[petsc-dev] PETSc release soon, request for input on needed fixes or enhancements

Wed Mar 9 06:21:20 CST 2016

Hi,

the make install error occurred again last night, but with a different 
error:

           CC arch-linux2-c-debug/obj/src/tao/leastsquares/impls/pounders/gqt.o
      CLINKER /pmi/cmpbib/compilation_BIB_gcc_redhat_petsc-master_debug/COMPILE_AUTO/petsc-master-debug/arch-linux2-c-debug/lib/libpetsc.so.3.06.3
gmake[2]: Leaving directory `/pmi/cmpbib/compilation_BIB_gcc_redhat_petsc-master_debug/COMPILE_AUTO/petsc-master-debug'
=========================================
gmake[1]: Leaving directory `/pmi/cmpbib/compilation_BIB_gcc_redhat_petsc-master_debug/COMPILE_AUTO/petsc-master-debug'
Now to install the libraries do:
make PETSC_DIR=/pmi/cmpbib/compilation_BIB_gcc_redhat_petsc-master_debug/COMPILE_AUTO/petsc-master-debug PETSC_ARCH=arch-linux2-c-debug install
=========================================
*** Using PETSC_DIR=/pmi/cmpbib/compilation_BIB_gcc_redhat_petsc-master_debug/COMPILE_AUTO/petsc-master-debug PETSC_ARCH=arch-linux2-c-debug ***
     No cached configure in RDict at /pmi/cmpbib/compilation_BIB_gcc_redhat_petsc-master_debug/COMPILE_AUTO/petsc-master-debug/arch-linux2-c-debug/lib/petsc/conf/RDict.db
Traceback (most recent call last):
   File "./config/install.py", line 380, in <module>
     Installer(sys.argv[1:]).run()
   File "./config/install.py", line 373, in run
     self.runsetup()
   File "./config/install.py", line 334, in runsetup
     self.setup()
   File "./config/install.py", line 60, in setup
     self.setupModules()
   File "./config/install.py", line 51, in setupModules
     self.setCompilers  = self.framework.require('config.setCompilers',         None)
AttributeError: 'NoneType' object has no attribute 'require'
make: *** [install] Error 1

Here are the logs:

http://www.giref.ulaval.ca/~cmpgiref/petsc-master-debug/2016.03.09.02h00m02s_configure.log
http://www.giref.ulaval.ca/~cmpgiref/petsc-master-debug/2016.03.09.02h00m02s_default.log
http://www.giref.ulaval.ca/~cmpgiref/petsc-master-debug/2016.03.09.02h00m02s_RDict.log
http://www.giref.ulaval.ca/~cmpgiref/petsc-master-debug/2016.03.09.02h00m02s_make.log
http://www.giref.ulaval.ca/~cmpgiref/petsc-master-debug/2016.03.09.02h00m02s_CMakeLists.txt

And the log from our scripts, with "env" output at the beginning:

http://www.giref.ulaval.ca/~cmpgiref/petsc-master-debug/2016.03.09.02h00m02s.script.log

May I launch a the make install with some debugging variables activated to see more output that might help?

I can also launch the scripts in a while loop until it bug again to help dig this problem...

Thanks,

Eric

Le 16-03-03 20:35, Éric Chamberland a écrit :
> Hi, it's weird.  I launched my compilation script on another computer, 
> and everything was fine with the petsc-master.tar.gz... So I launched 
> it on the faulty computer, but it installed with succes... however, 
> there is no "default.log" file after the make procedure...?? Here are 
> the non-problematic logs: 
> http://www.giref.ulaval.ca/~cmpgiref/petsc-master-debug/2016.03.03.18h22m37s_configure.log 
> http://www.giref.ulaval.ca/~cmpgiref/petsc-master-debug/2016.03.03.18h22m37s_RDict.log 
> What can I say? Some thoughts: - All this is completely automated by a 
> crontab... so the environment variables shall be the same each 
> night... - configure+make+make install are launched by the script 
> without any "human pause" in between the end of each command:  if 
> there is anything launched in the background by make, maybe it is not 
> ended before the beginning of make install? I saw that there are also 
> "make.log", "build.log" ".nagged", CMakeLists.txt files generated: do 
> they should be inspected? Unfortunately, be runing the script again, I 
> erased the faulty petsc directory... so I am only left with the 
> configure.log+RDict.log+default.log I put on the web page... :/ 
> Thanks, Eric Le 16-03-03 17:29, Satish Balay a écrit :
>> I tried a few builds - but I can't reproduce. So I do not know whats 
>> going on here [and what the trigger is].. Tried a configure command 
>> as close as possible to this one.. [and used the nightlytarball 
>> aswell] balay at es^/scratch/balay/petsc-master $ ./configure 
>> --prefix=/scratch/balay/petsc-install --with-mpi-compilers=1 
>> --with-make-np=12 --with-shared-libraries=1 --download-openmpi 
>> --with-debugging=yes --with-mkl_pardiso=1 --with-mkl_pardiso-dir=$MKL 
>> --download-ml=yes --download-mumps=yes --download-superlu=yes 
>> --download-superlu_dist=yes --download-parmetis=yes 
>> --download-ptscotch=yes --download-metis=yes 
>> --download-suitesparse=yes --download-hypre=yes --with-scalapack=1 
>> --with-scalapack-include=$MKL/include 
>> --with-scalapack-lib="-lmkl_scalapack_lp64 -lmkl_blacs_openmpi_lp64" 
>> --with-blas-lapack-dir=$MKL --download-cmake Satish On Thu, 3 Mar 
>> 2016, Matthew Knepley wrote:
>>> On Thu, Mar 3, 2016 at 5:42 AM, Éric Chamberland < 
>>> Eric.Chamberland at giref.ulaval.ca> wrote:
>>>> Hi, since last night, with petsc-master.tar.gz, it seems that "make 
>>>> install" is broken. Please have a look at these errors: 
>>> Satish, do you see this? We had something like this before, but it 
>>> was fixed.    Matt
>>>> ==================== ...           CC 
>>>> arch-linux2-c-debug/obj/src/tao/leastsquares/impls/pounders/gqt.o 
>>>>       CLINKER 
>>>> /pmi/cmpbib/compilation_BIB_gcc_redhat_petsc-master_debug/COMPILE_AUTO/petsc-master-debug/arch-linux2-c-debug/lib/libpetsc.so.3.06.3 
>>>> gmake[2]: Leaving directory 
>>>> `/pmi/cmpbib/compilation_BIB_gcc_redhat_petsc-master_debug/COMPILE_AUTO/petsc-master-debug' 
>>>> ========================================= gmake[1]: Leaving 
>>>> directory 
>>>> `/pmi/cmpbib/compilation_BIB_gcc_redhat_petsc-master_debug/COMPILE_AUTO/petsc-master-debug' 
>>>> Now to install the libraries do: make 
>>>> PETSC_DIR=/pmi/cmpbib/compilation_BIB_gcc_redhat_petsc-master_debug/COMPILE_AUTO/petsc-master-debug 
>>>> PETSC_ARCH=arch-linux2-c-debug install 
>>>> ========================================= *** Using 
>>>> PETSC_DIR=/pmi/cmpbib/compilation_BIB_gcc_redhat_petsc-master_debug/COMPILE_AUTO/petsc-master-debug 
>>>> PETSC_ARCH=arch-linux2-c-debug ***      No cached configure in 
>>>> RDict at 
>>>> /pmi/cmpbib/compilation_BIB_gcc_redhat_petsc-master_debug/COMPILE_AUTO/petsc-master-debug/arch-linux2-c-debug/lib/petsc/conf/RDict.db 
>>>> Traceback (most recent call last):    File "./config/install.py", 
>>>> line 380, in <module>      Installer(sys.argv[1:]).run()    File 
>>>> "./config/install.py", line 373, in run      self.runsetup()    
>>>> File "./config/install.py", line 334, in runsetup      self.setup() 
>>>>    File "./config/install.py", line 60, in setup      
>>>> self.setupModules()    File "./config/install.py", line 51, in 
>>>> setupModules      self.setCompilers  = 
>>>> self.framework.require('config.setCompilers',         None) 
>>>> AttributeError: 'NoneType' object has no attribute 'require' make: 
>>>> *** [install] Error 1 ================================= Some useful 
>>>> logs: 
>>>> http://www.giref.ulaval.ca/~cmpgiref/petsc-master-debug/configure_2016.03.03.02h00m01s.loghttp://www.giref.ulaval.ca/~cmpgiref/petsc-master-debug/2016.03.03.02h00m01s_RDict.loghttp://www.giref.ulaval.ca/~cmpgiref/petsc-master-debug/2016.03.03.02h00m01s_default.log 
>>>> Thanks! Eric Le 16-02-27 15:36, Barry Smith a écrit :     PETSc 
>>>> Users,      We are planning the PETSc release 3.7 shortly. If you 
>>>> know of any bugs that need to be fixed or enhancements added before 
>>>> the release please let us know.      You can think of the master 
>>>> branch of the PETSc repository obtainable with git clone 
>>>> https://bitbucket.org/petsc/petsc petsc as a release candidate for 
>>>> 3.7. Changes for the release are listed at 
>>>> http://www.mcs.anl.gov/petsc/documentation/changes/dev.html     
>>>> Thanks     Barry 
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.mcs.anl.gov/pipermail/petsc-dev/attachments/20160309/eb363987/attachment.html>