Messages in this thread | | | Date | Sun, 21 Mar 2004 18:06:24 -0500 | From | Jeff Garzik <> | Subject | Re: can device drivers return non-ram via vm_ops->nopage? |
| |
David Woodhouse wrote: > On Sun, 2004-03-21 at 17:34 -0500, Jeff Garzik wrote: > >>Tell driver writers to call a standard platform function with a >>{dma|mmio|pio|vmalloc} handle+size+len for {dma|mmio|pio|vmalloc} mmap >>setup, and {fault|nopage} handler. ;-) IMO they shouldn't have to care >>about the details. > > > Don't let drivers see the {fault|nopage} handler. On most arches it can > probably continue to be nopage(); other arches may use the > newly-proposed fault() or perhaps just put all the PTEs in place up > front. The driver shouldn't be given an opportunity to care.
If that's possible within the MM APIs... certainly. Have a standard struct vm_operations_struct for dma, dma s/g, mmio, ... I presume?
Jeff
- To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html Please read the FAQ at http://www.tux.org/lkml/
| |