In message <200010182207.SAA01956@multics.mit.edu>, jhawk writes: 
				  ^^^ Heh, neat!

-> I think there are issues with this "fix" that I implemented.
-> For one thing, the pckbc the attachment still doesn't work on the Vaio.
-> Other laptops seem to have other quirks on particular pnpbios
-> indexes.

As well as some desktops... I found this out when I rebuilt a kernel for
my Gateway E5200 and it (again!) panic'ed in the PNPBIOS attach code.  
Turns out I had blown away the pnpbios.c that skipped index 14, which
when queried under the current code always causes the box to reboot,
though a PNPBIOS query tool for DOS I found works A-OK under NT on the
same box (with the caveat that it uses real mode, so I'm not sure if
NT is providing a shadow copy of the BIOS or simply redirects the BIOS
calls to some internal layer or does some other magic).

-> So I think we need some more generic quirk code, and probably
-> to pull out the static queries and go back the dynamic queries.

Agreed, though I think that will reduce the usefulness of pnpbios by
about 1000% (imagine this: sure, we can detect all those pnp devices
on your system's motherboard, but first you have to go through and 
figure out which ones you can query and which ones reboot your box 8-).

Anyway, I can't suppose I can blame that on you or NetBSD 8-)

-> But not in time for 1.5.

I hear you.

Thanks!
--rafal

----
Rafal Boni                                                  rafal@mediaone.net