linux-kernel - Re: [PATCH 1/2] dt-bindings: chrome: Add cros-ec-typec mux props

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 for Android: free password hash cracker in your pocket

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

Message-ID: <20200610153356.GC3213128@kuha.fi.intel.com>
Date:   Wed, 10 Jun 2020 18:33:56 +0300
From:   Heikki Krogerus <heikki.krogerus@...ux.intel.com>
To:     Prashant Malani <pmalani@...omium.org>,
        Rob Herring <robh@...nel.org>
Cc:     "linux-kernel@...r.kernel.org" <linux-kernel@...r.kernel.org>,
        Tim Wawrzynczak <twawrzynczak@...omium.org>,
        Benson Leung <bleung@...omium.org>,
        "open list:OPEN FIRMWARE AND FLATTENED DEVICE TREE BINDINGS" 
        <devicetree@...r.kernel.org>,
        Enric Balletbo i Serra <enric.balletbo@...labora.com>,
        Guenter Roeck <groeck@...omium.org>
Subject: Re: [PATCH 1/2] dt-bindings: chrome: Add cros-ec-typec mux props

On Tue, Jun 09, 2020 at 04:57:40PM -0700, Prashant Malani wrote:
> Hi Rob,
> 
> Thanks again for the comments and feedback. Kindly see responses inline:
> 
> (Trimming unrelated text from thread):
> 
> On Tue, Jun 09, 2020 at 02:30:11PM -0600, Rob Herring wrote:
> > On Fri, May 29, 2020 at 5:30 PM Prashant Malani <pmalani@...omium.org> wrote:
> > >
> > > Nodes truncated and unrelated fields omitted in the interest of brevity:
> > >
> > > // Chrome OS EC Type C Port Manager.
> > > typec {
> > >     compatible = "google,cros-ec-typec";
> > >     #address-cells = <1>;
> > >     #size-cells = <0>;
> > >
> > >     connector@0 {
> > >         compatible = "usb-c-connector";
> > >         reg = <0>;
> > >         power-role = "dual";
> > >         data-role = "dual";
> > >         try-power-role = "source";
> > >         mode-switch = <&foo_mux>;
> > >         // Other switches can point to the same mux.
> > >         ....
> > 
> > The connector is supposed to have 'ports' for USB2, USB3, and Aux
> > unless the parent is the USB controller.
> Understood; so, coupled with Heikki's explanation (see below for where
> I've pasted it), would it be something like so? (adding inline to the connector
> node definition):
> 
>             ports {
>                 #address-cells = <1>;
>                 #size-cells = <0>;
> 
>                 port@0 {
>                     reg = <0>;
>                     usb_con_hs: endpoint {
>                         remote-endpoint = <&foo_usb_hs_controller>;
>                     };
>                 };
> 
>                 port@1 {
>                     reg = <1>;
>                     usb_con0_ss: endpoint@0 {
>                         remote-endpoint = <&mode_mux_in>;
>                     };
>                 };
> 
>                 port@2 {
>                     reg = <2>;
>                     usb_con_sbu: endpoint {
>                         remote-endpoint = <&foo_dp_aux>;
>                     };
>                 };
>             };

The pins that can be reassigned can in practice go anywhere. We can't
group them in any way. What do we do for example when the two sideband
pins go to different locations?

It looks like the OF graph for the USB Type-C connectors expects the
pins to be grouped like that, which is too bad, because unfortunately
it will not work. It would require that we have a separate port for
every pin that can be reassigned on the connector, and let's not even
consider that.

We need higher level description of the connections for the USB Type-C
connectors. For example, a connector can be connected to this (or
these) DisplayPort(s), this USB 2.0 port, this USB 3.x port, this USB4
port, etc. And this is the mux that handles the pins on this
connector, and these are the retimers, etc. etc.

We also need a way to identify those connections, and relying on
something like fixed port node addresses, so indexes in practice,
feels really risky to me. A conflict may seem unlikely now, but we
seen those so many times in the past other things like GPIOs, IRQs,
etc. We really need to define string identifiers for these
connections.


thanks,

-- 
heikki