linux-kernel - Re: [PATCH 1/3] soundwire: qcom: drop unneeded DAI .set

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: <df6715a5-9e8d-4cbd-a654-2a18a021ff99@linux.intel.com>
Date:   Fri, 3 Nov 2023 10:29:13 -0500
From:   Pierre-Louis Bossart <pierre-louis.bossart@...ux.intel.com>
To:     Krzysztof Kozlowski <krzysztof.kozlowski@...aro.org>,
        Andy Gross <agross@...nel.org>,
        Bjorn Andersson <andersson@...nel.org>,
        Konrad Dybcio <konrad.dybcio@...aro.org>,
        Vinod Koul <vkoul@...nel.org>,
        Bard Liao <yung-chuan.liao@...ux.intel.com>,
        Sanyog Kale <sanyog.r.kale@...el.com>,
        Srinivas Kandagatla <srinivas.kandagatla@...aro.org>,
        Banajit Goswami <bgoswami@...cinc.com>,
        Liam Girdwood <lgirdwood@...il.com>,
        Mark Brown <broonie@...nel.org>,
        Jaroslav Kysela <perex@...ex.cz>,
        Takashi Iwai <tiwai@...e.com>, linux-arm-msm@...r.kernel.org,
        alsa-devel@...a-project.org, linux-kernel@...r.kernel.org
Subject: Re: [PATCH 1/3] soundwire: qcom: drop unneeded DAI .set_stream
 callback


>>> Qualcomm Soundwire controller drivers do not support multi-link setups,
>>> so DAI .set_stream() callback will not be used.  What's more, if called
>>> it will overwrite the sdw_stream_runtime runtime set in DAI .startup
>>> (qcom_swrm_startup()) causing issues (unsupported multi-link error) when
>>> two Soundwire controllers are passed as codec DAIs.
>>
>> This last sentence is confusing at best.
>>
>> A controller can have one or more managers, each of whom can have one or
>> more peripherals.
>>
>> only peripherals should expose codec DAIs, managers should expose CPU DAIs.
>>
>> Put differently, the controller is the host part while the peripheral is
>> the codec part. "controllers passed as codec DAIs" is not really
>> possible, or this was a typo?
> 
> No, it wasn't a typo. Take a look here:
> https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/tree/arch/arm64/boot/dts/qcom/sc8280xp-lenovo-thinkpad-x13s.dts#n1023
> 
> The <&swr0 0> is the controller, although probably I should call it
> manager, but in case of Qualcomm I think they are 1-to-1.

Is this a case where the SoundWire manager is part of a codec?

In that case, how are the SoundWire peripheral modeled?

The .set_stream callback was really meant to be used when you have a CPU
DAI for the manager and a codec DAI for the peripheral(s). This seems to
be a different configuration where CPU and codec DAIs are mixed.