From hsu@clinet.fi  Wed Jun  5 07:33:14 1996
Received: from hauki.clinet.fi (hauki.clinet.fi [194.100.0.1])
          by freefall.freebsd.org (8.7.5/8.7.3) with ESMTP id HAA04423
          for <FreeBSD-gnats-submit@freebsd.org>; Wed, 5 Jun 1996 07:33:09 -0700 (PDT)
Received: from katiska.clinet.fi (root@katiska.clinet.fi [194.100.0.4]) by hauki.clinet.fi (8.7.5/8.6.4) with ESMTP id RAA00921 for <FreeBSD-gnats-submit@freebsd.org>; Wed, 5 Jun 1996 17:32:36 +0300 (EET DST)
Received: (hsu@localhost) by katiska.clinet.fi (8.7.5/8.6.4) id RAA20913; Wed, 5 Jun 1996 17:32:35 +0300 (EET DST)
Message-Id: <199606051432.RAA20913@katiska.clinet.fi>
Date: Wed, 5 Jun 1996 17:32:35 +0300 (EET DST)
From: Heikki Suonsivu <hsu@clinet.fi>
Reply-To: hsu@clinet.fi
To: FreeBSD-gnats-submit@freebsd.org
Subject: Fatal trap 12: page fault while in kernel mode (PPP/modems/cyclades?)
X-Send-Pr-Version: 3.2

>Number:         1293
>Category:       kern
>Synopsis:       Fatal trap 12: page fault while in kernel mode (PPP/modems/cyclades?)
>Confidential:   no
>Severity:       serious
>Priority:       high
>Responsible:    freebsd-bugs
>State:          closed
>Quarter:        
>Keywords:       
>Date-Required:  
>Class:          sw-bug
>Submitter-Id:   current-users
>Arrival-Date:   Wed Jun  5 07:40:02 PDT 1996
>Closed-Date:    Fri Mar 19 04:37:30 PST 1999
>Last-Modified:  Fri Mar 19 04:38:21 PST 1999
>Originator:     Heikki Suonsivu
>Release:        FreeBSD 2.2-CURRENT i386
>Organization:
Clinet, Espoo, Finland
>Environment:

	We have 4 modem servers built on 486 and pentium machines.  They have
	several different motherboard configurations, common things are
	that they have ide drives and cyclades serial boards (32 or 16 ports
	each).  They have been running several versions of -stable and
	-current.  The problem has been there for long time.
	Most modems are mostly V34, some V32.

	We have identical configurations used as leased line asynchronous
	routers, and have no problems at all (they do never seem to crash). 

	Dialups are using both kernel ppp and rlogin connections to
	other machines (these are dedicated, users cannot login to the
	machines themselves).  The only programs running in these machines
	are pppd, rlogin, getty (patched agetty, actually) and gated.

>Description:

	The machine locks up with console containing either

	Fatal trap 12: page fault while in kernel mode

	Fatal trap 12: page fault while in kernel mode

	Fatal trap 12: page fault while in kernel mode

	...

	or some fatal trap message and a double panic message.

	It seems that about 1 crash out of 4 leads to a deadlock like
	one of these.  Other panics are reboots without trails, the systems
	do not seem to generate dumps.

	If it is a hardware problem, it would have to be a general design
	problem in one of the components (cyclades is the only
	"uncommon" piece of hardware).  But I doubt it, as the same
	configuration works fine with leased lines, and the machines
	are from varying ages and configurations.

	It seems that this has something to do with dialup modems.

>How-To-Repeat:

	I do not know.  I do not seem to ever get crash dumps.

>Fix:
	
>Release-Note:
>Audit-Trail:
State-Changed-From-To: open->feedback 
State-Changed-By: scrappy 
State-Changed-When: Mon Oct 21 22:52:41 PDT 1996 
State-Changed-Why:  

confirm status 

State-Changed-From-To: feedback->open 
State-Changed-By: scrappy 
State-Changed-When: Tue Oct 22 22:51:21 PDT 1996 
State-Changed-Why:  

ACK: 

From: Heikki Suonsivu <hsu@clinet.fi> 

This is still a problem (and is getting even more serious, up to extent we 
are just installing linux on one of the machines in question to try if it 
could solve the problem) 

State-Changed-From-To: open->feedback 
State-Changed-By: mpp 
State-Changed-When: Sun Feb 16 10:35:55 PST 1997 
State-Changed-Why:  
Checking with originator to see if this is still an issue. 

From: Brian Somers <brian@utell.co.uk>
To: freebsd-gnats-submit@freebsd.org, hsu@clinet.fi
Cc: Brian Somers <brian@awfulhak.org>
Subject: Re: kern/1293: Fatal trap 12: page fault while in kernel mode
Date: Mon, 23 Jun 1997 17:44:19 +0100

 Is this still happening ?
 
 It may have been fixed by an update to slcompress.c a few days ago.
 The update prevents mbufs with an original length of 40 bytes from
 getting a negative length after vj compression.
 -- 
 Brian <brian@awfulhak.org> <brian@freebsd.org>
       <http://www.awfulhak.org>
 Don't _EVER_ lose your sense of humour !
 
 
State-Changed-From-To: feedback->closed 
State-Changed-By: brian 
State-Changed-When: Fri Jul 4 01:53:19 PDT 1997 
State-Changed-Why:  
This is a mistake 
State-Changed-From-To: closed->feedback 
State-Changed-By: brian 
State-Changed-When: Fri Jul 4 01:59:11 PDT 1997 
State-Changed-Why:  
Undo mistake 

From: Brian Somers <brian@awfulhak.org>
To: freebsd-gnats-submit@freebsd.org, hsu@clinet.fi
Cc: Brian Somers <brian@awfulhak.org>
Subject: Re: kern/1293: Fatal trap 12: page fault while in kernel mode
Date: Fri, 04 Jul 1997 10:02:29 +0100

 The first transmission of this mail failed.... it seems clinet.fi
 has a somewhat smart MDA that refused the mail on accound of my
 (incorrect) From: line.
 
 This is now fixed, here's the original message (the few days is
 now "a few days" + 8):
 
 Is this still happening ?
 
 It may have been fixed by an update to slcompress.c a few days ago.
 The update prevents mbufs with an original length of 40 bytes from
 getting a negative length after vj compression.
 -- 
 Brian <brian@awfulhak.org> <brian@freebsd.org>
       <http://www.awfulhak.org>
 Don't _EVER_ lose your sense of humour !
 
 

From: Brian Somers <brian@utell.co.uk>
To: freebsd-gnats-submit@freebsd.org, hsu@clinet.fi
Cc: brian@awfulhak.org
Subject: Re: kern/1293: Fatal trap 12: page fault while in kernel mode
Date: Fri, 4 Jul 1997 10:08:13 +0100 (BST)

 The MDA isn't just looking at the From: line.  What
 a pain in the ass !  You can't send mail to clinet.fi
 from a private network.
 
 > The first transmission of this mail failed.... it seems clinet.fi
 > has a somewhat smart MDA that refused the mail on accound of my
 > (incorrect) From: line.
 > 
 > This is now fixed, here's the original message (the few days is
 > now "a few days" + 8):
 > 
 > Is this still happening ?
 > 
 > It may have been fixed by an update to slcompress.c a few days ago.
 > The update prevents mbufs with an original length of 40 bytes from
 > getting a negative length after vj compression.
 
 -- 
 Brian <brian@awfulhak.org> <brian@freebsd.org>
       <http://www.awfulhak.org>
 Don't _EVER_ lose your sense of humour !
 
 
State-Changed-From-To: feedback->closed 
State-Changed-By: brian 
State-Changed-When: Mon Sep 8 14:00:11 PDT 1997 
State-Changed-Why:  
I suspect this was the vjcompression thing.  I sent mail to 
the bug submitter, but never got a reply. 


Responsible-Changed-From-To: freebsd-bugs->brian 
Responsible-Changed-By: brian 
Responsible-Changed-When: Mon Sep 8 14:00:11 PDT 1997 
Responsible-Changed-Why:  
Pppd's kinda mine too. 
State-Changed-From-To: closed->open 
State-Changed-By: brian 
State-Changed-When: Tue Sep 9 15:24:05 PDT 1997 
State-Changed-Why:  
I've got the feedback & I was wrong. 
Responsible-Changed-From-To: brian->freebsd-bugs 
Responsible-Changed-By: brian 
Responsible-Changed-When: Fri Jan 23 07:04:15 PST 1998 
Responsible-Changed-Why:  
Pppd isn't mine 

From: "Vladimir B. Grebenschikoc" <vova@express.ru>
To: freebsd-gnats-submit@freebsd.org, hsu@clinet.fi
Cc:  Subject: Re: kern/1293: Fatal trap 12: page fault while in kernel mode (PPP/modems/cyclades?)
Date: Mon, 13 Apr 1998 22:53:06 +0000

 I have same problem with sio & ppp/pppd:
 
 page fault with panic when pppd or ppp hang up after session
 
 only when firewall is on ! (regardless firewall rules)
 I've compiled kernel with IPDIVERT, but no divert rules was 
 
 on 486DX 20Mb RAM I has this effect in ~20%
 on P75 64Mb ~50%, on 6x86 200+ 64Mb - 100%
 
 FreeBSD 3.0-971022-SNAP
 
 Hardware:
 Cirix 6x86 200+, RAM 64Mb, de0, ide_pci0, aic0, sio[01]: 16550A,
 external USR Courier, pnp sound blaster
 
 --
 
 Vladimir B. Grebenschikov
 
 vova@folco.lms.ru
 2:5020/302@fidonet.org

From: Michael Graziano <mikeg@hoflink.com>
To: freebsd-gnats-submit@freebsd.org, hsu@clinet.fi
Cc:  Subject: Re: kern/1293: Fatal trap 12: page fault while in kernel mode 
 (PPP/modems/cyclades?)
Date: Fri, 14 Aug 1998 01:07:57 -0400

 This is also showing up on one of my machines (a 486 running 2.2.7
 (Freshly installed))
 
 The problem shows up every time the machine is shutdown, halt-ed or
 reboot-ed, as well as at random (but frequent) intervals while the
 system is running, and is always in the form of a "Fatal Trap 12: Page
 fault while in kernel mode" message (followed by details of where the
 thing died, an attempt to sync disks (which always fails), and an
 auto-reboot.)
 This ONLY occours for me using my custom-built kernel.  kernel.GENERIC
 boots, runs, and shuts down fine on my system, and I am using this as a
 temporary fix.
 
 My box is not networked (at all) yet.
 
State-Changed-From-To: open->closed 
State-Changed-By: sheldonh 
State-Changed-When: Fri Mar 19 04:37:30 PST 1999 
State-Changed-Why:  
push 
Active development on the 2.2 branch is over. (Pls ignore "push" above) 
>Unformatted:
