In article <17713.932196449@coconut.itojun.org> you write:
>>The PR is '6928':
>>http://www.NetBSD.org/cgi-bin/query-pr-single.pl?number=6928
>>If you do get an SMP kernel built, perhaps you could make it available
>>somewhere?  A friend has an SMP box, and we have a number at work,
>>it would be cool to boot a test kernel and at least see the second
>>CPU spin up. :)  I haven't tried building my own yet, though it 
>>might not be bad, since it is now possible to checkout the exact
>>files the diff is for.

>	Is it considered a bad thing to commit the above, with necessary
>	care to keep non-SMP build as before? (make sure to #ifdef)

>itojun

I was able (with a little work) to get the patches to apply to a late 1.3H
tree, and did enjoy seeing the second processor spin up on the machine
I tested them on :)

However, as it is, the code has a few buglets that need to be cleaned
up before it should go in.  Among other things, in mainbus.c it should
add its own struct to union mainbus_attach_args.  Right now it simply
overloads apm_attach_args, thus making it impossible to compile a kernel
with options XXMP but without options APM.

jdarrow

--
John Darrow - Senior Technical Specialist               Office: 630/752-5201
Computing Services, Wheaton College, Wheaton, IL 60187  Fax:    630/752-5968
John.P.Darrow@wheaton.edu                               
Alphapage: 6303160707@alphapage.airtouch.com            Pager:  630/316-0707