Re: [PATCH v2] firmware: cs_dsp: fix missing pwr_lock in cs_dsp_stop()

[Richard Fitzgerald]

On 02/02/2026 2:41 pm, Ziyi Guo wrote:
> cs_dsp_signal_event_controls() calls cs_dsp_coeff_write_acked_control()
> which has lockdep_assert_held(&dsp->pwr_lock), but the lock is not held.
>
> cs_dsp_signal_event_controls() is invoked from cs_dsp_stop().
>
> Put mutex_lock() at the top of cs_dsp_stop() to fix the
> missing lock protection, also align with cs_dsp_run().
>
> Signed-off-by: Ziyi Guo <n7l8m4@xxxxxxxxxxxxxxxxxx>
> ---
> v2:
>   - Put mutex_lock() at the top of cs_dsp_stop() to fix and align with cs_dsp_run()
>
>   drivers/firmware/cirrus/cs_dsp.c | 3 ++-
>   1 file changed, 2 insertions(+), 1 deletion(-)
>
> diff --git a/drivers/firmware/cirrus/cs_dsp.c b/drivers/firmware/cirrus/cs_dsp.c
> index 525ac0f0a75d..a491e89c0c97 100644
> --- a/drivers/firmware/cirrus/cs_dsp.c
> +++ b/drivers/firmware/cirrus/cs_dsp.c
> @@ -2864,6 +2864,8 @@ EXPORT_SYMBOL_NS_GPL(cs_dsp_run, "FW_CS_DSP");
>    */
>   void cs_dsp_stop(struct cs_dsp *dsp)
>   {
> +	mutex_lock(&dsp->pwr_lock);
> +
>   	/* Tell the firmware to cleanup */
>   	cs_dsp_signal_event_controls(dsp, CS_DSP_FW_EVENT_SHUTDOWN);
>   
> @@ -2874,7 +2876,6 @@ void cs_dsp_stop(struct cs_dsp *dsp)
>   	if (dsp->ops->show_fw_status)
>   		dsp->ops->show_fw_status(dsp);
>   
> -	mutex_lock(&dsp->pwr_lock);
>   
>   	if (dsp->client_ops->pre_stop)
>   		dsp->client_ops->pre_stop(dsp);

Nitpick: Commit title "fix missing pwr_lock in cs_dsp_stop()", it's
not missing. Just in the wrong place.

Apart from that:
Reviewed-by: Richard Fitzgerald <rf@xxxxxxxxxxxxxxxxxxxxx>