Track down EHCI and companion errors on rk3xxx systems

[Diederik de Haas]

On Wed Jan 14, 2026 at 5:07 PM CET, Alan Stern wrote:
> However, I agree: 30 seconds is _much_ longer than it should take to 
> detect a newly plugged-in device.  It should take less than one second.  
> One possibility is that something involved in waking up the EHCI host 
> controller from its runtime suspend may have caused the delay -- that's 
> why I asked you to try turning off the controller's runtime suspend. 

It may be completely natural, but I was surprised runtime suspend was
involved at all. So I'll definitely do a test by turning it off.

> Also, I'm not at all familiar with the particular hardware used by your 
> platform for receiving wakeup signals.  It's possible that a GPIO 
> responsible for this wasn't working right and that's why you were 
> getting all those warning messages.  Another reason for wanting to know 
> what will happen if you take suspends and wakeups out of the picture.

I think GPIO is involved with wakeup signals ... sometimes at least.
I added the linux-rockchip ML precisely because I (strongly) suspect
there is Rockchip specific part to all this.

Cheers,
  Diederik

PS: I very much appreciate you taking the time for this 'debugging'
session :-)