[PATCH v3] ASoC: fsl_sai: Force a software reset when starting in consumer mode

[PATCH v3] ASoC: fsl_sai: Force a software reset when starting in consumer mode

[Arun Raghavan]

From: Arun Raghavan <arun@asymptotic.io>

On an imx8mm platform with an external clock provider, when running the
receiver (arecord) and triggering an xrun with xrun_injection, we see a
channel swap/offset. This happens sometimes when running only the
receiver, but occurs reliably if a transmitter (aplay) is also
concurrently running.

It seems that the SAI loses track of frame sync during the trigger stop
-> trigger start cycle that occurs during an xrun. Doing just a FIFO
reset in this case does not suffice, and only a software reset seems to
get it back on track.

This looks like the same h/w bug that is already handled for the
producer case, so we now do the reset unconditionally on config disable.

Signed-off-by: Arun Raghavan <arun@asymptotic.io>
Reported-by: Pieterjan Camerlynck <p.camerlynck@televic.com>
---

v3
- Incorporate feedback from Shengjiu Wang to consolidate with the
  existing handling of this issue in producer mode

 sound/soc/fsl/fsl_sai.c | 14 ++++++++------
 1 file changed, 8 insertions(+), 6 deletions(-)

diff --git a/sound/soc/fsl/fsl_sai.c b/sound/soc/fsl/fsl_sai.c
index af1a168d35e3..50af6b725670 100644
--- a/sound/soc/fsl/fsl_sai.c
+++ b/sound/soc/fsl/fsl_sai.c
@@ -803,13 +803,15 @@ static void fsl_sai_config_disable(struct fsl_sai *sai, int dir)
 	 * anymore. Add software reset to fix this issue.
 	 * This is a hardware bug, and will be fix in the
 	 * next sai version.
+	 *
+	 * In consumer mode, this can happen even after a
+	 * single open/close, especially if both tx and rx
+	 * are running concurrently.
 	 */
-	if (!sai->is_consumer_mode[tx]) {
-		/* Software Reset */
-		regmap_write(sai->regmap, FSL_SAI_xCSR(tx, ofs), FSL_SAI_CSR_SR);
-		/* Clear SR bit to finish the reset */
-		regmap_write(sai->regmap, FSL_SAI_xCSR(tx, ofs), 0);
-	}
+	/* Software Reset */
+	regmap_write(sai->regmap, FSL_SAI_xCSR(tx, ofs), FSL_SAI_CSR_SR);
+	/* Clear SR bit to finish the reset */
+	regmap_write(sai->regmap, FSL_SAI_xCSR(tx, ofs), 0);
 }
 
 static int fsl_sai_trigger(struct snd_pcm_substream *substream, int cmd,
-- 
2.49.0

Re: [PATCH v3] ASoC: fsl_sai: Force a software reset when starting in consumer mode

[Fabio Estevam]

Hi Arun,

On Thu, Jun 26, 2025 at 8:52 AM Arun Raghavan <arun@arunraghavan.net> wrote:
>
> From: Arun Raghavan <arun@asymptotic.io>
>
> On an imx8mm platform with an external clock provider, when running the
> receiver (arecord) and triggering an xrun with xrun_injection, we see a
> channel swap/offset. This happens sometimes when running only the
> receiver, but occurs reliably if a transmitter (aplay) is also
> concurrently running.
>
> It seems that the SAI loses track of frame sync during the trigger stop
> -> trigger start cycle that occurs during an xrun. Doing just a FIFO
> reset in this case does not suffice, and only a software reset seems to
> get it back on track.
>
> This looks like the same h/w bug that is already handled for the
> producer case, so we now do the reset unconditionally on config disable.
>
> Signed-off-by: Arun Raghavan <arun@asymptotic.io>
> Reported-by: Pieterjan Camerlynck <p.camerlynck@televic.com>

What about adding a Fixes tag and Cc stable so that it gets backported
to the stable trees?

Re: [PATCH v3] ASoC: fsl_sai: Force a software reset when starting in consumer mode

[Arun Raghavan]

On Thu, 26 Jun 2025, at 7:58 AM, Fabio Estevam wrote:
> Hi Arun,
>
> On Thu, Jun 26, 2025 at 8:52 AM Arun Raghavan
> <arun@arunraghavan.net> wrote:
>>
>> From: Arun Raghavan <arun@asymptotic.io>
>>
>> On an imx8mm platform with an external clock provider, when running
>> the receiver (arecord) and triggering an xrun with xrun_injection, we
>> see a channel swap/offset. This happens sometimes when running only
>> the receiver, but occurs reliably if a transmitter (aplay) is also
>> concurrently running.
>>
>> It seems that the SAI loses track of frame sync during the trigger
>> stop -> trigger start cycle that occurs during an xrun. Doing just a
>> FIFO reset in this case does not suffice, and only a software reset
>> seems to get it back on track.
>>
>> This looks like the same h/w bug that is already handled for the
>> producer case, so we now do the reset unconditionally on config
>> disable.
>>
>> Signed-off-by: Arun Raghavan <arun@asymptotic.io> Reported-by:
>> Pieterjan Camerlynck <p.camerlynck@televic.com>
>
> What about adding a Fixes tag and Cc stable so that it gets backported
> to the stable trees?

Sure! Will send in a v4 with both of these.

The commit that added the initial fix (3e3f8bd56955 ("ASoC: fsl_sai: fix
no frame clk in master mode")) refers to an errata, but I could find
anything in a couple of imx8* errata I checked.

I was wondering if there's any public documentation of this that I can
refer (and link) to?

Cheers,
Arun

Re: [PATCH v3] ASoC: fsl_sai: Force a software reset when starting in consumer mode

[Fabio Estevam]

On Thu, Jun 26, 2025 at 9:28 AM Arun Raghavan <arun@arunraghavan.net> wrote:

> Sure! Will send in a v4 with both of these.
>
> The commit that added the initial fix (3e3f8bd56955 ("ASoC: fsl_sai: fix
> no frame clk in master mode")) refers to an errata, but I could find
> anything in a couple of imx8* errata I checked.
>
> I was wondering if there's any public documentation of this that I can
> refer (and link) to?

I'm not aware of any public errata for this issue.