From owner-freebsd-questions  Sat Dec  5 14:08:39 1998
Return-Path: <owner-freebsd-questions@FreeBSD.ORG>
Received: (from majordom@localhost)
          by hub.freebsd.org (8.8.8/8.8.8) id OAA16912
          for freebsd-questions-outgoing; Sat, 5 Dec 1998 14:08:39 -0800 (PST)
          (envelope-from owner-freebsd-questions@FreeBSD.ORG)
Received: from marvin.ece.utexas.edu (marvin.ece.utexas.edu [128.83.52.151])
          by hub.freebsd.org (8.8.8/8.8.8) with ESMTP id OAA16907;
          Sat, 5 Dec 1998 14:08:34 -0800 (PST)
          (envelope-from bgrayson@marvin.ece.utexas.edu)
Received: (from bgrayson@localhost)
	by marvin.ece.utexas.edu (8.8.8/8.8.8) id QAA22054;
	Sat, 5 Dec 1998 16:08:24 -0600 (CST)
Message-ID: <19981205160824.A20455@marvin.ece.utexas.edu>
Date: Sat, 5 Dec 1998 16:08:24 -0600
From: "Brian C. Grayson" <bgrayson@marvin.ece.utexas.edu>
To: Doug White <dwhite@resnet.uoregon.edu>
Cc: jmutter@netwalk.com, phk@FreeBSD.ORG, thallgren@yahoo.com,
        gkshenaut@ucdavis.edu, questions@FreeBSD.ORG
Subject: SOLVED: calcru and XF86_3DLabs
References: <19981203125900.A27570@marvin.ece.utexas.edu> <Pine.BSF.4.03.9812031158490.12937-100000@resnet.uoregon.edu>
Mime-Version: 1.0
Content-Type: text/plain; charset=us-ascii
X-Mailer: Mutt 0.93.1i
In-Reply-To: <Pine.BSF.4.03.9812031158490.12937-100000@resnet.uoregon.edu>; from Doug White on Thu, Dec 03, 1998 at 12:03:07PM -0800
Sender: owner-freebsd-questions@FreeBSD.ORG
Precedence: bulk
X-Loop: FreeBSD.ORG

  I think I found the problem with calcru and Fire GL 1000 Pro cards
using the XFree86 XF86_3DLabs server -- the XFree86 FAQ says:

   Warning! This server has a bug that will most likely make your clock
   run significantly slower. This bug has been fixed, new binaries will
   be made available. If you are building from source, use this patch:
...
-       ( outb(80,0),*(unsigned int *)((char*)GLINTMMIOBase+r))
+       ( outb(0x80,0),*(unsigned int *)((char*)GLINTMMIOBase+r))

  (See Q.F15 at http://www.xfree86.org/FAQ/#GLINT for the whole patch.)

  From a disassembly of the XF86_3DLabs server still up on
ftp.xfree86.org, the server suffers from this bug.  It turns
out that 0x50 (80 decimal) is in the range associated with the
timer, if one of my ISA documents is correct, which explains a
lot of the symptoms.  

  To check if your XF86_3DLabs server is vulnerable, do
"objdump --disassemble <path>/XF86_3DLabs | grep -1 outb".  You
should see lots of lines like these if you are vulnerable:
 8076e35:       ba 50 00 00 00  movl   $0x50,%edx
 8076e3a:       ee              outb   %al,(%dx)
...

  I recompiled the server, with that modification, and now it looks
like:
 8076e35:       ba 80 00 00 00  movl   $0x80,%edx
 8076e3a:       ee              outb   %al,(%dx)

  I will not have physical access to the machine in question for a few
days, but starting up X remotely does not appear to cause time to slow
down.  I've put the server for 3.0 up at
http://lore.ece.utexas.edu/~bgrayson/XF86_3DLabs.gz, but with
absolutely no warranty, etc. etc.  I assume XFree86 will have new
binaries up some time within the next few days.

  Brian Grayson

To Unsubscribe: send mail to majordomo@FreeBSD.org
with "unsubscribe freebsd-questions" in the body of the message