On Mon, Feb 11, 2019 at 09:51:58AM +0100, Martin Kepplinger wrote: > On 11.02.19 09:45, 'dmitry.torokhov@xxxxxxxxx' wrote: > > Hi Matthias, > > > > On Mon, Feb 11, 2019 at 07:15:12AM +0000, Matthias Fend wrote: > >> Hi Dmitry, > >> > >> You asked me if I could also include the conversion to touchscreen_report_pos() - which I did in this patch version. > >> Is there anything else missing? > > > > Sorry, gmail decided to mark the v2 patches as spam for some reason :( > > I fished them out and applied. > > > > Matthias, Martin, I noticed that the driver is still using MT-A > > protocol. Could you convert it to the proper MT-B (slotted) protocol? > > > > Yes, I thought about that actually. type A protocol is what matches the > controller here... It stores touch points consecutively in memory, but I > doubt that it tracks them (by keeping them at one position in its > memory, for instance). But yes, I want to test that. We still can do MT-B for such devices by employing in-kernel contact tracking. See input_mt_assign_slots(). Thanks. -- Dmitry
