Re: Add skeleton "generic IO mapping" infrastructure.

Re: Add skeleton "generic IO mapping" infrastructure.

[David Woodhouse]

< moved to linux-arch >

On Tue, 2004-09-14 at 10:30 -0400, Jeff Garzik wrote:
> > Argh! Please no. You can't infer the IO space from the address. Provide
> > a cookie containing {space, address} instead -- or indeed {bus,
> > address}. Let some architectures optimise that by ignoring the bus and
> > working it out from the address if you must, but don't put that in the
> > generic version.
> 
> Override it in your arch if you don't like the generic version ;-)

No, seriously, let's not introduce a new API which is _already_
insufficient to disambiguate between the various things the caller might
want. A 'void *' is not sufficient to give both bus and address. Use a
tuple of { bus, addr } or { dev, addr } instead.

For platforms where it really doesn't matter you can optimise it out by
not looking at the device argument.

-- 
dwmw2

Re: Add skeleton "generic IO mapping" infrastructure.

[Jeff Garzik]

David Woodhouse wrote:
> No, seriously, let's not introduce a new API which is _already_
> insufficient to disambiguate between the various things the caller might
> want. A 'void *' is not sufficient to give both bus and address. Use a
> tuple of { bus, addr } or { dev, addr } instead.

The existing setup is perfectly sufficient.

You pass (dev,addr) to an arch-specific function which creates the 
cookie in an arch-specific manner.

	Jeff