lubuntu-qa team mailing list archive

[Erick Brunzell <lbsolost@xxxxxxxxx>]

On 09/02/2013 02:56 AM, JM wrote:
> On Mon, 02 Sep 2013 02:24:57 -0500
> Erick Brunzell <lbsolost@xxxxxxxxx> wrote:
>
>> With Beta 1 testing approaching - maybe as soon as tomorrow - I wanted
>> to let everyone know that I'm still unable to test as much as usual
>> because of this bug:
>>
>> https://bugs.launchpad.net/ubuntu/+source/xorg/+bug/1205643
>>
>> I have made a few changes so I can run some additional test-cases on my
>> other test box, but my participation will be limited since half of my
>> test-suite is down :^(
>>
>> Lance

Wanted to follow up on this a bit more since JM sent me a PM requesting
more info.

> Hi,
>
> The bug report you point to mentions several desktop environments where it does not work:
> ******
> I also want to say that I'm fully aware that graphics chip is not supported by
> Ubuntu/compiz itself, or even GNOME's mutter window manager, but I think it should still
> work with the openbox, metacity, and Xfwm window managers ...... at least I hope so :^)
> ******
>
> however, Graphics are not related to DE, or to window managers, but depend on the
> driver provided to work with Xorg.

That's not entirely true. That VIA P4M800 chip has never worked with
Compiz due to a lack of 3d rendering, so up through Precise Ubuntu would
always fall back to using the Metacity WM. The same was true of GNOME 3
until they dropped the fallback-session entirely. But Metacity is going
to be maintained by Edubuntu for their LTSP installs.

> Your Chipset is a Chrome9 chipset, which is the/one of the cheapest (and bad) out
> there.
>
> The driver for it is unichrome: 
> http://unichrome.sourceforge.net

Yes, Xorgs 'openchrome' I believe. I didn't see anything in the recent
changelog:

xserver-xorg-video-openchrome (1:0.3.1-0ubuntu2.1) saucy; urgency=low

  * Rebuild for new video api.

 -- Maarten Lankhorst <maarten.lankhorst@xxxxxxxxxx>  Thu, 13 Jun 2013
12:49:28 +0200

xserver-xorg-video-openchrome (1:0.3.1-0ubuntu2) saucy; urgency=low

  * SECURITY UPDATE: denial of service and possible code execution via
    incorrect memory size calculations
    - 68bf50ce4903ec93da59cea78e063ed7c3882d3e
    - db309e3cd87a1279e8b592a692390755c528de4f
    - CVE-2013-1994

 -- Marc Deslauriers <marc.deslauriers@xxxxxxxxxx>  Mon, 10 Jun 2013
11:25:03 -0400

And it still worked in Saucy Alpha 1 (20130626) but not Alpha 2 or beyond.

>
> I had a motherboard setup with this chipset, I just added a nVidia pci-e GPU (a cheap one
> solved the display issues I was meeting at this time).
>
> >From your bug report again:
> ******
> So onto what actually happens. When I try to boot either Lubuntu or Xubuntu Saucy Alpha
> 2, either the live image or an installed version, I just get a frozen progress bar. So
> it's just 5 frozen dots on the screen. If I try booting through the recovery mode or
> otherwise trying to "startx" or "'sudo service lightdm start' I just get a blank screen.
> I can see the backlight is working but that's all.
> ******
>
> It seems according to your report that only the progress bar is a problem to you. And you
> can't start X from the recovery mode of course, it's an init mode which is too low for
> that. 
>
> If it is only the progress bar which bothers you: the progress bar does not matter. You
> can hit a key before booting (it might be F6, look at the lower part of the screen to
> check) and replace "quiet splash" in the kernel command line with "noquiet" (leave the
> '--' at the end of the line) and you will have a text mode boot instead. Or just ignore
> what the progress bar looks like and go do something else by the time the system finishes
> the boot?
>
> Regards, Mélodie
>

One thing has changed between Alpha 2 and now. Rather than a frozen
progress bar, I now boot the live image directly to a blank screen.
Server installs seem to work OK, as does booting in "text" mode. I'll
try the "noquiet" option rather than "text" and see what happens.

Oh, one thing I tried with Alpha 2 to help rule out a kernel problem was
upgrading from Raring to Saucy and then booting the old Raring kernel
which was still a no-go. Since it effects both the live image and
installed images I believe I should be able to collect info if an Xorg
dev tells me what to do, and what to look for.

Lance