On Fri, Jul 09, 2004 at 12:19:46PM -0700, Luck, Tony wrote: > >So what really happens is: calls from init to 'core' functions > >and calls from 'exit' to core functions. > > Hmmm. Neither of those should have tripped the check in the > kernel module loader. It should only go bang when you have a > call from "core" to "init". Does "exit" end up in a separate > section? Maybe I was mistakenly triggering on the exit->core? Apparently yes: 0 .text 00003f78 00000000 00000000 00000034 2**2 CONTENTS, ALLOC, LOAD, RELOC, READONLY, CODE 1 .init.text 0000029c 00000000 00000000 00003fac 2**2 CONTENTS, ALLOC, LOAD, RELOC, READONLY, CODE 2 .exit.text 00000008 00000000 00000000 00004248 2**2 CONTENTS, ALLOC, LOAD, RELOC, READONLY, CODE But also, we don't call __init from __exit (which would be crash once __init has been purged from memory). > Which module actually failed to load? yes, I would be interested in that, too :) > -Tony -- - Harald Welte http://www.netfilter.org/ ============================================================================ "Fragmentation is like classful addressing -- an interesting early architectural error that shows how much experimentation was going on while IP was being designed." -- Paul Vixie