> I went all in and did a 100ms delay before returning from the resets of
> 3 and 4 you mention. Sure enough, everything worked! It certainly should
> be understood and optimized. I added the linux-omap list to this thread
> (please let me know if there were others I should've CC'd on any of
> these emails).
It probably has little to do with the OMAP, it is the Microchip PHY
you need more information about.
100ms is a very long time. The data sheet says:
Note: For the first 16us after coming out of reset, the RMII
interface will run at 2.5 MHz. After this time, it will switch to 25
MHz if auto-negotiation is enabled.
I did not find anything else. I would try a deassert time the same as
the asset time as a starting point. It could also be that you ask
Linux for 100us sleep, and Linux actually gives your 1ms because of
the resolution of the timers. But i doubt many application will care
about a 1ms delay.
Andrew
