keyboard.c fix in 2.4.21 breaks my (buggy?) keyboard

keyboard.c fix in 2.4.21 breaks my (buggy?) keyboard

[Andreas U. Trottmann]

Hello linux-kernel,

patch-2.4.21 contains the following patch to keyboard.c:

diff -urN linux-2.4.20/drivers/char/keyboard.c linux-2.4.21/drivers/char/keyboard.c
--- linux-2.4.20/drivers/char/keyboard.c        2002-08-02 17:39:43.000000000 -0700
+++ linux-2.4.21/drivers/char/keyboard.c        2003-06-13 07:51:33.000000000 -0700
@@ -213,7 +215,17 @@
        }
        kbd = kbd_table + fg_console;
        if ((raw_mode = (kbd->kbdmode == VC_RAW))) {
-               put_queue(scancode | up_flag);
+               /*
+                *      The following is a workaround for hardware
+                *      which sometimes send the key release event twice
+                */
+               unsigned char next_scancode = scancode|up_flag;
+               if (up_flag && next_scancode==prev_scancode) {
+                       /* unexpected 2nd release event */
+               } else {
+                       prev_scancode=next_scancode;
+                       put_queue(next_scancode);
+               }
                /* we do not return yet, because we want to maintain
                   the key_down array, so that we have the correct
                   values when finishing RAW mode or when changing VT's */


But apparently, there is not only hardware that sends the key release
event twice, but also hardware that sends the key press event twice 
in certain circumstances.

I've got a keyboard ("Unikey model KWD-205"), which, for the left
control key (and apparently only for this key) sends the key down
event twice, followed by two key up events. For certain applications 
using VC_RAW (most important XFree86), this means that the ctrl
key gets "stuck" once it's been pressed.


I don't know an "appropriate" fix to this problem. Obviously, the
patch got inserted in 2.4.21 because it fixes broken keyboards
owned by many people. As I didn't find any reports about my
problem with google, it seems that keyboards like my one are
quite rare. One possible solution would be to ignore any events
where next_scancode==prev_scancode, regardless of up_flag - but
this breaks autorepeat, which then would maybe need to be
emulated.

Another possible fix would be to only ignore multiple identical
"up" events for non-modifier keys, but this would require
keyboard.c to have a list of key symbols of keys like "control",
"shift" etc.


So, are there other people having the same problem? Is it worth
to find a "good" fix for the linux kernel, or should I just
remove this patch by hand every time I compile a new kernel,
until my keyboard dies and I get a non-broken one? ;-)

-- 
Andreas Trottmann
Ideen Werft22 GmbH
Tel    +41 (0)56 210 91 37
Fax    +41 (0)56 210 91 34
Mobile +41 (0)79 229 88 55

Re: keyboard.c fix in 2.4.21 breaks my (buggy?) keyboard

[Catalin BOIE]

Hi!

I have this problem with left shift key.
X 4.2.0, 2.4.21, epox motherboard, several keyboards.
But this happen from 6 to 6 hours. And seems to last 10-20 seconds.


> Hello linux-kernel,
>
> patch-2.4.21 contains the following patch to keyboard.c:
>
> diff -urN linux-2.4.20/drivers/char/keyboard.c linux-2.4.21/drivers/char/keyboard.c
> --- linux-2.4.20/drivers/char/keyboard.c        2002-08-02 17:39:43.000000000 -0700
> +++ linux-2.4.21/drivers/char/keyboard.c        2003-06-13 07:51:33.000000000 -0700
> @@ -213,7 +215,17 @@
>         }
>         kbd = kbd_table + fg_console;
>         if ((raw_mode = (kbd->kbdmode == VC_RAW))) {
> -               put_queue(scancode | up_flag);
> +               /*
> +                *      The following is a workaround for hardware
> +                *      which sometimes send the key release event twice
> +                */
> +               unsigned char next_scancode = scancode|up_flag;
> +               if (up_flag && next_scancode==prev_scancode) {
> +                       /* unexpected 2nd release event */
> +               } else {
> +                       prev_scancode=next_scancode;
> +                       put_queue(next_scancode);
> +               }
>                 /* we do not return yet, because we want to maintain
>                    the key_down array, so that we have the correct
>                    values when finishing RAW mode or when changing VT's */
>
>
> But apparently, there is not only hardware that sends the key release
> event twice, but also hardware that sends the key press event twice
> in certain circumstances.
>
> I've got a keyboard ("Unikey model KWD-205"), which, for the left
> control key (and apparently only for this key) sends the key down
> event twice, followed by two key up events. For certain applications
> using VC_RAW (most important XFree86), this means that the ctrl
> key gets "stuck" once it's been pressed.
>
>
> I don't know an "appropriate" fix to this problem. Obviously, the
> patch got inserted in 2.4.21 because it fixes broken keyboards
> owned by many people. As I didn't find any reports about my
> problem with google, it seems that keyboards like my one are
> quite rare. One possible solution would be to ignore any events
> where next_scancode==prev_scancode, regardless of up_flag - but
> this breaks autorepeat, which then would maybe need to be
> emulated.
>
> Another possible fix would be to only ignore multiple identical
> "up" events for non-modifier keys, but this would require
> keyboard.c to have a list of key symbols of keys like "control",
> "shift" etc.
>
>
> So, are there other people having the same problem? Is it worth
> to find a "good" fix for the linux kernel, or should I just
> remove this patch by hand every time I compile a new kernel,
> until my keyboard dies and I get a non-broken one? ;-)
>
> --
> Andreas Trottmann
> Ideen Werft22 GmbH
> Tel    +41 (0)56 210 91 37
> Fax    +41 (0)56 210 91 34
> Mobile +41 (0)79 229 88 55
> -
> To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
> the body of a message to majordomo@vger.kernel.org
> More majordomo info at  http://vger.kernel.org/majordomo-info.html
> Please read the FAQ at  http://www.tux.org/lkml/
>

---
Catalin(ux) BOIE
catab@deuroconsult.ro