PC keyboard rate/delay

PC keyboard rate/delay

[Sergey Tursanov]

[-- Attachment #1: Type: text/plain, Size: 596 bytes --]

In file include/linux/kd.h was declared KDKBDREP ioctl number
to tune up keyboard rate/delay with struct kbd_repeat.
But in 2.4.x kernel there is only m68k version for that.
I wrote some code for implement this feature on x86 machines.
Gzipped and uuencoded patch for kernel 2.4.5 is attached.
To setup keyboard rate/delay on x86 you should use code like that:

struct kbd_repeat kbd_rep={
       1000,      /* delay in ms */
       30         /* repeat rate in cps */
};
ioctl(0,KDKBDREP,&kbd_rep);

After that ioctl kbd_rep is filled with previous values.

I hope it will be useful for someone.

[-- Attachment #2: patch-kbdrate-2.4.5.gz --]
[-- Type: application/x-gzip, Size: 2355 bytes --]

Re: PC keyboard rate/delay

[Alan Cox]

> In file include/linux/kd.h was declared KDKBDREP ioctl number
> to tune up keyboard rate/delay with struct kbd_repeat.
> But in 2.4.x kernel there is only m68k version for that.
> I wrote some code for implement this feature on x86 machines.
> Gzipped and uuencoded patch for kernel 2.4.5 is attached.
> To setup keyboard rate/delay on x86 you should use code like that:
> 
> struct kbd_repeat kbd_rep={
>        1000,      /* delay in ms */
>        30         /* repeat rate in cps */
> };
> ioctl(0,KDKBDREP,&kbd_rep);
> 
> After that ioctl kbd_rep is filled with previous values.
> I hope it will be useful for someone.

You must have been reading my mind. Yesterday I traced at least one X11
hang down to the kernel and X server both frobbing with the port at the same
time and crashing the microcontroller on my PC110.

Alan

Re: PC keyboard rate/delay

[James Simmons]

> > In file include/linux/kd.h was declared KDKBDREP ioctl number
> > to tune up keyboard rate/delay with struct kbd_repeat.
> > But in 2.4.x kernel there is only m68k version for that.
> > I wrote some code for implement this feature on x86 machines.
> > Gzipped and uuencoded patch for kernel 2.4.5 is attached.
> > To setup keyboard rate/delay on x86 you should use code like that:
> > 
> > struct kbd_repeat kbd_rep={
> >        1000,      /* delay in ms */
> >        30         /* repeat rate in cps */
> > };
> > ioctl(0,KDKBDREP,&kbd_rep);
> > 
> > After that ioctl kbd_rep is filled with previous values.
> > I hope it will be useful for someone.
> 
> You must have been reading my mind. Yesterday I traced at least one X11
> hang down to the kernel and X server both frobbing with the port at the same
> time and crashing the microcontroller on my PC110.

Also the beeper is a problem. Have X and a console app set off a beep at
the same time. I'm glad the input api will make these problems go away :-)

Re[2]: PC keyboard rate/delay

[Sergey Tursanov]

AC> You must have been reading my mind. Yesterday I traced at least one X11
AC> hang down to the kernel and X server both frobbing with the port at the same
AC> time and crashing the microcontroller on my PC110.

I think it would be better to place all of kbd controller code
into the kernel instead of using various userspace programs
such as kbdrate. Otherwise why KDKBDREP was defined ?-)

Sergey Tursanov                       mailto:__gsr@mail.ru

Re: Re[2]: PC keyboard rate/delay

[Geert Uytterhoeven]

On Tue, 12 Jun 2001, Sergey Tursanov wrote:
> AC> You must have been reading my mind. Yesterday I traced at least one X11
> AC> hang down to the kernel and X server both frobbing with the port at the same
> AC> time and crashing the microcontroller on my PC110.
> 
> I think it would be better to place all of kbd controller code
> into the kernel instead of using various userspace programs
> such as kbdrate. Otherwise why KDKBDREP was defined ?-)

Because (almost?) all m68k machines don't have PC style keyboard controllers,
so we _had_ to invent some other way to implement it in a portable (across all
m68k machines) way.

<wishful thinking>
Of course it would be nice if all architectures would want to use it.
</wishful thinking>

Gr{oetje,eeting}s,

						Geert

--
Geert Uytterhoeven -- There's lots of Linux beyond ia32 -- geert@linux-m68k.org

In personal conversations with technical people, I call myself a hacker. But
when I'm talking to journalists I just say "programmer" or something like that.
							    -- Linus Torvalds

Re: Re[2]: PC keyboard rate/delay

[James Simmons]

> Because (almost?) all m68k machines don't have PC style keyboard controllers,
> so we _had_ to invent some other way to implement it in a portable (across all
> m68k machines) way.

This stuff is such a mess :-( Sparc has its own routines as well. 

> <wishful thinking>
> Of course it would be nice if all architectures would want to use it.
> </wishful thinking>

It will for 2.5.X :-)

Re: PC keyboard rate/delay

[Guest section DW]

On Tue, Jun 12, 2001 at 06:19:49PM +0400, Sergey Tursanov wrote:

> In file include/linux/kd.h was declared KDKBDREP ioctl number
> But in 2.4.x kernel there is only m68k version for that.
> I wrote some code for implement this feature on x86 machines.

Yes, it is unfortunate that many ioctls are needlessly different
between various architectures. But the utility kbdrate
works on all architectures I have tried it on.

Andries

Re: PC keyboard rate/delay

[Guest section DW]

On Tue, Jun 12, 2001 at 09:23:37PM +0400, Sergey Tursanov wrote:
> AC> You must have been reading my mind. Yesterday I traced at least one X11
> AC> hang down to the kernel and X server both frobbing with the port at the same
> AC> time and crashing the microcontroller on my PC110.
> 
> I think it would be better to place all of kbd controller code
> into the kernel instead of using various userspace programs
> such as kbdrate. Otherwise why KDKBDREP was defined ?-)

KDKBDREP on m68k
KIOCSRATE on sparc
/dev/port on i386 etc.

Yes, we want to get rid of /dev/port.

Re: Re[2]: PC keyboard rate/delay

[Geert Uytterhoeven]

On Tue, 12 Jun 2001, James Simmons wrote:
> > Because (almost?) all m68k machines don't have PC style keyboard controllers,
> > so we _had_ to invent some other way to implement it in a portable (across all
> > m68k machines) way.
> 
> This stuff is such a mess :-( Sparc has its own routines as well. 

Ours is older :-)

> > <wishful thinking>
> > Of course it would be nice if all architectures would want to use it.
> > </wishful thinking>
> 
> It will for 2.5.X :-)

Good!

Gr{oetje,eeting}s,

						Geert

--
Geert Uytterhoeven -- There's lots of Linux beyond ia32 -- geert@linux-m68k.org

In personal conversations with technical people, I call myself a hacker. But
when I'm talking to journalists I just say "programmer" or something like that.
							    -- Linus Torvalds

Re: Re[2]: PC keyboard rate/delay

[Alan Cox]

> > It will for 2.5.X :-)
> Good!

Since it lets anyone hang many boxes by careful combined use of X and gpm I
suspect it wants fixing in 2.4