I'm running the 2.6.22.6 kernel with the berry_charge module loaded.  I connect my 8700c via USB, but the blackberry does not appear to charge. The power icon does not change, and even after staying connected for an extended period, the device does not appear to be gaining any power.  The cable is good, as I can plug it into my Windows machine and it begins charging immediately.  I see the following in dmesg:

[ 6638.606493] usb 1-2: new full speed USB device using uhci_hcd and address 2
[ 6638.776524] usb 1-2: configuration #1 chosen from 1 choice
[ 6639.639611] usbcore: registered new interface driver berry_charge
[ 6639.837097] usb 1-2: Second magic command failed: -71.
[ 6639.866430] usb 1-2: USB disconnect, address 2
[ 6640.610385] usb 1-2: new full speed USB device using uhci_hcd and address 3
[ 6640.775947] usb 1-2: configuration #1 chosen from 1 choice

I'd be glad to provide more information if needed.

-- 
Matt LaPlante
-

Ick, that's not nice.  Can you enable CONFIG_USB_DEBUG and debugging for
the berry_charge driver:
	modprobe berry_charge debug=1

and let us know what happens?

thanks,

greg k-h
-

On Sun, 9 Sep 2007 03:52:12 -0700

Here you go...

Sep  9 13:47:58 prizm kernel: [  521.256235] usbcore: registered new interface driver berry_charge
Sep  9 13:49:00 prizm kernel: [  582.750808] usb usb5: usb resume
Sep  9 13:49:00 prizm kernel: [  582.750814] usb usb5: finish resume
Sep  9 13:49:00 prizm kernel: [  582.750825] hub 5-0:1.0: hub_resume
Sep  9 13:49:00 prizm kernel: [  582.750828] ehci_hcd 0000:00:10.4: resume root hub
Sep  9 13:49:00 prizm kernel: [  582.793723] hub 5-0:1.0: state 7 ports 8 chg 0000 evt 0000
Sep  9 13:49:00 prizm kernel: [  582.793864] ehci_hcd 0000:00:10.4: GetStatus port 2 status 001803 POWER sig=j CSC CONNECT
Sep  9 13:49:00 prizm kernel: [  582.793978] hub 5-0:1.0: port 2, status 0501, change 0001, 480 Mb/s
Sep  9 13:49:00 prizm kernel: [  582.923607] hub 5-0:1.0: debounce: port 2: total 100ms stable 100ms status 0x501
Sep  9 13:49:00 prizm kernel: [  582.981556] ehci_hcd 0000:00:10.4: port 2 full speed --> companion
Sep  9 13:49:00 prizm kernel: [  582.981565] ehci_hcd 0000:00:10.4: GetStatus port 2 status 003801 POWER OWNER sig=j CONNECT
Sep  9 13:49:00 prizm kernel: [  582.981676] usb usb1: usb resume
Sep  9 13:49:00 prizm kernel: [  582.981678] usb usb1: finish resume
Sep  9 13:49:00 prizm kernel: [  582.981684] hub 1-0:1.0: hub_resume
Sep  9 13:49:00 prizm kernel: [  582.981686] usb usb1: wakeup_rh
Sep  9 13:49:00 prizm kernel: [  582.993728] hub 5-0:1.0: port 2 not reset yet, waiting 50ms
Sep  9 13:49:00 prizm kernel: [  583.049727] ehci_hcd 0000:00:10.4: GetStatus port 2 status 003802 POWER OWNER sig=j CSC
Sep  9 13:49:00 prizm kernel: [  583.049906] hub 5-0:1.0: state 7 ports 8 chg 0000 evt 0004
Sep  9 13:49:00 prizm kernel: [  583.049982] hub 1-0:1.0: state 7 ports 2 chg 0000 evt 0004
Sep  9 13:49:00 prizm kernel: [  583.050058] uhci_hcd 0000:00:10.0: port 2 portsc 0093,00
Sep  9 13:49:00 prizm kernel: [  583.050128] hub 1-0:1.0: port 2, status 0101, change 0001, 12 Mb/s
Sep  9 13:49:00 prizm kernel: [  583.177674] hub 1-0:1.0: debounce: po...
[ message continues ]

And so on, numerous times.




And the device is properly charging.

So, it should all be working now, it's just that we kept sending things

Ah, oh wait, now we just turned the power off.

Try disabling CONFIG_USB_SUSPEND and see if that fixes this issue.  Or
you can manually turn the power back on to your blackberry by writing to
the autosuspend file for the usb device in sysfs, but that can be a
pain.

Let me know if just changing that config option works for you.

thanks,

greg k-h
-

On Mon, 10 Sep 2007 23:35:02 -0700

And now for the dramatic conclusion...

To begin, I have no access to the original machine at the moment, as I'm now out of that area for a couple weeks.  I built a similar kernel (same version) on another box that I have at my current location.  The new machine is different hardware, so some kernel re-configuring was required, but I kept with the same USB settings (and similar overall design).  Interestingly, this machine didn't reproduce the "magic command failed" error, but it did fail very similarly to the original at charging the device.  I disabled CONFIG_USB_SUSPEND as suggested, and lo and behold, it now charges the berry.  Looks like an excellent diagnosis to me, doctor.  


-- 
Matt
-