[Spice_discussion] 64bit Linux
peje at irf.se
Thu Nov 17 12:46:27 PST 2005
Ed Wright wrote:
> To: Pär-Ola Nilsson
> From: Ed Wright
> On Nov 17, 2005, at 4:50 AM, Pär-Ola Nilsson wrote:
>> I'm trying to compile the cspice library on a 64bit Debian Linux
>> system. compiling the library and executables works flawless but
>> running does not.
>> commnt -r VEX/kernels/ck/ATPV_P______051109_00001.BC
>> I get:
> Thank you for including the error message.
>> Toolkit version: N0058
>> This version of SPICELIB was originally packaged by NAIF for PC
>> hardware using
>> LINUX with the GCC compiler. This environment has a binary file
>> format of
>> LTL-IEEE; however the software is running on an environment that has
>> a binary
>> file format of UNKNOWN. This is a severe problem and may be because the
>> software package was intended for use on a different computer
>> system. It also
>> may be the result of an improper port; please contact NAIF.
>> A traceback follows. The name of the highest level module is first.
>> COMMNT --> CLCOMMNT --> DAFOPR --> ZZDDHOPN --> ZZPLTCHK
>> Is there an easy way to fix this or do I have to wait for the next
> I assume you used the standard NAIF Linux toolkit for this compile.
> SPICELIB includes platform/architecture dependent routines with build
> scripts cast for particular compiler. I have little experience with
> 64 bit debian and so cannot competently diagnose the problem. An
> educate guess is that the record size coded into SPICELIB differs
> from that used by the 64 bit debian.
> I'll try to obtain access to a 64 bit debian for some tests.
>> Pär-Ola Nilsson
>> Swedish Institute of Space Physics, Teknikhuset, SE-901 87 Umeå, SWEDEN
>> Web: www.umea.irf.se, Phone: +46-(0)90-7869206, Fax: +46-(0)90-7869203
> As always,
> Ed Wright
> ed.wright at jpl.nasa.gov
I've mucked about in the source and found out that 64bit linux uses 64
the same way as the ALPHA processor.
I tried to replace all
# ifdef CSPICE_ALPHA_DIGITAL_UNIX
#if defined( CSPICE_ALPHA_DIGITAL_UNIX) ||defined(CSPICE_PC_LINUX64)
and a corresponding
and gained some succes.
The limited tests I've done suggests it might work.
Will do more testing later.
Is there some kind of validation tests available one could try??
This is just a pet project so far.
Swedish Institute of Space Physics, Teknikhuset, SE-901 87 Umeå, SWEDEN
Web: www.umea.irf.se, Phone: +46-(0)90-7869206, Fax: +46-(0)90-7869203
More information about the Spice_discussion