Open MPI logo

Open MPI User's Mailing List Archives

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

From: Jeff Squyres (jsquyres_at_[hidden])
Date: 2007-01-29 20:59:42


I'm quite sure that we have since fixed the command line parsing
problem, and I *think* we fixed the mmap problem.

Is there any way that you can upgrade to v1.1.3?

On Jan 29, 2007, at 3:24 PM, Avishay Traeger wrote:

> Hello,
>
> I have just installed Open MPI 1.1 on a 64-bit FC6 machine using yum.
> The packages that were installed are:
> openmpi-devel-1.1-7.fc6
> openmpi-libs-1.1-7.fc6
> openmpi-1.1-7.fc6
>
> I tried running ompi_info, but it results in a segmentation fault.
> Running strace shows this at the end:
>
> mmap(NULL, 4294967296, PROT_READ|PROT_WRITE, MAP_PRIVATE|
> MAP_ANONYMOUS,
> -1, 0) = -1 ENOMEM (Cannot allocate memory)
> --- SIGSEGV (Segmentation fault) @ 0 (0) ---
> +++ killed by SIGSEGV +++
>
> The full output of ompi_info is:
> # ompi_info
> Open MPI: 1.1
> Open MPI SVN revision: r10477
> Open RTE: 1.1
> Open RTE SVN revision: r10477
> OPAL: 1.1
> OPAL SVN revision: r10477
> Prefix: /usr
> Configured architecture: x86_64-redhat-linux-gnu
> Configured by: brewbuilder
> Configured on: Fri Oct 13 14:34:07 EDT 2006
> Configure host: hs20-bc1-7.build.redhat.com
> Built by: brewbuilder
> Built on: Fri Oct 13 14:44:39 EDT 2006
> Built host: hs20-bc1-7.build.redhat.com
> C bindings: yes
> C++ bindings: yes
> Fortran77 bindings: yes (single underscore)
> Fortran90 bindings: yes
> Fortran90 bindings size: small
> C compiler: gcc
> C compiler absolute: /usr/bin/gcc
> C++ compiler: g++
> C++ compiler absolute: /usr/bin/g++
> Fortran77 compiler: gfortran
> Fortran77 compiler abs: /usr/bin/gfortran
> Fortran90 compiler: gfortran
> Fortran90 compiler abs: /usr/bin/gfortran
> C profiling: yes
> C++ profiling: yes
> Fortran77 profiling: yes
> Fortran90 profiling: yes
> C++ exceptions: no
> Thread support: posix (mpi: no, progress: no)
> Internal debug support: no
> MPI parameter check: runtime
> Memory profiling support: no
> Memory debugging support: no
> libltdl support: yes
> Segmentation fault
>
> It seems that at this point in the program, it tries to map 4GB of
> memory, which results in ENOMEM. I'm guessing that the return
> value of
> mmap isn't checked, which results in this segmentation fault.
>
> Also, I tried running "mpirun", and the output was:
> # mpirun
> *** buffer overflow detected ***: mpirun terminated
> ======= Backtrace: =========
> /lib64/libc.so.6(__chk_fail+0x2f)[0x3f59ce0dff]
> /lib64/libc.so.6[0x3f59ce065b]
> /lib64/libc.so.6(__snprintf_chk+0x7b)[0x3f59ce052b]
> /usr/lib64/openmpi/libopal.so.0(opal_cmd_line_get_usage_msg
> +0x20a)[0x304901963a]
> mpirun[0x403c7c]
> mpirun(orterun+0xa4)[0x40260c]
> mpirun(main+0x1b)[0x402563]
> /lib64/libc.so.6(__libc_start_main+0xf4)[0x3f59c1da44]
> mpirun[0x4024b9]
>
> It also included a "Memory map", which I left out.
>
> Any suggestions?
>
> Thanks in advance,
> Avishay
>
> _______________________________________________
> users mailing list
> users_at_[hidden]
> http://www.open-mpi.org/mailman/listinfo.cgi/users

-- 
Jeff Squyres
Server Virtualization Business Unit
Cisco Systems