Re: [OMPI users] SilverStorm IB

[Troy Telford]

Just a quick follow-up:  Now even /without/ the debug flag, it just  
works.  I'm wondering if there was some sort of weird action going on with  
the subnet manager... but that's not Open MPI's problem.  Still, I'm quite  
pleased that it works, with almost no effort on my (or hopefully anybody  
else's) part.

On Wed, 12 Apr 2006 10:56:24 -0600, Troy Telford  
<ttelf...@linuxnetworx.com> wrote:

On Wed, 12 Apr 2006 10:04:18 -0600, Brian Barrett <brbar...@open-mpi.org>
wrote:

We've tested against the SilverStorm drivers for OS X with success,
but I don't think anyone has tried the Linux drivers.  A quick poll
of the developers show that none of us has access to a Linux cluster
using the SilverStorm stack, so we can't really look too deeply at
the problem.  If you compile with --enable-debug, are there any error
messages that show up?

O.o ROFL.  Schrödinger's bug strikes again.  Turning on debugging has
changed the outcome.

I re-compiled, but this time with the following (additional) configure
options:
        --enable-debug
        --enable-mem-debug
        --enable-mem-profile

Now it is working.

(I'll try it again with only the --enable-debug, and no mem-debug,
mem-profile).  Still...

One other thing I did is re-compile the benchmark (in this case, IMB, the
Intel MPI Benchmark).  The original benchmark (compiled with Open MPI,  
but
that particular Open MPI was not compiled with the above options) does
generate some errors, but I may just need to take its advice and  
re-link...

***Begin Errors***
IMB-MPI1.ss: Symbol `ompi_mpi_errhandler_null' has different size in
shared object, consider re-linking
IMB-MPI1.ss: Symbol `ompi_mpi_comm_null' has different size in shared
object, consider re-linking
IMB-MPI1.ss: Symbol `ompi_mpi_datatype_null' has different size in shared
object, consider re-linking
IMB-MPI1.ss: Symbol `ompi_mpi_int' has different size in shared object,
consider re-linking
IMB-MPI1.ss: Symbol `ompi_mpi_byte' has different size in shared object,
consider re-linking
IMB-MPI1.ss: Symbol `ompi_mpi_op_sum' has different size in shared  
object,
consider re-linking
IMB-MPI1.ss: Symbol `ompi_mpi_float' has different size in shared object,
consider re-linking
IMB-MPI1.ss: Symbol `ompi_mpi_comm_world' has different size in shared
object, consider re-linking
IMB-MPI1.ss: Symbol `ompi_mpi_double' has different size in shared  
object,
consider re-linking
IMB-MPI1.ss: Symbol `ompi_mpi_op_null' has different size in shared
object, consider re-linking
IMB-MPI1.ss: Symbol `ompi_mpi_comm_self' has different size in shared
object, consider re-linking
IMB-MPI1.ss: Symbol `ompi_mpi_errhandler_null' has different size in
shared object, consider re-linking
IMB-MPI1.ss: Symbol `ompi_mpi_comm_null' has different size in shared
object, consider re-linking
IMB-MPI1.ss: Symbol `ompi_mpi_datatype_null' has different size in shared
object, consider re-linking
IMB-MPI1.ss: Symbol `ompi_mpi_int' has different size in shared object,
consider re-linking
IMB-MPI1.ss: Symbol `ompi_mpi_byte' has different size in shared object,
consider re-linking
IMB-MPI1.ss: Symbol `ompi_mpi_op_sum' has different size in shared  
object,
consider re-linking
IMB-MPI1.ss: Symbol `ompi_mpi_float' has different size in shared object,
consider re-linking
IMB-MPI1.ss: Symbol `ompi_mpi_comm_world' has different size in shared
object, consider re-linking
IMB-MPI1.ss: Symbol `ompi_mpi_double' has different size in shared  
object,
consider re-linking
IMB-MPI1.ss: Symbol `ompi_mpi_op_null' has different size in shared
object, consider re-linking
IMB-MPI1.ss: Symbol `ompi_mpi_comm_self' has different size in shared
object, consider re-linking
IMB-MPI1.ss: Symbol `ompi_mpi_errhandler_null' has different size in
shared object, consider re-linking
IMB-MPI1.ss: Symbol `ompi_mpi_comm_null' has different size in shared
object, consider re-linking
IMB-MPI1.ss: Symbol `ompi_mpi_datatype_null' has different size in shared
object, consider re-linking
IMB-MPI1.ss: Symbol `ompi_mpi_int' has different size in shared object,
consider re-linking
IMB-MPI1.ss: Symbol `ompi_mpi_byte' has different size in shared object,
consider re-linking
IMB-MPI1.ss: Symbol `ompi_mpi_op_sum' has different size in shared  
object,
consider re-linking
IMB-MPI1.ss: Symbol `ompi_mpi_float' has different size in shared object,
consider re-linking
IMB-MPI1.ss: Symbol `ompi_mpi_comm_world' has different size in shared
object, consider re-linking
IMB-MPI1.ss: Symbol `ompi_mpi_double' has different size in shared  
object,
consider re-linking
IMB-MPI1.ss: Symbol `ompi_mpi_op_null' has different size in shared
object, consider re-linking
IMB-MPI1.ss: Symbol `ompi_mpi_comm_self' has different size in shared
object, consider re-linking
IMB-MPI1.ss: Symbol `ompi_mpi_errhandler_null' has different size in
shared object, consider re-linking
IMB-MPI1.ss: Symbol `ompi_mpi_comm_null' has different size in shared
object, consider re-linking
IMB-MPI1.ss: Symbol `ompi_mpi_datatype_null' has different size in shared
object, consider re-linking
IMB-MPI1.ss: Symbol `ompi_mpi_int' has different size in shared object,
consider re-linking
IMB-MPI1.ss: Symbol `ompi_mpi_byte' has different size in shared object,
consider re-linking
IMB-MPI1.ss: Symbol `ompi_mpi_op_sum' has different size in shared  
object,
consider re-linking
IMB-MPI1.ss: Symbol `ompi_mpi_float' has different size in shared object,
consider re-linking
IMB-MPI1.ss: Symbol `ompi_mpi_comm_world' has different size in shared
object, consider re-linking
IMB-MPI1.ss: Symbol `ompi_mpi_double' has different size in shared  
object,
consider re-linking
IMB-MPI1.ss: Symbol `ompi_mpi_op_null' has different size in shared
object, consider re-linking
IMB-MPI1.ss: Symbol `ompi_mpi_comm_self' has different size in shared
object, consider re-linking
IMB-MPI1.ss: Symbol `ompi_mpi_errhandler_null' has different size in
shared object, consider re-linking
IMB-MPI1.ss: Symbol `ompi_mpi_comm_null' has different size in shared
object, consider re-linking
IMB-MPI1.ss: Symbol `ompi_mpi_datatype_null' has different size in shared
object, consider re-linking
IMB-MPI1.ss: Symbol `ompi_mpi_int' has different size in shared object,
consider re-linking
IMB-MPI1.ss: Symbol `ompi_mpi_byte' has different size in shared object,
consider re-linking
IMB-MPI1.ss: Symbol `ompi_mpi_op_sum' has different size in shared  
object,
consider re-linking
IMB-MPI1.ss: Symbol `ompi_mpi_float' has different size in shared object,
consider re-linking
IMB-MPI1.ss: Symbol `ompi_mpi_comm_world' has different size in shared
object, consider re-linking
IMB-MPI1.ss: Symbol `ompi_mpi_double' has different size in shared  
object,
consider re-linking
IMB-MPI1.ss: Symbol `ompi_mpi_op_null' has different size in shared
object, consider re-linking
IMB-MPI1.ss: Symbol `ompi_mpi_comm_self' has different size in shared
object, consider re-linking
IMB-MPI1.ss: Symbol `ompi_mpi_errhandler_null' has different size in
shared object, consider re-linking
IMB-MPI1.ss: Symbol `ompi_mpi_comm_null' has different size in shared
object, consider re-linking
IMB-MPI1.ss: Symbol `ompi_mpi_datatype_null' has different size in shared
object, consider re-linking
IMB-MPI1.ss: Symbol `ompi_mpi_int' has different size in shared object,
consider re-linking
IMB-MPI1.ss: Symbol `ompi_mpi_byte' has different size in shared object,
consider re-linking
IMB-MPI1.ss: Symbol `ompi_mpi_op_sum' has different size in shared  
object,
consider re-linking
IMB-MPI1.ss: Symbol `ompi_mpi_float' has different size in shared object,
consider re-linking
IMB-MPI1.ss: Symbol `ompi_mpi_comm_world' has different size in shared
object, consider re-linking
IMB-MPI1.ss: Symbol `ompi_mpi_double' has different size in shared  
object,
consider re-linking
IMB-MPI1.ss: Symbol `ompi_mpi_op_null' has different size in shared
object, consider re-linking
IMB-MPI1.ss: Symbol `ompi_mpi_comm_self' has different size in shared
object, consider re-linking
IMB-MPI1.ss: Symbol `ompi_mpi_errhandler_null' has different size in
shared object, consider re-linking
IMB-MPI1.ss: Symbol `ompi_mpi_comm_null' has different size in shared
object, consider re-linking
IMB-MPI1.ss: Symbol `ompi_mpi_datatype_null' has different size in shared
object, consider re-linking
IMB-MPI1.ss: Symbol `ompi_mpi_int' has different size in shared object,
consider re-linking
IMB-MPI1.ss: Symbol `ompi_mpi_byte' has different size in shared object,
consider re-linking
IMB-MPI1.ss: Symbol `ompi_mpi_op_sum' has different size in shared  
object,
consider re-linking
IMB-MPI1.ss: Symbol `ompi_mpi_float' has different size in shared object,
consider re-linking
IMB-MPI1.ss: Symbol `ompi_mpi_comm_world' has different size in shared
object, consider re-linking
IMB-MPI1.ss: Symbol `ompi_mpi_double' has different size in shared  
object,
consider re-linking
IMB-MPI1.ss: Symbol `ompi_mpi_op_null' has different size in shared
object, consider re-linking
IMB-MPI1.ss: Symbol `ompi_mpi_comm_self' has different size in shared
object, consider re-linking
IMB-MPI1.ss: Symbol `ompi_mpi_errhandler_null' has different size in
shared object, consider re-linking
IMB-MPI1.ss: Symbol `ompi_mpi_comm_null' has different size in shared
object, consider re-linking
IMB-MPI1.ss: Symbol `ompi_mpi_datatype_null' has different size in shared
object, consider re-linking
IMB-MPI1.ss: Symbol `ompi_mpi_int' has different size in shared object,
consider re-linking
IMB-MPI1.ss: Symbol `ompi_mpi_byte' has different size in shared object,
consider re-linking
IMB-MPI1.ss: Symbol `ompi_mpi_op_sum' has different size in shared  
object,
consider re-linking
IMB-MPI1.ss: Symbol `ompi_mpi_float' has different size in shared object,
consider re-linking
IMB-MPI1.ss: Symbol `ompi_mpi_comm_world' has different size in shared
object, consider re-linking
IMB-MPI1.ss: Symbol `ompi_mpi_double' has different size in shared  
object,
consider re-linking
IMB-MPI1.ss: Symbol `ompi_mpi_op_null' has different size in shared
object, consider re-linking
IMB-MPI1.ss: Symbol `ompi_mpi_comm_self' has different size in shared
object, consider re-linking
IMB-MPI1.ss: Symbol `ompi_mpi_errhandler_null' has different size in
shared object, consider re-linking
IMB-MPI1.ss: Symbol `ompi_mpi_comm_null' has different size in shared
object, consider re-linking
IMB-MPI1.ss: Symbol `ompi_mpi_datatype_null' has different size in shared
object, consider re-linking
IMB-MPI1.ss: Symbol `ompi_mpi_int' has different size in shared object,
consider re-linking
IMB-MPI1.ss: Symbol `ompi_mpi_byte' has different size in shared object,
consider re-linking
IMB-MPI1.ss: Symbol `ompi_mpi_op_sum' has different size in shared  
object,
consider re-linking
IMB-MPI1.ss: Symbol `ompi_mpi_float' has different size in shared object,
consider re-linking
IMB-MPI1.ss: Symbol `ompi_mpi_comm_world' has different size in shared
object, consider re-linking
IMB-MPI1.ss: Symbol `ompi_mpi_double' has different size in shared  
object,
consider re-linking
IMB-MPI1.ss: Symbol `ompi_mpi_op_null' has different size in shared
object, consider re-linking
IMB-MPI1.ss: Symbol `ompi_mpi_comm_self' has different size in shared
object, consider re-linking
IMB-MPI1.ss: Symbol `ompi_mpi_errhandler_null' has different size in
shared object, consider re-linking
IMB-MPI1.ss: Symbol `ompi_mpi_comm_null' has different size in shared
object, consider re-linking
IMB-MPI1.ss: Symbol `ompi_mpi_datatype_null' has different size in shared
object, consider re-linking
IMB-MPI1.ss: Symbol `ompi_mpi_int' has different size in shared object,
consider re-linking
IMB-MPI1.ss: Symbol `ompi_mpi_byte' has different size in shared object,
consider re-linking
IMB-MPI1.ss: Symbol `ompi_mpi_op_sum' has different size in shared  
object,
consider re-linking
IMB-MPI1.ss: Symbol `ompi_mpi_float' has different size in shared object,
consider re-linking
IMB-MPI1.ss: Symbol `ompi_mpi_comm_world' has different size in shared
object, consider re-linking
IMB-MPI1.ss: Symbol `ompi_mpi_double' has different size in shared  
object,
consider re-linking
IMB-MPI1.ss: Symbol `ompi_mpi_op_null' has different size in shared
object, consider re-linking
IMB-MPI1.ss: Symbol `ompi_mpi_comm_self' has different size in shared
object, consider re-linking
Signal:11 info.si_errno:0(Success) si_code:2(SEGV_ACCERR)
Failing at addr:0x2a99610600
Signal:11 info.si_errno:0(Success) si_code:1(SEGV_MAPERR)
Failing at addr:0xa8
****



--
Troy Telford
Linux Networx
ttelf...@linuxnetworx.com
(801) 649-1356