On Thu, Dec 14, 2006 at 06:26:26PM +0000, Alan wrote: > > Think of uio as just a "class" of driver, like input or v4l. It's still > > up to the driver writer to provide a proper bus interface to the > > hardware (pci, usb, etc.) in order for the device to work at all. > > Understood. That leads me to ask another question of the folks who deal > with a lot of these cards. How many could reasonably be described by the > following > > bar to map, offset, length, ro/rw, root/user, local-offset > (x n ?) > interrupt function or null > > It seems if we have a lot of this kind of card that all fit that pattern > it might actually get more vendors submitting updates if we had a single > pci driver that took a struct of the above as the device_id field so > vendors had to write five lines of IRQ code, a struct and update a PCI > table ? That seems to have mostly worked with all the parallel/serial > boards.
I think that something like this might work out, and it would be a good goal to get there eventually. But I would like to see a few drivers using the uio core to see where we can consolidate things like this first. thanks, greg k-h - To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to [EMAIL PROTECTED] More majordomo info at http://vger.kernel.org/majordomo-info.html Please read the FAQ at http://www.tux.org/lkml/
