Re: bug #13249 - problem with intelfb on i845gm

Re: bug #13249 - problem with intelfb on i845gm

[Krzysztof Helt]

On Mon, 18 May 2009 12:48:18 +0200 (CEST)
Werner LEMBERG <wl@gnu.org> wrote:

> 
> > Could you send to me (or attach to the bugzilla report) output of
> > the dmesg command on the system with the bug?
> 
> Done.
> 

What I see here is that for 2.6.27 you use the VESA framebuffer
while you do not use it for the 2.6.28

Compare your configs:
2.6.27 : CONFIG_FB_VESA=y
2.6.28: # CONFIG_FB_VESA is not set

and messages from the 2.6.27:
<6>vesafb: framebuffer at 0xd0000000, mapped to 0xf8880000, using 5120k, total 7872k
<6>vesafb: mode is 1280x1024x16, linelength=2560, pages=2
<6>vesafb: scrolling: redraw
<6>vesafb: Truecolor: size=0:5:6:5, shift=0:11:5:0
<4>Console: switching to colour frame buffer device 160x64
<6>fb0: VESA VGA frame buffer device

and nothing like that from the 2.6.28.

My advice is try using the Intel framebuffer (change it to Y not 'M' like module). It should
be faster than VESA on your hardware. Unselect VESA. 

Please report if it works for you after the config change.

Kind regards,
Krzysztof


----------------------------------------------------------------------
Oryginalne przepisy na grilla. Zaskocz swoich gosci!
Sprawdz >>> http://link.interia.pl/f217c





------------------------------------------------------------------------------
Crystal Reports - New Free Runtime and 30 Day Trial
Check out the new simplified licensing option that enables 
unlimited royalty-free distribution of the report engine 
for externally facing server and web deployment. 
http://p.sf.net/sfu/businessobjects

Re: bug #13249 - problem with intelfb on i845gm

[Werner LEMBERG]

> What I see here is that for 2.6.27 you use the VESA framebuffer
> while you do not use it for the 2.6.28
> 
> Compare your configs:
> 2.6.27 : CONFIG_FB_VESA=y
> 2.6.28: # CONFIG_FB_VESA is not set

This is intentional.  Reason is that in 2.6.28 option
`CONFIG_FB_BOOT_VESA_SUPPORT' has been added, and it seems that this
change introduced a bug.

> My advice is try using the Intel framebuffer (change it to Y not 'M'
> like module).  It should be faster than VESA on your hardware.
> Unselect VESA.

I'm doing a git bisect right now to find out why
CONFIG_FB_BOOT_VESA_SUPPORT doesn't work as advertised.


    Werner

------------------------------------------------------------------------------
Crystal Reports - New Free Runtime and 30 Day Trial
Check out the new simplified licensing option that enables 
unlimited royalty-free distribution of the report engine 
for externally facing server and web deployment. 
http://p.sf.net/sfu/businessobjects

Re: bug #13249 - problem with intelfb on i845gm

[Krzysztof Helt]

On Mon, 18 May 2009 15:29:03 +0200 (CEST)
Werner LEMBERG <wl@gnu.org> wrote:

> 
> > What I see here is that for 2.6.27 you use the VESA framebuffer
> > while you do not use it for the 2.6.28
> > 
> > Compare your configs:
> > 2.6.27 : CONFIG_FB_VESA=y
> > 2.6.28: # CONFIG_FB_VESA is not set
> 
> This is intentional.  Reason is that in 2.6.28 option
> `CONFIG_FB_BOOT_VESA_SUPPORT' has been added, and it seems that this
> change introduced a bug.
> 
> > My advice is try using the Intel framebuffer (change it to Y not 'M'
> > like module).  It should be faster than VESA on your hardware.
> > Unselect VESA.
> 
> I'm doing a git bisect right now to find out why
> CONFIG_FB_BOOT_VESA_SUPPORT doesn't work as advertised.
> 
> 

Probably, it does work as supposed to. The CONFIG_FB_BOOT_VESA_SUPPORT
tells Linux to allow VESA modes values for the "vga=" switch. It also switches
on the VESA graphics modes. However, without a framebuffer which makes 
use of the graphics display (vesafb, intelfb, sisfb) it only garbles screen.

Regards,
Krzysztof

----------------------------------------------------------------------
Fantastyczne nagrody do zgarniecia!
Zagraj >> http://link.interia.pl/f2177 



------------------------------------------------------------------------------
Crystal Reports - New Free Runtime and 30 Day Trial
Check out the new simplified licensing option that enables 
unlimited royalty-free distribution of the report engine 
for externally facing server and web deployment. 
http://p.sf.net/sfu/businessobjects

Re: bug #13249 - problem with intelfb on i845gm

[Werner LEMBERG]

>> I'm doing a git bisect right now to find out why
>> CONFIG_FB_BOOT_VESA_SUPPORT doesn't work as advertised.
> 
> Probably, it does work as supposed to.  The
> CONFIG_FB_BOOT_VESA_SUPPORT tells Linux to allow VESA modes values
> for the "vga=" switch.  It also switches on the VESA graphics modes.
> However, without a framebuffer which makes use of the graphics
> display (vesafb, intelfb, sisfb) it only garbles screen.

Well, I use the intelfb driver (CONFIG_FB_INTEL=m), but I get a black
screen for any VESA mode...


    Werner

------------------------------------------------------------------------------
Crystal Reports - New Free Runtime and 30 Day Trial
Check out the new simplified licensing option that enables 
unlimited royalty-free distribution of the report engine 
for externally facing server and web deployment. 
http://p.sf.net/sfu/businessobjects

Re: bug #13249 - problem with intelfb on i845gm

[Krzysztof Helt]

On Wed, 20 May 2009 08:25:48 +0200 (CEST)
Werner LEMBERG <wl@gnu.org> wrote:

> >> I'm doing a git bisect right now to find out why
> >> CONFIG_FB_BOOT_VESA_SUPPORT doesn't work as advertised.
> > 
> > Probably, it does work as supposed to.  The
> > CONFIG_FB_BOOT_VESA_SUPPORT tells Linux to allow VESA modes values
> > for the "vga=" switch.  It also switches on the VESA graphics modes.
> > However, without a framebuffer which makes use of the graphics
> > display (vesafb, intelfb, sisfb) it only garbles screen.
> 
> Well, I use the intelfb driver (CONFIG_FB_INTEL=m), but I get a black
> screen for any VESA mode...
> 

You have to compile it into the kernel image (CONFIG_FB_INTEL=y) 
otherwise no VESA mode is used.

Regards,
Krzysztof

----------------------------------------------------------------------
Zrob sobie prezent. Wygraj nagrode!
Sprawdz >>  http://link.interia.pl/f2176 



------------------------------------------------------------------------------
Crystal Reports - New Free Runtime and 30 Day Trial
Check out the new simplified licensing option that enables 
unlimited royalty-free distribution of the report engine 
for externally facing server and web deployment. 
http://p.sf.net/sfu/businessobjects

Re: bug #13249 - problem with intelfb on i845gm

[Werner LEMBERG]

>> Well, I use the intelfb driver (CONFIG_FB_INTEL=m), but I get a black
>> screen for any VESA mode...
> 
> You have to compile it into the kernel image (CONFIG_FB_INTEL=y) 
> otherwise no VESA mode is used.

Aaah.  Will try soon, thanks.


   Werner

------------------------------------------------------------------------------
Crystal Reports - New Free Runtime and 30 Day Trial
Check out the new simplified licensing option that enables 
unlimited royalty-free distribution of the report engine 
for externally facing server and web deployment. 
http://p.sf.net/sfu/businessobjects

Re: bug #13249 - problem with intelfb on i845gm

[Krzysztof Helt]

On Wed, 20 May 2009 08:25:48 +0200 (CEST)
Werner LEMBERG <wl@gnu.org> wrote:

> >> I'm doing a git bisect right now to find out why
> >> CONFIG_FB_BOOT_VESA_SUPPORT doesn't work as advertised.
> > 
> > Probably, it does work as supposed to.  The
> > CONFIG_FB_BOOT_VESA_SUPPORT tells Linux to allow VESA modes values
> > for the "vga=" switch.  It also switches on the VESA graphics modes.
> > However, without a framebuffer which makes use of the graphics
> > display (vesafb, intelfb, sisfb) it only garbles screen.
> 
> Well, I use the intelfb driver (CONFIG_FB_INTEL=m), but I get a black
> screen for any VESA mode...
> 
> 

I see a problem here. If the intelfb is selected as a module the CONFIG_FB_BOOT_VESA_SUPPORT  should not be selected.
I'll prepare a patch for it.

Regards,
Krzysztof

----------------------------------------------------------------------
Fantastyczne nagrody do zgarniecia!
Zagraj >> http://link.interia.pl/f2177 



------------------------------------------------------------------------------
Crystal Reports - New Free Runtime and 30 Day Trial
Check out the new simplified licensing option that enables 
unlimited royalty-free distribution of the report engine 
for externally facing server and web deployment. 
http://p.sf.net/sfu/businessobjects