linux-fbdev-users — Generic mailing list for users of frame buffers

Re: [Linux-fbdev-users] fbset doesn't change resolution for all consoles

[Grzegorz A. H. <gr...@ti...>]

> Can you test this patch?  This is against my tree which is based
> on 2.6.12-rc2-mm3 for this patch, but you can use other kernels
> if you're willing to fix hunks.

This makes fbset -a work again. Thanks. I had to modify it for
2.6.10. The worst change was I had to get rid of this check:

> +		if (!vc || vc->vc_mode != KD_TEXT ||

Apparently my version doesn't have a vc_mode field. Not a problem
for me, since by the time I use fbset all my consoles are text mode.

Re: [Linux-fbdev-users] fbset doesn't change resolution for all consoles

[Grzegorz A. H. <gr...@ti...>]

On 2005-04-25, Petr Vandrovec <van...@vc...> wrote:
> As far as I can tell it behaves as designed.

Well, if fbset has an "--all" parameter which doesn't affect all
consoles...  I guess you mean that the current kernel fb support
is designed to not allow this, and it's just fbset which hasn't
been updated?

> If you are not satisifed, try some of matroxfb-* patches from
> http://platan.vc.cvut.cz/ftp/pub/linux/matrox-latest (either
> matroxfb-2.6.10.gz, or upgrade to 2.6.12-rc3 and apply relevant
> matroxfb-2.6.12-rc3.gz).  This patch contains 2.4.x fbcon ported
> to 2.6.x fbdev layer, besides random matroxfb updates, and so
> fbset should work in a way you are used to after that.

Thanks. I've tried using the matroxfb-2.6.10 patch, but at some
points complains about tileops structure not being a complete
type. A grep in the subdirectory doesn't show where this structure
is being defined, so I don't know how to fix this:

In file included from /linux-2.6.10/drivers/video/matrox/matroxfb_base.c:=
106:
/linux-2.6.10/drivers/video/matrox/matroxfb_base.h:519: error: field `til=
eops' has incomplete type
/linux-2.6.10/drivers/video/matrox/matroxfb_base.c: En la funci=F3n `init=
Matrox2':
/linux-2.6.10/drivers/video/matrox/matroxfb_base.c:1980: error: structure=
 has no member named `tileops'

matroxfb_accel complains too if I use "make -k".

> On Linus's 2.6.x you are supposed to use stty cols & rows to
> change resolution, and you are not supposed to change color depth.

I'm shocked. Is he going to ban changing color depth at all in the
future? Right now I boot into some vesa mode and need fbset to set
the same resolution with hardware acceleration to avoid it being
so sluggish. Is there a way to do this with stty rather than fbset?

Re: [Linux-fbdev-users] fbset doesn't change resolution for all consoles

[Antonino A. D. <ad...@ho...>]

On Tuesday 26 April 2005 22:30, Grzegorz Adam Hankiewicz wrote:
> On 2005-04-26, "Antonino A. Daplas" <ad...@ho...> wrote:
> > Not a bug in fbset, but is due to changes in fbcon code from
> > 2.4->2.6.  Now, only the visible console will respond to fbset
> > requests.
> >
> > I'll see what I can do about this bug.
>
> Tell me if I can help in any way.

Can you test this patch?  This is against my tree which is based on
2.6.12-rc2-mm3 for this patch, but you can use other kernels if you're
willing to fix hunks.

Tony

diff -Nru a/drivers/video/console/fbcon.c b/drivers/video/console/fbcon.c
--- a/drivers/video/console/fbcon.c	2005-03-23 11:53:44 +08:00
+++ b/drivers/video/console/fbcon.c	2005-04-27 11:37:20 +08:00
@@ -2598,6 +2598,51 @@
 	}
 }
 
+static void fbcon_set_all_vcs(struct fb_info *info)
+{
+	struct fbcon_ops *ops = info->fbcon_par;
+	struct vc_data *vc;
+	struct display *p;
+	int i, rows, cols;
+
+	if (!ops || ops->currcon < 0)
+		return;
+	
+	for (i = 0; i < MAX_NR_CONSOLES; i++) {
+		vc = vc_cons[i].d;
+		if (!vc || vc->vc_mode != KD_TEXT ||
+		    registered_fb[con2fb_map[i]] != info)
+			continue;
+		
+		p = &fb_display[vc->vc_num];
+
+		info->var.xoffset = info->var.yoffset = p->yscroll = 0;
+		var_to_display(p, &info->var, info);
+		cols = info->var.xres / vc->vc_font.width;
+		rows = info->var.yres / vc->vc_font.height;
+		vc_resize(vc, cols, rows);
+
+		if (CON_IS_VISIBLE(vc)) {
+			updatescrollmode(p, info, vc);
+			scrollback_max = 0;
+			scrollback_current = 0;
+			update_var(vc->vc_num, info);
+			fbcon_set_palette(vc, color_table);
+			update_screen(vc);
+			if (softback_buf) {
+				int l = fbcon_softback_size / vc->vc_size_row;
+				if (l > 5)
+					softback_end = softback_buf + l * vc->vc_size_row;
+				else {
+					/* Smaller scrollback makes no sense, and 0
+					   would screw the operation totally */
+					softback_top = 0;
+				}
+			}
+		}
+	}
+}	
+
 static int fbcon_mode_deleted(struct fb_info *info,
 			      struct fb_videomode *mode)
 {
@@ -2712,6 +2757,9 @@
 		break;
 	case FB_EVENT_MODE_CHANGE:
 		fbcon_modechanged(info);
+		break;
+	case FB_EVENT_MODE_CHANGE_ALL:
+		fbcon_set_all_vcs(info);
 		break;
 	case FB_EVENT_MODE_DELETE:
 		mode = event->data;
diff -Nru a/drivers/video/fbmem.c b/drivers/video/fbmem.c
--- a/drivers/video/fbmem.c	2005-04-14 12:06:43 +08:00
+++ b/drivers/video/fbmem.c	2005-04-27 11:37:20 +08:00
@@ -729,11 +729,13 @@
 
 			if (!err && info->flags & FBINFO_MISC_USEREVENT) {
 				struct fb_event event;
+				int evnt = (var->activate & FB_ACTIVATE_ALL) ?
+					FB_EVENT_MODE_CHANGE_ALL :
+					FB_EVENT_MODE_CHANGE;
 
 				info->flags &= ~FBINFO_MISC_USEREVENT;
 				event.info = info;
-				notifier_call_chain(&fb_notifier_list,
-						    FB_EVENT_MODE_CHANGE,
+				notifier_call_chain(&fb_notifier_list, evnt,
 						    &event);
 			}
 		}
diff -Nru a/include/linux/fb.h b/include/linux/fb.h
--- a/include/linux/fb.h	2005-04-14 12:44:11 +08:00
+++ b/include/linux/fb.h	2005-04-27 11:37:20 +08:00
@@ -495,6 +495,9 @@
 #define FB_EVENT_BLANK                  0x08
 /*      Private modelist is to be replaced */
 #define FB_EVENT_NEW_MODELIST           0x09
+/*	The resolution of the passed in fb_info about to change and
+        all vc's should be changed         */ 
+#define FB_EVENT_MODE_CHANGE_ALL	0x0A
 
 struct fb_event {
 	struct fb_info *info;

Re: [Linux-fbdev-users] fbset doesn't change resolution for all consoles

[James S. <jsi...@ww...>]

> On Linus's 2.6.x you are supposed to use stty cols & rows to change resolution,
> and you are not supposed to change color depth.

You can use fbset. It works.

Re: [Linux-fbdev-users] fbset doesn't change resolution for all consoles

[James S. <jsi...@ww...>]

> > Matrox g450 support built into the kernel as well as fbcon. Standing
> > on a console I try running "fbset -a resolution" or "fbset
> > --all resolution" where resolution is one of the modes listed
> > in /etc/fb.modes. However, only the current console's resolution
> > is changed.
> >
> > Is this a bug in fbset? Is there any new version of this software? I
> > remember using fbset on a 2.4.x kernel with the same video card
> > without problems.
> >
> 
> Not a bug in fbset, but is due to changes in fbcon code from 2.4->2.6.
> Now, only the visible console will respond to fbset requests.
> 
> I'll see what I can do about this bug.

This was done because sometimes the user entered a bad mode which the 
driver could not handle. As a result all the VCs went blank and became 
useless. The only fix was to reboot the box. 

Re: [Linux-fbdev-users] fbset doesn't change resolution for all consoles

[Grzegorz A. H. <gr...@ti...>]

On 2005-04-26, "Antonino A. Daplas" <ad...@ho...> wrote:
> Not a bug in fbset, but is due to changes in fbcon code from
> 2.4->2.6.  Now, only the visible console will respond to fbset
> requests.
> 
> I'll see what I can do about this bug.

Tell me if I can help in any way.

Re: [Linux-fbdev-users] fbset doesn't change resolution for all consoles

[Antonino A. D. <ad...@ho...>]

On Monday 25 April 2005 18:17, Grzegorz Adam Hankiewicz wrote:
> Hi.
>
> I have fbset 2.1 (23/06/1999) and I'm running kernel 2.6.10 with
> Matrox g450 support built into the kernel as well as fbcon. Standing
> on a console I try running "fbset -a resolution" or "fbset
> --all resolution" where resolution is one of the modes listed
> in /etc/fb.modes. However, only the current console's resolution
> is changed.
>
> Is this a bug in fbset? Is there any new version of this software? I
> remember using fbset on a 2.4.x kernel with the same video card
> without problems.
>

Not a bug in fbset, but is due to changes in fbcon code from 2.4->2.6.
Now, only the visible console will respond to fbset requests.

I'll see what I can do about this bug.

Tony

[Linux-fbdev-users] intelfb on 82852/855GM and cursor problem

[Stanislav S. <sse...@cs...>]

Hello,

For those who are on the:

 intelfb: Non-CRT device is enabled ( LVDS port ).  Disabling mode switching.
 intelfb: Video mode must be programmed at boot time.

stage read this: http://www.xfree86.org/~dawes/intelfb.html
and svga.txt in the kernel source documentation. I managed to start
video mode needed, but with tricks: I had to define this

#define CONFIG_FB_VESA_STD

in the beginning of the arch/i386/boot/video.S file. Without it I couldn't
start video mode via "vga=" or etc. though I played a lot with all
that stuff described in svga.txt, the definition I used is not mentioned
in svga.txt. What was wrong?

Ok, now I have it running but my dmesg is full of

the cursor was killed - restore it !!

and there is no cursor in console (without X started). I tried turning off hardware
cursor support, didn't help, could you suggest what to do next?

Thanks in advance

Regards,
Stas

Re: [Linux-fbdev-users] fbset doesn't change resolution for all consoles

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

Grzegorz Adam Hankiewicz wrote:
> Hi.
> 
> I have fbset 2.1 (23/06/1999) and I'm running kernel 2.6.10 with
> Matrox g450 support built into the kernel as well as fbcon. Standing
> on a console I try running "fbset -a resolution" or "fbset
> --all resolution" where resolution is one of the modes listed
> in /etc/fb.modes. However, only the current console's resolution
> is changed.
> 
> Is this a bug in fbset? Is there any new version of this software? I
> remember using fbset on a 2.4.x kernel with the same video card
> without problems.

As far as I can tell it behaves as designed.  If you are not satisifed,
try some of matroxfb-* patches from http://platan.vc.cvut.cz/ftp/pub/linux/matrox-latest
(either matroxfb-2.6.10.gz, or upgrade to 2.6.12-rc3 and apply relevant
matroxfb-2.6.12-rc3.gz).  This patch contains 2.4.x fbcon ported to 2.6.x
fbdev layer, besides random matroxfb updates, and so fbset should work
in a way you are used to after that.

On Linus's 2.6.x you are supposed to use stty cols & rows to change resolution,
and you are not supposed to change color depth.
							Petr Vandrovec

[Linux-fbdev-users] fbset doesn't change resolution for all consoles

[Grzegorz A. H. <gr...@ti...>]

Hi.

I have fbset 2.1 (23/06/1999) and I'm running kernel 2.6.10 with
Matrox g450 support built into the kernel as well as fbcon. Standing
on a console I try running "fbset -a resolution" or "fbset
--all resolution" where resolution is one of the modes listed
in /etc/fb.modes. However, only the current console's resolution
is changed.

Is this a bug in fbset? Is there any new version of this software? I
remember using fbset on a 2.4.x kernel with the same video card
without problems.

Re: [Linux-fbdev-users] intelfb on 82852/855GM

[Sylvain M. <syl...@wo...>]

        Hi,
  =20
    Local Flat Panel support isn't fully implemented in intelfb. Only=20
CRT output is.  The workaround is to use the BIOS to program the graphic=20
chip at boot time using the 'vga=3D' option. One restriction is the drive=
r=20
cannot change video modes.

Regards
Sylvain

themuse a =E9crit:

>Hello,=20
>
>I'm trying to get intelfb compiled into kernel 2.6.11(-gentoo-r5) with=20
>
>video=3Dintelfb:1024x768-16@60=20
>
>in my grub.conf kernel command line to work with my fujitsu-siemens
>lifebook (with intel 82852/855GM graphics) and keep getting the
>following in dmesg:
>
>intelfb: Framebuffer driver for Intel(R) 830M/845G/852GM/855GM/865G
>chipsets
>intelfb: Version 0.9.2
>intelfb: 00:02.0: Intel(R) 855GM, aperture size 128MB, stolen memory
>8060kB
>intelfb: Non-CRT device is enabled ( LVDS port ).  Disabling mode
>switching.
>intelfb: Video mode must be programmed at boot time.
>
>And I'm left with the text console. I've also tried
>other modes (theoretically, the system supports 1400x1050-24@60), but
>with no success so far. (Research in Gentoo forums makes me think I'm
>not the only person getting these error messages, but I didn't find
>anything usable there.) Any help would be welcome, I'd help gladly with
>testing fixes (if it's the driver's and not my fault).=20
>
>Greetings
>themuse
>
>
>
>-------------------------------------------------------
>This SF.Net email is sponsored by: New Crystal Reports XI.
>Version 11 adds new functionality designed to reduce time involved in
>creating, integrating, and deploying reporting solutions. Free runtime i=
nfo,
>new features, or free trial, at: http://www.businessobjects.com/devxi/72=
8
>_______________________________________________
>Linux-fbdev-users mailing list
>Lin...@li...
>https://lists.sourceforge.net/lists/listinfo/linux-fbdev-users
>
> =20
>

[Linux-fbdev-users] intelfb on 82852/855GM

[themuse <th...@gm...>]

Hello, 

I'm trying to get intelfb compiled into kernel 2.6.11(-gentoo-r5) with 

video=intelfb:1024x768-16@60 

in my grub.conf kernel command line to work with my fujitsu-siemens
lifebook (with intel 82852/855GM graphics) and keep getting the
following in dmesg:

intelfb: Framebuffer driver for Intel(R) 830M/845G/852GM/855GM/865G
chipsets
intelfb: Version 0.9.2
intelfb: 00:02.0: Intel(R) 855GM, aperture size 128MB, stolen memory
8060kB
intelfb: Non-CRT device is enabled ( LVDS port ).  Disabling mode
switching.
intelfb: Video mode must be programmed at boot time.

And I'm left with the text console. I've also tried
other modes (theoretically, the system supports 1400x1050-24@60), but
with no success so far. (Research in Gentoo forums makes me think I'm
not the only person getting these error messages, but I didn't find
anything usable there.) Any help would be welcome, I'd help gladly with
testing fixes (if it's the driver's and not my fault). 

Greetings
themuse

[Linux-fbdev-users] tridentfb with cyberblade i7

[Olaf L. <le...@ne...>]

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

Hello everybody!

I just installed ubuntu-linux on an older notebook with a trident
cyberblade i7 onboard graphic card and I have several problems with
tridentfb from vanilla-linux-2.6.11

1. The kernel commandline seems to be ignored:

both
video=tridentfb:800x600,bpp=16
and
video=tridentfb,mode:800x600-16@60,fb

deserve 640x480-8

2. No matter what mode being selected the cursor is fixed to the last
line (after touching it the first time) and scrolling is only done in
this line (does not apply to ncurses apps which control the terminal).

3. clear doesn't clear the screen but puts the cursor to the first line.
 (ncurses apps also seem to have this problem)

I took some screenshots:

http://leidola.newcon.de/pixmap/laptop-fb/aptitude.png
http://leidola.newcon.de/pixmap/laptop-fb/refreshed.png
http://leidola.newcon.de/pixmap/laptop-fb/refreshed-fbset.png


As you can see, there is some strange noise at the bottom-right corner
in 800x600x16 mode (set via fbset).

lscpi:
0000:01:00.0 VGA compatible controller: Trident Microsystems
CyberBlade/i7d (rev 5c)



dmesg:

Apr 18 15:41:40 laptop kernel: Kernel command line: root=/dev/hda2 ro
quiet splash video=tridentfb:800x600,bpp=16,fb
...
Apr 18 15:41:40 laptop kernel: tridentfb: Trident framebuffer
0.7.8-NEWAPI initializing
Apr 18 15:41:40 laptop kernel: ACPI: PCI Interrupt Link [LNKA] enabled
at IRQ 9
Apr 18 15:41:40 laptop kernel: PCI: setting IRQ 9 as level-triggered
Apr 18 15:41:40 laptop kernel: ACPI: PCI interrupt 0000:01:00.0[A] ->
GSI 9 (level, low) -> IRQ 9
Apr 18 15:41:40 laptop kernel: tridentfb: framebuffer size = 4096 Kb
Apr 18 15:41:40 laptop kernel: tridentfb: 0000:01:00.0 board found
Apr 18 15:41:40 laptop kernel: tridentfb: 800x600 flat panel found
Apr 18 15:41:40 laptop kernel: Console: switching to colour frame buffer
device
80x25
Apr 18 15:41:40 laptop kernel: tridentfb: fb0: Trident frame buffer
device 640x400-16bpp
...



What could I do about it?

Ciao,

Olaf
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.1 (GNU/Linux)
Comment: Using GnuPG with Thunderbird - http://enigmail.mozdev.org

iD8DBQFCZE4p4Yz1Ukx83/kRAngiAJ908zRqnHRxRNiomqfKZRVY89/FwgCgs2Qy
zenHZm5euItFLeeI/GdU0Ak=
=cePQ
-----END PGP SIGNATURE-----

[Linux-fbdev-users] intelfb on 82852/855GM

[themuse <th...@gm...>]

Hello, 

I'm trying to get intelfb compiled into kernel 2.6.11(-gentoo-r5) with 

video=intelfb:1024x768-16@60 

in my grub.conf kernel command line to work with my fujitsu-siemens
lifebook (with intel 82852/855GM graphics) and keep getting the
following in dmesg:

intelfb: Framebuffer driver for Intel(R) 830M/845G/852GM/855GM/865G
chipsets
intelfb: Version 0.9.2
intelfb: 00:02.0: Intel(R) 855GM, aperture size 128MB, stolen memory
8060kB
intelfb: Non-CRT device is enabled ( LVDS port ).  Disabling mode
switching.
intelfb: Video mode must be programmed at boot time.

Needless to say, the fb doesn't work for me so far. I've also tried
other modes (theoretically, the system supports 1400x1050-24@60), but
with no success so far. (Research in Gentoo forums makes me think I'm
not the only person getting these error messages, but I didn't find
anything usable there.) Any help would be welcome, I'd help gladly with
testing fixes (if it's the driver's and not my fault). 

Greetings
themuse

Re: [Linux-fbdev-users] 16/9 fb resolution

[Dely N. <Nic...@th...>]

On Fri, 2005-04-08 at 11:10 +0100, Torgeir Veimo wrote:

> Of course! I use 1280x720, not 768 though. I have these in
> my /etc/fb.modes file;
> 
> # 50 Hz
> mode "1280x720-50"
>    # D: 61.80 MHz, H: 37.50 kHz, V: 50.00 Hz
>    geometry 1280 720 1280 720 32
>    timings 16181 218 110 20 5 40 5
> endmode
> 
> # 60 Hz 
> mode "1280x720-60"
>    # D: 74.25 MHz, H: 45.00 kHz, V: 60.00 Hz
>    geometry 1280 720 1280 720 16
>    timings 13468 220 110 20 5 40 5
> endmode
> 
> Then an fbset "1280x720-50" does the trick. Or if you use DirectFB
> applications, just put in your /etc/directfb file;
> 
> mode=1280x720
> depth=32
> 
> to use this resolution by default.
> 
> To use 1280x768, you need to find the right timing parameters. You might
> find a modeline for X, and calculate the correct values for fb.modes;
> 
> http://www.faqs.org/docs/Linux-HOWTO/Framebuffer-HOWTO.html#s18
> 
> Using google usually gives you something usable, eg. 
> 
> http://www.mail-archive.com/lin...@li.../msg15331.html
> 
> mode "1280x768 50Hz 32bit"
>     geometry 1280 768 1280 768 32
>     timings 11415 248 16 144 30 1 3
> endmode

Oh thank you very much, it's right 1280x720 is exactly 16/9, but I try
with a Thomson Scenium and its resolution seems to be 1280x768, so thank
you for these last lines too.

Re: [Linux-fbdev-users] 16/9 fb resolution

[Torgeir V. <to...@po...>]

On Fri, 2005-04-08 at 11:11 +0200, Dely Nicolas wrote:
> Hi,
> 
> Is it possible to have 16:9 resolution in framebuffer?
> 
> maybe 1280x768 about ?
> 
> Has anybody link or documentation on it?

Of course! I use 1280x720, not 768 though. I have these in
my /etc/fb.modes file;

# 50 Hz
mode "1280x720-50"
   # D: 61.80 MHz, H: 37.50 kHz, V: 50.00 Hz
   geometry 1280 720 1280 720 32
   timings 16181 218 110 20 5 40 5
endmode

# 60 Hz 
mode "1280x720-60"
   # D: 74.25 MHz, H: 45.00 kHz, V: 60.00 Hz
   geometry 1280 720 1280 720 16
   timings 13468 220 110 20 5 40 5
endmode

Then an fbset "1280x720-50" does the trick. Or if you use DirectFB
applications, just put in your /etc/directfb file;

mode=1280x720
depth=32

to use this resolution by default.

To use 1280x768, you need to find the right timing parameters. You might
find a modeline for X, and calculate the correct values for fb.modes;

http://www.faqs.org/docs/Linux-HOWTO/Framebuffer-HOWTO.html#s18

Using google usually gives you something usable, eg. 

http://www.mail-archive.com/lin...@li.../msg15331.html

mode "1280x768 50Hz 32bit"
    geometry 1280 768 1280 768 32
    timings 11415 248 16 144 30 1 3
endmode




-- 
Torgeir Veimo <to...@po...>

[Linux-fbdev-users] 16/9 fb resolution

[Dely N. <Nic...@th...>]

Hi,

Is it possible to have 16:9 resolution in framebuffer?

maybe 1280x768 about ?

Has anybody link or documentation on it?

Thanks

Regards

Nicolas

[Linux-fbdev-users] Oops with Rage XL init.

[Richard S. <smi...@gm...>]

On Thu, 31 Mar 2005 10:53:46 -0600, Richard Smith <smi...@gm...> wrote:

Sorry about the incorrect subject.  I forgot to change it before I sent.

> Who is the maintainer of the atyfb?
> 
> I get an Div error (division by zero I think)  and kernel oops when I
> try to load the module in my system with 2.6.11.4
> 
> Its currently a secondary card but I have the Rage XL no bios init
> enabled.  I'm not sure yet if its being called.
> 
> Is the Rage XL no bios stuff supposed to work?
> 
> I would send an ksymoops procesed log but it ksymoops seems to issues
> finding everything on my system currently.  I'll send one along when I
> get that worked out.
> 
> --
> Richard A. Smith
> 


-- 
Richard A. Smith

Re: [Linux-fbdev-users] voodoo5 tdfxfb mmap problems

[Richard S. <smi...@gm...>]

Who is the maintainer of the atyfb? 

I get an Div error (division by zero I think)  and kernel oops when I
try to load the module in my system with 2.6.11.4

Its currently a secondary card but I have the Rage XL no bios init
enabled.  I'm not sure yet if its being called.

Is the Rage XL no bios stuff supposed to work?

I would send an ksymoops procesed log but it ksymoops seems to issues
finding everything on my system currently.  I'll send one along when I
get that worked out.

-- 
Richard A. Smith

[Linux-fbdev-users] voodoo5 tdfxfb mmap problems

[Marlon de B. <ma...@su...>]

I'm trying to get mplayer to work with tdfxfb on a 2.6 kernel, when looking at 
the source I found out that

	/* Open up a window to the hardware */
        memBase1 = mmap(0, fb_finfo.smem_len, PROT_READ | PROT_WRITE,
                                        MAP_SHARED, fd, 0);
        memBase0 = mmap(0, fb_finfo.mmio_len, PROT_READ | PROT_WRITE,
                                        MAP_SHARED, fd, fb_finfo.smem_len);

was causing the problem. The second mmap (memBase0) returned EINVAL. 

man mmap says " EINVAL We don't like start or length or offset.  (E.g., they 
are too large, or not aligned on  a  PAGE-SIZE boundary".

So I thought that it must be some problem with the mmio mem length. I wrote a 
small program that printed out some of the fb_fix_screeninfo structure.

smem_start d0000000 
mmio_start d8000000 
mmio_len 67108864 (64Mb)
smem_len 33554432 (32Mb)

My voodoo5 has 2 gpu's each with 32Mb memory (according to the hardware 
specs).

lspci:

0000:01:00.0 VGA compatible controller: 3Dfx Interactive, Inc. Voodoo 4 / 
Voodoo 5 (rev 01) (prog-if 00 [VGA])
        Subsystem: 3Dfx Interactive, Inc.: Unknown device 0002
        Flags: 66Mhz, fast devsel, IRQ 16
        Memory at d8000000 (32-bit, non-prefetchable)
        Memory at d0000000 (32-bit, prefetchable) [size=128M]
        I/O ports at a000 [size=256]
        Capabilities: [54] AGP version 2.0
        Capabilities: [60] Power Management version 2

dmesg:

ACPI: PCI Interrupt 0000:01:00.0[A] -> GSI 16 (level, low) -> IRQ 16
fb: 3Dfx Voodoo5 memory = 32768K
Console: switching to colour frame buffer device 80x30

/proc/mtrr

reg00: base=0x00000000 (   0MB), size= 128MB: write-back, count=1
reg01: base=0xd0000000 (3328MB), size=  64MB: write-combining, count=2

Some useful info from Xorg.0.log

II) TDFX(0): Softbooting the board (through the int10 interface).
(EE) TDFX(0): shmget(lowmem) error: Unknown error 999
(WW) TDFX(0): Softbooting the board failed.
(II) TDFX(0): TDFXFindChips: found 1 chip(s)
(**) TDFX(0): Depth 24, (--) framebuffer bpp 32
(==) TDFX(0): RGB weight 888
(==) TDFX(0): Default visual is TrueColor
(--) TDFX(0): Chipset: "3dfx Voodoo5"
(--) TDFX(0): Linear framebuffer at 0xD0000000
(--) TDFX(0): MMIO registers at addr 0xD8000000
(--) TDFX(0): PIO registers at addr 0xA000
(II) TDFX(0): DRAMINIT1 read 0x40240031, programming 0x40202031 (not Banshee)
(II) TDFX(0): TDFXInitChips: numchips = 1
(II) TDFX(0): TDFXInitChips: cfgbits = 0x00000004, initbits = 0x00000301
(II) TDFX(0): TDFXInitChips: mem0base = 0xd8000000, mem1base = 0xd0000008
(II) TDFX(0): TDFXInitChips: mem0size = 0x02000000, mem1size = 0x04000000
(II) TDFX(0): TDFXInitChips: mem0bits = 0x00000005, mem1bits = 0x00000040
(II) TDFX(0): TDFXInitChips: cfgbits = 0x00000045
(II) TDFX(0): TDFXInitChips: MMIOAddr[0] = 0xd8000000
(II) TDFX(0): TDFXInitChips: LinearAddr[0] = 0xd0000008
(--) TDFX(0): VideoRAM: 32768 kByte Mapping 65536 kByte

(==) TDFX(0): Write-combining range (0xd0000000,0x4000000)
(II) TDFX(0): vgaHWGetIOBase: hwp->IOBase is 0x03d0, hwp->PIOOffset is 0x9d00
(II) TDFX(0): Changing back offset from 0x019ff000 to 0x019fe000
(II) TDFX(0): Textures Memory 21.42 MB
(II) TDFX(0): Cursor Offset: [0x00000000,0x00001000)
(II) TDFX(0): Fifo Offset: [0x00001000, 0x00041000)
(II) TDFX(0): Front Buffer Offset: [0x00041000, 0x00493000)
(II) TDFX(0): Texture Offset: [0x00493000, 0x019FE000)
(II) TDFX(0): BackOffset: [0x019FE000, 0x01CFE000)
(II) TDFX(0): DepthOffset: [0x01CFF000, 0x01FFF000)
(II) TDFX(0): Minimum 338, Maximum 3327 lines of offscreen memory available
(II) TDFX(0): [dri] VideoRAM = 32768, VirtualXres = 1024, VirtualYres= 768,

I'm running Linux freakypeople 2.6.11.6-grsec #1 SMP Wed Mar 30 09:48:16 CEST 
2005 i686 Pentium III (Coppermine) GenuineIntel GNU/Linux

I know that's kind of old hardware and 3dfx is dead for long time now, but 
does anybody has an idea how I prevent

memBase0 = mmap(0, fb_finfo.mmio_len, PROT_READ | PROT_WRITE,
                                        MAP_SHARED, fd, fb_finfo.smem_len);

from returning EINVAL? I'm not really a framebuffer / graphics programmer, so 
I don't know much about the internal functions of the framebuffer, (did a 
quick scan of the tdfxfb source in the kernel though). My guess is I'm 
mmapping the wrong area. I tried to change memBase0 to

memBase0 = mmap((unsigned long *)&fb_finfo.mmio_start, fb_finfo.mmio_len, 
PROT_READ | PROT_WRITE, MAP_SHARED, fd, 0); , without any luck

from the mmap man

The  mmap() function asks to map length bytes starting at offset offset from 
the file (or other object) specified by the file descriptor fd into memory, 
preferably at address start.  This latter address is a  hint  only,  and is 
usually specified as 0.  

So I did an strace on my program and found out that mmap didn't use the hint 
at all.

Anybody an idea?

Regards,

Marlon.

[Linux-fbdev-users] nvidiafb: known to work yet with AdvanceMAME?

[John W. C. <jo...@co...>]

Has anyone got the new nvidiafb driver working with AdvanceMAME, or with
anything other than fbset?  AdvanceMAME will crash with a seg fault when
interacting with the fb APIs, whereas it won't with other fb drivers
(rivafb and savagefb).

I know the driver is very new.  What can I do to help interested parties
track down this likely nvidiafb bug?

Thanks,
John

[Linux-fbdev-users] X server fails after i810fb insertion

[Mukund JB. <muk...@es...>]

Hi Adaplas,

I am unable to activate the X server on my Desktop after inserting the
i810fb driver as follows.

	modprobe i810fb vram=3D2 xres=3D1024 bpp=3D16 hsync1=3D30 hsync2=3D55
vsync1=3D50 	vsync2=3D85 accel=3D1
After that I get no GUI on the Desktop & when I say startx, I get some
errors & I have attached the relevant Xserver_log file as well as the
configuration file to this mail.

I am specific to i810fb driver as I want to rotate the screen using X
server

After that on trial basis when I modies the Device section as follows,
trhe GUI is working with a light shift in the screen. I thing that can
be rectified with modification is resolution but what is "fbdev" which
made my GUI work.

	Section " Device"
		Identifier "videocard0"
		Driver "fbdev"
		VendorName "videocard vendor"
		BoardName "Intel 810"
		VideoRam 16384
	EndSection

How do I solve this problem?=20
Thanks in advance.

Regards,
Mukund jampala




> > Is it a must that hardware should support the screen rotation. Can't
we
> > handle that in an fb driver? And, if possible do explain me that
like
> > How & where the fb driver allows changes to support screen rotation.
>=20
> It doesn't.
>=20
> >
> > Please suggest me how can I activate the frambuffer on RedHat 9. It
is
> > not active on my system. When I say fbset -i, it responds saying
> > open /dev/fb0: No such device
> >
> > I have a Intel Corp. 82810E DC-133 Graphic controller on my system.
Is a
> > fb driver specific to this. Does the generic fbdev work with this
> > controller?
>=20
> vesafb does not work for the i810.  You have to enable the specific
driver
> for it, i810fb, in your kernel config.  The i810fb is only available
for
> 2.6
> kernels, but you can grab an old patch for 2.4 in
>=20
> http://i810fb.sf.net
>=20
> Grab the latest one, and it should apply to any recent 2.4 kernel.
You
> might
> get hunk failures but it should be easy to fix.
>=20
> Tony
>=20
> BTW, the i810 does not support hardware rotation. It can be done in
> software,
> if you enable the fbdev driver for X.
>=20

[Linux-fbdev-users] intelfb 855gm memory issue

[Baard K. <ba...@in...>]

Hello

I have a Dell Inspiron 510M with Intel 855GM chip.
510M has no video memory option in bios(with the latest A09) , so agpgart
allocates 892KB video memory. In other words, I can't run at any higher
res. than 1024x768-8. 
I run 1400x1050 in Xfree86, but it seems to allocate memory on its own:

(II) I810(0): detected 892 kB stolen memory.
(II) I810(0): I830CheckAvailableMemory: 705532 kB available
(II) I810(0): Will attempt to tell the BIOS that there is 12288 kB VideoRAM

How can I make intelfb(or agpgart?) allocate more ram for higher res.? 

intelfb has an option
MODULE_PARM_DESC(vram, "System RAM to allocate to framebuffer in MiB");
but passing "video=intelfb:1024x768@60-16,mode=1024x768@60-16,vram=16,hwcursor=0"
doesn't help(intelfb says vram=0).

Any way to fix this? I don't think Dell is going to update their bios,
so that's a dead end.

I run 2.6.11.2

Baard

[Linux-fbdev-users] Re: [Linux-fbdev-devel] "remap_page_range" i810 compile time problem ..

[Antonino A. D. <ad...@ho...>]

On Friday 11 March 2005 14:36, Mukund JB. wrote:
> Hi adaplas,
>
> Thankful to u for making things clear at every level.
>
> > > But when I try this without accel=1 rotation fails. Can u just
>
> narrate
>
> > > me what is accel accentually doing?
> >
> > Because if accel is on, the driver is using it's own version of
>
> drawing
>
> > functions which are flexible enough that additional features can be
>
> added,
>
> > such as screen rotation.
> >
> > If accel is off, the driver will use the generic functions which
>
> cannot do
>
> > screen rotation.  Of course, one can write a software version that can
>
> do
>
> > screen rotation.
>
> So, u mean the i810fb driver will process the command line parameters &
> if finds the accel is "1" in the list, it will use its own functions or
> else it will use the generic functions provided in the fbgen.c file.

Above is correct, except that i810fb, if accel is disabled will use the
functions in fbcon-cfb*.c, not fbgen.c.

>
> This is my understanding about fb driver architecture
>
> 			Application
> 			-----------
> 		  XFree86 / Directfb (Embedded)
> 			-----------
> 			 FB driver
> 			-----------
> 			 HARDWARE
>
> Are these files below a part of fb driver or do they come under a
> different level in fb driver architecture?

The above is correct, assuming that XFree86 uses the 'fbdev' driver.

>
> fbcon.c - low level framebuffer based console driver
> fbcon-cfb* - low level framebuffer based operations for 8 bpp packed
> pixels
> How do I differentiate between the above two?

fbcon.c is the glue between console and fbdev.  fbcon-cfb* contains
generic drawing functions for framebuffers with packed pixel format.

>
> fbcmap.c 	- It handles color mapping
> fbcon-plan2p*
>
> I find fbdevhw.a & I find groups referring to it.

fbdevhw, if I'm not mistaken, came from XFree86/Xorg.  It contains
helper functions used by XFree86 to talk to fbdev.

> I think it is a library. If so, give me some inputs where we can use it.
> Please do correct if I am going OFF-TRACK understanding FB drivers.
>
> Where exactly should I place the fbdevhw.a in the above shown
> architecture?

Probably between  XFree86 and fbdev.

Tony

[Linux-fbdev-users] RE: [Linux-fbdev-devel] "remap_page_range" i810 compile time problem ..

[Mukund JB. <muk...@es...>]

Hi adaplas,

Thankful to u for making things clear at every level.

> > But when I try this without accel=3D1 rotation fails. Can u just
narrate
> > me what is accel accentually doing?
>=20
> Because if accel is on, the driver is using it's own version of
drawing
> functions which are flexible enough that additional features can be
added,
> such as screen rotation.
>=20
> If accel is off, the driver will use the generic functions which
cannot do
> screen rotation.  Of course, one can write a software version that can
do
> screen rotation.

So, u mean the i810fb driver will process the command line parameters &
if finds the accel is "1" in the list, it will use its own functions or
else it will use the generic functions provided in the fbgen.c file.

This is my understanding about fb driver architecture

			Application
			-----------
		  XFree86 / Directfb (Embedded)
			-----------
			 FB driver
			-----------
			 HARDWARE

Are these files below a part of fb driver or do they come under a
different level in fb driver architecture?

fbcon.c - low level framebuffer based console driver
fbcon-cfb* - low level framebuffer based operations for 8 bpp packed
pixels
How do I differentiate between the above two?

fbcmap.c 	- It handles color mapping
fbcon-plan2p*

I find fbdevhw.a & I find groups referring to it.
I think it is a library. If so, give me some inputs where we can use it.
Please do correct if I am going OFF-TRACK understanding FB drivers.

Where exactly should I place the fbdevhw.a in the above shown
architecture?

Regards,
Mukund jampala