From owner-freebsd-hackers Sun Jan 14 09:06:14 1996 Return-Path: owner-hackers Received: (from root@localhost) by freefall.freebsd.org (8.7.3/8.7.3) id JAA12358 for hackers-outgoing; Sun, 14 Jan 1996 09:06:14 -0800 (PST) Received: from peedub.gj.org (ns092.munich.netsurf.de [194.64.166.92]) by freefall.freebsd.org (8.7.3/8.7.3) with ESMTP id JAA12352 for ; Sun, 14 Jan 1996 09:06:09 -0800 (PST) Received: from localhost (localhost [127.0.0.1]) by peedub.gj.org (8.7.3/8.6.9) with SMTP id SAA12195; Sun, 14 Jan 1996 18:05:58 GMT Message-Id: <199601141805.SAA12195@peedub.gj.org> X-Authentication-Warning: peedub.gj.org: Host localhost [127.0.0.1] didn't use HELO protocol X-Mailer: exmh version 1.6 4/21/95 To: Bruce Evans cc: freebsd-hackers@freefall.freebsd.org, isdn@muc.ditec.de Subject: Re: Status of ISDN drivers Reply-To: Gary Jennejohn In-reply-to: Your message of "Mon, 15 Jan 1996 00:13:06 +1100." <199601141313.AAA05940@godzilla.zeta.org.au> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Date: Sun, 14 Jan 1996 18:05:58 +0000 From: Gary Jennejohn Sender: owner-hackers@FreeBSD.ORG Precedence: bulk Bruce Evans writes: >>I've become aware of a horrible bug somewhere in the ISDN code in the >>last few days which apparently trashes the kernel stack somehow. > >Someone actually uses the isdn drivers in FreeBSD? :-) Perhaps they >have been broken by recent cleanups. OTOH, perhaps they have been >fixed by recent cleanups. What version are you using? > I'm using a version is is not based on the 2.1R sources. May be that the changes/fixes in 2.1R could eliminate the problem. I'm planning to merge the modifications in the sources I'm using into -stable and see what happens. >>This bug is extrememly hard to track down beacuse the fault address is >>totally bogus, e.g. 8:0. That's why I suspect that the stack is getting >>trashed. It also crops up under circumstances which I can't identify. > >This fault address (but not a trashed stack) is normal for a call or >jump through a null pointer. Calls through null pointers are very >easy to debug since the return address is on the stack, while jumps >through null pointers are hard to debug. > actually, it's the instruction pointer that's 0x8:0x0, as I noted in a followup I posted. --- Gary Jennejohn Home - Gary.Jennejohn@munich.netsurf.de Work - gjennejohn@frt.dec.com