How do I make Xglamo read directly from the touchscreen?

Nelson Castillo nelsoneci at
Thu Jan 8 23:32:34 CET 2009

On Thu, Jan 8, 2009 at 3:59 PM, Holger Freyther <zecke at> wrote:
> On Thursday 08 January 2009 20:38:42 you wrote:
>> | have and maintain code that is doing that.
>> The implementation already exists.
> Cool what prevents the kernel team from presenting the patches upstream[1] and
> asking for comments?

I thought that we needed to have a small mach-gta02.c that works in mainline.
Right now this file is quite big and we have some chicken and egg problems.
Thus we might have to ask for feedback about the filters, even with no
mach-gta02.c in mainline.

> I'm puzzled on the part that you have done 4x to 5x times
> the work to create a generic framework (instead of just fixing the issue for
> our GTA02 customers) but seem to stop there.

Mmm. I do not really get it... drivers usually have filtering in the
same driver code.
We moved filtering outside the driver code so that you can use more filters
without making the driver ugly.
The filters send feedback to the interrupt function, something that
you cannot do from user-space right now.

I fail to see your point :-(

The last time I wrote that I saw a problem in porting the filters to
tslib because
of blocking calls. I thought about it and it might be solved with a
thread. Even in
that case, you cannot send feedback to the interrupt function. Not big deal,
just ignore a bunch of points instead of discarding the bugus click.
Sounds weird in userspace :-/

> What is the plan regarding the
> patches? Just keep them for ourselves?

No. We even wrote a document that explains the filters and asked for feedback
a few days ago. It has a few videos that explain what the filters do.
(the videos are not pretty but they show our point).
I hope the document will be useful for sending the patches upstream.

The first plan to send the patches upstream. If they get rejected I guess
we will get some feedback. When we get the feedback we'll decide what to do
next. Possible scenarios I can think of now:

1) Patches are accepted.
2) We get feedback and the hope that patches can be accepted
3) Patches are rejected and we are told that we are not cool people

If (3) we might just keep the patches to ourselves or port them to tslib.

I feel you're trying to tell us there is a single path to solving the
problem and that we
are failing to follow it...

More information about the devel mailing list