On Mon, Jan 25, 2010 at 09:27:18PM -0700, Alex Chiang wrote: > * Greg KH <greg@xxxxxxxxx>: > > On Mon, Jan 25, 2010 at 09:57:45AM -0800, Gary Hade wrote: > > > > > > Andreas, Would the physical slot number to PCI ID mapping > > > provided in /sys/bus/pci/slots by the 'pci_slot' driver possibly > > > be of any use to you? > > > > That mapping should be already there if it is provided by the BIOS. > > Where? Somewhere else in sysfs? > > /sys/bus/pci/slots/ only gets populated when a hotplug driver or > pci_slot are loaded. > > And if those slots aren't hotplug capable, then that directory > won't get populated. Ah, I was thinking of the pci_slot stuff. > Finally, if the slots aren't described by ACPI, that directory > also won't get populated. > > If it's a relatively modern board, maybe they implement some of > the new SMBIOS fields that associate PCI bus:devfn to physical > slots... What does dmidecode say? > > > Also, Andreas, I see a symlink for all v4l devices already created in a > > persistant manner on the more recent udev releases. So you could just > > use that instead of the "raw" device node, and all should be good. > > otoh, this solution sounds a little better. :) Yes, that's the recommended way, as it's now supported by all distros. thanks, greg k-h -- To unsubscribe from this list: send the line "unsubscribe linux-pci" in the body of a message to majordomo@xxxxxxxxxxxxxxx More majordomo info at http://vger.kernel.org/majordomo-info.html
