1 |
On 2009-06-30, Helmut Jarausch <jarausch@××××××××××××××××.de> wrote: |
2 |
> On 29 Jun, Grant Edwards wrote: |
3 |
|
4 |
>> After recent updates Xorg on several of my machines have |
5 |
>> started dropping keyboard events. Sometimes it's a keydown |
6 |
>> event -- in which case the keystroke is ignored. Sometimes |
7 |
>> it's a keyup event -- in which case the key repeats |
8 |
>> indefinitely until another key is pressed. |
9 |
|
10 |
[...] |
11 |
|
12 |
>> Has anybody else seen this problem pop up recently? |
13 |
> |
14 |
> Yes and No, |
15 |
> |
16 |
> I had similar problems but completely deterministic. When the |
17 |
> keyboard is handled by hal it produces different keycodes then |
18 |
> before. |
19 |
> |
20 |
> In my case the PgDn key now generates a code which was |
21 |
> previously generated by the key right to the right Windows |
22 |
> key. And in my .Xmodmap I had this code defined as Mode_switch |
23 |
> which made the key "dead" in my editor. Try using xev to see |
24 |
> which codes are generated by the keys which cause you |
25 |
> problems. |
26 |
|
27 |
AFAICT, my problem happens with equal frequency for all keys. |
28 |
|
29 |
> Then look at all places where X11 key codes are redefined i.e. |
30 |
> xmodmap is called. |
31 |
|
32 |
The problem isn't that a key produces the wrong result. The |
33 |
problem is that sometimes a key will be "missed" and sometimes |
34 |
a key will "stick" until another key is pressed. It only seems |
35 |
to happen under heavy load (e.g. when compiling programs), and |
36 |
it's now happening on multiple machines. |
37 |
|
38 |
-- |
39 |
Grant Edwards grante Yow! Make me look like |
40 |
at LINDA RONSTADT again!! |
41 |
visi.com |