Messages in this thread | | | From | (david parsons) | Subject | Re: unkown PCI device | Date | 30 Nov 1998 10:50:42 -0800 |
| |
In article <linux.kernel.19981130071504.A14818@tiktok.cygnus.com>, Michael Meissner <meissner@cygnus.com> wrote: >On Sun, Nov 29, 1998 at 07:08:46PM -0800, david parsons wrote: [/proc/pci; boon or menace?]
>> servers, unless the companies providing these servers obligingly >> provide free upgrades. If, instead, you rip all the vendor >> and card IDs out of the kernel (or put them into a config option, >> if the commercial servers require the actual text) and just have >> the Unknown vendor Unknown device, the commercial servers will >> be able to pick up the vendor and card ids (which I'd hope they're >> doing now, but who knows) until the last ia32 architecture machine >> rolls off to the junkyard. > >Another way to do it is to have a /proc/pci device driver that autoloads >whenever /proc/pci is loaded. Or allow userlevel programs to create a FIFO in >/proc, and hang a program off waiting for read requests.
That sounds like it would be the best of all possible worlds; move /proc/pci into a lkm, nail down the interface to it, then never touch it again. It does mean that the internal interfaces would need to be published, but that's a pretty trivial thing.
____ david parsons \bi/ UDI, anyone? \/
- To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to majordomo@vger.rutgers.edu Please read the FAQ at http://www.tux.org/lkml/
| |