Re: [PATCH 0/3] Input: synaptics - multitouch and multifinger support

[Takashi Iwai <tiwai@suse.de>]

At Sun, 10 Oct 2010 13:59:32 -0700,
Dmitry Torokhov wrote:
> 
> On Sun, Oct 10, 2010 at 09:49:45AM +0200, Henrik Rydberg wrote:
> > Takashi, Chase,
> > 
> > talk about Heinz effect. :-)
> > 
> > Rather than taking any of the patches, I was wondering if we could get a single
> > patch including all the later findings and considerations on what devices should
> > have the multitouch mode. Also, unless there is a really really good reason for
> > it, without kernel parameters. After all, these patches should only add new
> > functionality without regressions.
> 
> Yes, I think we should limit number of module options (and even more so
> Kconfig options). I believe that either MT support is ready for prime-time
> and then it should be enabled unconditionally, or it is not ready and
> should be kept as patches. his also means that we should not introduce
> regressions in ST (even when emulated in MT mode).

Right, but it's pretty hard to know whether it works really for all
machines, simply because the current MT-detection is nothing but a
pure guess work.  As mentioned in another mail, there are devices that
have no this bit but reports multi-finger detections without
tracking.  So, there are definitely other things to be checked.

Thus, I'm inclined to add a module option, at least, for *disabling*
MT for non-working machines.  Also, another question is whether it's
safe to enable MT even for user-space that don't understand MT.


Takashi