Re: USB device disconnects on resume

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

 



On Tue, Apr 19, 2022 at 11:28:21AM +0200, Martin Kepplinger wrote:
> hi,
> 
> I'm seeing resets and re-enumerations on runtime-resume for one device
> a lot. It's a modem connected to the USB2642 Microchip (SMSC) USB2 hub,
> that's connected to an xhci HC.
> 
> A remote wakeup *sometimes* makes the hub say "physically disconnected"
> during resume in hub_activate(), and thus sets reset_resume. Then the
> device comes up as low-speed device once, which again is not allowed
> during normal runtime resume, so would itself trigger a reset.

Does the reset-resume always fail in this way?

> The Hub and device is permanently connected on the PCB, so the hub is
> interpreting it in a wrong way.

What is the hub is interpreting in a wrong way?  Why should a permanent 
connection on the PCB have anything to do with whether the resume 
signals are misinterpreted?

>  I found an email that describes what I
> see from Sarah Sharp in 2013 here:
> https://marc.info/?l=linux-usb&m=137754385421825&w=2 Where she says:
> 
> "Occasionally, the host controller was sending the SoFs too soon on
> resume, and the device would interpret it as a low-speed chirp.  The
> device would disconnect, and transform from a high speed device to a
> low speed device.  I don't think increasing the 10 ms time out will
> help at all in this case, but you did ask what USB device disconnect
> scenarios I've seen."

Read the following messages in that email thread.  Sarah said that she 
would fix the SoF signal timing in xhci-hcd ("I agree that this seems 
like an xHCI driver issue, and I'll fix it in the driver").  I have no 
idea whether this helped the faulty devices; my guess is that it didn't.

(She never did respond to my comment that there is no such thing as a 
"low-speed chirp".)

> I can't find any reference to that bug yet. Has anyone experienced
> something (similar) and knows where I can at least try to work around
> that?
> 
> I don't know Sarahs' current email, could we forward this question to
> her?

Sarah hasn't worked on Linux or xhci-hcd for many years.  It's probably 
not worth trying to ask her about this.

Alan Stern



[Index of Archives]     [Linux Media]     [Linux Input]     [Linux Audio Users]     [Yosemite News]     [Linux Kernel]     [Linux SCSI]     [Old Linux USB Devel Archive]

  Powered by Linux