[PATCH v3 1/3] dt-bindings: clk: rs9: Add Renesas 9-series I2C PCIe clock generator

[PATCH v3 1/3] dt-bindings: clk: rs9: Add Renesas 9-series I2C PCIe clock generator

[Marek Vasut]

Add binding for Renesas 9-series PCIe clock generators. This binding
is designed to support 9FGV/9DBV/9DMV/9FGL/9DML/9QXL/9SQ series I2C
PCIe clock generators, currently the only tested and supported chip
is 9FGV0241.

Signed-off-by: Marek Vasut <marex@denx.de>
Cc: Michael Turquette <mturquette@baylibre.com>
Cc: Rob Herring <robh+dt@kernel.org>
Cc: Stephen Boyd <sboyd@kernel.org>
Cc: devicetree@vger.kernel.org
To: linux-clk@vger.kernel.org
---
V2: - Drop clock consumer from the binding example
    - Make clocks, i.e. xtal, mandatory
V3: - Rename renesas,out-amplitude to renesas,out-amplitude-microvolt
    - Drop type ref: from renesas,out-amplitude-microvolt property
    - Explain 'pcm' in description text as 1/1000 of percent
    - Add newlines
---
 .../bindings/clock/renesas,9series.yaml       | 97 +++++++++++++++++++
 1 file changed, 97 insertions(+)
 create mode 100644 Documentation/devicetree/bindings/clock/renesas,9series.yaml

diff --git a/Documentation/devicetree/bindings/clock/renesas,9series.yaml b/Documentation/devicetree/bindings/clock/renesas,9series.yaml
new file mode 100644
index 0000000000000..102eb95cb3fcd
--- /dev/null
+++ b/Documentation/devicetree/bindings/clock/renesas,9series.yaml
@@ -0,0 +1,97 @@
+# SPDX-License-Identifier: (GPL-2.0-only OR BSD-2-Clause)
+%YAML 1.2
+---
+$id: http://devicetree.org/schemas/clock/renesas,9series.yaml#
+$schema: http://devicetree.org/meta-schemas/core.yaml#
+
+title: Binding for Renesas 9-series I2C PCIe clock generators
+
+description: |
+  The Renesas 9-series are I2C PCIe clock generators providing
+  from 1 to 20 output clocks.
+
+  When referencing the provided clock in the DT using phandle
+  and clock specifier, the following mapping applies:
+
+  - 9FGV0241:
+    0 -- DIF0
+    1 -- DIF1
+
+maintainers:
+  - Marek Vasut <marex@denx.de>
+
+properties:
+  compatible:
+    enum:
+      - renesas,9fgv0241
+
+  reg:
+    description: I2C device address
+    enum: [ 0x68, 0x6a ]
+
+  '#clock-cells':
+    const: 1
+
+  clocks:
+    items:
+      - description: XTal input clock
+
+  renesas,out-amplitude-microvolt:
+    enum: [ 600000, 700000, 800000, 900000 ]
+    description: Output clock signal amplitude
+
+  renesas,out-spread-spectrum:
+    $ref: /schemas/types.yaml#/definitions/uint32
+    enum: [ 100000, 99750, 99500 ]
+    description: Output clock down spread in pcm (1/1000 of percent)
+
+patternProperties:
+  "^DIF[0-19]$":
+    type: object
+    description:
+      Description of one of the outputs (DIF0..DIF19).
+
+    properties:
+      renesas,slew-rate:
+        $ref: /schemas/types.yaml#/definitions/uint32
+        enum: [ 2000000, 3000000 ]
+        description: Output clock slew rate select in V/ns
+
+    additionalProperties: false
+
+required:
+  - compatible
+  - reg
+  - clocks
+  - '#clock-cells'
+
+additionalProperties: false
+
+examples:
+  - |
+    /* 25MHz reference crystal */
+    ref25: ref25m {
+        compatible = "fixed-clock";
+        #clock-cells = <0>;
+        clock-frequency = <25000000>;
+    };
+
+    i2c@0 {
+        reg = <0x0 0x100>;
+        #address-cells = <1>;
+        #size-cells = <0>;
+
+        rs9: clock-generator@6a {
+            compatible = "renesas,9fgv0241";
+            reg = <0x6a>;
+            #clock-cells = <1>;
+
+            clocks = <&ref25m>;
+
+            DIF0 {
+                renesas,slew-rate = <3000000>;
+            };
+        };
+    };
+
+...
-- 
2.34.1

[PATCH v3 2/3] clk: Make __clk_hw_register_fixed_factor non-static

[Marek Vasut]

Access to the full parameters of __clk_hw_register_fixed_factor()
is useful in case a driver is registering fixed clock with only
single parent, in which case the driver should set parent_name to
NULL and parent_index to 0, and access to this function permits it
to do just that.

Signed-off-by: Marek Vasut <marex@denx.de>
Cc: Michael Turquette <mturquette@baylibre.com>
Cc: Rob Herring <robh+dt@kernel.org>
Cc: Stephen Boyd <sboyd@kernel.org>
Cc: devicetree@vger.kernel.org
To: linux-clk@vger.kernel.org
---
V2: - New patch
V3: - No change
---
 drivers/clk/clk-fixed-factor.c | 2 +-
 include/linux/clk-provider.h   | 5 +++++
 2 files changed, 6 insertions(+), 1 deletion(-)

diff --git a/drivers/clk/clk-fixed-factor.c b/drivers/clk/clk-fixed-factor.c
index 4e4b6d3676126..81d8c9e430a23 100644
--- a/drivers/clk/clk-fixed-factor.c
+++ b/drivers/clk/clk-fixed-factor.c
@@ -76,7 +76,7 @@ static void devm_clk_hw_register_fixed_factor_release(struct device *dev, void *
 	clk_hw_unregister(&fix->hw);
 }
 
-static struct clk_hw *
+struct clk_hw *
 __clk_hw_register_fixed_factor(struct device *dev, struct device_node *np,
 		const char *name, const char *parent_name, int index,
 		unsigned long flags, unsigned int mult, unsigned int div,
diff --git a/include/linux/clk-provider.h b/include/linux/clk-provider.h
index 2faa6f7aa8a87..d216221448d31 100644
--- a/include/linux/clk-provider.h
+++ b/include/linux/clk-provider.h
@@ -992,6 +992,11 @@ struct clk_fixed_factor {
 #define to_clk_fixed_factor(_hw) container_of(_hw, struct clk_fixed_factor, hw)
 
 extern const struct clk_ops clk_fixed_factor_ops;
+struct clk_hw *
+__clk_hw_register_fixed_factor(struct device *dev, struct device_node *np,
+		const char *name, const char *parent_name, int index,
+		unsigned long flags, unsigned int mult, unsigned int div,
+		bool devm);
 struct clk *clk_register_fixed_factor(struct device *dev, const char *name,
 		const char *parent_name, unsigned long flags,
 		unsigned int mult, unsigned int div);
-- 
2.34.1

Re: [PATCH v3 2/3] clk: Make __clk_hw_register_fixed_factor non-static

[Stephen Boyd]

Quoting Marek Vasut (2022-02-25 20:07:22)
> Access to the full parameters of __clk_hw_register_fixed_factor()
> is useful in case a driver is registering fixed clock with only
> single parent, in which case the driver should set parent_name to
> NULL and parent_index to 0, and access to this function permits it
> to do just that.
> 
> Signed-off-by: Marek Vasut <marex@denx.de>
> Cc: Michael Turquette <mturquette@baylibre.com>
> Cc: Rob Herring <robh+dt@kernel.org>
> Cc: Stephen Boyd <sboyd@kernel.org>
> Cc: devicetree@vger.kernel.org
> To: linux-clk@vger.kernel.org
> ---

Applied to clk-next

Re: [PATCH v3 2/3] clk: Make __clk_hw_register_fixed_factor non-static

[Stephen Boyd]

Quoting Marek Vasut (2022-02-25 20:07:22)
> Access to the full parameters of __clk_hw_register_fixed_factor()
> is useful in case a driver is registering fixed clock with only
> single parent, in which case the driver should set parent_name to
> NULL and parent_index to 0, and access to this function permits it
> to do just that.
> 
> Signed-off-by: Marek Vasut <marex@denx.de>
> Cc: Michael Turquette <mturquette@baylibre.com>
> Cc: Rob Herring <robh+dt@kernel.org>
> Cc: Stephen Boyd <sboyd@kernel.org>
> Cc: devicetree@vger.kernel.org
> To: linux-clk@vger.kernel.org
> ---
> V2: - New patch
> V3: - No change

This isn't exported. Given that we don't typically export an internal
function (hence the double underscore) I'm going to change this to be a
new function. See the attached patch.

---8<----
diff --git a/drivers/clk/clk-fixed-factor.c b/drivers/clk/clk-fixed-factor.c
index 81d8c9e430a2..54942d758ee6 100644
--- a/drivers/clk/clk-fixed-factor.c
+++ b/drivers/clk/clk-fixed-factor.c
@@ -76,7 +76,7 @@ static void devm_clk_hw_register_fixed_factor_release(struct device *dev, void *
 	clk_hw_unregister(&fix->hw);
 }
 
-struct clk_hw *
+static struct clk_hw *
 __clk_hw_register_fixed_factor(struct device *dev, struct device_node *np,
 		const char *name, const char *parent_name, int index,
 		unsigned long flags, unsigned int mult, unsigned int div,
@@ -131,6 +131,28 @@ __clk_hw_register_fixed_factor(struct device *dev, struct device_node *np,
 	return hw;
 }
 
+/**
+ * devm_clk_hw_register_fixed_factor_index - Register a fixed factor clock with
+ * parent from DT index
+ * @dev: device that is registering this clock
+ * @name: name of this clock
+ * @index: index of phandle in @dev 'clocks' property
+ * @flags: fixed factor flags
+ * @mult: multiplier
+ * @div: divider
+ *
+ * Return: Pointer to fixed factor clk_hw structure that was registered or
+ * an error pointer.
+ */
+struct clk_hw *devm_clk_hw_register_fixed_factor_index(struct device *dev,
+		const char *name, unsigned int index, unsigned long flags,
+		unsigned int mult, unsigned int div)
+{
+	return __clk_hw_register_fixed_factor(dev, NULL, name, NULL, index,
+					      flags, mult, div, true);
+}
+EXPORT_SYMBOL_GPL(devm_clk_hw_register_fixed_factor_index);
+
 struct clk_hw *clk_hw_register_fixed_factor(struct device *dev,
 		const char *name, const char *parent_name, unsigned long flags,
 		unsigned int mult, unsigned int div)
diff --git a/drivers/clk/clk-renesas-pcie.c b/drivers/clk/clk-renesas-pcie.c
index cd8229dd8123..59d9cf0053eb 100644
--- a/drivers/clk/clk-renesas-pcie.c
+++ b/drivers/clk/clk-renesas-pcie.c
@@ -250,8 +250,8 @@ static int rs9_probe(struct i2c_client *client, const struct i2c_device_id *id)
 	/* Register clock */
 	for (i = 0; i < rs9->chip_info->num_clks; i++) {
 		snprintf(name, 5, "DIF%d", i);
-		hw = __clk_hw_register_fixed_factor(&client->dev, NULL, name,
-						    NULL, 0, 0, 4, 1, true);
+		hw = devm_clk_hw_register_fixed_factor_index(&client->dev, name,
+						    0, 0, 4, 1);
 		if (IS_ERR(hw))
 			return PTR_ERR(hw);
 
diff --git a/include/linux/clk-provider.h b/include/linux/clk-provider.h
index d216221448d3..b7a7923f6bbb 100644
--- a/include/linux/clk-provider.h
+++ b/include/linux/clk-provider.h
@@ -992,11 +992,6 @@ struct clk_fixed_factor {
 #define to_clk_fixed_factor(_hw) container_of(_hw, struct clk_fixed_factor, hw)
 
 extern const struct clk_ops clk_fixed_factor_ops;
-struct clk_hw *
-__clk_hw_register_fixed_factor(struct device *dev, struct device_node *np,
-		const char *name, const char *parent_name, int index,
-		unsigned long flags, unsigned int mult, unsigned int div,
-		bool devm);
 struct clk *clk_register_fixed_factor(struct device *dev, const char *name,
 		const char *parent_name, unsigned long flags,
 		unsigned int mult, unsigned int div);
@@ -1008,6 +1003,9 @@ void clk_hw_unregister_fixed_factor(struct clk_hw *hw);
 struct clk_hw *devm_clk_hw_register_fixed_factor(struct device *dev,
 		const char *name, const char *parent_name, unsigned long flags,
 		unsigned int mult, unsigned int div);
+struct clk_hw *devm_clk_hw_register_fixed_factor_index(struct device *dev,
+		const char *name, unsigned int index, unsigned long flags,
+		unsigned int mult, unsigned int div);
 /**
  * struct clk_fractional_divider - adjustable fractional divider clock
  *

Re: [PATCH v3 2/3] clk: Make __clk_hw_register_fixed_factor non-static

[Marek Vasut]

On 3/18/22 22:03, Stephen Boyd wrote:
> Quoting Marek Vasut (2022-02-25 20:07:22)
>> Access to the full parameters of __clk_hw_register_fixed_factor()
>> is useful in case a driver is registering fixed clock with only
>> single parent, in which case the driver should set parent_name to
>> NULL and parent_index to 0, and access to this function permits it
>> to do just that.
>>
>> Signed-off-by: Marek Vasut <marex@denx.de>
>> Cc: Michael Turquette <mturquette@baylibre.com>
>> Cc: Rob Herring <robh+dt@kernel.org>
>> Cc: Stephen Boyd <sboyd@kernel.org>
>> Cc: devicetree@vger.kernel.org
>> To: linux-clk@vger.kernel.org
>> ---
>> V2: - New patch
>> V3: - No change
> 
> This isn't exported. Given that we don't typically export an internal
> function (hence the double underscore) I'm going to change this to be a
> new function. See the attached patch.

I can confirm the change works and looks OK.

Do you want me to send a V4 or will you squash it into these patches 
yourself when applying?

Re: [PATCH v3 2/3] clk: Make __clk_hw_register_fixed_factor non-static

[Stephen Boyd]

Quoting Marek Vasut (2022-03-20 06:23:14)
> On 3/18/22 22:03, Stephen Boyd wrote:
> > Quoting Marek Vasut (2022-02-25 20:07:22)
> >> Access to the full parameters of __clk_hw_register_fixed_factor()
> >> is useful in case a driver is registering fixed clock with only
> >> single parent, in which case the driver should set parent_name to
> >> NULL and parent_index to 0, and access to this function permits it
> >> to do just that.
> >>
> >> Signed-off-by: Marek Vasut <marex@denx.de>
> >> Cc: Michael Turquette <mturquette@baylibre.com>
> >> Cc: Rob Herring <robh+dt@kernel.org>
> >> Cc: Stephen Boyd <sboyd@kernel.org>
> >> Cc: devicetree@vger.kernel.org
> >> To: linux-clk@vger.kernel.org
> >> ---
> >> V2: - New patch
> >> V3: - No change
> > 
> > This isn't exported. Given that we don't typically export an internal
> > function (hence the double underscore) I'm going to change this to be a
> > new function. See the attached patch.
> 
> I can confirm the change works and looks OK.
> 
> Do you want me to send a V4 or will you squash it into these patches 
> yourself when applying?

No need I fixed it up and pushed it out.

Re: [PATCH v3 2/3] clk: Make __clk_hw_register_fixed_factor non-static

[Marek Vasut]

On 3/21/22 20:26, Stephen Boyd wrote:
> Quoting Marek Vasut (2022-03-20 06:23:14)
>> On 3/18/22 22:03, Stephen Boyd wrote:
>>> Quoting Marek Vasut (2022-02-25 20:07:22)
>>>> Access to the full parameters of __clk_hw_register_fixed_factor()
>>>> is useful in case a driver is registering fixed clock with only
>>>> single parent, in which case the driver should set parent_name to
>>>> NULL and parent_index to 0, and access to this function permits it
>>>> to do just that.
>>>>
>>>> Signed-off-by: Marek Vasut <marex@denx.de>
>>>> Cc: Michael Turquette <mturquette@baylibre.com>
>>>> Cc: Rob Herring <robh+dt@kernel.org>
>>>> Cc: Stephen Boyd <sboyd@kernel.org>
>>>> Cc: devicetree@vger.kernel.org
>>>> To: linux-clk@vger.kernel.org
>>>> ---
>>>> V2: - New patch
>>>> V3: - No change
>>>
>>> This isn't exported. Given that we don't typically export an internal
>>> function (hence the double underscore) I'm going to change this to be a
>>> new function. See the attached patch.
>>
>> I can confirm the change works and looks OK.
>>
>> Do you want me to send a V4 or will you squash it into these patches
>> yourself when applying?
> 
> No need I fixed it up and pushed it out.

Thank you

[PATCH v3 3/3] clk: rs9: Add Renesas 9-series PCIe clock generator driver

[Marek Vasut]

Add driver for Renesas 9-series PCIe clock generators. This driver
is designed to support 9FGV/9DBV/9DMV/9FGL/9DML/9QXL/9SQ series I2C
PCIe clock generators, currently the only tested and supported chip
is 9FGV0241.

The driver is capable of configuring per-chip spread spectrum mode
and output amplitude, as well as per-output slew rate.

Signed-off-by: Marek Vasut <marex@denx.de>
Cc: Michael Turquette <mturquette@baylibre.com>
Cc: Rob Herring <robh+dt@kernel.org>
Cc: Stephen Boyd <sboyd@kernel.org>
Cc: devicetree@vger.kernel.org
To: linux-clk@vger.kernel.org
---
V2: - Drop unused includes
    - Use REGCACHE_FLAT for smaller reg file
    - Move of_node_put() in rs9_get_output_config() a bit higher up
    - Drop forward declaration of clk_rs9_of_match
    - Use device_get_match_data() instead of of_device_get_match_data()
      and check for its return value, verify it is non-NULL
    - Use newly available __clk_hw_register_fixed_factor() with
      parent_data index=0 and drop of_clk_get_parent_name() altogether
V3: - Rename renesas,out-amplitude to renesas,out-amplitude-microvolt
---
 drivers/clk/Kconfig            |   9 +
 drivers/clk/Makefile           |   1 +
 drivers/clk/clk-renesas-pcie.c | 322 +++++++++++++++++++++++++++++++++
 3 files changed, 332 insertions(+)
 create mode 100644 drivers/clk/clk-renesas-pcie.c

diff --git a/drivers/clk/Kconfig b/drivers/clk/Kconfig
index 6f03c29c40be2..fccc1760af210 100644
--- a/drivers/clk/Kconfig
+++ b/drivers/clk/Kconfig
@@ -343,6 +343,15 @@ config COMMON_CLK_OXNAS
 	help
 	  Support for the OXNAS SoC Family clocks.
 
+config COMMON_CLK_RS9_PCIE
+	tristate "Clock driver for Renesas 9-series PCIe clock generators"
+	depends on I2C
+	depends on OF
+	select REGMAP_I2C
+	help
+	  This driver supports the Renesas 9-series PCIe clock generator
+	  models 9FGV/9DBV/9DMV/9FGL/9DML/9QXL/9SQ.
+
 config COMMON_CLK_VC5
 	tristate "Clock driver for IDT VersaClock 5,6 devices"
 	depends on I2C
diff --git a/drivers/clk/Makefile b/drivers/clk/Makefile
index 6a98291350b64..3ec27842ec779 100644
--- a/drivers/clk/Makefile
+++ b/drivers/clk/Makefile
@@ -68,6 +68,7 @@ obj-$(CONFIG_COMMON_CLK_STM32MP157)	+= clk-stm32mp1.o
 obj-$(CONFIG_COMMON_CLK_TPS68470)      += clk-tps68470.o
 obj-$(CONFIG_CLK_TWL6040)		+= clk-twl6040.o
 obj-$(CONFIG_ARCH_VT8500)		+= clk-vt8500.o
+obj-$(CONFIG_COMMON_CLK_RS9_PCIE)	+= clk-renesas-pcie.o
 obj-$(CONFIG_COMMON_CLK_VC5)		+= clk-versaclock5.o
 obj-$(CONFIG_COMMON_CLK_WM831X)		+= clk-wm831x.o
 obj-$(CONFIG_COMMON_CLK_XGENE)		+= clk-xgene.o
diff --git a/drivers/clk/clk-renesas-pcie.c b/drivers/clk/clk-renesas-pcie.c
new file mode 100644
index 0000000000000..462cc2fe23a21
--- /dev/null
+++ b/drivers/clk/clk-renesas-pcie.c
@@ -0,0 +1,322 @@
+// SPDX-License-Identifier: GPL-2.0-or-later
+/*
+ * Driver for Renesas 9-series PCIe clock generator driver
+ *
+ * The following series can be supported:
+ *   - 9FGV/9DBV/9DMV/9FGL/9DML/9QXL/9SQ
+ * Currently supported:
+ *   - 9FGV0241
+ *
+ * Copyright (C) 2022 Marek Vasut <marex@denx.de>
+ */
+
+#include <linux/clk-provider.h>
+#include <linux/i2c.h>
+#include <linux/mod_devicetable.h>
+#include <linux/module.h>
+#include <linux/of.h>
+#include <linux/regmap.h>
+
+#define RS9_REG_OE				0x0
+#define RS9_REG_OE_DIF_OE(n)			BIT((n) + 1)
+#define RS9_REG_SS				0x1
+#define RS9_REG_SS_AMP_0V6			0x0
+#define RS9_REG_SS_AMP_0V7			0x1
+#define RS9_REG_SS_AMP_0V8			0x2
+#define RS9_REG_SS_AMP_0V9			0x3
+#define RS9_REG_SS_AMP_MASK			0x3
+#define RS9_REG_SS_SSC_100			0
+#define RS9_REG_SS_SSC_M025			(1 << 3)
+#define RS9_REG_SS_SSC_M050			(3 << 3)
+#define RS9_REG_SS_SSC_MASK			(3 << 3)
+#define RS9_REG_SS_SSC_LOCK			BIT(5)
+#define RS9_REG_SR				0x2
+#define RS9_REG_SR_2V0_DIF(n)			0
+#define RS9_REG_SR_3V0_DIF(n)			BIT((n) + 1)
+#define RS9_REG_SR_DIF_MASK(n)		BIT((n) + 1)
+#define RS9_REG_REF				0x3
+#define RS9_REG_REF_OE				BIT(4)
+#define RS9_REG_REF_OD				BIT(5)
+#define RS9_REG_REF_SR_SLOWEST			0
+#define RS9_REG_REF_SR_SLOW			(1 << 6)
+#define RS9_REG_REF_SR_FAST			(2 << 6)
+#define RS9_REG_REF_SR_FASTER			(3 << 6)
+#define RS9_REG_VID				0x5
+#define RS9_REG_DID				0x6
+#define RS9_REG_BCP				0x7
+
+/* Supported Renesas 9-series models. */
+enum rs9_model {
+	RENESAS_9FGV0241,
+};
+
+/* Structure to describe features of a particular 9-series model */
+struct rs9_chip_info {
+	const enum rs9_model	model;
+	unsigned int		num_clks;
+};
+
+struct rs9_driver_data {
+	struct i2c_client	*client;
+	struct regmap		*regmap;
+	const struct rs9_chip_info *chip_info;
+	struct clk		*pin_xin;
+	struct clk_hw		*clk_dif[2];
+	u8			pll_amplitude;
+	u8			pll_ssc;
+	u8			clk_dif_sr;
+};
+
+/*
+ * Renesas 9-series i2c regmap
+ */
+static const struct regmap_range rs9_readable_ranges[] = {
+	regmap_reg_range(RS9_REG_OE, RS9_REG_REF),
+	regmap_reg_range(RS9_REG_VID, RS9_REG_BCP),
+};
+
+static const struct regmap_access_table rs9_readable_table = {
+	.yes_ranges = rs9_readable_ranges,
+	.n_yes_ranges = ARRAY_SIZE(rs9_readable_ranges),
+};
+
+static const struct regmap_range rs9_writeable_ranges[] = {
+	regmap_reg_range(RS9_REG_OE, RS9_REG_REF),
+	regmap_reg_range(RS9_REG_BCP, RS9_REG_BCP),
+};
+
+static const struct regmap_access_table rs9_writeable_table = {
+	.yes_ranges = rs9_writeable_ranges,
+	.n_yes_ranges = ARRAY_SIZE(rs9_writeable_ranges),
+};
+
+static const struct regmap_config rs9_regmap_config = {
+	.reg_bits = 8,
+	.val_bits = 8,
+	.cache_type = REGCACHE_FLAT,
+	.max_register = 0x8,
+	.rd_table = &rs9_readable_table,
+	.wr_table = &rs9_writeable_table,
+};
+
+static int rs9_get_output_config(struct rs9_driver_data *rs9, int idx)
+{
+	struct i2c_client *client = rs9->client;
+	unsigned char name[5] = "DIF0";
+	struct device_node *np;
+	int ret;
+	u32 sr;
+
+	/* Set defaults */
+	rs9->clk_dif_sr &= ~RS9_REG_SR_DIF_MASK(idx);
+	rs9->clk_dif_sr |= RS9_REG_SR_3V0_DIF(idx);
+
+	snprintf(name, 5, "DIF%d", idx);
+	np = of_get_child_by_name(client->dev.of_node, name);
+	if (!np)
+		return 0;
+
+	/* Output clock slew rate */
+	ret = of_property_read_u32(np, "renesas,slew-rate", &sr);
+	of_node_put(np);
+	if (!ret) {
+		if (sr == 2000000) {		/* 2V/ns */
+			rs9->clk_dif_sr &= ~RS9_REG_SR_DIF_MASK(idx);
+			rs9->clk_dif_sr |= RS9_REG_SR_2V0_DIF(idx);
+		} else if (sr == 3000000) {	/* 3V/ns (default) */
+			rs9->clk_dif_sr &= ~RS9_REG_SR_DIF_MASK(idx);
+			rs9->clk_dif_sr |= RS9_REG_SR_3V0_DIF(idx);
+		} else
+			ret = dev_err_probe(&client->dev, -EINVAL,
+					    "Invalid renesas,slew-rate value\n");
+	}
+
+	return ret;
+}
+
+static int rs9_get_common_config(struct rs9_driver_data *rs9)
+{
+	struct i2c_client *client = rs9->client;
+	struct device_node *np = client->dev.of_node;
+	unsigned int amp, ssc;
+	int ret;
+
+	/* Set defaults */
+	rs9->pll_amplitude = RS9_REG_SS_AMP_0V7;
+	rs9->pll_ssc = RS9_REG_SS_SSC_100;
+
+	/* Output clock amplitude */
+	ret = of_property_read_u32(np, "renesas,out-amplitude-microvolt",
+				   &amp);
+	if (!ret) {
+		if (amp == 600000)	/* 0.6V */
+			rs9->pll_amplitude = RS9_REG_SS_AMP_0V6;
+		else if (amp == 700000)	/* 0.7V (default) */
+			rs9->pll_amplitude = RS9_REG_SS_AMP_0V7;
+		else if (amp == 800000)	/* 0.8V */
+			rs9->pll_amplitude = RS9_REG_SS_AMP_0V8;
+		else if (amp == 900000)	/* 0.9V */
+			rs9->pll_amplitude = RS9_REG_SS_AMP_0V9;
+		else
+			return dev_err_probe(&client->dev, -EINVAL,
+					     "Invalid renesas,out-amplitude-microvolt value\n");
+	}
+
+	/* Output clock spread spectrum */
+	ret = of_property_read_u32(np, "renesas,out-spread-spectrum", &ssc);
+	if (!ret) {
+		if (ssc == 100000)	/* 100% ... no spread (default) */
+			rs9->pll_ssc = RS9_REG_SS_SSC_100;
+		else if (ssc == 99750)	/* -0.25% ... down spread */
+			rs9->pll_ssc = RS9_REG_SS_SSC_M025;
+		else if (ssc == 99500)	/* -0.50% ... down spread */
+			rs9->pll_ssc = RS9_REG_SS_SSC_M050;
+		else
+			return dev_err_probe(&client->dev, -EINVAL,
+					     "Invalid renesas,out-spread-spectrum value\n");
+	}
+
+	return 0;
+}
+
+static void rs9_update_config(struct rs9_driver_data *rs9)
+{
+	int i;
+
+	/* If amplitude is non-default, update it. */
+	if (rs9->pll_amplitude != RS9_REG_SS_AMP_0V7) {
+		regmap_update_bits(rs9->regmap, RS9_REG_SS, RS9_REG_SS_AMP_MASK,
+				   rs9->pll_amplitude);
+	}
+
+	/* If SSC is non-default, update it. */
+	if (rs9->pll_ssc != RS9_REG_SS_SSC_100) {
+		regmap_update_bits(rs9->regmap, RS9_REG_SS, RS9_REG_SS_SSC_MASK,
+				   rs9->pll_ssc);
+	}
+
+	for (i = 0; i < rs9->chip_info->num_clks; i++) {
+		if (rs9->clk_dif_sr & RS9_REG_SR_3V0_DIF(i))
+			continue;
+
+		regmap_update_bits(rs9->regmap, RS9_REG_SR, RS9_REG_SR_3V0_DIF(i),
+				   rs9->clk_dif_sr & RS9_REG_SR_3V0_DIF(i));
+	}
+}
+
+static struct clk_hw *
+rs9_of_clk_get(struct of_phandle_args *clkspec, void *data)
+{
+	struct rs9_driver_data *rs9 = data;
+	unsigned int idx = clkspec->args[0];
+
+	return rs9->clk_dif[idx];
+}
+
+static int rs9_probe(struct i2c_client *client, const struct i2c_device_id *id)
+{
+	unsigned char name[5] = "DIF0";
+	struct rs9_driver_data *rs9;
+	struct clk_hw *hw;
+	int i, ret;
+
+	rs9 = devm_kzalloc(&client->dev, sizeof(*rs9), GFP_KERNEL);
+	if (!rs9)
+		return -ENOMEM;
+
+	i2c_set_clientdata(client, rs9);
+	rs9->client = client;
+	rs9->chip_info = device_get_match_data(&client->dev);
+	if (!rs9->chip_info)
+		return -EINVAL;
+
+	/* Fetch common configuration from DT (if specified) */
+	ret = rs9_get_common_config(rs9);
+	if (ret)
+		return ret;
+
+	/* Fetch DIFx output configuration from DT (if specified) */
+	for (i = 0; i < rs9->chip_info->num_clks; i++) {
+		ret = rs9_get_output_config(rs9, i);
+		if (ret)
+			return ret;
+	}
+
+	rs9->regmap = devm_regmap_init_i2c(client, &rs9_regmap_config);
+	if (IS_ERR(rs9->regmap))
+		return dev_err_probe(&client->dev, PTR_ERR(rs9->regmap),
+				     "Failed to allocate register map\n");
+
+	/* Register clock */
+	for (i = 0; i < rs9->chip_info->num_clks; i++) {
+		snprintf(name, 5, "DIF%d", i);
+		hw = __clk_hw_register_fixed_factor(&client->dev, NULL, name,
+						    NULL, 0, 0, 4, 1, true);
+		if (IS_ERR(hw))
+			return PTR_ERR(hw);
+
+		rs9->clk_dif[i] = hw;
+	}
+
+	ret = devm_of_clk_add_hw_provider(&client->dev, rs9_of_clk_get, rs9);
+	if (!ret)
+		rs9_update_config(rs9);
+
+	return ret;
+}
+
+static int __maybe_unused rs9_suspend(struct device *dev)
+{
+	struct rs9_driver_data *rs9 = dev_get_drvdata(dev);
+
+	regcache_cache_only(rs9->regmap, true);
+	regcache_mark_dirty(rs9->regmap);
+
+	return 0;
+}
+
+static int __maybe_unused rs9_resume(struct device *dev)
+{
+	struct rs9_driver_data *rs9 = dev_get_drvdata(dev);
+	int ret;
+
+	regcache_cache_only(rs9->regmap, false);
+	ret = regcache_sync(rs9->regmap);
+	if (ret)
+		dev_err(dev, "Failed to restore register map: %d\n", ret);
+	return ret;
+}
+
+static const struct rs9_chip_info renesas_9fgv0241_info = {
+	.model		= RENESAS_9FGV0241,
+	.num_clks	= 2,
+};
+
+static const struct i2c_device_id rs9_id[] = {
+	{ "9fgv0241", .driver_data = RENESAS_9FGV0241 },
+	{ }
+};
+MODULE_DEVICE_TABLE(i2c, rs9_id);
+
+static const struct of_device_id clk_rs9_of_match[] = {
+	{ .compatible = "renesas,9fgv0241", .data = &renesas_9fgv0241_info },
+	{ },
+};
+MODULE_DEVICE_TABLE(of, clk_rs9_of_match);
+
+static SIMPLE_DEV_PM_OPS(rs9_pm_ops, rs9_suspend, rs9_resume);
+
+static struct i2c_driver rs9_driver = {
+	.driver = {
+		.name = "clk-renesas-pcie-9series",
+		.pm	= &rs9_pm_ops,
+		.of_match_table = clk_rs9_of_match,
+	},
+	.probe		= rs9_probe,
+	.id_table	= rs9_id,
+};
+module_i2c_driver(rs9_driver);
+
+MODULE_AUTHOR("Marek Vasut <marex@denx.de>");
+MODULE_DESCRIPTION("Renesas 9-series PCIe clock generator driver");
+MODULE_LICENSE("GPL");
-- 
2.34.1

Re: [PATCH v3 3/3] clk: rs9: Add Renesas 9-series PCIe clock generator driver

[Marek Vasut]

On 2/26/22 05:07, Marek Vasut wrote:
> Add driver for Renesas 9-series PCIe clock generators. This driver
> is designed to support 9FGV/9DBV/9DMV/9FGL/9DML/9QXL/9SQ series I2C
> PCIe clock generators, currently the only tested and supported chip
> is 9FGV0241.
> 
> The driver is capable of configuring per-chip spread spectrum mode
> and output amplitude, as well as per-output slew rate.
> 
> Signed-off-by: Marek Vasut <marex@denx.de>
> Cc: Michael Turquette <mturquette@baylibre.com>
> Cc: Rob Herring <robh+dt@kernel.org>
> Cc: Stephen Boyd <sboyd@kernel.org>
> Cc: devicetree@vger.kernel.org
> To: linux-clk@vger.kernel.org
> ---
> V2: - Drop unused includes
>      - Use REGCACHE_FLAT for smaller reg file
>      - Move of_node_put() in rs9_get_output_config() a bit higher up
>      - Drop forward declaration of clk_rs9_of_match
>      - Use device_get_match_data() instead of of_device_get_match_data()
>        and check for its return value, verify it is non-NULL
>      - Use newly available __clk_hw_register_fixed_factor() with
>        parent_data index=0 and drop of_clk_get_parent_name() altogether
> V3: - Rename renesas,out-amplitude to renesas,out-amplitude-microvolt

Are there any news on this series ?

Re: [PATCH v3 3/3] clk: rs9: Add Renesas 9-series PCIe clock generator driver

[Stephen Boyd]

Quoting Marek Vasut (2022-02-25 20:07:23)
> Add driver for Renesas 9-series PCIe clock generators. This driver
> is designed to support 9FGV/9DBV/9DMV/9FGL/9DML/9QXL/9SQ series I2C
> PCIe clock generators, currently the only tested and supported chip
> is 9FGV0241.
> 
> The driver is capable of configuring per-chip spread spectrum mode
> and output amplitude, as well as per-output slew rate.
> 
> Signed-off-by: Marek Vasut <marex@denx.de>
> Cc: Michael Turquette <mturquette@baylibre.com>
> Cc: Rob Herring <robh+dt@kernel.org>
> Cc: Stephen Boyd <sboyd@kernel.org>
> Cc: devicetree@vger.kernel.org
> To: linux-clk@vger.kernel.org
> ---

Applied to clk-next

Re: [PATCH v3 1/3] dt-bindings: clk: rs9: Add Renesas 9-series I2C PCIe clock generator

[Rob Herring]

On Sat, 26 Feb 2022 05:07:21 +0100, Marek Vasut wrote:
> Add binding for Renesas 9-series PCIe clock generators. This binding
> is designed to support 9FGV/9DBV/9DMV/9FGL/9DML/9QXL/9SQ series I2C
> PCIe clock generators, currently the only tested and supported chip
> is 9FGV0241.
> 
> Signed-off-by: Marek Vasut <marex@denx.de>
> Cc: Michael Turquette <mturquette@baylibre.com>
> Cc: Rob Herring <robh+dt@kernel.org>
> Cc: Stephen Boyd <sboyd@kernel.org>
> Cc: devicetree@vger.kernel.org
> To: linux-clk@vger.kernel.org
> ---
> V2: - Drop clock consumer from the binding example
>     - Make clocks, i.e. xtal, mandatory
> V3: - Rename renesas,out-amplitude to renesas,out-amplitude-microvolt
>     - Drop type ref: from renesas,out-amplitude-microvolt property
>     - Explain 'pcm' in description text as 1/1000 of percent
>     - Add newlines
> ---
>  .../bindings/clock/renesas,9series.yaml       | 97 +++++++++++++++++++
>  1 file changed, 97 insertions(+)
>  create mode 100644 Documentation/devicetree/bindings/clock/renesas,9series.yaml
> 

Reviewed-by: Rob Herring <robh@kernel.org>

Re: [PATCH v3 1/3] dt-bindings: clk: rs9: Add Renesas 9-series I2C PCIe clock generator

[Stephen Boyd]

Quoting Marek Vasut (2022-02-25 20:07:21)
> Add binding for Renesas 9-series PCIe clock generators. This binding
> is designed to support 9FGV/9DBV/9DMV/9FGL/9DML/9QXL/9SQ series I2C
> PCIe clock generators, currently the only tested and supported chip
> is 9FGV0241.
> 
> Signed-off-by: Marek Vasut <marex@denx.de>
> Cc: Michael Turquette <mturquette@baylibre.com>
> Cc: Rob Herring <robh+dt@kernel.org>
> Cc: Stephen Boyd <sboyd@kernel.org>
> Cc: devicetree@vger.kernel.org
> To: linux-clk@vger.kernel.org
> ---

Applied to clk-next