Patch for powercomputing keyboards

Patch for powercomputing keyboards

[Jerry Quinn]

Hi all.  I've put together a hack that allows the CapsLock key on
PowerComputing keyboards to work reasonably as a control key.

I was trying to tackle two problems.  One is that 'a' was being generated
under X when releasing the capslock key.  This apparently went away when I
switched to Xfree68_fbdev.

However, the current code in 2.2.10 generates a down and up when capslock is
pressed and the 0x80 generated when it is released is not used.  To get
capslock to work correctly as a soft key, the capslock up has to be
interpreted correctly.  Since it returns 0x80, same as 'a' up, this makes it
more complicated.

What I did is create a little state machine that watches 'a' and capslock.  It
will generate the 0xb9 when capslock is released, the way a sane keyboard
should work.  The only place where things can get confused is if you press
capslock, press 'a', and then release capslock.  I made the assumption that
'a' would always be released first, since I had to guess at something.

With this patch in place, I can remap the capslock to be a control key and
everything behaves fine.

At the moment, I think this patch is incompatible with Apple hard-locking
keyboards. 

So anyone with a Powercomputing keyboard, please try this out and let me know
what you think.  The patch is against Linus' 2.2.10.  If it seems good, I
think it also will need a configuration, unless someone has a suggestion on
how to make it play nicely with Apple keyboards.

-- 
Jerry Quinn                             Tel: (514) 761-8737
jquinn@nortelnetworks.com               Fax: (514) 761-8505
Speech Recognition Research

--------------------------------
*** mac_keyb.c.orig	Sun Jun  6 11:05:00 1999
--- linux/drivers/macintosh/mac_keyb.c	Sat Aug 14 01:06:21 1999
***************
*** 320,328 ****
--- 320,397 ----
  static void
  input_keycode(int keycode, int repeat)
  {
+ 	static int pcc_caps_state = 0;
  	struct kbd_struct *kbd;
  	int up_flag;
  
+ 	/* Kludge to try to fix capslock behavior for powercomputing
+ 	   keyboards.  On these, 0xb9 is returned on alternating
+ 	   capslock down events.  capslock up returns 0x80 - the same
+ 	   as A.  Fun fun.  */
+ 	/* This is a state machine to untangle capslock up
+ 	   and 'a' up on powercomputing keyboards */
+ 	switch (keycode) {
+ 	case 0x39:
+ 	case 0xb9:
+ 	  switch (pcc_caps_state) {
+ 	  case 0:
+ 	    pcc_caps_state = 2;
+ 	    break;
+ 	  case 1:
+ 	    pcc_caps_state = 3;
+ 	    break;
+ 	  case 4:
+ 	    pcc_caps_state = 5;
+ 	    break;
+ 	  default:
+ 	    break;
+ 	  }
+ 	  keycode = 0x39;	/* Change to capslock down */
+ 	  break;
+ 	case 0x00:
+ 	  switch (pcc_caps_state) {
+ 	  case 0:
+ 	    pcc_caps_state = 1;
+ 	    break;
+ 	  case 2:
+ 	  case 4:	/* We guessed right in returning A up previously */
+ 	    pcc_caps_state = 3;
+ 	    break;
+ 	  default:
+ 	    break;
+ 	  }
+ 	  break;
+ 
+ 	  /* This is the broken 0x80 return code from capslock
+ 	     or A.  We always assume A release first. */
+ 	case 0x80:
+ 	  switch (pcc_caps_state) {
+ 	  case 0:
+ 	  case 1:		/* A up */
+ 	    pcc_caps_state = 0;
+ 	    break;
+ 	  case 2:		/* CapsLock up */
+ 	    pcc_caps_state = 0;
+ 	    keycode = 0xb9;
+ 	    break;
+ 	  case 3:		/* Assume A up - we don't really know */
+ 	    pcc_caps_state = 4;
+ 	    keycode = 0x80;
+ 	    break;
+ 	  case 4:		/* Now assume CapsLock up */
+ 	    pcc_caps_state = 0;
+ 	    keycode = 0xb9;
+ 	    break;
+ 	  case 5:		/* Return an A up */
+ 	    pcc_caps_state = 4;
+ 	    keycode = 0x80;
+ 	    break;
+ 	  }
+ 	  break;
+ 	default:
+ 	  break;
+ 	}
+ 		  
   	kbd = kbd_table + fg_console;
  	up_flag = (keycode & 0x80);
  	keycode &= 0x7f;
***************
*** 383,400 ****
  		 * is pressed/released. Also, makes sure that the
  		 * LED are handled.  atong@uiuc.edu
  		 */
! 		 switch (keycode) {
! 		 /*case 0xb9:*/
! 		 case 0x39:
! 			handle_scancode(0x39, 1);
! 			handle_scancode(0x39, 0);
! 		 	mark_bh(KEYBOARD_BH);
! 		 	return;
! 		 case 0x47:
! 		 /*case 0xc7:*/
! 		 	mark_bh(KEYBOARD_BH);
! 		 	break;
! 		 }
  	}
  
  	handle_scancode(keycode, !up_flag);
--- 452,473 ----
  		 * is pressed/released. Also, makes sure that the
  		 * LED are handled.  atong@uiuc.edu
  		 */
! 		
! 		/* 0xb9 and 0x39 commented out because they don't let
! 		   capslock work as anything else */
! 		switch (keycode) {
! 		  /*case 0xb9:*/
! 		  /*		 case 0x39:
! 				 handle_scancode(0x39, 1);
! 				 handle_scancode(0x39, 0);
! 				 mark_bh(KEYBOARD_BH);
! 				 return; */
! 		case 0x39:
! 		case 0x47:
! 		  /*case 0xc7:*/
! 		  mark_bh(KEYBOARD_BH);
! 		  break;
! 		}
  	}
  
  	handle_scancode(keycode, !up_flag);


[[ This message was sent via the linuxppc-dev mailing list.  Replies are ]]
[[ not  forced  back  to the list, so be sure to Cc linuxppc-dev if your ]]
[[ reply is of general interest. Please check http://lists.linuxppc.org/ ]]
[[ and http://www.linuxppc.org/ for useful information before posting.   ]]

Re: Patch for powercomputing keyboards

[Simon Piette]

Jerry Quinn wrote:
> 
> Hi all.  I've put together a hack that allows the CapsLock key on
> PowerComputing keyboards to work reasonably as a control key.
> 
> So anyone with a Powercomputing keyboard, please try this out and let me know
> what you think.  The patch is against Linus' 2.2.10.  If it seems good, I
> think it also will need a configuration, unless someone has a suggestion on
> how to make it play nicely with Apple keyboards.
> 

Hi,

I have a Umax J700 and the patch work perfectly. Thanks a lot for this
annoying bug fix! Do you think it will be integrated in the vger and/or samba
kernel?

Simon

[[ This message was sent via the linuxppc-dev mailing list.  Replies are ]]
[[ not  forced  back  to the list, so be sure to Cc linuxppc-dev if your ]]
[[ reply is of general interest. Please check http://lists.linuxppc.org/ ]]
[[ and http://www.linuxppc.org/ for useful information before posting.   ]]

Re: Patch for powercomputing keyboards

[Jerry Quinn]

>> "Simon" == Simon Piette <spiette@generation.net> writes:

 Simon> Jerry Quinn wrote:
 >> Hi all.  I've put together a hack that allows the CapsLock key on
 >> PowerComputing keyboards to work reasonably as a control key.
 >> 
 >> So anyone with a Powercomputing keyboard, please try this out and let me
 >> know what you think.  The patch is against Linus' 2.2.10.  If it seems
 >> good, I think it also will need a configuration, unless someone has a
 >> suggestion on how to make it play nicely with Apple keyboards.
 >> 

 Simon> Hi,

 Simon> I have a Umax J700 and the patch work perfectly. Thanks a lot for this
 Simon> annoying bug fix! Do you think it will be integrated in the vger
 Simon> and/or samba kernel?

Thanks for the feedback!It's very useful to me as well.  

I would like it to go in, especially to the mainline as well as development
kernel.  Can anyone offer some input on how to get it into the vger kernel?

Thanks,
Jerry


** Sent via the linuxppc-dev mail list. See http://lists.linuxppc.org/