linux-kernel - Re: [PATCH v4 10/23] ASoC: simple-card: Wrong daifmt for CPU end of DPCM DAI link

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: <841ac69f-1c3f-2b13-17f9-6f196811ce52@nvidia.com>
Date:   Tue, 30 Jun 2020 09:17:59 +0530
From:   Sameer Pujar <spujar@...dia.com>
To:     Kuninori Morimoto <kuninori.morimoto.gx@...esas.com>
CC:     <spujar@...dia.com>, <broonie@...nel.org>, <perex@...ex.cz>,
        <tiwai@...e.com>, <robh+dt@...nel.org>, <lgirdwood@...il.com>,
        <thierry.reding@...il.com>, <jonathanh@...dia.com>,
        <digetx@...il.com>, <alsa-devel@...a-project.org>,
        <linux-tegra@...r.kernel.org>, <linux-kernel@...r.kernel.org>,
        <sharadg@...dia.com>, <mkumard@...dia.com>,
        <viswanathl@...dia.com>, <rlokhande@...dia.com>,
        <dramesh@...dia.com>, <atalambedu@...dia.com>,
        <nwartikar@...dia.com>, <swarren@...dia.com>,
        <nicoleotsuka@...il.com>
Subject: Re: [PATCH v4 10/23] ASoC: simple-card: Wrong daifmt for CPU end of
 DPCM DAI link



On 6/30/2020 6:26 AM, Kuninori Morimoto wrote:
> External email: Use caution opening links or attachments
>
>
> Hi Sameer
>
>>   snd_soc_runtime_set_dai_fmt() {
>>       ...
>>
>>       if (cpu_dai->component->driver->non_legacy_dai_naming)
>>           fmt = inv_dai_fmt;
>>
>>       ...
>>   }
>>
>> Above flips polarity for 'cpu_dai' if 'non_legacy_dai_naming' flag is set.
>>
>> 1. Hence example mentioned in the commit message does not work if my 'cpu_dai'
>> driver does not have this flag set.
> ?
> Do you want fo flip it ? or don't flip?
> It is for Codec <-> Codec connection.

For DPCM links I don't want to flip based on one Codec reference. My 
goal was to make the binding work for multiple CPU/Codec link. Hence I 
thought it would be better to explicitly describe the 'Master' DAI. We 
can eventually get rid of 'codec' argument from simple_dai_link_of_dpcm().
>> 2. While it is true that we consider reference of 'Codec' mode for simple CPU<->
>> Codec DAI links, for DPCM this does not seem flexible. For DPCM links CPU and
>> Codec are not directly connected (CPU<->Dummy or Dummy<->Codec). Please
>> consider, for example, if the DAI link has multiple CPU/Codecs. Which 'Codec'
>> reference needs to be considered? Isn't it better if we explicitly mention which
>> DAI we want to operate as 'Master'?
> I think Lars-Peter has (had ?) plan for this SND_SOC_DAIFMT_CBx_CFx
> flag flexibility ? Yes maybe it is needed for multi CPU/Codec system.
>
> Thank you for your help !!
>
> Best regards
> ---
> Kuninori Morimoto