hardclock() sources

[Poul-Henning Kamp]

In message <20090506101301.GA17874 at server.vk2pj.dyndns.org>, Peter Jeremy write
s:

>A better solution would seem to be to use a HPET periodic timer.

Part of the problem is discovering the timers location early enough.

We have long (10 years) talked about doing (at least) two scans of
the busses:

First scan to allocate missing resources
Second scan to get "important" hardware (console, timers)
Third scan to find the rest.

-- 
Poul-Henning Kamp       | UNIX since Zilog Zeus 3.20
phk at FreeBSD.ORG         | TCP/IP since RFC 956
FreeBSD committer       | BSD since 4.3-tahoe    
Never attribute to malice what can adequately be explained by incompetence.