linux-fbdev-devel — linux-fbdev mailing list for project developers

Re: [Linux-fbdev-devel] S3Virge fb on PPC?

[Geert U. <ge...@li...>]

On Wed, 7 Aug 2002, Meelis Roos wrote:
> I have a mac (powermac 7300/166 upgraded to G3/280) that has broken
> onboard video (probaly a half-working address bit in its video RAM). The
> computer also has a S3Virge PCI card with fcode so it can boot MacOS
> with it and see the picture.

  [...]

> Also, we seem to have a s3virge framebuffer for Amigas, how difficult
> would it be to port it to PPC+x86 (I have a virge DX on a PC too where I
> would like to see accelerated fb for faster scrolling). I would like to
> do it if I can.

Up to 2.4.18 virgefb relied on the AmigaOS driver to have initialized the card.
Since your ViRGE card has a F-code ROM, adopting virgefb to support it cannot
be very difficult.

In 2.4.19 (and 2.5.x) virgefb can do the initialization itself (that is, on
Amiga CyberVision64/3D cards). What this means for PC cards is unknown, since
they may have been `screwed up' by the VGA BIOS code. For your F-code card it's
probably harmless since OF uses a linear graphics mode.

Conclusion: I'd say it's not that much work. Good luck and enjoy ;-)

Gr{oetje,eeting}s,

						Geert

--
Geert Uytterhoeven -- There's lots of Linux beyond ia32 -- ge...@li...

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

Radeon support (was Re: [Linux-fbdev-devel] (no subject))

[Adam K K. <ad...@vo...>]

Me again,

=09So I think I might have a better idea of what's going on... I did
some testing with the Radeon framebuffer tonight, and can only seem to get
it to work with just two modes: 1024x768@70 and 640x480@60.  I've tried a
number of others, which worked with the DVI output of my 7500, but don't
seem to work with the VGA port of the card.

=09Any ideas?

Adam

On Tue, 6 Aug 2002, Adam K Kirchhoff wrote:

>
> Ani and others :-)
>
> =09I was hoping you could lend me some insight :-)  I'm trying to do
> some work with DirectFB, and I'm now running into a slight problem
> concerning getting output on the VGA port.
>
> =09I've e-mailed the directfb-dev mailing list about it and had some
> interesting conversations, but the end result was basically: "You should
> check with Ani Joshi about that, and maybe ask on the linux-fbdev-devel
> mailing list".
>
> =09So now I'm asking here :-)  Below this I'm including the
> description of the problem I'm seeing that I sent to the directfb-dev
> mailing list.
>
> =09Can anyone help me out?  Thanks :-)
>
> Adam
>
> ---------- Forwarded message ----------
> Date: Tue, 6 Aug 2002 12:51:24 -0400 (EDT)
> From: Adam K Kirchhoff <ad...@vo...>
> To: "[iso-8859-1] Ville Syrj=E4l=E4" <sy...@sc...>
> Cc: directfb-devel <dir...@di...>
> Subject: Re: [directfb-dev] Re: Radeon support...
>
>
> > I'm getting really confused as to what's going on... What exactly happe=
ns
> > when you run a DirectFB application?
>
> It took me a while to figure it out what I know so far, and I'm still not
> 100% sure what's going on...
>
> If I run *any* DirectFB application, the screen blanks.  The system
> doesn't lock.  I can ssh in and see the application listed in the process
> table.  For example, running df_fire, the screen blanks and, after a few
> seconds, the monitor goes to sleep (as if it's no longer getting a
> signal)...  Hitting the 'esc' key brings back the vt, just as usual when
> runnin the application.
>
> Running dfbsee produces similar results.  However, with any video that I
> play, I can hear the audio stream :-)  I just can't see anything...
>
> Now, when I first experienced this, I posted about it on either
> directfb-dev or directfb-user.  A short while later, I figured out a
> workaround :-)
>
> Instead of booting with my CRT plugged into the VGA port, I used a
> DVI-to-VGA adapter (that came with my Radeon 7500), and plugged the
> monitor into the DVI-A port.  Lo' and behold, DirectFB applications were
> now showing up :-)
>
> Since then I've done some video card swapping and now have a "Powered By
> ATI" Radeon 8500 LE.  This card has both a VGA port and a DVI-D port, but
> no adapter.  The old adapter doesn't work for this card because it's
> designed for DVI-A as compared to DVI-D.
>
> I've encountered similar problems elsewhere, too.  For example, a hacked
> radeon 7500 driver for BeOS would only display on the DVI port, even if
> the computer booted up fine with a monitor on the VGA port.
>
> Of course, I just don't know why this is the case :-)
>
> I've been poking around on-line looking for a DVI-D to VGA adapter, but
> very few of the shops that sell them on-line clarify if it's DVI-A or
> DVI-D :-)
>
> Adam
>
>
>
>
>
> -------------------------------------------------------
> This sf.net email is sponsored by:ThinkGeek
> Welcome to geek heaven.
> http://thinkgeek.com/sf
> _______________________________________________
> Linux-fbdev-devel mailing list
> Lin...@li...
> https://lists.sourceforge.net/lists/listinfo/linux-fbdev-devel
>
>

[Linux-fbdev-devel] FYI: [BK PATCH] matroxfb update: G450/G550 DVI and TV support

[Petr V. <van...@vc...>]

Hello everybody,
  I just sent changeset below to the Linus. If you are using
my patches for TV-Out from ftp://platan.vc.cvut.cz/pub/linux/matrox-latest,
please note that V4L2 support for setting brightness/contrast and other
features is NOT part of this patch. I'll send it to Linus after V4L2
interface finds its way into the 2.5 kernel.

  And also, as you may guess, driver is not ported to the new fbdev API.

  Patch and bkpatch were removed from this copy of message - it is
129KB file. You can download it from 
ftp://platan.vc.cvut.cz/pub/linux/matrox-latest/bkpatch-for-linus-2.5.30,
or you can look at http://matroxfb.bkbits.net, or you can pull from
bk://matroxfb.bkbits.net/linux-2.5 or ... It applies cleanly to the currently
available 2.5 tree.
						Best regards,
							Petr Vandrovec
							van...@vc...

----- Forwarded message from Petr Vandrovec <van...@vc...> -----

Date: Thu, 8 Aug 2002 00:16:52 +0200
From: Petr Vandrovec <van...@vc...>
To: tor...@tr...
Subject: [BK PATCH] matroxfb update: G450/G550 DVI and TV support

Linus,
  please apply changesets below. It adds DVI and TVOut functionality to
the matroxfb. Without proper DVI support matroxfb does not work with
currently sold G550.

You can import this changeset into BK by piping this whole message to:
'| bk receive [path to repository]' or apply the patch as usual.
You can also do

     bk pull bk://matroxfb.bkbits.net/linux-2.5

						Thanks,
							Petr Vandrovec
							van...@vc...

===================================================================


ChangeSet@1.486, 2002-08-07 23:08:40+02:00, van...@vc...
  Set system PLL vcomax correctly in matroxfb. Discovered by Dirk Uffmann.

ChangeSet@1.485, 2002-08-07 22:56:59+02:00, van...@vc...
  matroxfb: Do not store results of bitwise AND directly in variables which are 
  treated as a booleans. Comparsion does not work correctly on them. 

ChangeSet@1.484, 2002-08-07 22:48:09+02:00, van...@vc...
  Add support for MGA-TVO-B into matroxfb. By Mike Pieper.

ChangeSet@1.483, 2002-08-07 22:17:53+02:00, van...@vc...
  Return ENOTTY instead of EINVAL for unknown ioctl ops in matroxfb.

ChangeSet@1.482, 2002-08-07 22:13:00+02:00, van...@vc...
  Use sizeof(*var) instead of sizeof(struct xxx) in matroxfb.

ChangeSet@1.481, 2002-08-07 22:07:39+02:00, van...@vc...
  Remove currcon field from private fb_info in matroxfb. It was moved to the generic
  layer long ago.

ChangeSet@1.480, 2002-08-07 21:55:40+02:00, van...@vc...
  Make debug printouts in matroxfb G400 TV-out disabled by default.
  OUTPUT_MODE are values, not a bitmap, so use compare instead of bitwise AND.

ChangeSet@1.479, 2002-08-07 21:46:46+02:00, van...@vc...
  Add TV-Out support for Matrox G450/G550. Due to the hardware only secondary CRTC
  can be used as a source for TV output.

ChangeSet@1.478, 2002-08-07 20:34:38+02:00, van...@vc...
  Handle DVI output on G450/G550.
  Powerdown unused portions of G450/G550 DAC.
  Split G450/G550 DAC from older DAC1064 handling.
  Modify PLL setting when both CRTCs use same pixel clocks.

ChangeSet@1.477, 2002-08-07 19:49:54+02:00, van...@vc...
  Initialize Matrox G100 and G400 hardware with values read from BIOS instead of
  with failsafe settings discovered in the past.
  Fixes corrupted screen display on some G100.

ChangeSet@1.476, 2002-08-07 19:31:42+02:00, van...@vc...
  Use container_of instead of simple typecast when we convert pointers from pointer
  to fb_info to pointers to matrox_fb_info.

ChangeSet@1.475, 2002-08-07 19:15:08+02:00, van...@vc...
  Store pointer to matroxfb specific fb information instead of universal fb_info*
  pointer for secondary head. Saves some typecasts.

ChangeSet@1.474, 2002-08-07 19:06:50+02:00, van...@vc...
  Use arrays for holding Matrox output drivers, it is nicer and more extensible
  than current solution with per-CRTC bitmaps.

ChangeSet@1.473, 2002-08-07 01:24:20+02:00, van...@vc...
  Simplify rules for writting secondary output drivers to matroxfb.
  Update some initializations to use C99 initializers.

ChangeSet@1.472, 2002-08-07 00:33:35+02:00, van...@vc...
  Verify validity of color depth passed to matroxfb by looking through table instead of using ifelseif... code.

ChangeSet@1.471, 2002-08-07 00:21:32+02:00, van...@vc...
  Merge secondary output state structure to main in matroxfb driver.

ChangeSet@1.470, 2002-08-06 23:58:36+02:00, van...@vc...
  Make secondary output support mandatory for Matrox G450/G550.

ChangeSet@1.469, 2002-08-06 22:52:03+02:00, van...@vc...
  Support secondary head DDC on G450/G550.
  Simplify i2c-matroxfb code.


 Config.help               |   58 ++---
 Config.in                 |    9 
 matrox/Makefile           |    6 
 matrox/g450_pll.c         |   85 ++++----
 matrox/g450_pll.h         |    2 
 matrox/matroxfb_DAC1064.c |  302 ++++++++++++++++++++---------
 matrox/matroxfb_DAC1064.h |    4 
 matrox/matroxfb_Ti3026.c  |   14 +
 matrox/matroxfb_base.c    |  474 ++++++++++++++++++++++++----------------------
 matrox/matroxfb_base.h    |   51 +++-
 matrox/matroxfb_crtc2.c   |  279 +++++++++++++++++----------
 matrox/matroxfb_crtc2.h   |    3 
 matrox/matroxfb_g450.c    |  467 +++++++++++++++++++++++++++++++++++----------
 matrox/matroxfb_g450.h    |   14 -
 matrox/matroxfb_maven.c   |  123 +++++------
 matrox/matroxfb_misc.c    |    6 
 16 files changed, 1208 insertions(+), 689 deletions(-)

[Linux-fbdev-devel] Re: Matrox Oopsed when using kernel with ACPI support...

[Dawid K. <qn...@at...>]

On Wed, 7 Aug 2002, Petr Vandrovec wrote:

> > Code;  c024d739 <matrox_cfbX_putcs+189/280>   <=3D=3D=3D=3D=3D
> >    0:   f3 a5                     repz movsl %ds:(%esi),%es:(%edi) =
  <=3D=3D=3D=3D=3D
>=20
> Uh-oh. Impossible. Do you use multihead, or do all your ttyS belong
> to /dev/fb0? If you are using /dev/fb1 (but you should do it anyway..=
=2E)
> please apply patch from
> ftp://platan.vc.cvut.cz/pub/linux/matrox-latest/fbset-through-vt-2.4.=
19-rc5.gz
> to your kernel. Besides that it fixes (oopses|kernel corruption) if y=
ou do=20
> 'fbset -fb /dev/fb0' while you are watching console on /dev/fb1, it a=
lso
> allows you to do 'fbset -fb /dev/tty*' - this could simplify your

Actually I have only one fbdev:
[root@atlantis ~]# cat /proc/fb
0 MATROX VGA

=2E..the trick with open/fbdev pair might

> script a lot. And no, patch is not going to the kernel. James vetoed =
it
> years ago...
Oh [ anyways the patch is great! and I runned the "ugly old script" and
it works fine too :) ]

> And while you are on it, please try
> ftp://platan.vc.cvut.cz/pub/linux/matrox-latest/mga-2.4.19-rc5-tvout.=
gz
> too. But I believe that fbset-through-vt patch will fix your oopses, =
and
> this mga patch will not bring any new feature to your old MGA.

I applied it and appears to be rock solid.  If I experience any problem=
s,
I'll write.  The card has no TV out, so only new feature are new things
in /proc:

[root@atlantis ~]# cat /proc/driver/mga/fb0/bios                       =
                      =20
BIOS:   0.0.0
Output: 0x00
TVOut:  no
PINS:   found
Info:   c03f2a0c
[root@atlantis ~]# hexdump -c /proc/driver/mga/fb0/pins
0000000   .   A   @   =FF  \0 002   .   =C4 003  \0  \0 021   A   B   C=
   1
0000010   7   3   5   1  \0 213   X  \0   4 213   X  \0   2   0   7  \0
0000020  \0  \0   =A2   &   =FF   =FF   =FF   =FF  \0   F   F   <   =FF=
   <   2   7
0000030   =FF   =FF   =FF   =FF   =EE   =FF   =FF   =FF   =FF   =FF   =FF=
   =FF   =FF   =FF   =FF   =E0
0000040

And once again lots of thanks for quick response and solving my
problem (and for fbset-through-vt which is defenatley going to stay
in my copy of kernel :)).

   Regards,
      Dawid

[Linux-fbdev-devel] controlfb bug in 2.4

[Meelis R. <mr...@li...>]

If this not the right place to ask, please redirect me to another list.

I use controlfb on a powermac. In 2.2.20 it works OK, modulo some
harware problem (half-broken bit in some video ram wire resulting in
vertical lines or color chanes depending on color depth). In 2.4 (tried
up to 2.4.19) there is a problem. Sometimes the display shifts right
half a screen after scrolling operations. The contents are of course
wrapped from left but it is more that 1 line so the last text line is
mostly shifted out of screen. The total vertical shift is about 1.5 text
lines in default mode (640x480). I first thought that it might be
because of the hardware problem but it works well with 2.2 kernels.

controlfb: VRAM Total = 2MB (0MB @ bank 1, 2MB @ bank 2)
controlfb: using video mode 6 and color mode 0.
Console: switching to colour frame buffer device 80x30
fb0: control display adapter

-- 
Meelis Roos (mr...@li...)

[Linux-fbdev-devel] S3Virge fb on PPC?

[Meelis R. <mr...@li...>]

I have a mac (powermac 7300/166 upgraded to G3/280) that has broken
onboard video (probaly a half-working address bit in its video RAM). The
computer also has a S3Virge PCI card with fcode so it can boot MacOS
with it and see the picture.

Now I want to have linux framebuffer & run accelerated X on this S3
Virge. There is some generic framebuffer (offb?) that somewhat works (no
acceleration and can't get X to work). XFree 4.1 with s3virge driver
also works somewhat but has strange bad effects: mouse cursor is bad in
HW mode, lower 1/3 of screen is full of gibberish, strange colors
sometimes, ... I have had similar problems on PC where matrox X driver
and matroxfb didn't play well, maybe something similar is happening
here.

Is there any hope that native s3virge framebuffer would work better
(besides being accelerated)?

Also, we seem to have a s3virge framebuffer for Amigas, how difficult
would it be to port it to PPC+x86 (I have a virge DX on a PC too where I
would like to see accelerated fb for faster scrolling). I would like to
do it if I can.

-- 
Meelis Roos (mr...@li...)

Re: [Linux-fbdev-devel] 2.5 atyfb on Sparc question

[Petr V. <VAN...@vc...>]

On  7 Aug 02 at 10:43, Jani Monoses wrote:
> On Tue, 6 Aug 2002 09:06:48 -0700 (PDT)
> James Simmons <cap...@ya...> wrote:
> > I push stuff. The next set of patches will break alot
> > of drivers but I need to do this to get people to port
> > there stuff over to the new api.
> 
> Could you outline what the next set of patches will change?
> Thanks

I also hope that performance problems will be solved before we are
forced to not use putcs.
                                                Petr Vandrovec
                                                van...@vc...

[Linux-fbdev-devel] Re: Matrox Oopsed when using kernel with ACPI support...

[Petr V. <VAN...@vc...>]

On  7 Aug 02 at 10:57, Dawid Kuroczko wrote:
> No modules in ksyms, skipping objects
> Warning (read_lsmod): no symbols in lsmod, is /proc/modules a valid lsmod file?
> Aug  7 02:08:34 atlantis kernel: Unable to handle kernel NULL pointer dereference at virtual address 00000200
> Aug  7 02:08:34 atlantis kernel: c024d739
> Aug  7 02:08:34 atlantis kernel: *pde = 00000000
> >>EIP; c024d739 <matrox_cfbX_putcs+189/280>   <=====
> Trace; c024d88d <matrox_cfb8_putcs+5d/70>
> Trace; c0246f44 <fbcon_putcs+94/100>
> Trace; c0205257 <do_update_region+127/190>

> Code;  c024d739 <matrox_cfbX_putcs+189/280>   <=====
>    0:   f3 a5                     repz movsl %ds:(%esi),%es:(%edi)   <=====

Uh-oh. Impossible. Do you use multihead, or do all your ttyS belong
to /dev/fb0? If you are using /dev/fb1 (but you should do it anyway...)
please apply patch from
ftp://platan.vc.cvut.cz/pub/linux/matrox-latest/fbset-through-vt-2.4.19-rc5.gz
to your kernel. Besides that it fixes (oopses|kernel corruption) if you do 
'fbset -fb /dev/fb0' while you are watching console on /dev/fb1, it also
allows you to do 'fbset -fb /dev/tty*' - this could simplify your
script a lot. And no, patch is not going to the kernel. James vetoed it
years ago...

And while you are on it, please try
ftp://platan.vc.cvut.cz/pub/linux/matrox-latest/mga-2.4.19-rc5-tvout.gz
too. But I believe that fbset-through-vt patch will fix your oopses, and
this mga patch will not bring any new feature to your old MGA.
                                                Petr Vandrovec
                                                van...@vc...
                                                

[Linux-fbdev-devel] Matrox Oopsed when using kernel with ACPI support...

[Dawid K. <qn...@at...>]

I'm running kernel 2.4.19 and I'm exeriencing repeatable kernel Oops
with Matrox Mystique card.  Kernel appears OK when compiled with
APM, but oopses whem compiled with ACPI...  More below oops dump.

ksymoops 2.4.4 on i686 2.4.19.  Options used
     -V (default)
     -k /proc/ksyms (default)
     -l /proc/modules (default)
     -o /lib/modules/2.4.19/ (default)
     -m /boot/System.map-2.4.19 (default)

Warning: You did not tell me where to find symbol information.  I will
assume that the log matches the kernel and modules that are running
right now and I'll use the default options above for symbol resolution.
If the current kernel and/or modules do not match the log, you can get
more accurate output by telling me the kernel version and where to find
map, modules, ksyms etc.  ksymoops -h explains the options.

No modules in ksyms, skipping objects
Warning (read_lsmod): no symbols in lsmod, is /proc/modules a valid lsm=
od file?
Aug  7 02:08:34 atlantis kernel: Unable to handle kernel NULL pointer d=
ereference at virtual address 00000200
Aug  7 02:08:34 atlantis kernel: c024d739
Aug  7 02:08:34 atlantis kernel: *pde =3D 00000000
Aug  7 02:08:34 atlantis kernel: Oops: 0000
Aug  7 02:08:34 atlantis kernel: CPU:    1
Aug  7 02:08:34 atlantis kernel: EIP:    0010:[matrox_cfbX_putcs+393/64=
0]    Not tainted
Aug  7 02:08:34 atlantis kernel: EIP:    0010:[<c024d739>]    Not taint=
ed
Using defaults from ksymoops -t elf32-i386 -a i386
Aug  7 02:08:34 atlantis kernel: EFLAGS: 00010202
Aug  7 02:08:34 atlantis kernel: eax: 00000010   ebx: f0809000   ecx: 0=
0000004   edx: 01200010
Aug  7 02:08:34 atlantis kernel: esi: 00000200   edi: f0809000   ebp: 0=
0000010   esp: c1857e60
Aug  7 02:08:34 atlantis kernel: ds: 0018   es: 0018   ss: 0018
Aug  7 02:08:34 atlantis kernel: Process keventd (pid: 2, stackpage=3Dc=
1857000)
Aug  7 02:08:34 atlantis kernel: Stack: 00000001 011f0118 0000007f 0000=
0010 01200010 00000001 07070707 c03ec65c=20
Aug  7 02:08:34 atlantis kernel:        e607ae10 00000012 c024d88d 0707=
0707 00000000 c03ec65c e607ae58 00000040=20
Aug  7 02:08:34 atlantis kernel:        00000012 00000000 c03f0970 c03e=
c65c 00000000 c0246f44 efde7600 c03ec65c=20
Aug  7 02:08:34 atlantis kernel: Call Trace:    [matrox_cfb8_putcs+93/1=
12] [fbcon_putcs+148/256] [do_update_region+295/400] [redraw_screen+244=
/336] [complete_change_console+43/192]
Aug  7 02:08:34 atlantis kernel: Call Trace:    [<c024d88d>] [<c0246f44=
>] [<c0205257>] [<c0205d74>] [<c020378b>]
Aug  7 02:08:34 atlantis kernel:   [<c020907a>] [<c011f8cd>] [<c01286aa=
>] [<c01284a0>] [<c0105000>] [<c01073c6>]
Aug  7 02:08:34 atlantis kernel:   [<c01284a0>]
Aug  7 02:08:34 atlantis kernel: Code: f3 a5 a8 02 74 02 66 a5 a8 01 74=
 01 a4 8b 74 24 34 0f b7 96=20

>>EIP; c024d739 <matrox_cfbX_putcs+189/280>   <=3D=3D=3D=3D=3D
Trace; c024d88d <matrox_cfb8_putcs+5d/70>
Trace; c0246f44 <fbcon_putcs+94/100>
Trace; c0205257 <do_update_region+127/190>
Trace; c0205d74 <redraw_screen+f4/150>
Trace; c020378b <complete_change_console+2b/c0>
Trace; c020907a <console_callback+3a/b0>
Trace; c011f8cd <__run_task_queue+5d/70>
Trace; c01286aa <context_thread+20a/220>
Trace; c01284a0 <context_thread+0/220>
Trace; c0105000 <_stext+0/0>
Trace; c01073c6 <kernel_thread+26/30>
Trace; c01284a0 <context_thread+0/220>
Code;  c024d739 <matrox_cfbX_putcs+189/280>
00000000 <_EIP>:
Code;  c024d739 <matrox_cfbX_putcs+189/280>   <=3D=3D=3D=3D=3D
   0:   f3 a5                     repz movsl %ds:(%esi),%es:(%edi)   <=3D=
=3D=3D=3D=3D
Code;  c024d73b <matrox_cfbX_putcs+18b/280>
   2:   a8 02                     test   $0x2,%al
Code;  c024d73d <matrox_cfbX_putcs+18d/280>
   4:   74 02                     je     8 <_EIP+0x8> c024d741 <matrox_=
cfbX_putcs+191/280>
Code;  c024d73f <matrox_cfbX_putcs+18f/280>
   6:   66 a5                     movsw  %ds:(%esi),%es:(%edi)
Code;  c024d741 <matrox_cfbX_putcs+191/280>
   8:   a8 01                     test   $0x1,%al
Code;  c024d743 <matrox_cfbX_putcs+193/280>
   a:   74 01                     je     d <_EIP+0xd> c024d746 <matrox_=
cfbX_putcs+196/280>
Code;  c024d745 <matrox_cfbX_putcs+195/280>
   c:   a4                        movsb  %ds:(%esi),%es:(%edi)
Code;  c024d746 <matrox_cfbX_putcs+196/280>
   d:   8b 74 24 34               mov    0x34(%esp,1),%esi
Code;  c024d74a <matrox_cfbX_putcs+19a/280>
  11:   0f b7 96 00 00 00 00      movzwl 0x0(%esi),%edx

OK, and now for more information.  The system is SMP 2 x PIII 733
machine with:
   matroxfb: Matrox Mystique (PCI) detected
   matroxfb: MTRR's turned on
   matroxfb: 640x480x8bpp (virtual: 640x3270)
   matroxfb: framebuffer at 0xD6000000, mapped to 0xc8805000, size 2097=
152
Everything works just fine, there is an fbdev X server running on one o=
f
the consoles.  The oops results when following script is run (please no=
te
that it works fine when ACPI is not compiled in):

--- >8 --- CUT here --- /etc/rc.d/init.d/fbset
#!/bin/sh
#
# hdpram        Sets up fbcon video modes.
#
#
# chkconfig:	2345 5 30
# description:	fbset is a utility with which fbcons video modes can be =
read\
#		and changed


# Source function library
=2E /etc/rc.d/init.d/functions

# Get service config
[ -f /etc/sysconfig/fbset ] && . /etc/sysconfig/fbset

# See how we were called.
case "$1" in
  start)
	# Check if we have framebuffer in kernel.
	if [ -f /proc/fb ]; then
		# /proc files show as files with size=3D0, this is a workaround.
		cat /proc/fb | grep -q "." || exit 0
	else
		exit 0
	fi

	if [ -n "${FBMODE_default}" ]; then
		show "Setting default video mode"
		busy
		/usr/sbin/fbset -a ${FBMODE_default}
		deltext
		ok
	fi
	for tty in /dev/tty[1-9] /dev/tty[1-6][0-9]; do
		eval FBMODE=3D\${FBMODE_${tty#/dev/}}
		if [ -n "${FBMODE}" ]; then
			echo "settking $tty to $FBMODE"
			# XXX fuser $tty <- sprawdzi=E6, czy tty jest wolna.
			show "Setting $tty video mode to $FBMODE"=20
			busy
			echo -n > $tty
			/usr/bin/open -s -c ${tty#/dev/tty} -- /usr/sbin/fbset ${FBMODE}
			deltext
			ok
		fi
	done
	touch /var/lock/subsys/fbset
	;;
    stop)=09
	rm -f /var/lock/subsys/fbset
	;;
  status)
	# Check if we have framebuffer in kernel.
	if [ -f /proc/fb ]; then
		# /proc files show as files with size=3D0, this is a workaround.
		cat /proc/fb | grep -q "." || exit 0
		echo "Frame buffer present."
	fi
	;;
  restart|reload)
	$0 stop
	$0 start
	;;
  *)
	echo "Usage: $0 {start|stop|status|restart|reload}"
	exit 1
esac

exit 0
--- >8 --- END

--- >8 --- /etc/sysconfig/fbset
# This file lets you set your default console video mode and
# video modes for individual virtual consoles.
# See /etc/fb.modes for sample video mode definitions.

=46BMODE_default=3D"Q100x31"

=46BMODE_tty1=3D
=46BMODE_tty2=3D
=46BMODE_tty3=3D
=46BMODE_tty4=3D
=46BMODE_tty5=3D
=46BMODE_tty6=3D
=46BMODE_tty7=3D
=46BMODE_tty8=3D

=46BMODE_tty12=3D"Q128x48"

=46BMODE_tty21=3D"Q128x48"
=46BMODE_tty22=3D"Q80x25"
=46BMODE_tty23=3D"Q128x48"
=46BMODE_tty24=3D"Q128x48"
--- >8 --- END

If I missed any needed information, please ask. :-)

   Regards,
     Dawid
--=20
                 http://www.muppetlabs.com/~breadbox/bf/
+++++[>++++++<-]>++[>+>++>+++>++++<<<<-]>>++++.>+.>---------.<++++++++.=
--
---.<<.>+++++++.>>--.---.---.<-.>+++++++++++.<++++++++.++++.>>+++++++++=
+.

Re: [Linux-fbdev-devel] 2.5 atyfb on Sparc question

[Jani M. <ja...@iv...>]

On Tue, 6 Aug 2002 09:06:48 -0700 (PDT)
James Simmons <cap...@ya...> wrote:
> I push stuff. The next set of patches will break alot
> of drivers but I need to do this to get people to port
> there stuff over to the new api.

Could you outline what the next set of patches will change?
Thanks
Jani.

Re: [Linux-fbdev-devel] RFC: Optimizing putcs()

[Antonino D. <ad...@po...>]

One of the reason why 2.4 console performance is good especially at low
bit depths is its ability to process more than 1 pixel per iteration and
its usage of mask arrays.  

I tried to generalize the above in cfbimgblt.c by incorporating the idea
in fbcon-cfb*.c. It's significantly faster but still not as fast as the
2.4 API. 

time cat /usr/src/linux/MAINTAINERs (40K text file) 
1024x768-8bpp, y-panning disabled 

2.5 old (with offscreen buffers) 

real    0m10.708s 
user    0m0.001s 
sys     0m10.707s 

2.5 new 

real    0m4.378s 
user    0m0.002s 
sys     0m4.375s 

2.4 

real    0m2.098s 
user    0m0.000s 
sys     0m2.070s 

I've only tested the implementation at 8, 16, 24, and 32 bpp.  24bpp is
slightly slower than 32 bpp :( 

Tony 

Re: [Linux-fbdev-devel] RFC: Optimizing putcs()

[Antonino D. <ad...@po...>]

On Wed, 2002-08-07 at 04:08, Geert Uytterhoeven wrote:

> 
> Just for reference, did you run this benchmark on 2.4.x as well?
> 
> Gr{oetje,eeting}s,
> 
Sort of. The functions in fbcon-cfb*.c are already very fast, because
fbcon and character drawing are tightly integrated together, and
fbcon_cfb8_putcs() is very, very efficient, processing 4 bits per
iteration, instead of 1.  I'm getting numbers like this:

real    0m2.098s
user    0m0.000s
sys     0m2.070s

which was faster(!) than my hardware implementation of putcs, and 5x
faster than 2.5.  Since I'm using an i810 with Video in System RAM,
direct framebuffer access does not carry much overhead.  I just have to
beat fbcon-cfb8, so I thought of placing text data in offscreen graphics
memory to take full advantage of hardware blitting.

At high bit depths (32 bpp), 2.5 with an offscreen buffer is as fast as
2.4.

Tony    

Re: [Linux-fbdev-devel] radeonfb and dvi support

[Benjamin H. <be...@ke...>]

>My desktop has a radeon 7500 with a flat panel. How can I get a patch
>from your tree?

Just pick the entire radeonfb from my rsync

(rsync.penguinppc.org::linux-2.4-benh)

Ben.

Radeon support (Was: Re: [Linux-fbdev-devel] (no subject))

[Adam K K. <ad...@vo...>]

I thought that might be a possibility, too, except that the same monitor
works with the same applications when connected to the DVI output on the
7500 with the adapter.  So apparently the application isn't switching to
an unsupported frequency.

I might try to get the app to dump the registers anyway, though.

Thanks for the response.

Oh, and sorry for sending an e-mail without a subject to the list.

Adam

On Tue, 6 Aug 2002, Christopher P Wright <softpixel.com> wrote:

> i dont develop many fb drivers, but it sounds like the video out dfb sets
> your card to is out of sync range (causing the monitor to ignore it rather
> than toast itself).  perhaps see what the clocks are being set to, and
> maybe check configuration of dfb (i havnt used that in a while now, so i
> dont know too much about it)
> 
> if possible, get some app to dump your card's registers while the screen
> is blanked.  then send this to dfb developers, amybe they will be able to
> see what is getting done improperly.
> 
> best of wishes on getting this fixed,
> chris
> 
> ---
> if you got the will
> and you got the skill
> then you take a step.
> - Master Phung
> 
> 
> 
> 
> 

Re: [Linux-fbdev-devel] radeonfb and dvi support

[Louis G. <lou...@be...>]

My desktop has a radeon 7500 with a flat panel. How can I get a patch
from your tree?

--Lou

On Tue, 2002-08-06 at 11:50, Benjamin Herrenschmidt wrote:
> >Is the radeonfb driver going to have support for the dvi port any time soon.
> >I'm talking about the 2.4.x kernels.
> 
> If you are running single head (and not the LCD on DVI on the latest
> laptops), the one in my tree should work (thanks to Paulus fixes)
> 
> Ben.
> 
> 

Re: [Linux-fbdev-devel] RFC: Optimizing putcs()

[Geert U. <ge...@li...>]

On 6 Aug 2002, Antonino Daplas wrote:
> With fbcon-accel and the new drawing functions in linux-2.5, console
> performance degraded compared to the linux-2.4 implementation.  This is
> because putcs() has to to do 1 fb_imageblit() per character to be
> drawn.  

Yes, this will be shown badly after I'll have ported amifb to the new
framework, since chip RAM accesses are very slow and we use bitplanes...

> This can be optimized by letting putcs() initially construct the row of
> text to be drawn into an offscreen buffer, then do a single
> fb_imageblit() in the end.  Performance wil increase for several
> reasons:

Yes, this is very nice! I was thinking about passing an array of images to an
fb_imageblit_multiple() or so, but yours may be better.

> For drivers that uses cfb_imageblit or similar, a code such as the one
> below can be inserted during initialization:
> 
> info->pixmap.addr = (unsigned long) kmalloc(BUFFER_SIZE, GFP_KERNEL);
> info->pixmap.size = BUFFER_SIZE;
> info->pixmap.offset = 0;
> info->pixmap.buf_align = 1;
> info->pixmap.scan_align = 1;
> 
> Some benchmarks:
> 
> time cat /usr/src/linux/MAINTAINERS (40K text file)
> mode: 1024x768@8bpp, y-panning disabled.

  [...]

Just for reference, did you run this benchmark on 2.4.x as well?

Gr{oetje,eeting}s,

						Geert

--
Geert Uytterhoeven -- There's lots of Linux beyond ia32 -- ge...@li...

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

Re: [Linux-fbdev-devel] RFC: Optimizing putcs()

[Antonino D. <ad...@po...>]

On Tue, 2002-08-06 at 13:49, James Simmons wrote:

> Ah. Thank you for solving this problem for me. I
> haven't had time to figure it out. Also left to be
> done is 24 bpp support as well as drawing the penguin.
> 
I took a crack at adding support for bpp24 for the cfb_* drawing
functions.  I tried to keep the original code as much as possible, so
the result may not be optimal.  My test shows though that bpp24 should
be as fast as (maybe a tad slower than) bpp32.

As for drawing the logo, will the source be containing indices to the
palette?

Tony

[Linux-fbdev-devel] (no subject)

[Adam K K. <ad...@vo...>]

Ani and others :-)

=09I was hoping you could lend me some insight :-)  I'm trying to do
some work with DirectFB, and I'm now running into a slight problem
concerning getting output on the VGA port.

=09I've e-mailed the directfb-dev mailing list about it and had some
interesting conversations, but the end result was basically: "You should
check with Ani Joshi about that, and maybe ask on the linux-fbdev-devel
mailing list".

=09So now I'm asking here :-)  Below this I'm including the
description of the problem I'm seeing that I sent to the directfb-dev
mailing list.

=09Can anyone help me out?  Thanks :-)

Adam

---------- Forwarded message ----------
Date: Tue, 6 Aug 2002 12:51:24 -0400 (EDT)
From: Adam K Kirchhoff <ad...@vo...>
To: "[iso-8859-1] Ville Syrj=E4l=E4" <sy...@sc...>
Cc: directfb-devel <dir...@di...>
Subject: Re: [directfb-dev] Re: Radeon support...


> I'm getting really confused as to what's going on... What exactly happens
> when you run a DirectFB application?

It took me a while to figure it out what I know so far, and I'm still not
100% sure what's going on...

If I run *any* DirectFB application, the screen blanks.  The system
doesn't lock.  I can ssh in and see the application listed in the process
table.  For example, running df_fire, the screen blanks and, after a few
seconds, the monitor goes to sleep (as if it's no longer getting a
signal)...  Hitting the 'esc' key brings back the vt, just as usual when
runnin the application.

Running dfbsee produces similar results.  However, with any video that I
play, I can hear the audio stream :-)  I just can't see anything...

Now, when I first experienced this, I posted about it on either
directfb-dev or directfb-user.  A short while later, I figured out a
workaround :-)

Instead of booting with my CRT plugged into the VGA port, I used a
DVI-to-VGA adapter (that came with my Radeon 7500), and plugged the
monitor into the DVI-A port.  Lo' and behold, DirectFB applications were
now showing up :-)

Since then I've done some video card swapping and now have a "Powered By
ATI" Radeon 8500 LE.  This card has both a VGA port and a DVI-D port, but
no adapter.  The old adapter doesn't work for this card because it's
designed for DVI-A as compared to DVI-D.

I've encountered similar problems elsewhere, too.  For example, a hacked
radeon 7500 driver for BeOS would only display on the DVI port, even if
the computer booted up fine with a monitor on the VGA port.

Of course, I just don't know why this is the case :-)

I've been poking around on-line looking for a DVI-D to VGA adapter, but
very few of the shops that sell them on-line clarify if it's DVI-A or
DVI-D :-)

Adam

Re: [Linux-fbdev-devel] 2.5 atyfb on Sparc question

[James S. <cap...@ya...>]

> Quick question..  I was wondering if there is still
> a bk tree for fbdev
> work, or are you pushing direct to Linus?
> 
> I am getting the following errors when compiling on
> a sparc64 machine from
> the latest BK 2.5 tree, and was wondering if a fix
> may already be around.  I
> am not subscribed to this list, but a quick search
> didn't turn up anything.
> Thanks in advance..

Yes. I received a fix. I added to my BK repository but
I need to work on the driver a little bit more before
I push stuff. The next set of patches will break alot
of drivers but I need to do this to get people to port
there stuff over to the new api.

__________________________________________________
Do You Yahoo!?
Yahoo! Health - Feel better, live better
http://health.yahoo.com

Re: [Linux-fbdev-devel] radeonfb and dvi support

[Benjamin H. <be...@ke...>]

>Is the radeonfb driver going to have support for the dvi port any time soon.
>I'm talking about the 2.4.x kernels.

If you are running single head (and not the LCD on DVI on the latest
laptops), the one in my tree should work (thanks to Paulus fixes)

Ben.

[Linux-fbdev-devel] 2.5 atyfb on Sparc question

[Holzrichter, B. <bru...@mo...>]

Quick question..  I was wondering if there is still a bk tree for fbdev
work, or are you pushing direct to Linus?

I am getting the following errors when compiling on a sparc64 machine from
the latest BK 2.5 tree, and was wondering if a fix may already be around.  I
am not subscribed to this list, but a quick search didn't turn up anything.
Thanks in advance..

Thanks,
Bruce H. 

atyfb_base.c: In function `atyfb_set_dispsw':
atyfb_base.c:1098: warning: assignment discards `const' from pointer target
type
atyfb_base.c:1101: warning: assignment discards `const' from pointer target
type
atyfb_base.c:1105: warning: assignment discards `const' from pointer target
type
atyfb_base.c:1109: warning: assignment discards `const' from pointer target
type
atyfb_base.c: In function `atyfb_set_var':
atyfb_base.c:1191: warning: implicit declaration of function
`do_install_cmap'
atyfb_base.c: In function `atyfb_ioctl':
atyfb_base.c:1258: `info2' undeclared (first use in this function)
atyfb_base.c:1258: (Each undeclared identifier is reported only once
atyfb_base.c:1258: for each function it appears in.)
atyfb_base.c:1253: warning: `disp' might be used uninitialized in this
function
atyfb_base.c: In function `atyfb_save_palette':
atyfb_base.c:1447: warning: passing arg 2 of `aty_ld_8' from incompatible
pointer type
atyfb_base.c:1448: `par' undeclared (first use in this function)
atyfb_base.c:1450: warning: passing arg 3 of `aty_st_8' from incompatible
pointer type
atyfb_base.c:1451: warning: passing arg 3 of `aty_st_8' from incompatible
pointer type
atyfb_base.c: In function `atyfb_init':
atyfb_base.c:2204: parse error before `struct'
atyfb_base.c:2231: `default_par' undeclared (first use in this function)
atyfb_base.c:2261: `par' undeclared (first use in this function)
atyfb_base.c:2268: warning: assignment makes pointer from integer without a
cast
atyfb_base.c:2364: warning: passing arg 2 of `aty_ld_le32' from incompatible
pointer type
atyfb_base.c:2366: warning: passing arg 2 of `aty_ld_le32' from incompatible
pointer type
atyfb_base.c:2388: warning: passing arg 2 of `aty_ld_le32' from incompatible
pointer type
atyfb_base.c:2392: warning: passing arg 3 of `aty_st_le32' from incompatible
pointer type
atyfb_base.c:2434: warning: passing arg 2 of `aty_ld_le32' from incompatible
pointer type
atyfb_base.c:2437: warning: passing arg 2 of `aty_ld_le32' from incompatible
pointer type
atyfb_base.c:2439: warning: passing arg 2 of `aty_ld_le32' from incompatible
pointer type
atyfb_base.c:2442: warning: passing arg 2 of `aty_ld_le32' from incompatible
pointer type
atyfb_base.c:2444: warning: passing arg 2 of `aty_ld_le32' from incompatible
pointer type
atyfb_base.c:2455: warning: passing arg 2 of `aty_ld_8' from incompatible
pointer type
atyfb_base.c:2458: warning: passing arg 2 of `aty_ld_pll' from incompatible
pointer type
atyfb_base.c:2560: invalid operands to binary &
atyfb_base.c: At top level:
atyfb_base.c:136: warning: `atyfb_fix' defined but not used
make[3]: *** [atyfb_base.o] Error 1
make[3]: Leaving directory `/home/bruce/sparctest/drivers/video/aty'
make[2]: *** [aty] Error 2
make[2]: Leaving directory `/home/bruce/sparctest/drivers/video'
make[1]: *** [video] Error 2
make[1]: Leaving directory `/home/bruce/sparctest/drivers'
make: *** [drivers] Error 2

[Linux-fbdev-devel] Re: [Xpert]PPC -- XFree86 (ati) support for 1280x854-16 with Radeon Mobility M7 LW (new Apple Titanium)?

[Michel <mi...@da...>]

[ CC: to linux-fbdev-devel as this should also cover some of your
questions there ]

On Tue, 2002-08-06 at 09:13, Miles Lane wrote:
>=20
> What's the status with XFree86 support for the new
> 800MHz Titanium PowerBook?  Every time I have tried
> to start the XFree86 HEAD ati driver, it locks up
> my machine.
>=20
> I have tried starting the server at 8bpp, disabling
> accelleration and forcing PCIMode. =20
>=20
> XFree86 -configure gave me the following (commented out
> lines removed):
>=20
> Section "Device"
> 	Identifier  "Card0"
> 	Driver      "ati"
> 	VendorName  "ATI Technologies Inc"
> 	BoardName   "Radeon Mobility M7 LW"
> 	BusID       "PCI:0:16:0"
> EndSection
>=20
> dmesg (SuSE 2.4.16-npmac kernel):
>=20
> radeonfb: ref_clk=3D2700, ref_div=3D12, xclk=3D36675 from OF
> radeonfb: Failed to detect DFP panel size
> Using unsupported 1280x854 ATY,Crown_A at bc008000, depth=3D8, pitch=3D12=
80
> Console: switching to colour frame buffer device 160x53
> fb0: Open Firmware frame buffer device on
> /pci@f0000000/ATY,CrownParent@10/ATY,Crown_A
> no framebuffer address found for
> /pci@f0000000/ATY,CrownParent@10/ATY,Crown_B

You need a newer kernel.

> fbset:
>=20
> mode "1280x854-85"
>     # D: 100.000 MHz, H: 75.758 kHz, V: 84.740 Hz
>     geometry 1280 854 1280 854 8
>     timings 10000 16 16 16 16 8 8
>     rgba 8/0,8/0,8/0,0/0
> endmode

This is from OFfb and thus bogus.


> Should the radeonfb driver enable the XFree86 fbdev driver=20
> to operate at 16 bpp?  Where can I get the latest/best=20
> radeonfb code?

Recent benh kernel or at least 2.4.18.

> What are the pros and cons of using fbdev verses the XFree86 ati driver?

All cons, no pros really.

The 4.2.0 or later ati/radeon driver should work with Option "UseFBDev"
if you're running radeonfb. You'll have to provide a 1280x854 mode
definition (fbset -x). If you want DRI, you need to get the current DRI
CVS.


--=20
Earthling Michel D=E4nzer (MrCooper)/ Debian GNU/Linux (powerpc) developer
XFree86 and DRI project member   /  CS student, Free Software enthusiast

Re: [Linux-fbdev-devel] pm3fb driver is ok in 2.5.24

[Romain D. <do...@ir...>]

Romain Dolbeau wrote:

> Thanks, but this doesn't help: beside the not really
> needed serial ports and floppy drive, the SCSI
> drivers also don't compile. And this box has no
> IDE channel.

Good news: Only 53c94 doesn't compile, but MESH is OK,
           and linux is on the internal bus (w/ MESH)

Bad news: arch/ppc/kernel/process.c doesn't compile.

So the conlusion stand: 2.4.x only for my box ATM

-- 
DOLBEAU Romain           | I witness a time and a place that never dies,
ENS Cachan / Ker Lann    | still frozen in time, this darkness the only 
Thesard IRISA / CAPS     | place that I can hide. I witness, a dream.   
dol...@cl...|  -- Black Sabbath, 'I Witness'

Re: [Linux-fbdev-devel] pm3fb driver is ok in 2.5.24

[Romain D. <do...@ir...>]

Michel D=E4nzer wrote:

> It's at 2.5.30 now at least.

Thanks, but this doesn't help: beside the not really
needed serial ports and floppy drive, the SCSI
drivers also don't compile. And this box has no
IDE channel.

It's better than it once was, but it still 2.4.x
only for me :-(

Thanks for the help,

--=20
DOLBEAU Romain               | I tell you to enjoy life,               =20
ENS Cachan / Ker Lann        |  I wish I could                         =20
Thesard IRISA / CAPS         |        but it's too late.               =20
dol...@cl...    |  -- Black Sabbath, 'Paranoid'

[Linux-fbdev-devel] PPC -- radeonfb / XFree86 (ati) support for 1280x854-16 with Radeon Mobility M7 LW (new Apple Titanium)?

[Miles L. <mi...@me...>]

A few questions:

1)  What is the current state of syncing between benh's kernel
    tree with 2.4 (latest vanilla and ac) and with 2.5 (
    latest vanilla and dj)?  Should the 2.5 kernels have any
    chance of working with the newest Apple Titanium (88MHz)?

2)  Ani, what's the status with XFree86 support?  Should the
    radeonfb driver enable the XFree86 fbdev driver to operate
    at 16 bpp?  Where can I get the latest/best radeonfb code?

3)  What are the pros and cons of using fbdev verses the 
    XFree86 ati driver?

If anyone has new code they'd like me to test, please let me know.

Thanks,
	Miles