Re: can device drivers return non-ram via vm_ops->nopage?

[David Woodhouse]

On Sat, 2004-03-20 at 14:45 -0800, William Lee Irwin III wrote:
> This is the exact opposite of what I'd hoped come of this discussion.
> ISTR something about remap_area_pages() missing several pieces, but
> I pretty much need some kind of clarification to know what. Well, that,
> and I presumed your fixups for ALSA were headed toward mainline
> regardless after coping with whatever issue dwmw2 had (e.g. returning
> pfn's or something).

My request was that we shouldn't assume an architecture will have a
'struct page' corresponding to whatever it chooses to return from
dma_alloc_coherent(). 

There are machines where DMA to/from main memory _cannot_ be coherent
but we have some memory elsewhere, perhaps some SRAM which itself is
hanging off an I/O bus somewhere, which can be used. One of my toys is
currently running with dma_alloc_coherent() giving out memory from a PCI
video card, in fact.

Using a PFN should be OK.

-- 
dwmw2


-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/