The PCI ID Repository

The PCI ID Repository

[Martin Mares]

Hello, world!\n

I've created a public repository for PCI ID's to avoid the long delays
caused by me being unable to cope with the flood of ID updates I was
receiving.

The repository lives at http://pciids.sourceforge.net/ and you can download
the daily snapshots of pci.ids, browse the ID lists interactively and also
submit new entries via Web forms. Alternatively, you can mail your submissions
as unified diffs (no base64 encoded attachments, please) to pci-ids@ucw.cz
where they get processed by an e-mail robot and also automatically added
to the database, awaiting approval by one of the maintainers.

So share and enjoy and submit your ID's.

(There is still a part of my e-mail backlog unprocessed, so don't worry,
I'll send it to the mailbot soon.)

Also, I've released a new version of pciutils containing a pci.ids file
synchronized with the repository and sent a patch to Linus which will
get the kernel in sync as well.

				Have a nice fortnight
-- 
Martin `MJ' Mares   <mj@ucw.cz>   http://atrey.karlin.mff.cuni.cz/~mj/
Faculty of Math and Physics, Charles University, Prague, Czech Rep., Earth
"Dijkstra probably hates me." -- /usr/src/linux/kernel/sched.c

Re: The PCI ID Repository

[Albert D. Cahalan]

Martin Mares writes:

> The repository lives at http://pciids.sourceforge.net/ and you can download

What about revision codes?

Vendor and device really isn't enough to identify something.
There may be completely different chips with the same vendor
and device IDs.

Tundra provides an example: Universe, Universe II, Universe IIB

These chips are the same device in some sense; they are all
PCI-to-VME bridges. (damn popular too) The programming interface
is incompatible, so they get different revisions. Tundra isn't
alone in interpreting the PCI spec this way.

Re: The PCI ID Repository

[H. Peter Anvin]

Followup to:  <200111050044.fA50i8o182130@saturn.cs.uml.edu>
By author:    "Albert D. Cahalan" <acahalan@cs.uml.edu>
In newsgroup: linux.dev.kernel
> 
> Tundra provides an example: Universe, Universe II, Universe IIB
> 
> These chips are the same device in some sense; they are all
> PCI-to-VME bridges. (damn popular too) The programming interface
> is incompatible, so they get different revisions. Tundra isn't
> alone in interpreting the PCI spec this way.
> 

Another interpretation, which seems pretty common (we use this one at
Transmeta, for example) is to treat the version ID as the "minor
revision" (indicating an upward compatible change) and the device ID
as the "major revision" (change this to indicate an incompatible
change in the programming interface.)

Unfortunately PCI doesn't have the very nice "compatible with" list
that ISAPnP has -- that, and the "human readable string" were very
useful features of the ISAPnP spec.

	-hpa
-- 
<hpa@transmeta.com> at work, <hpa@zytor.com> in private!
"Unix gives you enough rope to shoot yourself in the foot."
http://www.zytor.com/~hpa/puzzle.txt	<amsp@zytor.com>