Re: [PATCH RFC net-next 03/12] net: dsa: replace phylink_get_interfaces() with phylink_get_caps()

["Russell King (Oracle)" <linux@armlinux.org.uk>]

On Wed, Nov 24, 2021 at 06:26:48PM +0000, Russell King (Oracle) wrote:
> On Wed, Nov 24, 2021 at 06:15:08PM +0000, Vladimir Oltean wrote:
> > On Wed, Nov 24, 2021 at 05:52:38PM +0000, Russell King (Oracle) wrote:
> > > Phylink needs slightly more information than phylink_get_interfaces()
> > > allows us to get from the DSA drivers - we need the MAC capabilities.
> > > Replace the phylink_get_interfaces() method with phylink_get_caps() to
> > > allow DSA drivers to fill in the phylink_config MAC capabilities field
> > > as well.
> > > 
> > > Signed-off-by: Russell King (Oracle) <rmk+kernel@armlinux.org.uk>
> > > ---
> > 
> > The effects of submitting new API without any user :)
> 
> It had users at the time, but they were not submitted, and the addition
> of MAC capabilities was a future development. Had they been submitted at
> the time, then they would have required updating too.

That was a bit rushed... to explain more fully.

Prior to the merge window, the development work was centered around
only eliminating the PHY_INTERFACE_MODE_xxx checks and the complexity
that the PHY_INTERFACE_MODE_NA technique brought into the many
validation functions. Users of this had already been merged, and
included mvneta and mvpp2. See these commits, which are all in
v5.16-rc1:

b63f1117aefc net: mvpp2: clean up mvpp2_phylink_validate()
76947a635874 net: mvpp2: drop use of phylink_helper_basex_speed()
6c0c4b7ac06f net: mvpp2: remove interface checks in mvpp2_phylink_validate()
8498e17ed4c5 net: mvpp2: populate supported_interfaces member

099cbfa286ab net: mvneta: drop use of phylink_helper_basex_speed()
d9ca72807ecb net: mvneta: remove interface checks in mvneta_validate()
fdedb695e6a8 net: mvneta: populate supported_interfaces member

The original commit adding phylink_get_interfaces() extended this
into DSA, and the intention was to submit at least mv88e6xxx, but
it was too close to the merge window to do so.

Through making that change and eventually eliminating the basex helper
from all drivers that were using it, thereby making the validate()
behaviour much cleaner, it then became clear that it was possible to
push this cleanup further by also introducing a MAC capabilities field
to phylink_config.

The addition of the supported_interfaces member and the addition of the
mac_capabilities member are two entirely separate developments, but I
have now chosen to combine the two after the merge window in order to
reduce the number of patches. They were separate patches in my tree up
until relatively recently, and still are for the mt7530 and b53 drivers
currently.

So no, this is not "The effects of submitting new API without any user".

Thanks.

-- 
RMK's Patch system: https://www.armlinux.org.uk/developer/patches/
FTTP is here! 40Mbps down 10Mbps up. Decent connectivity at last!