Re: Resuming rt73usb wireless network dongle: ep0in 3strikes

On Mon, 2007-09-03 at 17:26 -0400, Alan Stern wrote:
> On Mon, 3 Sep 2007, Mattias Nissler wrote:
> 
> > > The problem is that either your device or else the EHCI controller it 
> > > is attached to dropped the USB connection during the suspend.  The 
> > > same thing happened to the Alcor Micro mass storage device attached to 
> > > the same controller, which suggests the problem lies in the controller.  
> > > Or maybe it lies in the BIOS, which could reset the controller during 
> > > the initial stages of resuming.
> > 
> > Just for my understanding: Do you mean the system BIOS or some firmware
> > in the ICH8?
> 
> In principle it could be either one, but I doubt the firmware would do 
> it.
> 
> > Note that I'm *not* doing a complete software-suspend
> > cycle, but only a test run (cause I write "test" to /sys/power/disk).
> > How could the system BIOS be involved when the EHCI controller
> > suspends/resumes without actually rebooting the machine?
> 
> I don't know -- that was just a guess.  But clearly _something_ is 
> messing you up.

:-)

> 
> > Another question: If the controller drops the connection, shouldn't the
> > ehci driver just do a reconnect instead of trying to resume the
> > interface driver?
> 
> Yes, I can't tell exactly what happened there.  Did you or did you not
> have CONFIG_USB_DEBUG enabled?  The log contains some messages that
> won't be generated without it, but it's missing others.  For instance,
> there would be messages saying "finish resume" or "can't resume".  The
> missing messages should explain exactly what happened.

I switched CONFIG_USB_DEBUG on and recompiled just before I did the test
run that turned up the 3strike errors. Just to be sure, I'll do a
complete kernel recompile to see if it gives more messages.

> 
> It seems that the connections did remain intact long enough for the 
> system to think the resumes had succeeded.  Otherwise you wouldn't get 
> all those "3strikes" errors; you'd get -19 errors instead.  But it's 
> hard to be sure of the sequence.
> 
> > And how does CONFIG_USB_PERSIST relate to all that?
> 
> It doesn't come into effect unless the host controller driver
> recognizes at resume time that power was lost, which apparently didn't
> happen.  There were no "root hub lost power" messages in the log.

Ok, can you give me some pointers where in the usb code this is to be
detected and handled? I'll try to debug this, but I haven't worked with
the USB subsystem before, so any help I can get is welcome ;-)

Thanks for all the information so far!

Mattias


-------------------------------------------------------------------------
This SF.net email is sponsored by: Splunk Inc.
Still grepping through log files to find problems?  Stop.
Now Search log events and configuration files using AJAX and a browser.
Download your FREE copy of Splunk now >>  http://get.splunk.com/
_______________________________________________
[email protected]
To unsubscribe, use the last form field at:
https://lists.sourceforge.net/lists/listinfo/linux-usb-devel