geodynamics.org
>>
>>
>> You can reach the person managing the list at
>>
>> cig-short-owner at geodynamics.org
>>
>>
>> When replying, please edit your Subject line so it is more specific
>>
>> than "Re: Contents of CIG-SHORT digest..."
>>
>>
>>
>> Today's Topics:
>>
>>
>>   1. Re: Git Version of PETSc does not compile (Matthew Knepley)
>>
>>
>>
>> ----------------------------------------------------------------------
>>
>>
>> Message: 1
>>
>> Date: Thu, 12 Feb 2015 10:39:32 -0600
>>
>> From: Matthew Knepley <knepley at mcs.anl.gov>
>>
>> To: "cig-short at geodynamics.org" <cig-short at geodynamics.org>
>>
>> Subject: Re: [CIG-SHORT] Git Version of PETSc does not compile
>>
>> Message-ID:
>>
>> <CAMYG4G=yAZ9hNL8uqaH+8aC2ybsgKXm7oNe4dUuakKg2ZOJ5Xg at mail.gmail.com>
>>
>> Content-Type: text/plain; charset="utf-8"
>>
>>
>> On Thu, Feb 12, 2015 at 10:28 AM, Romain Jolivet <jolivetinsar at gmail.com>
>>
>> wrote:
>>
>>
>> In order to do that, I had to configure, so I ran:
>>
>>
>> ./configure --prefix=/Users/rjolivet/MYBIN/PyLith/pylith-2
>>
>> --with-c2html=0 --with-x=0 --with-clanguage=C --with-mpicompilers=1
>>
>> --with-debugging=0 --with-shared-libraries=1 --download-chaco=1
>>
>> --download-ml=1 --download-fblaslapack=1 --with-hdf5=1
>>
>> --with-hdf5-include=/opt/local/include
>>
>> --with-hdf5-lib=/opt/local/lib/libhdf5.dylib --LIBS=-lz
>>
>> CPPFLAGS="-I/Users/rjolivet/MYBIN/pylith/pylith-2/include "
>>
>> LDFLAGS="-L/Users/rjolivet/MYBIN/pylith/pylith-2/lib " CFLAGS="-g -O2"
>>
>> CXXFLAGS="-g -O2 -DMPICH_IGNORE_CXX_SEEK" FCFLAGS="-g -O2"
>>
>> PETSC_DIR=/Users/rjolivet/MYBIN/PyLith/pylith-build/petsc-pylith
>>
>> PETSC_ARCH=arch-pylith
>>
>>
>> A (minor) thing is that, on my mac, I have to remove the .git directory in
>>
>> order to do that. I don’t know why but when I do:
>>
>>
>> git describe
>>
>> fatal: No names found, cannot describe anything.
>>
>>
>>
>> You have destroyed the repository, since all history is kept under the
>> .git
>>
>> directory. I would start over.
>>
>> You definitely do not have to remove it to build.
>>
>>
>>
>> So I removed the .git directory and it configures fine.
>>
>>
>> I then type
>>
>> make V=1 -f ./gmakefile
>>
>>
>> Please find the output attached.
>>
>>
>>
>> So you have extraneous include directories:
>>
>>
>> mpicc -c -g -O2 -fPIC -I/Users/rjolivet/MYBIN/pylith/pylith-2/include
>>
>> -I/Users/rjolivet/MYBIN/PyLith/pylith-build/petsc-pylith/include
>>
>>
>> -I/Users/rjolivet/MYBIN/PyLith/pylith-build/petsc-pylith/arch-pylith/include
>>
>> -I/Users/rjolivet/MYBIN/PyLith/pylith-2/include -I/opt/local/include
>>
>> -I/opt/local/include/openmpi-gcc49
>>
>> -I/Users/rjolivet/MYBIN/pylith/pylith-2/include  -MMD -MP
>>
>>
>> /Users/rjolivet/MYBIN/PyLith/pylith-build/petsc-pylith/src/mat/interface/dlregismat.c
>>
>> -o arch-pylith/obj/src/mat/interface/dlregismat.o
>>
>>
>> which are probably coming in from you defining CFLAGS or passing them to
>>
>> PETSc somehow. It sounds like
>>
>> one of those has an old PETSc version.
>>
>>
>>  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 12 Feb 2015, at 16:13, cig-short-request at geodynamics.org wrote:
>>
>>
>> Send CIG-SHORT mailing list submissions to
>>
>> cig-short at geodynamics.org
>>
>>
>> To subscribe or unsubscribe via the World Wide Web, visit
>>
>> http://lists.geodynamics.org/cgi-bin/mailman/listinfo/cig-short
>>
>> or, via email, send a message with subject or body 'help' to
>>
>> cig-short-request at geodynamics.org
>>
>>
>> You can reach the person managing the list at
>>
>> cig-short-owner at geodynamics.org
>>
>>
>> When replying, please edit your Subject line so it is more specific
>>
>> than "Re: Contents of CIG-SHORT digest..."
>>
>>
>>
>> Today's Topics:
>>
>>
>>  1. Re: Git Version of PETSc does not compile (Romain Jolivet)
>>
>>  2. Re: Git Version of PETSc does not compile (Matthew Knepley)
>>
>>
>>
>> ----------------------------------------------------------------------
>>
>>
>> Message: 1
>>
>> Date: Thu, 12 Feb 2015 15:57:40 +0000
>>
>> From: Romain Jolivet <jolivetinsar at gmail.com>
>>
>> To: cig-short at geodynamics.org
>>
>> Subject: Re: [CIG-SHORT] Git Version of PETSc does not compile
>>
>> Message-ID: <61E15CC8-6107-404D-A00E-2275FEBA2EBA at gmail.com>
>>
>> Content-Type: text/plain; charset="windows-1252"
>>
>>
>> I have that line in matimpl.h.
>>
>> I just pulled a couple of hours ago by doing:
>>
>>
>> git clone --depth 1 -b knepley/pylith
>>
>> https://bitbucket.org/petsc/petsc.git petsc-pylith
>>
>>
>>
>> When I do git log, I see:
>>
>>
>> git log
>>
>>
>> commit 04fcc16b6fc1fef768ea84f574fa77a92547ade2
>>
>> Author: Matthew G. Knepley <knepley at gmail.com>
>>
>> Date:   Tue Feb 10 07:14:54 2015 -0600
>>
>>
>>   Merge branch 'knepley/fix-veclock-fftw' into knepley/pylith
>>
>>
>>   * knepley/fix-veclock-fftw:
>>
>>     DM: Fix more Vec lock issues
>>
>>     FFTW: Updated for vector locking
>>
>>
>> So I guess we are on the same version…
>>
>> 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 12 Feb 2015, at 15:51, cig-short-request at geodynamics.org wrote:
>>
>>
>> Message: 1
>>
>> Date: Thu, 12 Feb 2015 09:35:54 -0600
>>
>> From: Matthew Knepley <knepley at mcs.anl.gov>
>>
>> To: "cig-short at geodynamics.org" <cig-short at geodynamics.org>
>>
>> Subject: Re: [CIG-SHORT] Git Version of PETSc does not compile
>>
>> Message-ID:
>>
>> <CAMYG4Gn_7_G5ZVP_RnrsJAvD9+Je5mzt1mCgJLHB2rZ=DpxaTg at mail.gmail.com>
>>
>> Content-Type: text/plain; charset="utf-8"
>>
>>
>> On Thu, Feb 12, 2015 at 8:27 AM, Romain Jolivet <jolivetinsar at gmail.com>
>>
>> wrote:
>>
>>
>> Hello,
>>
>>
>> As I tried to compile the newest version of PETSc from the git repository,
>>
>> I ran into some compiling problems.
>>
>> It says that MAT_GetSubMatrix (which was not present in the previous
>>
>> versions of pets I was using) is undeclared when compiling
>>
>> src/mat/interface/dlregismat.c
>>
>> I checked and it seems to be declared the same way MAT_GetSubMatrices is.
>>
>> Any thoughts? Please find attached the make and configure log files.
>>
>>
>>
>> The declaration of that value is here:
>>
>>
>>
>>
>>
>> https://bitbucket.org/petsc/petsc/src/04fcc16b6fc1fef768ea84f574fa77a92547ade2/include/petsc-private/matimpl.h?at=knepley%2Fpylith#cl-1534
>>
>>
>> where you can see I am on the knepley/pylth branch. What branch are you
>> on?
>>
>>
>> Thanks,
>>
>>
>>   Matt
>>
>>
>>
>> -------------- next part --------------
>>
>> An HTML attachment was scrubbed...
>>
>> URL: <
>>
>>
>> http://lists.geodynamics.org/pipermail/cig-short/attachments/20150212/1d11d491/attachment-0001.html
>>
>>
>>
>> ------------------------------
>>
>>
>> Message: 2
>>
>> Date: Thu, 12 Feb 2015 10:12:34 -0600
>>
>> From: Matthew Knepley <knepley at mcs.anl.gov>
>>
>> To: "cig-short at geodynamics.org" <cig-short at geodynamics.org>
>>
>> Subject: Re: [CIG-SHORT] Git Version of PETSc does not compile
>>
>> Message-ID:
>>
>> <CAMYG4GmXVw3qVK8e2_8Jeiz0VO6autQje25E+6nsx=DGTDaFFg at mail.gmail.com>
>>
>> Content-Type: text/plain; charset="utf-8"
>>
>>
>> On Thu, Feb 12, 2015 at 9:57 AM, Romain Jolivet <jolivetinsar at gmail.com>
>>
>> wrote:
>>
>>
>> I have that line in matimpl.h.
>>
>> I just pulled a couple of hours ago by doing:
>>
>>
>> git clone --depth 1 -b knepley/pylith
>>
>>
>> https://bitbucket.org/petsc/petsc.git petsc-pylith
>>
>>
>> When I do git log, I see:
>>
>>
>> git log
>>
>>
>> commit 04fcc16b6fc1fef768ea84f574fa77a92547ade2
>>
>> Author: Matthew G. Knepley <knepley at gmail.com>
>>
>> Date:   Tue Feb 10 07:14:54 2015 -0600
>>
>>
>>   Merge branch 'knepley/fix-veclock-fftw' into knepley/pylith
>>
>>
>>
>>   * knepley/fix-veclock-fftw:
>>
>>     DM: Fix more Vec lock issues
>>
>>     FFTW: Updated for vector locking
>>
>>
>> So I guess we are on the same version…
>>
>>
>>
>> Next lets check that you are not picking up some old version of PETSc
>>
>> somewhere else on
>>
>> your machine through include paths.
>>
>>
>> make V=1 -f ./gmakefile
>>
>>
>> and send all output
>>
>>
>> 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 12 Feb 2015, at 15:51, cig-short-request at geodynamics.org wrote:
>>
>>
>> Message: 1
>>
>> Date: Thu, 12 Feb 2015 09:35:54 -0600
>>
>> From: Matthew Knepley <knepley at mcs.anl.gov>
>>
>> To: "cig-short at geodynamics.org" <cig-short at geodynamics.org>
>>
>> Subject: Re: [CIG-SHORT] Git Version of PETSc does not compile
>>
>> Message-ID:
>>
>> <CAMYG4Gn_7_G5ZVP_RnrsJAvD9+Je5mzt1mCgJLHB2rZ=DpxaTg at mail.gmail.com>
>>
>> Content-Type: text/plain; charset="utf-8"
>>
>>
>> On Thu, Feb 12, 2015 at 8:27 AM, Romain Jolivet <jolivetinsar at gmail.com>
>>
>> wrote:
>>
>>
>> Hello,
>>
>>
>> As I tried to compile the newest version of PETSc from the git repository,
>>
>> I ran into some compiling problems.
>>
>> It says that MAT_GetSubMatrix (which was not present in the previous
>>
>> versions of pets I was using) is undeclared when compiling
>>
>> src/mat/interface/dlregismat.c
>>
>> I checked and it seems to be declared the same way MAT_GetSubMatrices is.
>>
>> Any thoughts? Please find attached the make and configure log files.
>>
>>
>>
>> The declaration of that value is here:
>>
>>
>>
>>
>>
>>
>> https://bitbucket.org/petsc/petsc/src/04fcc16b6fc1fef768ea84f574fa77a92547ade2/include/petsc-private/matimpl.h?at=knepley%2Fpylith#cl-1534
>>
>>
>> where you can see I am on the knepley/pylth branch. What branch are you
>> on?
>>
>>
>> Thanks,
>>
>>
>>   Matt
>>
>>
>>
>>
>> _______________________________________________
>>
>> 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
>>
>> -------------- next part --------------
>>
>> An HTML attachment was scrubbed...
>>
>> URL: <
>>
>>
>> http://lists.geodynamics.org/pipermail/cig-short/attachments/20150212/d7532d3b/attachment.html
>>
>>
>>
>> ------------------------------
>>
>>
>> Subject: Digest Footer
>>
>>
>> _______________________________________________
>>
>> CIG-SHORT mailing list
>>
>> CIG-SHORT at geodynamics.org
>>
>> http://lists.geodynamics.org/cgi-bin/mailman/listinfo/cig-short
>>
>>
>> ------------------------------
>>
>>
>> End of CIG-SHORT Digest, Vol 101, Issue 12
>>
>> ******************************************
>>
>>
>>
>>
>> _______________________________________________
>>
>> 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
>>
>> -------------- next part --------------
>>
>> An HTML attachment was scrubbed...
>>
>> URL: <
>> http://lists.geodynamics.org/pipermail/cig-short/attachments/20150212/3598b455/attachment-0001.html
>> >
>>
>>
>> ------------------------------
>>
>>
>> Subject: Digest Footer
>>
>>
>> _______________________________________________
>>
>> CIG-SHORT mailing list
>>
>> CIG-SHORT at geodynamics.org
>>
>> http://lists.geodynamics.org/cgi-bin/mailman/listinfo/cig-short
>>
>>
>> ------------------------------
>>
>>
>> End of CIG-SHORT Digest, Vol 101, Issue 14
>>
>> ******************************************
>>
>>
>>
>>
>> N �n�r����)em�h�yhiם�w^��
>>
>>
>>
>> _______________________________________________
>> 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
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.geodynamics.org/pipermail/cig-short/attachments/20150213/e08e75e4/attachment-0001.html>

More information about the CIG-SHORT mailing list