Messages in this thread | | | Subject | strange core dumps | Date | Wed, 23 Oct 1996 09:17:44 +0200 | From | Axel Kohlmeyer <> |
| |
hi everybody,
can anybody explain to me what this kernel message is supposed to mean?
Oct 21 15:41:37 grapevine kernel: elf_core_dump: file.f_pos (172032) != offset (4096) Oct 21 23:36:32 grapevine kernel: elf_core_dump: file.f_pos (172032) != offset (4096) Oct 22 08:15:23 grapevine kernel: elf_core_dump: file.f_pos (172032) != offset (4096)
we're getting tons of them caused by our molecular dynamics simulation program. the program is (mostly) written in fortran and compiled with f2c/gcc. what really confuses me is the fact that the core dump happens AFTER the program has terminated correctly and only when it is started from our queueing system. if you start the program interactively or repeatedly from a script: no core dump.
everything else is running just fine. does anyone know, where to start looking for the CAUSE of these messages? is this a fault of f2c, gcc, binutils, libc or the kernel? of course, it would be easy to comment out the corresponding printk, but i'd like to find out, what is happening.
here's the usual info kernel is: 2.0.21 (but the kernel version doesn't seem to matter) gcc: 2.7.2 binutils: 2.6.0.14 libc: 2.3.12 f2c: VERSION 19960301
thanks in advance, axel.
========================================================================= Axel Kohlmeyer email: axel.kohlmeyer@chemie.uni-ulm.de Abteilung fuer Theoretische Chemie Universitaet Ulm D-89069 Ulm/Donau =========================================================================
| |