NetBSD Problem Report #18025

Received: (qmail 20319 invoked by uid 605); 22 Aug 2002 00:30:06 -0000
Message-Id: <200208220027.g7M0ROb00327@osprey.buzzard.freeserve.co.uk>
Date: Thu, 22 Aug 2002 01:27:24 +0100 (BST)
From: Richard Earnshaw <rearnsha@buzzard.freeserve.co.uk>
Sender: gnats-bugs-owner@netbsd.org
Reply-To: rearnsha@buzzard.freeserve.co.uk
To: gnats-bugs@gnats.netbsd.org
Subject: cats kernel data abort in non-usr mode
X-Send-Pr-Version: 3.95

>Number:         18025
>Category:       port-arm
>Synopsis:       cats kernel data abort in non-usr mode
>Confidential:   no
>Severity:       critical
>Priority:       high
>Responsible:    briggs
>State:          closed
>Class:          sw-bug
>Submitter-Id:   net
>Arrival-Date:   Thu Aug 22 00:31:01 +0000 2002
>Closed-Date:    Thu Aug 22 20:19:37 +0000 2002
>Last-Modified:  Thu Aug 22 20:19:37 +0000 2002
>Originator:     Richard Earnshaw
>Release:        NetBSD 1.6F (2002/08/21)
>Organization:
none

>Environment:


System: NetBSD osprey.buzzard.freeserve.co.uk 1.6F NetBSD 1.6F (OSPREY) #44: Wed Aug 21 20:59:30 BST 2002 rearnsha@osprey.buzzard.freeserve.co.uk:/usr/src/sys/arch/cats/compile/OSPREY cats
Architecture: arm
Machine: cats
>Description:
	Machine takes a data abort inside uiomove (or in a non-frame 
	generating subroutine of it) when trying to load gv
	postscript viewer.  Bug is repeatable on 2/2 attempts.

	My suspicions would be the new copyin/copyout code, that is called 
	directly from the above
>How-To-Repeat:
	Try to run gv on a current kernel.

>Fix:
	None known.

>Release-Note:
>Audit-Trail:

From: Allen Briggs <briggs@wasabisystems.com>
To: Richard Earnshaw <rearnsha@buzzard.freeserve.co.uk>
Cc: gnats-bugs@gnats.netbsd.org
Subject: Re: port-arm/18025: cats kernel data abort in non-usr mode
Date: Thu, 22 Aug 2002 01:04:52 -0400

 On Thu, Aug 22, 2002 at 01:27:24AM +0100, Richard Earnshaw wrote:
 > 	My suspicions would be the new copyin/copyout code, that is called 
 > 	directly from the above

 I suspect you're right.  It was doing Bad Things on unaligned copies.
 I made up some tests for a version of it in userland, and it failed
 while a corrected version passes.

 My apologies for not getting it run through this kind of test
 before it was committed.

 Could you please try revision 1.6 of arch/arm/arm/bcopyinout.S ?

 Thanks,
 -allen

 -- 
  Allen Briggs                     briggs@wasabisystems.com
  http://www.wasabisystems.com/    Quality NetBSD CDs, Sales, Support, Service
 NetBSD development for Alpha, ARM, M68K, MIPS, PowerPC, SuperH, XScale, etc...
State-Changed-From-To: open->feedback 
State-Changed-By: briggs 
State-Changed-When: Wed Aug 21 22:05:36 PDT 2002 
State-Changed-Why:  
Updated bcopyinout.S and requested that submitted test revision 1.6 
of same on his failure case. 


Responsible-Changed-From-To: port-arm-maintainer->briggs 
Responsible-Changed-By: briggs 
Responsible-Changed-When: Wed Aug 21 22:05:36 PDT 2002 
Responsible-Changed-Why:  
bcopyinout.S is mine. 

From: Richard Earnshaw <rearnsha@buzzard.freeserve.co.uk>
To: Allen Briggs <briggs@wasabisystems.com>
Cc: Richard Earnshaw <rearnsha@buzzard.freeserve.co.uk>,
  gnats-bugs@gnats.netbsd.org
Subject: Re: port-arm/18025: cats kernel data abort in non-usr mode 
Date: Thu, 22 Aug 2002 20:35:52 +0100

 > Could you please try revision 1.6 of arch/arm/arm/bcopyinout.S ?

 Seems to be fine now, thanks.

 R.

State-Changed-From-To: feedback->closed 
State-Changed-By: briggs 
State-Changed-When: Thu Aug 22 13:19:16 PDT 2002 
State-Changed-Why:  
Submitter confirms that it's fixed. 
>Unformatted:

NetBSD Home
NetBSD PR Database Search

(Contact us) $NetBSD: query-full-pr,v 1.39 2013/11/01 18:47:49 spz Exp $
$NetBSD: gnats_config.sh,v 1.8 2006/05/07 09:23:38 tsutsui Exp $
Copyright © 1994-2007 The NetBSD Foundation, Inc. ALL RIGHTS RESERVED.