Skip to contentSkip to conversation

hwmon: (pmbus) Add support for MPS MP2949A

PATCH Superseded [email protected] 2 participants 7 messages Jun 23, 2021 – Jun 29, 2021

Subsystems (2)

v5 of 6 Superseded Latest is v6 →

lore
Share & tools
lore
#0 lore raw

PATCH v5 1/1 hwmon: (pmbus) Add support for MPS MP2949A

From: "Ainux.Wang" <[email protected]>

Add support for MP2949A device from Monolithic Power Systems, Inc. (MPS). This is a triple-loop, digital, multi-phase controller.
This device:
- Supports up to three power rail.
- Provides 6 pulse-width modulations (PWMs), and can be configured up to 6-phase operation for Rail A , up to 2-phase operation for Rail B and up to 1-phase operation for Rail C.
- The PMBus registers are distributed into three pages: Page 0, Page 1, Page 2. Page 0 contains the registers for Rail A and most of the common settings for all of the rails. Page 1 contains register information for Rail B. Page 2 contains register information for Rail C.
- The MP2949A supports both 5mV VID step and 10mv VID step for IMVP8 and IMVP9.

---
v5:
- Moved change log to right here.
v4:
- Removed mp2949a_read_byte_data().
- Added space before and after '-' and fixed a bug that is '~' on line 35.

v3:
- Added change log here.
- Errors in the document have been modified.
- Fixed PMBUS_VOUT_MODE to MP2949A_MFR_VR_CONFIG.
- Removed unnecessary vout_params and empty line.

v2:
- Reference has been added to Documentation/hwmon/index.rst.
- Droped 'vendor'.
- Removed change codes of the PMBus core and added mp2949a_identify() and mp2949a_read_byte_data() to this driver.

v1:
- Add support for MPS MP2949A.

 Documentation/hwmon/index.rst   |   1 +
 Documentation/hwmon/mp2949a.rst |  44 ++++++++++++++
 drivers/hwmon/pmbus/Kconfig     |   9 +++
 drivers/hwmon/pmbus/Makefile    |   1 +
 drivers/hwmon/pmbus/mp2949a.c   | 102 ++++++++++++++++++++++++++++++++
 5 files changed, 157 insertions(+)
 create mode 100644 Documentation/hwmon/mp2949a.rst
 create mode 100644 drivers/hwmon/pmbus/mp2949a.c
Documentation/hwmon/index.rst +1 -0
index 9ed60fa84cbe..56aac3b1678d 100644
@@ -137,6 +137,7 @@ Hardware Monitoring Kernel Drivers
    mcp3021
    menf21bmc
    mlxreg-fan
+   mp2949a
    mp2975
    nct6683
    nct6775
Documentation/hwmon/mp2949a.rst new file +44 -0

44 lines not on this page · raw message

drivers/hwmon/pmbus/Kconfig +9 -0
index 37a5c39784fa..b1344b265976 100644
@@ -248,6 +248,15 @@ config SENSORS_MAX8688
 	  This driver can also be built as a module. If so, the module will
 	  be called max8688.
 
+config SENSORS_MP2949A
+	tristate "MPS MP2949A"
+	help
+	  If you say yes here you get hardware monitoring support for MPS
+	  MP2949A Triple Loop Digital Multi-Phase Controller.
+
+	  This driver can also be built as a module. If so, the module will
+	  be called mp2949a.
+
 config SENSORS_MP2975
 	tristate "MPS MP2975"
 	help
drivers/hwmon/pmbus/Makefile +1 -0
index f8dcc27cd56a..bfb55ab12da1 100644
@@ -28,6 +28,7 @@ obj-$(CONFIG_SENSORS_MAX20751)	+= max20751.o
 obj-$(CONFIG_SENSORS_MAX31785)	+= max31785.o
 obj-$(CONFIG_SENSORS_MAX34440)	+= max34440.o
 obj-$(CONFIG_SENSORS_MAX8688)	+= max8688.o
+obj-$(CONFIG_SENSORS_MP2949A)	+= mp2949a.o
 obj-$(CONFIG_SENSORS_MP2975)	+= mp2975.o
 obj-$(CONFIG_SENSORS_PM6764TR)	+= pm6764tr.o
 obj-$(CONFIG_SENSORS_PXE1610)	+= pxe1610.o
drivers/hwmon/pmbus/mp2949a.c new file +102 -0

102 lines not on this page · raw message

Signature · 1 line
2.18.1
maintainer
#1 lore raw
Quoted 16 lines · On Wed, Jun 23, 2021 at 08:54:26AM +0800, [email protected] wrote:

From: "Ainux.Wang" <[email protected]>

Add support for MP2949A device from Monolithic Power Systems, Inc. (MPS). This is a triple-loop, digital, multi-phase controller.
This device:
- Supports up to three power rail.
- Provides 6 pulse-width modulations (PWMs), and can be configured up to 6-phase operation for Rail A , up to 2-phase operation for Rail B and up to 1-phase operation for Rail C.
- The PMBus registers are distributed into three pages: Page 0, Page 1, Page 2. Page 0 contains the registers for Rail A and most of the common settings for all of the rails. Page 1 contains register information for Rail B. Page 2 contains register information for Rail C.
- The MP2949A supports both 5mV VID step and 10mv VID step for IMVP8 and IMVP9.

Quoted 6 lines

---
v5:
- Moved change log to right here.
v4:
- Removed mp2949a_read_byte_data().

Your other question left me confused. I had previously asked to provide a rationale for filtering out the PMBUS_VOUT_MODE command, and I had stated that "the chip does not support it" is not a valid reason. However, "the chip does not support it but does not report an error when reading it" _is_ a valid reason. So what happens when the PMBus core reads PMBUS_VOUT_MODE ? Does the chip return an error, or some random data ?

Thanks,
Guenter

Quoted 227 lines

- Added space before and after '-' and fixed a bug that is '~' on line 35.

v3:
- Added change log here.
- Errors in the document have been modified.
- Fixed PMBUS_VOUT_MODE to MP2949A_MFR_VR_CONFIG.
- Removed unnecessary vout_params and empty line.

v2:
- Reference has been added to Documentation/hwmon/index.rst.
- Droped 'vendor'.
- Removed change codes of the PMBus core and added mp2949a_identify() and mp2949a_read_byte_data() to this driver.

v1:
- Add support for MPS MP2949A.

 Documentation/hwmon/index.rst   |   1 +
 Documentation/hwmon/mp2949a.rst |  44 ++++++++++++++
 drivers/hwmon/pmbus/Kconfig     |   9 +++
 drivers/hwmon/pmbus/Makefile    |   1 +
 drivers/hwmon/pmbus/mp2949a.c   | 102 ++++++++++++++++++++++++++++++++
 5 files changed, 157 insertions(+)
 create mode 100644 Documentation/hwmon/mp2949a.rst
 create mode 100644 drivers/hwmon/pmbus/mp2949a.c
Documentation/hwmon/index.rst +1 -0
index 9ed60fa84cbe..56aac3b1678d 100644
@@ -137,6 +137,7 @@ Hardware Monitoring Kernel Drivers
    mcp3021
    menf21bmc
    mlxreg-fan
+   mp2949a
    mp2975
    nct6683
    nct6775
Documentation/hwmon/mp2949a.rst new file +44 -0
new file mode 100644
index 000000000000..0235dec70f63
@@ -0,0 +1,44 @@
+.. SPDX-License-Identifier: GPL-2.0
+
+Kernel driver mp2949a
+====================
+
+Supported chips:
+
+  * MPS MP2949A
+
+    Prefix: 'mp2949a'
+
+Author:
+
+	Ainux Wang <[email protected]>
+
+Description
+-----------
+
+This driver implements support for Monolithic Power Systems, Inc. (MPS)
+triple-loop, digital, multi-phase controller MP2949A.
+
+This device:
+
+- Supports up to three power rails.
+- Provides 6 pulse-width modulations (PWMs), and can be configured for
+  to 6-phase operation for Rail A , up to 2-phase operation for Rail B,
+  and up to 1-phase operation for Rail C.
+- The PMBus registers are distributed into three pages: Page 0, Page 1,
+  Page 2. Page 0 contains the registers for Rail A and most of the common
+  settings for all of the rails. Page 1 contains register information for
+  Rail B. Page 2 contains register information for Rail C.
+- The MP2949A supports both 5mV VID step and 10mv VID step for IMVP8 and
+  IMVP9.
+
+Device supports:
+
+- SVID interface.
+- PMBus rev 1.2 interface.
+
+Device supports direct format for reading output power.
+Device supports linear format for reading input voltage, output current,
+and temperature.
+Device supports VID for reading output voltage.
+The below VID modes are supported: VR12, VR13, IMVP8, IMVP9.
drivers/hwmon/pmbus/Kconfig +9 -0
index 37a5c39784fa..b1344b265976 100644
@@ -248,6 +248,15 @@ config SENSORS_MAX8688
 	  This driver can also be built as a module. If so, the module will
 	  be called max8688.
 
+config SENSORS_MP2949A
+	tristate "MPS MP2949A"
+	help
+	  If you say yes here you get hardware monitoring support for MPS
+	  MP2949A Triple Loop Digital Multi-Phase Controller.
+
+	  This driver can also be built as a module. If so, the module will
+	  be called mp2949a.
+
 config SENSORS_MP2975
 	tristate "MPS MP2975"
 	help
drivers/hwmon/pmbus/Makefile +1 -0
index f8dcc27cd56a..bfb55ab12da1 100644
@@ -28,6 +28,7 @@ obj-$(CONFIG_SENSORS_MAX20751)	+= max20751.o
 obj-$(CONFIG_SENSORS_MAX31785)	+= max31785.o
 obj-$(CONFIG_SENSORS_MAX34440)	+= max34440.o
 obj-$(CONFIG_SENSORS_MAX8688)	+= max8688.o
+obj-$(CONFIG_SENSORS_MP2949A)	+= mp2949a.o
 obj-$(CONFIG_SENSORS_MP2975)	+= mp2975.o
 obj-$(CONFIG_SENSORS_PM6764TR)	+= pm6764tr.o
 obj-$(CONFIG_SENSORS_PXE1610)	+= pxe1610.o
drivers/hwmon/pmbus/mp2949a.c new file +102 -0
new file mode 100644
index 000000000000..d1f36c008e58
@@ -0,0 +1,102 @@
+// SPDX-License-Identifier: GPL-2.0-or-later
+/*
+ * Hardware monitoring driver for Monolithic Power Systems MP2949A
+ *
+ * Copyright (c) 2021 Lemote Technologies. All rights reserved.
+ * Copyright (c) 2021 Ainux <[email protected]>
+ */
+
+#include <linux/err.h>
+#include <linux/i2c.h>
+#include <linux/init.h>
+#include <linux/kernel.h>
+#include <linux/module.h>
+#include "pmbus.h"
+
+#define MP2949A_PAGE_NUM		3
+
+#define MP2949A_MFR_VR_CONFIG	0xE4
+
+static int mp2949a_identify(struct i2c_client *client,
+			    struct pmbus_driver_info *info)
+{
+	int i, ret;
+
+	for (i = 0; i < MP2949A_PAGE_NUM; i++) {
+		ret = pmbus_read_byte_data(client, i, MP2949A_MFR_VR_CONFIG);
+		if (ret < 0)
+			return ret;
+
+		/*
+		 * Rail A bit 5, Rail B bit 4, Rail C bit 3.
+		 * 1'b1: 5mV  (vr12/imvp8)
+		 * 1'b0: 10mv (imvp9)
+		 */
+		info->vrm_version[i] = (ret & BIT(5 - i)) ? vr12 : imvp9;
+	}
+
+	return 0;
+}
+
+static struct pmbus_driver_info mp2949a_info = {
+	.pages = MP2949A_PAGE_NUM,
+	.format[PSC_VOLTAGE_IN] = linear,
+	.format[PSC_VOLTAGE_OUT] = vid,
+	.format[PSC_CURRENT_OUT] = linear,
+	.format[PSC_TEMPERATURE] = linear,
+	.format[PSC_POWER] = direct,
+	.m[PSC_POWER] = 1,
+	.b[PSC_POWER] = 0,
+	.R[PSC_POWER] = 0,
+	.func[0] = PMBUS_HAVE_VIN | PMBUS_HAVE_VOUT | PMBUS_HAVE_STATUS_VOUT |
+		PMBUS_HAVE_IOUT | PMBUS_HAVE_STATUS_IOUT |
+		PMBUS_HAVE_TEMP | PMBUS_HAVE_POUT,
+	.func[1] = PMBUS_HAVE_VOUT | PMBUS_HAVE_STATUS_VOUT |
+		PMBUS_HAVE_IOUT | PMBUS_HAVE_STATUS_IOUT |
+		PMBUS_HAVE_POUT,
+	.func[2] = PMBUS_HAVE_VOUT | PMBUS_HAVE_STATUS_VOUT |
+		PMBUS_HAVE_IOUT | PMBUS_HAVE_STATUS_IOUT |
+		PMBUS_HAVE_POUT,
+	.identify = mp2949a_identify,
+};
+
+static int mp2949a_probe(struct i2c_client *client,
+			  const struct i2c_device_id *id)
+{
+	struct pmbus_driver_info *info;
+
+	info = devm_kmemdup(&client->dev, &mp2949a_info, sizeof(*info),
+			    GFP_KERNEL);
+	if (!info)
+		return -ENOMEM;
+
+	return pmbus_do_probe(client, info);
+}
+
+static const struct i2c_device_id mp2949a_id[] = {
+	{"mp2949a", 0},
+	{}
+};
+
+MODULE_DEVICE_TABLE(i2c, mp2949a_id);
+
+static const struct of_device_id mp2949a_of_match[] = {
+	{.compatible = "mps,mp2949a"},
+	{}
+};
+MODULE_DEVICE_TABLE(of, mp2949a_of_match);
+
+static struct i2c_driver mp2949a_driver = {
+	.driver = {
+		.name = "mp2949a",
+		.of_match_table = of_match_ptr(mp2949a_of_match),
+	},
+	.probe = mp2949a_probe,
+	.id_table = mp2949a_id,
+};
+
+module_i2c_driver(mp2949a_driver);
+
+MODULE_AUTHOR("Ainux <[email protected]>");
+MODULE_DESCRIPTION("PMBus driver for Monolithic Power Systems MP2949A");
+MODULE_LICENSE("GPL");
maintainer
#2 lore raw
Quoted 24 lines · Guenter Roeck
Quoted 16 lines · On Wed, Jun 23, 2021 at 08:54:26AM +0800, [email protected] wrote:

From: "Ainux.Wang" <[email protected]>

Add support for MP2949A device from Monolithic Power Systems, Inc. (MPS). This is a triple-loop, digital, multi-phase controller.
This device:
- Supports up to three power rail.
- Provides 6 pulse-width modulations (PWMs), and can be configured up to 6-phase operation for Rail A , up to 2-phase operation for Rail B and up to 1-phase operation for Rail C.
- The PMBus registers are distributed into three pages: Page 0, Page 1, Page 2. Page 0 contains the registers for Rail A and most of the common settings for all of the rails. Page 1 contains register information for Rail B. Page 2 contains register information for Rail C.
- The MP2949A supports both 5mV VID step and 10mv VID step for IMVP8 and IMVP9.

Quoted 6 lines

---
v5:
- Moved change log to right here.
v4:
- Removed mp2949a_read_byte_data().

Quoted 7 lines

Your other question left me confused. I had previously asked to provide a rationale for filtering out the PMBUS_VOUT_MODE command, and I had stated that "the chip does not support it" is not a valid reason. However, "the chip does not support it but does not report an error when reading it" _is_ a valid reason. So what happens when the PMBus core reads PMBUS_VOUT_MODE ? Does the chip return an error, or some random data ?

Still waiting for a reply.

Guenter

author
#3 lore raw
Quoted 34 lines · Guenter Roeck
Quoted 24 lines · Guenter Roeck
Quoted 16 lines · On Wed, Jun 23, 2021 at 08:54:26AM +0800, [email protected] wrote:

From: "Ainux.Wang" <[email protected]>

Add support for MP2949A device from Monolithic Power Systems, Inc. (MPS). This is a triple-loop, digital, multi-phase controller.
This device:
- Supports up to three power rail.
- Provides 6 pulse-width modulations (PWMs), and can be configured up to 6-phase operation for Rail A , up to 2-phase operation for Rail B and up to 1-phase operation for Rail C.
- The PMBus registers are distributed into three pages: Page 0, Page 1, Page 2. Page 0 contains the registers for Rail A and most of the common settings for all of the rails. Page 1 contains register information for Rail B. Page 2 contains register information for Rail C.
- The MP2949A supports both 5mV VID step and 10mv VID step for IMVP8 and IMVP9.

Quoted 6 lines

---
v5:
- Moved change log to right here.
v4:
- Removed mp2949a_read_byte_data().

Quoted 7 lines

Your other question left me confused. I had previously asked to provide a rationale for filtering out the PMBUS_VOUT_MODE command, and I had stated that "the chip does not support it" is not a valid reason. However, "the chip does not support it but does not report an error when reading it" _is_ a valid reason. So what happens when the PMBus core reads PMBUS_VOUT_MODE ? Does the chip return an error, or some random data ?

Quoted 4 lines

Still waiting for a reply.

Guenter

Hi, Guenter

Sorry to have kept you waiting.

Best regards,
Ainux Wang.

author
#4 lore raw
In reply to Guenter Roeck
Quoted 32 lines · Guenter Roeck
Quoted 16 lines · On Wed, Jun 23, 2021 at 08:54:26AM +0800, [email protected] wrote:

From: "Ainux.Wang" <[email protected]>

Add support for MP2949A device from Monolithic Power Systems, Inc. (MPS). This is a triple-loop, digital, multi-phase controller.
This device:
- Supports up to three power rail.
- Provides 6 pulse-width modulations (PWMs), and can be configured up to 6-phase operation for Rail A , up to 2-phase operation for Rail B and up to 1-phase operation for Rail C.
- The PMBus registers are distributed into three pages: Page 0, Page 1, Page 2. Page 0 contains the registers for Rail A and most of the common settings for all of the rails. Page 1 contains register information for Rail B. Page 2 contains register information for Rail C.
- The MP2949A supports both 5mV VID step and 10mv VID step for IMVP8 and IMVP9.

Quoted 6 lines

---
v5:
- Moved change log to right here.
v4:
- Removed mp2949a_read_byte_data().

Your other question left me confused. I had previously asked to provide a rationale for filtering out the PMBUS_VOUT_MODE command, and I had stated that "the chip does not support it" is not a valid reason. However, "the chip does not support it but does not report an error when reading it" _is_ a valid reason. So what happens when the PMBus core reads PMBUS_VOUT_MODE ? Does the chip return an error, or some random data ?

Quoted 3 lines

Thanks,
Guenter

Hi, Guenter

I have not been clear about the cause of this problem that "the chip does not support it is not a valid reason".

However, i have added some "printk", i found the chip will return some random data,
when the PMbus core reads PMBUS_STATUS and PMBUS_VOUT_MODE.

So, I do not known what should i do, when the PMbus core reads PMBUS_STATUS. and i have knwon that use "This chip do not support the VOUT_MODE command, the chip does not support it but return some random data when reading " instead of
"
/*
* This chip do not support the VOUT_MODE command.
* There is not VOUT_MODE command in MP2949A datasheet P29~P31.
* So this is EINVAL in here.
*/
"
Now, there is only one question left, what should the PMbus do, the chip return random data,
when it reads PMBUS_STATUS by i2c_smbus_read_word_data() or i2c_smbus_read_byte_data()?
Can the PMbus core use pmbus_read_status_byte() and pmbus_read_status_word() instead of
i2c_smbus_read_byte_data() and i2c_smbus_read_word_data() ?

Best regards,
Ainux Wang.

Quoted 227 lines
Quoted 227 lines

- Added space before and after '-' and fixed a bug that is '~' on line 35.

v3:
- Added change log here.
- Errors in the document have been modified.
- Fixed PMBUS_VOUT_MODE to MP2949A_MFR_VR_CONFIG.
- Removed unnecessary vout_params and empty line.

v2:
- Reference has been added to Documentation/hwmon/index.rst.
- Droped 'vendor'.
- Removed change codes of the PMBus core and added mp2949a_identify() and mp2949a_read_byte_data() to this driver.

v1:
- Add support for MPS MP2949A.

 Documentation/hwmon/index.rst   |   1 +
 Documentation/hwmon/mp2949a.rst |  44 ++++++++++++++
 drivers/hwmon/pmbus/Kconfig     |   9 +++
 drivers/hwmon/pmbus/Makefile    |   1 +
 drivers/hwmon/pmbus/mp2949a.c   | 102 ++++++++++++++++++++++++++++++++
 5 files changed, 157 insertions(+)
 create mode 100644 Documentation/hwmon/mp2949a.rst
 create mode 100644 drivers/hwmon/pmbus/mp2949a.c
Documentation/hwmon/index.rst +1 -0
index 9ed60fa84cbe..56aac3b1678d 100644
@@ -137,6 +137,7 @@ Hardware Monitoring Kernel Drivers
    mcp3021
    menf21bmc
    mlxreg-fan
+   mp2949a
    mp2975
    nct6683
    nct6775
Documentation/hwmon/mp2949a.rst new file +44 -0
new file mode 100644
index 000000000000..0235dec70f63
@@ -0,0 +1,44 @@
+.. SPDX-License-Identifier: GPL-2.0
+
+Kernel driver mp2949a
+====================
+
+Supported chips:
+
+  * MPS MP2949A
+
+    Prefix: 'mp2949a'
+
+Author:
+
+     Ainux Wang <[email protected]>
+
+Description
+-----------
+
+This driver implements support for Monolithic Power Systems, Inc. (MPS)
+triple-loop, digital, multi-phase controller MP2949A.
+
+This device:
+
+- Supports up to three power rails.
+- Provides 6 pulse-width modulations (PWMs), and can be configured for
+  to 6-phase operation for Rail A , up to 2-phase operation for Rail B,
+  and up to 1-phase operation for Rail C.
+- The PMBus registers are distributed into three pages: Page 0, Page 1,
+  Page 2. Page 0 contains the registers for Rail A and most of the common
+  settings for all of the rails. Page 1 contains register information for
+  Rail B. Page 2 contains register information for Rail C.
+- The MP2949A supports both 5mV VID step and 10mv VID step for IMVP8 and
+  IMVP9.
+
+Device supports:
+
+- SVID interface.
+- PMBus rev 1.2 interface.
+
+Device supports direct format for reading output power.
+Device supports linear format for reading input voltage, output current,
+and temperature.
+Device supports VID for reading output voltage.
+The below VID modes are supported: VR12, VR13, IMVP8, IMVP9.
drivers/hwmon/pmbus/Kconfig +9 -0
index 37a5c39784fa..b1344b265976 100644
@@ -248,6 +248,15 @@ config SENSORS_MAX8688
        This driver can also be built as a module. If so, the module will
        be called max8688.
 
+config SENSORS_MP2949A
+     tristate "MPS MP2949A"
+     help
+       If you say yes here you get hardware monitoring support for MPS
+       MP2949A Triple Loop Digital Multi-Phase Controller.
+
+       This driver can also be built as a module. If so, the module will
+       be called mp2949a.
+
 config SENSORS_MP2975
      tristate "MPS MP2975"
      help
drivers/hwmon/pmbus/Makefile +1 -0
index f8dcc27cd56a..bfb55ab12da1 100644
@@ -28,6 +28,7 @@ obj-$(CONFIG_SENSORS_MAX20751)      += max20751.o
 obj-$(CONFIG_SENSORS_MAX31785)       += max31785.o
 obj-$(CONFIG_SENSORS_MAX34440)       += max34440.o
 obj-$(CONFIG_SENSORS_MAX8688)        += max8688.o
+obj-$(CONFIG_SENSORS_MP2949A)        += mp2949a.o
 obj-$(CONFIG_SENSORS_MP2975) += mp2975.o
 obj-$(CONFIG_SENSORS_PM6764TR)       += pm6764tr.o
 obj-$(CONFIG_SENSORS_PXE1610)        += pxe1610.o
drivers/hwmon/pmbus/mp2949a.c new file +102 -0
new file mode 100644
index 000000000000..d1f36c008e58
@@ -0,0 +1,102 @@
+// SPDX-License-Identifier: GPL-2.0-or-later
+/*
+ * Hardware monitoring driver for Monolithic Power Systems MP2949A
+ *
+ * Copyright (c) 2021 Lemote Technologies. All rights reserved.
+ * Copyright (c) 2021 Ainux <[email protected]>
+ */
+
+#include <linux/err.h>
+#include <linux/i2c.h>
+#include <linux/init.h>
+#include <linux/kernel.h>
+#include <linux/module.h>
+#include "pmbus.h"
+
+#define MP2949A_PAGE_NUM             3
+
+#define MP2949A_MFR_VR_CONFIG        0xE4
+
+static int mp2949a_identify(struct i2c_client *client,
+                         struct pmbus_driver_info *info)
+{
+     int i, ret;
+
+     for (i = 0; i < MP2949A_PAGE_NUM; i++) {
+             ret = pmbus_read_byte_data(client, i, MP2949A_MFR_VR_CONFIG);
+             if (ret < 0)
+                     return ret;
+
+             /*
+              * Rail A bit 5, Rail B bit 4, Rail C bit 3.
+              * 1'b1: 5mV  (vr12/imvp8)
+              * 1'b0: 10mv (imvp9)
+              */
+             info->vrm_version[i] = (ret & BIT(5 - i)) ? vr12 : imvp9;
+     }
+
+     return 0;
+}
+
+static struct pmbus_driver_info mp2949a_info = {
+     .pages = MP2949A_PAGE_NUM,
+     .format[PSC_VOLTAGE_IN] = linear,
+     .format[PSC_VOLTAGE_OUT] = vid,
+     .format[PSC_CURRENT_OUT] = linear,
+     .format[PSC_TEMPERATURE] = linear,
+     .format[PSC_POWER] = direct,
+     .m[PSC_POWER] = 1,
+     .b[PSC_POWER] = 0,
+     .R[PSC_POWER] = 0,
+     .func[0] = PMBUS_HAVE_VIN | PMBUS_HAVE_VOUT | PMBUS_HAVE_STATUS_VOUT |
+             PMBUS_HAVE_IOUT | PMBUS_HAVE_STATUS_IOUT |
+             PMBUS_HAVE_TEMP | PMBUS_HAVE_POUT,
+     .func[1] = PMBUS_HAVE_VOUT | PMBUS_HAVE_STATUS_VOUT |
+             PMBUS_HAVE_IOUT | PMBUS_HAVE_STATUS_IOUT |
+             PMBUS_HAVE_POUT,
+     .func[2] = PMBUS_HAVE_VOUT | PMBUS_HAVE_STATUS_VOUT |
+             PMBUS_HAVE_IOUT | PMBUS_HAVE_STATUS_IOUT |
+             PMBUS_HAVE_POUT,
+     .identify = mp2949a_identify,
+};
+
+static int mp2949a_probe(struct i2c_client *client,
+                       const struct i2c_device_id *id)
+{
+     struct pmbus_driver_info *info;
+
+     info = devm_kmemdup(&client->dev, &mp2949a_info, sizeof(*info),
+                         GFP_KERNEL);
+     if (!info)
+             return -ENOMEM;
+
+     return pmbus_do_probe(client, info);
+}
+
+static const struct i2c_device_id mp2949a_id[] = {
+     {"mp2949a", 0},
+     {}
+};
+
+MODULE_DEVICE_TABLE(i2c, mp2949a_id);
+
+static const struct of_device_id mp2949a_of_match[] = {
+     {.compatible = "mps,mp2949a"},
+     {}
+};
+MODULE_DEVICE_TABLE(of, mp2949a_of_match);
+
+static struct i2c_driver mp2949a_driver = {
+     .driver = {
+             .name = "mp2949a",
+             .of_match_table = of_match_ptr(mp2949a_of_match),
+     },
+     .probe = mp2949a_probe,
+     .id_table = mp2949a_id,
+};
+
+module_i2c_driver(mp2949a_driver);
+
+MODULE_AUTHOR("Ainux <[email protected]>");
+MODULE_DESCRIPTION("PMBus driver for Monolithic Power Systems MP2949A");
+MODULE_LICENSE("GPL");
maintainer
#5 lore raw
Quoted 55 lines · Ainux Wang
Quoted 32 lines · Guenter Roeck
Quoted 16 lines · On Wed, Jun 23, 2021 at 08:54:26AM +0800, [email protected] wrote:

From: "Ainux.Wang" <[email protected]>

Add support for MP2949A device from Monolithic Power Systems, Inc. (MPS). This is a triple-loop, digital, multi-phase controller.
This device:
- Supports up to three power rail.
- Provides 6 pulse-width modulations (PWMs), and can be configured up to 6-phase operation for Rail A , up to 2-phase operation for Rail B and up to 1-phase operation for Rail C.
- The PMBus registers are distributed into three pages: Page 0, Page 1, Page 2. Page 0 contains the registers for Rail A and most of the common settings for all of the rails. Page 1 contains register information for Rail B. Page 2 contains register information for Rail C.
- The MP2949A supports both 5mV VID step and 10mv VID step for IMVP8 and IMVP9.

Quoted 6 lines

---
v5:
- Moved change log to right here.
v4:
- Removed mp2949a_read_byte_data().

Your other question left me confused. I had previously asked to provide a rationale for filtering out the PMBUS_VOUT_MODE command, and I had stated that "the chip does not support it" is not a valid reason. However, "the chip does not support it but does not report an error when reading it" _is_ a valid reason. So what happens when the PMBus core reads PMBUS_VOUT_MODE ? Does the chip return an error, or some random data ?

Quoted 3 lines

Thanks,
Guenter

Hi, Guenter

I have not been clear about the cause of this problem that "the chip does not support it is not a valid reason".

However, i have added some "printk", i found the chip will return some random data,
when the PMbus core reads PMBUS_STATUS and PMBUS_VOUT_MODE.

So, I do not known what should i do, when the PMbus core reads PMBUS_STATUS. and i have knwon that use "This chip do not support the VOUT_MODE command, the chip does not support it but return some random data when reading " instead of
"
/*
* This chip do not support the VOUT_MODE command.
* There is not VOUT_MODE command in MP2949A datasheet P29~P31.
* So this is EINVAL in here.
*/

Quoted 9 lines

"
Now, there is only one question left, what should the PMbus do, the chip return random data,
when it reads PMBUS_STATUS by i2c_smbus_read_word_data() or i2c_smbus_read_byte_data()?
Can the PMbus core use pmbus_read_status_byte() and pmbus_read_status_word() instead of i2c_smbus_read_byte_data() and i2c_smbus_read_word_data() ?

The driver should implement a read_byte_data() function and either simulate PMBUS_VOUT_MODE and let it return whatever makes sense for the driver, or have it return -EINVAL.

For the status register, handling is a bit more difficult. For this, we will need an introductory patch. That patch needs to change pmbus_init_common() to call pmbus_read_status_word() and pmbus_read_status_byte() instead of i2c_smbus_read_word_data() and i2c_smbus_read_byte_data() (with page set to -1).
The driver patch then needs to implement a read_word_data() function and have it simulate the PMBUS_STATUS_WORD command. Or, alternatively, have the read_word_data() function return -EINVAL for PMBUS_STATUS_WORD, and simulate PMBUS_STATUS_BYTE in the read_byte_data callback.

Thanks,
Guenter

author
#6 lore raw
Quoted 81 lines · Guenter Roeck
Quoted 55 lines · Ainux Wang
Quoted 32 lines · Guenter Roeck
Quoted 16 lines · On Wed, Jun 23, 2021 at 08:54:26AM +0800, [email protected] wrote:

From: "Ainux.Wang" <[email protected]>

Add support for MP2949A device from Monolithic Power Systems, Inc. (MPS). This is a triple-loop, digital, multi-phase controller.
This device:
- Supports up to three power rail.
- Provides 6 pulse-width modulations (PWMs), and can be configured up to 6-phase operation for Rail A , up to 2-phase operation for Rail B and up to 1-phase operation for Rail C.
- The PMBus registers are distributed into three pages: Page 0, Page 1, Page 2. Page 0 contains the registers for Rail A and most of the common settings for all of the rails. Page 1 contains register information for Rail B. Page 2 contains register information for Rail C.
- The MP2949A supports both 5mV VID step and 10mv VID step for IMVP8 and IMVP9.

Quoted 6 lines

---
v5:
- Moved change log to right here.
v4:
- Removed mp2949a_read_byte_data().

Your other question left me confused. I had previously asked to provide a rationale for filtering out the PMBUS_VOUT_MODE command, and I had stated that "the chip does not support it" is not a valid reason. However, "the chip does not support it but does not report an error when reading it" _is_ a valid reason. So what happens when the PMBus core reads PMBUS_VOUT_MODE ? Does the chip return an error, or some random data ?

Quoted 3 lines

Thanks,
Guenter

Hi, Guenter

I have not been clear about the cause of this problem that "the chip does not support it is not a valid reason".

However, i have added some "printk", i found the chip will return some random data,
when the PMbus core reads PMBUS_STATUS and PMBUS_VOUT_MODE.

So, I do not known what should i do, when the PMbus core reads PMBUS_STATUS. and i have knwon that use "This chip do not support the VOUT_MODE command, the chip does not support it but return some random data when reading " instead of
"
/*
* This chip do not support the VOUT_MODE command.
* There is not VOUT_MODE command in MP2949A datasheet P29~P31.
* So this is EINVAL in here.
*/

Quoted 9 lines

"
Now, there is only one question left, what should the PMbus do, the chip return random data,
when it reads PMBUS_STATUS by i2c_smbus_read_word_data() or i2c_smbus_read_byte_data()?
Can the PMbus core use pmbus_read_status_byte() and pmbus_read_status_word() instead of i2c_smbus_read_byte_data() and i2c_smbus_read_word_data() ?

The driver should implement a read_byte_data() function and either simulate PMBUS_VOUT_MODE and let it return whatever makes sense for the driver, or have it return -EINVAL.

For the status register, handling is a bit more difficult. For this, we will need an introductory patch. That patch needs to change pmbus_init_common() to call pmbus_read_status_word() and pmbus_read_status_byte() instead of i2c_smbus_read_word_data() and i2c_smbus_read_byte_data() (with page set to -1).
The driver patch then needs to implement a read_word_data() function and have it simulate the PMBUS_STATUS_WORD command. Or, alternatively, have the read_word_data() function return -EINVAL for PMBUS_STATUS_WORD, and simulate PMBUS_STATUS_BYTE in the read_byte_data callback.

Quoted 2 lines

Thanks,
Guenter

Hi, Guenter

Very nice, that is what i mean, we have the same idea, I am going to revise pmbus_init_common(),
pmbus_check_register and so on.

Best regards,
Ainux Wang.