lubuntu-qa team mailing list archive

[Str8bs <str8bslinux@xxxxxxxxx>]

Thanks Lars. It is looking more and more like a frame buffer issue.
If you have the time to devote, you may want to remove quiet splash
from the append in yaboot and add
"drm.debug=0x4 log_buf_len=16M" so as not to have to type those every
reboot. On the ones we can get to a shell, save a copy of dmesg. On
the ones that x starts, whether viewable or not, drop to a shell and
save Xorg.0.log as well. Idea being to clearly make a case to the devs
once we find it and the logs to demonstrate.
See in line.

On Tue, May 14, 2013 at 2:40 PM, Lars Nooden <lars.nooden@xxxxxxxxx> wrote:
> On Tue, 14 May 2013, Str8bs wrote:
>> Does booting with: "Linux snd-powermac.pinkelephant=yes" get any
>> further before errors?
> ...
>> Can you get beyond error or further before the error with:
>> 1. Linux single
>> 2. Linux single snd-powermac.pinkelephant=yes (Or
>> snd-powermac.blacklist=yes if you don't have a warped sense of humor.
>> :) )
>> 3. Linux single nomodeset
>
> 0.  I tried pinkelephant above and it boot as far as giving me the mouse
> pointer on a strangely colored and patterned screen.  But no evidence of
> LCDE was present.
Sounds like x is starting using FBDEV even though Radeon KMS is present.
Could you ctl-alt-f1 to shell? If so, output of cat /proc/fb ?
If you add video=radeonfb:yada to pinkelephant, is desktop viewable?

>
> 1. Linux single by itself give the recursive fault.
>
> 2. Linux single with pinkelephant boots to a root shell with no X and no
> login.  So I guess that works as expected.
>
Yes

> 3. Linux single nomodeset also ends with a recursive fault.
Could you repeat and rule out framebuffer?
Linux single nomodeset video=offb:off
Linux single nomodeset video=radeonfb:off


>
> Regards,
> /Lars
>
> --
> Mailing list: https://launchpad.net/~lubuntu-qa
> Post to     : lubuntu-qa@xxxxxxxxxxxxxxxxxxx
> Unsubscribe : https://launchpad.net/~lubuntu-qa
> More help   : https://help.launchpad.net/ListHelp