[RFC PATCH 00/11] Rules for simple-mfd child nodes

[RFC PATCH 00/11] Rules for simple-mfd child nodes

[Conor Dooley]

From: Conor Dooley <conor.dooley@microchip.com>

Yo,

Please ignore the whole thing, except for the "dt-bindings: soc" patch.
As I mentioned on IRC earlier, I just do not understand when it is or is
not appropriate to have a child node in a simple-mfd and what the rules
about it are. The patch itself repeats the questions in more detail. The
rest of this is here for context and is very much something that is
WIP...

Cheers,
Conor.

CC: Lee Jones <lee@kernel.org>
CC: Rob Herring <robh@kernel.org>
CC: Krzysztof Kozlowski <krzk+dt@kernel.org>
CC: Conor Dooley <conor+dt@kernel.org>
CC: devicetree@vger.kernel.org
CC: linux-kernel@vger.kernel.org

Conor Dooley (10):
  dt-bindings: mailbox: mpfs: fix reg properties
  mailbox: mpfs: support fixed binding (TODO: always use regmap)
  riscv: dts: microchip: fix mailbox description (TODO drop 3rd syscon
    from here)
  dt-bindings: mfd: syscon document the non simple-mfd syscon on
    PolarFire SoC
  dt-bindings: soc: microchip: document the two simple-mfd syscons on
    PolarFire SoC
  reset: mpfs: add non-auxiliary bus probing
  copy meson clk-regmap for now
  clk: microchip: mpfs: use regmap clock types
  dt-bindings: clk: microchip: mpfs: remove first reg region
  riscv: dts: microchip: convert clock and reset (TODO: fixup phandle)

Lars Randers (1):
  hwmon: add a driver for the temp/voltage sensor on PolarFire SoC

 .../bindings/clock/microchip,mpfs-clkcfg.yaml |  33 +-
 .../mailbox/microchip,mpfs-mailbox.yaml       |  10 +-
 .../devicetree/bindings/mfd/syscon.yaml       |   2 +
 .../microchip/microchip,mpfs-control-scb.yaml |  54 +++
 .../microchip,mpfs-mss-top-sysreg.yaml        |  53 +++
 arch/riscv/boot/dts/microchip/mpfs.dtsi       |  44 +-
 drivers/clk/microchip/Makefile                |   1 +
 drivers/clk/microchip/clk-mpfs.c              |  81 ++--
 drivers/clk/microchip/clk-regmap.c            | 186 +++++++++
 drivers/clk/microchip/clk-regmap.h            | 137 +++++++
 drivers/hwmon/Kconfig                         |  12 +
 drivers/hwmon/Makefile                        |   1 +
 drivers/hwmon/tvs-mpfs.c                      | 379 ++++++++++++++++++
 drivers/mailbox/mailbox-mpfs.c                |  46 ++-
 drivers/reset/reset-mpfs.c                    |  86 +++-
 15 files changed, 1055 insertions(+), 70 deletions(-)
 create mode 100644 Documentation/devicetree/bindings/soc/microchip/microchip,mpfs-control-scb.yaml
 create mode 100644 Documentation/devicetree/bindings/soc/microchip/microchip,mpfs-mss-top-sysreg.yaml
 create mode 100644 drivers/clk/microchip/clk-regmap.c
 create mode 100644 drivers/clk/microchip/clk-regmap.h
 create mode 100644 drivers/hwmon/tvs-mpfs.c

-- 
2.43.0

[RFC PATCH 01/11] dt-bindings: mailbox: mpfs: fix reg properties

[Conor Dooley]

From: Conor Dooley <conor.dooley@microchip.com>

When the binding for this was originally written, and later modified,
mistakes were made - and the precise nature of the later modification
should have been a giveaway, but alas I was naive at the time.

A more correct modelling of the hardware is to use two syscons and have
a single reg entry for the mailbox, containing the mailbox region. The
two syscons contain the general control/status registers for the mailbox
and the interrupt related registers respectively. The reason for two
syscons is that the same mailbox is present on the non-SoC version of
the FPGA, which has no interrupt controller, and the shared part of the
rtl was unchanged between devices.

This is now coming to a head, because the control/status registers share
a register region with the "tvs" (temperature & voltage sensors)
registers and, as it turns out, people do want to monitor temperatures
and voltages...

Signed-off-by: Conor Dooley <conor.dooley@microchip.com>
---
 .../bindings/mailbox/microchip,mpfs-mailbox.yaml       | 10 +++++-----
 1 file changed, 5 insertions(+), 5 deletions(-)

diff --git a/Documentation/devicetree/bindings/mailbox/microchip,mpfs-mailbox.yaml b/Documentation/devicetree/bindings/mailbox/microchip,mpfs-mailbox.yaml
index 404477910f02..3af599efd359 100644
--- a/Documentation/devicetree/bindings/mailbox/microchip,mpfs-mailbox.yaml
+++ b/Documentation/devicetree/bindings/mailbox/microchip,mpfs-mailbox.yaml
@@ -15,6 +15,8 @@ properties:
 
   reg:
     oneOf:
+      - items:
+          - description: mailbox data registers
       - items:
           - description: mailbox control & data registers
           - description: mailbox interrupt registers
@@ -23,6 +25,7 @@ properties:
           - description: mailbox control registers
           - description: mailbox interrupt registers
           - description: mailbox data registers
+        deprecated: true
 
   interrupts:
     maxItems: 1
@@ -41,12 +44,9 @@ additionalProperties: false
 examples:
   - |
     soc {
-      #address-cells = <2>;
-      #size-cells = <2>;
-      mbox: mailbox@37020000 {
+      mailbox@37020800 {
         compatible = "microchip,mpfs-mailbox";
-        reg = <0x0 0x37020000 0x0 0x58>, <0x0 0x2000318C 0x0 0x40>,
-              <0x0 0x37020800 0x0 0x100>;
+        reg = <0x37020800 0x100>;
         interrupt-parent = <&L1>;
         interrupts = <96>;
         #mbox-cells = <1>;
-- 
2.43.0

[RFC PATCH 02/11] hwmon: add a driver for the temp/voltage sensor on PolarFire SoC

[Conor Dooley]

From: Lars Randers <lranders@mail.dk>

Signed-off-by: Lars Randers <lranders@mail.dk>
Signed-off-by: Conor Dooley <conor.dooley@microchip.com>
---
 drivers/hwmon/Kconfig          |  12 ++
 drivers/hwmon/Makefile         |   1 +
 drivers/hwmon/tvs-mpfs.c       | 379 +++++++++++++++++++++++++++++++++
 drivers/mailbox/mailbox-mpfs.c |   1 +
 4 files changed, 393 insertions(+)
 create mode 100644 drivers/hwmon/tvs-mpfs.c

diff --git a/drivers/hwmon/Kconfig b/drivers/hwmon/Kconfig
index b60fe2e58ad6..2d2bed68dcad 100644
--- a/drivers/hwmon/Kconfig
+++ b/drivers/hwmon/Kconfig
@@ -2295,6 +2295,18 @@ config SENSORS_TMP513
 	  This driver can also be built as a module. If so, the module
 	  will be called tmp513.
 
+config SENSORS_TVS_MPFS
+	tristate "PolarFire SoC (MPFS) temperature and voltage sensor"
+	depends on POLARFIRE_SOC_MAILBOX
+	help
+	  This driver adds support for the PolarFire SoC (MPFS) Temperature and
+	  Voltage Sensor.
+
+	  To compile this driver as a module, choose M here. the
+	  module will be called tvs-mpfs.
+
+	  If unsure, say N.
+
 config SENSORS_VEXPRESS
 	tristate "Versatile Express"
 	depends on VEXPRESS_CONFIG
diff --git a/drivers/hwmon/Makefile b/drivers/hwmon/Makefile
index b1c7056c37db..7f44c2567008 100644
--- a/drivers/hwmon/Makefile
+++ b/drivers/hwmon/Makefile
@@ -220,6 +220,7 @@ obj-$(CONFIG_SENSORS_TMP401)	+= tmp401.o
 obj-$(CONFIG_SENSORS_TMP421)	+= tmp421.o
 obj-$(CONFIG_SENSORS_TMP464)	+= tmp464.o
 obj-$(CONFIG_SENSORS_TMP513)	+= tmp513.o
+obj-$(CONFIG_SENSORS_TVS_MPFS)  += tvs-mpfs.o
 obj-$(CONFIG_SENSORS_VEXPRESS)	+= vexpress-hwmon.o
 obj-$(CONFIG_SENSORS_VIA_CPUTEMP)+= via-cputemp.o
 obj-$(CONFIG_SENSORS_VIA686A)	+= via686a.o
diff --git a/drivers/hwmon/tvs-mpfs.c b/drivers/hwmon/tvs-mpfs.c
new file mode 100644
index 000000000000..6f117d0b7db6
--- /dev/null
+++ b/drivers/hwmon/tvs-mpfs.c
@@ -0,0 +1,379 @@
+// SPDX-License-Identifier: GPL-2.0+
+/*
+ *
+ * Author: Lars Randers <lranders@mail.dk>
+ *
+ */
+
+#include <linux/err.h>
+#include <linux/freezer.h>
+#include <linux/kthread.h>
+#include <linux/hwmon.h>
+#include <linux/io.h>
+#include <linux/mfd/syscon.h>
+#include <linux/module.h>
+#include <linux/of_address.h>
+#include <linux/platform_device.h>
+#include <linux/regmap.h>
+
+#define PFSOC_CONTROL_SCB_TVS_CONTROL 0x08
+#define PFSOC_CONTROL_SCB_TVS_OUTPUT0 0x24
+#define PFSOC_CONTROL_SCB_TVS_OUTPUT1 0x28
+
+#define CTRL_POWEROFF BIT(5)
+#define CTRL_ABORT    BIT(4)
+#define CTRL_TEMP     BIT(3)
+#define CTRL_2P5      BIT(2)
+#define CTRL_1P8      BIT(1)
+#define CTRL_1P05     BIT(0)
+
+#define OUTPUT0_U1P8_MASK GENMASK(30, 16)
+#define OUTPUT0_U1P8_OFF  16
+#define OUTPUT0_U1P0_MASK GENMASK(14, 0)
+#define OUTPUT0_U1P0_OFF  0
+#define OUTPUT1_TEMP_MASK GENMASK(31, 16)
+#define OUTPUT1_TEMP_OFF  16
+#define OUTPUT1_U2P5_MASK GENMASK(14, 0)
+#define OUTPUT1_U2P5_OFF  0
+
+#define MPFS_TVS_MIN_POLL_INTERVAL_IN_MILLIS 2000
+
+/* The following constant is 273.5 in (16.4) fixedpoint notation */
+#define MPFS_TVS_MIN_TEMP_IN_K 0x1112
+
+typedef struct {
+	long min;
+	long actual;
+	long max;
+} mpfs_tvs_sensor_t;
+
+typedef enum {
+	SN_V1P05 = 0,
+	SN_V1P8,
+	SN_V2P5,
+	SN_TEMP,
+
+	SN_MAX
+} mpfs_tvs_sn_t;
+
+static const char *mpfs_tvs_voltage_labels[] = { "U1P05", "U1P8", "U2P5" };
+
+struct mpfs_tvs {
+	struct device *dev;
+	struct device *hwmon_dev;
+	struct task_struct *poll_task;
+	struct regmap *regmap;
+	bool kthread_running;
+	long update_interval;	/* in milli-seconds */
+	mpfs_tvs_sensor_t sensors[SN_MAX];
+};
+
+static int mpfs_tvs_update_sensors(struct mpfs_tvs *data) {
+	u32 temp;
+	u32 work;
+	int ret;
+
+	ret = regmap_read(data->regmap, PFSOC_CONTROL_SCB_TVS_OUTPUT1, &temp);
+	if (ret)
+		return ret;
+
+	work = temp;
+
+	temp = (temp & OUTPUT1_TEMP_MASK) >> OUTPUT1_TEMP_OFF;
+	temp = clamp_val(temp, MPFS_TVS_MIN_TEMP_IN_K, INT_MAX);
+	temp = temp - MPFS_TVS_MIN_TEMP_IN_K; /* Kelvin to Celsius */
+	temp = (1000 * temp) >> 4; /* fixed point (10.4) to millicentigrade */
+	data->sensors[SN_TEMP].actual = temp;
+	data->sensors[SN_TEMP].max =
+		max(data->sensors[SN_TEMP].actual, data->sensors[SN_TEMP].max);
+	data->sensors[SN_TEMP].min =
+		min(data->sensors[SN_TEMP].min, data->sensors[SN_TEMP].actual);
+
+	work &= OUTPUT1_U2P5_MASK;
+	/* fixed point (11.3) adjust; value is already millivolts */
+	work  = (1 * work) >> 3;
+	data->sensors[SN_V2P5].actual = work;
+	data->sensors[SN_V2P5].max =
+		max(data->sensors[SN_V2P5].actual, data->sensors[SN_V2P5].max);
+	data->sensors[SN_V2P5].min =
+		min(data->sensors[SN_V2P5].min, data->sensors[SN_V2P5].actual);
+
+	ret = regmap_read(data->regmap, PFSOC_CONTROL_SCB_TVS_OUTPUT0, &temp);
+	if (ret)
+		return ret;
+
+	work = temp;
+	temp = (OUTPUT0_U1P8_MASK & temp) >> OUTPUT0_U1P8_OFF;
+	/* fixed point (11.3) adjust; value is already millivolts */
+	temp = (1 * temp) >> 3;
+	data->sensors[SN_V1P8].actual = temp;
+	data->sensors[SN_V1P8].max =
+		max(data->sensors[SN_V1P8].actual, data->sensors[SN_V1P8].max);
+	data->sensors[SN_V2P5].min =
+		min(data->sensors[SN_V1P8].min, data->sensors[SN_V1P8].actual);
+
+	work &= OUTPUT0_U1P0_MASK;
+	 /* fixed point (11.3) adjust; value is already millivolts */
+	work  = (1 * work) >> 3;
+	data->sensors[SN_V1P05].actual = work;
+	data->sensors[SN_V1P05].max =
+		max(data->sensors[SN_V1P05].actual, data->sensors[SN_V1P05].max);
+	data->sensors[SN_V1P05].min =
+		min(data->sensors[SN_V1P05].min, data->sensors[SN_V1P05].actual);
+
+	return 0;
+}
+
+
+static int mpfs_tvs_chip_read(struct mpfs_tvs *data, long *val)
+{
+	*val = data->update_interval;
+	return 0;
+}
+
+static int mpfs_tvs_temp_read(struct mpfs_tvs *data, u32 attr,
+			      int channel, long *val)
+{
+	switch(attr) {
+	case hwmon_temp_input:
+		*val = data->sensors[SN_TEMP].actual;
+		break;
+
+	case hwmon_temp_max:
+		*val = data->sensors[SN_TEMP].max;
+		break;
+
+	default:
+		return -EOPNOTSUPP;
+	}
+	return 0;
+}
+
+static int mpfs_tvs_voltage_read(struct mpfs_tvs *data, u32 attr,
+				 int channel, long *val)
+{
+	dev_dbg(data->dev, "read voltage chan %d\n", channel);
+	switch(attr) {
+	case hwmon_in_input:
+		*val = data->sensors[channel].actual;
+		break;
+
+	case hwmon_in_max:
+		*val = data->sensors[channel].max;
+		break;
+
+	default:
+		return -EOPNOTSUPP;
+	}
+	return 0;
+}
+
+
+static ssize_t mpfs_tvs_interval_write(struct mpfs_tvs *data, long val)
+{
+	data->update_interval =
+		clamp_val(val, MPFS_TVS_MIN_POLL_INTERVAL_IN_MILLIS, INT_MAX);
+	return 0;
+}
+
+
+static umode_t mpfs_tvs_is_visible(const void *data,
+				   enum hwmon_sensor_types type,
+				   u32 attr, int channel)
+{
+	if(type == hwmon_chip && attr == hwmon_chip_update_interval)
+		return 0644;
+
+	if(type == hwmon_temp) {
+		switch(attr) {
+		case hwmon_temp_input:
+		case hwmon_temp_max:
+		case hwmon_temp_label:
+			return 0444;
+
+		default:
+			return 0;
+		}
+	} else if(type == hwmon_in) {
+		switch(attr) {
+		case hwmon_in_input:
+		case hwmon_in_label:
+			return 0444;
+
+		default:
+			return 0;
+		}
+	}
+	return 0;
+}
+
+static int mpfs_tvs_read(struct device *dev, enum hwmon_sensor_types type,
+			 u32 attr, int channel, long *val)
+{
+	struct mpfs_tvs *data = dev_get_drvdata(dev);
+
+	switch(type) {
+	case hwmon_temp:
+		return mpfs_tvs_temp_read(data, attr, channel, val);
+	case hwmon_in:
+		return mpfs_tvs_voltage_read(data, attr, channel, val);
+	case hwmon_chip:
+		return mpfs_tvs_chip_read(data, val);
+
+	default:
+		return -EOPNOTSUPP;
+	}
+}
+
+static int mpfs_tvs_write(struct device *dev, enum hwmon_sensor_types type,
+			  u32 attr, int channel, long val)
+{
+	struct mpfs_tvs *data = dev_get_drvdata(dev);
+
+	switch(type) {
+	case hwmon_chip:
+		return mpfs_tvs_interval_write(data, val);
+	default:
+		return -EOPNOTSUPP;
+	}
+}
+
+static int mpfs_tvs_read_labels(struct device *dev,
+				enum hwmon_sensor_types type,
+				u32 attr, int channel,
+				const char **str)
+{
+	switch(type) {
+	case hwmon_temp:
+		*str = "CPU Temp";
+		return 0;
+	case hwmon_in:
+		*str = mpfs_tvs_voltage_labels[channel];
+		return 0;
+	default:
+		return -EOPNOTSUPP;
+	}
+}
+
+
+static const struct hwmon_ops mpfs_tvs_ops = {
+	.is_visible = mpfs_tvs_is_visible,
+	.read_string = mpfs_tvs_read_labels,
+	.read = mpfs_tvs_read,
+	.write = mpfs_tvs_write,
+};
+
+static const struct hwmon_channel_info *mpfs_tvs_info[] = {
+	HWMON_CHANNEL_INFO(chip,
+			   HWMON_C_REGISTER_TZ | HWMON_C_UPDATE_INTERVAL),
+	HWMON_CHANNEL_INFO(temp,
+			   HWMON_T_INPUT | HWMON_T_MIN |
+			   HWMON_T_MAX | HWMON_T_LABEL),
+	HWMON_CHANNEL_INFO(in,
+			   HWMON_I_INPUT | HWMON_I_LABEL,
+			   HWMON_I_INPUT | HWMON_I_LABEL,
+			   HWMON_I_INPUT | HWMON_I_LABEL),
+	NULL
+};
+
+static const struct hwmon_chip_info mpfs_tvs_chip_info = {
+	.ops = &mpfs_tvs_ops,
+	.info = mpfs_tvs_info,
+};
+
+
+static int mpfs_tvs_poll_task(void *ptr)
+{
+	struct mpfs_tvs *data = ptr;
+	int ret = 0;
+
+	data->kthread_running = true;
+
+	set_freezable();
+
+	while(!kthread_should_stop()) {
+		schedule_timeout_interruptible(data->update_interval);
+		try_to_freeze();
+		ret = mpfs_tvs_update_sensors(data);
+		if(ret)
+			break;
+	}
+
+	data->kthread_running = false;
+	return ret;
+}
+
+static int mpfs_tvs_probe(struct platform_device *pdev)
+{
+	struct device *hwmon_dev;
+	struct mpfs_tvs *data;
+	struct task_struct *task;
+	int err;
+
+	data = devm_kzalloc(&pdev->dev, sizeof(*data), GFP_KERNEL);
+	if(!data)
+		return -ENOMEM;
+
+	data->dev = &pdev->dev;
+
+	data->regmap = syscon_node_to_regmap(data->dev->of_node->parent);
+
+	data->kthread_running = false;
+
+	hwmon_dev = devm_hwmon_device_register_with_info(data->dev, "mpfs_tvs",
+							 data,
+							 &mpfs_tvs_chip_info,
+							 NULL);
+
+	if(IS_ERR(hwmon_dev)) {
+		err = PTR_ERR(hwmon_dev);
+		dev_err(data->dev, "Class registration failed (%d)\n", err);
+		return err;
+	}
+
+	/* enable HW sensor */
+	err = regmap_write(data->regmap, PFSOC_CONTROL_SCB_TVS_CONTROL,
+			   CTRL_1P05 | CTRL_1P8 | CTRL_2P5 | CTRL_TEMP);
+
+	data->hwmon_dev = hwmon_dev;
+	data->sensors[SN_TEMP].max = 0;
+	data->sensors[SN_V1P05].min =
+		data->sensors[SN_V1P8].min =
+		data->sensors[SN_V2P5].min = 20000;
+	data->sensors[SN_V1P05].max =
+		data->sensors[SN_V1P8].max =
+		data->sensors[SN_V2P5].max = 0;
+	data->update_interval = MPFS_TVS_MIN_POLL_INTERVAL_IN_MILLIS;
+	mpfs_tvs_update_sensors(data);
+
+	task = kthread_run(mpfs_tvs_poll_task, data, "tvs-mpfs-kthread");
+	if (IS_ERR(task)) {
+		err = PTR_ERR(task);
+		dev_err(data->dev, "Unable to run kthread err %d\n", err);
+		return err;
+	}
+
+	data->poll_task = task;
+
+	dev_info(data->dev, "Registered MPFS TVS auxiliary driver\n");
+	return 0;
+}
+
+static const struct of_device_id mpfs_tvs_of_match[] = {
+	{ .compatible = "microchip,mpfs-tvs", },
+	{},
+};
+MODULE_DEVICE_TABLE(of, mpfs_tvs_of_match);
+
+static struct platform_driver mpfs_tvs_driver = {
+	.probe		= mpfs_tvs_probe,
+	.driver = {
+		.name = "mpfs-tvs",
+		.of_match_table = mpfs_tvs_of_match,
+	},
+};
+module_platform_driver(mpfs_tvs_driver);
+
+MODULE_AUTHOR("Lars Randers <lranders@mail.dk>");
+MODULE_DESCRIPTION("PolarFire SoC temperature & voltage sensor driver");
+MODULE_LICENSE("GPL");
diff --git a/drivers/mailbox/mailbox-mpfs.c b/drivers/mailbox/mailbox-mpfs.c
index 20ee283a04cc..0fd83bdd4cee 100644
--- a/drivers/mailbox/mailbox-mpfs.c
+++ b/drivers/mailbox/mailbox-mpfs.c
@@ -262,6 +262,7 @@ static int mpfs_mbox_probe(struct platform_device *pdev)
 		dev_err(&pdev->dev, "Registering MPFS mailbox controller failed\n");
 		return ret;
 	}
+
 	dev_info(&pdev->dev, "Registered MPFS mailbox controller driver\n");
 
 	return 0;
-- 
2.43.0

[RFC PATCH 03/11] mailbox: mpfs: support fixed binding (TODO: always use regmap)

[Conor Dooley]

From: Conor Dooley <conor.dooley@microchip.com>

The two previous bindings for this hardware were incorrect, as the
control/status and interrupt register regions should have been described
as syscons and dealt with via regmap in the driver. Add support for
accessing these registers using that method now, so that the hwmon
driver can be supported without using auxdev or hacks with io_remap().

Signed-off-by: Conor Dooley <conor.dooley@microchip.com>
---
 drivers/mailbox/mailbox-mpfs.c | 45 +++++++++++++++++++++++++++++++---
 1 file changed, 41 insertions(+), 4 deletions(-)

diff --git a/drivers/mailbox/mailbox-mpfs.c b/drivers/mailbox/mailbox-mpfs.c
index 0fd83bdd4cee..65aa466ffe8b 100644
--- a/drivers/mailbox/mailbox-mpfs.c
+++ b/drivers/mailbox/mailbox-mpfs.c
@@ -13,12 +13,15 @@
 #include <linux/init.h>
 #include <linux/module.h>
 #include <linux/kernel.h>
+#include <linux/regmap.h>
 #include <linux/interrupt.h>
+#include <linux/mfd/syscon.h>
 #include <linux/mod_devicetable.h>
 #include <linux/platform_device.h>
 #include <linux/mailbox_controller.h>
 #include <soc/microchip/mpfs.h>
 
+#define MESSAGE_INT_OFFSET		0x18cu
 #define SERVICES_CR_OFFSET		0x50u
 #define SERVICES_SR_OFFSET		0x54u
 #define MAILBOX_REG_OFFSET		0x800u
@@ -68,6 +71,7 @@ struct mpfs_mbox {
 	void __iomem *int_reg;
 	struct mbox_chan chans[1];
 	struct mpfs_mss_response *response;
+	struct regmap *sysreg_scb, *control_scb;
 	u16 resp_offset;
 };
 
@@ -75,7 +79,10 @@ static bool mpfs_mbox_busy(struct mpfs_mbox *mbox)
 {
 	u32 status;
 
-	status = readl_relaxed(mbox->ctrl_base + SERVICES_SR_OFFSET);
+	if (mbox->control_scb)
+		regmap_read(mbox->control_scb, SERVICES_SR_OFFSET, &status);
+	else
+		status = readl_relaxed(mbox->ctrl_base + SERVICES_SR_OFFSET);
 
 	return status & SCB_STATUS_BUSY_MASK;
 }
@@ -95,7 +102,11 @@ static bool mpfs_mbox_last_tx_done(struct mbox_chan *chan)
 	 * Failed services are intended to generated interrupts, but in reality
 	 * this does not happen, so the status must be checked here.
 	 */
-	val = readl_relaxed(mbox->ctrl_base + SERVICES_SR_OFFSET);
+	if (mbox->control_scb)
+		regmap_read(mbox->control_scb, SERVICES_SR_OFFSET, &val);
+	else
+		val = readl_relaxed(mbox->ctrl_base + SERVICES_SR_OFFSET);
+
 	response->resp_status = (val & SCB_STATUS_MASK) >> SCB_STATUS_POS;
 
 	return true;
@@ -143,7 +154,12 @@ static int mpfs_mbox_send_data(struct mbox_chan *chan, void *data)
 
 	tx_trigger = (opt_sel << SCB_CTRL_POS) & SCB_CTRL_MASK;
 	tx_trigger |= SCB_CTRL_REQ_MASK | SCB_STATUS_NOTIFY_MASK;
-	writel_relaxed(tx_trigger, mbox->ctrl_base + SERVICES_CR_OFFSET);
+
+	if (mbox->control_scb)
+		regmap_write(mbox->control_scb, SERVICES_CR_OFFSET, tx_trigger);
+	else
+		writel_relaxed(tx_trigger, mbox->ctrl_base + SERVICES_CR_OFFSET);
+
 
 	return 0;
 }
@@ -185,7 +201,10 @@ static irqreturn_t mpfs_mbox_inbox_isr(int irq, void *data)
 	struct mbox_chan *chan = data;
 	struct mpfs_mbox *mbox = (struct mpfs_mbox *)chan->con_priv;
 
-	writel_relaxed(0, mbox->int_reg);
+	if (mbox->control_scb)
+		regmap_write(mbox->sysreg_scb, MESSAGE_INT_OFFSET, 0);
+	else
+		writel_relaxed(0, mbox->int_reg);
 
 	mpfs_mbox_rx_data(chan);
 
@@ -231,6 +250,23 @@ static int mpfs_mbox_probe(struct platform_device *pdev)
 	if (!mbox)
 		return -ENOMEM;
 
+	mbox->control_scb = syscon_regmap_lookup_by_compatible("microchip,mpfs-control-scb");
+	if (IS_ERR(mbox->control_scb)) {
+		mbox->control_scb = NULL;
+		goto old_format;
+	}
+
+	mbox->sysreg_scb = syscon_regmap_lookup_by_compatible("microchip,mpfs-sysreg-scb");
+	if (IS_ERR(mbox->sysreg_scb))
+		return -1;
+
+	mbox->mbox_base = devm_platform_get_and_ioremap_resource(pdev, 0, &regs);
+	if (IS_ERR(mbox->ctrl_base))
+		return PTR_ERR(mbox->mbox_base);
+
+	goto done;
+
+old_format:
 	mbox->ctrl_base = devm_platform_get_and_ioremap_resource(pdev, 0, &regs);
 	if (IS_ERR(mbox->ctrl_base))
 		return PTR_ERR(mbox->ctrl_base);
@@ -243,6 +279,7 @@ static int mpfs_mbox_probe(struct platform_device *pdev)
 	if (IS_ERR(mbox->mbox_base)) // account for the old dt-binding w/ 2 regs
 		mbox->mbox_base = mbox->ctrl_base + MAILBOX_REG_OFFSET;
 
+done:
 	mbox->irq = platform_get_irq(pdev, 0);
 	if (mbox->irq < 0)
 		return mbox->irq;
-- 
2.43.0

[RFC PATCH 04/11] riscv: dts: microchip: fix mailbox description (TODO drop 3rd syscon from here)

[Conor Dooley]

From: Conor Dooley <conor.dooley@microchip.com>

When the binding for the mailbox on PolarFire SoC was originally
written, and later modified, mistakes were made - and the precise
nature of the later modification should have been a giveaway, but alas
I was naive at the time.

A more correct modelling of the hardware is to use two syscons and have
a single reg entry for the mailbox, containing the mailbox region. The
two syscons contain the general control/status registers for the mailbox
and the interrupt related registers respectively. The reason for two
syscons is that the same mailbox is present on the non-SoC version of
the FPGA, which has no interrupt controller, and the shared part of the
rtl was unchanged between devices.

Signed-off-by: Conor Dooley <conor.dooley@microchip.com>
---
 arch/riscv/boot/dts/microchip/mpfs.dtsi | 24 +++++++++++++++++++++---
 1 file changed, 21 insertions(+), 3 deletions(-)

diff --git a/arch/riscv/boot/dts/microchip/mpfs.dtsi b/arch/riscv/boot/dts/microchip/mpfs.dtsi
index 9883ca3554c5..1d655126b66f 100644
--- a/arch/riscv/boot/dts/microchip/mpfs.dtsi
+++ b/arch/riscv/boot/dts/microchip/mpfs.dtsi
@@ -251,6 +251,11 @@ pdma: dma-controller@3000000 {
 			#dma-cells = <1>;
 		};
 
+		mss_top_scb: syscon@20002000 {
+			compatible = "microchip,mpfs-mss-top-sysreg", "syscon", "simple-mfd";
+			reg = <0x0 0x20002000 0x0 0x1000>;
+		};
+
 		clkcfg: clkcfg@20002000 {
 			compatible = "microchip,mpfs-clkcfg";
 			reg = <0x0 0x20002000 0x0 0x1000>, <0x0 0x3E001000 0x0 0x1000>;
@@ -259,6 +264,11 @@ clkcfg: clkcfg@20002000 {
 			#reset-cells = <1>;
 		};
 
+		sysreg_scb: syscon@20003000 {
+			compatible = "microchip,mpfs-sysreg-scb", "syscon";
+			reg = <0x0 0x20003000 0x0 0x1000>;
+		};
+
 		ccc_se: clock-controller@38010000 {
 			compatible = "microchip,mpfs-ccc";
 			reg = <0x0 0x38010000 0x0 0x1000>, <0x0 0x38020000 0x0 0x1000>,
@@ -521,10 +531,18 @@ usb: usb@20201000 {
 			status = "disabled";
 		};
 
-		mbox: mailbox@37020000 {
+		control_scb: syscon@37020000 {
+			compatible = "microchip,mpfs-control-scb", "syscon", "simple-mfd";
+			reg = <0x0 0x37020000 0x0 0x100>;
+
+			sensor {
+				compatible = "microchip,mpfs-tvs";
+			};
+		};
+
+		mbox: mailbox@37020800 {
 			compatible = "microchip,mpfs-mailbox";
-			reg = <0x0 0x37020000 0x0 0x58>, <0x0 0x2000318C 0x0 0x40>,
-			      <0x0 0x37020800 0x0 0x100>;
+			reg = <0x0 0x37020800 0x0 0x100>;
 			interrupt-parent = <&plic>;
 			interrupts = <96>;
 			#mbox-cells = <1>;
-- 
2.43.0

[RFC PATCH 05/11] dt-bindings: mfd: syscon document the non simple-mfd syscon on PolarFire SoC

[Conor Dooley]

From: Conor Dooley <conor.dooley@microchip.com>

The "mss_top_scb" register region on PolarFire SoC contains many
different functions, including controls for the AXI bus and other things
mainly of interest to the bootloader. The interrupt register for the
system controller's mailbox is also in here, which is needed by the
operating system.

Signed-off-by: Conor Dooley <conor.dooley@microchip.com>
---
 Documentation/devicetree/bindings/mfd/syscon.yaml | 2 ++
 1 file changed, 2 insertions(+)

diff --git a/Documentation/devicetree/bindings/mfd/syscon.yaml b/Documentation/devicetree/bindings/mfd/syscon.yaml
index 9dc594ea3654..6e6eda8afeed 100644
--- a/Documentation/devicetree/bindings/mfd/syscon.yaml
+++ b/Documentation/devicetree/bindings/mfd/syscon.yaml
@@ -88,6 +88,7 @@ select:
           - mediatek,mt8173-pctl-a-syscfg
           - mediatek,mt8365-syscfg
           - microchip,lan966x-cpu-syscon
+          - microchip,mpfs-sysreg-scb
           - microchip,sam9x60-sfr
           - microchip,sama7g5-ddr3phy
           - mscc,ocelot-cpu-syscon
@@ -183,6 +184,7 @@ properties:
           - mediatek,mt8173-pctl-a-syscfg
           - mediatek,mt8365-syscfg
           - microchip,lan966x-cpu-syscon
+          - microchip,mpfs-sysreg-scb
           - microchip,sam9x60-sfr
           - microchip,sama7g5-ddr3phy
           - mscc,ocelot-cpu-syscon
-- 
2.43.0

[RFC PATCH 06/11] dt-bindings: soc: microchip: document the two simple-mfd syscons on PolarFire SoC

[Conor Dooley]

From: Conor Dooley <conor.dooley@microchip.com>

There are two syscons on PolarFire SoC that provide various functionality of
use to the OS.

The first of these is the "control-scb" region, that contains the "tvs"
temperature and voltage sensors and the control/status registers for the
system controller's mailbox. The mailbox has a dedicated node, so
there's no need for a child node describing it, looking the syscon up by
compatible is sufficient.

The second, "mss-top-sysreg", contains clocks, pinctrl, resets, and
interrupt controller and more. For this RFC, only the reset controller
child is described as that's all that is described by the existing
bindings. The clock controller already has a dedicated node, and will
retain it as there are other clock regions, so like the mailbox,
a compatible-based lookup of the syscon is sufficient to keep the clock
driver working as before so no child is needed.

Signed-off-by: Conor Dooley <conor.dooley@microchip.com>
---
(I'll split this in two later, it's just easier when I have the same
questions about both...)

Are these things entitled to have child nodes for the reset and sensor
nodes, or should the properties be in the parent and the OS probe the
drivers for the functions? That's something that, despite supposedly
being a maintainer, I do not understand the rules (of thumb?) for.

Secondly, is it okay to make the "pragmatic" decision to not have a
child clock node and keep routing the clocks via the existing & retained
clock node (and therefore not update the various clocks nodes in the
consumers)? Doing so would require a lot more hocus pocus with the clock
driver than this series does, as the same driver would no longer be
suitable for the before/after bindings.
---
 .../microchip/microchip,mpfs-control-scb.yaml | 54 +++++++++++++++++++
 .../microchip,mpfs-mss-top-sysreg.yaml        | 53 ++++++++++++++++++
 2 files changed, 107 insertions(+)
 create mode 100644 Documentation/devicetree/bindings/soc/microchip/microchip,mpfs-control-scb.yaml
 create mode 100644 Documentation/devicetree/bindings/soc/microchip/microchip,mpfs-mss-top-sysreg.yaml

diff --git a/Documentation/devicetree/bindings/soc/microchip/microchip,mpfs-control-scb.yaml b/Documentation/devicetree/bindings/soc/microchip/microchip,mpfs-control-scb.yaml
new file mode 100644
index 000000000000..3673bf139ce8
--- /dev/null
+++ b/Documentation/devicetree/bindings/soc/microchip/microchip,mpfs-control-scb.yaml
@@ -0,0 +1,54 @@
+# SPDX-License-Identifier: GPL-2.0-only OR BSD-2-Clause
+%YAML 1.2
+---
+$id: http://devicetree.org/schemas/soc/microchip/microchip,mpfs-control-scb.yaml#
+$schema: http://devicetree.org/meta-schemas/core.yaml#
+
+title: Microchip PolarFire SoC System Controller Bus (SCB) Control Register region
+
+maintainers:
+  - Conor Dooley <conor.dooley@microchip.com>
+
+description:
+  An assortment of system controller related registers, including voltage and
+  temperature sensors and the status/control registers for the system
+  controller's mailbox.
+
+properties:
+  compatible:
+    items:
+      - const: microchip,mpfs-control-scb
+      - const: syscon
+      - const: simple-mfd
+
+  reg:
+    maxItems: 1
+
+  sensor:
+    type: object
+
+    properties:
+      compatible:
+        const: microchip,mpfs-tvs
+
+    additionalProperties: false
+
+required:
+  - compatible
+  - reg
+
+additionalProperties: false
+
+examples:
+  - |
+    soc {
+      syscon@37020000 {
+        compatible = "microchip,mpfs-control-scb", "syscon", "simple-mfd";
+        reg = <0x37020000 0x100>;
+
+        sensor {
+          compatible = "microchip,mpfs-tvs";
+        };
+      };
+    };
+
diff --git a/Documentation/devicetree/bindings/soc/microchip/microchip,mpfs-mss-top-sysreg.yaml b/Documentation/devicetree/bindings/soc/microchip/microchip,mpfs-mss-top-sysreg.yaml
new file mode 100644
index 000000000000..d70c9c3348ac
--- /dev/null
+++ b/Documentation/devicetree/bindings/soc/microchip/microchip,mpfs-mss-top-sysreg.yaml
@@ -0,0 +1,53 @@
+# SPDX-License-Identifier: GPL-2.0-only OR BSD-2-Clause
+%YAML 1.2
+---
+$id: http://devicetree.org/schemas/soc/microchip/microchip,mpfs-mss-top-sysreg.yaml#
+$schema: http://devicetree.org/meta-schemas/core.yaml#
+
+title: Microchip PolarFire SoC Microprocessor Subsystem (MSS) sysreg Register region
+
+maintainers:
+  - Conor Dooley <conor.dooley@microchip.com>
+
+description:
+  An wide assortment of registers that control elements of the MSS on PolarFire
+  SoC, including pinmuxing, resets and clocks among others.
+
+properties:
+  compatible:
+    items:
+      - const: microchip,mpfs-mss-top-sysreg
+      - const: syscon
+      - const: simple-mfd
+
+  reg:
+    maxItems: 1
+
+  reset-controller:
+    type: object
+
+    properties:
+      compatible:
+        const: microchip,mpfs-reset
+
+      '#reset-cells':
+        description:
+          The AHB/AXI peripherals on the PolarFire SoC have reset support, so
+          from CLK_ENVM to CLK_CFM. The reset consumer should specify the
+          desired peripheral via the clock ID in its "resets" phandle cell.
+          See include/dt-bindings/clock/microchip,mpfs-clock.h for the full list
+          of PolarFire clock/reset IDs.
+        const: 1
+
+    additionalProperties: false
+required:
+  - compatible
+  - reg
+
+additionalProperties: false
+
+examples:
+  - |
+    soc {
+    };
+
-- 
2.43.0

[RFC PATCH 07/11] reset: mpfs: add non-auxiliary bus probing

[Conor Dooley]

From: Conor Dooley <conor.dooley@microchip.com>

While the auxiliary bus was a nice bandaid, and meant that re-writing
the representation of the clock regions in devicetree was not required,
it has run its course. The "mss_top_sysreg" region that contains the
clock and reset regions, also contains pinctrl and an interrupt
controller, so the time has come rewrite the devicetree and probe the
reset controller from a dedicated devicetree node, rather than implement
those drivers using the auxiliary bus. Wanting to avoid propagating this
naive/incorrect description of the hardware to the new pic64gx SoC is a
major motivating factor here.

Signed-off-by: Conor Dooley <conor.dooley@microchip.com>
---
 drivers/reset/reset-mpfs.c | 86 +++++++++++++++++++++++++++++++++-----
 1 file changed, 75 insertions(+), 11 deletions(-)

diff --git a/drivers/reset/reset-mpfs.c b/drivers/reset/reset-mpfs.c
index 710f9c1676f9..3b5c1f680e79 100644
--- a/drivers/reset/reset-mpfs.c
+++ b/drivers/reset/reset-mpfs.c
@@ -9,10 +9,12 @@
 #include <linux/auxiliary_bus.h>
 #include <linux/delay.h>
 #include <linux/io.h>
+#include <linux/mfd/syscon.h>
 #include <linux/module.h>
 #include <linux/of.h>
 #include <linux/platform_device.h>
 #include <linux/slab.h>
+#include <linux/regmap.h>
 #include <linux/reset-controller.h>
 #include <dt-bindings/clock/microchip,mpfs-clock.h>
 #include <soc/microchip/mpfs.h>
@@ -27,11 +29,14 @@
 #define MPFS_SLEEP_MIN_US	100
 #define MPFS_SLEEP_MAX_US	200
 
+#define REG_SUBBLK_RESET_CR	0x88u
+
 /* block concurrent access to the soft reset register */
 static DEFINE_SPINLOCK(mpfs_reset_lock);
 
 struct mpfs_reset {
 	void __iomem *base;
+	struct regmap *regmap;
 	struct reset_controller_dev rcdev;
 };
 
@@ -51,9 +56,17 @@ static int mpfs_assert(struct reset_controller_dev *rcdev, unsigned long id)
 
 	spin_lock_irqsave(&mpfs_reset_lock, flags);
 
-	reg = readl(rst->base);
+	if (rst->regmap)
+		regmap_read(rst->regmap, REG_SUBBLK_RESET_CR, &reg);
+	else
+		reg = readl(rst->base);
+
 	reg |= BIT(id);
-	writel(reg, rst->base);
+
+	if (rst->regmap)
+		regmap_write(rst->regmap, REG_SUBBLK_RESET_CR, reg);
+	else
+		writel(reg, rst->base);
 
 	spin_unlock_irqrestore(&mpfs_reset_lock, flags);
 
@@ -68,9 +81,17 @@ static int mpfs_deassert(struct reset_controller_dev *rcdev, unsigned long id)
 
 	spin_lock_irqsave(&mpfs_reset_lock, flags);
 
-	reg = readl(rst->base);
+	if (rst->regmap)
+		regmap_read(rst->regmap, REG_SUBBLK_RESET_CR, &reg);
+	else
+		reg = readl(rst->base);
+
 	reg &= ~BIT(id);
-	writel(reg, rst->base);
+
+	if (rst->regmap)
+		regmap_write(rst->regmap, REG_SUBBLK_RESET_CR, reg);
+	else
+		writel(reg, rst->base);
 
 	spin_unlock_irqrestore(&mpfs_reset_lock, flags);
 
@@ -130,11 +151,54 @@ static int mpfs_reset_xlate(struct reset_controller_dev *rcdev,
 	return index - MPFS_PERIPH_OFFSET;
 }
 
-static int mpfs_reset_probe(struct auxiliary_device *adev,
-			    const struct auxiliary_device_id *id)
+static int mpfs_reset_of_probe(struct platform_device *pdev)
 {
-	struct device *dev = &adev->dev;
 	struct reset_controller_dev *rcdev;
+	struct device *dev = &pdev->dev;
+	struct mpfs_reset *rst;
+
+	rst = devm_kzalloc(dev, sizeof(*rst), GFP_KERNEL);
+	if (!rst)
+		return -ENOMEM;
+
+	rst->regmap = syscon_node_to_regmap(dev->of_node->parent);
+	if (IS_ERR(rst->regmap))
+		dev_err_probe(dev, PTR_ERR(rst->regmap), "Failed to find syscon regmap\n");
+
+	rcdev = &rst->rcdev;
+	rcdev->dev = dev;
+	rcdev->ops = &mpfs_reset_ops;
+
+	rcdev->of_node = dev->of_node;
+	rcdev->of_reset_n_cells = 1;
+	rcdev->of_xlate = mpfs_reset_xlate;
+	rcdev->nr_resets = MPFS_NUM_RESETS;
+
+	printk("of probe\n");
+
+	return devm_reset_controller_register(dev, rcdev);
+}
+
+static const struct of_device_id mpfs_reset_of_match[] = {
+	{ .compatible = "microchip,mpfs-reset", },
+	{},
+};
+MODULE_DEVICE_TABLE(of, mpfs_reset_of_match);
+
+static struct platform_driver mpfs_reset_of_driver = {
+	.probe		= mpfs_reset_of_probe,
+	.driver = {
+		.name = "mpfs-reset",
+		.of_match_table = mpfs_reset_of_match,
+	},
+};
+module_platform_driver(mpfs_reset_of_driver);
+
+static int mpfs_reset_adev_probe(struct auxiliary_device *adev,
+				  const struct auxiliary_device_id *id)
+{
+	struct reset_controller_dev *rcdev;
+	struct device *dev = &adev->dev;
 	struct mpfs_reset *rst;
 
 	rst = devm_kzalloc(dev, sizeof(*rst), GFP_KERNEL);
@@ -145,8 +209,8 @@ static int mpfs_reset_probe(struct auxiliary_device *adev,
 
 	rcdev = &rst->rcdev;
 	rcdev->dev = dev;
-	rcdev->dev->parent = dev->parent;
 	rcdev->ops = &mpfs_reset_ops;
+
 	rcdev->of_node = dev->parent->of_node;
 	rcdev->of_reset_n_cells = 1;
 	rcdev->of_xlate = mpfs_reset_xlate;
@@ -222,12 +286,12 @@ static const struct auxiliary_device_id mpfs_reset_ids[] = {
 };
 MODULE_DEVICE_TABLE(auxiliary, mpfs_reset_ids);
 
-static struct auxiliary_driver mpfs_reset_driver = {
-	.probe		= mpfs_reset_probe,
+static struct auxiliary_driver mpfs_reset_aux_driver = {
+	.probe		= mpfs_reset_adev_probe,
 	.id_table	= mpfs_reset_ids,
 };
 
-module_auxiliary_driver(mpfs_reset_driver);
+module_auxiliary_driver(mpfs_reset_aux_driver);
 
 MODULE_DESCRIPTION("Microchip PolarFire SoC Reset Driver");
 MODULE_AUTHOR("Conor Dooley <conor.dooley@microchip.com>");
-- 
2.43.0

[RFC PATCH 08/11] copy meson clk-regmap for now

[Conor Dooley]

From: Conor Dooley <conor.dooley@microchip.com>

Signed-off-by: Conor Dooley <conor.dooley@microchip.com>
---
 drivers/clk/microchip/Makefile     |   1 +
 drivers/clk/microchip/clk-regmap.c | 186 +++++++++++++++++++++++++++++
 drivers/clk/microchip/clk-regmap.h | 137 +++++++++++++++++++++
 3 files changed, 324 insertions(+)
 create mode 100644 drivers/clk/microchip/clk-regmap.c
 create mode 100644 drivers/clk/microchip/clk-regmap.h

diff --git a/drivers/clk/microchip/Makefile b/drivers/clk/microchip/Makefile
index 13250e04e46c..6b463066c64e 100644
--- a/drivers/clk/microchip/Makefile
+++ b/drivers/clk/microchip/Makefile
@@ -3,3 +3,4 @@ obj-$(CONFIG_COMMON_CLK_PIC32) += clk-core.o
 obj-$(CONFIG_PIC32MZDA) += clk-pic32mzda.o
 obj-$(CONFIG_MCHP_CLK_MPFS) += clk-mpfs.o
 obj-$(CONFIG_MCHP_CLK_MPFS) += clk-mpfs-ccc.o
+obj-y += clk-regmap.o
diff --git a/drivers/clk/microchip/clk-regmap.c b/drivers/clk/microchip/clk-regmap.c
new file mode 100644
index 000000000000..ad116d24f700
--- /dev/null
+++ b/drivers/clk/microchip/clk-regmap.c
@@ -0,0 +1,186 @@
+// SPDX-License-Identifier: GPL-2.0
+/*
+ * Copyright (c) 2018 BayLibre, SAS.
+ * Author: Jerome Brunet <jbrunet@baylibre.com>
+ */
+
+#include <linux/module.h>
+#include "clk-regmap.h"
+
+static int clk_regmap_gate_endisable(struct clk_hw *hw, int enable)
+{
+	struct clk_regmap *clk = to_clk_regmap(hw);
+	struct clk_regmap_gate_data *gate = clk_get_regmap_gate_data(clk);
+	int set = gate->flags & CLK_GATE_SET_TO_DISABLE ? 1 : 0;
+
+	set ^= enable;
+
+	return regmap_update_bits(clk->map, gate->offset, BIT(gate->bit_idx),
+				  set ? BIT(gate->bit_idx) : 0);
+}
+
+static int clk_regmap_gate_enable(struct clk_hw *hw)
+{
+	return clk_regmap_gate_endisable(hw, 1);
+}
+
+static void clk_regmap_gate_disable(struct clk_hw *hw)
+{
+	clk_regmap_gate_endisable(hw, 0);
+}
+
+static int clk_regmap_gate_is_enabled(struct clk_hw *hw)
+{
+	struct clk_regmap *clk = to_clk_regmap(hw);
+	struct clk_regmap_gate_data *gate = clk_get_regmap_gate_data(clk);
+	unsigned int val;
+
+	regmap_read(clk->map, gate->offset, &val);
+	if (gate->flags & CLK_GATE_SET_TO_DISABLE)
+		val ^= BIT(gate->bit_idx);
+
+	val &= BIT(gate->bit_idx);
+
+	return val ? 1 : 0;
+}
+
+const struct clk_ops clk_regmap_gate_ops = {
+	.enable = clk_regmap_gate_enable,
+	.disable = clk_regmap_gate_disable,
+	.is_enabled = clk_regmap_gate_is_enabled,
+};
+EXPORT_SYMBOL_GPL(clk_regmap_gate_ops);
+
+const struct clk_ops clk_regmap_gate_ro_ops = {
+	.is_enabled = clk_regmap_gate_is_enabled,
+};
+EXPORT_SYMBOL_GPL(clk_regmap_gate_ro_ops);
+
+static unsigned long clk_regmap_div_recalc_rate(struct clk_hw *hw,
+						unsigned long prate)
+{
+	struct clk_regmap *clk = to_clk_regmap(hw);
+	struct clk_regmap_div_data *div = clk_get_regmap_div_data(clk);
+	unsigned int val;
+	int ret;
+
+	ret = regmap_read(clk->map, div->offset, &val);
+	if (ret)
+		/* Gives a hint that something is wrong */
+		return 0;
+
+	val >>= div->shift;
+	val &= clk_div_mask(div->width);
+	return divider_recalc_rate(hw, prate, val, div->table, div->flags,
+				   div->width);
+}
+
+static int clk_regmap_div_determine_rate(struct clk_hw *hw,
+					 struct clk_rate_request *req)
+{
+	struct clk_regmap *clk = to_clk_regmap(hw);
+	struct clk_regmap_div_data *div = clk_get_regmap_div_data(clk);
+	unsigned int val;
+	int ret;
+
+	/* if read only, just return current value */
+	if (div->flags & CLK_DIVIDER_READ_ONLY) {
+		ret = regmap_read(clk->map, div->offset, &val);
+		if (ret)
+			return ret;
+
+		val >>= div->shift;
+		val &= clk_div_mask(div->width);
+
+		return divider_ro_determine_rate(hw, req, div->table,
+						 div->width, div->flags, val);
+	}
+
+	return divider_determine_rate(hw, req, div->table, div->width,
+				      div->flags);
+}
+
+static int clk_regmap_div_set_rate(struct clk_hw *hw, unsigned long rate,
+				   unsigned long parent_rate)
+{
+	struct clk_regmap *clk = to_clk_regmap(hw);
+	struct clk_regmap_div_data *div = clk_get_regmap_div_data(clk);
+	unsigned int val;
+	int ret;
+
+	ret = divider_get_val(rate, parent_rate, div->table, div->width,
+			      div->flags);
+	if (ret < 0)
+		return ret;
+
+	val = (unsigned int)ret << div->shift;
+	return regmap_update_bits(clk->map, div->offset,
+				  clk_div_mask(div->width) << div->shift, val);
+};
+
+/* Would prefer clk_regmap_div_ro_ops but clashes with qcom */
+
+const struct clk_ops clk_regmap_divider_ops = {
+	.recalc_rate = clk_regmap_div_recalc_rate,
+	.determine_rate = clk_regmap_div_determine_rate,
+	.set_rate = clk_regmap_div_set_rate,
+};
+EXPORT_SYMBOL_GPL(clk_regmap_divider_ops);
+
+const struct clk_ops clk_regmap_divider_ro_ops = {
+	.recalc_rate = clk_regmap_div_recalc_rate,
+	.determine_rate = clk_regmap_div_determine_rate,
+};
+EXPORT_SYMBOL_GPL(clk_regmap_divider_ro_ops);
+
+static u8 clk_regmap_mux_get_parent(struct clk_hw *hw)
+{
+	struct clk_regmap *clk = to_clk_regmap(hw);
+	struct clk_regmap_mux_data *mux = clk_get_regmap_mux_data(clk);
+	unsigned int val;
+	int ret;
+
+	ret = regmap_read(clk->map, mux->offset, &val);
+	if (ret)
+		return ret;
+
+	val >>= mux->shift;
+	val &= mux->mask;
+	return clk_mux_val_to_index(hw, mux->table, mux->flags, val);
+}
+
+static int clk_regmap_mux_set_parent(struct clk_hw *hw, u8 index)
+{
+	struct clk_regmap *clk = to_clk_regmap(hw);
+	struct clk_regmap_mux_data *mux = clk_get_regmap_mux_data(clk);
+	unsigned int val = clk_mux_index_to_val(mux->table, mux->flags, index);
+
+	return regmap_update_bits(clk->map, mux->offset,
+				  mux->mask << mux->shift,
+				  val << mux->shift);
+}
+
+static int clk_regmap_mux_determine_rate(struct clk_hw *hw,
+					 struct clk_rate_request *req)
+{
+	struct clk_regmap *clk = to_clk_regmap(hw);
+	struct clk_regmap_mux_data *mux = clk_get_regmap_mux_data(clk);
+
+	return clk_mux_determine_rate_flags(hw, req, mux->flags);
+}
+
+const struct clk_ops clk_regmap_mux_ops = {
+	.get_parent = clk_regmap_mux_get_parent,
+	.set_parent = clk_regmap_mux_set_parent,
+	.determine_rate = clk_regmap_mux_determine_rate,
+};
+EXPORT_SYMBOL_GPL(clk_regmap_mux_ops);
+
+const struct clk_ops clk_regmap_mux_ro_ops = {
+	.get_parent = clk_regmap_mux_get_parent,
+};
+EXPORT_SYMBOL_GPL(clk_regmap_mux_ro_ops);
+
+MODULE_DESCRIPTION("Amlogic regmap backed clock driver");
+MODULE_AUTHOR("Jerome Brunet <jbrunet@baylibre.com>");
+MODULE_LICENSE("GPL");
diff --git a/drivers/clk/microchip/clk-regmap.h b/drivers/clk/microchip/clk-regmap.h
new file mode 100644
index 000000000000..e365312da54e
--- /dev/null
+++ b/drivers/clk/microchip/clk-regmap.h
@@ -0,0 +1,137 @@
+/* SPDX-License-Identifier: GPL-2.0 */
+/*
+ * Copyright (c) 2018 BayLibre, SAS.
+ * Author: Jerome Brunet <jbrunet@baylibre.com>
+ */
+
+#ifndef __CLK_REGMAP_H
+#define __CLK_REGMAP_H
+
+#include <linux/clk-provider.h>
+#include <linux/regmap.h>
+
+/**
+ * struct clk_regmap - regmap backed clock
+ *
+ * @hw:		handle between common and hardware-specific interfaces
+ * @map:	pointer to the regmap structure controlling the clock
+ * @data:	data specific to the clock type
+ *
+ * Clock which is controlled by regmap backed registers. The actual type of
+ * of the clock is controlled by the clock_ops and data.
+ */
+struct clk_regmap {
+	struct clk_hw	hw;
+	struct regmap	*map;
+	void		*data;
+};
+
+static inline struct clk_regmap *to_clk_regmap(struct clk_hw *hw)
+{
+	return container_of(hw, struct clk_regmap, hw);
+}
+
+/**
+ * struct clk_regmap_gate_data - regmap backed gate specific data
+ *
+ * @offset:	offset of the register controlling gate
+ * @bit_idx:	single bit controlling gate
+ * @flags:	hardware-specific flags
+ *
+ * Flags:
+ * Same as clk_gate except CLK_GATE_HIWORD_MASK which is ignored
+ */
+struct clk_regmap_gate_data {
+	unsigned int	offset;
+	u8		bit_idx;
+	u8		flags;
+};
+
+static inline struct clk_regmap_gate_data *
+clk_get_regmap_gate_data(struct clk_regmap *clk)
+{
+	return (struct clk_regmap_gate_data *)clk->data;
+}
+
+extern const struct clk_ops clk_regmap_gate_ops;
+extern const struct clk_ops clk_regmap_gate_ro_ops;
+
+/**
+ * struct clk_regmap_div_data - regmap backed adjustable divider specific data
+ *
+ * @offset:	offset of the register controlling the divider
+ * @shift:	shift to the divider bit field
+ * @width:	width of the divider bit field
+ * @table:	array of value/divider pairs, last entry should have div = 0
+ *
+ * Flags:
+ * Same as clk_divider except CLK_DIVIDER_HIWORD_MASK which is ignored
+ */
+struct clk_regmap_div_data {
+	unsigned int	offset;
+	u8		shift;
+	u8		width;
+	u8		flags;
+	const struct clk_div_table	*table;
+};
+
+static inline struct clk_regmap_div_data *
+clk_get_regmap_div_data(struct clk_regmap *clk)
+{
+	return (struct clk_regmap_div_data *)clk->data;
+}
+
+extern const struct clk_ops clk_regmap_divider_ops;
+extern const struct clk_ops clk_regmap_divider_ro_ops;
+
+/**
+ * struct clk_regmap_mux_data - regmap backed multiplexer clock specific data
+ *
+ * @hw:		handle between common and hardware-specific interfaces
+ * @offset:	offset of theregister controlling multiplexer
+ * @table:	array of parent indexed register values
+ * @shift:	shift to multiplexer bit field
+ * @mask:	mask of mutliplexer bit field
+ * @flags:	hardware-specific flags
+ *
+ * Flags:
+ * Same as clk_divider except CLK_MUX_HIWORD_MASK which is ignored
+ */
+struct clk_regmap_mux_data {
+	unsigned int	offset;
+	u32		*table;
+	u32		mask;
+	u8		shift;
+	u8		flags;
+};
+
+static inline struct clk_regmap_mux_data *
+clk_get_regmap_mux_data(struct clk_regmap *clk)
+{
+	return (struct clk_regmap_mux_data *)clk->data;
+}
+
+extern const struct clk_ops clk_regmap_mux_ops;
+extern const struct clk_ops clk_regmap_mux_ro_ops;
+
+#define __MESON_PCLK(_name, _reg, _bit, _ops, _pname)			\
+struct clk_regmap _name = {						\
+	.data = &(struct clk_regmap_gate_data){				\
+		.offset = (_reg),					\
+		.bit_idx = (_bit),					\
+	},								\
+	.hw.init = &(struct clk_init_data) {				\
+		.name = #_name,						\
+		.ops = _ops,						\
+		.parent_hws = (const struct clk_hw *[]) { _pname },	\
+		.num_parents = 1,					\
+		.flags = (CLK_SET_RATE_PARENT | CLK_IGNORE_UNUSED),	\
+	},								\
+}
+
+#define MESON_PCLK(_name, _reg, _bit, _pname)	\
+	__MESON_PCLK(_name, _reg, _bit, &clk_regmap_gate_ops, _pname)
+
+#define MESON_PCLK_RO(_name, _reg, _bit, _pname)	\
+	__MESON_PCLK(_name, _reg, _bit, &clk_regmap_gate_ro_ops, _pname)
+#endif /* __CLK_REGMAP_H */
-- 
2.43.0

[RFC PATCH 09/11] clk: microchip: mpfs: use regmap clock types

[Conor Dooley]

From: Conor Dooley <conor.dooley@microchip.com>

Convert the PolarFire SoC clock driver to use regmap clock types as a
preparatory work for supporting the new binding for this device that
will only provide the second of the two register regions, and will
require the use of syscon regmap to access the "cfg" and "periph" clocks
currently supported by the driver.

Signed-off-by: Conor Dooley <conor.dooley@microchip.com>
---
 drivers/clk/microchip/clk-mpfs.c | 81 ++++++++++++++++++++++----------
 1 file changed, 56 insertions(+), 25 deletions(-)

diff --git a/drivers/clk/microchip/clk-mpfs.c b/drivers/clk/microchip/clk-mpfs.c
index 28ec0da88cb3..e288c1729a23 100644
--- a/drivers/clk/microchip/clk-mpfs.c
+++ b/drivers/clk/microchip/clk-mpfs.c
@@ -6,10 +6,13 @@
  */
 #include <linux/clk-provider.h>
 #include <linux/io.h>
+#include <linux/mfd/syscon.h>
 #include <linux/module.h>
 #include <linux/platform_device.h>
+#include <linux/regmap.h>
 #include <dt-bindings/clock/microchip,mpfs-clock.h>
 #include <soc/microchip/mpfs.h>
+#include "clk-regmap.h"
 
 /* address offset of control registers */
 #define REG_MSSPLL_REF_CR	0x08u
@@ -30,6 +33,14 @@
 #define MSSPLL_POSTDIV_WIDTH	0x07u
 #define MSSPLL_FIXED_DIV	4u
 
+static const struct regmap_config clk_mpfs_regmap_config = {
+	.reg_bits = 32,
+	.reg_stride = 4,
+	.val_bits = 32,
+	.val_format_endian = REGMAP_ENDIAN_LITTLE,
+	.max_register = REG_SUBBLK_CLOCK_CR,
+};
+
 /*
  * This clock ID is defined here, rather than the binding headers, as it is an
  * internal clock only, and therefore has no consumers in other peripheral
@@ -39,6 +50,7 @@
 
 struct mpfs_clock_data {
 	struct device *dev;
+	struct regmap *regmap;
 	void __iomem *base;
 	void __iomem *msspll_base;
 	struct clk_hw_onecell_data hw_data;
@@ -68,14 +80,14 @@ struct mpfs_msspll_out_hw_clock {
 #define to_mpfs_msspll_out_clk(_hw) container_of(_hw, struct mpfs_msspll_out_hw_clock, hw)
 
 struct mpfs_cfg_hw_clock {
-	struct clk_divider cfg;
-	struct clk_init_data init;
+	struct clk_regmap sigh;
+	struct clk_regmap_div_data cfg;
 	unsigned int id;
-	u32 reg_offset;
 };
 
 struct mpfs_periph_hw_clock {
-	struct clk_gate periph;
+	struct clk_regmap sigh;
+	struct clk_regmap_gate_data periph;
 	unsigned int id;
 };
 
@@ -225,10 +237,9 @@ static int mpfs_clk_register_msspll_outs(struct device *dev,
 	.cfg.shift = _shift,								\
 	.cfg.width = _width,								\
 	.cfg.table = _table,								\
-	.reg_offset = _offset,								\
+	.cfg.offset = _offset,								\
 	.cfg.flags = _flags,								\
-	.cfg.hw.init = CLK_HW_INIT(_name, _parent, &clk_divider_ops, 0),		\
-	.cfg.lock = &mpfs_clk_lock,							\
+	.sigh.hw.init = CLK_HW_INIT(_name, _parent, &clk_regmap_divider_ops, 0),	\
 }
 
 #define CLK_CPU_OFFSET		0u
@@ -248,10 +259,10 @@ static struct mpfs_cfg_hw_clock mpfs_cfg_clks[] = {
 		.cfg.shift = 0,
 		.cfg.width = 12,
 		.cfg.table = mpfs_div_rtcref_table,
-		.reg_offset = REG_RTC_CLOCK_CR,
+		.cfg.offset = REG_RTC_CLOCK_CR,
 		.cfg.flags = CLK_DIVIDER_ONE_BASED,
-		.cfg.hw.init =
-			CLK_HW_INIT_PARENTS_DATA("clk_rtcref", mpfs_ext_ref, &clk_divider_ops, 0),
+		.sigh.hw.init =
+			CLK_HW_INIT_PARENTS_DATA("clk_rtcref", mpfs_ext_ref, &clk_regmap_divider_ops, 0),
 	}
 };
 
@@ -264,14 +275,16 @@ static int mpfs_clk_register_cfgs(struct device *dev, struct mpfs_cfg_hw_clock *
 	for (i = 0; i < num_clks; i++) {
 		struct mpfs_cfg_hw_clock *cfg_hw = &cfg_hws[i];
 
-		cfg_hw->cfg.reg = data->base + cfg_hw->reg_offset;
-		ret = devm_clk_hw_register(dev, &cfg_hw->cfg.hw);
+		cfg_hw->sigh.map = data->regmap;
+		cfg_hw->sigh.data = &cfg_hw->cfg;
+
+		ret = devm_clk_hw_register(dev, &cfg_hw->sigh.hw);
 		if (ret)
 			return dev_err_probe(dev, ret, "failed to register clock id: %d\n",
 					     cfg_hw->id);
 
 		id = cfg_hw->id;
-		data->hw_data.hws[id] = &cfg_hw->cfg.hw;
+		data->hw_data.hws[id] = &cfg_hw->sigh.hw;
 	}
 
 	return 0;
@@ -283,13 +296,13 @@ static int mpfs_clk_register_cfgs(struct device *dev, struct mpfs_cfg_hw_clock *
 
 #define CLK_PERIPH(_id, _name, _parent, _shift, _flags) {			\
 	.id = _id,								\
+	.periph.offset = REG_SUBBLK_CLOCK_CR,					\
 	.periph.bit_idx = _shift,						\
-	.periph.hw.init = CLK_HW_INIT_HW(_name, _parent, &clk_gate_ops,		\
-				  _flags),					\
-	.periph.lock = &mpfs_clk_lock,						\
+	.sigh.hw.init = CLK_HW_INIT_HW(_name, _parent, &clk_regmap_gate_ops,	\
+					 _flags),				\
 }
 
-#define PARENT_CLK(PARENT) (&mpfs_cfg_clks[CLK_##PARENT##_OFFSET].cfg.hw)
+#define PARENT_CLK(PARENT) (&mpfs_cfg_clks[CLK_##PARENT##_OFFSET].sigh.hw)
 
 /*
  * Critical clocks:
@@ -346,14 +359,15 @@ static int mpfs_clk_register_periphs(struct device *dev, struct mpfs_periph_hw_c
 	for (i = 0; i < num_clks; i++) {
 		struct mpfs_periph_hw_clock *periph_hw = &periph_hws[i];
 
-		periph_hw->periph.reg = data->base + REG_SUBBLK_CLOCK_CR;
-		ret = devm_clk_hw_register(dev, &periph_hw->periph.hw);
+		periph_hw->sigh.map = data->regmap;
+		periph_hw->sigh.data = &periph_hw->periph;
+		ret = devm_clk_hw_register(dev, &periph_hw->sigh.hw);
 		if (ret)
 			return dev_err_probe(dev, ret, "failed to register clock id: %d\n",
 					     periph_hw->id);
 
 		id = periph_hws[i].id;
-		data->hw_data.hws[id] = &periph_hw->periph.hw;
+		data->hw_data.hws[id] = &periph_hw->sigh.hw;
 	}
 
 	return 0;
@@ -374,6 +388,19 @@ static int mpfs_clk_probe(struct platform_device *pdev)
 	if (!clk_data)
 		return -ENOMEM;
 
+	clk_data->regmap = syscon_regmap_lookup_by_compatible("microchip,mpfs-mss-top-sysreg");
+	if (IS_ERR(clk_data->regmap)) {
+		clk_data->regmap = NULL;
+		goto old_format;
+	}
+
+	clk_data->msspll_base = devm_platform_ioremap_resource(pdev, 0);
+	if (IS_ERR(clk_data->msspll_base))
+		return PTR_ERR(clk_data->msspll_base);
+
+	goto done;
+
+old_format:
 	clk_data->base = devm_platform_ioremap_resource(pdev, 0);
 	if (IS_ERR(clk_data->base))
 		return PTR_ERR(clk_data->base);
@@ -382,6 +409,14 @@ static int mpfs_clk_probe(struct platform_device *pdev)
 	if (IS_ERR(clk_data->msspll_base))
 		return PTR_ERR(clk_data->msspll_base);
 
+	clk_data->regmap = devm_regmap_init_mmio(dev, clk_data->base, &clk_mpfs_regmap_config);
+	if (IS_ERR(clk_data->regmap))
+		return PTR_ERR(clk_data->regmap);
+
+	ret = mpfs_reset_controller_register(dev, clk_data->base + REG_SUBBLK_RESET_CR);
+	if (ret)
+		return ret;
+done:
 	clk_data->hw_data.num = num_clks;
 	clk_data->dev = dev;
 	dev_set_drvdata(dev, clk_data);
@@ -406,11 +441,7 @@ static int mpfs_clk_probe(struct platform_device *pdev)
 	if (ret)
 		return ret;
 
-	ret = devm_of_clk_add_hw_provider(dev, of_clk_hw_onecell_get, &clk_data->hw_data);
-	if (ret)
-		return ret;
-
-	return mpfs_reset_controller_register(dev, clk_data->base + REG_SUBBLK_RESET_CR);
+	return devm_of_clk_add_hw_provider(dev, of_clk_hw_onecell_get, &clk_data->hw_data);
 }
 
 static const struct of_device_id mpfs_clk_of_match_table[] = {
-- 
2.43.0

[RFC PATCH 10/11] dt-bindings: clk: microchip: mpfs: remove first reg region

[Conor Dooley]

From: Conor Dooley <conor.dooley@microchip.com>

The first reg region in this binding is not exclusively for clocks, as
evidenced by the dual role of this device as a reset controller at
present. The first region is however better described by a simple-mfd
syscon, but this would have require a significant re-write of the
devicetree for the platform, so the easy way out was chosen when reset
support was first introduced. The region doesn't just contain clock and
reset registers, it also contains pinctrl and interrupt controller
functionality, so drop the region from the clock binding so that it can
be described instead by a simple-mfd syscon rather than propagate this
incorrect description of the hardware to the new pic64gx SoC.

Signed-off-by: Conor Dooley <conor.dooley@microchip.com>
---
 .../bindings/clock/microchip,mpfs-clkcfg.yaml | 33 +++++++++++--------
 1 file changed, 19 insertions(+), 14 deletions(-)

diff --git a/Documentation/devicetree/bindings/clock/microchip,mpfs-clkcfg.yaml b/Documentation/devicetree/bindings/clock/microchip,mpfs-clkcfg.yaml
index e4e1c31267d2..df861eb73e86 100644
--- a/Documentation/devicetree/bindings/clock/microchip,mpfs-clkcfg.yaml
+++ b/Documentation/devicetree/bindings/clock/microchip,mpfs-clkcfg.yaml
@@ -22,16 +22,23 @@ properties:
     const: microchip,mpfs-clkcfg
 
   reg:
-    items:
-      - description: |
-          clock config registers:
-          These registers contain enable, reset & divider tables for the, cpu,
-          axi, ahb and rtc/mtimer reference clocks as well as enable and reset
-          for the peripheral clocks.
-      - description: |
-          mss pll dri registers:
-          Block of registers responsible for dynamic reconfiguration of the mss
-          pll
+    oneOf:
+      - items:
+          - description: |
+              clock config registers:
+              These registers contain enable, reset & divider tables for the, cpu,
+              axi, ahb and rtc/mtimer reference clocks as well as enable and reset
+              for the peripheral clocks.
+          - description: |
+              mss pll dri registers:
+              Block of registers responsible for dynamic reconfiguration of the mss
+              pll
+      - items:
+          - description: |
+              mss pll dri registers:
+              Block of registers responsible for dynamic reconfiguration of the mss
+              pll
+
 
   clocks:
     maxItems: 1
@@ -69,11 +76,9 @@ examples:
   - |
     #include <dt-bindings/clock/microchip,mpfs-clock.h>
     soc {
-            #address-cells = <2>;
-            #size-cells = <2>;
-            clkcfg: clock-controller@20002000 {
+            clkcfg: clock-controller@3E001000 {
                 compatible = "microchip,mpfs-clkcfg";
-                reg = <0x0 0x20002000 0x0 0x1000>, <0x0 0x3E001000 0x0 0x1000>;
+                reg = <0x3E001000 0x1000>;
                 clocks = <&ref>;
                 #clock-cells = <1>;
         };
-- 
2.43.0

[RFC PATCH 11/11] riscv: dts: microchip: convert clock and reset (TODO: fixup phandle)

[Conor Dooley]

From: Conor Dooley <conor.dooley@microchip.com>

Signed-off-by: Conor Dooley <conor.dooley@microchip.com>
---
 arch/riscv/boot/dts/microchip/mpfs.dtsi | 22 +++++++++++++---------
 1 file changed, 13 insertions(+), 9 deletions(-)

diff --git a/arch/riscv/boot/dts/microchip/mpfs.dtsi b/arch/riscv/boot/dts/microchip/mpfs.dtsi
index 1d655126b66f..6c5b6fef313f 100644
--- a/arch/riscv/boot/dts/microchip/mpfs.dtsi
+++ b/arch/riscv/boot/dts/microchip/mpfs.dtsi
@@ -254,14 +254,11 @@ pdma: dma-controller@3000000 {
 		mss_top_scb: syscon@20002000 {
 			compatible = "microchip,mpfs-mss-top-sysreg", "syscon", "simple-mfd";
 			reg = <0x0 0x20002000 0x0 0x1000>;
-		};
 
-		clkcfg: clkcfg@20002000 {
-			compatible = "microchip,mpfs-clkcfg";
-			reg = <0x0 0x20002000 0x0 0x1000>, <0x0 0x3E001000 0x0 0x1000>;
-			clocks = <&refclk>;
-			#clock-cells = <1>;
-			#reset-cells = <1>;
+			rst: reset-controller {
+				compatible = "microchip,mpfs-reset";
+				#reset-cells = <1>;
+			};
 		};
 
 		sysreg_scb: syscon@20003000 {
@@ -457,7 +454,7 @@ mac0: ethernet@20110000 {
 			local-mac-address = [00 00 00 00 00 00];
 			clocks = <&clkcfg CLK_MAC0>, <&clkcfg CLK_AHB>;
 			clock-names = "pclk", "hclk";
-			resets = <&clkcfg CLK_MAC0>;
+			resets = <&rst CLK_MAC0>;
 			status = "disabled";
 		};
 
@@ -471,7 +468,7 @@ mac1: ethernet@20112000 {
 			local-mac-address = [00 00 00 00 00 00];
 			clocks = <&clkcfg CLK_MAC1>, <&clkcfg CLK_AHB>;
 			clock-names = "pclk", "hclk";
-			resets = <&clkcfg CLK_MAC1>;
+			resets = <&rst CLK_MAC1>;
 			status = "disabled";
 		};
 
@@ -559,5 +556,12 @@ syscontroller_qspi: spi@37020100 {
 			clocks = <&scbclk>;
 			status = "disabled";
 		};
+
+		clkcfg: clkcfg@3e001000 {
+			compatible = "microchip,mpfs-clkcfg";
+			reg = <0x0 0x3e001000 0x0 0x1000>;
+			clocks = <&refclk>;
+			#clock-cells = <1>;
+		};
 	};
 };
-- 
2.43.0

Re: [RFC PATCH 01/11] dt-bindings: mailbox: mpfs: fix reg properties

[Rob Herring (Arm)]

On Thu, 15 Aug 2024 15:01:03 +0100, Conor Dooley wrote:
> From: Conor Dooley <conor.dooley@microchip.com>
> 
> When the binding for this was originally written, and later modified,
> mistakes were made - and the precise nature of the later modification
> should have been a giveaway, but alas I was naive at the time.
> 
> A more correct modelling of the hardware is to use two syscons and have
> a single reg entry for the mailbox, containing the mailbox region. The
> two syscons contain the general control/status registers for the mailbox
> and the interrupt related registers respectively. The reason for two
> syscons is that the same mailbox is present on the non-SoC version of
> the FPGA, which has no interrupt controller, and the shared part of the
> rtl was unchanged between devices.
> 
> This is now coming to a head, because the control/status registers share
> a register region with the "tvs" (temperature & voltage sensors)
> registers and, as it turns out, people do want to monitor temperatures
> and voltages...
> 
> Signed-off-by: Conor Dooley <conor.dooley@microchip.com>
> ---
>  .../bindings/mailbox/microchip,mpfs-mailbox.yaml       | 10 +++++-----
>  1 file changed, 5 insertions(+), 5 deletions(-)
> 

My bot found errors running 'make dt_binding_check' on your patch:

yamllint warnings/errors:

dtschema/dtc warnings/errors:
Documentation/devicetree/bindings/mailbox/microchip,mpfs-mailbox.example.dts:27.13-38: Warning (reg_format): /example-0/soc/mailbox@37020800:reg: property has invalid length (8 bytes) (#address-cells == 2, #size-cells == 1)
Documentation/devicetree/bindings/mailbox/microchip,mpfs-mailbox.example.dtb: Warning (pci_device_reg): Failed prerequisite 'reg_format'
Documentation/devicetree/bindings/mailbox/microchip,mpfs-mailbox.example.dtb: Warning (pci_device_bus_num): Failed prerequisite 'reg_format'
Documentation/devicetree/bindings/mailbox/microchip,mpfs-mailbox.example.dtb: Warning (simple_bus_reg): Failed prerequisite 'reg_format'
Documentation/devicetree/bindings/mailbox/microchip,mpfs-mailbox.example.dtb: Warning (i2c_bus_reg): Failed prerequisite 'reg_format'
Documentation/devicetree/bindings/mailbox/microchip,mpfs-mailbox.example.dtb: Warning (spi_bus_reg): Failed prerequisite 'reg_format'
Documentation/devicetree/bindings/mailbox/microchip,mpfs-mailbox.example.dts:25.28-31.13: Warning (avoid_default_addr_size): /example-0/soc/mailbox@37020800: Relying on default #address-cells value
Documentation/devicetree/bindings/mailbox/microchip,mpfs-mailbox.example.dts:25.28-31.13: Warning (avoid_default_addr_size): /example-0/soc/mailbox@37020800: Relying on default #size-cells value
Documentation/devicetree/bindings/mailbox/microchip,mpfs-mailbox.example.dtb: Warning (unique_unit_address_if_enabled): Failed prerequisite 'avoid_default_addr_size'

doc reference errors (make refcheckdocs):

See https://patchwork.ozlabs.org/project/devicetree-bindings/patch/20240815-premiere-given-1dab82e67eba@spud

The base for the series is generally the latest rc1. A different dependency
should be noted in *this* patch.

If you already ran 'make dt_binding_check' and didn't see the above
error(s), then make sure 'yamllint' is installed and dt-schema is up to
date:

pip3 install dtschema --upgrade

Please check and re-submit after running the above command yourself. Note
that DT_SCHEMA_FILES can be set to your schema file to speed up checking
your schema. However, it must be unset to test all examples with your schema.

Re: [RFC PATCH 06/11] dt-bindings: soc: microchip: document the two simple-mfd syscons on PolarFire SoC

[Rob Herring (Arm)]

On Thu, 15 Aug 2024 15:01:09 +0100, Conor Dooley wrote:
> From: Conor Dooley <conor.dooley@microchip.com>
> 
> There are two syscons on PolarFire SoC that provide various functionality of
> use to the OS.
> 
> The first of these is the "control-scb" region, that contains the "tvs"
> temperature and voltage sensors and the control/status registers for the
> system controller's mailbox. The mailbox has a dedicated node, so
> there's no need for a child node describing it, looking the syscon up by
> compatible is sufficient.
> 
> The second, "mss-top-sysreg", contains clocks, pinctrl, resets, and
> interrupt controller and more. For this RFC, only the reset controller
> child is described as that's all that is described by the existing
> bindings. The clock controller already has a dedicated node, and will
> retain it as there are other clock regions, so like the mailbox,
> a compatible-based lookup of the syscon is sufficient to keep the clock
> driver working as before so no child is needed.
> 
> Signed-off-by: Conor Dooley <conor.dooley@microchip.com>
> ---
> (I'll split this in two later, it's just easier when I have the same
> questions about both...)
> 
> Are these things entitled to have child nodes for the reset and sensor
> nodes, or should the properties be in the parent and the OS probe the
> drivers for the functions? That's something that, despite supposedly
> being a maintainer, I do not understand the rules (of thumb?) for.
> 
> Secondly, is it okay to make the "pragmatic" decision to not have a
> child clock node and keep routing the clocks via the existing & retained
> clock node (and therefore not update the various clocks nodes in the
> consumers)? Doing so would require a lot more hocus pocus with the clock
> driver than this series does, as the same driver would no longer be
> suitable for the before/after bindings.
> ---
>  .../microchip/microchip,mpfs-control-scb.yaml | 54 +++++++++++++++++++
>  .../microchip,mpfs-mss-top-sysreg.yaml        | 53 ++++++++++++++++++
>  2 files changed, 107 insertions(+)
>  create mode 100644 Documentation/devicetree/bindings/soc/microchip/microchip,mpfs-control-scb.yaml
>  create mode 100644 Documentation/devicetree/bindings/soc/microchip/microchip,mpfs-mss-top-sysreg.yaml
> 

My bot found errors running 'make dt_binding_check' on your patch:

yamllint warnings/errors:

dtschema/dtc warnings/errors:
Documentation/devicetree/bindings/soc/microchip/microchip,mpfs-control-scb.example.dts:21.13-38: Warning (reg_format): /example-0/soc/syscon@37020000:reg: property has invalid length (8 bytes) (#address-cells == 2, #size-cells == 1)
Documentation/devicetree/bindings/soc/microchip/microchip,mpfs-control-scb.example.dtb: Warning (pci_device_reg): Failed prerequisite 'reg_format'
Documentation/devicetree/bindings/soc/microchip/microchip,mpfs-control-scb.example.dtb: Warning (pci_device_bus_num): Failed prerequisite 'reg_format'
Documentation/devicetree/bindings/soc/microchip/microchip,mpfs-control-scb.example.dtb: Warning (simple_bus_reg): Failed prerequisite 'reg_format'
Documentation/devicetree/bindings/soc/microchip/microchip,mpfs-control-scb.example.dtb: Warning (i2c_bus_reg): Failed prerequisite 'reg_format'
Documentation/devicetree/bindings/soc/microchip/microchip,mpfs-control-scb.example.dtb: Warning (spi_bus_reg): Failed prerequisite 'reg_format'
Documentation/devicetree/bindings/soc/microchip/microchip,mpfs-control-scb.example.dts:19.27-26.13: Warning (avoid_default_addr_size): /example-0/soc/syscon@37020000: Relying on default #address-cells value
Documentation/devicetree/bindings/soc/microchip/microchip,mpfs-control-scb.example.dts:19.27-26.13: Warning (avoid_default_addr_size): /example-0/soc/syscon@37020000: Relying on default #size-cells value
Documentation/devicetree/bindings/soc/microchip/microchip,mpfs-control-scb.example.dtb: Warning (unique_unit_address_if_enabled): Failed prerequisite 'avoid_default_addr_size'

doc reference errors (make refcheckdocs):

See https://patchwork.ozlabs.org/project/devicetree-bindings/patch/20240815-pending-sacrifice-f2569ed756fe@spud

The base for the series is generally the latest rc1. A different dependency
should be noted in *this* patch.

If you already ran 'make dt_binding_check' and didn't see the above
error(s), then make sure 'yamllint' is installed and dt-schema is up to
date:

pip3 install dtschema --upgrade

Please check and re-submit after running the above command yourself. Note
that DT_SCHEMA_FILES can be set to your schema file to speed up checking
your schema. However, it must be unset to test all examples with your schema.

Re: [RFC PATCH 10/11] dt-bindings: clk: microchip: mpfs: remove first reg region

[Rob Herring (Arm)]

On Thu, 15 Aug 2024 15:01:13 +0100, Conor Dooley wrote:
> From: Conor Dooley <conor.dooley@microchip.com>
> 
> The first reg region in this binding is not exclusively for clocks, as
> evidenced by the dual role of this device as a reset controller at
> present. The first region is however better described by a simple-mfd
> syscon, but this would have require a significant re-write of the
> devicetree for the platform, so the easy way out was chosen when reset
> support was first introduced. The region doesn't just contain clock and
> reset registers, it also contains pinctrl and interrupt controller
> functionality, so drop the region from the clock binding so that it can
> be described instead by a simple-mfd syscon rather than propagate this
> incorrect description of the hardware to the new pic64gx SoC.
> 
> Signed-off-by: Conor Dooley <conor.dooley@microchip.com>
> ---
>  .../bindings/clock/microchip,mpfs-clkcfg.yaml | 33 +++++++++++--------
>  1 file changed, 19 insertions(+), 14 deletions(-)
> 

My bot found errors running 'make dt_binding_check' on your patch:

yamllint warnings/errors:

dtschema/dtc warnings/errors:
Documentation/devicetree/bindings/clock/microchip,mpfs-clkcfg.example.dts:22.21-47: Warning (reg_format): /example-0/soc/clock-controller@3E001000:reg: property has invalid length (8 bytes) (#address-cells == 2, #size-cells == 1)
Documentation/devicetree/bindings/clock/microchip,mpfs-clkcfg.example.dtb: Warning (pci_device_reg): Failed prerequisite 'reg_format'
Documentation/devicetree/bindings/clock/microchip,mpfs-clkcfg.example.dtb: Warning (pci_device_bus_num): Failed prerequisite 'reg_format'
Documentation/devicetree/bindings/clock/microchip,mpfs-clkcfg.example.dtb: Warning (simple_bus_reg): Failed prerequisite 'reg_format'
Documentation/devicetree/bindings/clock/microchip,mpfs-clkcfg.example.dtb: Warning (i2c_bus_reg): Failed prerequisite 'reg_format'
Documentation/devicetree/bindings/clock/microchip,mpfs-clkcfg.example.dtb: Warning (spi_bus_reg): Failed prerequisite 'reg_format'
Documentation/devicetree/bindings/clock/microchip,mpfs-clkcfg.example.dts:20.51-25.15: Warning (avoid_default_addr_size): /example-0/soc/clock-controller@3E001000: Relying on default #address-cells value
Documentation/devicetree/bindings/clock/microchip,mpfs-clkcfg.example.dts:20.51-25.15: Warning (avoid_default_addr_size): /example-0/soc/clock-controller@3E001000: Relying on default #size-cells value
Documentation/devicetree/bindings/clock/microchip,mpfs-clkcfg.example.dtb: Warning (unique_unit_address_if_enabled): Failed prerequisite 'avoid_default_addr_size'

doc reference errors (make refcheckdocs):

See https://patchwork.ozlabs.org/project/devicetree-bindings/patch/20240815-fernlike-levitate-6004f5f46d66@spud

The base for the series is generally the latest rc1. A different dependency
should be noted in *this* patch.

If you already ran 'make dt_binding_check' and didn't see the above
error(s), then make sure 'yamllint' is installed and dt-schema is up to
date:

pip3 install dtschema --upgrade

Please check and re-submit after running the above command yourself. Note
that DT_SCHEMA_FILES can be set to your schema file to speed up checking
your schema. However, it must be unset to test all examples with your schema.

Re: [RFC PATCH 06/11] dt-bindings: soc: microchip: document the two simple-mfd syscons on PolarFire SoC

[Conor Dooley]

[-- Attachment #1: Type: text/plain, Size: 4408 bytes --]

On Thu, Aug 15, 2024 at 09:34:11AM -0600, Rob Herring (Arm) wrote:
> 
> On Thu, 15 Aug 2024 15:01:09 +0100, Conor Dooley wrote:
> > From: Conor Dooley <conor.dooley@microchip.com>
> > 
> > There are two syscons on PolarFire SoC that provide various functionality of
> > use to the OS.
> > 
> > The first of these is the "control-scb" region, that contains the "tvs"
> > temperature and voltage sensors and the control/status registers for the
> > system controller's mailbox. The mailbox has a dedicated node, so
> > there's no need for a child node describing it, looking the syscon up by
> > compatible is sufficient.
> > 
> > The second, "mss-top-sysreg", contains clocks, pinctrl, resets, and
> > interrupt controller and more. For this RFC, only the reset controller
> > child is described as that's all that is described by the existing
> > bindings. The clock controller already has a dedicated node, and will
> > retain it as there are other clock regions, so like the mailbox,
> > a compatible-based lookup of the syscon is sufficient to keep the clock
> > driver working as before so no child is needed.
> > 
> > Signed-off-by: Conor Dooley <conor.dooley@microchip.com>
> > ---
> > (I'll split this in two later, it's just easier when I have the same
> > questions about both...)
> > 
> > Are these things entitled to have child nodes for the reset and sensor
> > nodes, or should the properties be in the parent and the OS probe the
> > drivers for the functions? That's something that, despite supposedly
> > being a maintainer, I do not understand the rules (of thumb?) for.
> > 
> > Secondly, is it okay to make the "pragmatic" decision to not have a
> > child clock node and keep routing the clocks via the existing & retained
> > clock node (and therefore not update the various clocks nodes in the
> > consumers)? Doing so would require a lot more hocus pocus with the clock
> > driver than this series does, as the same driver would no longer be
> > suitable for the before/after bindings.
> > ---
> >  .../microchip/microchip,mpfs-control-scb.yaml | 54 +++++++++++++++++++
> >  .../microchip,mpfs-mss-top-sysreg.yaml        | 53 ++++++++++++++++++
> >  2 files changed, 107 insertions(+)
> >  create mode 100644 Documentation/devicetree/bindings/soc/microchip/microchip,mpfs-control-scb.yaml
> >  create mode 100644 Documentation/devicetree/bindings/soc/microchip/microchip,mpfs-mss-top-sysreg.yaml
> > 
> 
> My bot found errors running 'make dt_binding_check' on your patch:
> 
> yamllint warnings/errors:
> 
> dtschema/dtc warnings/errors:
> Documentation/devicetree/bindings/soc/microchip/microchip,mpfs-control-scb.example.dts:21.13-38: Warning (reg_format): /example-0/soc/syscon@37020000:reg: property has invalid length (8 bytes) (#address-cells == 2, #size-cells == 1)
> Documentation/devicetree/bindings/soc/microchip/microchip,mpfs-control-scb.example.dtb: Warning (pci_device_reg): Failed prerequisite 'reg_format'
> Documentation/devicetree/bindings/soc/microchip/microchip,mpfs-control-scb.example.dtb: Warning (pci_device_bus_num): Failed prerequisite 'reg_format'
> Documentation/devicetree/bindings/soc/microchip/microchip,mpfs-control-scb.example.dtb: Warning (simple_bus_reg): Failed prerequisite 'reg_format'
> Documentation/devicetree/bindings/soc/microchip/microchip,mpfs-control-scb.example.dtb: Warning (i2c_bus_reg): Failed prerequisite 'reg_format'
> Documentation/devicetree/bindings/soc/microchip/microchip,mpfs-control-scb.example.dtb: Warning (spi_bus_reg): Failed prerequisite 'reg_format'
> Documentation/devicetree/bindings/soc/microchip/microchip,mpfs-control-scb.example.dts:19.27-26.13: Warning (avoid_default_addr_size): /example-0/soc/syscon@37020000: Relying on default #address-cells value
> Documentation/devicetree/bindings/soc/microchip/microchip,mpfs-control-scb.example.dts:19.27-26.13: Warning (avoid_default_addr_size): /example-0/soc/syscon@37020000: Relying on default #size-cells value
> Documentation/devicetree/bindings/soc/microchip/microchip,mpfs-control-scb.example.dtb: Warning (unique_unit_address_if_enabled): Failed prerequisite 'avoid_default_addr_size'

Yeah, these are all known. One of the bindings doesn't even have an
example. I know this is automated, but just to point out that my only
objective here is figuring out whether or not child nodes are okay here.

[-- Attachment #2: signature.asc --]
[-- Type: application/pgp-signature, Size: 228 bytes --]

Re: [RFC PATCH 06/11] dt-bindings: soc: microchip: document the two simple-mfd syscons on PolarFire SoC

[Conor Dooley]

[-- Attachment #1: Type: text/plain, Size: 6415 bytes --]

On Thu, Aug 15, 2024 at 03:01:09PM +0100, Conor Dooley wrote:
> From: Conor Dooley <conor.dooley@microchip.com>
> 
> There are two syscons on PolarFire SoC that provide various functionality of
> use to the OS.
> 
> The first of these is the "control-scb" region, that contains the "tvs"
> temperature and voltage sensors and the control/status registers for the
> system controller's mailbox. The mailbox has a dedicated node, so
> there's no need for a child node describing it, looking the syscon up by
> compatible is sufficient.
> 
> The second, "mss-top-sysreg", contains clocks, pinctrl, resets, and
> interrupt controller and more. For this RFC, only the reset controller
> child is described as that's all that is described by the existing
> bindings. The clock controller already has a dedicated node, and will
> retain it as there are other clock regions, so like the mailbox,
> a compatible-based lookup of the syscon is sufficient to keep the clock
> driver working as before so no child is needed.
> 
> Signed-off-by: Conor Dooley <conor.dooley@microchip.com>
> ---
> (I'll split this in two later, it's just easier when I have the same
> questions about both...)
> 
> Are these things entitled to have child nodes for the reset and sensor
> nodes, or should the properties be in the parent and the OS probe the
> drivers for the functions? That's something that, despite supposedly
> being a maintainer, I do not understand the rules (of thumb?) for.

After posting a link to this on the devicetree irc channel, I had some
discussion with Mark Brown about whether or not the functions described
by the child nodes were ever likely to be independently reused. Given that
they're pretty closely tied to the integration of this particular SoC-FPGA
I think it is unlikely to happen. Reading between the lines, I'm going
to chalk that up as one vote against child nodes being suitable here.

Cheers,
Conor.

> 
> Secondly, is it okay to make the "pragmatic" decision to not have a
> child clock node and keep routing the clocks via the existing & retained
> clock node (and therefore not update the various clocks nodes in the
> consumers)? Doing so would require a lot more hocus pocus with the clock
> driver than this series does, as the same driver would no longer be
> suitable for the before/after bindings.
>
> ---
>  .../microchip/microchip,mpfs-control-scb.yaml | 54 +++++++++++++++++++
>  .../microchip,mpfs-mss-top-sysreg.yaml        | 53 ++++++++++++++++++
>  2 files changed, 107 insertions(+)
>  create mode 100644 Documentation/devicetree/bindings/soc/microchip/microchip,mpfs-control-scb.yaml
>  create mode 100644 Documentation/devicetree/bindings/soc/microchip/microchip,mpfs-mss-top-sysreg.yaml
> 
> diff --git a/Documentation/devicetree/bindings/soc/microchip/microchip,mpfs-control-scb.yaml b/Documentation/devicetree/bindings/soc/microchip/microchip,mpfs-control-scb.yaml
> new file mode 100644
> index 000000000000..3673bf139ce8
> --- /dev/null
> +++ b/Documentation/devicetree/bindings/soc/microchip/microchip,mpfs-control-scb.yaml
> @@ -0,0 +1,54 @@
> +# SPDX-License-Identifier: GPL-2.0-only OR BSD-2-Clause
> +%YAML 1.2
> +---
> +$id: http://devicetree.org/schemas/soc/microchip/microchip,mpfs-control-scb.yaml#
> +$schema: http://devicetree.org/meta-schemas/core.yaml#
> +
> +title: Microchip PolarFire SoC System Controller Bus (SCB) Control Register region
> +
> +maintainers:
> +  - Conor Dooley <conor.dooley@microchip.com>
> +
> +description:
> +  An assortment of system controller related registers, including voltage and
> +  temperature sensors and the status/control registers for the system
> +  controller's mailbox.
> +
> +properties:
> +  compatible:
> +    items:
> +      - const: microchip,mpfs-control-scb
> +      - const: syscon
> +      - const: simple-mfd
> +
> +  reg:
> +    maxItems: 1
> +
> +  sensor:
> +    type: object
> +
> +    properties:
> +      compatible:
> +        const: microchip,mpfs-tvs
> +
> +    additionalProperties: false
> +
> +required:
> +  - compatible
> +  - reg
> +
> +additionalProperties: false
> +
> +examples:
> +  - |
> +    soc {
> +      syscon@37020000 {
> +        compatible = "microchip,mpfs-control-scb", "syscon", "simple-mfd";
> +        reg = <0x37020000 0x100>;
> +
> +        sensor {
> +          compatible = "microchip,mpfs-tvs";
> +        };
> +      };
> +    };
> +
> diff --git a/Documentation/devicetree/bindings/soc/microchip/microchip,mpfs-mss-top-sysreg.yaml b/Documentation/devicetree/bindings/soc/microchip/microchip,mpfs-mss-top-sysreg.yaml
> new file mode 100644
> index 000000000000..d70c9c3348ac
> --- /dev/null
> +++ b/Documentation/devicetree/bindings/soc/microchip/microchip,mpfs-mss-top-sysreg.yaml
> @@ -0,0 +1,53 @@
> +# SPDX-License-Identifier: GPL-2.0-only OR BSD-2-Clause
> +%YAML 1.2
> +---
> +$id: http://devicetree.org/schemas/soc/microchip/microchip,mpfs-mss-top-sysreg.yaml#
> +$schema: http://devicetree.org/meta-schemas/core.yaml#
> +
> +title: Microchip PolarFire SoC Microprocessor Subsystem (MSS) sysreg Register region
> +
> +maintainers:
> +  - Conor Dooley <conor.dooley@microchip.com>
> +
> +description:
> +  An wide assortment of registers that control elements of the MSS on PolarFire
> +  SoC, including pinmuxing, resets and clocks among others.
> +
> +properties:
> +  compatible:
> +    items:
> +      - const: microchip,mpfs-mss-top-sysreg
> +      - const: syscon
> +      - const: simple-mfd
> +
> +  reg:
> +    maxItems: 1
> +
> +  reset-controller:
> +    type: object
> +
> +    properties:
> +      compatible:
> +        const: microchip,mpfs-reset
> +
> +      '#reset-cells':
> +        description:
> +          The AHB/AXI peripherals on the PolarFire SoC have reset support, so
> +          from CLK_ENVM to CLK_CFM. The reset consumer should specify the
> +          desired peripheral via the clock ID in its "resets" phandle cell.
> +          See include/dt-bindings/clock/microchip,mpfs-clock.h for the full list
> +          of PolarFire clock/reset IDs.
> +        const: 1
> +
> +    additionalProperties: false
> +required:
> +  - compatible
> +  - reg
> +
> +additionalProperties: false
> +
> +examples:
> +  - |
> +    soc {
> +    };
> +
> -- 
> 2.43.0
> 

[-- Attachment #2: signature.asc --]
[-- Type: application/pgp-signature, Size: 228 bytes --]

Re: [RFC PATCH 06/11] dt-bindings: soc: microchip: document the two simple-mfd syscons on PolarFire SoC

[Rob Herring]

On Thu, Aug 15, 2024 at 03:01:09PM +0100, Conor Dooley wrote:
> From: Conor Dooley <conor.dooley@microchip.com>
> 
> There are two syscons on PolarFire SoC that provide various functionality of
> use to the OS.
> 
> The first of these is the "control-scb" region, that contains the "tvs"
> temperature and voltage sensors and the control/status registers for the
> system controller's mailbox. The mailbox has a dedicated node, so
> there's no need for a child node describing it, looking the syscon up by
> compatible is sufficient.
> 
> The second, "mss-top-sysreg", contains clocks, pinctrl, resets, and
> interrupt controller and more. For this RFC, only the reset controller
> child is described as that's all that is described by the existing
> bindings. The clock controller already has a dedicated node, and will
> retain it as there are other clock regions, so like the mailbox,
> a compatible-based lookup of the syscon is sufficient to keep the clock
> driver working as before so no child is needed.

I'm confused. The reset controller is reused from somewhere else? I 
thought you didn't expect any reuse of the IP happening. If a child node 
makes it possible to enable the h/w without any s/w changes, then that 
is a compelling argument for having a child node.

> Signed-off-by: Conor Dooley <conor.dooley@microchip.com>
> ---
> (I'll split this in two later, it's just easier when I have the same
> questions about both...)
> 
> Are these things entitled to have child nodes for the reset and sensor
> nodes, or should the properties be in the parent and the OS probe the
> drivers for the functions? That's something that, despite supposedly
> being a maintainer, I do not understand the rules (of thumb?) for.

Besides the is it an independent, reusable IP block test, my test 
generally is do the child nodes have their own DT resources? Say 
you have phy registers mixed in some syscon and clocks which only go to 
the phy. Then a child node with "clocks" makes sense. If your only 
property is #phy-cells, then a child node doesn't make sense. Of course 
you could reach different conclusions based on the completeness of the 
binding.

> 
> Secondly, is it okay to make the "pragmatic" decision to not have a
> child clock node and keep routing the clocks via the existing & retained
> clock node (and therefore not update the various clocks nodes in the
> consumers)? Doing so would require a lot more hocus pocus with the clock
> driver than this series does, as the same driver would no longer be
> suitable for the before/after bindings.

In the 2 cases here, I don't think you need child nodes. I would expect 
pinctrl to have one though if only as a container for all the pinctrl 
child nodes.

Rob

Re: [RFC PATCH 06/11] dt-bindings: soc: microchip: document the two simple-mfd syscons on PolarFire SoC

[Conor Dooley]

[-- Attachment #1: Type: text/plain, Size: 3869 bytes --]

On Thu, Aug 15, 2024 at 02:00:03PM -0600, Rob Herring wrote:
> On Thu, Aug 15, 2024 at 03:01:09PM +0100, Conor Dooley wrote:
> > From: Conor Dooley <conor.dooley@microchip.com>
> > 
> > There are two syscons on PolarFire SoC that provide various functionality of
> > use to the OS.
> > 
> > The first of these is the "control-scb" region, that contains the "tvs"
> > temperature and voltage sensors and the control/status registers for the
> > system controller's mailbox. The mailbox has a dedicated node, so
> > there's no need for a child node describing it, looking the syscon up by
> > compatible is sufficient.
> > 
> > The second, "mss-top-sysreg", contains clocks, pinctrl, resets, and
> > interrupt controller and more. For this RFC, only the reset controller
> > child is described as that's all that is described by the existing
> > bindings. The clock controller already has a dedicated node, and will
> > retain it as there are other clock regions, so like the mailbox,
> > a compatible-based lookup of the syscon is sufficient to keep the clock
> > driver working as before so no child is needed.
> 
> I'm confused. The reset controller is reused from somewhere else?

There's already a driver for it on this device, but probed via the
auxiliary bus, and the #reset-cells property is in the clock controller
node. The only devices that use this driver are the various different
logic element SKUs (which all share a compatible, they're identical as
far as an OS is concerned) and an upcoming SoC that is effectively a
zero logic element SKU.

> I 
> thought you didn't expect any reuse of the IP happening.

> If a child node 
> makes it possible to enable the h/w without any s/w changes, then that 
> is a compelling argument for having a child node.

No, in both cases there'd be software changes - they're just simpler
with a child node.

> 
> > Signed-off-by: Conor Dooley <conor.dooley@microchip.com>
> > ---
> > (I'll split this in two later, it's just easier when I have the same
> > questions about both...)
> > 
> > Are these things entitled to have child nodes for the reset and sensor
> > nodes, or should the properties be in the parent and the OS probe the
> > drivers for the functions? That's something that, despite supposedly
> > being a maintainer, I do not understand the rules (of thumb?) for.
> 
> Besides the is it an independent, reusable IP block test, my test 
> generally is do the child nodes have their own DT resources? Say 
> you have phy registers mixed in some syscon and clocks which only go to 
> the phy. Then a child node with "clocks" makes sense. If your only 
> property is #phy-cells, then a child node doesn't make sense. Of course 
> you could reach different conclusions based on the completeness of the 
> binding.

AFAIK, none of these things are consumers of resources like that, other
than the interrupt controller, which has an interrupts property. I think
that could justify a child node (and I think a dedicated binding,
because it is a confusing irq mux that that kernel doesn't appear to
have anything else similar to).

> > Secondly, is it okay to make the "pragmatic" decision to not have a
> > child clock node and keep routing the clocks via the existing & retained
> > clock node (and therefore not update the various clocks nodes in the
> > consumers)? Doing so would require a lot more hocus pocus with the clock
> > driver than this series does, as the same driver would no longer be
> > suitable for the before/after bindings.
> 
> In the 2 cases here, I don't think you need child nodes. I would expect 
> pinctrl to have one though if only as a container for all the pinctrl 
> child nodes.

Good to know for when that gets written. Hopefully not by me, I have
enough messes to sort out as is!

Cheers,
Conor.

[-- Attachment #2: signature.asc --]
[-- Type: application/pgp-signature, Size: 228 bytes --]