Date: Sun, 21 Dec 1997 11:16:36 +0100 (MET) From: j@uriah.heep.sax.de (J Wunsch) To: freebsd-current@freebsd.org Subject: Re: panics when stopping pppd Message-ID: <199712211016.LAA00416@uriah.heep.sax.de> References: <19971215020638.WG24374@@> <199712151524.KAA08128@khavrinen.lcs.mit.edu>
next in thread | previous in thread | raw e-mail | index | archive | help
Garrett Wollman <wollman@khavrinen.lcs.mit.edu> wrote: > This seems to confirm my theory. Here's what I think is happening: > > 1) An interface is downed. All of its non-static routes get deleted > automatically. > > 2) One of those interface routes was a cloning route. rtrequest() > notices this, and deletes all of its children. > > 3) Oops, a pointer to one of those children was held by rn_walktree() > as the next node to examine... blam! Any news on this front? I'm also suffering from that problem, the kernel always panics at the very same (totally invalid) %eip address. In my situation, it happens whenever i shut down the SLIP interface to one of my possible Internet connections. The shutdown wakes up GateD which in turn deletes the default route out the SLIP interface, and reinstalls the (lower prioritized) default route out my ISDN (sppp) interface. slattach isn't even given the time to hangup the modem, the kernel coredumps before. Default routes out a p2p interface are cloning? In this case, it's pretty sure there are cloned routes out the SLIP interface around. After all, i've been opening the SLIP line in order to reach some host, so at least one cloned route is likely to lurk around. Any workaround that at least prevents the worst? Why didn't this happen in previous kernels? -- cheers, J"org joerg_wunsch@uriah.heep.sax.de -- http://www.sax.de/~joerg/ -- NIC: JW11-RIPE Never trust an operating system you don't have sources for. ;-)
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?199712211016.LAA00416>