linux-kernel - Re: [PATCH 1/2] dt-bindings: usb: Introduce GPIO-based SBU mux

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]

Date:   Fri, 19 Aug 2022 23:04:28 -0500
From:   Bjorn Andersson <bjorn.andersson@...aro.org>
To:     Prashant Malani <pmalani@...omium.org>
Cc:     Stephen Boyd <swboyd@...omium.org>, Rob Herring <robh@...nel.org>,
        Krzysztof Kozlowski <krzysztof.kozlowski@...aro.org>,
        Greg Kroah-Hartman <gregkh@...uxfoundation.org>,
        Krzysztof Kozlowski <krzysztof.kozlowski+dt@...aro.org>,
        Heikki Krogerus <heikki.krogerus@...ux.intel.com>,
        linux-usb@...r.kernel.org, devicetree@...r.kernel.org,
        linux-kernel@...r.kernel.org, linux-arm-msm@...r.kernel.org,
        Pin-yen Lin <treapking@...omium.org>
Subject: Re: [PATCH 1/2] dt-bindings: usb: Introduce GPIO-based SBU mux

On Fri 19 Aug 17:18 CDT 2022, Prashant Malani wrote:

> On Fri, Aug 19, 2022 at 2:39 PM Bjorn Andersson
> <bjorn.andersson@...aro.org> wrote:
> >
> > On Fri 19 Aug 15:49 CDT 2022, Stephen Boyd wrote:
> >
> > > > I like 2 endpoints to represent the usb-c-connector, but that doesn't seem
> > > > to be compatible (without introducing `data-lanes`, at least) with all
> > > > the various
> > > > combinations on the remote side, if that remote side is a DRM bridge with DP
> > > > output capability (like it6505 or anx7625).
> > > > That type of DRM bridge supports 1, 2 or 4 lane DP connections.
> > >
> > > Why can't the remote side that's a pure DP bridge (it6505) bundle
> > > however many lanes it wants into one endpoint? If it's a pure DP bridge
> > > we should design the bridge binding to have up to 4 endpoints, but
> > > sometimes 2 or 1 and then overlay data-lanes onto that binding so that
> > > we can tell the driver how to remap the lanes if it can. If the hardware
> > > can't support remapping lanes then data-lanes shouldn't be in the
> > > binding.
> 
> 2 endpoints sounds fine to me. The overloading of the bridge-side endpoint
> to mean different things depending on what it's connected to seemed odd to
> me, but if that is acceptable for the bridge binding, then great.
> 
> > The existing implementation provides the interfaces usb_role_switch,
> > usb_typec_mux and usb_typec_switch. These works based on the concept
> > that the USB Type-C controller will request the endpoints connected to
> > the usb-c-connector about changes such as "switch to host mode", "switch
> > to 2+2 USB/DP combo" and "switch orientation to reverse". We use this
> > same operations to inform any endpoint at any port about these events
> > and they all react accordingly.
> 
> Right, but that implementation/assumption doesn't work so well when you
> have 2 Type-C ports which might route to the same bridge (2 lane from each).
> The other 2 lanes from the other endpoints can go to (say) a USB HUB.
> 

Are you saying that two of your SS-lanes in connector A are connected to
directly to the QMP PHY at the same time as two SS-lanes from connector
B are connected to the same two pads on the QMP PHY - without any
mux etc inbetween?

I.e. that there are a set of pins in connector A which is directly
connected to a set of pins in connector B?


I was under the impression that in your hardware there's some component
muxing the single DP output to one of the connectors. If so there should
be no graph-link directly connecting the two usb-c-connectors and the
one QMP PHY.

Is this not the case?

> >
> > Perhaps I'm misunderstanding your suggestion, but if you start
> > representing each individual lane in the SuperSpeed interface I believe
> > you would have to just abandon this interface and replace it with
> > something like "give me USB on port@...ndpoint@0 and port@...ndpoint@1
> > and give me DP on port@...ndpoint@2 and port@...ndpoint@3".
> 
> I don't think that is necessary. The switch driver can register the switches (
> and it can find out which end-points map to the same usb-c-connector).
> 
> From the port driver, the port driver just needs to tell each switch
> registered for it's port that "I want
> DP Pin assignment C/ DP Pin assignment D / Plain USB3.x" and the
> switch driver(s) can figure out what to output on its pins (since
> the Type-C binding will specify ep0 = A2-A3 (TX1), ep1 = B10-B11 , etc)
> 
> orientation-switch can tell the switch if the signals need to be swapped around.
> 

In a typical Qualcomm design the QMP PHY is directly connected to the
usb-c-connector and as such it is the component that implements
usb_typec_mux and usb_typec_switch.

Regards,
Bjorn

> The above notwithstanding, it sounds like the 2-ep approach has more support
> than 4 ep-approach, so this specific example is moot.