* vgacon regression
@ 2005-12-20 20:03 Jesper Juhl
2005-12-20 21:46 ` Antonino A. Daplas
0 siblings, 1 reply; 2+ messages in thread
From: Jesper Juhl @ 2005-12-20 20:03 UTC (permalink / raw)
To: Antonino A. Daplas
Cc: Andrew Morton, Linus Torvalds, Linux Kernel Mailing List
Hi Antonino,
[starting a new thread and adding Linus to CC since this problem is
now also in his tree as well as Andrews]
The problem I observed in 2.6.15-rc5-mm2 with the (vgacon) console
getting corrupted after starting X is still present in 2.6.15-rc5-mm3
and is now also present in mainline (2.6.15-rc6-git1).
Latest kernel that I know of that is problem free is 2.6.15-rc5-git4.
The problem in a nutshell:
If I boot with vga=791 everything is fine with both mainline kernels
and -mm and
I don't see any of the problems noted below.
If I boot with vga=normal then I see these two problems :
1) Just after starts and switches resolution but before the login
manager is
displayed I briefly see a corrupted graphical image that is
sometimes what
looks like a garbled snapshot of the text-mode console and sometimes a
garbled version of the X login screen or a previous KDE desktop session.
2) If I switch back to a text mode console with ctrl+alt+f6 from
within X (or
even shut down X), then the text console is completely
garbled. It still has
the resolution it had at boot but it's displaying lots of
coloured (some of
them blinking) boxes and random characters all over, including
some bits
that look like partial boot-up messages (half a word here, two
words there,
with random bits inserted, etc) - in short, it's completely messed up.
The console stil remains functional in the sense that you can
log in blind and
run commands just fine, but no amount of "reset" or "clear"
will fix the
breakage.
Any advice as to what patches I can try to revert?
While I wait for a reply I'll test the mainline kernels between
2.6.15-rc5-git4 and 2.6.15-rc6-git1 to try and determine the first one
that breaks.
--
Jesper Juhl <jesper.juhl@gmail.com>
Don't top-post http://www.catb.org/~esr/jargon/html/T/top-post.html
Plain text mails only, please http://www.expita.com/nomime.html
^ permalink raw reply [flat|nested] 2+ messages in thread
* Re: vgacon regression
2005-12-20 20:03 vgacon regression Jesper Juhl
@ 2005-12-20 21:46 ` Antonino A. Daplas
0 siblings, 0 replies; 2+ messages in thread
From: Antonino A. Daplas @ 2005-12-20 21:46 UTC (permalink / raw)
To: Jesper Juhl; +Cc: Andrew Morton, Linus Torvalds, Linux Kernel Mailing List
Jesper Juhl wrote:
> Hi Antonino,
>
> Any advice as to what patches I can try to revert?
>
> While I wait for a reply I'll test the mainline kernels between
> 2.6.15-rc5-git4 and 2.6.15-rc6-git1 to try and determine the first one
> that breaks.
More info. The vgacon is actually restored by X as vgacon does not
touch the hardware anymore (unless you resize the console with stty). So
any patch that affects X is a culprit. More specifically, the X driver
used was vesa, which means that the vgacon restoration would be done
using VBE state save and state restore function calls. With X, this would
involve lrmi (the same library used by vbetool).
I wasn't able to duplicate this problem, using Xorg vesa and vgacon 80x25.
Also, there are no changes to vgacon or to the console that I know of between
the versions he mentioned.
To Jesper,
>From your description, I would presume that the VGA fontmap was not
adequately restored. Can you try changing your system font using whatever
utility is available for your system, such as setfont? Choose any 8x16 font
in the consolefont directory (It's usually in /usr/share/kbd/consolefont)
Tony
PS: I'll be offline from the 22nd of December to early January 2006. I
_might_ be able to occasionally answer by e-mail.
^ permalink raw reply [flat|nested] 2+ messages in thread
end of thread, other threads:[~2005-12-20 21:47 UTC | newest]
Thread overview: 2+ messages (download: mbox.gz follow: Atom feed
-- links below jump to the message on this page --
2005-12-20 20:03 vgacon regression Jesper Juhl
2005-12-20 21:46 ` Antonino A. Daplas
This is an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.