>>>>> On Mon, 16 Jul 2001 20:55:11 -0700,
>>>>> [EMAIL PROTECTED] (Bruce A. Mah) said:
>> > (kgdb) print rt
>> > $1 = (struct rtentry *) 0x3
>> > (kgdb) print ln
>> > $2 = (struct llinfo_nd6 *) 0x620000
>>
>> Then rt and ln are surely broken. I have not got the reason yet,
>> though...
> Yes, rt in particular caught my eye.
I've almost got the problem. Unfortunately, all IPv6 autoconfigured
nodes would panic every 24 hours. Since I guess I need some more work
for a complete fix to this, I'll attach a workaround.
Please try the following patch for the moment. The patch suppresses
garbage collection of stale neighbor cache entries, but it would not
matter for most users.
Thanks,
JINMEI, Tatuya
Communication Platform Lab.
Corporate R&D Center, Toshiba Corp.
[EMAIL PROTECTED]
*** nd6.c.orig Wed Jul 18 00:06:30 2001
--- nd6.c Wed Jul 18 00:06:57 2001
***************
*** 500,507 ****
--- 500,509 ----
case ND6_LLINFO_STALE:
/* Garbage Collection(RFC 2461 5.3) */
+ #if 0 /* XXX: short-term workaround to avoid panic */
if (ln->ln_expire)
next = nd6_free(rt);
+ #endif
break;
case ND6_LLINFO_DELAY:
To Unsubscribe: send mail to [EMAIL PROTECTED]
with "unsubscribe freebsd-net" in the body of the message
