On Tue, Sep 22, 2015 at 04:30:04PM +0200, Hans de Goede wrote:

Hi,

On 22-09-15 16:04, Maxime Ripard wrote:

quoted

On Tue, Sep 22, 2015 at 03:01:58PM +0200, Hans de Goede wrote:

quoted

quoted

In short, this is not about "power supply" but VBUS detection. IIRC,
if no VBUS detection method is provided, the phy driver just waits a
period of time after an ID pin change and then considers VBUS invalid.

Right, but that is a hack for boards with no / broken vbus detection
(or vbus control), we really want to use vbus-det where available,
so I agree that a TODO comment here would be good.

What's so special in our SoCs that makes that we can't just rely on
the ID pin ? (which seems to be working just fine here)

There is nothing special, AFAIK all OTG ports (also for other SoCs)
have some sort of vbus detection mechanism. We need the hack because
vbus-det is broken on some of our boards.

And where it is not broken we should be using vbus-det normallu.

Ah, and most controllers embed that directly into the controller
itself and don't rely on a GPIO to do that. Understood.

Thanks!
Maxime

-- 
Maxime Ripard, Free Electrons
Embedded Linux, Kernel and Android engineering
http://free-electrons.com
-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 819 bytes
Desc: Digital signature
URL: <http://lists.infradead.org/pipermail/linux-arm-kernel/attachments/20150929/f3a3e957/attachment-0001.sig>