linux-kernel - Re: [RFC PATCH 1/2] soundwire: add support for static port mapping

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]

Message-ID: <64d2537a-bd60-e0a3-c4aa-4f802c34102b@linux.intel.com>
Date:   Fri, 19 Feb 2021 13:52:04 -0600
From:   Pierre-Louis Bossart <pierre-louis.bossart@...ux.intel.com>
To:     Srinivas Kandagatla <srinivas.kandagatla@...aro.org>,
        Vinod Koul <vkoul@...nel.org>
Cc:     alsa-devel@...a-project.org, gregkh@...uxfoundation.org,
        linux-kernel@...r.kernel.org, sanyog.r.kale@...el.com,
        yung-chuan.liao@...ux.intel.com
Subject: Re: [RFC PATCH 1/2] soundwire: add support for static port mapping

>>>> It seems you are in a different solution-space, where the codec driver
>>>> needs to notify the master of which ports it needs to use?
>>> Correct! As Codec is the place where we have mixer controls ant it can
>>> clearly tell which master ports should be used for that particular
>>> configuration.
>> And that should come from firmware (DT etc) and driver should pass on
>> this info
> 
> Are you okay with the patch as it is, provided this information is 
> populated from DT?

I am fine with the direction at a high-level. The premise for SoundWire 
is that the bus is simple enough that it can be used in different 
contexts and architectures, so if Qualcomm need something that differs 
from what is needed for Intel we are really not in a position to object.

That said, I could use more explanations on how the mapping is defined: 
I don't think we have the same definition of 'static port mapping'. For 
SDCA integration, we plan to have a static mapping in some sort of ACPI 
table that will describe which port on the Manager side is connected to 
which ports on Peripheral XYZ. That's static as in set in stone in 
platform firmware. I understand the reference to DT settings as the same 
idea.

But if the mapping depends on the value of mixer controls as you 
describe it, then it's not static and defined by DT settings, but 
run-time defined.

Also maybe we'd want to have a more opaque way of passing the 
information, maybe with a stream private data or a callback, instead of 
hard-coding fields that are only used by Qualcomm.