Re: Proposal for new generic device API: dma_get_required_mask()

[James Bottomley <James.Bottomley@steeleye.com>]

On Fri, 2004-06-18 at 18:07, Krzysztof Halasa wrote:
> James Bottomley <James.Bottomley@steeleye.com> writes:
> I don't think so. We already have separate masks for coherent
> and non-coherent mappings (in PCI API, and I'm told it's to be extended
> to DMA API as well). And we need them.
> 
> The problem is we're missing DMA masks for non-alloc calls (depending
> on the platform) and thus that it isn't very reliable. Drivers which
> need this are forced to bounce buffers themselves, and many of them
> will not work on 64-bit platforms (as of ~ 2.6.0, I don't check that
> regularly). And yes, we really need reliable masks for non-alloc
> mappings.

Could you elaborate on this?  In the current scheme the coherent mask is
for descriptor allocation (i.e. dma_alloc_coherent()) and the dma_mask
represents the bus physical addresses to which the device can DMA
directly; there's not much more the DMA API really does, what do you
think is missing?

James