Re: Yet more display troubles with 2.6.15-rc5-mm2

["Antonino A. Daplas" <adaplas@gmail.com>]

Jesper Juhl wrote:
> On 12/12/05, Antonino A. Daplas <adaplas@gmail.com> wrote:
>> Jesper Juhl wrote:
>>> In addition to the problem I reported earlier about 2.6.15-rc5-mm2
>>> hanging at boot with vga=791 I've just discovered another problem.
>>>
>>> If I boot with vga=normal (which is aparently all that works), then I
>>> can boot up to a nice lain text login and run startx, but if I then
>>> switch away from X back to a text console with ctrl+alt+f6 or if I
>>> shut down X, then I'm presented with a completely garbled text mode
>>> screen - flashing coloured blocks all over, random bits of text at
>>> random locations etc.
>>>
>>> Also, when starting X, just before the cursor appears I normally just
>>> have a black screen. With this kernel I first get a short blink of a
>>> garbled graphics mode screeen with either what looks like just random
>>> pixels or sometimes with something that looks like a mangled snapshot
>>> of my text mode console, or if I kill X with ctrl+alt+backspace and
>>> then start it again (the garbled text mode console does work, although
>>> I'm glad I know how to touch type ;) then I sometimes get what looks
>>> like a snapshot of my previous X session with random pixels on top.
>>> The garbled graphical screen stays for just a blink of an eye, then
>>> it's replaced with the normal black screen and the mouse cursor.
>>>
>>> 2.6.15-rc5-git1 works perfectly without these issues.
>> I cannot reproduce your problem...
>>
> 
>> Can you try another X driver, ie, vesa?
>>
> I'm already using the vesa driver. It seems to be the only Open Source
> driver that'll work with this card, so i don't have any other to try.
> 

Ah, vesa Xorg driver with vgacon.  I haven't tried that yet. Let me check if
I can reproduce the problem.

>> Also, these 2 patches are present in mm but not in Linus' tree.  Can
>> you check which of these are the culprit, if any?
>>
>> http://kernel.org/pub/linux/kernel/people/akpm/patches/2.6/2.6.15-rc2/2.6.15-rc2-mm1/broken-out/vgacon-fix-doublescan-mode.patch
>> http://kernel.org/pub/linux/kernel/people/akpm/patches/2.6/2.6.15-rc2/2.6.15-rc2-mm1/broken-out/vgacon-workaround-for-resize-bug-in-some-chipsets.patch
>>
> 
> Since this is 2.6.15-rc5-mm2 I grabbed these two instead:
> http://kernel.org/pub/linux/kernel/people/akpm/patches/2.6/2.6.15-rc5/2.6.15-rc5-mm2/broken-out/vgacon-fix-doublescan-mode.patch
> http://kernel.org/pub/linux/kernel/people/akpm/patches/2.6/2.6.15-rc5/2.6.15-rc5-mm2/broken-out/vgacon-workaround-for-resize-bug-in-some-chipsets.patch
> 
> Reverting both patches didn't fix the problem. Starting X then
> switching back to a text mode console still results in a completely
> messed up text console. X is fine, I can switch back to it no problem,
> but text consoles go bye-bye...
> 
> Would there be any point in trying a kernel with just one of the
> patches reverted?

Those 2 patches are independent, so it's possible that their side
effects can cancel each other out.  So yes, try reversing one patch
at a time.

Tony