[PATCH] ASoC: max98373: Added 10ms delay after amp software reset

[PATCH] ASoC: max98373: Added 10ms delay after amp software reset

[Ryan Lee]

Signed-off-by: Ryan Lee <ryans.lee@maximintegrated.com>
---
 Changes : Applied 10ms delay after amp software reset.
           10ms guard time is required for stability.
 sound/soc/codecs/max98373.c | 2 ++
 1 file changed, 2 insertions(+)

diff --git a/sound/soc/codecs/max98373.c b/sound/soc/codecs/max98373.c
index 2764fae..6aaff03 100644
--- a/sound/soc/codecs/max98373.c
+++ b/sound/soc/codecs/max98373.c
@@ -730,6 +730,7 @@ static int max98373_probe(struct snd_soc_component *component)
 	/* Software Reset */
 	regmap_write(max98373->regmap,
 		MAX98373_R2000_SW_RESET, MAX98373_SOFT_RESET);
+	mdelay(10);
 
 	/* IV default slot configuration */
 	regmap_write(max98373->regmap,
@@ -818,6 +819,7 @@ static int max98373_resume(struct device *dev)
 
 	regmap_write(max98373->regmap,
 		MAX98373_R2000_SW_RESET, MAX98373_SOFT_RESET);
+	mdelay(10);
 	regcache_cache_only(max98373->regmap, false);
 	regcache_sync(max98373->regmap);
 	return 0;
-- 
2.7.4

Re: [PATCH] ASoC: max98373: Added 10ms delay after amp software reset

[Dmitry Torokhov]

On Wed, Aug 22, 2018 at 5:21 PM Ryan Lee <ryans.lee@maximintegrated.com> wrote:
>
> Signed-off-by: Ryan Lee <ryans.lee@maximintegrated.com>
> ---
>  Changes : Applied 10ms delay after amp software reset.
>            10ms guard time is required for stability.
>  sound/soc/codecs/max98373.c | 2 ++
>  1 file changed, 2 insertions(+)
>
> diff --git a/sound/soc/codecs/max98373.c b/sound/soc/codecs/max98373.c
> index 2764fae..6aaff03 100644
> --- a/sound/soc/codecs/max98373.c
> +++ b/sound/soc/codecs/max98373.c
> @@ -730,6 +730,7 @@ static int max98373_probe(struct snd_soc_component *component)
>         /* Software Reset */
>         regmap_write(max98373->regmap,
>                 MAX98373_R2000_SW_RESET, MAX98373_SOFT_RESET);
> +       mdelay(10);

Is it really necessary for the CPU to spin for 10msec here?
usleep_range() would be better solution.

>
>         /* IV default slot configuration */
>         regmap_write(max98373->regmap,
> @@ -818,6 +819,7 @@ static int max98373_resume(struct device *dev)
>
>         regmap_write(max98373->regmap,
>                 MAX98373_R2000_SW_RESET, MAX98373_SOFT_RESET);
> +       mdelay(10);
>         regcache_cache_only(max98373->regmap, false);
>         regcache_sync(max98373->regmap);
>         return 0;
> --
> 2.7.4
>

Thanks.

-- 
Dmitry

Re: [PATCH] ASoC: max98373: Added 10ms delay after amp software reset

[Mark Brown]

[-- Attachment #1: Type: text/plain, Size: 377 bytes --]

On Wed, Aug 22, 2018 at 05:31:04PM -0700, Dmitry Torokhov wrote:
> On Wed, Aug 22, 2018 at 5:21 PM Ryan Lee <ryans.lee@maximintegrated.com> wrote:
> > +       mdelay(10);

> Is it really necessary for the CPU to spin for 10msec here?
> usleep_range() would be better solution.

Or msleep() for that matter, it's not going ot be the end of the world
if we delay a bit too long.

[-- Attachment #2: signature.asc --]
[-- Type: application/pgp-signature, Size: 488 bytes --]

Re: [PATCH] ASoC: max98373: Added 10ms delay after amp software reset

[Dmitry Torokhov]

On Thu, Aug 23, 2018 at 10:51:07AM +0100, Mark Brown wrote:
> On Wed, Aug 22, 2018 at 05:31:04PM -0700, Dmitry Torokhov wrote:
> > On Wed, Aug 22, 2018 at 5:21 PM Ryan Lee <ryans.lee@maximintegrated.com> wrote:
> > > +       mdelay(10);
> 
> > Is it really necessary for the CPU to spin for 10msec here?
> > usleep_range() would be better solution.
> 
> Or msleep() for that matter, it's not going ot be the end of the world
> if we delay a bit too long.

Frankly, I'd rather we did not introduce extra delays. I do not know if
we use this part, but on ChromeOS we are interested in shaving as much
from the boot time as possible and 10 msecs here and 10 msecs there do
add up.

Thanks.

-- 
Dmitry

RE: [PATCH] ASoC: max98373: Added 10ms delay after amp software reset

[Ryan Lee]

>-----Original Message-----
>From: Dmitry Torokhov <dmitry.torokhov@gmail.com>
>Sent: Thursday, August 23, 2018 5:08 PM
>To: Mark Brown <broonie@kernel.org>
>Cc: Ryan Lee <RyanS.Lee@maximintegrated.com>; Liam Girdwood
><lgirdwood@gmail.com>; Jaroslav Kysela <perex@perex.cz>; Takashi Iwai
><tiwai@suse.com>; Kuninori Morimoto <kuninori.morimoto.gx@renesas.com>;
>alsa-devel@alsa-project.org; lkml <linux-kernel@vger.kernel.org>;
>ryan.lee.maxim@gmail.com
>Subject: Re: [PATCH] ASoC: max98373: Added 10ms delay after amp software
>reset
>
>EXTERNAL EMAIL
>
>
>
>On Thu, Aug 23, 2018 at 10:51:07AM +0100, Mark Brown wrote:
>> On Wed, Aug 22, 2018 at 05:31:04PM -0700, Dmitry Torokhov wrote:
>> > On Wed, Aug 22, 2018 at 5:21 PM Ryan Lee
><ryans.lee@maximintegrated.com> wrote:
>> > > +       mdelay(10);
>>
>> > Is it really necessary for the CPU to spin for 10msec here?
>> > usleep_range() would be better solution.
>>
>> Or msleep() for that matter, it's not going ot be the end of the world
>> if we delay a bit too long.
>
>Frankly, I'd rather we did not introduce extra delays. I do not know if we use
>this part, but on ChromeOS we are interested in shaving as much from the boot
>time as possible and 10 msecs here and 10 msecs there do add up.

Thanks for all your feedback.
Let me apply usleep_range() or msleep() instead of using mdelay().

>
>Thanks.
>
>--
>Dmitry