does the attached oops makes sense or it is just messed up?
AFAICT the ksymoops is using right System.map, yet the stack
trace does not seem to follow logical order. it is from 2.4.6-pre1
for that matter is "defensive" programming the rule in kernel?
this ops could be avoided if the net/ipv4/ipmr.c:ipmr_new_tunel()
function was changed to check whether 'v' is null and if it is
true then just return.
I did not submit this patch though since I couldn't figure how
in the the first place code ended up there.
-- Adam http://www.eax.com The Supreme Headquarters of the 32 bit registers
This archive was generated by hypermail 2b29 : Fri Jun 15 2001 - 21:00:10 EST