Re: [PATCH v6 1/5] irqchip/aspeed-i2c-ic: binding docs for Aspeed I2C Interrupt Controller

[Benjamin Herrenschmidt]

On Wed, 2017-03-29 at 13:51 -0700, Brendan Higgins wrote:
> so maybe instead of setting a hard limit like I did, maybe the best
> thing is to just check and see what the base_clk gets set to and if
> it gets set to zero, we turn on high speed mode. What do you think?

Ah maybe. Did you scope it to see if clock_hi/low do indeed apply in
high speed mode ?

I wonder if that bit does other things.. I would be interesting to
check. Ohterwise why have the bit rather than just have the driver
write 0 to the divisor ?

The doc for the high speed mode bit says "high speed mode (3.4Mbps)"
which is why I, maybe incorrectly, assumed it was a fixed frequency.

Anyway, not a huge deal at this point, but something to look into
at some stage.

Cheers,
Ben.