Re: [-next] staging/vme: various fixes + new driver model for VME

From: Emilio G. Cota
Date: Fri Oct 22 2010 - 19:03:31 EST


On Fri, Oct 22, 2010 at 10:26:11 +0100, Martyn Welch wrote:
> On 22/10/10 07:36, Emilio G. Cota wrote:
>
> Hi Emilio,
>
> Thank you for the fixes. After a quick glance, there seem to be a number
> of valid fixes here, but I'm very concerned by the patches that change
> the driver model. We discussed this approach in August last year, I am
> still yet to be convinced by the approach you wish to take.

Yes I remember we discussed it privately. I didn't provide a solution
at that point, I just told you that I believed the model could
be improved--and this is my attempt.

> >
> > - Make DMA work on the tsi148 (it's the only bridge I've got).
> > This will probably involve changing or extending the
> > current API.
>
> The DMA is already working on the tsi148. In what way do you feel that
> the current API needs changing or extending for DMA?

Sorry I may be confused here. That's the impression I got after reading
the DMA code a few months ago.

Anyway forget about this for the time being.

> > - Test the whole thing with real hardware and a real VME driver
> > (currently out of tree as well, see [2]), which I'll try
> > to get merged, too--currently we just have vme_user.c which
> > really isn't a kosher driver.
> >
>
> The current API has been tested with real hardware, for both supported
> vme bridges, on multiple cards, by multiple people. If the changes to
> the API are to be applied, they would need to be throughly tested before
> they are applied. As I've said above - I am still not convinced by the
> change in approach.

Well what we have merged is just vme_user, which is an unusual driver
that is not even finished yet. I can't speak of what's not merged.

Don't get me wrong, I didn't mean to say that the current code has
never been tested.

What I wanted to say is that I could eventually merge a driver, and
that could validate the whole thing. However I wouldn't like to
port everything, test it, and then get a NAK because of the changed
driver model. So that's why I'm sending this first--I don't mind
if it's not yet merged, but at least I'd like to know if I'm on
the right track.

Thanks

Emilio
--
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/