NetBSD Problem Report #17258

Received: (qmail 16080 invoked by uid 605); 14 Jun 2002 22:24:44 -0000
Message-Id: <Pine.SOL.4.33.0206142315570.25070-100000@draco.cus.cam.ac.uk>
Date: Fri, 14 Jun 2002 23:24:42 +0100 (BST)
From: Ben Harris <bjh21@netbsd.org>
Sender: gnats-bugs-owner@netbsd.org
To: <gnats-bugs@gnats.netbsd.org>
Subject: vidcvideo initialisation corrupts display

>Number:         17258
>Category:       port-acorn32
>Synopsis:       vidcvideo initialisation corrupts display
>Confidential:   no
>Severity:       non-critical
>Priority:       low
>Responsible:    port-acorn32-maintainer
>State:          open
>Class:          sw-bug
>Submitter-Id:   net
>Arrival-Date:   Fri Jun 14 22:25:02 +0000 2002
>Closed-Date:    
>Last-Modified:  Mon Feb 23 15:57:00 +0000 2004
>Originator:     Ben Harris
>Release:        2002-06-06
>Organization:
>Environment:
NetBSD 1.6A (RPC_WSCONS) #0: Thu Jun  6 23:42:21 UTC 2002
    bjh21@southeast-iceland:/usr/src/sys/arch/acorn32/compile/RPC_WSCONS

>Description:
In /var/log/dmesg.out, I have (excerpted):

cd0 at atapibus0 drive 1: <PHILIPS CDRW2412A, 5VO1214DS22435, 0.01000> type 5 cdrom removable
cd0: drive supports PIO mode 4, DMA mode 2, Ultra-DMA mode 2 (Ultra/33)
wd0 at wdc0 channel 0 drive 0: <FUJITSU MPA3017ATU>
wd0: drive supports 32-sector PIO transfers, LBA addressing
wd0: 1668 MB, 3390 cyl, 16 head, 63 sec, 512 bytes/sect x 3417976 sectors
wd0: drive supports PIO mode 4, DMA mode 2, Ultra-DMA mode 2 (Ultra/33)
fdc0 at pioc0 offset 0x3f0-0x3f7 irq 12 drq 0x00002000
fd0 at fdc0 drive 0: 1.44MB 80 cyl, 2 head, 18 sec
com0 at pioc0 offset 0x3f8-0x3ff irq 10: ns16550a, working fifo
lpt0 at pioc0 offset 0x278-0x27b irq 0
iomd0 at mainbus0: IOMD20 version 2
iomd0: DRAM refresh=16us
clock0 at iomd0
rpckbd0 at iomd0
wskbd0 at rpckbd0: console keyboard
iic0 at iomd0
rtc0 at iic0 addr 0xa0: PCF8583 clock base 32.768KHz
todclock0 at rtc0
wsqms0 at iomd0
wsmouse0 at wsqms0 mux 0
vidc0 at mainbus0: VIDC20
vidcvideo0 at vidc0: refclk=24MHz 2048KB VRAM : using 640 x 480, 16bpp
wsdisplay0 at vidcvideo0 kbdmux 1: console (std, vt100 emulation), using wskbd0

On the console, however, the text between the "cd0" and "wsdisplay0" lines
was missing.  It looked to me as if the screen jumped when vidcvideo0
initialised, so my guess is that it reset the hardware scrolling or
somesuch.

>How-To-Repeat:
Boot a Risc PC in a 640x480 screen mode and watch the console carefully.
Having unclean filesystems or booting single-user probably helps.

>Fix:



>Release-Note:
>Audit-Trail:

From: Richard Earnshaw <rearnsha@arm.com>
To: gnats-bugs@gnats.netbsd.org, Ben Harris <bjh21@netbsd.org>
Cc: Richard.Earnshaw@arm.com
Subject: Re: port-acorn32/17258 vidcvideo initialisation corrupts display
Date: Wed, 09 Apr 2003 17:32:01 +0100

 > On the console, however, the text between the "cd0" and "wsdisplay0"
 > lines was missing.  It looked to me as if the screen jumped when
 > vidcvideo0 initialised, so my guess is that it reset the hardware
 > scrolling or somesuch.

 This is on a StrongARM, right? I think it's a failure to clean the cache 
 after a screen update.  I've seen it as well on my RPC with the new 
 console code.  It's particularly bad if you drop into ddb; but if you 
 execute a few commands, eventually earlier text gets flushed out of the 
 cache into video RAM.

 R.



From: Ben Harris <bjh21@netbsd.org>
To: Richard.Earnshaw@arm.com
Cc: gnats-bugs@gnats.netbsd.org
Subject: Re: port-acorn32/17258 vidcvideo initialisation corrupts display
Date: Thu, 10 Apr 2003 10:10:34 +0100 (BST)

 On Wed, 9 Apr 2003, Richard Earnshaw wrote:

 > > On the console, however, the text between the "cd0" and "wsdisplay0"
 > > lines was missing.  It looked to me as if the screen jumped when
 > > vidcvideo0 initialised, so my guess is that it reset the hardware
 > > scrolling or somesuch.
 >
 > This is on a StrongARM, right? I think it's a failure to clean the cache
 > after a screen update.  I've seen it as well on my RPC with the new
 > console code.  It's particularly bad if you drop into ddb; but if you
 > execute a few commands, eventually earlier text gets flushed out of the
 > cache into video RAM.

 No, it's not that problem.  Well, it's not the problem I think you're
 talking about, which is where characters appear in the right place, but
 aren't complete until the cache gets cleaned.

 The problem I'm talking about manifests as lines which simply aren't
 present, so the screen display just went:

 cd0 at atapibus0 drive 1: <PHILIPS CDRW2412A, 5VO1214DS22435, 0.01000> type 5 cd
 wsdisplay0 at vidcvideo0 kbdmux 1: console (std, vt100 emulation), using wskbd0

 with no gap in between.  I'm fairly certain that what happened was that
 vidcvideo reset the hardware scroll position when it initialised (after
 the screen had already scrolled), so the "wsdisplay" line appeared on top
 of lines that had already been printed.

 When I get a chance, I'll check this on a non-StrongARM machine, but that
 may not be for a while.

 -- 
 Ben Harris                                                   <bjh21@netbsd.org>
 Portmaster, NetBSD/acorn26           <URL:http://www.netbsd.org/Ports/acorn26/>

State-Changed-From-To: open->feedback 
State-Changed-By: reinoud 
State-Changed-When: Sun Jul 20 07:12:24 UTC 2003 
State-Changed-Why:  
Could you please check if the version currently checked in cvs.netbsd.org 
still displays this problem? 

From: Ben Harris <bjh21@NetBSD.org>
To: gnats-bugs@NetBSD.org
Cc:  
Subject: port-acorn32/17258: still present
Date: Tue, 29 Jul 2003 23:41:41 +0100 (BST)

 I've just had cause to test this (seeing if the new I2C framework works on
 acorn32), and this bug is still present as of 1.6U.

 -- 
 Ben Harris                                                   <bjh21@NetBSD.org>
 Portmaster, NetBSD/acorn26           <URL:http://www.NetBSD.org/Ports/acorn26/>

State-Changed-From-To: feedback->open 
State-Changed-By: bjh21 
State-Changed-When: Mon Oct 20 10:56:08 UTC 2003 
State-Changed-Why:  
Still present in recent snapshots. 

From: Reinoud Zandijk <reinoud@netbsd.org>
To: gnats-bugs@gnats.netbsd.org
Cc:  
Subject: Re: port-acorn32/17258
Date: Mon, 23 Feb 2004 16:56:15 +0100

 Sorry for responding so late but i've tried and tried with my RiscPC with
 different colourdepths but i can't seem to be able to reproduce it. It 
 scrolls very fast after the detection of the wd0 since i dont have a cd-rom 
 but i dont see any details being skipped....

 Could you please reverify it with bootloader 3.20 or newer?

 Cheers,
 Reinoud

>Unformatted:

Home	PR Database Search