tech-kern archive

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index][Old Index]

Re: Inconsistencies in usage of "locators" argument to config (*ca_rescan)() functions



On Sat 27 Mar 2021 at 02:12:36 +0300, Valery Ushakov wrote:
> On Fri, Mar 26, 2021 at 13:18:16 -0700, Jason Thorpe wrote:
> 
> > I think it may have been the terminology used by Chris Torek in his
> >  paper on the new 4.4BSD device auto configuration framework [...].
> >  Sadly, that paper is somewhat hard to find, and I don't know if it
> >  was ever actually published anywhere.
> 
> This one? :)
> 
> http://www.netbsd.org/docs/kernel/config-torek.ps

Seems so! The file is a bit annoying postscript (backwards, and pages
cannot be viewed independent from others). Here are some quotes:

p.2: In many cases, each bus driver can probe the bus all by itself. This
is called /direct configuration/, and is preferred because it finds
physically-connected hardware regardless of the presence of proper
drivers and configurations. Sometimes, however, the driver needs "hints"
as to what to try. This sort of /indirect configuration/ is more general
and is also provided.

p.14: [A]utoconfiguration can be run "forwards" (directly) or "sideways"
(indirectly). Direct configuration is better, because it detects
hardware that is physically present even if is not configured into the
system, but it can be used only on "self-describing" buses. The
SnailBus, for instance, [ explains the SnailBus, which seems similar to
PCI, and then p.15ff explains the configuration process more deeply  ] 

> -uwe
-Olaf.
-- 
___ Q: "What's an anagram of Banach-Tarski?"  -- Olaf "Rhialto" Seibert
\X/ A: "Banach-Tarski Banach-Tarski."         -- rhialto at falu dot nl

Attachment: signature.asc
Description: PGP signature



Home | Main Index | Thread Index | Old Index