Hello,
I figured out in which kernel the problem I was having appeared. It seems to
work with 2.5.50 and the problems starts with 2.5.51.
I've looked at the changelog and showed these changes that might have
contibuted to my problem;
-There are some changes in fbcon. Made modular along with some fixes and some
vesafb fixes.
-Accel wrapper is now intergarted into fbcon.c. VESA fb fixes
-Made fbcon modular.
-New NVIDIA and Radeon cards pci ids. Soon I will add support for these :-)
Also a needed fix for fbcon.c.
-Moved AGP and DRM code back to drivers/char until a proper solution is done
for handling AGP/DMA based framebuffer devices.
Almost all changes by the same developer <jsimmons>. I've been looking
at the source differences but the introduction of the fbcon module really
complicates the matter.
I was wondering where to go about reporting the
problem or where this issue might stem from. (i.e. the includes shown in
vesafb.h, vesafb.h itself, fb.h or the fbcon module related code.
Or would it be best to just mail the developer? :)
Any help is appreciated.
Regards,
P.S.: Here is the part of my original mail describing the problem. It has been
confirmed by two other submits to this mailinglist regarding the 2.6.0-test4
kernel. I'm using the 2.6.0.-bk3 kernel myself.
>The most apparent occurrence of the problem for me is when
>running mc (Midnight Commander). The blue blackground isn't drawn in the
>parts where there is no text. When using mcedit this comes all the more clear
>and ?particularly when scrolling is done. The space which hasn't got any text
>is replaced with RGB pixels (looks grey from a distance). If you jump back
>from one console and back again the screen is redrawn correctly.
Hi,
I'm mailing to inform you of a possible bug in the vesa framebuffer module
which has been appearing in the kernel since 2.5.51. I posted at the
[email protected] mailinglist but wasn't able to get feedback so
I'm directing it directly to the developers of the framebuffer layer and to
you specifically since you were involved with most of the changes in 2.5.51.
I have attached grabbed framebuffer distortion pics (png) since I think those
speak clearer than my previous attempts to put the problem into words through
the mailing list.
Hope you can shed more light on the situation.
Thanks,
S Ait-Kassi
On Friday 02 January 2004 14:22, S Ait-Kassi wrote:
> Hello,
>
> I figured out in which kernel the problem I was having appeared. It seems
> to work with 2.5.50 and the problems starts with 2.5.51.
>
> I've looked at the changelog and showed these changes that might have
> contibuted to my problem;
>
> -There are some changes in fbcon. Made modular along with some fixes and
> some vesafb fixes.
>
> -Accel wrapper is now intergarted into fbcon.c. VESA fb fixes
>
> -Made fbcon modular.
>
> -New NVIDIA and Radeon cards pci ids. Soon I will add support for these :-)
> Also a needed fix for fbcon.c.
>
> -Moved AGP and DRM code back to drivers/char until a proper solution is
> done for handling AGP/DMA based framebuffer devices.
>
> Almost all changes by the same developer <jsimmons>. I've been looking
> at the source differences but the introduction of the fbcon module really
> complicates the matter.
>
> I was wondering where to go about reporting the
> problem or where this issue might stem from. (i.e. the includes shown in
> vesafb.h, vesafb.h itself, fb.h or the fbcon module related code.
>
> Or would it be best to just mail the developer? :)
> Any help is appreciated.
>
> Regards,
>
> P.S.: Here is the part of my original mail describing the problem. It has
> been confirmed by two other submits to this mailinglist regarding the
> 2.6.0-test4 kernel. I'm using the 2.6.0.-bk3 kernel myself.
>
> >The most apparent occurrence of the problem for me is when
> >running mc (Midnight Commander). The blue blackground isn't drawn in the
> >parts where there is no text. When using mcedit this comes all the more
> > clear and ?particularly when scrolling is done. The space which hasn't
> > got any text is replaced with RGB pixels (looks grey from a distance). If
> > you jump back from one console and back again the screen is redrawn
> > correctly.
>
> -
> To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
> the body of a message to [email protected]
> More majordomo info at http://vger.kernel.org/majordomo-info.html
> Please read the FAQ at http://www.tux.org/lkml/
On Sun, 04 Jan 2004 01:20:09 +0100
S Ait-Kassi <[email protected]> wrote:
> Hi,
>
> I'm mailing to inform you of a possible bug in the vesa framebuffer
> module which has been appearing in the kernel since 2.5.51. I posted at
> the [email protected] mailinglist but wasn't able to get
> feedback so I'm directing it directly to the developers of the
> framebuffer layer and to you specifically since you were involved with
> most of the changes in 2.5.51.
>
> I have attached grabbed framebuffer distortion pics (png) since I think
> those speak clearer than my previous attempts to put the problem into
> words through the mailing list.
After starting tuxcard or the program "fracplanet" (creates a 3d model of
a planet) and possibly others, my screen looks exactly the same as you can
see on the screen shots of S Ait-Kassi. Beside this, random pixels still
appearing.
My Hardware:
CPU: AMD 2800+
Graphic card: Radeon 7000 (with 64 Mb RAM)
Motherboard: MSI KT4V
Chipset: KT400
PS: I got several tips already and they made it a little bit better, but
it's still not ok.
--
Andreas Theofilu
http://www.TheosSoft.net/
--==| Enjoy the science of Linux! |==--
> I'm directing it directly to the developers of the framebuffer layer and to
> you specifically since you were involved with most of the changes in 2.5.51.
>
> I have attached grabbed framebuffer distortion pics (png) since I think those
> speak clearer than my previous attempts to put the problem into words through
> the mailing list.
Please try my latest patch. I tested midnight commander on my system and
my system is okay. This is using the vesa framebuffer.
http://phoenix.infradead.org/~jsimmons/fbdev.diff.gz
On Monday 05 January 2004 23:13, you wrote:
> Please try my latest patch. I tested midnight commander on my system and
> my system is okay. This is using the vesa framebuffer.
>
> http://phoenix.infradead.org/~jsimmons/fbdev.diff.gz
Just tried the patch with the 2.6.0 kernel and it didn't help. As a bonus
"feature" the cursor also makes the character it is underlining blink.
I assume that something broke in 2.5.51 on some cards (maybe while making
fb-con modular?). Three of the four people who stated they had this problem
had a ATI card (two had a Radeon 7500 and I have a Radeon 9500 and one didn't
include the videocard). :(
I have no idea what change could be the cause.
If I could help you with more info please say so.
Thanks again,
S Ait-Kassi
> > Please try my latest patch. I tested midnight commander on my system and
> > my system is okay. This is using the vesa framebuffer.
> >
> > http://phoenix.infradead.org/~jsimmons/fbdev.diff.gz
>
> Just tried the patch with the 2.6.0 kernel and it didn't help. As a bonus
> "feature" the cursor also makes the character it is underlining blink.
The cursor is busted but I think the latest patch from Pavel might fix it.
> I assume that something broke in 2.5.51 on some cards (maybe while making
> fb-con modular?). Three of the four people who stated they had this problem
> had a ATI card (two had a Radeon 7500 and I have a Radeon 9500 and one didn't
> include the videocard). :(
What is your config?
On Wednesday 07 January 2004 00:06, James Simmons wrote:
> > Just tried the patch with the 2.6.0 kernel and it didn't help. As a bonus
> > "feature" the cursor also makes the character it is underlining blink.
>
> The cursor is busted but I think the latest patch from Pavel might fix it.
>
> > I assume that something broke in 2.5.51 on some cards (maybe while making
> > fb-con modular?). Three of the four people who stated they had this
> > problem had a ATI card (two had a Radeon 7500 and I have a Radeon 9500
> > and one didn't include the videocard). :(
>
> What is your config?
That is my kernel config? I attached it to this mail.
That is a normal config. I have the same thing. Try my latest patch. I did
a up data again. This time the patch is against 2.6.0-rc3.
P.S
Have tried the Radeon fbdev driver ?
On Thu, 8 Jan 2004, S Ait-Kassi wrote:
> > > fb-con modular?). Three of the four people who stated they had this
> > > problem had a ATI card (two had a Radeon 7500 and I have a Radeon 9500
> > > and one didn't include the videocard). :(
> >
> > What is your config?
>
> That is my kernel config? I attached it to this mail.
Hi!
> > Just tried the patch with the 2.6.0 kernel and it didn't help. As a bonus
> > "feature" the cursor also makes the character it is underlining blink.
>
> The cursor is busted but I think the latest patch from Pavel might fix it.
Unless he is using softcursor (by issuing appropriate escape
sequence), that is unlikely.
Pavel