Updating PCI vendors database

[Garrett Cooper]

On Mon, Apr 4, 2011 at 11:04 AM, Doug Barton <dougb at freebsd.org> wrote:
> On 04/04/2011 07:10, Philip Paeps wrote:
>>
>> It looks like our /usr/share/misc/pci_vendors list (used only by pciconf
>> as
>> far as I can tell) has become rather stale.  We also appear to be tracking
>> sources which no longer exist.
>>
>> Would anyone object if I updated this list to source the same database
>> used by
>> Linux distributions at http://pciids.sourceforge.net/v2.2/pci.ids?
>>
>> It helps that our pciconf looks to be compatible with that format.  We
>> just
>> ignore subvendor and subdevice, but it doesn't appear to matter that the
>> file
>> contains this information.
>>
>> I could cull the subvendor/subdevice from the list though.
>>
>> Any views?
>
> Having read this thread, and the last one, my opinion is, let's do it
> already. :)  Repo churn should not, under any circumstances, be a
> consideration in technical improvements. I agree with those who have said
> that the new list should be confirmed to be a superset of the old, and
> anything missing should be merged in. Checking with Jack about Intel stuff
> is also reasonable, as would be cross-checking with what NetBSD and OpenBSD
> are doing (and perhaps communicating with them about your work).

1. People may have automation that depends on this output.
2. Some braindead SCMs may be problematic with this change (p4? *COUGH*).

Someone at $WORK recently backported a copy of pci_vendors from
CURRENT to STABLE without any issues (needed a few PCI IDs for new
hardware), but that might not have been as true with this shift to
using pci.ids.

> So, not a slam-dunk, but definitely a clear path forward. Oh, and I
> personally don't see a problem with MFC'ing this, but I'm willing to be
> convinced.

Thanks,
-Garrett