Open MPI logo

Open MPI User's Mailing List Archives

  |   Home   |   Support   |   FAQ   |   all Open MPI User's mailing list

From: Bas van der Vlies (basv_at_[hidden])
Date: 2007-04-04 07:59:19


Jeff Squyres wrote:
> On Apr 4, 2007, at 4:28 AM, Bas van der Vlies wrote:
>
>> Is the fix in trunk or also in the nighly build release. When i
>> download
>> the trunk version ./autogen.sh fails.
>
> You only need to use autogen.sh when building from an SVN checkout.
> Did you follow the instructions for SVN builds listed here:
>
> http://www.open-mpi.org/svn/building.php
>
Yes i get this error message:

*** Running GNU tools
[Running] autom4te --language=m4sh ompi_get_version.m4sh -o
ompi_get_version.sh
[Running] aclocal
[Running] autoheader
[Running] autoconf
[Running] libtoolize --automake --copy --ltdl
Adjusting libltdl for OMPI :-(
   -- patching for argz bugfix in libtool 1.5
      ==> your libtool doesn't need this! yay!
   -- patching 64-bit OS X bug in ltmain.sh
      ==> your libtool doesn't need this! yay!
   -- RTLD_GLOBAL in libltdl
      ==> your libltdl doesn't need this! yay!
   -- patching configure for broken -c/-o compiler test
[Running] automake --foreign -a --copy --include-deps
configure.ac:51: require Automake 1.10, but have 1.9.6

-------------------------------------------------------------------------
It seems that the execution of "automake --foreign -a --copy
--include-deps" has failed. See above for
the specific error message that caused it to abort.

This *MAY* be caused by an older version of one of the required
packages. Please make sure you are using at least the following
versions:

GNU Autoconf 2.59
GNU Automake 1.9.6
GNU Libtool 1.5.16
-------------------------------------------------------------------------

>> If i use the nighyly build version of openmpi-1.2.1a0r14212.tar.gz. We
>> still observe a memory leak. Is this the right version?
>
> Yes, this version has Mohamad's fixes in it. Can you be specific
> about the memory leak that you are seeing? Do you have some valgrind
> output or something similar?
>
I do not have valgrind output. What i have done is run this job on 4
nodes with two cpus (8 tasks) and run it for a long period and monitor
the memory usage with top. What we see is that it consumes more and more
memory.

-- 
********************************************************************
*                                                                  *
*  Bas van der Vlies                     e-mail: basv_at_[hidden]      *
*  SARA - Academic Computing Services    phone:  +31 20 592 8012   *
*  Kruislaan 415                         fax:    +31 20 6683167    *
*  1098 SJ Amsterdam                                               *
*                                                                  *
********************************************************************