View Single Post
javispedro's Avatar
Posts: 2,355 | Thanked: 5,249 times | Joined on Jan 2009 @ Barcelona
#867
Originally Posted by egoshin View Post
EDIT: I do not enable charge pump, it is impossible (read joerg). But to work as HOST some VBUS voltage is needed (read blue_led)
That's my issue here, I though you don't need vbus at all for certain devices. In fact I'm not still sure about this. I know that the driver as is disables the PHY when gaia's vbus sense is not asserted -- you must have seen this.
And of course without the (even if it's broken) pump and without a host connection it's surely not asserted. Pumping even a measly 3,0 V might mean that the useless-for-charger-detection vbus sense in gaia gets asserted, and the PHY turns on.

(Of course the less changes to the driver the better)

Originally Posted by egoshin View Post
And of course - the procedure... it is not suitable for normal use because it requires PC initially. I am working on that stuff right now, and first - try to use FORCE_HOST. However, it is not simple as a success experiment, it doesn't see anything at all, so I am looking into code and acquiring logic.
You mention that you need to select PC Suite mode. Does it work with USB STorage Mode? If it works only with PC Suite mode, I think it might have to do with g_mass_storage (weird suspicion). rmmod it and modprobe g_zero before switching to host.

Originally Posted by egoshin View Post
BTW, if joerg is successfull in setting VBUS charge voltage and workaround BME intervention it could be a solution for non-selfpowered hubs/HDs. I did an experiment with his command (first version) and it seems it works, at least until DSME shutdowns N900.
DSME reboots can be entirely disabled via a /etc file. The issue iirc is the rest of watchdogs, and even those can be disabled by a r&d flag.
 

The Following 2 Users Say Thank You to javispedro For This Useful Post: