linuxconsole-dev

Re: 2.6.0-test6 ruby success with matrox G550, and some questions

[Aivils S. <Aiv...@un...>]

>> Current ruby support straight in kernel compiled fb device
>> drivers and fbcon. To test Your matrox figure fbcon into
>> kernel and reboot.
>2.6.0-test6-ruby dies before the first printk if I do that,
>while 2.6.0-test6 works.  fbcon on matrox breaks somehow
>with ruby.  I don't know what to do to debug this
>further - adding printk's is probably not an option.

Strange. Normal is 5-10 sec VGA console output until
fb_console_init is called. After fb_console_init You
can read printk only on serial console terminal in case
fbcon is broken.

Aivils

Re: 2.6.0-test6 ruby success with matrox G550, and some questions

[Aivils S. <Aiv...@un...>]

>> Under ruby patched only hardware independ abstract layer -
>> frambuffer console fbcon.c .
>> If matroxfb will run under vanilla 2.6.0-test6, then it
>> should do same under 2.6.0-test6-ruby
>>
>Unfortunately it fails with ruby. :-(

Ok. I will be glad if You test G550 with single console only under ruby.

Please apply this patch, witch allow one console per fbdev.
Secondary console init may lead oops on matrox g550.

diff -Nu8rp ruby-260t9-CVS-20031028/drivers/video/console/fbcon.c
r-CHG/drivers/video/console/fbcon.c
--- ruby-260t9-CVS-20031028/drivers/video/console/fbcon.c    2003-10-28
10:08:58.000000000 +0000
+++ r-CHG/drivers/video/console/fbcon.c  2003-10-29 12:03:59.000000000
+0000
@@ -2337,17 +2337,17 @@ int __init fb_console_init(void)

     if(!vt2fb[0])
          vt2fb[0] = TAKE_OVER_CONSOLE;

     for(unit = 1; unit < num_registered_fb; unit++)
          if(!vt2fb[unit])
               vt2fb[unit] = MAX_NR_USER_CONSOLES;

-    for(unit = 0; unit < num_registered_fb; unit++)
+    for(unit = 0; unit < 1; unit++)
          if(vt2fb[unit] == TAKE_OVER_CONSOLE) {
                     admin_vt->data_hook = (void *)registered_fb[unit];
               take_over_console(admin_vt, &fb_con);
          }
          else
               fbcon_add(unit, vt2fb[unit]);
     return 0;
 }

2.6.0-test9-ruby available on CVS.

On success i should think up solution.

Aivils

Re: 2.6.0-test6 ruby success with matrox G550, and some questions

[Helge H. <hel...@ai...>]

Aivils Stoss wrote:
>>>Under ruby patched only hardware independ abstract layer -
>>>frambuffer console fbcon.c .
>>>If matroxfb will run under vanilla 2.6.0-test6, then it
>>>should do same under 2.6.0-test6-ruby
>>>
>>
>>Unfortunately it fails with ruby. :-(
> 
> 
> Ok. I will be glad if You test G550 with single console only under ruby.
> 
> Please apply this patch, witch allow one console per fbdev.
> Secondary console init may lead oops on matrox g550.

Thanks, I'll try.  This will give me console on one screen only, b


The ruby version I tried were ruby-260t6-20030930.diff.bz2
I just tried downloading ruby-260t9-CVS-20031028.tar.bz2
Copying the files into a linux-2.6.0-test9 kernel tree
didn't compile, it failed with:

drivers/char/keyboard.c: In function `kbd_keycode':
drivers/char/keyboard.c:1008: error: `handle' undeclared (first use in this function)
drivers/char/keyboard.c:1008: error: (Each undeclared identifier is reported only once
drivers/char/keyboard.c:1008: error: for each function it appears in.)
make[2]: *** [drivers/char/keyboard.o] Error 1
make[1]: *** [drivers/char] Error 2
make: *** [drivers] Error 2
make: *** Waiting for unfinished jobs....



Helge Hafting

Re: 2.6.0-test6 ruby success with matrox G550, and some questions

[Helge H. <hel...@ai...>]

Helge Hafting wrote:
> Aivils Stoss wrote:
> 
>>>> Under ruby patched only hardware independ abstract layer -
>>>> frambuffer console fbcon.c .
>>>> If matroxfb will run under vanilla 2.6.0-test6, then it
>>>> should do same under 2.6.0-test6-ruby
>>>>
>>>
>>> Unfortunately it fails with ruby. :-(
>>
>>
>>
>> Ok. I will be glad if You test G550 with single console only under ruby.
>>
>> Please apply this patch, witch allow one console per fbdev.
>> Secondary console init may lead oops on matrox g550.
> 
The patch doesn't apply to ruby-260t6-20030930, the fbcon.c
file seems very different.  What do I do to get a 
ruby-260t9 that compiles?

Helge Hafting

Re: 2.6.0-test6 ruby success with matrox G550, and some questions

[Svetoslav S. <sv...@gm...>]

> Helge Hafting wrote:
> > Aivils Stoss wrote:
> > 
> >>>> Under ruby patched only hardware independ abstract layer -
> >>>> frambuffer console fbcon.c .
> >>>> If matroxfb will run under vanilla 2.6.0-test6, then it
> >>>> should do same under 2.6.0-test6-ruby
> >>>>
> >>>
> >>> Unfortunately it fails with ruby. :-(
> >>
> >>
> >>
> >> Ok. I will be glad if You test G550 with single console only under
> ruby.
> >>
> >> Please apply this patch, witch allow one console per fbdev.
> >> Secondary console init may lead oops on matrox g550.
> > 
> The patch doesn't apply to ruby-260t6-20030930, the fbcon.c
> file seems very different.

it's for later relases with multi-user VT support

> What do I do to get a  ruby-260t9 that compiles?

the patch that i just send (latest cvs)
does compile for me, but i haven't yet tried booting it

for cvs 
(which on sf.net is sometimes brocken or behind,but currently looks Ok) 

cvs -d:pserver:ano...@cv...:/cvsroot/linuxconsole login
[ press Enter, no pass]

cvs -z3 -d:pserver:ano...@cv...:/cvsroot/linuxconsole
checkout ruby/ruby-2.6

best,

svetljo

-- 
NEU FÜR ALLE - GMX MediaCenter - für Fotos, Musik, Dateien...
Fotoalbum, File Sharing, MMS, Multimedia-Gruß, GMX FotoService

Jetzt kostenlos anmelden unter http://www.gmx.net

+++ GMX - die erste Adresse für Mail, Message, More! +++

Re: 2.6.0-test6 ruby success with matrox G550, and some questions

[Svetoslav S. <sv...@gm...>]

> Aivils Stoss wrote:
> >>>Under ruby patched only hardware independ abstract layer -
> >>>frambuffer console fbcon.c .
> >>>If matroxfb will run under vanilla 2.6.0-test6, then it
> >>>should do same under 2.6.0-test6-ruby
> >>>
> >>
> >>Unfortunately it fails with ruby. :-(
> > 
> > 
> > Ok. I will be glad if You test G550 with single console only under ruby.
> > 
> > Please apply this patch, witch allow one console per fbdev.
> > Secondary console init may lead oops on matrox g550.
> 
> Thanks, I'll try.  This will give me console on one screen only, b
> 
> 
> The ruby version I tried were ruby-260t6-20030930.diff.bz2
> I just tried downloading ruby-260t9-CVS-20031028.tar.bz2
> Copying the files into a linux-2.6.0-test9 kernel tree
> didn't compile, it failed with:
> 
> drivers/char/keyboard.c: In function `kbd_keycode':
> drivers/char/keyboard.c:1008: error: `handle' undeclared (first use in
> this function)
> drivers/char/keyboard.c:1008: error: (Each undeclared identifier is
> reported only once
> drivers/char/keyboard.c:1008: error: for each function it appears in.)
> make[2]: *** [drivers/char/keyboard.o] Error 1
> make[1]: *** [drivers/char] Error 2
> make: *** [drivers] Error 2
> make: *** Waiting for unfinished jobs....
> 

you probably need this (already fixed in cvs by Aivils )

Index: keyboard.c
===================================================================
RCS file: /cvsroot/linuxconsole/ruby/ruby-2.6/drivers/char/keyboard.c,v
retrieving revision 1.9
retrieving revision 1.10
diff -u -d -r1.9 -r1.10
--- keyboard.c    28 Oct 2003 07:17:36 -0000    1.9
+++ keyboard.c    29 Oct 2003 10:45:27 -0000    1.10
@@ -1005,7 +1005,7 @@
         return;
     }
     if (sysrq_down && down && !rep) {
-        handle_sysrq(kbd_sysrq_xlate[keycode], handle->dev->regs, tty);
+        handle_sysrq(kbd_sysrq_xlate[keycode], vt->keyboard->dev->regs,
tty);
         return;
     }

and  attached patch 
wich already contains the change :-)


-- 
NEU FÜR ALLE - GMX MediaCenter - für Fotos, Musik, Dateien...
Fotoalbum, File Sharing, MMS, Multimedia-Gruß, GMX FotoService

Jetzt kostenlos anmelden unter http://www.gmx.net

+++ GMX - die erste Adresse für Mail, Message, More! +++

Re: 2.6.0-test6 ruby success with matrox G550, and some questions

[Svetoslav S. <sv...@gm...>]

> I tried booting ruby-2.6-test9 with the single console patch
> and framebuffer console again.
> 
> ruby without the patch hangs as soon as it switch away
> >from vga.
> 
> runy with the patch gets further.  The framebuffer shows
> only garbage, similar to ruby with framebuffer but without
> fbcon.  (Plain 2.6.0-test9 shows text on the framebuffer
> console.) So there are more problems than just the
> number of framebuffers on a G550.
> 
> I can't see printk's, but I see from disk lights that
> the scsi bus does the usual lengthy initialization.
> The kernel hangs shortly after that.  I know it hangs
> because the keyboards stop reacting to numlock/capslock.
> (No LED reaction, and no sysrq either.)

thats with tha patch for single fbcon oopses in swapper

the shorter is without fbcon=..., the longer with fbcon=vt:toc.8 (i'mdumb,
am i not :( )


-- 
NEU FÜR ALLE - GMX MediaCenter - für Fotos, Musik, Dateien...
Fotoalbum, File Sharing, MMS, Multimedia-Gruß, GMX FotoService

Jetzt kostenlos anmelden unter http://www.gmx.net

+++ GMX - die erste Adresse für Mail, Message, More! +++

Re: 2.6.0-test6 ruby success with matrox G550, and some questions

[Helge H. <hel...@ai...>]

On Tue, Oct 28, 2003 at 02:03:58PM +0200, Aivils Stoss wrote:
> 
> >> Current ruby support straight in kernel compiled fb device
> >> drivers and fbcon. To test Your matrox figure fbcon into
> >> kernel and reboot.
> >2.6.0-test6-ruby dies before the first printk if I do that,
> >while 2.6.0-test6 works.  fbcon on matrox breaks somehow
> >with ruby.  I don't know what to do to debug this
> >further - adding printk's is probably not an option.
> 
> Strange. Normal is 5-10 sec VGA console output until
> fb_console_init is called. After fb_console_init You
> can read printk only on serial console terminal in case
> fbcon is broken.
> 
You are right, there is a few lines of VGA console
output.  I only see half a second if I'm lucky,
because the flat panel screen takes a long time
to switch modes.

2.6test6-ruby with fbcon and 2.6test9-ryby with fbcon
are the same - they die when switching away from vga.
The screen goes black/garbage and then its over.

2.6test6-ruby and 2.6test9-ruby without fbcon
works fine, except that I have no text console.
All I can do if I quit X is to type
blind while looking at a garbage framebuffer.
The garbage change as I type, indicating
prograss but not all that helpful.

2.6test9-ruby with the single console patch
seemed to crash, although it got a little
further.  It managed to corrupt filesystems,
so I believe it got as far as mounting while
the screen was showing garbage.  The keyboard
didn't react to numlock though, so I reset it.
Perhaps I try it again.

Helge Hafting

Re: 2.6.0-test6 ruby success with matrox G550, and some questions

[Helge H. <hel...@ai...>]

I tried booting ruby-2.6-test9 with the single console patch
and framebuffer console again.

ruby without the patch hangs as soon as it switch away
from vga.

runy with the patch gets further.  The framebuffer shows
only garbage, similar to ruby with framebuffer but without
fbcon.  (Plain 2.6.0-test9 shows text on the framebuffer
console.) So there are more problems than just the
number of framebuffers on a G550.

I can't see printk's, but I see from disk lights that
the scsi bus does the usual lengthy initialization.
The kernel hangs shortly after that.  I know it hangs
because the keyboards stop reacting to numlock/capslock.
(No LED reaction, and no sysrq either.)

Helge Hafting