From: schiers@tpki.toppoint.de (Carsten Schiers) Subject: Re: [Q] XF86 1.3, ATI, restore text screen Date: Wed, 23 Jun 1993 20:51:02 GMT
haefele@atlas.gis.univie.ac.at () writes:
>After exiting from X, the restoration of the text screen
>fails. I can type in commands but see no echo (only stripes).
>The use of runx from vgalib fails also.
>Is there a workaround for this problem.
>(Xfree89 Version 1.2 worked fine)
Yes. I got the following posting and it worked...
============================ cut here ============================
Newsgroups: comp.windows.x.i386unix
Path: tpki.toppoint.de!fub!math.fu-berlin.de!xlink.net!howland.reston.ans.net!noc.near.net!uunet!munnari.oz.au!metro!news
From: dawes@physics.su.OZ.AU (David Dawes)
Subject: Re: Xfree86 1.3 ATI GUP problem
Message-ID: <1993Jun22.134609.29337@ucc.su.OZ.AU>
Sender: news@ucc.su.OZ.AU
Nntp-Posting-Host: physics.su.oz.au
Organization: School of Physics, University of Sydney, Australia
References: <2012ga$ptq@calvin.nyu.edu> <2026tu$rdv@Lightning.McRCIM.McGill.EDU>
Date: Tue, 22 Jun 1993 13:46:09 GMT
Lines: 89
In article <2026tu$rdv@Lightning.McRCIM.McGill.EDU> panisset@McRCIM.McGill.EDU (Jean-Francois Panisset) writes:
>In article <2012ga$ptq@calvin.nyu.edu> reznick@acf2.nyu.edu ( ) writes:
>>
>>I have XFree86 up and running under Linux, works great, thank you!
>>I have an ATI Graphics Ultra PRO, which uses the Mach 32 chip set.
>>The server works fine, and it even detecs the mach-32 which it never
>>did before.
>>
>>The only problem is that it seems to be unable to restore my SVGA
>>text mode. It worked fine under 1.2.
>>
>
>I am having the same problem with my ATI VGA Wonder XL (chip version 28800-5).
>My configuration:
>Linux 0.99PL10
>libc 4.4
>XFree86-1.3 from tsx-11.mit.edu, dated June 18 (thus not the corrupted files)
>
>The X server starts up fine, I can even switch back to a text-mode VT and
>switch back to X, but when the server exits, it messes up the text-mode
>display: all I get are white vertical lines a few pixel wide, one per
>character position (I can still see the cursor). My default text mode
>is 80x25 (i.e. NORMAL_VGA in the kernel configuration file), so that
>should not be a problem.
>
>For now, I've just switched back to the 1.2 server, but it would be nice if
>this could be fixed in the 1.3 server.
Attached is a patch for the ATI driver which should fix this problem
(the original poster has tested it).
If you have the full XFree86 source, the patch should be applied in
mit/server/ddx/x386/vga256/drivers/ati/.
The easiest way to make use of this patch is to get the LinkKit, apply
this patch in /usr/X386/lib/Server/drivers/vga256/ati/, then rebuild
the server (there are instructions for using the LinkKit in
/usr/X386/lib/Server/).
[This and other late fixes can be found in the README.late file on
ftp.physics.su.oz.au:/XFree86 and export.lcs.mit.edu:/contrib/XFree86]
=============================================================================
*** ati/driver.c-1.3 Tue Jun 22 21:32:03 1993
--- ati/driver.c Tue Jun 22 21:35:12 1993
***************
*** 191,196 ****
--- 191,198 ----
#define TRACE(a) /* Disable TRACE statements */
#endif
+ #define ATI_PATCHLEVEL 1
+
typedef struct {
vgaHWRec std; /* good old IBM VGA */
unsigned char ATIExtRegBank[11]; /* ATI Registers B0,B1,B2,B3,B5,B6,B8,B9,BE,A6,A7 */
***************
*** 316,321 ****
--- 318,325 ----
/* Load Miscellaneous Output External Register */
outb(0x3C2, restore->std.MiscOutReg);
+ outw(ATIExtReg, 0x00b2); /* segment select 0 */
+
/* For text modes, download Character Generator into Plane 2 */
if (restore->std.FontInfo1 || restore->std.FontInfo2 ||
restore->std.TextInfo) {
***************
*** 1112,1117 ****
--- 1116,1124 ----
close( fd );
#endif /* MACH386 */
+ ErrorF("%s: ATI: driver patchlevel: %d\n", vga256InfoRec.name,
+ ATI_PATCHLEVEL);
+
ErrorF( "ATI BIOS Information Block:\n" );
ErrorF( " Signature code: %c%c = ",
bios_data[ 0x40 ], bios_data[ 0x41 ] );
=============================================================================
David