From: Padraig Brady (padraig at domain antefacto.com)
Date: Thu 20 Sep 2001 - 19:07:31 IST
Default RH7.1 install it works fine for me (gcc version 2.96).
Note I've glibc 2.2.2 also which gives the same nm output as
you have.
Padraig.
Dave Neary wrote:
>Hi all,
>
>Has anyone else had issues with a RedHat glibc (2.2.2-10)? Mine doesn't
>seem to define atexit(3), or doesn't define it properly... I'm just
>wondering whether I've messed something up, or this is something I
>should tell redhat about - I haven't found anything in their errata.
>
>The following program shows it up, if the problem's there - this came
>about from upgrading my glibc from an older (2.1) glibc, so there may be
>confusion as to what's defining what.
>--- test_atexit.c ---
>#include <stdlib.h>
>#include <stdio.h>
>
>void my_exit(void)
>{
> printf("Exitting now.\n");
>}
>int main(void)
>{
> atexit(my_exit);
> return 0;
>}
>--- EOF ---
>
>nm /lib/i686/libc.so.6 | grep atexit
>gives
>00030420 T __cxa_atexit
>00030514 t __dyn_atexit
>00129b8c ? __elf_set___libc_atexit_element__cleanup__
>00129b8c a __start___libc_atexit
>00129b90 a __stop___libc_atexit
>0012c160 b added_atexit_handler.0
>00030514 T atexit at domain GLIBC_2.0
> ^^^^^^^^^^^^^^^^^
>I think the last line's the important one. Somehow it hasn't defined
>atexit, but only atexit at domain GLIBC_2.0 - I don't know what this all means -
>any of you have any idea? If there's something wrong here, there could
>be something wrong elsewhere that I haven't spotted, so I'd kind of like
>to figure it out :)
>
>Thanks for the help,
>Dave.
>
This archive was generated by hypermail 2.1.6 : Thu 06 Feb 2003 - 13:12:14 GMT