Subject: USB keyboards, boot selector, boot program.
To: None <port-i386@netbsd.org>
From: Richard Rauch <rauch@eecs.ukans.edu>
List: port-i386
Date: 02/24/2001 16:30:19
The following is in the context of NetBSD/i386 1.5.  Since the affected
programs run independant of the kernel, I'm posting to port-i386, rather
than netbsd-help.  (The iMac's must handle this---USB is all that they
have, isn't it?)

I recently had my old PS/2 keyboard die on me.  (Well, it took a tumble
and the space bar came off---the switch mechanism below the key vanished,
and the keyboard isn't worth a whole lot to me without a space bar.  (^&)

Well, I attached a USB keyboard.  (My computer has two MicroSoft products
on it, now: An IntelliMouse and a USB ``wave'' keyboard.  IMHO, they put
the ``6'' key on the wrong side of the wave, but otherwise it's a
servicable product.  MS hardware is actually usable.  This never ceases to
amaze me.  *grin*)

Anyway, due to some thunderstorms, I had my computer off.  When I came
back online, I noticed two pieces of NetBSD software didn't pay attention
to the USB keyboard:

  The boot selector.  I have a NetBSD and a LINUX partition.  The boot
  selector ignored my pressing F4 for NetBSD to boot.  Thankfully,
  NetBSD is on the default boot partition.  (^&

  The boot program (where you're allowed 5 seconds to override the
  default multiuser boot to /netbsd).  I wanted to skip the countdown,
  per usual, but it ignored my keypresses.

Both of these shortcomings are significant, IMHO.  Especially since the
BIOS setup program has enough smarts to detect and use the USB
keyboard.  (I happened to actually enter setup before booting NetBSD.)

If this isn't fixed in -current (I assume not, since it's not been
discussed that I know of), how much trouble is it to address?  Is it too
much to expect these programs to either handle USB, or use BIOS
routines?  (Again, since BIOS setup worked with the keyboard, I assume
that BIOS can handle it fairly transparently.)

Or should this have actually worked for me, and I've stumbled upon a
heretofore unknown bug?  If so, dmesg output can be provided.  (^&


(I havn't send-pr'ed it since I don't know if it's a known---or even
fixed---problem, or if it's beyond the scope of the boot selector and boot
programs to deal with USB keyboards.)


Thanks for any answers.

  "I probably don't know what I'm talking about." --rauch@eecs.ukans.edu