linux-kernel - Re: broonie-sound/for-next bisection: baseline.bootrr.asoc-simple-card-probed on kontron-sl28-var3-ads2

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: <a1ec388def4febd9af6ef477245ef2d3@walle.cc>
Date:   Tue, 27 Apr 2021 17:29:35 +0200
From:   Michael Walle <michael@...le.cc>
To:     Mark Brown <broonie@...nel.org>
Cc:     Guillaume Tucker <guillaume.tucker@...labora.com>,
        kernelci-results@...ups.io,
        Kuninori Morimoto <kuninori.morimoto.gx@...esas.com>,
        Heiko Thiery <heiko.thiery@...il.com>,
        alsa-devel@...a-project.org, Jaroslav Kysela <perex@...ex.cz>,
        Liam Girdwood <lgirdwood@...il.com>,
        Takashi Iwai <tiwai@...e.com>,
        Stephan Gerhold <stephan@...hold.net>,
        Thierry Reding <treding@...dia.com>,
        Jon Hunter <jonathanh@...dia.com>, linux-kernel@...r.kernel.org
Subject: Re: broonie-sound/for-next bisection:
 baseline.bootrr.asoc-simple-card-probed on kontron-sl28-var3-ads2

Am 2021-04-27 16:16, schrieb Michael Walle:
> Am 2021-04-27 15:57, schrieb Mark Brown:
>> On Tue, Apr 27, 2021 at 03:43:42PM +0200, Michael Walle wrote:
>>> Am 2021-04-27 14:51, schrieb Guillaume Tucker:
>>> > On 27/04/2021 11:19, Mark Brown wrote:
>>> > > On Tue, Apr 27, 2021 at 07:28:34AM +0900, Kuninori Morimoto wrote:
>> 
>>> > > > > > If so, all sai1 - sai6 are using "fsl,vf610-sai",
>>> > > > > > all saiX doesn't have .name. I think it should have different name.
>>> > > > > > In your case, at least, sai5 / sai6 needs to have
> 
> Where does that (null) come from? I've briefly tried the following
> patch and I'd assume I get a duplicate for "abc-wm8904-hifi", but I 
> still
> get the old (null)-wm8904-hifi.
> 
> diff --git a/sound/soc/fsl/fsl_sai.c b/sound/soc/fsl/fsl_sai.c
> index 407a45e48eee..36280008b5c6 100644
> --- a/sound/soc/fsl/fsl_sai.c
> +++ b/sound/soc/fsl/fsl_sai.c
> @@ -1081,6 +1081,7 @@ static int fsl_sai_probe(struct platform_device 
> *pdev)
>         sai->cpu_dai_drv.symmetric_rate = 1;
>         sai->cpu_dai_drv.symmetric_channels = 1;
>         sai->cpu_dai_drv.symmetric_sample_bits = 1;
> +       sai->cpu_dai_drv.name = "abc";
> 
>         if (of_find_property(np, "fsl,sai-synchronous-rx", NULL) &&
>             of_find_property(np, "fsl,sai-asynchronous", NULL)) {

Ok here it is whats going on:
We have a simple-audio-card. We use
freescale/fsl-ls1028a-kontron-sl28-var3-ads2.dts. The codec name is
parsed correctly and set to either f140000.audio-controller or
f150000.audio-controller. Thus have different names. _But_ later
asoc_simple_canonicalize_cpu() will reset the dai_name pointer to
NULL in simple_parse_node() if "single" is 1 and then we end up
having two links with the same name.

Which begs the two questions:
  (1) What is "single" actually and when should it be 1?
  (2) If single is 1, then the sysfs file will be named 
"(null)-codec-name".
      Do we want that?

I guess there is a reason for it to be set to NULL, see [1].

-michael

[1] 
https://elixir.bootlin.com/linux/v5.12/source/sound/soc/generic/simple-card-utils.c#L420