linux-kernel - Re: [PATCH 3/4] ASoC: codecs: wsa884x: fix codec initialisation

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: <aVevbmfwoHCqrnQF@hovoldconsulting.com>
Date: Fri, 2 Jan 2026 12:43:42 +0100
From: Johan Hovold <johan@...nel.org>
To: Krzysztof Kozlowski <krzk@...nel.org>
Cc: Srinivas Kandagatla <srini@...nel.org>, Mark Brown <broonie@...nel.org>,
	Liam Girdwood <lgirdwood@...il.com>, linux-sound@...r.kernel.org,
	linux-arm-msm@...r.kernel.org, linux-kernel@...r.kernel.org,
	stable@...r.kernel.org
Subject: Re: [PATCH 3/4] ASoC: codecs: wsa884x: fix codec initialisation

On Fri, Jan 02, 2026 at 12:31:21PM +0100, Krzysztof Kozlowski wrote:
> On 02/01/2026 12:14, Johan Hovold wrote:
> > The soundwire update_status() callback may be called multiple times with
> > the same ATTACHED status but initialisation should only be done when
> > transitioning from UNATTACHED to ATTACHED.
> > 
> > Fix the inverted hw_init flag which was set to false instead of true
> > after initialisation which defeats its purpose and may result in
> > repeated unnecessary initialisation.
> 
> Either it results or it does not, not "may". 

No, it depends on whether update_status() is called with the same status
more than once. So "may" is correct here.

> If the device moves to
> UNATTACHED state flag should be probably set to "true". This is the bug.

No, update_status() has:

	if (wsa884x->hw_init || status != SDW_SLAVE_ATTACHED)
		return 0;

	...

	wsa884x_init(wsa884x);

so if you set hw_init to true then init is never called when status is
changed to ATTACHED.

> > 
> > Similarly, the initial state of the flag was also inverted so that the
> > codec would only be initialised and brought out of regmap cache only
> > mode if its status first transitions to UNATTACHED.
> 
> Maybe that's confusing wording but existing code was intentional and IMO
> almost correct. The flag is saying - we need hw init - that's why it is
> set to true in the probe and to false AFTER the proper hw initialization
> which is done after ATTACHED state.

All other codec drivers have hw_init mean that init has been done and
the check in update_status() reflects that too so this driver is still
broken.

Johan