[PATCH 0/7][MV643XX_ETH] various cleanups

[PATCH 0/7][MV643XX_ETH] various cleanups

[Lennert Buytenhek]

Hi all,

This patch series performs various cleanups of mv643xx_eth, and moves
it into the direction of being able to be instantiated multiple times
-- which is currently not possible due to the driver using static
variables, identifying ports by their port index within the single
supported instance[*] instead of a pointer to a describing structure,
and such.

The goal of multiple instantiation isn't achieved yet by this patch
set, since it requires some changes in how mv643xx_eth platform
devices are instantiated which in turn requires some changes in
arch/{ppc,powerpc} which haven't been fully sorted out and agreed
upon yet -- but this patch set is a step in the general direction,
and the maintainer has informally ACKed these patches previously
(in private mail), so I'm seeking wider feedback on them now.

Comments?


thanks,
Lennert

[*] A single mv643xx_eth instance can support up to 3 ethernet
    ports, but multiple mv643xx_eth Si blocks can exist within a
    single chip, requiring instantiation of multiple mv643xx_eth
    instances.

Re: [PATCH 0/7][MV643XX_ETH] various cleanups

[Dale Farnsworth]

On Mon, Mar 17, 2008 at 08:15:46PM +0100, Lennert Buytenhek wrote:
> Hi all,
> 
> This patch series performs various cleanups of mv643xx_eth, and moves
> it into the direction of being able to be instantiated multiple times
> -- which is currently not possible due to the driver using static
> variables, identifying ports by their port index within the single
> supported instance[*] instead of a pointer to a describing structure,
> and such.
> 
> The goal of multiple instantiation isn't achieved yet by this patch
> set, since it requires some changes in how mv643xx_eth platform
> devices are instantiated which in turn requires some changes in
> arch/{ppc,powerpc} which haven't been fully sorted out and agreed
> upon yet -- but this patch set is a step in the general direction,
> and the maintainer has informally ACKed these patches previously
> (in private mail), so I'm seeking wider feedback on them now.

Hi Lennert,

I'm happy with the end result, but the current patch series will
not build after patches 1-3 are applied.  Please reorder your
code changes so the result will be git-bisectable.

-Dale

Re: [PATCH 0/7][MV643XX_ETH] various cleanups

[Lennert Buytenhek]

On Tue, Mar 18, 2008 at 09:14:06AM -0700, Dale Farnsworth wrote:

> > This patch series performs various cleanups of mv643xx_eth, and moves
> > it into the direction of being able to be instantiated multiple times
> > -- which is currently not possible due to the driver using static
> > variables, identifying ports by their port index within the single
> > supported instance[*] instead of a pointer to a describing structure,
> > and such.
> > 
> > The goal of multiple instantiation isn't achieved yet by this patch
> > set, since it requires some changes in how mv643xx_eth platform
> > devices are instantiated which in turn requires some changes in
> > arch/{ppc,powerpc} which haven't been fully sorted out and agreed
> > upon yet -- but this patch set is a step in the general direction,
> > and the maintainer has informally ACKed these patches previously
> > (in private mail), so I'm seeking wider feedback on them now.
> 
> Hi Lennert,

Hi Dale,


> I'm happy with the end result, but the current patch series will
> not build after patches 1-3 are applied.  Please reorder your
> code changes so the result will be git-bisectable.

Urgh.  I could have sworn that I checked that.  Anyway, sorry.

Fixed patch series here:
	http://www.wantstofly.org/~buytenh/mv643xx_eth/

Does that look better?


thanks,
Lennert

Re: [PATCH 0/7][MV643XX_ETH] various cleanups

[Dale Farnsworth]

On Tue, Mar 18, 2008 at 05:40:45PM +0100, Lennert Buytenhek wrote:
> On Tue, Mar 18, 2008 at 09:14:06AM -0700, Dale Farnsworth wrote:
> > I'm happy with the end result, but the current patch series will
> > not build after patches 1-3 are applied.  Please reorder your
> > code changes so the result will be git-bisectable.
> 
> Urgh.  I could have sworn that I checked that.  Anyway, sorry.
> 
> Fixed patch series here:
> 	http://www.wantstofly.org/~buytenh/mv643xx_eth/

Yes.  :)

I've applied all 1-7 of these to the for-2.6.26 branch of
	git://git.farnsworth.org/dale/linux-2.6-mv643xx_eth.git for-2.6.26

It's also viewable at
<http://www.farnsworth.org/git?p=dale/linux-2.6-mv643xx_eth.git;a=shortlog;h=for-2.6.26>

-Dale