All of lore.kernel.org
 help / color / mirror / Atom feed
From: Nicolin Chen <nicoleotsuka@gmail.com>
To: Arnaud Ferraris <arnaud.ferraris@collabora.com>
Cc: alsa-devel@alsa-project.org, Timur Tabi <timur@kernel.org>,
	Xiubo Li <Xiubo.Lee@gmail.com>,
	Fabio Estevam <festevam@gmail.com>,
	Liam Girdwood <lgirdwood@gmail.com>,
	linux-kernel@vger.kernel.org, Mark Brown <broonie@kernel.org>,
	kernel@collabora.com, Shengjiu Wang <shengjiu.wang@gmail.com>
Subject: Re: [PATCH v3 1/1] ASoC: fsl_asrc: make sure the input and output clocks are different
Date: Fri, 17 Jul 2020 16:05:12 -0700	[thread overview]
Message-ID: <20200717230511.GA8495@Asurada-Nvidia> (raw)
In-Reply-To: <20200717103857.31877-2-arnaud.ferraris@collabora.com>

On Fri, Jul 17, 2020 at 12:38:58PM +0200, Arnaud Ferraris wrote:
> The current clock selection algorithm might select the same clock for
> both input and output when, for instance, the output sample rate is a
> multiple of the input rate.
> 
> In that case, both selectable clocks will be multiples of both the input
> and output sample rates, and therefore the first of these clocks will be
> selected for both input and output, leading to miscalculation of the
> dividers for either the input or output side.
> 
> Example:
>   Input uses clock A (512kHz) and has a sample rate of 8kHz
>   Output uses clock B (1024kHz) and has a sample rate of 16kHz
> 
>   In this case, the algorithm will select clock A for both input and
>   output: the input divider will therefore be calculated properly
>   (512 / 8 => 64), but the output divider's value will be only half
>   the expected value (512 / 16 => 32 instead of 1024 / 16 => 64).
> 
> This patch makes sure it always selects distinct input and output
> clocks.

Please allow me to take some time to review the use case and
the changes. And I'd love to wait for a review from Shengjiu
also, as he introduced this auto-selection function and he's
more familiar with internal ratio mode.

Thanks

> Signed-off-by: Arnaud Ferraris <arnaud.ferraris@collabora.com>
> ---
>  sound/soc/fsl/fsl_asrc.c | 16 ++++++++++------
>  1 file changed, 10 insertions(+), 6 deletions(-)
> 
> diff --git a/sound/soc/fsl/fsl_asrc.c b/sound/soc/fsl/fsl_asrc.c
> index 02c81d2e34ad..de10c208d3c8 100644
> --- a/sound/soc/fsl/fsl_asrc.c
> +++ b/sound/soc/fsl/fsl_asrc.c
> @@ -608,8 +608,8 @@ static void fsl_asrc_select_clk(struct fsl_asrc_priv *asrc_priv,
>  {
>  	struct fsl_asrc_pair_priv *pair_priv = pair->private;
>  	struct asrc_config *config = pair_priv->config;
> -	int rate[2], select_clk[2]; /* Array size 2 means IN and OUT */
> -	int clk_rate, clk_index;
> +	int rate[2], select_clk[2], clk_index[2]; /* Array size 2 means IN and OUT */
> +	int clk_rate;
>  	int i = 0, j = 0;
>  
>  	rate[IN] = in_rate;
> @@ -618,11 +618,15 @@ static void fsl_asrc_select_clk(struct fsl_asrc_priv *asrc_priv,
>  	/* Select proper clock source for internal ratio mode */
>  	for (j = 0; j < 2; j++) {
>  		for (i = 0; i < ASRC_CLK_MAP_LEN; i++) {
> -			clk_index = asrc_priv->clk_map[j][i];
> -			clk_rate = clk_get_rate(asrc_priv->asrck_clk[clk_index]);
> -			/* Only match a perfect clock source with no remainder */
> +			clk_index[j] = asrc_priv->clk_map[j][i];
> +			clk_rate = clk_get_rate(asrc_priv->asrck_clk[clk_index[j]]);
> +			/*
> +			 * Only match a perfect clock source with no remainder
> +			 * and make sure the input & output clocks are different
> +			 */
>  			if (clk_rate != 0 && (clk_rate / rate[j]) <= 1024 &&
> -			    (clk_rate % rate[j]) == 0)
> +			    (clk_rate % rate[j]) == 0 &&
> +			    (j == 0 || clk_index[j] != clk_index[j - 1]))
>  				break;
>  		}
>  
> -- 
> 2.27.0
> 

WARNING: multiple messages have this Message-ID (diff)
From: Nicolin Chen <nicoleotsuka@gmail.com>
To: Arnaud Ferraris <arnaud.ferraris@collabora.com>
Cc: alsa-devel@alsa-project.org, Timur Tabi <timur@kernel.org>,
	Xiubo Li <Xiubo.Lee@gmail.com>,
	Fabio Estevam <festevam@gmail.com>,
	Shengjiu Wang <shengjiu.wang@gmail.com>,
	Liam Girdwood <lgirdwood@gmail.com>,
	Mark Brown <broonie@kernel.org>,
	linux-kernel@vger.kernel.org, kernel@collabora.com
Subject: Re: [PATCH v3 1/1] ASoC: fsl_asrc: make sure the input and output clocks are different
Date: Fri, 17 Jul 2020 16:05:12 -0700	[thread overview]
Message-ID: <20200717230511.GA8495@Asurada-Nvidia> (raw)
In-Reply-To: <20200717103857.31877-2-arnaud.ferraris@collabora.com>

On Fri, Jul 17, 2020 at 12:38:58PM +0200, Arnaud Ferraris wrote:
> The current clock selection algorithm might select the same clock for
> both input and output when, for instance, the output sample rate is a
> multiple of the input rate.
> 
> In that case, both selectable clocks will be multiples of both the input
> and output sample rates, and therefore the first of these clocks will be
> selected for both input and output, leading to miscalculation of the
> dividers for either the input or output side.
> 
> Example:
>   Input uses clock A (512kHz) and has a sample rate of 8kHz
>   Output uses clock B (1024kHz) and has a sample rate of 16kHz
> 
>   In this case, the algorithm will select clock A for both input and
>   output: the input divider will therefore be calculated properly
>   (512 / 8 => 64), but the output divider's value will be only half
>   the expected value (512 / 16 => 32 instead of 1024 / 16 => 64).
> 
> This patch makes sure it always selects distinct input and output
> clocks.

Please allow me to take some time to review the use case and
the changes. And I'd love to wait for a review from Shengjiu
also, as he introduced this auto-selection function and he's
more familiar with internal ratio mode.

Thanks

> Signed-off-by: Arnaud Ferraris <arnaud.ferraris@collabora.com>
> ---
>  sound/soc/fsl/fsl_asrc.c | 16 ++++++++++------
>  1 file changed, 10 insertions(+), 6 deletions(-)
> 
> diff --git a/sound/soc/fsl/fsl_asrc.c b/sound/soc/fsl/fsl_asrc.c
> index 02c81d2e34ad..de10c208d3c8 100644
> --- a/sound/soc/fsl/fsl_asrc.c
> +++ b/sound/soc/fsl/fsl_asrc.c
> @@ -608,8 +608,8 @@ static void fsl_asrc_select_clk(struct fsl_asrc_priv *asrc_priv,
>  {
>  	struct fsl_asrc_pair_priv *pair_priv = pair->private;
>  	struct asrc_config *config = pair_priv->config;
> -	int rate[2], select_clk[2]; /* Array size 2 means IN and OUT */
> -	int clk_rate, clk_index;
> +	int rate[2], select_clk[2], clk_index[2]; /* Array size 2 means IN and OUT */
> +	int clk_rate;
>  	int i = 0, j = 0;
>  
>  	rate[IN] = in_rate;
> @@ -618,11 +618,15 @@ static void fsl_asrc_select_clk(struct fsl_asrc_priv *asrc_priv,
>  	/* Select proper clock source for internal ratio mode */
>  	for (j = 0; j < 2; j++) {
>  		for (i = 0; i < ASRC_CLK_MAP_LEN; i++) {
> -			clk_index = asrc_priv->clk_map[j][i];
> -			clk_rate = clk_get_rate(asrc_priv->asrck_clk[clk_index]);
> -			/* Only match a perfect clock source with no remainder */
> +			clk_index[j] = asrc_priv->clk_map[j][i];
> +			clk_rate = clk_get_rate(asrc_priv->asrck_clk[clk_index[j]]);
> +			/*
> +			 * Only match a perfect clock source with no remainder
> +			 * and make sure the input & output clocks are different
> +			 */
>  			if (clk_rate != 0 && (clk_rate / rate[j]) <= 1024 &&
> -			    (clk_rate % rate[j]) == 0)
> +			    (clk_rate % rate[j]) == 0 &&
> +			    (j == 0 || clk_index[j] != clk_index[j - 1]))
>  				break;
>  		}
>  
> -- 
> 2.27.0
> 

  reply	other threads:[~2020-07-17 23:06 UTC|newest]

Thread overview: 36+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2020-07-16 14:51 [PATCH 0/2] ASoC: fsl_asrc: improve clock selection and quality Arnaud Ferraris
2020-07-16 14:51 ` Arnaud Ferraris
2020-07-16 14:52 ` [PATCH 1/2] ASoC: fsl_asrc: make sure the input and output clocks are different Arnaud Ferraris
2020-07-16 14:52   ` Arnaud Ferraris
2020-07-16 15:07   ` Arnaud Ferraris
2020-07-16 15:07     ` Arnaud Ferraris
2020-07-16 14:52 ` [PATCH 2/2] ASoC: fsl_asrc: always use internal ratio Arnaud Ferraris
2020-07-16 14:52   ` Arnaud Ferraris
2020-07-16 15:13 ` [PATCH v2 0/2] ASoC: fsl_asrc: improve clock selection and quality Arnaud Ferraris
2020-07-16 15:13   ` Arnaud Ferraris
2020-07-16 15:13   ` [PATCH v2 1/2] ASoC: fsl_asrc: make sure the input and output clocks are different Arnaud Ferraris
2020-07-16 15:13     ` Arnaud Ferraris
2020-07-16 23:20     ` Nicolin Chen
2020-07-16 23:20       ` Nicolin Chen
2020-07-17 10:38       ` [PATCH v3 0/1] ASoC: fsl_asrc: always select different clocks Arnaud Ferraris
2020-07-17 10:38         ` Arnaud Ferraris
2020-07-17 10:38         ` [PATCH v3 1/1] ASoC: fsl_asrc: make sure the input and output clocks are different Arnaud Ferraris
2020-07-17 10:38           ` Arnaud Ferraris
2020-07-17 23:05           ` Nicolin Chen [this message]
2020-07-17 23:05             ` Nicolin Chen
2020-07-20 10:34           ` Shengjiu Wang
2020-07-20 10:34             ` Shengjiu Wang
2020-07-17 11:21         ` [PATCH v3 0/1] ASoC: fsl_asrc: always select different clocks Mark Brown
2020-07-17 11:21           ` Mark Brown
2020-07-17 11:34           ` Arnaud Ferraris
2020-07-17 11:34             ` Arnaud Ferraris
2020-07-17 11:45             ` Mark Brown
2020-07-17 11:45               ` Mark Brown
2020-07-16 15:13   ` [PATCH v2 2/2] ASoC: fsl_asrc: always use internal ratio Arnaud Ferraris
2020-07-16 15:13     ` Arnaud Ferraris
2020-07-16 23:37     ` Nicolin Chen
2020-07-16 23:37       ` Nicolin Chen
2020-07-17  9:58       ` Arnaud Ferraris
2020-07-17  9:58         ` Arnaud Ferraris
2020-07-20 10:20         ` Shengjiu Wang
2020-07-20 10:20           ` Shengjiu Wang

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=20200717230511.GA8495@Asurada-Nvidia \
    --to=nicoleotsuka@gmail.com \
    --cc=Xiubo.Lee@gmail.com \
    --cc=alsa-devel@alsa-project.org \
    --cc=arnaud.ferraris@collabora.com \
    --cc=broonie@kernel.org \
    --cc=festevam@gmail.com \
    --cc=kernel@collabora.com \
    --cc=lgirdwood@gmail.com \
    --cc=linux-kernel@vger.kernel.org \
    --cc=shengjiu.wang@gmail.com \
    --cc=timur@kernel.org \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.