Comment 183 for bug 561802

wow :)

i did this in the backwards manner : installed 10.10, which has the problem,
and since downgraded a kernel to 2.6.32.11 which worked fine, but no wifi...

From what I can tell from the forums, there is a kernel patch that fixed
this and that may have been unpatched since......

hoping someone finds and fixes for good.......
david

2010/11/30 CVTECH <email address hidden>

> I have access to qty3 E6410ATGs with the 1280x800 display (reports
> itself as "seiko-epson corporation 14"), only one is experiencing this
> issue and this is after update manager updating from the base Ubuntu
> 10.04.1 installation with 2.6.32-24 kernel. The machine having troubles
> is a core5-i540 vs core5-i520 on the other two machines not having
> troubles.
>
> The symptoms are very similar to post #168. The machine worked fine
> until updating a couple days ago. Included in the update was update from
> kernel 2.6.32-24 to 2.6.32-26 but I suspect this is not the issue for
> reasons I will explain.
>
> I am experiencing similar issues to post #168. At login screen after
> updates including the 2.6.32-26 kernel the blank screen happens.
> Plugging in any external monitor (on Dsub 15 pin, or DisplayPort)
> instantly the internal monitor will come to life. So the workaround is
> to pick the kernel from Grub, wait for the blank screen which would
> normally be the login/welcome screen on Ubuntu, quickly plug in a cable
> (there is about a 5 second window that this works), then unplug the
> cable and now you can take the laptop and walk around the house like you
> wanted to in the first place when you bought a laptop. This is very
> reliable, with about 96 of 100 reboots (hot or cold) able to start the
> internal display by plugging a monitor cable in (Dsub or Displayport,
> doesn't matter). The other 4 reboots vary from won't work (maybe took
> too long to plug) to about 1 in 100 will randomly just boot without
> needing the cable plug procedure!
>
> I note also that the monitor must be present but not required to be
> powered on, an unterminated cable does not have the same effect. So,
> something in the process sees the presence of a new external monitor
> connection and this then triggers the internal monitor to come on (at
> correct resolution) as DP1 and everything is good after that until the
> next reboot, even after disconnecting the external cable.
>
> I suspect something has changed unrelated to the kernel update. The
> reason is that the two machines I have that are working fine one is
> running 2.6.32-24, the other 2.6.32-25. The machine having problems was
> working fine on 2.6.32-24 and started having issues immediately after
> the update (via update manager) that included 2.6.32-26 amongst many
> other things. I removed (via package manager) the 2.6.32-26 kernel image
> and headers and installed the 2.6.32-25. This did not solve the issue
> and now even if I boot (from grub) the old previously working 2.6.32-24
> or the 2.6.32-25 kernel that works on another machine the same issue
> happens on the problem machine, even in recovery mode!
>
> I am reasonably familiar with Ubuntu as a user, but no idea how to
> gather info to track this down. If someone has an idea of a path to
> follow, then let me know how to compile logs or config files I can
> gather info from the working (2.6.32-25) vs non-working (2.6.32.-25)
> machines, only difference being i520 vs i540 processors.
>
> These are live (not development) machines so unfortunately I can't try
> too much in the way of installing source code and patches but on the
> third machine if we can come up with a way to replicate the problem, it
> can be a total sandbox to test anything out.
>
> I note also that this is the "Daylight Viewable" E6410 ATG display which
> has a max resolution of 1280x800 and when working is correctly
> identified by xrandr on DP1. Dell makes the E6410 with at least 1 other
> display.
>
> Also intel graphics gma4500 (i915) on all machines and kernels noted are
> the "-generic" versions.
>
> --
> [i915] blank screen on Latitude E6410 (NOT E6510)
> https://bugs.launchpad.net/bugs/561802
> You received this bug notification because you are a direct subscriber
> of the bug.
>