Groups | Search | Server Info | Keyboard shortcuts | Login | Register [http] [https] [nntp] [nntps]


Groups > linux.kernel > #1680896 > unrolled thread

[PATCH v2 6/9] clk: at91: clk-generated: make gclk determine audio_pll rate

Started byQuentin Schulz <quentin.schulz@free-electrons.com>
First post2017-07-04 14:10 +0200
Last post2017-07-06 18:10 +0200
Articles 4 — 4 participants

Back to article view | Back to linux.kernel

This discussion starts older than the indexed window; earlier articles aren't shown. The article labeled Started by below is the oldest one visible, not the original post.


Contents

  [PATCH v2 6/9] clk: at91: clk-generated: make gclk determine audio_pll rate Quentin Schulz <quentin.schulz@free-electrons.com> - 2017-07-04 14:10 +0200
    Re: [PATCH v2 6/9] clk: at91: clk-generated: make gclk determine  audio_pll rate Boris Brezillon <boris.brezillon@free-electrons.com> - 2017-07-04 22:50 +0200
    Re: [PATCH v2 6/9] clk: at91: clk-generated: make gclk determine  audio_pll rate Nicolas Ferre <nicolas.ferre@microchip.com> - 2017-07-06 17:30 +0200
      Re: [PATCH v2 6/9] clk: at91: clk-generated: make gclk determine  audio_pll rate Alexandre Belloni <alexandre.belloni@free-electrons.com> - 2017-07-06 18:10 +0200

#1680896 — [PATCH v2 6/9] clk: at91: clk-generated: make gclk determine audio_pll rate

FromQuentin Schulz <quentin.schulz@free-electrons.com>
Date2017-07-04 14:10 +0200
Subject[PATCH v2 6/9] clk: at91: clk-generated: make gclk determine audio_pll rate
Message-ID<tZuOt-83k-9@gated-at.bofh.it>
This allows gclk to determine audio_pll rate and set the parent rate
accordingly.

However, there are multiple children clocks that could technically
change the rate of audio_pll (via gck). With the rate locking, the first
consumer to enable the clock will be the one definitely setting the rate
of the clock.

Since audio IPs are most likely to request the same rate, we enforce
that the only clks able to modify gck rate are those of audio IPs.

To remain consistent, we deny other clocks to be children of audio_pll.

Signed-off-by: Quentin Schulz <quentin.schulz@free-electrons.com>
---

v2:
  - added conditions for audio pll rate setting restriction for SSC and
I2S,

 drivers/clk/at91/clk-generated.c | 48 +++++++++++++++++++++++++++++++++++-----
 1 file changed, 42 insertions(+), 6 deletions(-)

diff --git a/drivers/clk/at91/clk-generated.c b/drivers/clk/at91/clk-generated.c
index 6530a2e7e84d..87866786a6ab 100644
--- a/drivers/clk/at91/clk-generated.c
+++ b/drivers/clk/at91/clk-generated.c
@@ -26,6 +26,13 @@
 #define GENERATED_SOURCE_MAX	6
 #define GENERATED_MAX_DIV	255
 
+#define GCK_ID_SSC0		43
+#define GCK_ID_SSC1		44
+#define GCK_ID_I2S0		54
+#define GCK_ID_I2S1		55
+#define GCK_ID_CLASSD		59
+#define GCK_INDEX_DT_AUDIO_PLL	5
+
 struct clk_generated {
 	struct clk_hw hw;
 	struct regmap *regmap;
@@ -126,15 +133,14 @@ static int clk_generated_determine_rate(struct clk_hw *hw,
 {
 	struct clk_generated *gck = to_clk_generated(hw);
 	struct clk_hw *parent = NULL;
+	struct clk_rate_request req_parent = *req;
 	long best_rate = -EINVAL;
-	unsigned long min_rate;
+	unsigned long min_rate, parent_rate;
 	int best_diff = -1;
 	int i;
+	u32 div;
 
-	for (i = 0; i < clk_hw_get_num_parents(hw); i++) {
-		u32 div;
-		unsigned long parent_rate;
-
+	for (i = 0; i < clk_hw_get_num_parents(hw) - 1; i++) {
 		parent = clk_hw_get_parent_by_index(hw, i);
 		if (!parent)
 			continue;
@@ -150,11 +156,40 @@ static int clk_generated_determine_rate(struct clk_hw *hw,
 		clk_generated_best_diff(req, parent, parent_rate, div,
 					&best_diff, &best_rate);
 
+		if (!best_diff)
+			break;
+	}
+
+	/*
+	 * The audio_pll rate can be modified, unlike the five others clocks
+	 * that should never be altered.
+	 * The audio_pll can technically be used by multiple consumers. However,
+	 * with the rate locking, the first consumer to enable to clock will be
+	 * the one definitely setting the rate of the clock.
+	 * Since audio IPs are most likely to request the same rate, we enforce
+	 * that the only clks able to modify gck rate are those of audio IPs.
+	 */
+
+	if (gck->id != GCK_ID_SSC0 && gck->id != GCK_ID_SSC1 &&
+	    gck->id != GCK_ID_I2S0 && gck->id != GCK_ID_I2S1 &&
+	    gck->id != GCK_ID_CLASSD)
+		goto end;
+
+	parent = clk_hw_get_parent_by_index(hw, GCK_INDEX_DT_AUDIO_PLL);
+	if (!parent)
+		goto end;
+
+	for (div = 1; div < GENERATED_MAX_DIV + 2; div++) {
+		req_parent.rate = req->rate * div;
+		__clk_determine_rate(parent, &req_parent);
+		clk_generated_best_diff(req, parent, req_parent.rate, div,
+					&best_diff, &best_rate);
 
 		if (!best_diff)
 			break;
 	}
 
+end:
 	pr_debug("GCLK: %s, best_rate = %ld, parent clk: %s @ %ld\n",
 		 __func__, best_rate,
 		 __clk_get_name((req->best_parent_hw)->clk),
@@ -264,7 +299,8 @@ at91_clk_register_generated(struct regmap *regmap, spinlock_t *lock,
 	init.ops = &generated_ops;
 	init.parent_names = parent_names;
 	init.num_parents = num_parents;
-	init.flags = CLK_SET_RATE_GATE | CLK_SET_PARENT_GATE;
+	init.flags = CLK_SET_RATE_GATE | CLK_SET_PARENT_GATE |
+		CLK_SET_RATE_PARENT;
 
 	gck->id = id;
 	gck->hw.init = &init;
-- 
2.11.0

[toc] | [next] | [standalone]


#1681179 — Re: [PATCH v2 6/9] clk: at91: clk-generated: make gclk determine audio_pll rate

FromBoris Brezillon <boris.brezillon@free-electrons.com>
Date2017-07-04 22:50 +0200
SubjectRe: [PATCH v2 6/9] clk: at91: clk-generated: make gclk determine audio_pll rate
Message-ID<tZCVI-4Yl-3@gated-at.bofh.it>
In reply to#1680896
Le Tue,  4 Jul 2017 13:59:24 +0200,
Quentin Schulz <quentin.schulz@free-electrons.com> a écrit :

> This allows gclk to determine audio_pll rate and set the parent rate
> accordingly.
> 
> However, there are multiple children clocks that could technically
> change the rate of audio_pll (via gck). With the rate locking, the first
> consumer to enable the clock will be the one definitely setting the rate
> of the clock.
> 
> Since audio IPs are most likely to request the same rate, we enforce
> that the only clks able to modify gck rate are those of audio IPs.
> 
> To remain consistent, we deny other clocks to be children of audio_pll.
> 
> Signed-off-by: Quentin Schulz <quentin.schulz@free-electrons.com>

Acked-by: Boris Brezillon <boris.brezillon@free-electrons.com>

> ---
> 
> v2:
>   - added conditions for audio pll rate setting restriction for SSC and
> I2S,
> 
>  drivers/clk/at91/clk-generated.c | 48 +++++++++++++++++++++++++++++++++++-----
>  1 file changed, 42 insertions(+), 6 deletions(-)
> 
> diff --git a/drivers/clk/at91/clk-generated.c b/drivers/clk/at91/clk-generated.c
> index 6530a2e7e84d..87866786a6ab 100644
> --- a/drivers/clk/at91/clk-generated.c
> +++ b/drivers/clk/at91/clk-generated.c
> @@ -26,6 +26,13 @@
>  #define GENERATED_SOURCE_MAX	6
>  #define GENERATED_MAX_DIV	255
>  
> +#define GCK_ID_SSC0		43
> +#define GCK_ID_SSC1		44
> +#define GCK_ID_I2S0		54
> +#define GCK_ID_I2S1		55
> +#define GCK_ID_CLASSD		59
> +#define GCK_INDEX_DT_AUDIO_PLL	5
> +
>  struct clk_generated {
>  	struct clk_hw hw;
>  	struct regmap *regmap;
> @@ -126,15 +133,14 @@ static int clk_generated_determine_rate(struct clk_hw *hw,
>  {
>  	struct clk_generated *gck = to_clk_generated(hw);
>  	struct clk_hw *parent = NULL;
> +	struct clk_rate_request req_parent = *req;
>  	long best_rate = -EINVAL;
> -	unsigned long min_rate;
> +	unsigned long min_rate, parent_rate;
>  	int best_diff = -1;
>  	int i;
> +	u32 div;
>  
> -	for (i = 0; i < clk_hw_get_num_parents(hw); i++) {
> -		u32 div;
> -		unsigned long parent_rate;
> -
> +	for (i = 0; i < clk_hw_get_num_parents(hw) - 1; i++) {
>  		parent = clk_hw_get_parent_by_index(hw, i);
>  		if (!parent)
>  			continue;
> @@ -150,11 +156,40 @@ static int clk_generated_determine_rate(struct clk_hw *hw,
>  		clk_generated_best_diff(req, parent, parent_rate, div,
>  					&best_diff, &best_rate);
>  
> +		if (!best_diff)
> +			break;
> +	}
> +
> +	/*
> +	 * The audio_pll rate can be modified, unlike the five others clocks
> +	 * that should never be altered.
> +	 * The audio_pll can technically be used by multiple consumers. However,
> +	 * with the rate locking, the first consumer to enable to clock will be
> +	 * the one definitely setting the rate of the clock.
> +	 * Since audio IPs are most likely to request the same rate, we enforce
> +	 * that the only clks able to modify gck rate are those of audio IPs.
> +	 */
> +
> +	if (gck->id != GCK_ID_SSC0 && gck->id != GCK_ID_SSC1 &&
> +	    gck->id != GCK_ID_I2S0 && gck->id != GCK_ID_I2S1 &&
> +	    gck->id != GCK_ID_CLASSD)
> +		goto end;
> +
> +	parent = clk_hw_get_parent_by_index(hw, GCK_INDEX_DT_AUDIO_PLL);
> +	if (!parent)
> +		goto end;
> +
> +	for (div = 1; div < GENERATED_MAX_DIV + 2; div++) {
> +		req_parent.rate = req->rate * div;
> +		__clk_determine_rate(parent, &req_parent);
> +		clk_generated_best_diff(req, parent, req_parent.rate, div,
> +					&best_diff, &best_rate);
>  
>  		if (!best_diff)
>  			break;
>  	}
>  
> +end:
>  	pr_debug("GCLK: %s, best_rate = %ld, parent clk: %s @ %ld\n",
>  		 __func__, best_rate,
>  		 __clk_get_name((req->best_parent_hw)->clk),
> @@ -264,7 +299,8 @@ at91_clk_register_generated(struct regmap *regmap, spinlock_t *lock,
>  	init.ops = &generated_ops;
>  	init.parent_names = parent_names;
>  	init.num_parents = num_parents;
> -	init.flags = CLK_SET_RATE_GATE | CLK_SET_PARENT_GATE;
> +	init.flags = CLK_SET_RATE_GATE | CLK_SET_PARENT_GATE |
> +		CLK_SET_RATE_PARENT;
>  
>  	gck->id = id;
>  	gck->hw.init = &init;

[toc] | [prev] | [next] | [standalone]


#1682512 — Re: [PATCH v2 6/9] clk: at91: clk-generated: make gclk determine audio_pll rate

FromNicolas Ferre <nicolas.ferre@microchip.com>
Date2017-07-06 17:30 +0200
SubjectRe: [PATCH v2 6/9] clk: at91: clk-generated: make gclk determine audio_pll rate
Message-ID<u0gT7-6UV-3@gated-at.bofh.it>
In reply to#1680896
On 04/07/2017 at 13:59, Quentin Schulz wrote:
> This allows gclk to determine audio_pll rate and set the parent rate
> accordingly.
> 
> However, there are multiple children clocks that could technically
> change the rate of audio_pll (via gck). With the rate locking, the first
> consumer to enable the clock will be the one definitely setting the rate
> of the clock.
> 
> Since audio IPs are most likely to request the same rate, we enforce
> that the only clks able to modify gck rate are those of audio IPs.
> 
> To remain consistent, we deny other clocks to be children of audio_pll.
> 
> Signed-off-by: Quentin Schulz <quentin.schulz@free-electrons.com>
> ---
> 
> v2:
>   - added conditions for audio pll rate setting restriction for SSC and
> I2S,
> 
>  drivers/clk/at91/clk-generated.c | 48 +++++++++++++++++++++++++++++++++++-----
>  1 file changed, 42 insertions(+), 6 deletions(-)
> 
> diff --git a/drivers/clk/at91/clk-generated.c b/drivers/clk/at91/clk-generated.c
> index 6530a2e7e84d..87866786a6ab 100644
> --- a/drivers/clk/at91/clk-generated.c
> +++ b/drivers/clk/at91/clk-generated.c
> @@ -26,6 +26,13 @@
>  #define GENERATED_SOURCE_MAX	6
>  #define GENERATED_MAX_DIV	255
>  
> +#define GCK_ID_SSC0		43
> +#define GCK_ID_SSC1		44
> +#define GCK_ID_I2S0		54
> +#define GCK_ID_I2S1		55
> +#define GCK_ID_CLASSD		59
> +#define GCK_INDEX_DT_AUDIO_PLL	5

hum...

> +
>  struct clk_generated {
>  	struct clk_hw hw;
>  	struct regmap *regmap;
> @@ -126,15 +133,14 @@ static int clk_generated_determine_rate(struct clk_hw *hw,
>  {
>  	struct clk_generated *gck = to_clk_generated(hw);
>  	struct clk_hw *parent = NULL;
> +	struct clk_rate_request req_parent = *req;
>  	long best_rate = -EINVAL;
> -	unsigned long min_rate;
> +	unsigned long min_rate, parent_rate;
>  	int best_diff = -1;
>  	int i;
> +	u32 div;
>  
> -	for (i = 0; i < clk_hw_get_num_parents(hw); i++) {
> -		u32 div;
> -		unsigned long parent_rate;
> -
> +	for (i = 0; i < clk_hw_get_num_parents(hw) - 1; i++) {
>  		parent = clk_hw_get_parent_by_index(hw, i);
>  		if (!parent)
>  			continue;
> @@ -150,11 +156,40 @@ static int clk_generated_determine_rate(struct clk_hw *hw,
>  		clk_generated_best_diff(req, parent, parent_rate, div,
>  					&best_diff, &best_rate);
>  
> +		if (!best_diff)
> +			break;
> +	}
> +
> +	/*
> +	 * The audio_pll rate can be modified, unlike the five others clocks
> +	 * that should never be altered.
> +	 * The audio_pll can technically be used by multiple consumers. However,
> +	 * with the rate locking, the first consumer to enable to clock will be
> +	 * the one definitely setting the rate of the clock.
> +	 * Since audio IPs are most likely to request the same rate, we enforce
> +	 * that the only clks able to modify gck rate are those of audio IPs.
> +	 */
> +
> +	if (gck->id != GCK_ID_SSC0 && gck->id != GCK_ID_SSC1 &&
> +	    gck->id != GCK_ID_I2S0 && gck->id != GCK_ID_I2S1 &&
> +	    gck->id != GCK_ID_CLASSD)
> +		goto end;

...well... maybe for a temporary solution, but what about the next
product that will also have such feature but without the same set of IDs?

I know that the main use of this clock is the "audio" subsystem (so its
name) but in a machine without audio needs, it can be interesting to use
this fractional PLL for other devices that can handle a GCK (and I'm
thinking about the SDHCI interface that then could be configured with
maximum clock speed).

> +	parent = clk_hw_get_parent_by_index(hw, GCK_INDEX_DT_AUDIO_PLL);
> +	if (!parent)
> +		goto end;

Here, if the audio pll id changes in the DT, we don't retrieve the
proper parent... And again, not very future proof.

> +	for (div = 1; div < GENERATED_MAX_DIV + 2; div++) {
> +		req_parent.rate = req->rate * div;
> +		__clk_determine_rate(parent, &req_parent);
> +		clk_generated_best_diff(req, parent, req_parent.rate, div,
> +					&best_diff, &best_rate);
>  
>  		if (!best_diff)
>  			break;
>  	}
>  
> +end:
>  	pr_debug("GCLK: %s, best_rate = %ld, parent clk: %s @ %ld\n",
>  		 __func__, best_rate,
>  		 __clk_get_name((req->best_parent_hw)->clk),
> @@ -264,7 +299,8 @@ at91_clk_register_generated(struct regmap *regmap, spinlock_t *lock,
>  	init.ops = &generated_ops;
>  	init.parent_names = parent_names;
>  	init.num_parents = num_parents;
> -	init.flags = CLK_SET_RATE_GATE | CLK_SET_PARENT_GATE;
> +	init.flags = CLK_SET_RATE_GATE | CLK_SET_PARENT_GATE |
> +		CLK_SET_RATE_PARENT;
>  
>  	gck->id = id;
>  	gck->hw.init = &init;
> 


-- 
Nicolas Ferre

[toc] | [prev] | [next] | [standalone]


#1682549 — Re: [PATCH v2 6/9] clk: at91: clk-generated: make gclk determine audio_pll rate

FromAlexandre Belloni <alexandre.belloni@free-electrons.com>
Date2017-07-06 18:10 +0200
SubjectRe: [PATCH v2 6/9] clk: at91: clk-generated: make gclk determine audio_pll rate
Message-ID<u0hvR-7HB-31@gated-at.bofh.it>
In reply to#1682512
On 06/07/2017 at 17:29:22 +0200, Nicolas Ferre wrote:
> > +	/*
> > +	 * The audio_pll rate can be modified, unlike the five others clocks
> > +	 * that should never be altered.
> > +	 * The audio_pll can technically be used by multiple consumers. However,
> > +	 * with the rate locking, the first consumer to enable to clock will be
> > +	 * the one definitely setting the rate of the clock.
> > +	 * Since audio IPs are most likely to request the same rate, we enforce
> > +	 * that the only clks able to modify gck rate are those of audio IPs.
> > +	 */
> > +
> > +	if (gck->id != GCK_ID_SSC0 && gck->id != GCK_ID_SSC1 &&
> > +	    gck->id != GCK_ID_I2S0 && gck->id != GCK_ID_I2S1 &&
> > +	    gck->id != GCK_ID_CLASSD)
> > +		goto end;
> 
> ...well... maybe for a temporary solution, but what about the next
> product that will also have such feature but without the same set of IDs?
> 

It will have a different compatible and that can be used to encode the
difference.

> I know that the main use of this clock is the "audio" subsystem (so its
> name) but in a machine without audio needs, it can be interesting to use
> this fractional PLL for other devices that can handle a GCK (and I'm
> thinking about the SDHCI interface that then could be configured with
> maximum clock speed).
> 
> > +	parent = clk_hw_get_parent_by_index(hw, GCK_INDEX_DT_AUDIO_PLL);
> > +	if (!parent)
> > +		goto end;
> 
> Here, if the audio pll id changes in the DT, we don't retrieve the
> proper parent... And again, not very future proof.
> 

One more reason to remove all the crap we have in the DT, as suggested
by RobH.


-- 
Alexandre Belloni, Free Electrons
Embedded Linux and Kernel engineering
http://free-electrons.com

[toc] | [prev] | [standalone]


Back to top | Article view | linux.kernel


csiph-web