Re: [PATCH 0/4] musb host improvments mostly for omap2430 glue

[Pavel Machek <pavel@ucw.cz>]

[-- Attachment #1: Type: text/plain, Size: 2608 bytes --]

Hi!

On Mon 2019-09-02 09:06:51, Tony Lindgren wrote:
> * Pavel Machek <pavel@denx.de> [190902 09:44]:
> > On Mon 2019-09-02 11:23:44, Pavel Machek wrote:
> > Hmm. I guess CONFIG_USB_MUSB_DUAL_ROLE=y might be useful.
> > 
> > And now... if I unplug/replug the usb after the boot, USB hub and
> > mouse are recognized. Good!
> > 
> > Less than minute later:
> > 
> > mmusb-hdrc.0.auto: Babble
> > USB disconnect
> 
> The babble is most likely caused by some kind of signaling issue.
> 
> > I unplug, replug usb (not at the phone, between hub and dongle, and
> > green LED indincating charging starts blinking rapidly.
> > 
> > cpcap-core spi0.0: EOT timed out.
> > 
> > I try plug/replug, and now green led is on.
> > 
> > I unplug replug at the phone, and get bunch more of messages:
> > 
> > musm _set_peripheral: already in peripheral mode: 99
> > musm _set_peripheral: already in peripheral mode: 81
> > musm _set_peripheral: already in peripheral mode: 81
> > 
> > musb_set_host: could not set host: 99
> > musb_set_host: could not set host: 99
> > musb_set_host: could not set host: 99
> > musb_set_host: could not set host: 99
> > musb_set_host: could not set host: 99
> > musb_set_host: could not set host: 99
> > musb_set_host: could not set host: 99
> > musb_set_host: could not set host: 99
> > 
> > Unplug/replug at host, and again, hub+mouse is detected.
> > 
> > I unplug power connected to one of USB hub's ports... and find out
> > that phone was _not_ powering it.
> > 
> > Ok... so something somehow works.... sometimes :-).
> 
> My guess is you're missing a USB micro-B cable with ID pin
> grounded, with that things should just work automagically.
> 
> So no need for hubs feeding back VBUS and no need to
> try to force host mode via sysfs unlike on n900.

I don't think so... I got it to run in the end (and I have to
apologize, it seems to work at least as long as it is plugged it an
boot and not touched).

So... I actually have two cables.

#1 definitely does not have ID pin grounded. That does not work, not
 even in original android.

#2 definitely has _something_, because it does work in original
 android. But not even original android provides VBUS (5V on USB) in
 that configuration. It also looks like hardware _can_ provide at
 least VBAT on VBUS, because I seen that during some of the crashes.

Thanks for the patches, BTW.

Best regards,
							Pavel

-- 
DENX Software Engineering GmbH,      Managing Director: Wolfgang Denk
HRB 165235 Munich, Office: Kirchenstr.5, D-82194 Groebenzell, Germany

[-- Attachment #2: Digital signature --]
[-- Type: application/pgp-signature, Size: 181 bytes --]