[ADMB Users] Memory Dump while running an admb program

Jared Blashka evilamarant7x at gmail.com
Wed Jan 5 06:55:36 PST 2011

While analyzing a data set I encounted what appears to be a memory dump.
I've included the terminal output at the end. I've been able to recreate the
issue on two different machines, one which is RHEL 5.4 and the other which
is Fedora 6 (old, yes I know). I believe that it's an issue with gcc, as the
same version is on both the problem machines (4.1.2 20070626 (Red Hat
4.1.2-13)). Both machines are also 32-bit. I've complied and run the same
template on a 64-bit 10.10 ubuntu laptop with gcc 4.4 and the program runs
normally. What strikes me as interesting is that the dump doesn't occur with
two other data sets that i've used; these two run without any issues
regardless of the machine. I've included the compiled program, the template,
and the three datasets in the zip file below.


Anyways, I was just wondering if anyone had any insight into the situation.
I've tried to upgrade gcc to 4.4, but the upgrade isn't officially available
in the repository for those versions of linux, so I'm out of luck there. I
then tried to install it as a second instance of gcc (as gcc4.4), compiled
the program specifying gcc4.4 to be used and then tried to run the result,
but I was only greeted with errors specifying that certain functions were
unavailable. I'm guessing it was trying to run the gcc4.4 compiled program
with the 4.1 libraries that were found by default.

[root at localhost rwh]# ./single
*** glibc detected *** ./single: free(): invalid next size (fast):
0x092eb808 ***
======= Backtrace: =========
======= Memory map: ========
00110000-00111000 r-xp 00110000 00:00 0          [vdso]
001a1000-00281000 r-xp 00000000 fd:00 336611     /usr/lib/libstdc++.so.6.0.8
00281000-00285000 r-xp 000df000 fd:00 336611     /usr/lib/libstdc++.so.6.0.8
00285000-00286000 rwxp 000e3000 fd:00 336611     /usr/lib/libstdc++.so.6.0.8
00286000-0028c000 rwxp 00286000 00:00 0
0095e000-00977000 r-xp 00000000 fd:00 946562     /lib/ld-2.5.so
00977000-00978000 r-xp 00019000 fd:00 946562     /lib/ld-2.5.so
00978000-00979000 rwxp 0001a000 fd:00 946562     /lib/ld-2.5.so
0097b000-00ab5000 r-xp 00000000 fd:00 1338251    /lib/libc-2.5.so
00ab5000-00ab7000 r-xp 0013a000 fd:00 1338251    /lib/libc-2.5.so
00ab7000-00ab8000 rwxp 0013c000 fd:00 1338251    /lib/libc-2.5.so
00ab8000-00abb000 rwxp 00ab8000 00:00 0
00abd000-00ae2000 r-xp 00000000 fd:00 1338256    /lib/libm-2.5.so
00ae2000-00ae3000 r-xp 00024000 fd:00 1338256    /lib/libm-2.5.so
00ae3000-00ae4000 rwxp 00025000 fd:00 1338256    /lib/libm-2.5.so
00d64000-00d6f000 r-xp 00000000 fd:00 946585
00d6f000-00d70000 rwxp 0000a000 fd:00 946585
08048000-0811c000 r-xp 00000000 fd:00 1209914    /var/admb/single
0811c000-0811e000 rw-p 000d3000 fd:00 1209914    /var/admb/single
0811e000-0811f000 rw-p 0811e000 00:00 0
09290000-092f9000 rw-p 09290000 00:00 0
b7b00000-b7b21000 rw-p b7b00000 00:00 0
b7b21000-b7c00000 ---p b7b21000 00:00 0
b7cc5000-b7f7b000 rw-p b7cc5000 00:00 0
bf8bf000-bf8d4000 rw-p bf8bf000 00:00 0          [stack]

Thanks for any input,
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.admb-project.org/pipermail/users/attachments/20110105/79a0dd8f/attachment.html>

More information about the Users mailing list