Re: ODROID-C1/-C2 USB Detection only triggered by some devices dwc2

From: Martin Blumenstingl <martin.blumenstingl@googlemail.com>
To: Minas Harutyunyan <Minas.Harutyunyan@synopsys.com>
Cc: Alan Stern <stern@rowland.harvard.edu>,
	Matt Corallo <oc2udbzfd@mattcorallo.com>,
	 "linux-usb@vger.kernel.org" <linux-usb@vger.kernel.org>,
	 "linux-amlogic@lists.infradead.org"
	<linux-amlogic@lists.infradead.org>,
	 "linux.amoon@gmail.com" <linux.amoon@gmail.com>,
	Artur Petrosyan <Arthur.Petrosyan@synopsys.com>
Subject: Re: ODROID-C1/-C2 USB Detection only triggered by some devices dwc2
Date: Sun, 12 Dec 2021 17:08:55 +0100	[thread overview]
Message-ID: <CAFBinCAvFbLXoobcTWiOoUrRSFnNzhOdn-=dkxvJOMkeOC+EAg@mail.gmail.com> (raw)
In-Reply-To: <20210803213216.GA376608@rowland.harvard.edu>

Hi Minas,

first of all, I want to apologize for the big delay in this discussion.

On Tue, Aug 3, 2021 at 11:32 PM Alan Stern <stern@rowland.harvard.edu> wrote:
>
> Minas, you should look at analysis of the usbmon information below.
Minas, it would be great if you could take the time to go through
Alan's findings.
Since it's been a while I am trying to summarize the current state of
the findings here.

In [0] Alan mentioned:
> Minas, some of the things noted below may require your attention. In
> particular, the usbmon traces don't show the root hub doing what it should,
> which leads me to wonder whether the Genesys Logic hub attached to port 1
> really is getting suspended and resumed properly.

Then in [1] Alan goes through the usbmon traces from my Meson8b
Odroid-C1+ and comes to the following conclusion:
> [...] the trace shows that the GL hub was never fully suspended!
>
> So the question is: Why doesn't the port status on the DWC2 root hub
> show the suspend feature set, and why doesn't the root hub port get
> suspended?  This looks like a bug in the dwc2 driver. [...]

Linux 5.16-rc4 is still affected by this issue, meaning that it hasn't
been fixed in the meantime.

I am happy about any feedback which can help resolve this issue. That
means: even if you have an idea (like: flipping bit X in register Y
could help) which you are unable to test yourself then please let me
know - I can take care of the testing part (and share debug
logs/usbmon traces if needed).

Thank you and best regards,
Martin

[0] https://lore.kernel.org/linux-usb/20210721155817.GC633399@rowland.harvard.edu/
[1] https://lore.kernel.org/linux-usb/20210803213216.GA376608@rowland.harvard.edu/

_______________________________________________
linux-amlogic mailing list
linux-amlogic@lists.infradead.org
http://lists.infradead.org/mailman/listinfo/linux-amlogic