Re: [PATCH] hid Logitech G13 Driver 0.0.5

["Rick L. Vinyard, Jr." <rvinyard@cs.nmsu.edu>]

> On Tue, 22 June 2010 "Rick L. Vinyard, Jr." <rvinyard@cs.nmsu.edu> wrote:
>> I've been waiting to resubmit until I get some of the userspace
>> libraries
>> and utilities a little more mature. There were some things I was
>> exposing
>> to sysfs that I didn't absolutely need, and some others that I needed to
>> add.
>>
>> The cairo and papyrus libraries are (preliminarily) fully working. Once
>> I
>> make sure I have everything the daemon will need I'll resubmit the
>> driver,
>> and once that is done I was planning on submitting the libs/utils to
>> Fedora.
>>
>> For now, I've been dropping Fedora 12/13 patches and kernel SRPMS here:
>> http://miskatonic.cs.nmsu.edu/g13/
>>
>> And the userspace libraries/utilities are at:
>> http://g13.sourceforge.net
>>
>> As for the g13fb on the G15... I would imagine it could be done, but I
>> don't have a G15 so I can't say for sure.
>>
>> If I had to guess I think you'll have to take the core functionality of
>> the keyboard driver and extend it with the fb code in the _same_ driver
>> if
>> the keyboard and display use a similar design to the G13.
>>
>> The G13 uses the same endpoint for both keyboard input and display
>> messages, and since a USB endpoint can't be shared across drivers both
>> will have to reside in the same driver.
>>
> Bruno Prémont wrote:
> Hi Rick,
>
> Please check how your driver behaves if the framebuffer is still in use
> (open and/or mmapped) by some userspace process when you unplug the
> keyboard.
>
> Probably you will get hit be the same complaints from kernel I was seeing
> with PicoLCD driver (see my patch http://lkml.org/lkml/2010/5/30/15 that
> is expected to fix it on PicoLCD side, the refcounting part of the patch)
>

Will do. Thanks for the heads up.