aty128 and offb and linux-2.3.99pre6

aty128 and offb and linux-2.3.99pre6

[Chas Williams]

i think i have spotted an oversite, offb.c, the aty128 probing should
come after the offb probing otherwise, my aty128 comes up as an
unsupported framebuffer.  the following patch 'correctly' recoginizes
my aty128 now:

--- fbmem.c.000 Tue Apr 18 17:51:12 2000
+++ fbmem.c     Tue Apr 18 17:52:39 2000
@@ -160,15 +160,15 @@
 #ifdef CONFIG_FB_ATY
        { "atyfb", atyfb_init, atyfb_setup },
 #endif
+#ifdef CONFIG_FB_ATY128
+       { "aty128fb", aty128fb_init, aty128fb_setup },
+#endif
 #ifdef CONFIG_FB_OF
        /*
         * Offb must be initialized _after_ all other frame buffer devices
         * that use PCI probing and PCI resources! [ Geert ]
         */
        { "offb", offb_init, offb_setup },
-#endif
-#ifdef CONFIG_FB_ATY128
-       { "aty128fb", aty128fb_init, aty128fb_setup },
 #endif
 #ifdef CONFIG_FB_IGA
         { "igafb", igafb_init, igafb_setup },


** Sent via the linuxppc-dev mail list. See http://lists.linuxppc.org/

Re: aty128 and offb and linux-2.3.99pre6

[Nelson Abramson]

Chas Williams wrote:

> i think i have spotted an oversite, offb.c, the aty128 probing should
> come after the offb probing otherwise, my aty128 comes up as an
> unsupported framebuffer.  the following patch 'correctly' recoginizes
> my aty128 now:

<snip:patch>

Does this fix the problem w/ the broken 2.3.99 aty128pro driver?

--Nelson Abramson


--
"Apparently if you play the Windows NT CD backwards you hear satanic messages"
"You think that's bad, if you play it forwards it installs Windows NT!"
--orlando


** Sent via the linuxppc-dev mail list. See http://lists.linuxppc.org/

Re: aty128 and offb and linux-2.3.99pre6

[Kevin B. Hendricks]

Hi,

I submitted my use_of_pll.patch to to Brad and Anthony (the maintainers of
the aty128fb.c driver) and the changes to allow the rage128pro card to work
will eventually make it into the latest development 2.3.99X kernel but it
has not as of yet (today's rsync from Paul devel tree) made it in yet.

You can pretty much hand apply the patch easily and that should help make
things work with 2.3.99preX kernels and rage128pros.

Kevin


At 13:45 -0400 4/19/00, Nelson Abramson wrote:
>Chas Williams wrote:
>
>> i think i have spotted an oversite, offb.c, the aty128 probing should
>> come after the offb probing otherwise, my aty128 comes up as an
>> unsupported framebuffer.  the following patch 'correctly' recoginizes
>> my aty128 now:
>
><snip:patch>
>
>Does this fix the problem w/ the broken 2.3.99 aty128pro driver?
>
>--Nelson Abramson
>
>
>--
>"Apparently if you play the Windows NT CD backwards you hear satanic messages"
>"You think that's bad, if you play it forwards it installs Windows NT!"
>--orlando
>
>


--
Kevin B. Hendricks, Associate Professor of Operations and Information
Technology
Richard Ivey School of Business, University of Western Ontario
London, Ontario  N6A-3K7  CANADA
khendricks@ivey.uwo.ca, (519) 661-3874, fax: 519-661-3959


** Sent via the linuxppc-dev mail list. See http://lists.linuxppc.org/

Re: aty128 and offb and linux-2.3.99pre6

[Geert Uytterhoeven]

On Wed, 19 Apr 2000, Chas Williams wrote:
> i think i have spotted an oversite, offb.c, the aty128 probing should
> come after the offb probing otherwise, my aty128 comes up as an
> unsupported framebuffer.  the following patch 'correctly' recoginizes
> my aty128 now:

I thought that was already fixed, but it wasn't, so I've just sent it to Linus.

What about other frame buffer devices, like e.g. matroxfb? Does it still work
after this patch? If yes, we can get rid of matrox_of_init() as well.

--- geert-2.3.99-pre5/drivers/video/fbmem.c.orig	Wed Apr 19 22:48:17 2000
+++ geert-2.3.99-pre5/drivers/video/fbmem.c	Wed Apr 19 22:48:26 2000
@@ -163,6 +163,9 @@
 #ifdef CONFIG_FB_ATY128
 	{ "aty128fb", aty128fb_init, aty128fb_setup },
 #endif
+#ifdef CONFIG_FB_MATROX
+	{ "matrox", matroxfb_init, matroxfb_setup },
+#endif
 #ifdef CONFIG_FB_OF
 	/*
 	 * Offb must be initialized _after_ all other frame buffer devices
@@ -203,9 +206,6 @@
 #ifdef CONFIG_FB_HGA
 	{ "hga", hgafb_init, hgafb_setup },
 #endif
-#ifdef CONFIG_FB_MATROX
-	{ "matrox", matroxfb_init, matroxfb_setup },
-#endif
 #ifdef CONFIG_FB_HP300
 	{ "hpfb", hpfb_init, hpfb_setup },
 #endif

I'd really like to get rid of all these:

    s3triofb_init_of()
    imsttfb_of_init()
    chips_of_init()
    matrox_of_init()
    control_of_init()
    valkyrie_of_init()
    platinum_of_init()
    clgen_of_init()

These are all PCI based, so they can be found using PCI probing and marked
busy using PCI resource management.

When creating this list, I noticed that

  - clgenfb is already initialized before offb, so I guess clgen_of_init() is
    dead code already.

  - chipsfb is always initialized from offb, there's no corresponding entry in
    fbmem.c:fb_drivers[].

Gr{oetje,eeting}s,

						Geert

--
Geert Uytterhoeven -- Linux/{m68k~Amiga,PPC~CHRP} -- geert@linux-m68k.org

In personal conversations with technical people, I call myself a hacker. But
when I'm talking to journalists I just say "programmer" or something like that.
							    -- Linus Torvalds


** Sent via the linuxppc-dev mail list. See http://lists.linuxppc.org/

Re: aty128 and offb and linux-2.3.99pre6

[Martin Costabel]

Geert Uytterhoeven wrote:

> I'd really like to get rid of all these:
>
>     s3triofb_init_of()
>     imsttfb_of_init()
>     chips_of_init()
>     matrox_of_init()
>     control_of_init()
>     valkyrie_of_init()
>     platinum_of_init()
>     clgen_of_init()
>
> These are all PCI based, so they can be found using PCI probing and marked
> busy using PCI resource management.

Please leave the support for valkyriefb as it is. It may not be perfect,
but it works well enough, even with 2.3.99-pre6. My attempts to  switch
it to the new PCI resource management failed miserably. I am not sure if
it has everything that makes a true PCI card.

--
Martin

** Sent via the linuxppc-dev mail list. See http://lists.linuxppc.org/

Re: aty128 and offb and linux-2.3.99pre6

[Michel Lanners]

Hi all,

On  19 Apr, this message from Geert Uytterhoeven echoed through cyberspace:
> What about other frame buffer devices, like e.g. matroxfb? Does it still work
> after this patch? If yes, we can get rid of matrox_of_init() as well.

It does work... but it changes the order my fbs are detected. Grrr....
;-)

> I'd really like to get rid of all these:
>
>     s3triofb_init_of()
>     imsttfb_of_init()
>     chips_of_init()
>     matrox_of_init()
>     control_of_init()
>     valkyrie_of_init()
>     platinum_of_init()
>     clgen_of_init()
>
> These are all PCI based, so they can be found using PCI probing and marked
> busy using PCI resource management.

control is only PCI-based with some abstractions.... It can, however,
be made to look entirely like a PCI device, if you can avoid poking at
it afterwards.

I have patches for control anyway (adding it to the PCI device chain,
precisely); so I might try and implement the changes for control. Or
are you already working on this, Dan?

Cheers


Michel

-------------------------------------------------------------------------
Michel Lanners                 |  " Read Philosophy.  Study Art.
23, Rue Paul Henkes            |    Ask Questions.  Make Mistakes.
L-1710 Luxembourg              |
email   mlan@cpu.lu            |
http://www.cpu.lu/~mlan        |                     Learn Always. "


** Sent via the linuxppc-dev mail list. See http://lists.linuxppc.org/

Re: aty128 and offb and linux-2.3.99pre6

[Geert Uytterhoeven]

On Mon, 24 Apr 2000, Michel Lanners wrote:
> On  19 Apr, this message from Geert Uytterhoeven echoed through cyberspace:
> > What about other frame buffer devices, like e.g. matroxfb? Does it still work
> > after this patch? If yes, we can get rid of matrox_of_init() as well.
>
> It does work... but it changes the order my fbs are detected. Grrr....
> ;-)

But now you can throw out offb, right? :-) (unless you have controlfb, ...)

Gr{oetje,eeting}s,

						Geert

--
Geert Uytterhoeven -- Linux/{m68k~Amiga,PPC~CHRP} -- geert@linux-m68k.org

In personal conversations with technical people, I call myself a hacker. But
when I'm talking to journalists I just say "programmer" or something like that.
							    -- Linus Torvalds


** Sent via the linuxppc-dev mail list. See http://lists.linuxppc.org/

Re: aty128 and offb and linux-2.3.99pre6

[Chas Williams]

In message <200004240651.IAA00313@piglet.grunz.lu>,Michel Lanners writes:
>It does work... but it changes the order my fbs are detected. Grrr....

i have to agree.  this is fairly annoying since i have 2 rage pro's
(one on the agp and one on the pci bus) and the console should be on
the agp!

** Sent via the linuxppc-dev mail list. See http://lists.linuxppc.org/

Re: aty128 and offb and linux-2.3.99pre6

[Geert Uytterhoeven]

On Mon, 24 Apr 2000, Chas Williams wrote:
> In message <200004240651.IAA00313@piglet.grunz.lu>,Michel Lanners writes:
> >It does work... but it changes the order my fbs are detected. Grrr....
>
> i have to agree.  this is fairly annoying since i have 2 rage pro's
> (one on the agp and one on the pci bus) and the console should be on
> the agp!

Huh? This has nothing to do with the order of 2 RAGE PROs. Atyfb is probed for
before offb (like it should be) since a long time, and the order is always the
same.

Try `video=map:1' to map all your virtual consoles to the second card.


Gr{oetje,eeting}s,

						Geert

--
Geert Uytterhoeven -- Linux/{m68k~Amiga,PPC~CHRP} -- geert@linux-m68k.org

In personal conversations with technical people, I call myself a hacker. But
when I'm talking to journalists I just say "programmer" or something like that.
							    -- Linus Torvalds


** Sent via the linuxppc-dev mail list. See http://lists.linuxppc.org/

Re: aty128 and offb and linux-2.3.99pre6

[Chas Williams]

In message <Pine.LNX.4.10.10004242008100.316-100000@cassiopeia.home>,Geert Uytt
erhoeven writes:
>Huh? This has nothing to do with the order of 2 RAGE PROs. Atyfb is probed for
>before offb (like it should be) since a long time, and the order is always the
>same.

with a 2.2 kernel, the console appears on the agp card.
with a 2.3 kernel, the console appears on the pci card. [yes, it tooks
me a while i figure this out since the other card was not connected to
a monitor]


** Sent via the linuxppc-dev mail list. See http://lists.linuxppc.org/

Re: aty128 and offb and linux-2.3.99pre6

[Michel Lanners]

On  24 Apr, this message from Geert Uytterhoeven echoed through cyberspace:
> On Mon, 24 Apr 2000, Chas Williams wrote:
>> In message <200004240651.IAA00313@piglet.grunz.lu>,Michel Lanners writes:
>> >It does work... but it changes the order my fbs are detected. Grrr....
>>
>> i have to agree.  this is fairly annoying since i have 2 rage pro's
>> (one on the agp and one on the pci bus) and the console should be on
>> the agp!
>
> Huh? This has nothing to do with the order of 2 RAGE PROs. Atyfb is probed for
> before offb (like it should be) since a long time, and the order is always the
> same.
>
> Try `video=map:1' to map all your virtual consoles to the second card.

Geert, is there any documentation available about these things? Like
how you go about 'advanced' framebuffer use (multiple fb's, switching
consoles between fb's, ...)? Where are the userland tools, and what do
they do (fbset and others)?

I do have some things figured out, but am still not really comfortable
with framebuffers...

Thanks

Michel

-------------------------------------------------------------------------
Michel Lanners                 |  " Read Philosophy.  Study Art.
23, Rue Paul Henkes            |    Ask Questions.  Make Mistakes.
L-1710 Luxembourg              |
email   mlan@cpu.lu            |
http://www.cpu.lu/~mlan        |                     Learn Always. "


** Sent via the linuxppc-dev mail list. See http://lists.linuxppc.org/

Re: aty128 and offb and linux-2.3.99pre6

[Geert Uytterhoeven]

On Mon, 24 Apr 2000, Michel Lanners wrote:
> On  24 Apr, this message from Geert Uytterhoeven echoed through cyberspace:
> > On Mon, 24 Apr 2000, Chas Williams wrote:
> >> In message <200004240651.IAA00313@piglet.grunz.lu>,Michel Lanners writes:
> >> >It does work... but it changes the order my fbs are detected. Grrr....
> >>
> >> i have to agree.  this is fairly annoying since i have 2 rage pro's
> >> (one on the agp and one on the pci bus) and the console should be on
> >> the agp!
> >
> > Huh? This has nothing to do with the order of 2 RAGE PROs. Atyfb is probed for
> > before offb (like it should be) since a long time, and the order is always the
> > same.
> >
> > Try `video=map:1' to map all your virtual consoles to the second card.
>
> Geert, is there any documentation available about these things? Like
> how you go about 'advanced' framebuffer use (multiple fb's, switching
> consoles between fb's, ...)? Where are the userland tools, and what do
> they do (fbset and others)?
>
> I do have some things figured out, but am still not really comfortable
> with framebuffers...

Unfortunately there's still no real documentation, besides the few things in
linux/Documentation/fb/.

Kernel options: read drivers/video/fbmem.c

Tools: fbset, con2fbmap.

Gr{oetje,eeting}s,

						Geert

--
Geert Uytterhoeven -- Linux/{m68k~Amiga,PPC~CHRP} -- geert@linux-m68k.org

In personal conversations with technical people, I call myself a hacker. But
when I'm talking to journalists I just say "programmer" or something like that.
							    -- Linus Torvalds


** Sent via the linuxppc-dev mail list. See http://lists.linuxppc.org/

Re: aty128 and offb and linux-2.3.99pre6

[Daniel Jacobowitz]

On Mon, Apr 24, 2000 at 08:51:12AM +0200, Michel Lanners wrote:
> control is only PCI-based with some abstractions.... It can, however,
> be made to look entirely like a PCI device, if you can avoid poking at
> it afterwards.
>
> I have patches for control anyway (adding it to the PCI device chain,
> precisely); so I might try and implement the changes for control. Or
> are you already working on this, Dan?

I've been too busy with Debian to touch controlfb lately; patches would
be appreciated.

Also, what is the current state of your memory testing?  Or was that
not you?

Dan

/--------------------------------\  /--------------------------------\
|       Daniel Jacobowitz        |__|        SCS Class of 2002       |
|   Debian GNU/Linux Developer    __    Carnegie Mellon University   |
|         dan@debian.org         |  |       dmj+@andrew.cmu.edu      |
\--------------------------------/  \--------------------------------/

** Sent via the linuxppc-dev mail list. See http://lists.linuxppc.org/

Re: aty128 and offb and linux-2.3.99pre6

[Michel Lanners]

On  24 Apr, this message from Daniel Jacobowitz echoed through cyberspace:
> On Mon, Apr 24, 2000 at 08:51:12AM +0200, Michel Lanners wrote:
>> control is only PCI-based with some abstractions.... It can, however,
>> be made to look entirely like a PCI device, if you can avoid poking at
>> it afterwards.
>>
>> I have patches for control anyway (adding it to the PCI device chain,
>> precisely); so I might try and implement the changes for control. Or
>> are you already working on this, Dan?
>
> I've been too busy with Debian to touch controlfb lately; patches would
> be appreciated.

I'll see what I can do ;-)

> Also, what is the current state of your memory testing?  Or was that
> not you?

Yes, that was me. I got some reports, but I couldn't conclude anything
out of them. I still haven't found a way to identify the VRAM config in
all cases... We might need to do two different detections, based on the
board type: 7x00 (TNT) or 8x00 (Nitro); but I have no idea how to
differentiate them except by looking for additional device(s) in the
device tree (Sixty6, video out?).

Or get some info out of Apple, either directly or reverse-engineered....

Michel

-------------------------------------------------------------------------
Michel Lanners                 |  " Read Philosophy.  Study Art.
23, Rue Paul Henkes            |    Ask Questions.  Make Mistakes.
L-1710 Luxembourg              |
email   mlan@cpu.lu            |
http://www.cpu.lu/~mlan        |                     Learn Always. "


** Sent via the linuxppc-dev mail list. See http://lists.linuxppc.org/