[Clas_offline] CERNLIB in 64bit ?

[Cole Smith]

Kijun, read this link:

https://userweb.jlab.org/~brads/cernlib-20061220/cernlib-20061220+dfsg3/debian/aaREADME.64-bit

COMIS is not 64-bit safe if you want to dynamically link functions.  You 
might try 'call file.f' instead of 'call file.f77'.  The former is slower
but doesn't try load a sharable library like the latter.

Cole


On Wed, 21 Nov 2012, Kijun wrote:

>
> Hello everyone,
>
> I am currently having a problem with CERN lib under running PAW/ROOT
> under 64bit machine (ifarm).
> I don't have any problem for simple plotting.
> However, I try to run the analysis code using PAW/ROOT interactive mode,
> I got an error.
> The error seems a mis-linking cernlib or something like that or maybe
> other issue.
> The older one(32bit: ifarml5) does not have this problem.
> Does any one have an idea how I can resolve this ? Here is the error
> message that I got.
>
>
> ifarm12s01 <31> ./npip_run01fake.exe -r0
> Test output ::  /u/site/cernlib/x86_64_rhel6/2005/bin/pawX11 -b
> gopip_37673.for.kumac
>  /u/site/cernlib/x86_64_rhel6/2005/bin/pawX11 -b gopip_37673.for.kumac
> Version 1.29/04 of HIGZ started
>  e1f_ana_npip_bg_fine_for.f 0 0 1
>  All events in the N-tuples will be processed
>  All files will be processed    Name of the function ->
> e1f_ana_npip_bg_fine_for.f
>  Unit   1 is not active
>  Startfile 1 21
>  Proccessing file  1 --> nt/ntp_eposorneg_37673_pass1.a00.hbook
>
>  *** Break *** Segmentation violation
>  Traceq lun = 0, level = 99
>
>  TRACEQ.  In-line trace-back still not available.
> Signal 11 caught without longjmp target
>
>
>
> Thank you in advance.
> Sincerely yours,
> Kijun
> _______________________________________________
> Clas_offline mailing list
> Clas_offline at jlab.org
> https://mailman.jlab.org/mailman/listinfo/clas_offline
>