Hi,
On 21-03-16 12:49, Chen-Yu Tsai wrote:
Hi,
I've been looking at the new EMAC driver and help cleaning
up the platform related (probe/remove/DT) bits.
The new hardware has the same glue layer like we had with
earlier SoCs with the GMAC. So for the A83T we could just
re-use the clock driver for that bit.
With the H3's internal PHY things get complicated. The
controls for the PHY are mapped to the upper bits of the
same register.
Should we just map the whole register in the EMAC driver,
and not use the clock driver in this case? (Maybe the clock
approach for the GMAC was bad?) Or try to share the register
with a regmap or syscon? The latter has its own difficulties
in the kernel. Or maybe a generic PHY driver that also exports
the clock?
Either way I think we'll end up with more than 1 way to deal
with the GMAC/EMAC clock settings.
Any suggestions?
I think that a phy driver which also registers a clock is
probably the best solution.
Note that the orangepi plus has an external gbit mac connected
to the H3, let me know if you want me to run some tests on that
one.
Regards,
Hans
--
You received this message because you are subscribed to the Google Groups
"linux-sunxi" group.
To unsubscribe from this group and stop receiving emails from it, send an email
to [email protected].
For more options, visit https://groups.google.com/d/optout.
