NetBSD Problem Report #2136

From gnats  Mon Feb 26 18:31:59 1996
Received: from darkwatch.flame.org by pain.lcs.mit.edu (8.7.4/8.6.12) with ESMTP id SAA12760 for <gnats-bugs@gnats.netbsd.org>; Mon, 26 Feb 1996 18:24:12 -0500 (EST)
Message-Id: <199602262324.RAA00489@darkwatch.flame.org>
Date: Mon, 26 Feb 1996 17:24:14 -0600 (CST)
From: Michael Graff <explorer@flame.org>
Reply-To: explorer@flame.org
To: gnats-bugs@gnats.netbsd.org
Subject: kernels with -O2 break pcvt (at least)
X-Send-Pr-Version: 3.95

>Number:         2136
>Category:       port-i386
>Synopsis:       kernels with -O2 break pcvt (at least)
>Confidential:   no
>Severity:       critical
>Priority:       high
>Responsible:    gnats-admin
>State:          closed
>Class:          sw-bug
>Submitter-Id:   net
>Arrival-Date:   Mon Feb 26 18:35:02 +0000 1996
>Closed-Date:    Fri May 16 01:15:24 +0000 1997
>Last-Modified:  Fri May 16 01:16:24 +0000 1997
>Originator:     Michael Graff
>Release:        26-Feb-1996
>Organization:
flame.org:  yes, we do know everything
>Environment:

System: NetBSD darkwatch.flame.org 1.1A NetBSD 1.1A (DARKWATCH) #0: Mon Feb 26 15:55:39 CST 1996 explorer@packrat.flame.org:/usr/src/sys/arch/i386/compile/DARKWATCH i386


>Description:
It seems that compiling a kernel with -O2 still does not do the right
thing on the i386.  I notice this problem mostly on the i486 I have, but
even the pentathing I also run has screen problems from time to time.

My main concern is that if this is a compiler problem, what else may
be broken, in the kernel or in userland, because of it?

>How-To-Repeat:
Varies ; the 486 always has the problem, the pentathing, once in a while,
especially when leaving X or the machine is up for a long time.

>Fix:
Compiling a kernel with -O fixes it, or at least appears to hide it.

>Release-Note:
>Audit-Trail:

From: Chris_G_Demetriou@NIAGARA.NECTAR.CS.CMU.EDU
To: explorer@flame.org
Cc: gnats-bugs@NetBSD.ORG
Subject: Re: port-i386/2136: kernels with -O2 break pcvt (at least) 
Date: Mon, 26 Feb 96 20:59:41 EST

 I've seen this with on the i386 as well, but never got a chance to
 track it down.  (Not worth it, on my home machine...)  This made pcvt
 just about completely unusable for me on that machine...


From: "Michael Graff" <explorer@flame.org>
To: Chris_G_Demetriou@niagara.nectar.cs.cmu.edu
Cc: gnats-bugs@NetBSD.ORG
Subject: Re: port-i386/2136: kernels with -O2 break pcvt (at least) 
Date: Mon, 26 Feb 1996 20:15:48 -0600

 >I've seen this with on the i386 as well, but never got a chance to
 >track it down.  (Not worth it, on my home machine...)  This made pcvt
 >just about completely unusable for me on that machine...

 Yea...

 Do you have any suggestions on what to look for?  The pcvt code is
 utterly disgusting to read, and I have limited time to attack this
 sort of thing.

 --Michael

From: "Michael Graff" <explorer@flame.org>
Subject:  More info

I can work around this bug by compiling the pcvt_*.c files using
-O.  I will try to find out which file is causing the problem
later tonight.

--Michael

State-Changed-From-To: open->feedback 
State-Changed-By: lukem 
State-Changed-When: Wed May 14 20:29:55 PDT 1997 
State-Changed-Why:  

Michael, is this still a problem under -current (1.2E, circa may 1997)? 
I run pcvt on my 486dx4-100 and my p5-100, and my kernels are compiled 
with -O2 

State-Changed-From-To: feedback->closed 
State-Changed-By: lukem 
State-Changed-When: Thu May 15 18:15:24 PDT 1997 
State-Changed-Why:  
Not a problem anymore 
>Unformatted:

NetBSD Home
NetBSD PR Database Search

(Contact us) $NetBSD: query-full-pr,v 1.51 2026/08/10 02:28:17 riastradh Exp $
$NetBSD: gnats_config.sh,v 1.10 2026/05/13 22:00:09 riastradh Exp $
Copyright © 1994-2026 The NetBSD Foundation, Inc. ALL RIGHTS RESERVED.