Since last July there have been questions about the message
inetd in realloc(): warning: junk pointer, too low to make sense.
appearing to users who try to telnet, ftp, or rlogin to a FreeBSD machine.
If swap ever fills up, inetd does naughty things with stuff called
"pointers" and "realloc()" and after that it just can't ever get back on the
We need to get a core dump or attach gdb to a running inetd process to
find out where this realloc() is occuring, which (we hope) would lead us
to discover & fix the actual corruption problem.
The command "ln -s A /etc/malloc.conf" will create a link that sets
configuration options for malloc (see malloc(3)'s man page) that will
cause the malloc library routines to dump the process' core upon warnings
or errors. If someone can get this core image, then they can use gdb to
get a stack trace and find out what routine had called realloc. This
stack trace would at least be a good clue, if not a smoking gun.