linux-arm-kernel.lists.infradead.org archive mirror
 help / color / mirror / Atom feed
* [PATCH 2/3] clk: conf: Support assigned-clock-sscs
  2025-01-24 14:25 Peng Fan (OSS)
@ 2025-01-24 14:25 ` Peng Fan (OSS)
  0 siblings, 0 replies; 11+ messages in thread
From: Peng Fan (OSS) @ 2025-01-24 14:25 UTC (permalink / raw)
  To: Michael Turquette, Stephen Boyd, Russell King, Sudeep Holla,
	Cristian Marussi
  Cc: linux-clk, linux-kernel, arm-scmi, linux-arm-kernel, Rob Herring,
	Krzysztof Kozlowski, Dario Binacchi, Shawn Guo, Sascha Hauer,
	Pengutronix Kernel Team, Fabio Estevam, imx, Peng Fan

From: Peng Fan <peng.fan@nxp.com>

Parse the Spread Spectrum Configuration(SSC) from device tree and configure
them before using the clock.

Each SSC is three u32 elements which means '<modfreq spreadpercent
modmethod>', so assigned-clock-sscs is an array of multiple three u32
elements.

Signed-off-by: Peng Fan <peng.fan@nxp.com>
---
 drivers/clk/clk-conf.c | 68 ++++++++++++++++++++++++++++++++++++++++++++++++++
 1 file changed, 68 insertions(+)

diff --git a/drivers/clk/clk-conf.c b/drivers/clk/clk-conf.c
index 303a0bb26e54a95655ce094a35b989c97ebc6fd8..1ca4222caa829223d65dd37c7a776d43dc421944 100644
--- a/drivers/clk/clk-conf.c
+++ b/drivers/clk/clk-conf.c
@@ -155,6 +155,70 @@ static int __set_clk_rates(struct device_node *node, bool clk_supplier)
 	return 0;
 }
 
+static int __set_clk_spread_spectrum(struct device_node *node, bool clk_supplier)
+{
+	u32 *sscs __free(kfree) = NULL;
+	u32 elem_size = sizeof(u32) * 3;
+	struct of_phandle_args clkspec;
+	int rc, count, index;
+	struct clk *clk;
+
+	/* modfreq, spreadPercent, modmethod */
+	count = of_property_count_elems_of_size(node, "assigned-clock-sscs", elem_size);
+	if (count > 0) {
+		sscs = kcalloc(count, elem_size, GFP_KERNEL);
+		if (!sscs)
+			return -ENOMEM;
+		rc = of_property_read_u32_array(node,
+						"assigned-clock-sscs",
+						sscs, count * 3);
+	} else {
+		return 0;
+	}
+
+	if (rc)
+		return rc;
+
+	for (index = 0; index < count; index++) {
+		u32 modfreq = sscs[index * 3], spreadpercent = sscs[index * 3 + 1];
+		u32 method = sscs[index * 3 + 2];
+
+		if (modfreq || spreadpercent || method) {
+			rc = of_parse_phandle_with_args(node, "assigned-clocks",
+					"#clock-cells",	index, &clkspec);
+			if (rc < 0) {
+				/* skip empty (null) phandles */
+				if (rc == -ENOENT)
+					continue;
+				else
+					return rc;
+			}
+
+			if (clkspec.np == node && !clk_supplier) {
+				of_node_put(clkspec.np);
+				return 0;
+			}
+
+			clk = of_clk_get_from_provider(&clkspec);
+			of_node_put(clkspec.np);
+			if (IS_ERR(clk)) {
+				if (PTR_ERR(clk) != -EPROBE_DEFER)
+					pr_warn("clk: couldn't get clock %d for %pOF\n",
+						index, node);
+				return PTR_ERR(clk);
+			}
+
+			rc = clk_set_spread_spectrum(clk, modfreq, spreadpercent, method, true);
+			if (rc < 0)
+				pr_err("clk: couldn't set %s clk spread spectrum %u %u %u: %d\n",
+				       __clk_get_name(clk), modfreq, spreadpercent, method, rc);
+			clk_put(clk);
+		}
+	}
+
+	return 0;
+}
+
 /**
  * of_clk_set_defaults() - parse and set assigned clocks configuration
  * @node: device node to apply clock settings for
@@ -174,6 +238,10 @@ int of_clk_set_defaults(struct device_node *node, bool clk_supplier)
 	if (!node)
 		return 0;
 
+	rc = __set_clk_spread_spectrum(node, clk_supplier);
+	if (rc < 0)
+		return rc;
+
 	rc = __set_clk_parents(node, clk_supplier);
 	if (rc < 0)
 		return rc;

-- 
2.35.3



^ permalink raw reply related	[flat|nested] 11+ messages in thread

* [PATCH 0/3] clk: Support spread spectrum and use it in clk-scmi
@ 2025-08-12 12:17 Peng Fan
  2025-08-12 12:17 ` [PATCH 1/3] clk: Introduce clk_hw_set_spread_spectrum Peng Fan
                   ` (3 more replies)
  0 siblings, 4 replies; 11+ messages in thread
From: Peng Fan @ 2025-08-12 12:17 UTC (permalink / raw)
  To: Michael Turquette, Stephen Boyd, Sudeep Holla, Cristian Marussi,
	Marco Felsch
  Cc: Geert Uytterhoeven, linux-clk, linux-kernel, arm-scmi,
	linux-arm-kernel, Peng Fan

Since the assigned-clock-sscs property [1] has been accepted into the device
tree schema, we can now support it in the Linux clock driver. Therefore,
I’ve picked up the previously submitted work [2] titled “clk: Support
spread spectrum and use it in clk-pll144x and clk-scmi.”
As more than six months have passed since [2] was posted, I’m treating this
patchset as a new submission rather than a v3.

- Introduce clk_set_spread_spectrum to set the parameters for enabling
  spread spectrum of a clock.
- Parse 'assigned-clock-sscs' and configure it by default before using the
  clock. This property is parsed before parsing clock rate.
- Enable this feature for clk-scmi on i.MX95.

Because SCMI spec will not include spread spectrum as a standard
extension, we still need to use NXP i.MX OEM extension.

[1] https://github.com/devicetree-org/dt-schema/pull/154
[2] https://lore.kernel.org/all/20250205-clk-ssc-v2-0-fa73083caa92@nxp.com/

Signed-off-by: Peng Fan <peng.fan@nxp.com>
---
Peng Fan (3):
      clk: Introduce clk_hw_set_spread_spectrum
      clk: conf: Support assigned-clock-sscs
      clk: scmi: Support Spread Spectrum for NXP i.MX95

 drivers/clk/clk-conf.c        | 70 +++++++++++++++++++++++++++++++++++++++++++
 drivers/clk/clk-scmi.c        | 64 ++++++++++++++++++++++++++++++++++++---
 drivers/clk/clk.c             | 32 ++++++++++++++++++++
 include/linux/clk-provider.h  | 29 ++++++++++++++++++
 include/linux/scmi_protocol.h |  5 ++++
 5 files changed, 196 insertions(+), 4 deletions(-)
---
base-commit: b1549501188cc9eba732c25b033df7a53ccc341f
change-id: 20250812-clk-ssc-version1-acf6f6efbd96

Best regards,
-- 
Peng Fan <peng.fan@nxp.com>



^ permalink raw reply	[flat|nested] 11+ messages in thread

* [PATCH 1/3] clk: Introduce clk_hw_set_spread_spectrum
  2025-08-12 12:17 [PATCH 0/3] clk: Support spread spectrum and use it in clk-scmi Peng Fan
@ 2025-08-12 12:17 ` Peng Fan
  2025-08-27 15:45   ` Brian Masney
  2025-08-12 12:17 ` [PATCH 2/3] clk: conf: Support assigned-clock-sscs Peng Fan
                   ` (2 subsequent siblings)
  3 siblings, 1 reply; 11+ messages in thread
From: Peng Fan @ 2025-08-12 12:17 UTC (permalink / raw)
  To: Michael Turquette, Stephen Boyd, Sudeep Holla, Cristian Marussi,
	Marco Felsch
  Cc: Geert Uytterhoeven, linux-clk, linux-kernel, arm-scmi,
	linux-arm-kernel, Peng Fan

Add clk_hw_set_spread_spectrum to configure a clock to enable spread
spectrum feature. set_spread_spectrum ops is added for clk drivers to
have their own hardware specific implementation.

Signed-off-by: Peng Fan <peng.fan@nxp.com>
---
 drivers/clk/clk.c            | 32 ++++++++++++++++++++++++++++++++
 include/linux/clk-provider.h | 29 +++++++++++++++++++++++++++++
 2 files changed, 61 insertions(+)

diff --git a/drivers/clk/clk.c b/drivers/clk/clk.c
index b821b2cdb155331c85fafbd2fac8ab3703a08e4d..48c7a301b72b30fd824dae7ada2c44ee84d40867 100644
--- a/drivers/clk/clk.c
+++ b/drivers/clk/clk.c
@@ -2802,6 +2802,38 @@ int clk_set_max_rate(struct clk *clk, unsigned long rate)
 }
 EXPORT_SYMBOL_GPL(clk_set_max_rate);
 
+int clk_hw_set_spread_spectrum(struct clk_hw *hw, unsigned int modfreq_hz,
+			       unsigned int spread_bp, unsigned int method)
+{
+	struct clk_spread_spectrum clk_ss;
+	struct clk_core *core;
+	int ret;
+
+	if (!hw)
+		return 0;
+
+	core = hw->core;
+
+	clk_ss.modfreq_hz = modfreq_hz;
+	clk_ss.spread_bp = spread_bp;
+	clk_ss.method = method;
+
+	clk_prepare_lock();
+
+	ret = clk_pm_runtime_get(core);
+	if (ret)
+		goto fail;
+
+	if (core->ops->set_spread_spectrum)
+		ret = core->ops->set_spread_spectrum(hw, &clk_ss);
+
+	clk_pm_runtime_put(core);
+
+fail:
+	clk_prepare_unlock();
+	return ret;
+}
+
 /**
  * clk_get_parent - return the parent of a clk
  * @clk: the clk whose parent gets returned
diff --git a/include/linux/clk-provider.h b/include/linux/clk-provider.h
index 630705a47129453c241f1b1755f2c2f2a7ed8f77..2b6cebe8b12268f537b3c92aa0bbadea601f0eb0 100644
--- a/include/linux/clk-provider.h
+++ b/include/linux/clk-provider.h
@@ -84,6 +84,26 @@ struct clk_duty {
 	unsigned int den;
 };
 
+enum clk_ssc_method {
+	CLK_SSC_NO_SPREAD,
+	CLK_SSC_CENTER_SPREAD,
+	CLK_SSC_UP_SPREAD,
+	CLK_SSC_DOWN_SPREAD,
+};
+
+/**
+ * struct clk_spread_spectrum - Structure encoding spread spectrum of a clock
+ *
+ * @modfreq_hz:		Modulation frequency
+ * @spread_bp:		Modulation percent in permyriad
+ * @method:		Modulation method
+ */
+struct clk_spread_spectrum {
+	unsigned int modfreq_hz;
+	unsigned int spread_bp;
+	enum clk_ssc_method method;
+};
+
 /**
  * struct clk_ops -  Callback operations for hardware clocks; these are to
  * be provided by the clock implementation, and will be called by drivers
@@ -178,6 +198,11 @@ struct clk_duty {
  *		separately via calls to .set_parent and .set_rate.
  *		Returns 0 on success, -EERROR otherwise.
  *
+ * @set_spread_spectrum: Configure the modulation frequency, modulation percentage
+ *		and method. This callback is optional for clocks that does not
+ *		support spread spectrum feature or no need to enable this feature.
+ *		Returns 0 on success, -EERROR otherwise.
+ *
  * @recalc_accuracy: Recalculate the accuracy of this clock. The clock accuracy
  *		is expressed in ppb (parts per billion). The parent accuracy is
  *		an input parameter.
@@ -255,6 +280,8 @@ struct clk_ops {
 	int		(*set_rate_and_parent)(struct clk_hw *hw,
 				    unsigned long rate,
 				    unsigned long parent_rate, u8 index);
+	int		(*set_spread_spectrum)(struct clk_hw *hw,
+					       struct clk_spread_spectrum *clk_ss);
 	unsigned long	(*recalc_accuracy)(struct clk_hw *hw,
 					   unsigned long parent_accuracy);
 	int		(*get_phase)(struct clk_hw *hw);
@@ -1430,6 +1457,8 @@ void clk_hw_get_rate_range(struct clk_hw *hw, unsigned long *min_rate,
 			   unsigned long *max_rate);
 void clk_hw_set_rate_range(struct clk_hw *hw, unsigned long min_rate,
 			   unsigned long max_rate);
+int clk_hw_set_spread_spectrum(struct clk_hw *hw, unsigned int modfreq_hz,
+			       unsigned int spread_bp, unsigned int method);
 
 static inline void __clk_hw_set_clk(struct clk_hw *dst, struct clk_hw *src)
 {

-- 
2.37.1



^ permalink raw reply related	[flat|nested] 11+ messages in thread

* [PATCH 2/3] clk: conf: Support assigned-clock-sscs
  2025-08-12 12:17 [PATCH 0/3] clk: Support spread spectrum and use it in clk-scmi Peng Fan
  2025-08-12 12:17 ` [PATCH 1/3] clk: Introduce clk_hw_set_spread_spectrum Peng Fan
@ 2025-08-12 12:17 ` Peng Fan
  2025-08-13  5:48   ` Dan Carpenter
  2025-08-12 12:17 ` [PATCH 3/3] clk: scmi: Support Spread Spectrum for NXP i.MX95 Peng Fan
  2025-08-27  7:05 ` [PATCH 0/3] clk: Support spread spectrum and use it in clk-scmi Peng Fan
  3 siblings, 1 reply; 11+ messages in thread
From: Peng Fan @ 2025-08-12 12:17 UTC (permalink / raw)
  To: Michael Turquette, Stephen Boyd, Sudeep Holla, Cristian Marussi,
	Marco Felsch
  Cc: Geert Uytterhoeven, linux-clk, linux-kernel, arm-scmi,
	linux-arm-kernel, Peng Fan

Parse the Spread Spectrum Configuration(SSC) from device tree and configure
them before using the clock.

Each SSC is three u32 elements which means '<modfreq spreaddepth
modmethod>', so assigned-clock-sscs is an array of multiple three u32
elements.

Signed-off-by: Peng Fan <peng.fan@nxp.com>
---
 drivers/clk/clk-conf.c | 70 ++++++++++++++++++++++++++++++++++++++++++++++++++
 1 file changed, 70 insertions(+)

diff --git a/drivers/clk/clk-conf.c b/drivers/clk/clk-conf.c
index 303a0bb26e54a95655ce094a35b989c97ebc6fd8..81a2c1f8ca4c44df2c54c1e51f800f533c9453b3 100644
--- a/drivers/clk/clk-conf.c
+++ b/drivers/clk/clk-conf.c
@@ -155,6 +155,72 @@ static int __set_clk_rates(struct device_node *node, bool clk_supplier)
 	return 0;
 }
 
+static int __set_clk_spread_spectrum(struct device_node *node, bool clk_supplier)
+{
+	u32 *sscs __free(kfree) = NULL;
+	u32 elem_size = sizeof(u32) * 3;
+	struct of_phandle_args clkspec;
+	int rc, count, index;
+	struct clk *clk;
+
+	/* modfreq, spreadPercent, modmethod */
+	count = of_property_count_elems_of_size(node, "assigned-clock-sscs", elem_size);
+	if (count > 0) {
+		sscs = kcalloc(count, elem_size, GFP_KERNEL);
+		if (!sscs)
+			return -ENOMEM;
+		rc = of_property_read_u32_array(node,
+						"assigned-clock-sscs",
+						sscs, count * 3);
+	} else {
+		return 0;
+	}
+
+	if (rc)
+		return rc;
+
+	for (index = 0; index < count; index++) {
+		u32 modfreq_hz = sscs[index * 3], spread_bp = sscs[index * 3 + 1];
+		u32 method = sscs[index * 3 + 2];
+		struct clk_hw *hw;
+
+		if (modfreq_hz || spread_bp || method) {
+			rc = of_parse_phandle_with_args(node, "assigned-clocks",
+					"#clock-cells",	index, &clkspec);
+			if (rc < 0) {
+				/* skip empty (null) phandles */
+				if (rc == -ENOENT)
+					continue;
+				else
+					return rc;
+			}
+
+			if (clkspec.np == node && !clk_supplier) {
+				of_node_put(clkspec.np);
+				return 0;
+			}
+
+			clk = of_clk_get_from_provider(&clkspec);
+			of_node_put(clkspec.np);
+			if (IS_ERR(clk)) {
+				if (PTR_ERR(clk) != -EPROBE_DEFER)
+					pr_warn("clk: couldn't get clock %d for %pOF\n",
+						index, node);
+				return PTR_ERR(clk);
+			}
+
+			hw = __clk_get_hw(clk);
+			rc = clk_hw_set_spread_spectrum(hw, modfreq_hz, spread_bp, method);
+			if (rc < 0)
+				pr_err("clk: couldn't set %s clk spread spectrum %u %u %u: %d\n",
+				       __clk_get_name(clk), modfreq_hz, spread_bp, method, rc);
+			clk_put(clk);
+		}
+	}
+
+	return 0;
+}
+
 /**
  * of_clk_set_defaults() - parse and set assigned clocks configuration
  * @node: device node to apply clock settings for
@@ -174,6 +240,10 @@ int of_clk_set_defaults(struct device_node *node, bool clk_supplier)
 	if (!node)
 		return 0;
 
+	rc = __set_clk_spread_spectrum(node, clk_supplier);
+	if (rc < 0)
+		return rc;
+
 	rc = __set_clk_parents(node, clk_supplier);
 	if (rc < 0)
 		return rc;

-- 
2.37.1



^ permalink raw reply related	[flat|nested] 11+ messages in thread

* [PATCH 3/3] clk: scmi: Support Spread Spectrum for NXP i.MX95
  2025-08-12 12:17 [PATCH 0/3] clk: Support spread spectrum and use it in clk-scmi Peng Fan
  2025-08-12 12:17 ` [PATCH 1/3] clk: Introduce clk_hw_set_spread_spectrum Peng Fan
  2025-08-12 12:17 ` [PATCH 2/3] clk: conf: Support assigned-clock-sscs Peng Fan
@ 2025-08-12 12:17 ` Peng Fan
  2025-08-27  7:05 ` [PATCH 0/3] clk: Support spread spectrum and use it in clk-scmi Peng Fan
  3 siblings, 0 replies; 11+ messages in thread
From: Peng Fan @ 2025-08-12 12:17 UTC (permalink / raw)
  To: Michael Turquette, Stephen Boyd, Sudeep Holla, Cristian Marussi,
	Marco Felsch
  Cc: Geert Uytterhoeven, linux-clk, linux-kernel, arm-scmi,
	linux-arm-kernel, Peng Fan

The PLL clocks on NXP i.MX95 SoCs support Spread Spectrum (SS).
This patch introduces scmi_clk_imx_set_spread_spectrum to pass SS
configuration to the SCMI firmware, which handles the actual
implementation.

To ensure this feature is only enabled on NXP platforms,
scmi_clk_imx_extended_config_oem is added. Since SS is only applicable
to PLL clocks, config_oem_get is used to verify SS support for a given
clock.

i.MX95 SCMI firmware Spread Spectrum extConfigValue definition is as
below, no modulation method because firmware forces to use down spread.
	 extConfigValue[7:0]   - spread percentage (%)
	 extConfigValue[23:8]  - Modulation Frequency (KHz)
	 extConfigValue[24]    - Enable/Disable
	 extConfigValue[31:25] - Reserved

Signed-off-by: Peng Fan <peng.fan@nxp.com>
---
 drivers/clk/clk-scmi.c        | 64 ++++++++++++++++++++++++++++++++++++++++---
 include/linux/scmi_protocol.h |  5 ++++
 2 files changed, 65 insertions(+), 4 deletions(-)

diff --git a/drivers/clk/clk-scmi.c b/drivers/clk/clk-scmi.c
index d2408403283fc72f0cf902e65f4c08bcbc7b4b0b..bb5e20dab18e92932ab4b99192b496e0c4d96417 100644
--- a/drivers/clk/clk-scmi.c
+++ b/drivers/clk/clk-scmi.c
@@ -12,6 +12,7 @@
 #include <linux/of.h>
 #include <linux/module.h>
 #include <linux/scmi_protocol.h>
+#include <linux/scmi_imx_protocol.h>
 #include <asm/div64.h>
 
 #define NOT_ATOMIC	false
@@ -23,6 +24,7 @@ enum scmi_clk_feats {
 	SCMI_CLK_RATE_CTRL_SUPPORTED,
 	SCMI_CLK_PARENT_CTRL_SUPPORTED,
 	SCMI_CLK_DUTY_CYCLE_SUPPORTED,
+	SCMI_CLK_IMX_SSC_SUPPORTED,
 	SCMI_CLK_FEATS_COUNT
 };
 
@@ -98,6 +100,35 @@ static int scmi_clk_set_parent(struct clk_hw *hw, u8 parent_index)
 	return scmi_proto_clk_ops->parent_set(clk->ph, clk->id, parent_index);
 }
 
+static int scmi_clk_imx_set_spread_spectrum(struct clk_hw *hw,
+					    struct clk_spread_spectrum *clk_ss)
+{
+	struct scmi_clk *clk = to_scmi_clk(hw);
+	int ret;
+	u32 val;
+
+	/*
+	 * extConfigValue[7:0]   - spread percentage (%)
+	 * extConfigValue[23:8]  - Modulation Frequency
+	 * extConfigValue[24]    - Enable/Disable
+	 * extConfigValue[31:25] - Reserved
+	 */
+	val = FIELD_PREP(SCMI_CLOCK_IMX_SS_PERCENTAGE_MASK, clk_ss->spread_bp / 10000);
+	val |= FIELD_PREP(SCMI_CLOCK_IMX_SS_MOD_FREQ_MASK, clk_ss->modfreq_hz);
+	if (clk_ss->method != CLK_SSC_NO_SPREAD)
+		val |= SCMI_CLOCK_IMX_SS_ENABLE_MASK;
+	ret = scmi_proto_clk_ops->config_oem_set(clk->ph, clk->id,
+						 SCMI_CLOCK_CFG_IMX_SSC,
+						 val, false);
+	if (ret)
+		dev_warn(clk->dev,
+			 "Failed to set spread spectrum(%u,%u,%u) for clock ID %d\n",
+			 clk_ss->modfreq_hz, clk_ss->spread_bp, clk_ss->method,
+			 clk->id);
+
+	return ret;
+}
+
 static u8 scmi_clk_get_parent(struct clk_hw *hw)
 {
 	struct scmi_clk *clk = to_scmi_clk(hw);
@@ -316,11 +347,33 @@ scmi_clk_ops_alloc(struct device *dev, unsigned long feats_key)
 		ops->set_duty_cycle = scmi_clk_set_duty_cycle;
 	}
 
+	if (feats_key & BIT(SCMI_CLK_IMX_SSC_SUPPORTED))
+		ops->set_spread_spectrum = scmi_clk_imx_set_spread_spectrum;
+
 	return ops;
 }
 
+static void scmi_clk_imx_extended_config_oem(const struct scmi_handle *handle,
+					     struct scmi_clk *sclk,
+					     unsigned int *feats_key)
+{
+	int ret;
+	u32 val;
+
+	if (strcmp(handle->version->vendor_id, SCMI_IMX_VENDOR) ||
+	    strcmp(handle->version->sub_vendor_id, SCMI_IMX_SUBVENDOR))
+		return;
+
+	ret = scmi_proto_clk_ops->config_oem_get(sclk->ph, sclk->id,
+						 SCMI_CLOCK_CFG_IMX_SSC,
+						 &val, NULL, false);
+	if (!ret)
+		*feats_key |= BIT(SCMI_CLK_IMX_SSC_SUPPORTED);
+}
+
 /**
  * scmi_clk_ops_select() - Select a proper set of clock operations
+ * @handle: A reference to an SCMI entity
  * @sclk: A reference to an SCMI clock descriptor
  * @atomic_capable: A flag to indicate if atomic mode is supported by the
  *		    transport
@@ -345,8 +398,8 @@ scmi_clk_ops_alloc(struct device *dev, unsigned long feats_key)
  *	   NULL otherwise.
  */
 static const struct clk_ops *
-scmi_clk_ops_select(struct scmi_clk *sclk, bool atomic_capable,
-		    unsigned int atomic_threshold_us,
+scmi_clk_ops_select(const struct scmi_handle *handle, struct scmi_clk *sclk,
+		    bool atomic_capable, unsigned int atomic_threshold_us,
 		    const struct clk_ops **clk_ops_db, size_t db_size)
 {
 	const struct scmi_clock_info *ci = sclk->info;
@@ -370,9 +423,12 @@ scmi_clk_ops_select(struct scmi_clk *sclk, bool atomic_capable,
 	if (!ci->parent_ctrl_forbidden)
 		feats_key |= BIT(SCMI_CLK_PARENT_CTRL_SUPPORTED);
 
-	if (ci->extended_config)
+	if (ci->extended_config) {
 		feats_key |= BIT(SCMI_CLK_DUTY_CYCLE_SUPPORTED);
 
+		scmi_clk_imx_extended_config_oem(handle, sclk, &feats_key);
+	}
+
 	if (WARN_ON(feats_key >= db_size))
 		return NULL;
 
@@ -459,7 +515,7 @@ static int scmi_clocks_probe(struct scmi_device *sdev)
 		 * to avoid sharing the devm_ allocated clk_ops between multiple
 		 * SCMI clk driver instances.
 		 */
-		scmi_ops = scmi_clk_ops_select(sclk, transport_is_atomic,
+		scmi_ops = scmi_clk_ops_select(handle, sclk, transport_is_atomic,
 					       atomic_threshold_us,
 					       scmi_clk_ops_db,
 					       ARRAY_SIZE(scmi_clk_ops_db));
diff --git a/include/linux/scmi_protocol.h b/include/linux/scmi_protocol.h
index 688466a0e816247d24704f7ba109667a14226b67..6f9f197ee671540e38470a5666eb5ba8ec9de439 100644
--- a/include/linux/scmi_protocol.h
+++ b/include/linux/scmi_protocol.h
@@ -80,9 +80,14 @@ enum scmi_clock_oem_config {
 	SCMI_CLOCK_CFG_DUTY_CYCLE = 0x1,
 	SCMI_CLOCK_CFG_PHASE,
 	SCMI_CLOCK_CFG_OEM_START = 0x80,
+	SCMI_CLOCK_CFG_IMX_SSC = 0x80,
 	SCMI_CLOCK_CFG_OEM_END = 0xFF,
 };
 
+#define SCMI_CLOCK_IMX_SS_PERCENTAGE_MASK	GENMASK(7, 0)
+#define SCMI_CLOCK_IMX_SS_MOD_FREQ_MASK		GENMASK(23, 8)
+#define SCMI_CLOCK_IMX_SS_ENABLE_MASK		BIT(24)
+
 /**
  * struct scmi_clk_proto_ops - represents the various operations provided
  *	by SCMI Clock Protocol

-- 
2.37.1



^ permalink raw reply related	[flat|nested] 11+ messages in thread

* Re: [PATCH 2/3] clk: conf: Support assigned-clock-sscs
  2025-08-12 12:17 ` [PATCH 2/3] clk: conf: Support assigned-clock-sscs Peng Fan
@ 2025-08-13  5:48   ` Dan Carpenter
  2025-08-15  8:50     ` Peng Fan
  0 siblings, 1 reply; 11+ messages in thread
From: Dan Carpenter @ 2025-08-13  5:48 UTC (permalink / raw)
  To: Peng Fan
  Cc: Michael Turquette, Stephen Boyd, Sudeep Holla, Cristian Marussi,
	Marco Felsch, Geert Uytterhoeven, linux-clk, linux-kernel,
	arm-scmi, linux-arm-kernel

On Tue, Aug 12, 2025 at 08:17:06PM +0800, Peng Fan wrote:
> Parse the Spread Spectrum Configuration(SSC) from device tree and configure
> them before using the clock.
> 
> Each SSC is three u32 elements which means '<modfreq spreaddepth
> modmethod>', so assigned-clock-sscs is an array of multiple three u32
> elements.
> 
> Signed-off-by: Peng Fan <peng.fan@nxp.com>
> ---
>  drivers/clk/clk-conf.c | 70 ++++++++++++++++++++++++++++++++++++++++++++++++++
>  1 file changed, 70 insertions(+)
> 
> diff --git a/drivers/clk/clk-conf.c b/drivers/clk/clk-conf.c
> index 303a0bb26e54a95655ce094a35b989c97ebc6fd8..81a2c1f8ca4c44df2c54c1e51f800f533c9453b3 100644
> --- a/drivers/clk/clk-conf.c
> +++ b/drivers/clk/clk-conf.c
> @@ -155,6 +155,72 @@ static int __set_clk_rates(struct device_node *node, bool clk_supplier)
>  	return 0;
>  }
>  
> +static int __set_clk_spread_spectrum(struct device_node *node, bool clk_supplier)
> +{
> +	u32 *sscs __free(kfree) = NULL;
> +	u32 elem_size = sizeof(u32) * 3;
> +	struct of_phandle_args clkspec;
> +	int rc, count, index;
> +	struct clk *clk;
> +
> +	/* modfreq, spreadPercent, modmethod */
> +	count = of_property_count_elems_of_size(node, "assigned-clock-sscs", elem_size);
> +	if (count > 0) {
> +		sscs = kcalloc(count, elem_size, GFP_KERNEL);
> +		if (!sscs)
> +			return -ENOMEM;
> +		rc = of_property_read_u32_array(node,
> +						"assigned-clock-sscs",
> +						sscs, count * 3);
> +	} else {
> +		return 0;
> +	}
> +
> +	if (rc)
> +		return rc;

Nit pick: Please, flip these conditions around.

	count = of_property_count_elems_of_size(node, "assigned-clock-sscs", elem_size);
	if (count <= 0)
		return 0;

	sscs = kcalloc(count, elem_size, GFP_KERNEL);
	if (!sscs)
		return -ENOMEM;

	rc = of_property_read_u32_array(node, "assigned-clock-sscs", sscs,
					count * 3);
	if (rc)
		return rc;

> +
> +	for (index = 0; index < count; index++) {
> +		u32 modfreq_hz = sscs[index * 3], spread_bp = sscs[index * 3 + 1];
> +		u32 method = sscs[index * 3 + 2];

This math would be nicer if you created a struct:

struct spread_config {
	u32 modfreq_hz;
	u32 spread_depth;
	u32 method;
};

Then you could use that instead of sscs.

	for (i = 0; i < count; i++) {
		struct spread_config *conf = &configs[i];
		struct clk_hw *hw;

		if (!conf->modfreq_hz && !conf->spread_depth && !conf->method)
			continue;

> +		struct clk_hw *hw;
> +
> +		if (modfreq_hz || spread_bp || method) {
> +			rc = of_parse_phandle_with_args(node, "assigned-clocks",
> +					"#clock-cells",	index, &clkspec);
> +			if (rc < 0) {
> +				/* skip empty (null) phandles */
> +				if (rc == -ENOENT)
> +					continue;
> +				else
> +					return rc;
> +			}
> +
> +			if (clkspec.np == node && !clk_supplier) {

Could you add a comment for this condition?  It's strange to me that we
don't iterate through the whole array.

regards,
dan carpenter

> +				of_node_put(clkspec.np);
> +				return 0;
> +			}
> +
> +			clk = of_clk_get_from_provider(&clkspec);
> +			of_node_put(clkspec.np);
> +			if (IS_ERR(clk)) {
> +				if (PTR_ERR(clk) != -EPROBE_DEFER)
> +					pr_warn("clk: couldn't get clock %d for %pOF\n",
> +						index, node);
> +				return PTR_ERR(clk);
> +			}
> +
> +			hw = __clk_get_hw(clk);
> +			rc = clk_hw_set_spread_spectrum(hw, modfreq_hz, spread_bp, method);
> +			if (rc < 0)
> +				pr_err("clk: couldn't set %s clk spread spectrum %u %u %u: %d\n",
> +				       __clk_get_name(clk), modfreq_hz, spread_bp, method, rc);
> +			clk_put(clk);
> +		}
> +	}
> +
> +	return 0;
> +}



^ permalink raw reply	[flat|nested] 11+ messages in thread

* Re: [PATCH 2/3] clk: conf: Support assigned-clock-sscs
  2025-08-15  8:50     ` Peng Fan
@ 2025-08-15  7:51       ` Dan Carpenter
  0 siblings, 0 replies; 11+ messages in thread
From: Dan Carpenter @ 2025-08-15  7:51 UTC (permalink / raw)
  To: Peng Fan
  Cc: Peng Fan, Michael Turquette, Stephen Boyd, Sudeep Holla,
	Cristian Marussi, Marco Felsch, Geert Uytterhoeven, linux-clk,
	linux-kernel, arm-scmi, linux-arm-kernel

On Fri, Aug 15, 2025 at 04:50:42PM +0800, Peng Fan wrote:
> >> +		if (modfreq_hz || spread_bp || method) {
> >> +			rc = of_parse_phandle_with_args(node, "assigned-clocks",
> >> +					"#clock-cells",	index, &clkspec);
> >> +			if (rc < 0) {
> >> +				/* skip empty (null) phandles */
> >> +				if (rc == -ENOENT)
> >> +					continue;
> >> +				else
> >> +					return rc;
> >> +			}
> >> +
> >> +			if (clkspec.np == node && !clk_supplier) {
> >
> >Could you add a comment for this condition?  It's strange to me that we
> >don't iterate through the whole array.
> 
> I just follow the logic in __set_clk_parents and __set_clk_rates, nothing
> special here.
> 
> It is just like to phase out cases as below:
>   node-x {
> 	/* node-x is not a clk provider, but assigned-clocks uses node-x phandle */
> 	assigned-clocks = <&node-x   XYZ>;
>   }
> 

Ah.  Great.  Thanks.

regards,
dan carpenter



^ permalink raw reply	[flat|nested] 11+ messages in thread

* Re: [PATCH 2/3] clk: conf: Support assigned-clock-sscs
  2025-08-13  5:48   ` Dan Carpenter
@ 2025-08-15  8:50     ` Peng Fan
  2025-08-15  7:51       ` Dan Carpenter
  0 siblings, 1 reply; 11+ messages in thread
From: Peng Fan @ 2025-08-15  8:50 UTC (permalink / raw)
  To: Dan Carpenter
  Cc: Peng Fan, Michael Turquette, Stephen Boyd, Sudeep Holla,
	Cristian Marussi, Marco Felsch, Geert Uytterhoeven, linux-clk,
	linux-kernel, arm-scmi, linux-arm-kernel

Hi Dan,

On Wed, Aug 13, 2025 at 08:48:15AM +0300, Dan Carpenter wrote:
>On Tue, Aug 12, 2025 at 08:17:06PM +0800, Peng Fan wrote:
>> Parse the Spread Spectrum Configuration(SSC) from device tree and configure
>> them before using the clock.
>> 
>> Each SSC is three u32 elements which means '<modfreq spreaddepth
>> modmethod>', so assigned-clock-sscs is an array of multiple three u32
>> elements.
>> 
>> Signed-off-by: Peng Fan <peng.fan@nxp.com>
>> ---
>>  drivers/clk/clk-conf.c | 70 ++++++++++++++++++++++++++++++++++++++++++++++++++
>>  1 file changed, 70 insertions(+)
>> 
>> diff --git a/drivers/clk/clk-conf.c b/drivers/clk/clk-conf.c
>> index 303a0bb26e54a95655ce094a35b989c97ebc6fd8..81a2c1f8ca4c44df2c54c1e51f800f533c9453b3 100644
>> --- a/drivers/clk/clk-conf.c
>> +++ b/drivers/clk/clk-conf.c
>> @@ -155,6 +155,72 @@ static int __set_clk_rates(struct device_node *node, bool clk_supplier)
>>  	return 0;
>>  }
>>  
>> +static int __set_clk_spread_spectrum(struct device_node *node, bool clk_supplier)
>> +{
>> +	u32 *sscs __free(kfree) = NULL;
>> +	u32 elem_size = sizeof(u32) * 3;
>> +	struct of_phandle_args clkspec;
>> +	int rc, count, index;
>> +	struct clk *clk;
>> +
>> +	/* modfreq, spreadPercent, modmethod */
>> +	count = of_property_count_elems_of_size(node, "assigned-clock-sscs", elem_size);
>> +	if (count > 0) {
>> +		sscs = kcalloc(count, elem_size, GFP_KERNEL);
>> +		if (!sscs)
>> +			return -ENOMEM;
>> +		rc = of_property_read_u32_array(node,
>> +						"assigned-clock-sscs",
>> +						sscs, count * 3);
>> +	} else {
>> +		return 0;
>> +	}
>> +
>> +	if (rc)
>> +		return rc;
>
>Nit pick: Please, flip these conditions around.

ok. Fix in next version.

>
>	count = of_property_count_elems_of_size(node, "assigned-clock-sscs", elem_size);
>	if (count <= 0)
>		return 0;
>
>	sscs = kcalloc(count, elem_size, GFP_KERNEL);
>	if (!sscs)
>		return -ENOMEM;
>
>	rc = of_property_read_u32_array(node, "assigned-clock-sscs", sscs,
>					count * 3);
>	if (rc)
>		return rc;
>
>> +
>> +	for (index = 0; index < count; index++) {
>> +		u32 modfreq_hz = sscs[index * 3], spread_bp = sscs[index * 3 + 1];
>> +		u32 method = sscs[index * 3 + 2];
>
>This math would be nicer if you created a struct:
>
>struct spread_config {
>	u32 modfreq_hz;
>	u32 spread_depth;
>	u32 method;
>};
>
>Then you could use that instead of sscs.
>
>	for (i = 0; i < count; i++) {
>		struct spread_config *conf = &configs[i];
>		struct clk_hw *hw;
>
>		if (!conf->modfreq_hz && !conf->spread_depth && !conf->method)
>			continue;

Thanks for the tips. Update in next version.

>
>> +		struct clk_hw *hw;
>> +
>> +		if (modfreq_hz || spread_bp || method) {
>> +			rc = of_parse_phandle_with_args(node, "assigned-clocks",
>> +					"#clock-cells",	index, &clkspec);
>> +			if (rc < 0) {
>> +				/* skip empty (null) phandles */
>> +				if (rc == -ENOENT)
>> +					continue;
>> +				else
>> +					return rc;
>> +			}
>> +
>> +			if (clkspec.np == node && !clk_supplier) {
>
>Could you add a comment for this condition?  It's strange to me that we
>don't iterate through the whole array.

I just follow the logic in __set_clk_parents and __set_clk_rates, nothing
special here.

It is just like to phase out cases as below:
  node-x {
	/* node-x is not a clk provider, but assigned-clocks uses node-x phandle */
	assigned-clocks = <&node-x   XYZ>;
  }

Thanks for reviewing the patch. I will wait to collect more comments before
posting next version.

Regards,
Peng

>
>regards,
>dan carpenter
>
>> +				of_node_put(clkspec.np);
>> +				return 0;
>> +			}
>> +
>> +			clk = of_clk_get_from_provider(&clkspec);
>> +			of_node_put(clkspec.np);
>> +			if (IS_ERR(clk)) {
>> +				if (PTR_ERR(clk) != -EPROBE_DEFER)
>> +					pr_warn("clk: couldn't get clock %d for %pOF\n",
>> +						index, node);
>> +				return PTR_ERR(clk);
>> +			}
>> +
>> +			hw = __clk_get_hw(clk);
>> +			rc = clk_hw_set_spread_spectrum(hw, modfreq_hz, spread_bp, method);
>> +			if (rc < 0)
>> +				pr_err("clk: couldn't set %s clk spread spectrum %u %u %u: %d\n",
>> +				       __clk_get_name(clk), modfreq_hz, spread_bp, method, rc);
>> +			clk_put(clk);
>> +		}
>> +	}
>> +
>> +	return 0;
>> +}
>


^ permalink raw reply	[flat|nested] 11+ messages in thread

* Re: [PATCH 0/3] clk: Support spread spectrum and use it in clk-scmi
  2025-08-12 12:17 [PATCH 0/3] clk: Support spread spectrum and use it in clk-scmi Peng Fan
                   ` (2 preceding siblings ...)
  2025-08-12 12:17 ` [PATCH 3/3] clk: scmi: Support Spread Spectrum for NXP i.MX95 Peng Fan
@ 2025-08-27  7:05 ` Peng Fan
  3 siblings, 0 replies; 11+ messages in thread
From: Peng Fan @ 2025-08-27  7:05 UTC (permalink / raw)
  To: Peng Fan
  Cc: Michael Turquette, Stephen Boyd, Sudeep Holla, Cristian Marussi,
	Marco Felsch, Geert Uytterhoeven, linux-clk, linux-kernel,
	arm-scmi, linux-arm-kernel

Hi Stephen, Sudeep, Cristian

On Tue, Aug 12, 2025 at 08:17:04PM +0800, Peng Fan wrote:
>Since the assigned-clock-sscs property [1] has been accepted into the device
>tree schema, we can now support it in the Linux clock driver. Therefore,
>I’ve picked up the previously submitted work [2] titled “clk: Support
>spread spectrum and use it in clk-pll144x and clk-scmi.”
>As more than six months have passed since [2] was posted, I’m treating this
>patchset as a new submission rather than a v3.
>
>- Introduce clk_set_spread_spectrum to set the parameters for enabling
>  spread spectrum of a clock.
>- Parse 'assigned-clock-sscs' and configure it by default before using the
>  clock. This property is parsed before parsing clock rate.
>- Enable this feature for clk-scmi on i.MX95.
>
>Because SCMI spec will not include spread spectrum as a standard
>extension, we still need to use NXP i.MX OEM extension.

I only got a comment from Dan until now, before I post V2, I would like
to see if CLK and SCMI maintainer's have any comments. Or I need to wait
a bit more time, please just drop me a message.

Thanks,
Peng

>
>[1] https://github.com/devicetree-org/dt-schema/pull/154
>[2] https://lore.kernel.org/all/20250205-clk-ssc-v2-0-fa73083caa92@nxp.com/
>
>Signed-off-by: Peng Fan <peng.fan@nxp.com>
>---
>Peng Fan (3):
>      clk: Introduce clk_hw_set_spread_spectrum
>      clk: conf: Support assigned-clock-sscs
>      clk: scmi: Support Spread Spectrum for NXP i.MX95
>
> drivers/clk/clk-conf.c        | 70 +++++++++++++++++++++++++++++++++++++++++++
> drivers/clk/clk-scmi.c        | 64 ++++++++++++++++++++++++++++++++++++---
> drivers/clk/clk.c             | 32 ++++++++++++++++++++
> include/linux/clk-provider.h  | 29 ++++++++++++++++++
> include/linux/scmi_protocol.h |  5 ++++
> 5 files changed, 196 insertions(+), 4 deletions(-)
>---
>base-commit: b1549501188cc9eba732c25b033df7a53ccc341f
>change-id: 20250812-clk-ssc-version1-acf6f6efbd96
>
>Best regards,
>-- 
>Peng Fan <peng.fan@nxp.com>
>


^ permalink raw reply	[flat|nested] 11+ messages in thread

* Re: [PATCH 1/3] clk: Introduce clk_hw_set_spread_spectrum
  2025-08-12 12:17 ` [PATCH 1/3] clk: Introduce clk_hw_set_spread_spectrum Peng Fan
@ 2025-08-27 15:45   ` Brian Masney
  2025-08-29  9:08     ` Peng Fan
  0 siblings, 1 reply; 11+ messages in thread
From: Brian Masney @ 2025-08-27 15:45 UTC (permalink / raw)
  To: Peng Fan
  Cc: Michael Turquette, Stephen Boyd, Sudeep Holla, Cristian Marussi,
	Marco Felsch, Geert Uytterhoeven, linux-clk, linux-kernel,
	arm-scmi, linux-arm-kernel

On Tue, Aug 12, 2025 at 08:17:05PM +0800, Peng Fan wrote:
> Add clk_hw_set_spread_spectrum to configure a clock to enable spread
> spectrum feature. set_spread_spectrum ops is added for clk drivers to
> have their own hardware specific implementation.
> 
> Signed-off-by: Peng Fan <peng.fan@nxp.com>
> ---
>  drivers/clk/clk.c            | 32 ++++++++++++++++++++++++++++++++
>  include/linux/clk-provider.h | 29 +++++++++++++++++++++++++++++
>  2 files changed, 61 insertions(+)
> 
> diff --git a/drivers/clk/clk.c b/drivers/clk/clk.c
> index b821b2cdb155331c85fafbd2fac8ab3703a08e4d..48c7a301b72b30fd824dae7ada2c44ee84d40867 100644
> --- a/drivers/clk/clk.c
> +++ b/drivers/clk/clk.c
> @@ -2802,6 +2802,38 @@ int clk_set_max_rate(struct clk *clk, unsigned long rate)
>  }
>  EXPORT_SYMBOL_GPL(clk_set_max_rate);
>  
> +int clk_hw_set_spread_spectrum(struct clk_hw *hw, unsigned int modfreq_hz,
> +			       unsigned int spread_bp, unsigned int method)
                                                       ^^^^^^^^^^^^
Should this be 'enum clk_ssc_method'?

Also can you add kernel docs for all of the parameters? I know it's
documented on 'struct clk_spread_spectrum' below.

What do you think about having this function take that struct instead as
a parameter to match what's on the clk op?

Brian



^ permalink raw reply	[flat|nested] 11+ messages in thread

* Re: [PATCH 1/3] clk: Introduce clk_hw_set_spread_spectrum
  2025-08-27 15:45   ` Brian Masney
@ 2025-08-29  9:08     ` Peng Fan
  0 siblings, 0 replies; 11+ messages in thread
From: Peng Fan @ 2025-08-29  9:08 UTC (permalink / raw)
  To: Brian Masney
  Cc: Peng Fan, Michael Turquette, Stephen Boyd, Sudeep Holla,
	Cristian Marussi, Marco Felsch, Geert Uytterhoeven, linux-clk,
	linux-kernel, arm-scmi, linux-arm-kernel

On Wed, Aug 27, 2025 at 11:45:19AM -0400, Brian Masney wrote:
>On Tue, Aug 12, 2025 at 08:17:05PM +0800, Peng Fan wrote:
>> Add clk_hw_set_spread_spectrum to configure a clock to enable spread
>> spectrum feature. set_spread_spectrum ops is added for clk drivers to
>> have their own hardware specific implementation.
>> 
>> Signed-off-by: Peng Fan <peng.fan@nxp.com>
>> ---
>>  drivers/clk/clk.c            | 32 ++++++++++++++++++++++++++++++++
>>  include/linux/clk-provider.h | 29 +++++++++++++++++++++++++++++
>>  2 files changed, 61 insertions(+)
>> 
>> diff --git a/drivers/clk/clk.c b/drivers/clk/clk.c
>> index b821b2cdb155331c85fafbd2fac8ab3703a08e4d..48c7a301b72b30fd824dae7ada2c44ee84d40867 100644
>> --- a/drivers/clk/clk.c
>> +++ b/drivers/clk/clk.c
>> @@ -2802,6 +2802,38 @@ int clk_set_max_rate(struct clk *clk, unsigned long rate)
>>  }
>>  EXPORT_SYMBOL_GPL(clk_set_max_rate);
>>  
>> +int clk_hw_set_spread_spectrum(struct clk_hw *hw, unsigned int modfreq_hz,
>> +			       unsigned int spread_bp, unsigned int method)
>                                                       ^^^^^^^^^^^^
>Should this be 'enum clk_ssc_method'?
>
>Also can you add kernel docs for all of the parameters? I know it's
>documented on 'struct clk_spread_spectrum' below.
>
>What do you think about having this function take that struct instead as
>a parameter to match what's on the clk op?

Yeah. Dan has raised similar comment, I will change to use
struct clk_spread_spectrum as the 2nd param.

Thanks,
Peng

>
>Brian
>


^ permalink raw reply	[flat|nested] 11+ messages in thread

end of thread, other threads:[~2025-08-29  8:29 UTC | newest]

Thread overview: 11+ messages (download: mbox.gz follow: Atom feed
-- links below jump to the message on this page --
2025-08-12 12:17 [PATCH 0/3] clk: Support spread spectrum and use it in clk-scmi Peng Fan
2025-08-12 12:17 ` [PATCH 1/3] clk: Introduce clk_hw_set_spread_spectrum Peng Fan
2025-08-27 15:45   ` Brian Masney
2025-08-29  9:08     ` Peng Fan
2025-08-12 12:17 ` [PATCH 2/3] clk: conf: Support assigned-clock-sscs Peng Fan
2025-08-13  5:48   ` Dan Carpenter
2025-08-15  8:50     ` Peng Fan
2025-08-15  7:51       ` Dan Carpenter
2025-08-12 12:17 ` [PATCH 3/3] clk: scmi: Support Spread Spectrum for NXP i.MX95 Peng Fan
2025-08-27  7:05 ` [PATCH 0/3] clk: Support spread spectrum and use it in clk-scmi Peng Fan
  -- strict thread matches above, loose matches on Subject: below --
2025-01-24 14:25 Peng Fan (OSS)
2025-01-24 14:25 ` [PATCH 2/3] clk: conf: Support assigned-clock-sscs Peng Fan (OSS)

This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).