> I tried this approach and it works as expected. Means whenever there is
> a c45 register access, it directly uses the
> oa_tc6_read_c45()/oa_tc6_write_c45() functions. Herewith I have attached
> the patch
> (v4-0006-net-ethernet-oa_tc6-implement-internal-PHY-initia.patch) which
> has this new implementation for your reference. Is this you expected?
> Can you comment on this?
Please just post a new patch series. I will then review it just like
other patches. Its O.K. to send patch series frequently, not just more
than one per day.
> I tried this approach by setting up is_c45 flag when I use
> phy_read_mmd() function but ended up with the kernel call trace
> (c45_kernel_call_trace.png) attached here for your reference.
Please post plain ASCII. I assume you have a serial port, so you
should be able to capture it. I'm not too surprised though, no other
driver plays with is_c45.
Andrew
