Subject: Re: PCI and subsystem IDs.
To: None <>
From: None <>
List: tech-kern
Date: 10/06/1998 21:10:39

>>>>> "Chris" == Chris G Demetriou <> writes:
    Chris> Specifically, the things i'm wondering are:

    Chris> (1) what is the namespace for subsystem vendor IDs?  (I've heard
    Chris> "same as vendor IDs," but haven't had C&V quoted at me, so i'm
    Chris> skeptical. 8-)

  I'll look up C&V tomorrow at work for you.
  An AMCC S5920 databook that I was reading today and the AMD7992 databook
that I read yesterday confirms this.

    Chris> (2) what is the namespace for subsystem IDs?  I.e. is it per
    Chris> subsystem vendor, or per prod/vendor/subsystem_vendor?

  It is per-vendor ID, since the subsystem vendor id/Vendor ID space are

    Chris> The answers to those questions are necessary before anything
    Chris> 'useful' can be done around recording them in a consistent manner.

    Chris> There's also the issue that, in the largest share of devices i've
    Chris> seen, if they're present at all they're completely ignorable.
    Chris> There're only two cases i've heard of where paying attention to
    Chris> them are important, and of those, only one has a driver in the
    Chris> source tree.

  Apparently Microsoft has demanded that devices have them in order to be
win98 compliant. It is a good idea since it allows an LE (amd) driver to work
with anyone's AMD-based core, regardless of what *other* stuff might have
been incorporated by the "subsystem" vendor (i.e. the people who made the
card). The extra stuff simply isn't activated. I will be involved in
such an effort soon.

   :!mcr!:            |  Solidum Systems Corporation,
   Michael Richardson |For a better connected world,where data flows faster<tm>
 Personal: PGP key available.
 Corporate: <A HREF=""></A>. 

Version: 2.6.3ia
Charset: latin1
Comment: Processed by Mailcrypt 3.4, an Emacs/PGP interface