Strange PCI IDs

Strange PCI IDs

[Nicolas Léveillé]

Hello,

While discovering the bttv code (I'd like to add support to a yet 
unsupported bttv derivative, the PMS PDI Deluxe) I found a strange commit:

changeset:   10944:6b90215088f0
user:        Mauro Carvalho Chehab <mchehab@redhat.com>
date:        Wed Mar 11 08:18:53 2009 -0300
summary:     Conceptronic CTVFMI2 PCI Id

within: linux/drivers/media/video/bt8xx/bttv-cards.c

         { 0x109e036e, BTTV_BOARD_CONCEPTRONIC_CTVFMI2,  "Conceptronic 
CTVFMi v2"},

Does the above work at all for autodetecting this card?

The PCI ID looks suspiciously like a generic PCI ID, especially 
considering that:

   109e : vendor id for "Brooktree Corporation"
   036e : device id for "Bt878 Video Capture"

So I immediately was surprised to see the above PCI ID constant.

It does not appear it would cause any problem since the vendor and 
device ids are reversed anyway, however is this really correct? 
Shouldn't there be a comment about it?

Cheers,
Nicolas Léveillé

Re: Strange PCI IDs

[hermann pitton]

Hi,

Am Montag, den 25.05.2009, 23:26 +0200 schrieb Nicolas Léveillé:
> Hello,
> 
> While discovering the bttv code (I'd like to add support to a yet 
> unsupported bttv derivative, the PMS PDI Deluxe) I found a strange commit:
> 
> changeset:   10944:6b90215088f0
> user:        Mauro Carvalho Chehab <mchehab@redhat.com>
> date:        Wed Mar 11 08:18:53 2009 -0300
> summary:     Conceptronic CTVFMI2 PCI Id
> 
> within: linux/drivers/media/video/bt8xx/bttv-cards.c
> 
>          { 0x109e036e, BTTV_BOARD_CONCEPTRONIC_CTVFMI2,  "Conceptronic 
> CTVFMi v2"},
> 
> Does the above work at all for autodetecting this card?
> 
> The PCI ID looks suspiciously like a generic PCI ID, especially 
> considering that:
> 
>    109e : vendor id for "Brooktree Corporation"
>    036e : device id for "Bt878 Video Capture"
> 
> So I immediately was surprised to see the above PCI ID constant.
> 
> It does not appear it would cause any problem since the vendor and 
> device ids are reversed anyway, however is this really correct? 
> Shouldn't there be a comment about it?
> 
> Cheers,
> Nicolas Léveillé
> 

on the saa7134 driver we treat these sort of subsystem IDs, lots of
them, as invalid. One patch once made use of them and that part was
removed.

Cheers,
Hermann