Messages in this thread | | | Date | Mon, 04 Apr 2011 14:20:21 +0200 | From | Simon Budig <> | Subject | Re: [PATCH resend] hid-magicmouse: Increase evdev buffer size |
| |
-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1
On 04/02/2011 06:36 PM, Chase Douglas wrote: > On 04/01/2011 07:51 PM, Jeffrey Brown wrote: >> I've got another change in the works that fixes this problem more >> systematically. > > That would be great :). I still would like to see this patch added so > it's released with the 2.6.39 kernel and backported to earlier stable > releases. I hope that's not a problem.
Wouldn't it make sense to change the buffer logic in the input layer that in the case of a buffer overrun the head gets pushed forward to the next SYN_REPORT?
One symptom I witnessed with the bug was, that I got partial reports, which especially in the case of a multitouch report (A-Type) creates false release-events. This would be solved by the head-forwarding.
(one could even think about more fancy stuff like combining relative events and keeping track of the absolute events, but this seems a bit like overkill)
Thanks, Simon - -- Simon Budig kernel concepts GbR simon.budig@kernelconcepts.de Sieghuetter Hauptweg 48 +49-271-771091-17 D-57072 Siegen
-----BEGIN PGP SIGNATURE----- Version: GnuPG v1.4.10 (GNU/Linux) Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org/
iEYEARECAAYFAk2Zt4UACgkQO2O/RXesiHByIgCgmHsujFhsYcM0wp7qEl+HytNh wSMAoLygXVkU1Lx+jNgBA++H8OL6dNVR =9+3z -----END PGP SIGNATURE-----
| |