linux-ipsec: Re: Pluto/main.c/creat_lock BUG + FIX

| Prob No1: Pluto deamon dies with a segfault in create_lock

Thanks very much for the bug report. I recommended that we pick up a cheap Alpha system for testing like this, but we've been busy enough without this. (I saw Multias for a really low price -- are they reasonable?)

Interestingly, this is just as much a bug on the x86 as on the Alpha. If the uninitialized pointer had been null on the x86, create_lock would have crashed. Somehow, it always gets some valid pointer (left over stack junk, I guess).

Fixed now (a little differently).

Hugh Redelmeier
hugh@mimosa.com voice: +1 416 482-8253 Received on Thu Dec 3 16:18:18 1998