Ming Lei wrote:

Ming Lei wrote:

quoted

2011/1/7 Anand Gadiyar [off-list ref]:

quoted

Hi Ming Lei,

I'm able to reproduce this on my panda, and I have it working as of
linux-next-20101221 (the last version I tested last year) and

failing

quoted

quoted

on linux-next-20101227 (which was the very next linux-next release).

Not sure why, but my Panda manages to get the VID:PID of the hub as
well, while yours does not even get there.

I may need a few more hours to debug this, unless someone beats me
to it. ;)

Is it caused by the below?

[   46.580200] ehci-omap ehci-omap.0: failed to get ehci port0

regulator

quoted

Nope. We don't set up any regulators for the PHY on Panda (at least not
yet).

And the working case (next-20101221) has the same log.

I suspect the PHY doesn't get reset for the proper duration; I'm looking
for something in that area. (Another option is to bisect between the two
versions, but I'm not sure if that's any easier).

Update: I disabled CONFIG_OMAP_RESET_CLOCKS and the PHY and other
connected devices enumerate just fine. Could you try this out on
your board as well?

I'll look deeper and see if there are some required clocks that are
accidentally getting turned off.

- Anand