Re: Bug on su using -c option

[James Youngman]

On 9/7/07, Mathieu Sivade <address@hidden> wrote:
>         *** /usr/local/ejabberd/bin/IperExternModule/dispatch_daemon:
>         realloc(): invalid next size: 0x09d79008 ***

I do not know why your "dispatch_daemon" program is experiencing a
memory corruption error, but I think it's very unlikely to have been
caused by any coreutils program.   That is unless "dispatch_daemon" is
actually a coreutils program you have modified or renamed or
something.



> Even though the tcsetpgrp errors which disappear when the Suid bit is
> set on the daemon, realloc keeps raising an error just like here.

... so you may have more than one different problem.

> By the way, gdb told me 0x09d79008 is the value of the data pointer, not
> the new size as the error message lead me to think at first.
>
> And finally, it appears that using su to log as ejabberd, then starting
> the daemon doesn't raise the error, as opposed to su -c.

To be honest, I don't think this is likely to be a coreutils problem.
 Try asking someone who knows something about the code you are working
with.  Sorry we couldn't help.

James.