Re: [Xpert]To xfs, or not to xfs? Is the question...(Was: Bug#147898: xserver-xfree86: freeze when using TrueType with Xinerama)

2002-05-25 Thread Derrik Pates

On Sat, May 25, 2002 at 09:14:02AM +0100, Jos? Fonseca wrote:
 What are the differences for a desktop user between using xfs or X's own 
 Xft? Which one is better?

Xft and xfs don't fill the same role at all. Xft is the client-side font
rendering component (used with the RENDER extension) for anti-aliased
fonts. xfs is a separate-process font server (as a complement to, or
replacement for, locally-specified fontpaths in XF86Config[-4]).
Supposedly in some configurations it can lower font-rendering latency
(of course, it doesn't affect apps that use the RENDER extension for
font drawing - the fonts there are drawn in the client process space).
It also introduces the liability that your X server can hang if it loses
its communication channel with the font server.

-- 
Derrik Pates  |   Sysadmin, Douglas School   |#linuxOS on EFnet
[EMAIL PROTECTED] | District (dsdk12.net)|#linuxOS on OPN
___
Xpert mailing list
[EMAIL PROTECTED]
http://XFree86.Org/mailman/listinfo/xpert



Re: [Xpert]RE: RADEON VE SIGNED DRIVERS UPDATES QUESTION

2002-05-18 Thread Derrik Pates

On Fri, May 17, 2002 at 07:20:47PM -0700, H. P. wrote:
 Have a Radeon VE, and Creative Savage4 video cards. 
 
 They work perfectly until I Try to load Commanche4
 
 Commanche4 plays the trailer then tries to load game but fails. 
 It crashes back to the desktop when game menu appears.
 
 Comanche4 C/S says that I need Signed or Certified Microsoft lab tested drivers 
even though the Directx drivers are 8.1 and pass the DXDIAG tests for Draw and 3d.
 
 None of the web sites have updates for this upgrade and fix.

Are you even running X? Doesn't sound like it to me. If you're running
Windows, as I suspect, you could call Microsoft, and see if they'll help
you (not likely), or if either of your video card manufacturers will
(also not likely).

-- 
Derrik Pates  |   Sysadmin, Douglas School   |#linuxOS on EFnet
[EMAIL PROTECTED] | District (dsdk12.net)|#linuxOS on OPN
___
Xpert mailing list
[EMAIL PROTECTED]
http://XFree86.Org/mailman/listinfo/xpert



Re: [Xpert]Howto turn off virtual screens?

2002-05-13 Thread Derrik Pates

On Mon, May 13, 2002 at 12:12:38PM +0200, Eric Noack wrote:
 Does anybody know how to turn of the bigger virtual desktop size, when changing to 
lower resolutions?
 
 Once upon a time(was it an 3.x Xserver, i dont remember) it was possible to set the
 Virtual
 Option in the Display Subsection of the Screen section
 in XFree86Config
 to a lower setting than your maximum mode, which lead to using of a higher virtual 
resolution only if i changed to a higher mode than i set in Virtual.

I don't think so. That's not how X works, afaik. With X, once the root
window (and its dimensions) are established, it stays that way
throughout the session.

 Now my question again, is it possible to turn the usage of a virtual desktop off? 
(and how)

Wait for the XRandR (resize/rotate) extension to make its way into a
release version of XFree86. It will add provisions for changing the real
desktop dimensions (and dynamically rotating the display in mid-session
too).

 PPS: Of course its possible to work around my specific problem by letting wine grab 
the cursor completely, or not to use Windoze games at all, but that's another 
question g

Use a recent version of Wine, install the WineSetup tool, and enable
DirectX cursor grabbing. That will allow Windows programs that use the
DirectInput APIs to keep the mouse cursor within their window (this is
off by default).

-- 
Derrik Pates  |   Sysadmin, Douglas School   |#linuxOS on EFnet
[EMAIL PROTECTED] | District (dsdk12.net)|#linuxOS on OPN
___
Xpert mailing list
[EMAIL PROTECTED]
http://XFree86.Org/mailman/listinfo/xpert



Re: [Xpert]Re: Testing opengl

2002-04-21 Thread Derrik Pates

On Fri, Apr 19, 2002 at 06:27:16AM -0500, Ian Truelsen wrote:
 Running Linux with 2.4.19 kernel and X 4.2.0 

 I am at 16bpp, as I read the notes on the card. 

 Yeah, I have read everything that I could find there. 

Have you tried running 'LIBGL_DEBUG=verbose glxinfo' at a prompt in X?
The dual-head setup may stop DRI from working correctly, unless you
don't have libglide3x.so (or don't have the right one).

-- 
Derrik Pates  |   Sysadmin, Douglas School   |#linuxOS on EFnet
[EMAIL PROTECTED] | District (dsdk12.net)|#linuxOS on OPN
___
Xpert mailing list
[EMAIL PROTECTED]
http://XFree86.Org/mailman/listinfo/xpert



Re: [Xpert]XFree86 version 4.2.0 and SiS 6326 compatibility

2002-04-11 Thread Derrik Pates

On Wed, Apr 10, 2002 at 07:49:32PM -0700, Dick Lewis wrote:
   1. While libglide2x.so wasn't present on my system, that
  in itself was likely not a problem;

Why would you need this anyway? libglide2x is only for standalone 3D
mode on Voodoo3/Banshee and for all 3D only Voodoo boards (Voodoo and
Voodoo2). Unless you want the X server to use the 'glide' driver to use
an old Voodoo/Voodoo2 board as a second head, you don't need the
libglide2x.so library. 

Repeat: Glide is 100% NOT relevant with an SiS chipset.

   2. That I could be using an obsolete X configuration file.
Could not load the shared library file for Glide: libglide2x.so! 

What are you getting this from? I'm not convinced you're getting this
from the X server itself. The error you got in your X server output log
seems to indicate the new SiS drivers just don't grok the chip for some
reason (old rev? old BIOS? something else?)

-- 
Derrik Pates  |   Sysadmin, Douglas School   |#linuxOS on EFnet
[EMAIL PROTECTED] | District (dsdk12.net)|#linuxOS on OPN
___
Xpert mailing list
[EMAIL PROTECTED]
http://XFree86.Org/mailman/listinfo/xpert



Re: [Xpert]xvideo overlay and xinerama

2002-03-24 Thread Derrik Pates

On Sat, Mar 23, 2002 at 06:30:25PM -0800, Mark Vojkovich wrote:
If in a non-Xinerama multihead situation the overlay on each
 head works individually, then it should work in Xinerama.
 
When you say you were able to get either head to work, was
 that with both cards running simultaneouly in a non-Xinerama
 multihead mode?

I've had dual Voodoos (a Voodoo3 2000 PCI and a Voodoo5 5500 AGP) with
XVideo working across both (could stretch a widescreen movie trailer,
playing in Xine, across the 2, and it worked great - CPU use  10% all
the time, on a PIII-650). Is this somehow specific to just having cards
of the same type? Or just paired Voodoo boards?

-- 
Derrik Pates  |   Sysadmin, Douglas School   |#linuxOS on EFnet
[EMAIL PROTECTED] | District (dsdk12.net)|#linuxOS on OPN
___
Xpert mailing list
[EMAIL PROTECTED]
http://XFree86.Org/mailman/listinfo/xpert



Re: [Xpert]Extension GLX missing

2002-03-12 Thread Derrik Pates

On Wed, Mar 13, 2002 at 02:46:31PM +1100, Dave Horsfall wrote:
 Yes, I searched the archives, but the solutions therein don't seem to
 apply here.

You're displaying X apps to an Xplora Xterminal. It only knows the X
extensions that are built into Xncdxpl (not a whole lot). The extensions
involve X-server-side code - since the X server doesn't support them,
you can't use them. Sorry.

-- 
Derrik Pates  |   Sysadmin, Douglas School   |#linuxOS on EFnet
[EMAIL PROTECTED] | District (dsdk12.net)|#linuxOS on OPN
___
Xpert mailing list
[EMAIL PROTECTED]
http://XFree86.Org/mailman/listinfo/xpert



Re: [Xpert]Cannot establish any listening sockets

2002-02-26 Thread Derrik Pates

On Sat, Feb 23, 2002 at 09:19:14AM -0800, Doug Fong wrote:
 and when i attempt to run startx I get the following error
 Fatal server error:
 Cannot establish any listening sockets - Make sure an X server isn't
 already running.

Have you used either 'lsof' or 'netstat -lnp --inet' to see what's
holding the socket open? It has to be something holding it open, and one
of those should be able to help you find out what the problem is.

-- 
Derrik Pates  |   Sysadmin, Douglas School   |#linuxOS on EFnet
[EMAIL PROTECTED] | District (dsdk12.net)|#linuxOS on OPN
___
Xpert mailing list
[EMAIL PROTECTED]
http://XFree86.Org/mailman/listinfo/xpert



Re: [Xpert]NVIDIA drivers on PowerPC

2002-02-09 Thread Derrik Pates

On Fri, 8 Feb 2002, Isreal Packard wrote:

I hope this is an appropriate question to direct here. I saw a reference
 on Usenet that indicates I might get some direction here. Basically, as a
 complete novice, I'd like to know how to get the display working on my
 GeForce 2?

Well, the complexity of that is going to afffected by a few things. What
type of display are you trying to drive? VGA, or a DVI flat panel? What
version of XFree86 do you have installed? What kernel?

Derrik Pates  |   Sysadmin, Douglas School   |#linuxOS on EFnet
[EMAIL PROTECTED] | District (dsdk12.net)|#linuxOS on OPN

___
Xpert mailing list
[EMAIL PROTECTED]
http://XFree86.Org/mailman/listinfo/xpert



Re: [Xpert]very confused about Direct rendering

2002-02-08 Thread Derrik Pates

On 8 Feb 2002, Brenden Conte wrote:

 So i have a banshee (Creative labs, 16M, AGP), on a 2.4.17 debian
 system.  and i'm trying to get as much out of it as i can, being an old
 card.  I've been trying to get direct rendering to work, using (if
 anyone is familiar with them) mesag3-glide2 and the device3dfx module.
 glxinfo works and says Direct rendering is on, but i need to set
 MESA_GLX_FX.  once i do that and run it again, it locks the system hard
 (no matter if i specify window or fullscreen)

You're using the wrong GL library then. Are you using XFree86 4.1.0? If
so, you should have 'xlibmesa3' installed on your system. You'll probably
also have to install the 'libglide3' package, and build and load the
'tdfx' DRM module (it's in the kernel).

 i've been fishing around the web for answers, which made me even more
 confused as ever... :-\  Do i need to use the kernel module for the
 banshee as well? should i not build it (its built as a module now)? am i
 doing it all wrong, using outdated methods?  Do i need a certain glide
 version for this card?

The 'libglide3' package will link the right Glide lib for your needs
automatically. But yes, the method you're using is severely outdated, if
you're using XFree 4.1.0 (which if you have glxinfo, I suspect is the
case).

 I would love to get glide working on this card, and any help (including
 directions to other people who would know more) would be very much
 appreciated...

You won't be using Glide itself, but OpenGL (which links the glide
library through the tdfx_drm.so module). Once you've installed libglide3,
put the whole concept of Glide as an application-level interface out of
your head, it's 100% totally irrelevant.

Derrik Pates  |   Sysadmin, Douglas School   |#linuxOS on EFnet
[EMAIL PROTECTED] | District (dsdk12.net)|#linuxOS on OPN

___
Xpert mailing list
[EMAIL PROTECTED]
http://XFree86.Org/mailman/listinfo/xpert



Re: [Xpert]How exactly do I get DRI working ?!?!

2002-02-01 Thread Derrik Pates

On Sun, 27 Jan 2002, OS wrote:

 Okay, here we go ! I've tried the following on an ATI Rage 128 RL and a Rage
 Mobility Pro chipset.

The R128 will do DRI, the Mobility Pro (which I believe is a
Mach64-derived design) will not - at least, not yet. Support for
Mach64-based 3D-enabled cores is in the works.

 I modprobed agpgart and it duly appeared with a useage counter of 0. I then
 modprobed r128. It too showed a useage count of 0, but the agpgart useage
 count had gone up to 1.

That's right - because the r128 DRM uses the AGPGART interface.

 I couldn't find a module called mtrr, perhaps Mandrake doesn't ship with that
 module ?

It's never a module, it's either built into your kernel or it's not. If
/proc/mtrr exists, then (a) it's built into the kernel, and (b) your CPU
supports MTRR (it's a CPU feature - K6-2s stepping 8 and up, K6-3s,
Athlons, some Cyrix chips, and P2/P3/P4 (including Celerons) have MTRRs or
a close analog).

 The XFree86 log file indicates that XF doesn't baulk at any of the modules it
 loads. It happily loads dri, drm, glx etc.. But no where does it say dri
 enabled. Anyway, it's attached.

It looks like you may need to add a line that says 'Load dri' in your
Module section. Otherwise the X server itself doesn't know about DRI, or
the setup for it or anything else.

Derrik Pates  |   Sysadmin, Douglas School   |#linuxOS on EFnet
[EMAIL PROTECTED] | District (dsdk12.net)|#linuxOS on OPN

___
Xpert mailing list
[EMAIL PROTECTED]
http://XFree86.Org/mailman/listinfo/xpert



Re: [Xpert]Re: Second card in the Mac G4

2002-01-30 Thread Derrik Pates

On Wed, 30 Jan 2002 [EMAIL PROTECTED] wrote:

 I wonder if we could use RAC for this, anyone with more knowledge of how RAC
 works cares to comment ?

 It's my understanding that the domain patches (which are now in the trunk
 afaik) will handle this. I haven't had a chance to try them out yet on
 PPC though.

Don't these cards still need to be softbooted somehow though? Or is that
really all it'll take to make them work?

Derrik Pates  |   Sysadmin, Douglas School   |#linuxOS on EFnet
[EMAIL PROTECTED] | District (dsdk12.net)|#linuxOS on OPN

___
Xpert mailing list
[EMAIL PROTECTED]
http://XFree86.Org/mailman/listinfo/xpert



Re: [Xpert]drm and kernel

2002-01-26 Thread Derrik Pates

On Sat, 26 Jan 2002, Jeremy West wrote:

 Please!! Please!!! I have been fighting dri for the last couple of
 months.  If someone could help me I would be really happy.  In my XFree
 log it says :

 What does this mean.  I went into my /dev/dri directory and found that I
 have a card1, card2, and card3.  So I used mknod, and created a card0
 based off of the major and minor numbers for the others.  I then
 restarted the Xserver.  I get the same error.  I went back into the
 /dev/dri folder, and  the node had been deleted.  What is up with that.
  The kernel module radeon is also not loaded, which is what I need.
  If I shut down X and run insmod on that module it comes back with lots
 of errors.

If you have a custom-built kernel, make sure you have AGP GART support,
and Radeon DRM support, either built into your kernel or built as modules.
DRI will not work without them. If you have the module already as you
claim, what missing symbols is it complaining about? Maybe you need to
rebuild the kernel?

Derrik Pates  |   Sysadmin, Douglas School   |#linuxOS on EFnet
[EMAIL PROTECTED] | District (dsdk12.net)|#linuxOS on OPN

___
Xpert mailing list
[EMAIL PROTECTED]
http://XFree86.Org/mailman/listinfo/xpert



Re: [Xpert]XFree 4.1.0 and DRM kernel modules for XFree86 4.2.0

2002-01-24 Thread Derrik Pates

On Thu, 24 Jan 2002, Cyril Zlachevsky wrote:

   I have 3dfx velocity 100 videocard, Athlon 1.33 GHz on SiS 735 chipset

3Dfx Velocity 100? I've never heard of this. How closely related to the
Voodoo3/5 is it? Does it use the 'tdfx' drivers? I find the same sort of
issues running UT on my home box with a Voodoo5 and a P3-650.

  Can this be related to the recently discovered AMD-AGP bug?

This would not be likely, since if it's a Voodoo-based card (as I
suspect), it doesn't actually make use of the GART anyway - it just uses
the AGP bus as a faster PCI bus, basically.

 About  wrong texture colours with Unreal Tournament 436 - on XFree 4.2.0 I
 can't found this problem - possibly, error fixed in 4.2.0. But I found
 another problem - UT freeze on high rendering quality in game video
 settings.

Do you see rainbowed textures in the opening sequence in particular?
Also, texture corruption in the menus?

Derrik Pates  |   Sysadmin, Douglas School   |#linuxOS on EFnet
[EMAIL PROTECTED] | District (dsdk12.net)|#linuxOS on OPN

___
Xpert mailing list
[EMAIL PROTECTED]
http://XFree86.Org/mailman/listinfo/xpert



Re: [Xpert]X crashes .

2002-01-17 Thread Derrik Pates

On Wed, 16 Jan 2002, Vincenzo Di Massa wrote:

 X 4.1.0 crashes when I try tu run 3d apps.
 I think I configured everything right.
 Any suggestion?
 I don't know what could it be wrong.
 Can you tell me where to look to find the problem?.

Can you be in the least specific? Like, what video card you're using? What
distribution? Maybe your XF86Config file or something? There are several
boards that have DRI (hardware 3D) support, and many possible
configurations, so you really need to narrow it down.

Derrik Pates  |   Sysadmin, Douglas School   |#linuxOS on EFnet
[EMAIL PROTECTED] | District (dsdk12.net)|#linuxOS on OPN

___
Xpert mailing list
[EMAIL PROTECTED]
http://XFree86.Org/mailman/listinfo/xpert



Re: [Xpert]ATI Radeon VE and xinerama

2002-01-14 Thread Derrik Pates

On 14 Jan 2002, Niclas Sodergard wrote:

 I have a ATI Radeon VE and I am trying to get xinerama to work with
 that. It seems to configure both screens but the second screen remains
 black. It doesn't matter which way I connect them. The second screen
 seems to get some kind of signal since the No signal-banner disappers
 when I start X.

Have you tried specifying 'Option CrtScreen' for the second-head
configuration? I remember seeing something about that being a necessary
option. (FYI, I don't have a laptop with a Radeon, so I don't know if this
does anything. Might work, might do nothing, might induce meltdown. :)

Derrik Pates  |   Sysadmin, Douglas School   |#linuxOS on EFnet
[EMAIL PROTECTED] | District (dsdk12.net)|#linuxOS on OPN


___
Xpert mailing list
[EMAIL PROTECTED]
http://XFree86.Org/mailman/listinfo/xpert



Re: [Xpert]radeon and glx problem (RTCW hangs system)

2002-01-14 Thread Derrik Pates

On Mon, 14 Jan 2002, Jacek [iso-8859-2] Pop³awski wrote:

 [drm:radeon_freelist_get] *ERROR* returning NULL!

I frequently see a similar message (but s/radeon/r128/) when I try to use
AGP on this iBook with a Rage Mobility M3, and it frequently leads up to a
system hang here as well (not running RtCW, of course, but trying to run
Quake or TuxRacer). I don't know if it's an AGPGART problem, or a DRM
problem. With a Rage128 descendant, of course, PCI GART can be used, but
it's a good bit slower than 2X AGP would be...

Derrik Pates  |   Sysadmin, Douglas School   |#linuxOS on EFnet
[EMAIL PROTECTED] | District (dsdk12.net)|#linuxOS on OPN

___
Xpert mailing list
[EMAIL PROTECTED]
http://XFree86.Org/mailman/listinfo/xpert



Re: [Xpert]Unresolved symbols in r128_drv

2002-01-12 Thread Derrik Pates

On Sat, 12 Jan 2002, Mike Heald wrote:

 Thanks, but I cannot find a DRM kernel module
 in CVS, I have checkedout xc and done
 make World, but there is no r128.o when its done.
 Is there another module i should be checking out?

It's in the CVS tree. Look under
xc/programs/Xserver/hw/xfree86/os-support/linux/drm/kernel - yes, it's a
long path, but it's in there. Do 'make -f Makefile.linux' in that
directory, and it'll make all the DRM modules (r128, radeon, tdfx, ...),
or do 'make -f Makefile.linux r128.o' for just the Rage128 DRM.

Derrik Pates  |   Sysadmin, Douglas School   |#linuxOS on EFnet
[EMAIL PROTECTED] | District (dsdk12.net)|#linuxOS on OPN

___
Xpert mailing list
[EMAIL PROTECTED]
http://XFree86.Org/mailman/listinfo/xpert



Re: [Xpert]DGA requires SUID/root access???

2002-01-10 Thread Derrik Pates

On Thu, 10 Jan 2002 [EMAIL PROTECTED] wrote:

 That's true for the install but not when the user is running the app, which
 is what I'm worried about.
 Surely, we don't need SUID just to display graphics?

For the program that's using DGA to open and mmap() /dev/mem, it has to
be root. That's part of doing DGA. That's why it's called _direct_.

Derrik Pates  |   Sysadmin, Douglas School   |#linuxOS on EFnet
[EMAIL PROTECTED] | District (dsdk12.net)|#linuxOS on OPN

___
Xpert mailing list
[EMAIL PROTECTED]
http://XFree86.Org/mailman/listinfo/xpert



Re: [Xpert]DGA requires SUID/root access???

2002-01-10 Thread Derrik Pates

On Thu, 10 Jan 2002, David Balazic wrote:

 Couldn't the X server process open /dev/mem and then pass the file descriptor
 to the X client ?

They're not running in the same process context. The X client communicates
with the X server via either TCP sockets or UNIX domain sockets (UNIX
domain sockets for local apps), but they run as completely separate
processes.

Derrik Pates  |   Sysadmin, Douglas School   |#linuxOS on EFnet
[EMAIL PROTECTED] | District (dsdk12.net)|#linuxOS on OPN

___
Xpert mailing list
[EMAIL PROTECTED]
http://XFree86.Org/mailman/listinfo/xpert



Re: [Xpert]DGA requires SUID/root access???

2002-01-10 Thread Derrik Pates

Sorry if this is a sort-of-repeat, I thought I sent it.

On Thu, 10 Jan 2002, Billy Biggs wrote:

   1. We don't have kernel-level video drivers, so we can't safely mmap
 to a user app sections of video memory.

Right.

   2. Giving a user full access to the framebuffer is dangerous so only
 let root do it.

Sort of. In theory, it could be made sufficiently safe (that's what the
GGI crew wanted to do with KGI).

   That is, there's no reason you need full access to /dev/mem just to
 write to the framebuffer, it's just some silliness because of the way
 everything is architectured.   I'd love to find out the real reason
 though.

If we had framebuffer drivers in the kernel for everything, it'd be OK
(then you just have to mediate access between X and the kernel framebuffer
driver). However, in the current situation, mapping /dev/mem is necessary.
Userspace processes don't know anything about the real physical addressing
of memory normally (and they don't have to care - the whole beauty of
virtual memory systems), allowing each process to kick around in its
littlee sandbox without bugging anybody else. However, to directly access
the framebuffer as it stands, via DGA, a process tells X it wants to do
DGA - then the X server tells it the framebuffer base address (an offset
into physical memory where the display device's framebuffer memory lives),
the size of it, and the line width. The process then opens /dev/mem,
mmap()s the framebuffer by offset and length. Unfortunately, allowing
access to /dev/mem opens us up to an untrusted program proceeding to
scribble on whatever it wants in system RAM. (Among all the other things
it can do as root.) Which is why you have to be picky about what programs
you make SUID root, of course. :)

One good reason why the DirectFB project wouldn't be such a bad thing.

Derrik Pates  |   Sysadmin, Douglas School   |#linuxOS on EFnet
[EMAIL PROTECTED] | District (dsdk12.net)|#linuxOS on OPN



___
Xpert mailing list
[EMAIL PROTECTED]
http://XFree86.Org/mailman/listinfo/xpert



Re: [Xpert]Radeon 7000 QD (32Mb DDR)

2002-01-05 Thread Derrik Pates

On Sat, 5 Jan 2002, Mike Mestnik wrote:

 Is this really worth trying, dose this enable direct rendering?
 Can you play Quake Arena (The only thing that matters!)?
 I have a Radeon 8500 might this work with my card to?

No. The Radeon 7000/7200/7500 boards's display controller is closely
related to the original Radeon, enough so that this should theoretically
work. The 8500, however, has some significant differences from both the
7x00 Radeons and the original Radeon that make it truly incompatible
with XFree86 4.1.0's Radeon driver. However, there's a 2D-only Radeon 8500
driver in XFree CVS now, which would at least get the board working.

I don't know ATI's intentions as far at Linux 3D support for the Radeon
8500 goes though. Anyone else know what the scoop is there?

Derrik Pates  |   Sysadmin, Douglas School   |#linuxOS on EFnet
[EMAIL PROTECTED] | District (dsdk12.net)|#linuxOS on OPN

___
Xpert mailing list
[EMAIL PROTECTED]
http://XFree86.Org/mailman/listinfo/xpert



[Xpert]Softbooting video card on PPC with OF ROMs

2002-01-04 Thread Derrik Pates

Has there been any development in this department? I was wondering if the
Mac-On-Linux OpenFirmware/OpenPROM implementation (minimal, but maybe
enough) would be a potential basis to make something like this work, as
part of an int10-like module. Any thoughts?

Derrik Pates  |   Sysadmin, Douglas School   |#linuxOS on EFnet
[EMAIL PROTECTED] | District (dsdk12.net)|#linuxOS on OPN

___
Xpert mailing list
[EMAIL PROTECTED]
http://XFree86.Org/mailman/listinfo/xpert



Re: [Xpert]Voodoo 4 4500 AGP

2002-01-03 Thread Derrik Pates

On Thu, 3 Jan 2002, Paulo Vitor Magacho da Silva wrote:

 First, sorry for the very long email that I am sending.
 So you are saying that I don't need the tdfx_dri.so file anymore ?

You need that library. XFree's libGL is just an intermediary, it uses that
library to talk to 3Dfx boards to do direct rendering.

 (II) TDFX(0): [DRI] installation complete 
 (==) TDFX(0): Direct rendering enabled

 What it's wrong, what else do I need ? I have the opengl library provided by the 
XFree 4.1.0, I
 also have installed the Mesa-utils 3.4.1 libs and the glide libraries. I can't use 
any program that
 need hardware acceleration, such as tuxracer.

Say what? Please say you didn't overwrite your libGL. First, do 'ldd
/usr/X11R6/bin/glxinfo' and make sure it's linking the correct libGL.
Second, as Michel said, run 'LIBGL_DEBUG=1 glxinfo' and see if it spits
out any errors. You may need to look at /dev/dri/* and see if the
permissions are set correctly for you to access the DRI devnode. If not,
'man XF86Config' is your friend. Look for info on the DRI section.

Derrik Pates  |   Sysadmin, Douglas School   |#linuxOS on EFnet
[EMAIL PROTECTED] | District (dsdk12.net)|#linuxOS on OPN

___
Xpert mailing list
[EMAIL PROTECTED]
http://XFree86.Org/mailman/listinfo/xpert



Re: [Xpert]Abridged: Total frustration with XFree86 4.1.0!

2001-12-17 Thread Derrik Pates

On Mon, 17 Dec 2001, Mac Cody wrote:

 Interesting. :-\  I would have thought that interlace support
 would be in place since XFree86 is at version 4.1.0 already
 and support was there in XFree86 3.3.6.  Anywhere I could find
 out whether that is the case or not (short of investigating the
 sources)?  If interlace support is missing, I guess I'll have
 to fall back to XFree 3.3.6 for now [sigh!].

Well, XFree86 underwent some serious reworking with the debut of the 4.x
line - moving away from monolithic X servers, and moving towards modules
for everything, with one smaller base server. Not a lot of attention had
initially been paid to moving the old-style 3.x drivers for many old
boards (pre-ViRGE S3 boards, older Cirrus and Trident boards had also
fallen victim to this), as much of the redesign was intended to allow
support for more-advanced features like the XVideo and DRI subsystems.
Only recently have those really old drivers been ported forward to the new
modular architecture.

So you may need to stay with 3.3.x for now, or if you're up to it, try
hacking in interlaced-mode support. That's about all you can do, until
someone eventually gets the drivers for the older boards fully ported up.

Derrik Pates  |   Sysadmin, Douglas School   |#linuxOS on EFnet
[EMAIL PROTECTED] | District (dsdk12.net)|#linuxOS on OPN

___
Xpert mailing list
[EMAIL PROTECTED]
http://XFree86.Org/mailman/listinfo/xpert



Re: [Xpert]Secrets to getting XFree86 working on PPC?

2001-12-08 Thread Derrik Pates

On Sat, 8 Dec 2001, Mark Vojkovich wrote:

   I've been trying to get XFree86 CVS to work on Linux PPC.
 Is this expected to work?  I'm using the nv driver with
 NV11 with Option NoAccel, something which others have
 allegedly gotten to work.  This is on a YDL 2.0 system.
 Basically, the system hardlocks sometime after finishing
 the PreInit.

Ani Joshi made some patches against both XFree86 and the kernel sources to
adapt the nVidia drivers in each to work on big-endian systems (esp. PPC
systems). I don't know if his patch against the 'nv' driver in XFree ever
made into the CVS mainline tree. I've attached both patches to this
message, so if it's not in CVS, you can try the patch instead.

Derrik Pates  |   Sysadmin, Douglas School   |#linuxOS on EFnet
[EMAIL PROTECTED] | District (dsdk12.net)|#linuxOS on OPN


diff -uNr nv.orig/nv_driver.c nv/nv_driver.c

--- nv.orig/nv_driver.c Sun Oct  7 22:28:30 2001

+++ nv/nv_driver.c  Sun Oct  7 22:28:35 2001

@@ -885,7 +885,7 @@

 

 pNv-Primary = xf86IsPrimaryPci(pNv-PciInfo);

 

-#ifndef __alpha__

+#if !defined(__alpha__) || !defined(__powerpc__)

 /* Initialize the card through int10 interface if needed */

 #if 0

  if ( !pNv-Primary )

@@ -1280,9 +1280,9 @@



 pNv-FbMapSize = pScrn-videoRam * 1024;

 

+#if 0

 /* Read and print the Monitor DDC info */

 pScrn-monitor-DDC = NVdoDDC(pScrn);

-#if 0

 /*

  * This code was for testing. It will be removed as soon

  * as this is integrated into the common level.

@@ -1610,6 +1610,18 @@

 if ( pNv-Restore )

 (*pNv-Restore)(pScrn, vgaReg, nvReg, FALSE);

 

+#if X_BYTE_ORDER == X_BIG_ENDIAN

+/* turn on LFB swapping */

+{

+   unsigned char tmp;

+

+   VGA_WR08(pNv-riva.PCIO, 0x3d4, 0x46);

+   tmp = VGA_RD08(pNv-riva.PCIO, 0x3d5);

+   tmp |= (1  7);

+   VGA_WR08(pNv-riva.PCIO, 0x3d5, tmp);

+}

+#endif

+

 NVResetGraphics(pScrn);

 

 vgaHWProtect(pScrn, FALSE);

@@ -1933,7 +1945,14 @@

 vgaRegPtr vgaReg = pVga-SavedReg;

 

 DEBUG(xf86DrvMsg(pScrn-scrnIndex, X_INFO, NVSave\n));

+#if defined(__powerpc__)

+/* temporary hack to get around PowerMac's inability to save

+ * vga fonts and cmap, will find a better solution later

+ */

+vgaHWSave(pScrn, vgaReg, VGA_SR_MODE);

+#else

 vgaHWSave(pScrn, vgaReg, VGA_SR_CMAP|VGA_SR_MODE|VGA_SR_FONTS);

+#endif

 pNv-riva.UnloadStateExt(pNv-riva, nvReg);

 }

 

diff -uNr nv.orig/nv_xaa.c nv/nv_xaa.c

--- nv.orig/nv_xaa.cSun Oct  7 22:28:30 2001

+++ nv/nv_xaa.c Sun Oct  7 22:28:35 2001

@@ -596,8 +596,12 @@

 NVSubsequentMono8x8PatternFillRect;

 

 /* Color expansion */

-infoPtr-ScanlineCPUToScreenColorExpandFillFlags = 

+infoPtr-ScanlineCPUToScreenColorExpandFillFlags =

+#if X_BYTE_ORDER == X_BIG_ENDIAN

+   BIT_ORDER_IN_BYTE_MSBFIRST | 

+#else

BIT_ORDER_IN_BYTE_LSBFIRST | 

+#endif

NO_PLANEMASK | 

CPU_TRANSFER_PAD_DWORD |

LEFT_EDGE_CLIPPING |

diff -uNr nv.orig/riva_hw.c nv/riva_hw.c

--- nv.orig/riva_hw.c   Sun Oct  7 22:28:30 2001

+++ nv/riva_hw.cSun Oct  7 22:28:35 2001

@@ -1656,6 +1656,11 @@

 int pan= (start  3)  1;

 unsigned char tmp;

 

+#if X_BYTE_ORDER == X_BIG_ENDIAN

+/* turn on big endian register access */

+chip-PMC[0x0004/4] = 0x0101;

+#endif

+

 /*

  * Unlock extended registers.

  */

diff -uNr nv.orig/riva_hw.h nv/riva_hw.h

--- nv.orig/riva_hw.h   Sun Oct  7 22:28:30 2001

+++ nv/riva_hw.hSun Oct  7 22:28:35 2001

@@ -60,8 +60,12 @@

 typedef volatile struct

 {

 U032 reserved00[4];

+#if X_BYTE_ORDER == X_BIG_ENDIAN

+U032 FifoFree;

+#else

 U016 FifoFree;

 U016 Nop;

+#endif

 U032 reserved01[0x0BB];

 U032 Rop3;

 } RivaRop;

@@ -71,8 +75,12 @@

 typedef volatile struct

 {

 U032 reserved00[4];

+#if X_BYTE_ORDER == X_BIG_ENDIAN

+U032 FifoFree;

+#else

 U016 FifoFree;

 U016 Nop;

+#endif

 U032 reserved01[0x0BD];

 U032 Shape;

 U032 reserved03[0x001];

@@ -86,8 +94,12 @@

 typedef volatile struct

 {

 U032 reserved00[4];

+#if X_BYTE_ORDER == X_BIG_ENDIAN

+U032 FifoFree;

+#else

 U016 FifoFree;

 U016 Nop;

+#endif

 U032 reserved01[0x0BB];

 U032 TopLeft;

 U032 WidthHeight;

@@ -98,8 +110,12 @@

 typedef volatile struct

 {

 U032 reserved00[4];

+#if X_BYTE_ORDER == X_BIG_ENDIAN

+U032 FifoFree;

+#else

 U016 FifoFree;

 U016 Nop[1];

+#endif

 U032 reserved01[0x0BC];

 U032 Color;

 U032 reserved03[0x03E];

@@ -112,8 +128,12 @@

 typedef volatile struct

Re: [Xpert]popup windows

2001-12-04 Thread Derrik Pates

On Tue, 4 Dec 2001, Daniel Secrieru wrote:

 They don't have a title bar.

And that's why there are WM hints, so you can tell the WM hey, don't
decorate me!

Derrik Pates  |   Sysadmin, Douglas School   |#linuxOS on EFnet
[EMAIL PROTECTED] | District (dsdk12.net)|#linuxOS on OPN

___
Xpert mailing list
[EMAIL PROTECTED]
http://XFree86.Org/mailman/listinfo/xpert



Re: [Xpert]Re: Rage128 (esp. RageM3) bad line length issue - Ifigured it out...

2001-12-04 Thread Derrik Pates

On 4 Dec 2001, Michel Dänzer wrote:

 If that's the problem, then the fix should be to query the pitch (i.e.
 displayWidth) from the framebuffer device. There is an
 fbdevHWGetLineLength() function for this. Derrik says it's also broken
 without Option UseFBDev though.

Yes, when the X server worked on PPC without UseFBDev - which it doesn't
right now - I did see the same effect, regardless of whether I specified
UseFBDev or not.

Also, is the fix that you came up with available as a patch against the
4.1.0.1 tree that Debian is using at present, so I can just build my own
local X server package and use that until 4.2.0 is released, and packages
for it make their way into Sid? I could, I suppose, use a local parallel X
install, but I'd rather not, if for no other reason than saving disk space
for other stuff. :)

Derrik Pates  |   Sysadmin, Douglas School   |#linuxOS on EFnet
[EMAIL PROTECTED] | District (dsdk12.net)|#linuxOS on OPN

___
Xpert mailing list
[EMAIL PROTECTED]
http://XFree86.Org/mailman/listinfo/xpert



Re: [Xpert]No Screens on XFree86

2001-12-01 Thread Derrik Pates

On Sun, 2 Dec 2001, David Loszewski wrote:

 I just installed XFree86-4.1.0, after I finished installing it I did the
 'XFree86 -configure' and it made my configuration file, so then I type
 'startx' and this is the error it gives me:

Did you copy the created config file (./XF86Config.new) to
/etc/X11/XF86Config? If not, that probably explains why the log is
claiming that it couldn't open the config file.

Derrik Pates  |   Sysadmin, Douglas School   |#linuxOS on EFnet
[EMAIL PROTECTED] | District (dsdk12.net)|#linuxOS on OPN

___
Xpert mailing list
[EMAIL PROTECTED]
http://XFree86.Org/mailman/listinfo/xpert



Re: [Xpert]ATI Question

2001-11-30 Thread Derrik Pates

On Thu, 29 Nov 2001, Gene Dascher wrote:

 Does anyone know who I need to talk to about seeing if it would be possible
 to get the ATI Rage128 Mobility card (I have a Dell Inspiron 5000e) in
 dual-display mode under XFree86?  It works wonderfully under Win98SE (ATI's
 drivers), so I don't see why it isn't POSSIBLE under X.

Others have discussed this very thing. (Check the list archives for more
info.) The only real problem is getting the right docs to the right people
to make it happen.

Derrik Pates  |   Sysadmin, Douglas School   |#linuxOS on EFnet
[EMAIL PROTECTED] | District (dsdk12.net)|#linuxOS on OPN


___
Xpert mailing list
[EMAIL PROTECTED]
http://XFree86.Org/mailman/listinfo/xpert



Re: Other multihead setups (was Re: [Xpert]Matrox G450AGP + G450PCI)

2001-09-20 Thread Derrik Pates

On Thu, 20 Sep 2001, Michel Dänzer wrote:

 It should work. What version are you trying? I seem to remember Xinerama
 problems in 4.1.0 which are fixed in CVS.

I'm using thhe 4.1.0-6 debs at present. The system I'm using I haven't
even gotten around to trying to enable Xinerama - it always dies with a
signal 11 (even avfter moving the int10 module out of
/usr/X11R6/lib/modules, as one message I'd seen about a similar setup
suggested).

Derrik Pates  |   Sysadmin, Douglas School   |#linuxOS on EFnet
[EMAIL PROTECTED] | District (dsdk12.net)|#linuxOS on OPN

___
Xpert mailing list
[EMAIL PROTECTED]
http://XFree86.Org/mailman/listinfo/xpert



Re: [Xpert]EGA-Cards on XFree ?

2001-09-19 Thread Derrik Pates

On Wed, 19 Sep 2001, Michael Hess wrote:

 I have a nice display (lcd) with 640x480 but it needs to be driven by
 digital signals. My Idee was to take an old EGA-Card i have here (AST3G)
 and if i'm right, EGA is digital. Now my question is, does anyone know,
 if EGA-cards are supported in anyway by XFree ? (And which one are
 supported ?)

EGA? EGA is _not_ digital, it's _very_ analog, and pretty old. Far as I
know, even XFree 3.3.x didn't support anything less than a standard VGA
display device. Are you sure it's an EGA card you have?

Derrik Pates  |   Sysadmin, Douglas School   |#linuxOS on EFnet
[EMAIL PROTECTED] | District (dsdk12.net)|#linuxOS on OPN

___
Xpert mailing list
[EMAIL PROTECTED]
http://XFree86.Org/mailman/listinfo/xpert



Other multihead setups (was Re: [Xpert]Matrox G450AGP + G450PCI)

2001-09-16 Thread Derrik Pates

Ok, not exactly related, but I wanted to find out if it would work - has
anyone been able to get more than one ATI Rage128 working in a system?
Particularly in a non-x86 system (in my case an Apple PPC G3 box)? I've
tried to get such a setup working, but XFree86 (including the 4.1.0
release bins) just dies with signal 11 every time (after coming up with
a blue root-weave-ish image). Has anyone gotten this to work? Will it ever
work?

Derrik Pates  |   Sysadmin, Douglas School   |#linuxOS on EFnet
[EMAIL PROTECTED] | District (dsdk12.net)|#linuxOS on OPN

___
Xpert mailing list
[EMAIL PROTECTED]
http://XFree86.Org/mailman/listinfo/xpert