[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
RE: Vendor ID
- To: Mailing List Recipients <pci-sig-request@znyx.com>
- Subject: RE: Vendor ID
- From: "Kimmery, Clifford (FL51)" <kimmery@space.honeywell.com>
- Date: Wed, 23 Apr 1997 14:54:13 -0400
- Cc: "'PCI SIG'" <pci-sig@znyx.com>
- Encoding: 39 TEXT
- Resent-Date: Wed, 23 Apr 1997 14:54:13 -0400
- Resent-From: pci-sig-request@znyx.com
- Resent-Message-Id: <"vcoLg.0.yf2.zgbNp"@dart>
- Resent-Sender: pci-sig-request@znyx.com
If you use the invalid value '0xFFFF', your software may not be able to
tell the difference between your device and a non-existent device. Page
186 allows (even encourages) host bridges to convert a Master Abort on a
configuration read into a return value of all ones. This allows
configuration software to scan the bus reading vendor IDs and marking
all that return 0xFFFF as non-existent without paying the penalty of an
interrupt or other fault every time a Master Abort occurs.
It isn't clear from your question that you need a vendor ID. The PCI
vendor ID is used to identify a PCI interface device, not the board
containing it. If you are developing an interface device you need a
vendor ID, otherwise the manufacturer of the device you are using will
already have one.
Cliff Kimmery
Honeywell Inc.
kimmery@space.honeywell.com
>----------
>From: Wilson Leung[SMTP:wilsonl@cs.sfu.ca]
>Sent: Wednesday, April 23, 1997 1:52 PM
>To: Mailing List Recipients
>Subject: Vendor ID
>
>A condition added to the previous question:
> we are not currently a PCI SIG member. i.e. no vendor ID has
> been assigned to us.
>>Hi PCI experts,
>>Is there a vendor ID for cards that are still in development stage?
>>I considered using someone else ID temporily. However, it is not
>>a good solution as it may confuse software that depends on the
>>ID to determine what the card is.
>>Also, there is an "invalid value" in the spec on p.188. Is it
>>okay to use that one? What does invalid means anyway?
>
>Thanks
>
>
P ˜ …