[Bioclusters] Opteron Perl64 segfault issues

[Joseph Landman]

Hi Nathan:

  I just ran into this myself.  I built the binaries on RedHat GinGin64
with the gcc-3.3 compiler.  I just tried a SUSE 8.2 platform, and had to
recompile the toolkit.  If you want to use the binary, I think you need
the relevant gcc tree installed.  I think I might have that here.  I
will look.

  For reasons that are not all that clear to me now, I had to use the
-m32 flag to generate a working binary on SUSE.  The -m32 forces the use
of the IA32 ABI, which means that you cannot use some of the nicer
features of the Opteron (large memory, 64 bit mode, etc).  I hope to
investigate this more.  

  A kind soul just gave me access to their opteron for some testing. 
Compiling with -m64 generates a binary, though upon completion of a test
run, it generates an ABORT.  Strace wasn't all that helpful here.  I am
trying to track this down.

  I can give you my -m32 binaries if you would like.  I would like to
get the -m64 binaries working (for SUSE).

Joe

On Fri, 2003-08-15 at 22:58, Nathan Siemers wrote:
> Hello Joe,
> 
> My default SUSE opteron configuration may not have the libs to run your 
> build:
> 
> 
> opt:/u1/gcgblast # formatdb -p F -i test &
> formatdb: /lib64/libc.so.6: version `GLIBC_2.3' not found (required by 
> formatdb)
> [1] 10941
> [1]   Exit 1                  formatdb -p F -i test
> opt:/u1/gcgblast # ldd /usr/local/bin/formatdb
> /usr/local/bin/formatdb: /lib64/libc.so.6: version `GLIBC_2.3' not found 
> (required by /usr/local/bin/formatdb)
>         libm.so.6 => /lib64/libm.so.6 (0x0000002a9566a000)
>         libc.so.6 => /lib64/libc.so.6 (0x0000002a957c2000)
>         /lib64/ld-linux-x86-64.so.2 => /lib64/ld-linux-x86-64.so.2 
> (0x0000002a95556000)
> 
> 
> I assume these are 64 bit executables?  Is there a newer libc needed?
> 
> Thanks for your (past and present) help with the opteron platform.
> 
>     Nathan
> 
> 
> 
> Joseph Landman wrote:
> 
> >Hi Doug:
> >
> >  These are 2.2.6.  The tree's VERSION (ncbi/VERSION) reads:
> >
> >[landman@squash.scalableinformatics.com:/big/t/ncbi]
> >8 >cat VERSION
> >Tue Apr 22 10:41:40 EDT 2003
> >
> >  Please let me know if you have problems with them.   This was a first
> >pass at getting working systems.
> >
> >Joe
> >
> >On Tue, 2003-08-12 at 20:17, Doug Shubert wrote:
> >  
> >
> >>Hello Joe,
> >>What version of ncbitools are you working with ?
> >>Doug
> >>
> >>Joe Landman wrote:
> >>
> >>    
> >>
> >>>Its up.  Please let me know if it works.  This is a snapshot from a
> >>>build done in June,  Download is
> >>>http://scalableinformatics.com/downloads/opteron.ncbitools.tar.gz .  It
> >>>is 29MB.
> >>>
> >>>Joe
> >>>
> >>>On Tue, 2003-08-12 at 15:46, Joe Landman wrote:
> >>>      
> >>>
> >>>>On Tue, 2003-08-12 at 13:11, Nathan Siemers wrote:
> >>>>        
> >>>>
> >>>>>That's great news, Joe.  Has this build passed any regression tests,
> >>>>>i.e. rigorous
> >>>>>comparison to output from blastall on other architectures?
> >>>>>          
> >>>>>
> >>>>I have not had an opteron to run on long enough to run my normal tests.
> >>>>It did generate the same results for my short tests.
> >>>>        
> >>>>
> >>>--
> >>>Joseph Landman, Ph.D
> >>>Scalable Informatics LLC,
> >>>email: landman@scalableinformatics.com
> >>>web  : http://scalableinformatics.com
> >>>phone: +1 734 612 4615
> >>>
> >>>_______________________________________________
> >>>Bioclusters maillist  -  Bioclusters@bioinformatics.org
> >>>https://bioinformatics.org/mailman/listinfo/bioclusters
> >>>      
> >>>
> >>_______________________________________________
> >>Bioclusters maillist  -  Bioclusters@bioinformatics.org
> >>https://bioinformatics.org/mailman/listinfo/bioclusters
> >>    
> >>
> 
> _______________________________________________
> Bioclusters maillist  -  Bioclusters@bioinformatics.org
> https://bioinformatics.org/mailman/listinfo/bioclusters
-- 
Joseph Landman, Ph.D
Scalable Informatics LLC
email: landman@scalableinformatics.com
  web: http://scalableinformatics.com
phone: +1 734 612 4615