[FLASH-USERS] Running Flash4 with OpenMPI+Intel

Michael Sabino michael.rocco.sabino at gmail.com
Fri May 24 22:16:00 EDT 2013


It looks like pointer memory allocation is failing for OpenMPI. I would try
a different version of OpenMPI.
Can you tell me what instruction it is attempting to execute at
opal_memory_ptmalloc2_int_malloc()+(~0x260-0x2aa) and
opal_memory_ptmalloc2_malloc+(~0x10-0x59)?
You can enter the offset in hte, or use gdb.

Http://hte(dot)sourceforge(dot)net

Thanks,

Michael Sabino

On Fri, May 24, 2013 at 11:05 AM, Asif ud-Doula <auu4 at psu.edu> wrote:

> Hi Everyone,
> I am having some trouble running the latest version of Flash on linux
> cluster with OpenMPI 1.6.1 version along with Intel
> (composer_xe_2011_sp1.8.27). There seems to be some memory allocation
> issues. Note that it compiles just fine.
>
>
> I have tried the standard ./setup Sedov -auto -2d and tried to run it with
> e.g. 4 processes with flash.par virtually unchanged. Here is the error I
> receive when I run it in an interactive mode:
>
> RuntimeParameters_read:  ignoring unknown parameter "order"...
>  RuntimeParameters_read:  ignoring unknown parameter "slopeLimiter"...
>  RuntimeParameters_read:  ignoring unknown parameter "LimitedSlopeBeta"...
>  RuntimeParameters_read:  ignoring unknown parameter "use_avisc"...
>  RuntimeParameters_read:  ignoring unknown parameter "use_flattening"...
>  RuntimeParameters_read:  ignoring unknown parameter "use_upwindTVD"...
>  RuntimeParameters_read:  ignoring unknown parameter "RiemannSolver"...
>  RuntimeParameters_read:  ignoring unknown parameter "entropy"...
>  RuntimeParameters_read:  ignoring unknown parameter "shockDetect"...
>
> flash4:27361 terminated with signal 11 at PC=2b66b15f539a SP=7fff34566480.
>  Backtrace:
>
> flash4:27362 terminated with signal 11 at PC=2b506c4cd39a SP=7fffd1113480.
>  Backtrace:
>
> flash4:27363 terminated with signal 11 at PC=2b812071539a SP=7fffd2f31780.
>  Backtrace:
> /opt/shared/openmpi/1.6.1-**intel64/lib/libmpi.so.1(opal_**
> memory_ptmalloc2_int_malloc+**0x2aa)[0x2b506c4cd39a]
> /opt/shared/openmpi/1.6.1-**intel64/lib/libmpi.so.1(opal_**
> memory_ptmalloc2_int_malloc+**0x2aa)[0x2b812071539a]
> /opt/shared/openmpi/1.6.1-**intel64/lib/libmpi.so.1(opal_**
> memory_ptmalloc2_malloc+0x59)[**0x2b506c4d0659]
> /opt/shared/openmpi/1.6.1-**intel64/lib/libmpi.so.1(opal_**
> memory_ptmalloc2_malloc+0x59)[**0x2b8120718659]
> /opt/shared/openmpi/1.6.1-**intel64/lib/libmpi.so.1(+**
> 0x110346)[0x2b8120718346]
> /opt/shared/openmpi/1.6.1-**intel64/lib/libmpi.so.1(+**
> 0x110346)[0x2b506c4d0346]
> /home/1232/FLASH4/object/**flash4(for_allocate+0x164)[**0x6a4544]
> /home/1232/FLASH4/object/**flash4(for_allocate+0x164)[**0x6a4544]
> /home/1232/FLASH4/object/**flash4(namevaluell_data_mp_**
> namevaluell_addreal_+0x35c)[**0x632bbc]
> /home/1232/FLASH4/object/**flash4(namevaluell_data_mp_**
> namevaluell_addreal_+0x35c)[**0x632bbc]
> /home/1232/FLASH4/object/**flash4(namevaluell_bcast_+**0x1e6f)[0x631f5f]
> /home/1232/FLASH4/object/**flash4(namevaluell_bcast_+**0x1e6f)[0x631f5f]
> /home/1232/FLASH4/object/**flash4(runtimeparameters_**
> bcast_+0x12)[0x46edb2]
> /home/1232/FLASH4/object/**flash4(runtimeparameters_**
> bcast_+0x12)[0x46edb2]
> /home/1232/FLASH4/object/**flash4(runtimeparameters_init_**
> +0x265)[0x46f3c5]
> /home/1232/FLASH4/object/**flash4(runtimeparameters_init_**
> +0x265)[0x46f3c5]
> /home/1232/FLASH4/object/**flash4(driver_initflash_+0x35)**[0x430245]
> /home/1232/FLASH4/object/**flash4(driver_initflash_+0x35)**[0x430245]
> /home/1232/FLASH4/object/**flash4(MAIN__+0x42)[0x435802]
> /home/1232/FLASH4/object/**flash4(MAIN__+0x42)[0x435802]
> /home/1232/FLASH4/object/**flash4(main+0x3c)[0x42b1dc]
> /home/1232/FLASH4/object/**flash4(main+0x3c)[0x42b1dc]
> /lib64/libc.so.6(__libc_start_**main+0xfd)[0x2b506db26cdd]
> /lib64/libc.so.6(__libc_start_**main+0xfd)[0x2b8121d6ecdd]
> /home/1232/FLASH4/object/**flash4[0x42b0d9]
> /home/1232/FLASH4/object/**flash4[0x42b0d9]
> /opt/shared/openmpi/1.6.1-**intel64/lib/libmpi.so.1(opal_**
> memory_ptmalloc2_int_malloc+**0x2aa)[0x2b66b15f539a]
> /opt/shared/openmpi/1.6.1-**intel64/lib/libmpi.so.1(opal_**
> memory_ptmalloc2_malloc+0x59)[**0x2b66b15f8659]
> /opt/shared/openmpi/1.6.1-**intel64/lib/libmpi.so.1(+**
> 0x110346)[0x2b66b15f8346]
> /home/1232/FLASH4/object/**flash4(for_allocate+0x164)[**0x6a4544]
> /home/1232/FLASH4/object/**flash4(namevaluell_data_mp_**
> namevaluell_addreal_+0x35c)[**0x632bbc]
> /home/1232/FLASH4/object/**flash4(namevaluell_bcast_+**0x1e6f)[0x631f5f]
> /home/1232/FLASH4/object/**flash4(runtimeparameters_**
> bcast_+0x12)[0x46edb2]
> /home/1232/FLASH4/object/**flash4(runtimeparameters_init_**
> +0x265)[0x46f3c5]
> /home/1232/FLASH4/object/**flash4(driver_initflash_+0x35)**[0x430245]
> /home/1232/FLASH4/object/**flash4(MAIN__+0x42)[0x435802]
> /home/1232/FLASH4/object/**flash4(main+0x3c)[0x42b1dc]
> /lib64/libc.so.6(__libc_start_**main+0xfd)[0x2b66b2c4ecdd]
> /home/1232/FLASH4/object/**flash4[0x42b0d9]
>
>
> Just to provide you with more information about the system:
> * uname -a
> Linux n106 2.6.32-279.19.1.el6.x86_64 #1 SMP Tue Dec 18 17:22:54 CST 2012
> x86_64 x86_64 x86_64 GNU/Linux
>
> * ldd flash4
>         linux-vdso.so.1 =>  (0x00007fff8d5d0000)
>         libhdf5.so.7 => /opt/shared/hdf5/1.8.8-**intel64/lib/libhdf5.so.7
> (0x00002ab4a4f58000)
>         libmpi.so.1 => /opt/shared/openmpi/1.6.1-**intel64/lib/libmpi.so.1
> (0x00002ab4a54f8000)
>         libm.so.6 => /lib64/libm.so.6 (0x00002ab4a5908000)
>         libpthread.so.0 => /lib64/libpthread.so.0 (0x00002ab4a5b90000)
>         libz.so.1 => /lib64/libz.so.1 (0x00002ab4a5db0000)
>         libmpi_f90.so.1 => /opt/shared/openmpi/1.6.1-**intel64/lib/libmpi_f90.so.1
> (0x00002ab4a5fc8000)
>         libmpi_f77.so.1 => /opt/shared/openmpi/1.6.1-**intel64/lib/libmpi_f77.so.1
> (0x00002ab4a61d0000)
>         libdl.so.2 => /lib64/libdl.so.2 (0x00002ab4a6408000)
>         librt.so.1 => /lib64/librt.so.1 (0x00002ab4a6610000)
>         libnsl.so.1 => /lib64/libnsl.so.1 (0x00002ab4a6818000)
>         libutil.so.1 => /lib64/libutil.so.1 (0x00002ab4a6a38000)
>         libc.so.6 => /lib64/libc.so.6 (0x00002ab4a6c40000)
>         libgcc_s.so.1 => /lib64/libgcc_s.so.1 (0x00002ab4a6fd8000)
>         libimf.so => /home/software/intel/composer_**
> xe_2011_sp1.8.273/compiler/**lib/intel64/libimf.so (0x00002ab4a71f0000)
>         libsvml.so => /home/software/intel/composer_**
> xe_2011_sp1.8.273/compiler/**lib/intel64/libsvml.so (0x00002ab4a75c0000)
>         libintlc.so.5 => /home/software/intel/composer_**
> xe_2011_sp1.8.273/compiler/**lib/intel64/libintlc.so.5
> (0x00002ab4a7d38000)
>         /lib64/ld-linux-x86-64.so.2 (0x00002ab4a4d30000)
>         libifport.so.5 => /home/software/intel/composer_**
> xe_2011_sp1.8.273/compiler/**lib/intel64/libifport.so.5
> (0x00002ab4a7e88000)
>         libifcore.so.5 => /home/software/intel/composer_**
> xe_2011_sp1.8.273/compiler/**lib/intel64/libifcore.so.5
> (0x00002ab4a7fc0000)
>         libifcoremt.so.5 => /home/software/intel/composer_**
> xe_2011_sp1.8.273/compiler/**lib/intel64/libifcoremt.so.5
> (0x00002ab4a8208000)
>
>
> I wonder if you have encountered a similar issue, or perhaps have a
> suggestion for me. Thank you in advance for your time and help,
>
> Asif
>
>
> --
> ______________________________**__
> Asif ud-Doula
> Assistant Professor of Physics
> Penn State Worthington Scranton
> 120 Ridge View Drive
> Dunmore, PA 18512
> USA
> tel. +1-570-963-2582
> ______________________________**__
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://flash.rochester.edu/pipermail/flash-users/attachments/20130524/88978db4/attachment.htm>


More information about the flash-users mailing list