Bug#594089: keyboard-configuration: caps lock keycode problem
On Wed, Oct 27, 2010 at 05:22:42PM +0200, Samuel Thibault wrote:
> Dmitry Torokhov, le Wed 27 Oct 2010 08:06:26 -0700, a écrit :
> > On Oct 27, 2010, at 5:05 AM, Samuel Thibault <email@example.com> wrote:
> > Dmitry Torokhov, le Wed 27 Oct 2010 02:10:29 -0700, a écrit :
> > Also, based on evtest data I only see presses/releases for one key
> > (Caps Lock, Right Shift, etc.) I do not see the presses for the additional
> > keys in the evtest stream so I am baffled as to where the additional
> > scancode is coming from... Do you have some funky keymap loaded?
> > Anything interesting in dumpkeys?
> > Err, since the duplicates appear in scancodes already, is the content of
> > dumpkeys really involved?
> > Evtest data does not show any additional events, so the additional scancodes
> > must be coming from the legacy keyboard driver.
> Sure, but dumpkeys shows the table from scancodes to keycodes, not from
> input events to scancodes.
I am just wondering if some of the translation is leaking even though
the keyboard is in raw mode...
OK, scratch that... Are there any other input devices? Anything
connected via uinput and listening? What does /proc/bus/input/devices
show? Maybe if Tom could compile and load evbug module it would show us
where the additional events are coming form, becaus ethey surely are not
generated by the keyboard itself (otherwise we'd have seen them in
evtest output for that device).