[Libdlo] displaylink-mod status for kernel 2.6.18

[Bernie Thompson]

Hi Alexander,

Thanks so much for testing and submitting the patch (and for Dan for
originally submitting it).

Although there's been some concern in other cases about inappropriate
use of vmalloc ... from my look at it, if we're going to have this
shadow/backbuffer at all, vmalloc (non-contiguous physical pages) is
the right fn.  We'll see what happens when we move the patch on.

The patch has been applied at:
http://git.plugable.com/gitphp/index.php?p=udlfb&a=shortlog

I'll start to generate the corresponding patches for the various
changes there against the full kernel shortly, for submission to
staging.  Part of the idea here is we give these patches some test
time before sending them out (e.g. big endian support, which still
appears to be incomplete even after Henry's patch), so that's probably
just as well.  Any testing and bug reports on the latest on
git.plugable.com is very welcome -- every commit on "master" is
expected to build and work without regressions.  On other branches
(not master), there may be incomplete work.

For bug tracking, we don't have a perfect answer.  Probably the best
answer is bugs.freedesktop.org with the Product set to "libdlo".

Thanks!
Bernie
http://plugable.com/

Some random notes of the day:
* The ConsoleKit/GDM patch set with new multiseat support has been
resubmitted yesterday:
http://lists.freedesktop.org/archives/consolekit/2009-November/date.html
That will be the foundation of usb multiseat stuff moving forward, so
any testing/help that effort can get, will also help usbseat.
* Also yesterday, Microsoft announced their equivalent to Linux
multiseat: http://plugable.com/2009/11/12/microsoft-windows-multipoint-server-2010-announced/
so it'd be great to get the Linux usbseat stuff rolling to keep ahead
of the game.