2010-06-24 22:06:14

by Guenter Roeck

[permalink] [raw]
Subject: [PATCH v3 0/4] hwmon: Driver for SMM665 Six-Channel Active DC Output Controller/Monitor

This patchset adds support for the monitoring features of the Summit
Microelectronics SMM665 Six-Channel Active DC Output Controller/Monitor.

------

v2 changes:
- Added (untested) support for SMM465, SMM764, and SMM766.
- Added vref module parameter to support non-default values of VREF_ADC.
- Added an explanation describing what the chip does with _min and _max limit
values.
- Removed alarm sysfs files (since the chip does not report alarm status in a
register, alarm detection was implemented by comparing limits with ADC readings,
which can as well be done in userland).
- Replaced register enums with defines.
- Made limit variable names better readable.
- Removed retries from smm665_read_adc().
- Check error reply from i2c_smbus_read_byte_data() and only accept ENXIO
when expecting NACK.
- Explained reason for byte swap after reading ADC registers in some more detail.
- Added debug messages to display the error reason when reading ADC registers
failed.

v3 changes:
- Return error codes from smm665_read_adc().
- Abort after first error in smm665_update_device() and return error code.
- Return ADC read errors to user.
- Added inX_lcrit and inX_crit attributes for min/max critical voltages.
- Added tempX_lcrit attribute for critical low temperature.
- Updated sysfs-interface to document support for new attributes.


2010-06-24 22:06:16

by Guenter Roeck

[permalink] [raw]
Subject: [PATCH v3 4/4] hwmon: Update MAINTAINERS for smm665 driver

Signed-off-by: Guenter Roeck <[email protected]>
---
MAINTAINERS | 7 +++++++
1 files changed, 7 insertions(+), 0 deletions(-)

diff --git a/MAINTAINERS b/MAINTAINERS
index 6d119c9..2e8334d 100644
--- a/MAINTAINERS
+++ b/MAINTAINERS
@@ -5205,6 +5205,13 @@ M: Nicolas Pitre <[email protected]>
S: Odd Fixes
F: drivers/net/smc91x.*

+SMM665 HARDWARE MONITOR DRIVER
+M: "Guenter Roeck" <[email protected]>
+L: [email protected]
+S: Maintained
+F: Documentation/hwmon/smm665
+F: drivers/hwmon/smm665.c
+
SMSC47B397 HARDWARE MONITOR DRIVER
M: "Mark M. Hoffman" <[email protected]>
L: [email protected]
--
1.7.0.87.g0901d

2010-06-24 22:06:25

by Guenter Roeck

[permalink] [raw]
Subject: [PATCH v3 1/4] hwmon: Driver for SMM665 Six-Channel Active DC Output Controller/Monitor

This driver adds support for the monitoring features of the Summit
Microelectronics SMM665 Six-Channel Active DC Output Controller/Monitor.

Signed-off-by: Guenter Roeck <[email protected]>
---
drivers/hwmon/Kconfig | 15 +
drivers/hwmon/Makefile | 1 +
drivers/hwmon/smm665.c | 744 ++++++++++++++++++++++++++++++++++++++++++++++++
3 files changed, 760 insertions(+), 0 deletions(-)
create mode 100644 drivers/hwmon/smm665.c

diff --git a/drivers/hwmon/Kconfig b/drivers/hwmon/Kconfig
index e19cf8e..c45c17e 100644
--- a/drivers/hwmon/Kconfig
+++ b/drivers/hwmon/Kconfig
@@ -739,6 +739,21 @@ config SENSORS_SIS5595
This driver can also be built as a module. If so, the module
will be called sis5595.

+config SENSORS_SMM665
+ tristate "Summit Microelectronics SMM665"
+ depends on I2C && EXPERIMENTAL
+ default n
+ help
+ If you say yes here you get support for the hardware monitoring
+ features of the Summit Microelectronics SMM665/SMM665B Six-Channel
+ Active DC Output Controller / Monitor.
+
+ Other supported chips are SMM465, SMM665C, SMM764, and SMM766.
+ Support for those chips is untested.
+
+ This driver can also be built as a module. If so, the module will
+ be called smm665.
+
config SENSORS_DME1737
tristate "SMSC DME1737, SCH311x and compatibles"
depends on I2C && EXPERIMENTAL
diff --git a/drivers/hwmon/Makefile b/drivers/hwmon/Makefile
index 2138ceb..55b0940 100644
--- a/drivers/hwmon/Makefile
+++ b/drivers/hwmon/Makefile
@@ -86,6 +86,7 @@ obj-$(CONFIG_SENSORS_PCF8591) += pcf8591.o
obj-$(CONFIG_SENSORS_S3C) += s3c-hwmon.o
obj-$(CONFIG_SENSORS_SHT15) += sht15.o
obj-$(CONFIG_SENSORS_SIS5595) += sis5595.o
+obj-$(CONFIG_SENSORS_SMM665) += smm665.o
obj-$(CONFIG_SENSORS_SMSC47B397)+= smsc47b397.o
obj-$(CONFIG_SENSORS_SMSC47M1) += smsc47m1.o
obj-$(CONFIG_SENSORS_SMSC47M192)+= smsc47m192.o
diff --git a/drivers/hwmon/smm665.c b/drivers/hwmon/smm665.c
new file mode 100644
index 0000000..d340f61
--- /dev/null
+++ b/drivers/hwmon/smm665.c
@@ -0,0 +1,744 @@
+/*
+ * Driver for SMM665 Power Controller / Monitor
+ *
+ * Copyright (C) 2010 Ericsson AB.
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License as published by
+ * the Free Software Foundation; version 2 of the License.
+ *
+ * This driver should also work for SMM465, SMM764, and SMM766, but is untested
+ * for those chips. Only monitoring functionality is implemented.
+ *
+ * Datasheets:
+ * http://www.summitmicro.com/prod_select/summary/SMM665/SMM665B_2089_20.pdf
+ * http://www.summitmicro.com/prod_select/summary/SMM766B/SMM766B_2122.pdf
+ */
+
+#include <linux/kernel.h>
+#include <linux/module.h>
+#include <linux/init.h>
+#include <linux/err.h>
+#include <linux/slab.h>
+#include <linux/i2c.h>
+#include <linux/hwmon.h>
+#include <linux/hwmon-sysfs.h>
+#include <linux/delay.h>
+
+/* Internal reference voltage (VREF, x 1000 */
+#define SMM665_VREF_ADC_X1000 1250
+
+/* module parameters */
+static int vref = SMM665_VREF_ADC_X1000;
+module_param(vref, int, 0);
+MODULE_PARM_DESC(vref, "Reference voltage in mV");
+
+enum chips { smm465, smm665, smm665c, smm764, smm766 };
+
+/*
+ * ADC channel addresses
+ */
+#define SMM665_MISC16_ADC_DATA_A 0x00
+#define SMM665_MISC16_ADC_DATA_B 0x01
+#define SMM665_MISC16_ADC_DATA_C 0x02
+#define SMM665_MISC16_ADC_DATA_D 0x03
+#define SMM665_MISC16_ADC_DATA_E 0x04
+#define SMM665_MISC16_ADC_DATA_F 0x05
+#define SMM665_MISC16_ADC_DATA_VDD 0x06
+#define SMM665_MISC16_ADC_DATA_12V 0x07
+#define SMM665_MISC16_ADC_DATA_INT_TEMP 0x08
+#define SMM665_MISC16_ADC_DATA_AIN1 0x09
+#define SMM665_MISC16_ADC_DATA_AIN2 0x0a
+
+/*
+ * Command registers
+ */
+#define SMM665_MISC8_CMD_STS 0x80
+#define SMM665_MISC8_STATUS1 0x81
+#define SMM665_MISC8_STATUSS2 0x82
+#define SMM665_MISC8_IO_POLARITY 0x83
+#define SMM665_MISC8_PUP_POLARITY 0x84
+#define SMM665_MISC8_ADOC_STATUS1 0x85
+#define SMM665_MISC8_ADOC_STATUS2 0x86
+#define SMM665_MISC8_WRITE_PROT 0x87
+#define SMM665_MISC8_STS_TRACK 0x88
+
+/*
+ * Configuration registers and register groups
+ */
+#define SMM665_ADOC_ENABLE 0x0d
+#define SMM665_LIMIT_BASE 0x80 /* First limit register */
+
+/*
+ * Limit register bit masks
+ */
+#define SMM665_TRIGGER_RST 0x8000
+#define SMM665_TRIGGER_HEALTHY 0x4000
+#define SMM665_TRIGGER_POWEROFF 0x2000
+#define SMM665_TRIGGER_SHUTDOWN 0x1000
+#define SMM665_ADC_MASK 0x03ff
+
+#define smm665_is_critical(lim) ((lim) & (SMM665_TRIGGER_RST \
+ | SMM665_TRIGGER_POWEROFF \
+ | SMM665_TRIGGER_SHUTDOWN))
+/*
+ * Fault register bit definitions
+ * Values are merged from status registers 1/2,
+ * with status register 1 providing the upper 8 bits.
+ */
+#define SMM665_FAULT_A 0x0001
+#define SMM665_FAULT_B 0x0002
+#define SMM665_FAULT_C 0x0004
+#define SMM665_FAULT_D 0x0008
+#define SMM665_FAULT_E 0x0010
+#define SMM665_FAULT_F 0x0020
+#define SMM665_FAULT_VDD 0x0040
+#define SMM665_FAULT_12V 0x0080
+#define SMM665_FAULT_TEMP 0x0100
+#define SMM665_FAULT_AIN1 0x0200
+#define SMM665_FAULT_AIN2 0x0400
+
+/*
+ * I2C Register addresses
+ *
+ * The configuration register needs to be the configured base register.
+ * The command/status register address is derived from it.
+ */
+#define SMM665_REGMASK 0x78
+#define SMM665_CMDREG_BASE 0x48
+#define SMM665_CONFREG_BASE 0x50
+
+/*
+ * Equations given by chip manufacturer to calculate voltage/temperature values
+ * vref = Reference voltage on VREF_ADC pin (module parameter)
+ * adc = 10bit ADC value read back from registers
+ */
+
+/* Voltage A-F and VDD */
+#define SMM665_VMON_ADC_TO_VOLTS(adc) ((adc) * vref / 256)
+
+/* Voltage 12VIN */
+#define SMM665_12VIN_ADC_TO_VOLTS(adc) ((adc) * vref * 3 / 256)
+
+/* Voltage AIN1, AIN2 */
+#define SMM665_AIN_ADC_TO_VOLTS(adc) ((adc) * vref / 512)
+
+/* Temp Sensor */
+#define SMM665_TEMP_ADC_TO_CELSIUS(adc) ((adc) <= 511) ? \
+ ((int)(adc) * 1000 / 4) : \
+ (((int)(adc) - 0x400) * 1000 / 4)
+
+#define SMM665_NUM_ADC 11
+
+/*
+ * Chip dependent ADC conversion time, in uS
+ */
+#define SMM665_ADC_WAIT_SMM665 70
+#define SMM665_ADC_WAIT_SMM766 185
+
+struct smm665_data {
+ enum chips type;
+ int conversion_time; /* ADC conversion time */
+ struct device *hwmon_dev;
+ struct mutex update_lock;
+ bool valid;
+ unsigned long last_updated; /* in jiffies */
+ u16 adc[SMM665_NUM_ADC]; /* adc values (raw) */
+ u16 faults; /* fault status */
+ /* The following values are in mV */
+ int critical_min_limit[SMM665_NUM_ADC];
+ int alarm_min_limit[SMM665_NUM_ADC];
+ int critical_max_limit[SMM665_NUM_ADC];
+ int alarm_max_limit[SMM665_NUM_ADC];
+ struct i2c_client *cmdreg;
+};
+
+/*
+ * smm665_read16()
+ *
+ * Read 16 bit value from <reg>, <reg+1>. Upper 8 bits are in <reg>.
+ */
+static int smm665_read16(struct i2c_client *client, int reg)
+{
+ int rv, val;
+
+ rv = i2c_smbus_read_byte_data(client, reg);
+ if (rv < 0)
+ return rv;
+ val = rv << 8;
+ rv = i2c_smbus_read_byte_data(client, reg + 1);
+ if (rv < 0)
+ return rv;
+ val |= rv;
+ return val;
+}
+
+/*
+ * Read adc value.
+ */
+static int smm665_read_adc(struct smm665_data *data, int adc)
+{
+ struct i2c_client *client = data->cmdreg;
+ int rv;
+ int radc;
+
+ /*
+ * Algorithm for reading ADC, per SMM665 datasheet
+ *
+ * {[S][addr][W][Ack]} {[offset][Ack]} {[S][addr][R][Nack]}
+ * [wait conversion time]
+ * {[S][addr][R][Ack]} {[datahi][Ack]} {[datalo][Ack][P]}
+ *
+ * To implement the first part of this exchange,
+ * do a full read transaction and expect a failure/Nack.
+ * This sets up the address pointer on the SMM665
+ * and starts the ADC conversion.
+ * Then do a two-byte read transaction.
+ */
+ rv = i2c_smbus_read_byte_data(client, adc << 3);
+ if (rv != -ENXIO) {
+ /*
+ * We expect ENXIO to reflect NACK
+ * (per Documentation/i2c/fault-codes).
+ * Everything else is an error.
+ */
+ dev_dbg(&client->dev,
+ "Unexpected return code %d when setting ADC index", rv);
+ return (rv < 0) ? rv : -EIO;
+ }
+
+ udelay(data->conversion_time);
+
+ /*
+ * Now read two bytes.
+ *
+ * Neither i2c_smbus_read_byte() nor
+ * i2c_smbus_read_block_data() worked here,
+ * so use i2c_smbus_read_word_data() instead.
+ * We could also try to use i2c_master_recv(),
+ * but that is not always supported.
+ */
+ rv = i2c_smbus_read_word_data(client, 0);
+ if (rv < 0) {
+ dev_dbg(&client->dev, "Failed to read ADC value: error %d", rv);
+ return -1;
+ }
+ /*
+ * Validate/verify readback adc channel (in bit 11..14).
+ * High byte is in lower 8 bit of rv, so only shift by 3.
+ */
+ radc = (rv >> 3) & 0x0f;
+ if (radc != adc) {
+ dev_dbg(&client->dev, "Unexpected RADC: Expected %d got %d",
+ adc, radc);
+ return -EIO;
+ }
+ /*
+ * Chip replies with H/L, while SMBus expects L/H.
+ * Thus, byte order is reversed, and we have to swap
+ * the result.
+ */
+ rv = swab16(rv) & SMM665_ADC_MASK;
+
+ return rv;
+}
+
+static struct smm665_data *smm665_update_device(struct device *dev)
+{
+ struct i2c_client *client = to_i2c_client(dev);
+ struct smm665_data *data = i2c_get_clientdata(client);
+
+ mutex_lock(&data->update_lock);
+
+ if (time_after(jiffies, data->last_updated + HZ) || !data->valid) {
+ int i, val;
+
+ /*
+ * read status registers
+ */
+ val = smm665_read16(client, SMM665_MISC8_STATUS1);
+ if (unlikely(val < 0))
+ return ERR_PTR(val);
+ data->faults = val;
+
+ /* Read adc registers */
+ for (i = 0; i < SMM665_NUM_ADC; i++) {
+ val = smm665_read_adc(data, i);
+ if (unlikely(val < 0))
+ return ERR_PTR(val);
+ data->adc[i] = val;
+ }
+ data->last_updated = jiffies;
+ data->valid = 1;
+ }
+
+ mutex_unlock(&data->update_lock);
+
+ return data;
+}
+
+/* Return converted value from given adc */
+static int smm665_convert(u16 adcval, int index)
+{
+ int val = 0;
+
+ switch (index) {
+ case SMM665_MISC16_ADC_DATA_12V:
+ val = SMM665_12VIN_ADC_TO_VOLTS(adcval & SMM665_ADC_MASK);
+ break;
+
+ case SMM665_MISC16_ADC_DATA_VDD:
+ case SMM665_MISC16_ADC_DATA_A:
+ case SMM665_MISC16_ADC_DATA_B:
+ case SMM665_MISC16_ADC_DATA_C:
+ case SMM665_MISC16_ADC_DATA_D:
+ case SMM665_MISC16_ADC_DATA_E:
+ case SMM665_MISC16_ADC_DATA_F:
+ val = SMM665_VMON_ADC_TO_VOLTS(adcval & SMM665_ADC_MASK);
+ break;
+
+ case SMM665_MISC16_ADC_DATA_AIN1:
+ case SMM665_MISC16_ADC_DATA_AIN2:
+ val = SMM665_AIN_ADC_TO_VOLTS(adcval & SMM665_ADC_MASK);
+ break;
+
+ case SMM665_MISC16_ADC_DATA_INT_TEMP:
+ val = SMM665_TEMP_ADC_TO_CELSIUS(adcval & SMM665_ADC_MASK);
+ break;
+
+ default:
+ /* If we get here, the developer messed up */
+ WARN_ON_ONCE(1);
+ break;
+ }
+
+ return val;
+}
+
+static int smm665_get_min(struct device *dev, int index)
+{
+ struct i2c_client *client = to_i2c_client(dev);
+ struct smm665_data *data = i2c_get_clientdata(client);
+
+ return data->alarm_min_limit[index];
+}
+
+static int smm665_get_max(struct device *dev, int index)
+{
+ struct i2c_client *client = to_i2c_client(dev);
+ struct smm665_data *data = i2c_get_clientdata(client);
+
+ return data->alarm_max_limit[index];
+}
+
+static int smm665_get_lcrit(struct device *dev, int index)
+{
+ struct i2c_client *client = to_i2c_client(dev);
+ struct smm665_data *data = i2c_get_clientdata(client);
+
+ return data->critical_min_limit[index];
+}
+
+static int smm665_get_crit(struct device *dev, int index)
+{
+ struct i2c_client *client = to_i2c_client(dev);
+ struct smm665_data *data = i2c_get_clientdata(client);
+
+ return data->critical_max_limit[index];
+}
+
+static ssize_t smm665_show_fault(struct device *dev,
+ struct device_attribute *da, char *buf)
+{
+ struct sensor_device_attribute *attr = to_sensor_dev_attr(da);
+ struct smm665_data *data = smm665_update_device(dev);
+ int val = 0;
+
+ if (IS_ERR(data))
+ return PTR_ERR(data);
+
+ if (data->faults & (1 << attr->index))
+ val = 1;
+
+ return snprintf(buf, PAGE_SIZE, "%d\n", val);
+}
+
+static ssize_t smm665_show_input(struct device *dev,
+ struct device_attribute *da, char *buf)
+{
+ struct sensor_device_attribute *attr = to_sensor_dev_attr(da);
+ struct smm665_data *data = smm665_update_device(dev);
+ int adc = attr->index;
+ int val;
+
+ if (IS_ERR(data))
+ return PTR_ERR(data);
+
+ val = smm665_convert(data->adc[adc], adc);
+ return snprintf(buf, PAGE_SIZE, "%d\n", val);
+}
+
+#define SMM665_SHOW(what) \
+ static ssize_t smm665_show_##what(struct device *dev, \
+ struct device_attribute *da, char *buf) \
+{ \
+ struct sensor_device_attribute *attr = to_sensor_dev_attr(da); \
+ const int val = smm665_get_##what(dev, attr->index); \
+ return snprintf(buf, PAGE_SIZE, "%d\n", val); \
+}
+
+SMM665_SHOW(min);
+SMM665_SHOW(max);
+SMM665_SHOW(lcrit);
+SMM665_SHOW(crit);
+
+/* These macros are used below in constructing device attribute objects
+ * for use with sysfs_create_group() to make a sysfs device file
+ * for each register.
+ */
+
+#define SMM665_ATTR(name, type, cmd_idx) \
+ static SENSOR_DEVICE_ATTR(name##_##type, S_IRUGO, \
+ smm665_show_##type, NULL, cmd_idx)
+
+/* Construct a sensor_device_attribute structure for each register */
+
+/* Input voltages */
+SMM665_ATTR(in1, input, SMM665_MISC16_ADC_DATA_12V);
+SMM665_ATTR(in2, input, SMM665_MISC16_ADC_DATA_VDD);
+SMM665_ATTR(in3, input, SMM665_MISC16_ADC_DATA_A);
+SMM665_ATTR(in4, input, SMM665_MISC16_ADC_DATA_B);
+SMM665_ATTR(in5, input, SMM665_MISC16_ADC_DATA_C);
+SMM665_ATTR(in6, input, SMM665_MISC16_ADC_DATA_D);
+SMM665_ATTR(in7, input, SMM665_MISC16_ADC_DATA_E);
+SMM665_ATTR(in8, input, SMM665_MISC16_ADC_DATA_F);
+SMM665_ATTR(in9, input, SMM665_MISC16_ADC_DATA_AIN1);
+SMM665_ATTR(in10, input, SMM665_MISC16_ADC_DATA_AIN2);
+
+/* Input voltages min */
+SMM665_ATTR(in1, min, SMM665_MISC16_ADC_DATA_12V);
+SMM665_ATTR(in2, min, SMM665_MISC16_ADC_DATA_VDD);
+SMM665_ATTR(in3, min, SMM665_MISC16_ADC_DATA_A);
+SMM665_ATTR(in4, min, SMM665_MISC16_ADC_DATA_B);
+SMM665_ATTR(in5, min, SMM665_MISC16_ADC_DATA_C);
+SMM665_ATTR(in6, min, SMM665_MISC16_ADC_DATA_D);
+SMM665_ATTR(in7, min, SMM665_MISC16_ADC_DATA_E);
+SMM665_ATTR(in8, min, SMM665_MISC16_ADC_DATA_F);
+SMM665_ATTR(in9, min, SMM665_MISC16_ADC_DATA_AIN1);
+SMM665_ATTR(in10, min, SMM665_MISC16_ADC_DATA_AIN2);
+
+/* Input voltages max */
+SMM665_ATTR(in1, max, SMM665_MISC16_ADC_DATA_12V);
+SMM665_ATTR(in2, max, SMM665_MISC16_ADC_DATA_VDD);
+SMM665_ATTR(in3, max, SMM665_MISC16_ADC_DATA_A);
+SMM665_ATTR(in4, max, SMM665_MISC16_ADC_DATA_B);
+SMM665_ATTR(in5, max, SMM665_MISC16_ADC_DATA_C);
+SMM665_ATTR(in6, max, SMM665_MISC16_ADC_DATA_D);
+SMM665_ATTR(in7, max, SMM665_MISC16_ADC_DATA_E);
+SMM665_ATTR(in8, max, SMM665_MISC16_ADC_DATA_F);
+SMM665_ATTR(in9, max, SMM665_MISC16_ADC_DATA_AIN1);
+SMM665_ATTR(in10, max, SMM665_MISC16_ADC_DATA_AIN2);
+
+/* Input voltages lcrit */
+SMM665_ATTR(in1, lcrit, SMM665_MISC16_ADC_DATA_12V);
+SMM665_ATTR(in2, lcrit, SMM665_MISC16_ADC_DATA_VDD);
+SMM665_ATTR(in3, lcrit, SMM665_MISC16_ADC_DATA_A);
+SMM665_ATTR(in4, lcrit, SMM665_MISC16_ADC_DATA_B);
+SMM665_ATTR(in5, lcrit, SMM665_MISC16_ADC_DATA_C);
+SMM665_ATTR(in6, lcrit, SMM665_MISC16_ADC_DATA_D);
+SMM665_ATTR(in7, lcrit, SMM665_MISC16_ADC_DATA_E);
+SMM665_ATTR(in8, lcrit, SMM665_MISC16_ADC_DATA_F);
+SMM665_ATTR(in9, lcrit, SMM665_MISC16_ADC_DATA_AIN1);
+SMM665_ATTR(in10, lcrit, SMM665_MISC16_ADC_DATA_AIN2);
+
+/* Input voltages crit */
+SMM665_ATTR(in1, crit, SMM665_MISC16_ADC_DATA_12V);
+SMM665_ATTR(in2, crit, SMM665_MISC16_ADC_DATA_VDD);
+SMM665_ATTR(in3, crit, SMM665_MISC16_ADC_DATA_A);
+SMM665_ATTR(in4, crit, SMM665_MISC16_ADC_DATA_B);
+SMM665_ATTR(in5, crit, SMM665_MISC16_ADC_DATA_C);
+SMM665_ATTR(in6, crit, SMM665_MISC16_ADC_DATA_D);
+SMM665_ATTR(in7, crit, SMM665_MISC16_ADC_DATA_E);
+SMM665_ATTR(in8, crit, SMM665_MISC16_ADC_DATA_F);
+SMM665_ATTR(in9, crit, SMM665_MISC16_ADC_DATA_AIN1);
+SMM665_ATTR(in10, crit, SMM665_MISC16_ADC_DATA_AIN2);
+
+/* Faults */
+SMM665_ATTR(in1, fault, SMM665_FAULT_12V);
+SMM665_ATTR(in2, fault, SMM665_FAULT_VDD);
+SMM665_ATTR(in3, fault, SMM665_FAULT_A);
+SMM665_ATTR(in4, fault, SMM665_FAULT_B);
+SMM665_ATTR(in5, fault, SMM665_FAULT_C);
+SMM665_ATTR(in6, fault, SMM665_FAULT_D);
+SMM665_ATTR(in7, fault, SMM665_FAULT_E);
+SMM665_ATTR(in8, fault, SMM665_FAULT_F);
+SMM665_ATTR(in9, fault, SMM665_FAULT_AIN1);
+SMM665_ATTR(in10, fault, SMM665_FAULT_AIN2);
+
+/* Temperature */
+SMM665_ATTR(temp1, input, SMM665_MISC16_ADC_DATA_INT_TEMP);
+SMM665_ATTR(temp1, min, SMM665_MISC16_ADC_DATA_INT_TEMP);
+SMM665_ATTR(temp1, max, SMM665_MISC16_ADC_DATA_INT_TEMP);
+SMM665_ATTR(temp1, lcrit, SMM665_MISC16_ADC_DATA_INT_TEMP);
+SMM665_ATTR(temp1, crit, SMM665_MISC16_ADC_DATA_INT_TEMP);
+SMM665_ATTR(temp1, fault, SMM665_FAULT_TEMP);
+
+/*
+ * Finally, construct an array of pointers to members of the above objects,
+ * as required for sysfs_create_group()
+ */
+static struct attribute *smm665_attributes[] = {
+ &sensor_dev_attr_in1_input.dev_attr.attr,
+ &sensor_dev_attr_in1_min.dev_attr.attr,
+ &sensor_dev_attr_in1_max.dev_attr.attr,
+ &sensor_dev_attr_in1_lcrit.dev_attr.attr,
+ &sensor_dev_attr_in1_crit.dev_attr.attr,
+ &sensor_dev_attr_in1_fault.dev_attr.attr,
+
+ &sensor_dev_attr_in2_input.dev_attr.attr,
+ &sensor_dev_attr_in2_min.dev_attr.attr,
+ &sensor_dev_attr_in2_max.dev_attr.attr,
+ &sensor_dev_attr_in2_lcrit.dev_attr.attr,
+ &sensor_dev_attr_in2_crit.dev_attr.attr,
+ &sensor_dev_attr_in2_fault.dev_attr.attr,
+
+ &sensor_dev_attr_in3_input.dev_attr.attr,
+ &sensor_dev_attr_in3_min.dev_attr.attr,
+ &sensor_dev_attr_in3_max.dev_attr.attr,
+ &sensor_dev_attr_in3_lcrit.dev_attr.attr,
+ &sensor_dev_attr_in3_crit.dev_attr.attr,
+ &sensor_dev_attr_in3_fault.dev_attr.attr,
+
+ &sensor_dev_attr_in4_input.dev_attr.attr,
+ &sensor_dev_attr_in4_min.dev_attr.attr,
+ &sensor_dev_attr_in4_max.dev_attr.attr,
+ &sensor_dev_attr_in4_lcrit.dev_attr.attr,
+ &sensor_dev_attr_in4_crit.dev_attr.attr,
+ &sensor_dev_attr_in4_fault.dev_attr.attr,
+
+ &sensor_dev_attr_in5_input.dev_attr.attr,
+ &sensor_dev_attr_in5_min.dev_attr.attr,
+ &sensor_dev_attr_in5_max.dev_attr.attr,
+ &sensor_dev_attr_in5_lcrit.dev_attr.attr,
+ &sensor_dev_attr_in5_crit.dev_attr.attr,
+ &sensor_dev_attr_in5_fault.dev_attr.attr,
+
+ &sensor_dev_attr_in6_input.dev_attr.attr,
+ &sensor_dev_attr_in6_min.dev_attr.attr,
+ &sensor_dev_attr_in6_max.dev_attr.attr,
+ &sensor_dev_attr_in6_lcrit.dev_attr.attr,
+ &sensor_dev_attr_in6_crit.dev_attr.attr,
+ &sensor_dev_attr_in6_fault.dev_attr.attr,
+
+ &sensor_dev_attr_in7_input.dev_attr.attr,
+ &sensor_dev_attr_in7_min.dev_attr.attr,
+ &sensor_dev_attr_in7_max.dev_attr.attr,
+ &sensor_dev_attr_in7_lcrit.dev_attr.attr,
+ &sensor_dev_attr_in7_crit.dev_attr.attr,
+ &sensor_dev_attr_in7_fault.dev_attr.attr,
+
+ &sensor_dev_attr_in8_input.dev_attr.attr,
+ &sensor_dev_attr_in8_min.dev_attr.attr,
+ &sensor_dev_attr_in8_max.dev_attr.attr,
+ &sensor_dev_attr_in8_lcrit.dev_attr.attr,
+ &sensor_dev_attr_in8_crit.dev_attr.attr,
+ &sensor_dev_attr_in8_fault.dev_attr.attr,
+
+ &sensor_dev_attr_in9_input.dev_attr.attr,
+ &sensor_dev_attr_in9_min.dev_attr.attr,
+ &sensor_dev_attr_in9_max.dev_attr.attr,
+ &sensor_dev_attr_in9_lcrit.dev_attr.attr,
+ &sensor_dev_attr_in9_crit.dev_attr.attr,
+ &sensor_dev_attr_in9_fault.dev_attr.attr,
+
+ &sensor_dev_attr_in10_input.dev_attr.attr,
+ &sensor_dev_attr_in10_min.dev_attr.attr,
+ &sensor_dev_attr_in10_max.dev_attr.attr,
+ &sensor_dev_attr_in10_lcrit.dev_attr.attr,
+ &sensor_dev_attr_in10_crit.dev_attr.attr,
+ &sensor_dev_attr_in10_fault.dev_attr.attr,
+
+ &sensor_dev_attr_temp1_input.dev_attr.attr,
+ &sensor_dev_attr_temp1_min.dev_attr.attr,
+ &sensor_dev_attr_temp1_max.dev_attr.attr,
+ &sensor_dev_attr_temp1_lcrit.dev_attr.attr,
+ &sensor_dev_attr_temp1_crit.dev_attr.attr,
+ &sensor_dev_attr_temp1_fault.dev_attr.attr,
+
+ NULL,
+};
+
+static const struct attribute_group smm665_group = {
+ .attrs = smm665_attributes,
+};
+
+static int smm665_probe(struct i2c_client *client,
+ const struct i2c_device_id *id)
+{
+ struct i2c_adapter *adapter = client->adapter;
+ struct smm665_data *data;
+ int i, ret;
+
+ if (!i2c_check_functionality(adapter, I2C_FUNC_SMBUS_BYTE_DATA
+ | I2C_FUNC_SMBUS_WORD_DATA))
+ return -ENODEV;
+
+ if (i2c_smbus_read_byte_data(client, SMM665_ADOC_ENABLE) < 0)
+ return -ENODEV;
+
+ ret = -ENOMEM;
+ data = kzalloc(sizeof(*data), GFP_KERNEL);
+ if (!data)
+ goto out_return;
+
+ i2c_set_clientdata(client, data);
+ mutex_init(&data->update_lock);
+
+ data->type = id->driver_data;
+ data->cmdreg = i2c_new_dummy(adapter, (client->addr & ~SMM665_REGMASK)
+ | SMM665_CMDREG_BASE);
+ if (!data->cmdreg)
+ goto out_kfree;
+
+ switch (data->type) {
+ case smm465:
+ case smm665:
+ data->conversion_time = SMM665_ADC_WAIT_SMM665;
+ break;
+ case smm665c:
+ case smm764:
+ case smm766:
+ data->conversion_time = SMM665_ADC_WAIT_SMM766;
+ break;
+ }
+
+ ret = -ENODEV;
+ if (i2c_smbus_read_byte_data(data->cmdreg, SMM665_MISC8_CMD_STS) < 0)
+ goto out_unregister;
+
+ /*
+ * Read limits.
+ *
+ * Limit registers start with register SMM665_LIMIT_BASE.
+ * Each channel uses 8 registers, providing four limit values
+ * per channel. Each limit value requires two registers, with the
+ * high byte in the first register and the low byte in the second
+ * register. The first two limits are under limit values, followed
+ * by two over limit values.
+ *
+ * Limit register order matches the ADC register order, so we use
+ * ADC register defines throughout the code to index limit registers.
+ *
+ * We save the first retrieved value both as "critical" and "alarm"
+ * value. The second value overwrites either the critical or the
+ * alarm value, depending on its configuration. This ensures that both
+ * critical and alarm values are initialized, even if both registers are
+ * configured as critical or non-critical.
+ *
+ * Note: Critical values for voltage channels are saved, even though
+ * this information is currently not used by the driver. This is mostly
+ * for consistency, though it might eventually be useful if future APIs
+ * support reporting "critical" voltage values.
+ */
+ for (i = 0; i < SMM665_NUM_ADC; i++) {
+ int val;
+
+ val = smm665_read16(client, SMM665_LIMIT_BASE + i * 8);
+ if (unlikely(val < 0))
+ goto out_unregister;
+ data->critical_min_limit[i] = data->alarm_min_limit[i]
+ = smm665_convert(val, i);
+ val = smm665_read16(client, SMM665_LIMIT_BASE + i * 8 + 2);
+ if (unlikely(val < 0))
+ goto out_unregister;
+ if (smm665_is_critical(val))
+ data->critical_min_limit[i] = smm665_convert(val, i);
+ else
+ data->alarm_min_limit[i] = smm665_convert(val, i);
+ val = smm665_read16(client, SMM665_LIMIT_BASE + i * 8 + 4);
+ if (unlikely(val < 0))
+ goto out_unregister;
+ data->critical_max_limit[i] = data->alarm_max_limit[i]
+ = smm665_convert(val, i);
+ val = smm665_read16(client, SMM665_LIMIT_BASE + i * 8 + 6);
+ if (unlikely(val < 0))
+ goto out_unregister;
+ if (smm665_is_critical(val))
+ data->critical_max_limit[i] = smm665_convert(val, i);
+ else
+ data->alarm_max_limit[i] = smm665_convert(val, i);
+ }
+
+ /* Register sysfs hooks */
+ ret = sysfs_create_group(&client->dev.kobj, &smm665_group);
+ if (ret)
+ goto out_unregister;
+
+ data->hwmon_dev = hwmon_device_register(&client->dev);
+ if (IS_ERR(data->hwmon_dev)) {
+ ret = PTR_ERR(data->hwmon_dev);
+ goto out_remove_group;
+ }
+
+ return 0;
+
+out_remove_group:
+ sysfs_remove_group(&client->dev.kobj, &smm665_group);
+out_unregister:
+ i2c_unregister_device(data->cmdreg);
+out_kfree:
+ kfree(data);
+out_return:
+ return ret;
+}
+
+static int smm665_remove(struct i2c_client *client)
+{
+ struct smm665_data *data = i2c_get_clientdata(client);
+
+ i2c_unregister_device(data->cmdreg);
+ hwmon_device_unregister(data->hwmon_dev);
+ sysfs_remove_group(&client->dev.kobj, &smm665_group);
+
+ kfree(data);
+
+ return 0;
+}
+
+static const struct i2c_device_id smm665_id[] = {
+ {"smm465", smm465},
+ {"smm665", smm665},
+ {"smm665c", smm665c},
+ {"smm764", smm764},
+ {"smm766", smm766},
+ {}
+};
+
+MODULE_DEVICE_TABLE(i2c, smm665_id);
+
+/* This is the driver that will be inserted */
+static struct i2c_driver smm665_driver = {
+ .driver = {
+ .name = "smm665",
+ },
+ .probe = smm665_probe,
+ .remove = smm665_remove,
+ .id_table = smm665_id,
+};
+
+static int __init smm665_init(void)
+{
+ return i2c_add_driver(&smm665_driver);
+}
+
+static void __exit smm665_exit(void)
+{
+ i2c_del_driver(&smm665_driver);
+}
+
+MODULE_AUTHOR("Guenter Roeck");
+MODULE_DESCRIPTION("SMM665 driver");
+MODULE_LICENSE("GPL");
+
+module_init(smm665_init);
+module_exit(smm665_exit);
--
1.7.0.87.g0901d

2010-06-24 22:06:12

by Guenter Roeck

[permalink] [raw]
Subject: [PATCH v3 2/4] hwmon: SMM665 driver documentation

Signed-off-by: Guenter Roeck <[email protected]>
---
Documentation/hwmon/smm665 | 126 ++++++++++++++++++++++++++++++++++++++++++++
1 files changed, 126 insertions(+), 0 deletions(-)
create mode 100644 Documentation/hwmon/smm665

diff --git a/Documentation/hwmon/smm665 b/Documentation/hwmon/smm665
new file mode 100644
index 0000000..4a4bfde
--- /dev/null
+++ b/Documentation/hwmon/smm665
@@ -0,0 +1,126 @@
+Kernel driver smm665
+====================
+
+Supported chips:
+ * Summit Microelectronics SMM465
+ Prefix: 'smm465'
+ Addresses scanned: -
+ Datasheet: Publicly available at the Summit Microelectronics website.
+ * Summit Microelectronics SMM665, SMM665B
+ Prefix: 'smm665'
+ Addresses scanned: -
+ Datasheet: Publicly available at the Summit Microelectronics website.
+ * Summit Microelectronics SMM665C
+ Prefix: 'smm665c'
+ Addresses scanned: -
+ Datasheet: Publicly available at the Summit Microelectronics website.
+ * Summit Microelectronics SMM764
+ Prefix: 'smm764'
+ Addresses scanned: -
+ Datasheet: Publicly available at the Summit Microelectronics website.
+ * Summit Microelectronics SMM766
+ Prefix: 'smm766'
+ Addresses scanned: -
+ Datasheet: Publicly available at the Summit Microelectronics website.
+
+Author: Guenter Roeck <[email protected]>
+
+
+Module Parameters
+-----------------
+
+* vref: int
+ Default: 1250 (mV)
+ Reference voltage on VREF_ADC pin in mV. It should not be necessary to set
+ this parameter unless a non-default reference voltage is used.
+
+
+Description
+-----------
+
+[From datasheet] The SMM665 is an Active DC Output power supply Controller
+that monitors, margins and cascade sequences power. The part monitors six
+power supply channels as well as VDD, 12V input, two general-purpose analog
+inputs and an internal temperature sensor using a 10-bit ADC.
+
+Each monitored channel has its own high and low limits, plus a critical
+limit.
+
+Support for SMM465, SMM764, and SMM766 has been implemented but is untested.
+
+
+Usage Notes
+-----------
+
+This driver does not probe for devices, since there is no register which
+can be safely used to identify the chip. You will have to instantiate
+the devices explicitly. When instantiating the device, you have to specify
+its configuration register address.
+
+Example: the following will load the driver for an SMM665 at address 0x57
+on I2C bus #1:
+$ modprobe smm665
+$ echo smm665 0x57 > /sys/bus/i2c/devices/i2c-1/new_device
+
+
+Sysfs entries
+-------------
+
+This driver uses the values in the datasheet to convert ADC register values
+into the values specified in the sysfs-interface document. All attributes are
+read only.
+
+Min and max values are used by the chip to trigger external signals and/or other
+activity. Triggered signals can include HEALTHY, RST, Power Off, or Fault depending
+on the chip configuration. For details please see the SMM665 datasheet.
+
+For SMM465 and SMM764, values for Channel E and F are reported but undefined.
+
+in1_input 12V input voltage (mV)
+in2_input 3.3V (VDD) input voltage (mV)
+in3_input Channel A voltage (mV)
+in4_input Channel B voltage (mV)
+in5_input Channel C voltage (mV)
+in6_input Channel D voltage (mV)
+in7_input Channel E voltage (mV)
+in8_input Channel F voltage (mV)
+in9_input AIN1 voltage (mV)
+in10_input AIN2 voltage (mV)
+
+in1_min 12v input minimum voltage (mV)
+in2_min 3.3V (VDD) input minimum voltage (mV)
+in3_min Channel A minimum voltage (mV)
+in4_min Channel B minimum voltage (mV)
+in5_min Channel C minimum voltage (mV)
+in6_min Channel D minimum voltage (mV)
+in7_min Channel E minimum voltage (mV)
+in8_min Channel F minimum voltage (mV)
+in9_min AIN1 minimum voltage (mV)
+in10_min AIN2 minimum voltage (mV)
+
+in1_max 12v input maximum voltage (mV)
+in2_max 3.3V (VDD) input maximum voltage (mV)
+in3_max Channel A maximum voltage (mV)
+in4_max Channel B maximum voltage (mV)
+in5_max Channel C maximum voltage (mV)
+in6_max Channel D maximum voltage (mV)
+in7_max Channel E maximum voltage (mV)
+in8_max Channel F maximum voltage (mV)
+in9_max AIN1 maximum voltage (mV)
+in10_max AIN2 maximum voltage (mV)
+
+in1_fault 12v input fault
+in2_fault 3.3V (VDD) input fault
+in3_fault Channel A fault
+in4_fault Channel B fault
+in5_fault Channel C fault
+in6_fault Channel D fault
+in7_fault Channel E fault
+in8_fault Channel F fault
+in9_fault AIN1 fault
+in10_fault AIN2 fault
+
+temp1_input Chip tempererature
+temp1_min Mimimum chip tempererature
+temp1_max Maximum chip tempererature
+temp1_crit Critical chip tempererature
--
1.7.0.87.g0901d

2010-06-24 22:06:10

by Guenter Roeck

[permalink] [raw]
Subject: [PATCH v3 3/4] hwmon: Update sysfs-interface to include new attributes

Added _lcrit and _crit to voltage attributes.
Added _lcrit to temperature attributes.

Signed-off-by: Guenter Roeck <[email protected]>
---
Documentation/hwmon/sysfs-interface | 21 ++++++++++++++++++++-
1 files changed, 20 insertions(+), 1 deletions(-)

diff --git a/Documentation/hwmon/sysfs-interface b/Documentation/hwmon/sysfs-interface
index d4e2917..7f60c0d 100644
--- a/Documentation/hwmon/sysfs-interface
+++ b/Documentation/hwmon/sysfs-interface
@@ -107,10 +107,24 @@ in[0-*]_min Voltage min value.
Unit: millivolt
RW

+in[0-*]_lcrit Voltage critical min value.
+ Unit: millivolt
+ RW
+ If voltage drops to or below this limit, the system may
+ take drastic action such as power down or reset. At the very
+ least, it should report a fault.
+
in[0-*]_max Voltage max value.
Unit: millivolt
RW

+in[0-*]_crit Voltage critical max value.
+ Unit: millivolt
+ RW
+ If voltage reaches or exceeds this limit, the system may
+ take drastic action such as power down or reset. At the very
+ least, it should report a fault.
+
in[0-*]_input Voltage input value.
Unit: millivolt
RO
@@ -284,11 +298,16 @@ temp[1-*]_input Temperature input value.
Unit: millidegree Celsius
RO

-temp[1-*]_crit Temperature critical value, typically greater than
+temp[1-*]_crit Temperature critical max value, typically greater than
corresponding temp_max values.
Unit: millidegree Celsius
RW

+temp[1-*]_lcrit Temperature critical min value, typically lower than
+ corresponding temp_min values.
+ Unit: millidegree Celsius
+ RW
+
temp[1-*]_crit_hyst
Temperature hysteresis value for critical limit.
Unit: millidegree Celsius
--
1.7.0.87.g0901d

2010-06-24 23:07:26

by Guenter Roeck

[permalink] [raw]
Subject: Re: [PATCH v3 1/4] hwmon: Driver for SMM665 Six-Channel Active DC Output Controller/Monitor

Couple of comments inline.

On Thu, Jun 24, 2010 at 06:00:58PM -0400, Guenter Roeck wrote:
> This driver adds support for the monitoring features of the Summit
> Microelectronics SMM665 Six-Channel Active DC Output Controller/Monitor.
>
> Signed-off-by: Guenter Roeck <[email protected]>
> ---
> drivers/hwmon/Kconfig | 15 +
> drivers/hwmon/Makefile | 1 +
> drivers/hwmon/smm665.c | 744 ++++++++++++++++++++++++++++++++++++++++++++++++
> 3 files changed, 760 insertions(+), 0 deletions(-)
> create mode 100644 drivers/hwmon/smm665.c
>
> diff --git a/drivers/hwmon/Kconfig b/drivers/hwmon/Kconfig
> index e19cf8e..c45c17e 100644
> --- a/drivers/hwmon/Kconfig
> +++ b/drivers/hwmon/Kconfig
> @@ -739,6 +739,21 @@ config SENSORS_SIS5595
> This driver can also be built as a module. If so, the module
> will be called sis5595.
>
> +config SENSORS_SMM665
> + tristate "Summit Microelectronics SMM665"
> + depends on I2C && EXPERIMENTAL
> + default n
> + help
> + If you say yes here you get support for the hardware monitoring
> + features of the Summit Microelectronics SMM665/SMM665B Six-Channel
> + Active DC Output Controller / Monitor.
> +
> + Other supported chips are SMM465, SMM665C, SMM764, and SMM766.
> + Support for those chips is untested.
> +
> + This driver can also be built as a module. If so, the module will
> + be called smm665.
> +
> config SENSORS_DME1737
> tristate "SMSC DME1737, SCH311x and compatibles"
> depends on I2C && EXPERIMENTAL
> diff --git a/drivers/hwmon/Makefile b/drivers/hwmon/Makefile
> index 2138ceb..55b0940 100644
> --- a/drivers/hwmon/Makefile
> +++ b/drivers/hwmon/Makefile
> @@ -86,6 +86,7 @@ obj-$(CONFIG_SENSORS_PCF8591) += pcf8591.o
> obj-$(CONFIG_SENSORS_S3C) += s3c-hwmon.o
> obj-$(CONFIG_SENSORS_SHT15) += sht15.o
> obj-$(CONFIG_SENSORS_SIS5595) += sis5595.o
> +obj-$(CONFIG_SENSORS_SMM665) += smm665.o
> obj-$(CONFIG_SENSORS_SMSC47B397)+= smsc47b397.o
> obj-$(CONFIG_SENSORS_SMSC47M1) += smsc47m1.o
> obj-$(CONFIG_SENSORS_SMSC47M192)+= smsc47m192.o
> diff --git a/drivers/hwmon/smm665.c b/drivers/hwmon/smm665.c
> new file mode 100644
> index 0000000..d340f61
> --- /dev/null
> +++ b/drivers/hwmon/smm665.c
> @@ -0,0 +1,744 @@
> +/*
> + * Driver for SMM665 Power Controller / Monitor
> + *
> + * Copyright (C) 2010 Ericsson AB.
> + *
> + * This program is free software; you can redistribute it and/or modify
> + * it under the terms of the GNU General Public License as published by
> + * the Free Software Foundation; version 2 of the License.
> + *
> + * This driver should also work for SMM465, SMM764, and SMM766, but is untested
> + * for those chips. Only monitoring functionality is implemented.
> + *
> + * Datasheets:
> + * http://www.summitmicro.com/prod_select/summary/SMM665/SMM665B_2089_20.pdf
> + * http://www.summitmicro.com/prod_select/summary/SMM766B/SMM766B_2122.pdf
> + */
> +
> +#include <linux/kernel.h>
> +#include <linux/module.h>
> +#include <linux/init.h>
> +#include <linux/err.h>
> +#include <linux/slab.h>
> +#include <linux/i2c.h>
> +#include <linux/hwmon.h>
> +#include <linux/hwmon-sysfs.h>
> +#include <linux/delay.h>
> +
> +/* Internal reference voltage (VREF, x 1000 */
> +#define SMM665_VREF_ADC_X1000 1250
> +
> +/* module parameters */
> +static int vref = SMM665_VREF_ADC_X1000;
> +module_param(vref, int, 0);
> +MODULE_PARM_DESC(vref, "Reference voltage in mV");
> +
> +enum chips { smm465, smm665, smm665c, smm764, smm766 };
> +
> +/*
> + * ADC channel addresses
> + */
> +#define SMM665_MISC16_ADC_DATA_A 0x00
> +#define SMM665_MISC16_ADC_DATA_B 0x01
> +#define SMM665_MISC16_ADC_DATA_C 0x02
> +#define SMM665_MISC16_ADC_DATA_D 0x03
> +#define SMM665_MISC16_ADC_DATA_E 0x04
> +#define SMM665_MISC16_ADC_DATA_F 0x05
> +#define SMM665_MISC16_ADC_DATA_VDD 0x06
> +#define SMM665_MISC16_ADC_DATA_12V 0x07
> +#define SMM665_MISC16_ADC_DATA_INT_TEMP 0x08
> +#define SMM665_MISC16_ADC_DATA_AIN1 0x09
> +#define SMM665_MISC16_ADC_DATA_AIN2 0x0a
> +
> +/*
> + * Command registers
> + */
> +#define SMM665_MISC8_CMD_STS 0x80
> +#define SMM665_MISC8_STATUS1 0x81
> +#define SMM665_MISC8_STATUSS2 0x82
> +#define SMM665_MISC8_IO_POLARITY 0x83
> +#define SMM665_MISC8_PUP_POLARITY 0x84
> +#define SMM665_MISC8_ADOC_STATUS1 0x85
> +#define SMM665_MISC8_ADOC_STATUS2 0x86
> +#define SMM665_MISC8_WRITE_PROT 0x87
> +#define SMM665_MISC8_STS_TRACK 0x88
> +
> +/*
> + * Configuration registers and register groups
> + */
> +#define SMM665_ADOC_ENABLE 0x0d
> +#define SMM665_LIMIT_BASE 0x80 /* First limit register */
> +
> +/*
> + * Limit register bit masks
> + */
> +#define SMM665_TRIGGER_RST 0x8000
> +#define SMM665_TRIGGER_HEALTHY 0x4000
> +#define SMM665_TRIGGER_POWEROFF 0x2000
> +#define SMM665_TRIGGER_SHUTDOWN 0x1000
> +#define SMM665_ADC_MASK 0x03ff
> +
> +#define smm665_is_critical(lim) ((lim) & (SMM665_TRIGGER_RST \
> + | SMM665_TRIGGER_POWEROFF \
> + | SMM665_TRIGGER_SHUTDOWN))
> +/*
> + * Fault register bit definitions
> + * Values are merged from status registers 1/2,
> + * with status register 1 providing the upper 8 bits.
> + */
> +#define SMM665_FAULT_A 0x0001
> +#define SMM665_FAULT_B 0x0002
> +#define SMM665_FAULT_C 0x0004
> +#define SMM665_FAULT_D 0x0008
> +#define SMM665_FAULT_E 0x0010
> +#define SMM665_FAULT_F 0x0020
> +#define SMM665_FAULT_VDD 0x0040
> +#define SMM665_FAULT_12V 0x0080
> +#define SMM665_FAULT_TEMP 0x0100
> +#define SMM665_FAULT_AIN1 0x0200
> +#define SMM665_FAULT_AIN2 0x0400
> +
> +/*
> + * I2C Register addresses
> + *
> + * The configuration register needs to be the configured base register.
> + * The command/status register address is derived from it.
> + */
> +#define SMM665_REGMASK 0x78
> +#define SMM665_CMDREG_BASE 0x48
> +#define SMM665_CONFREG_BASE 0x50
> +
> +/*
> + * Equations given by chip manufacturer to calculate voltage/temperature values
> + * vref = Reference voltage on VREF_ADC pin (module parameter)
> + * adc = 10bit ADC value read back from registers
> + */
> +
> +/* Voltage A-F and VDD */
> +#define SMM665_VMON_ADC_TO_VOLTS(adc) ((adc) * vref / 256)
> +
> +/* Voltage 12VIN */
> +#define SMM665_12VIN_ADC_TO_VOLTS(adc) ((adc) * vref * 3 / 256)
> +
> +/* Voltage AIN1, AIN2 */
> +#define SMM665_AIN_ADC_TO_VOLTS(adc) ((adc) * vref / 512)
> +
> +/* Temp Sensor */
> +#define SMM665_TEMP_ADC_TO_CELSIUS(adc) ((adc) <= 511) ? \
> + ((int)(adc) * 1000 / 4) : \
> + (((int)(adc) - 0x400) * 1000 / 4)
> +
> +#define SMM665_NUM_ADC 11
> +
> +/*
> + * Chip dependent ADC conversion time, in uS
> + */
> +#define SMM665_ADC_WAIT_SMM665 70
> +#define SMM665_ADC_WAIT_SMM766 185
> +
> +struct smm665_data {
> + enum chips type;
> + int conversion_time; /* ADC conversion time */
> + struct device *hwmon_dev;
> + struct mutex update_lock;
> + bool valid;
> + unsigned long last_updated; /* in jiffies */
> + u16 adc[SMM665_NUM_ADC]; /* adc values (raw) */
> + u16 faults; /* fault status */
> + /* The following values are in mV */
> + int critical_min_limit[SMM665_NUM_ADC];
> + int alarm_min_limit[SMM665_NUM_ADC];
> + int critical_max_limit[SMM665_NUM_ADC];
> + int alarm_max_limit[SMM665_NUM_ADC];
> + struct i2c_client *cmdreg;
> +};
> +
> +/*
> + * smm665_read16()
> + *
> + * Read 16 bit value from <reg>, <reg+1>. Upper 8 bits are in <reg>.
> + */
> +static int smm665_read16(struct i2c_client *client, int reg)
> +{
> + int rv, val;
> +
> + rv = i2c_smbus_read_byte_data(client, reg);
> + if (rv < 0)
> + return rv;
> + val = rv << 8;
> + rv = i2c_smbus_read_byte_data(client, reg + 1);
> + if (rv < 0)
> + return rv;
> + val |= rv;
> + return val;
> +}
> +
> +/*
> + * Read adc value.
> + */
> +static int smm665_read_adc(struct smm665_data *data, int adc)
> +{
> + struct i2c_client *client = data->cmdreg;
> + int rv;
> + int radc;
> +
> + /*
> + * Algorithm for reading ADC, per SMM665 datasheet
> + *
> + * {[S][addr][W][Ack]} {[offset][Ack]} {[S][addr][R][Nack]}
> + * [wait conversion time]
> + * {[S][addr][R][Ack]} {[datahi][Ack]} {[datalo][Ack][P]}
> + *
> + * To implement the first part of this exchange,
> + * do a full read transaction and expect a failure/Nack.
> + * This sets up the address pointer on the SMM665
> + * and starts the ADC conversion.
> + * Then do a two-byte read transaction.
> + */
> + rv = i2c_smbus_read_byte_data(client, adc << 3);
> + if (rv != -ENXIO) {
> + /*
> + * We expect ENXIO to reflect NACK
> + * (per Documentation/i2c/fault-codes).
> + * Everything else is an error.
> + */
> + dev_dbg(&client->dev,
> + "Unexpected return code %d when setting ADC index", rv);
> + return (rv < 0) ? rv : -EIO;
> + }
> +
> + udelay(data->conversion_time);
> +
> + /*
> + * Now read two bytes.
> + *
> + * Neither i2c_smbus_read_byte() nor
> + * i2c_smbus_read_block_data() worked here,
> + * so use i2c_smbus_read_word_data() instead.
> + * We could also try to use i2c_master_recv(),
> + * but that is not always supported.
> + */
> + rv = i2c_smbus_read_word_data(client, 0);
> + if (rv < 0) {
> + dev_dbg(&client->dev, "Failed to read ADC value: error %d", rv);
> + return -1;
> + }
> + /*
> + * Validate/verify readback adc channel (in bit 11..14).
> + * High byte is in lower 8 bit of rv, so only shift by 3.
> + */
> + radc = (rv >> 3) & 0x0f;
> + if (radc != adc) {
> + dev_dbg(&client->dev, "Unexpected RADC: Expected %d got %d",
> + adc, radc);
> + return -EIO;
> + }
> + /*
> + * Chip replies with H/L, while SMBus expects L/H.
> + * Thus, byte order is reversed, and we have to swap
> + * the result.
> + */
> + rv = swab16(rv) & SMM665_ADC_MASK;
> +
> + return rv;
> +}
> +
> +static struct smm665_data *smm665_update_device(struct device *dev)
> +{
> + struct i2c_client *client = to_i2c_client(dev);
> + struct smm665_data *data = i2c_get_clientdata(client);
> +
> + mutex_lock(&data->update_lock);
> +
> + if (time_after(jiffies, data->last_updated + HZ) || !data->valid) {
> + int i, val;
> +
> + /*
> + * read status registers
> + */
> + val = smm665_read16(client, SMM665_MISC8_STATUS1);
> + if (unlikely(val < 0))
> + return ERR_PTR(val);

unlock() missing. Will be fixed in v4.

> + data->faults = val;
> +
> + /* Read adc registers */
> + for (i = 0; i < SMM665_NUM_ADC; i++) {
> + val = smm665_read_adc(data, i);
> + if (unlikely(val < 0))
> + return ERR_PTR(val);

unlock() missing. Will be fixed in v4.

> + data->adc[i] = val;
> + }
> + data->last_updated = jiffies;
> + data->valid = 1;
> + }
> +
> + mutex_unlock(&data->update_lock);
> +
> + return data;
> +}
> +
> +/* Return converted value from given adc */
> +static int smm665_convert(u16 adcval, int index)
> +{
> + int val = 0;
> +
> + switch (index) {
> + case SMM665_MISC16_ADC_DATA_12V:
> + val = SMM665_12VIN_ADC_TO_VOLTS(adcval & SMM665_ADC_MASK);
> + break;
> +
> + case SMM665_MISC16_ADC_DATA_VDD:
> + case SMM665_MISC16_ADC_DATA_A:
> + case SMM665_MISC16_ADC_DATA_B:
> + case SMM665_MISC16_ADC_DATA_C:
> + case SMM665_MISC16_ADC_DATA_D:
> + case SMM665_MISC16_ADC_DATA_E:
> + case SMM665_MISC16_ADC_DATA_F:
> + val = SMM665_VMON_ADC_TO_VOLTS(adcval & SMM665_ADC_MASK);
> + break;
> +
> + case SMM665_MISC16_ADC_DATA_AIN1:
> + case SMM665_MISC16_ADC_DATA_AIN2:
> + val = SMM665_AIN_ADC_TO_VOLTS(adcval & SMM665_ADC_MASK);
> + break;
> +
> + case SMM665_MISC16_ADC_DATA_INT_TEMP:
> + val = SMM665_TEMP_ADC_TO_CELSIUS(adcval & SMM665_ADC_MASK);
> + break;
> +
> + default:
> + /* If we get here, the developer messed up */
> + WARN_ON_ONCE(1);
> + break;
> + }
> +
> + return val;
> +}
> +
> +static int smm665_get_min(struct device *dev, int index)
> +{
> + struct i2c_client *client = to_i2c_client(dev);
> + struct smm665_data *data = i2c_get_clientdata(client);
> +
> + return data->alarm_min_limit[index];
> +}
> +
> +static int smm665_get_max(struct device *dev, int index)
> +{
> + struct i2c_client *client = to_i2c_client(dev);
> + struct smm665_data *data = i2c_get_clientdata(client);
> +
> + return data->alarm_max_limit[index];
> +}
> +
> +static int smm665_get_lcrit(struct device *dev, int index)
> +{
> + struct i2c_client *client = to_i2c_client(dev);
> + struct smm665_data *data = i2c_get_clientdata(client);
> +
> + return data->critical_min_limit[index];
> +}
> +
> +static int smm665_get_crit(struct device *dev, int index)
> +{
> + struct i2c_client *client = to_i2c_client(dev);
> + struct smm665_data *data = i2c_get_clientdata(client);
> +
> + return data->critical_max_limit[index];
> +}
> +
> +static ssize_t smm665_show_fault(struct device *dev,
> + struct device_attribute *da, char *buf)
> +{
> + struct sensor_device_attribute *attr = to_sensor_dev_attr(da);
> + struct smm665_data *data = smm665_update_device(dev);
> + int val = 0;
> +
> + if (IS_ERR(data))
> + return PTR_ERR(data);
> +
> + if (data->faults & (1 << attr->index))
> + val = 1;
> +
> + return snprintf(buf, PAGE_SIZE, "%d\n", val);
> +}
> +
> +static ssize_t smm665_show_input(struct device *dev,
> + struct device_attribute *da, char *buf)
> +{
> + struct sensor_device_attribute *attr = to_sensor_dev_attr(da);
> + struct smm665_data *data = smm665_update_device(dev);
> + int adc = attr->index;
> + int val;
> +
> + if (IS_ERR(data))
> + return PTR_ERR(data);
> +
> + val = smm665_convert(data->adc[adc], adc);
> + return snprintf(buf, PAGE_SIZE, "%d\n", val);
> +}
> +
> +#define SMM665_SHOW(what) \
> + static ssize_t smm665_show_##what(struct device *dev, \
> + struct device_attribute *da, char *buf) \
> +{ \
> + struct sensor_device_attribute *attr = to_sensor_dev_attr(da); \
> + const int val = smm665_get_##what(dev, attr->index); \
> + return snprintf(buf, PAGE_SIZE, "%d\n", val); \
> +}
> +
> +SMM665_SHOW(min);
> +SMM665_SHOW(max);
> +SMM665_SHOW(lcrit);
> +SMM665_SHOW(crit);
> +
> +/* These macros are used below in constructing device attribute objects
> + * for use with sysfs_create_group() to make a sysfs device file
> + * for each register.
> + */
> +
> +#define SMM665_ATTR(name, type, cmd_idx) \
> + static SENSOR_DEVICE_ATTR(name##_##type, S_IRUGO, \
> + smm665_show_##type, NULL, cmd_idx)
> +
> +/* Construct a sensor_device_attribute structure for each register */
> +
> +/* Input voltages */
> +SMM665_ATTR(in1, input, SMM665_MISC16_ADC_DATA_12V);
> +SMM665_ATTR(in2, input, SMM665_MISC16_ADC_DATA_VDD);
> +SMM665_ATTR(in3, input, SMM665_MISC16_ADC_DATA_A);
> +SMM665_ATTR(in4, input, SMM665_MISC16_ADC_DATA_B);
> +SMM665_ATTR(in5, input, SMM665_MISC16_ADC_DATA_C);
> +SMM665_ATTR(in6, input, SMM665_MISC16_ADC_DATA_D);
> +SMM665_ATTR(in7, input, SMM665_MISC16_ADC_DATA_E);
> +SMM665_ATTR(in8, input, SMM665_MISC16_ADC_DATA_F);
> +SMM665_ATTR(in9, input, SMM665_MISC16_ADC_DATA_AIN1);
> +SMM665_ATTR(in10, input, SMM665_MISC16_ADC_DATA_AIN2);
> +
> +/* Input voltages min */
> +SMM665_ATTR(in1, min, SMM665_MISC16_ADC_DATA_12V);
> +SMM665_ATTR(in2, min, SMM665_MISC16_ADC_DATA_VDD);
> +SMM665_ATTR(in3, min, SMM665_MISC16_ADC_DATA_A);
> +SMM665_ATTR(in4, min, SMM665_MISC16_ADC_DATA_B);
> +SMM665_ATTR(in5, min, SMM665_MISC16_ADC_DATA_C);
> +SMM665_ATTR(in6, min, SMM665_MISC16_ADC_DATA_D);
> +SMM665_ATTR(in7, min, SMM665_MISC16_ADC_DATA_E);
> +SMM665_ATTR(in8, min, SMM665_MISC16_ADC_DATA_F);
> +SMM665_ATTR(in9, min, SMM665_MISC16_ADC_DATA_AIN1);
> +SMM665_ATTR(in10, min, SMM665_MISC16_ADC_DATA_AIN2);
> +
> +/* Input voltages max */
> +SMM665_ATTR(in1, max, SMM665_MISC16_ADC_DATA_12V);
> +SMM665_ATTR(in2, max, SMM665_MISC16_ADC_DATA_VDD);
> +SMM665_ATTR(in3, max, SMM665_MISC16_ADC_DATA_A);
> +SMM665_ATTR(in4, max, SMM665_MISC16_ADC_DATA_B);
> +SMM665_ATTR(in5, max, SMM665_MISC16_ADC_DATA_C);
> +SMM665_ATTR(in6, max, SMM665_MISC16_ADC_DATA_D);
> +SMM665_ATTR(in7, max, SMM665_MISC16_ADC_DATA_E);
> +SMM665_ATTR(in8, max, SMM665_MISC16_ADC_DATA_F);
> +SMM665_ATTR(in9, max, SMM665_MISC16_ADC_DATA_AIN1);
> +SMM665_ATTR(in10, max, SMM665_MISC16_ADC_DATA_AIN2);
> +
> +/* Input voltages lcrit */
> +SMM665_ATTR(in1, lcrit, SMM665_MISC16_ADC_DATA_12V);
> +SMM665_ATTR(in2, lcrit, SMM665_MISC16_ADC_DATA_VDD);
> +SMM665_ATTR(in3, lcrit, SMM665_MISC16_ADC_DATA_A);
> +SMM665_ATTR(in4, lcrit, SMM665_MISC16_ADC_DATA_B);
> +SMM665_ATTR(in5, lcrit, SMM665_MISC16_ADC_DATA_C);
> +SMM665_ATTR(in6, lcrit, SMM665_MISC16_ADC_DATA_D);
> +SMM665_ATTR(in7, lcrit, SMM665_MISC16_ADC_DATA_E);
> +SMM665_ATTR(in8, lcrit, SMM665_MISC16_ADC_DATA_F);
> +SMM665_ATTR(in9, lcrit, SMM665_MISC16_ADC_DATA_AIN1);
> +SMM665_ATTR(in10, lcrit, SMM665_MISC16_ADC_DATA_AIN2);
> +
> +/* Input voltages crit */
> +SMM665_ATTR(in1, crit, SMM665_MISC16_ADC_DATA_12V);
> +SMM665_ATTR(in2, crit, SMM665_MISC16_ADC_DATA_VDD);
> +SMM665_ATTR(in3, crit, SMM665_MISC16_ADC_DATA_A);
> +SMM665_ATTR(in4, crit, SMM665_MISC16_ADC_DATA_B);
> +SMM665_ATTR(in5, crit, SMM665_MISC16_ADC_DATA_C);
> +SMM665_ATTR(in6, crit, SMM665_MISC16_ADC_DATA_D);
> +SMM665_ATTR(in7, crit, SMM665_MISC16_ADC_DATA_E);
> +SMM665_ATTR(in8, crit, SMM665_MISC16_ADC_DATA_F);
> +SMM665_ATTR(in9, crit, SMM665_MISC16_ADC_DATA_AIN1);
> +SMM665_ATTR(in10, crit, SMM665_MISC16_ADC_DATA_AIN2);
> +
> +/* Faults */
> +SMM665_ATTR(in1, fault, SMM665_FAULT_12V);
> +SMM665_ATTR(in2, fault, SMM665_FAULT_VDD);
> +SMM665_ATTR(in3, fault, SMM665_FAULT_A);
> +SMM665_ATTR(in4, fault, SMM665_FAULT_B);
> +SMM665_ATTR(in5, fault, SMM665_FAULT_C);
> +SMM665_ATTR(in6, fault, SMM665_FAULT_D);
> +SMM665_ATTR(in7, fault, SMM665_FAULT_E);
> +SMM665_ATTR(in8, fault, SMM665_FAULT_F);
> +SMM665_ATTR(in9, fault, SMM665_FAULT_AIN1);
> +SMM665_ATTR(in10, fault, SMM665_FAULT_AIN2);
> +
> +/* Temperature */
> +SMM665_ATTR(temp1, input, SMM665_MISC16_ADC_DATA_INT_TEMP);
> +SMM665_ATTR(temp1, min, SMM665_MISC16_ADC_DATA_INT_TEMP);
> +SMM665_ATTR(temp1, max, SMM665_MISC16_ADC_DATA_INT_TEMP);
> +SMM665_ATTR(temp1, lcrit, SMM665_MISC16_ADC_DATA_INT_TEMP);
> +SMM665_ATTR(temp1, crit, SMM665_MISC16_ADC_DATA_INT_TEMP);
> +SMM665_ATTR(temp1, fault, SMM665_FAULT_TEMP);
> +
> +/*
> + * Finally, construct an array of pointers to members of the above objects,
> + * as required for sysfs_create_group()
> + */
> +static struct attribute *smm665_attributes[] = {
> + &sensor_dev_attr_in1_input.dev_attr.attr,
> + &sensor_dev_attr_in1_min.dev_attr.attr,
> + &sensor_dev_attr_in1_max.dev_attr.attr,
> + &sensor_dev_attr_in1_lcrit.dev_attr.attr,
> + &sensor_dev_attr_in1_crit.dev_attr.attr,
> + &sensor_dev_attr_in1_fault.dev_attr.attr,
> +
> + &sensor_dev_attr_in2_input.dev_attr.attr,
> + &sensor_dev_attr_in2_min.dev_attr.attr,
> + &sensor_dev_attr_in2_max.dev_attr.attr,
> + &sensor_dev_attr_in2_lcrit.dev_attr.attr,
> + &sensor_dev_attr_in2_crit.dev_attr.attr,
> + &sensor_dev_attr_in2_fault.dev_attr.attr,
> +
> + &sensor_dev_attr_in3_input.dev_attr.attr,
> + &sensor_dev_attr_in3_min.dev_attr.attr,
> + &sensor_dev_attr_in3_max.dev_attr.attr,
> + &sensor_dev_attr_in3_lcrit.dev_attr.attr,
> + &sensor_dev_attr_in3_crit.dev_attr.attr,
> + &sensor_dev_attr_in3_fault.dev_attr.attr,
> +
> + &sensor_dev_attr_in4_input.dev_attr.attr,
> + &sensor_dev_attr_in4_min.dev_attr.attr,
> + &sensor_dev_attr_in4_max.dev_attr.attr,
> + &sensor_dev_attr_in4_lcrit.dev_attr.attr,
> + &sensor_dev_attr_in4_crit.dev_attr.attr,
> + &sensor_dev_attr_in4_fault.dev_attr.attr,
> +
> + &sensor_dev_attr_in5_input.dev_attr.attr,
> + &sensor_dev_attr_in5_min.dev_attr.attr,
> + &sensor_dev_attr_in5_max.dev_attr.attr,
> + &sensor_dev_attr_in5_lcrit.dev_attr.attr,
> + &sensor_dev_attr_in5_crit.dev_attr.attr,
> + &sensor_dev_attr_in5_fault.dev_attr.attr,
> +
> + &sensor_dev_attr_in6_input.dev_attr.attr,
> + &sensor_dev_attr_in6_min.dev_attr.attr,
> + &sensor_dev_attr_in6_max.dev_attr.attr,
> + &sensor_dev_attr_in6_lcrit.dev_attr.attr,
> + &sensor_dev_attr_in6_crit.dev_attr.attr,
> + &sensor_dev_attr_in6_fault.dev_attr.attr,
> +
> + &sensor_dev_attr_in7_input.dev_attr.attr,
> + &sensor_dev_attr_in7_min.dev_attr.attr,
> + &sensor_dev_attr_in7_max.dev_attr.attr,
> + &sensor_dev_attr_in7_lcrit.dev_attr.attr,
> + &sensor_dev_attr_in7_crit.dev_attr.attr,
> + &sensor_dev_attr_in7_fault.dev_attr.attr,
> +
> + &sensor_dev_attr_in8_input.dev_attr.attr,
> + &sensor_dev_attr_in8_min.dev_attr.attr,
> + &sensor_dev_attr_in8_max.dev_attr.attr,
> + &sensor_dev_attr_in8_lcrit.dev_attr.attr,
> + &sensor_dev_attr_in8_crit.dev_attr.attr,
> + &sensor_dev_attr_in8_fault.dev_attr.attr,
> +
> + &sensor_dev_attr_in9_input.dev_attr.attr,
> + &sensor_dev_attr_in9_min.dev_attr.attr,
> + &sensor_dev_attr_in9_max.dev_attr.attr,
> + &sensor_dev_attr_in9_lcrit.dev_attr.attr,
> + &sensor_dev_attr_in9_crit.dev_attr.attr,
> + &sensor_dev_attr_in9_fault.dev_attr.attr,
> +
> + &sensor_dev_attr_in10_input.dev_attr.attr,
> + &sensor_dev_attr_in10_min.dev_attr.attr,
> + &sensor_dev_attr_in10_max.dev_attr.attr,
> + &sensor_dev_attr_in10_lcrit.dev_attr.attr,
> + &sensor_dev_attr_in10_crit.dev_attr.attr,
> + &sensor_dev_attr_in10_fault.dev_attr.attr,
> +
> + &sensor_dev_attr_temp1_input.dev_attr.attr,
> + &sensor_dev_attr_temp1_min.dev_attr.attr,
> + &sensor_dev_attr_temp1_max.dev_attr.attr,
> + &sensor_dev_attr_temp1_lcrit.dev_attr.attr,
> + &sensor_dev_attr_temp1_crit.dev_attr.attr,
> + &sensor_dev_attr_temp1_fault.dev_attr.attr,
> +
> + NULL,
> +};
> +
> +static const struct attribute_group smm665_group = {
> + .attrs = smm665_attributes,
> +};
> +
> +static int smm665_probe(struct i2c_client *client,
> + const struct i2c_device_id *id)
> +{
> + struct i2c_adapter *adapter = client->adapter;
> + struct smm665_data *data;
> + int i, ret;
> +
> + if (!i2c_check_functionality(adapter, I2C_FUNC_SMBUS_BYTE_DATA
> + | I2C_FUNC_SMBUS_WORD_DATA))
> + return -ENODEV;
> +
> + if (i2c_smbus_read_byte_data(client, SMM665_ADOC_ENABLE) < 0)
> + return -ENODEV;
> +
> + ret = -ENOMEM;
> + data = kzalloc(sizeof(*data), GFP_KERNEL);
> + if (!data)
> + goto out_return;
> +
> + i2c_set_clientdata(client, data);
> + mutex_init(&data->update_lock);
> +
> + data->type = id->driver_data;
> + data->cmdreg = i2c_new_dummy(adapter, (client->addr & ~SMM665_REGMASK)
> + | SMM665_CMDREG_BASE);
> + if (!data->cmdreg)
> + goto out_kfree;
> +
> + switch (data->type) {
> + case smm465:
> + case smm665:
> + data->conversion_time = SMM665_ADC_WAIT_SMM665;
> + break;
> + case smm665c:
> + case smm764:
> + case smm766:
> + data->conversion_time = SMM665_ADC_WAIT_SMM766;
> + break;
> + }
> +
> + ret = -ENODEV;
> + if (i2c_smbus_read_byte_data(data->cmdreg, SMM665_MISC8_CMD_STS) < 0)
> + goto out_unregister;
> +
> + /*
> + * Read limits.
> + *
> + * Limit registers start with register SMM665_LIMIT_BASE.
> + * Each channel uses 8 registers, providing four limit values
> + * per channel. Each limit value requires two registers, with the
> + * high byte in the first register and the low byte in the second
> + * register. The first two limits are under limit values, followed
> + * by two over limit values.
> + *
> + * Limit register order matches the ADC register order, so we use
> + * ADC register defines throughout the code to index limit registers.
> + *
> + * We save the first retrieved value both as "critical" and "alarm"
> + * value. The second value overwrites either the critical or the
> + * alarm value, depending on its configuration. This ensures that both
> + * critical and alarm values are initialized, even if both registers are
> + * configured as critical or non-critical.
> + *
> + * Note: Critical values for voltage channels are saved, even though
> + * this information is currently not used by the driver. This is mostly
> + * for consistency, though it might eventually be useful if future APIs
> + * support reporting "critical" voltage values.
> + */
> + for (i = 0; i < SMM665_NUM_ADC; i++) {
> + int val;
> +
> + val = smm665_read16(client, SMM665_LIMIT_BASE + i * 8);
> + if (unlikely(val < 0))
> + goto out_unregister;
> + data->critical_min_limit[i] = data->alarm_min_limit[i]
> + = smm665_convert(val, i);
> + val = smm665_read16(client, SMM665_LIMIT_BASE + i * 8 + 2);
> + if (unlikely(val < 0))
> + goto out_unregister;
> + if (smm665_is_critical(val))
> + data->critical_min_limit[i] = smm665_convert(val, i);
> + else
> + data->alarm_min_limit[i] = smm665_convert(val, i);
> + val = smm665_read16(client, SMM665_LIMIT_BASE + i * 8 + 4);
> + if (unlikely(val < 0))
> + goto out_unregister;
> + data->critical_max_limit[i] = data->alarm_max_limit[i]
> + = smm665_convert(val, i);
> + val = smm665_read16(client, SMM665_LIMIT_BASE + i * 8 + 6);
> + if (unlikely(val < 0))
> + goto out_unregister;
> + if (smm665_is_critical(val))
> + data->critical_max_limit[i] = smm665_convert(val, i);
> + else
> + data->alarm_max_limit[i] = smm665_convert(val, i);
> + }
> +
> + /* Register sysfs hooks */
> + ret = sysfs_create_group(&client->dev.kobj, &smm665_group);
> + if (ret)
> + goto out_unregister;
> +
> + data->hwmon_dev = hwmon_device_register(&client->dev);
> + if (IS_ERR(data->hwmon_dev)) {
> + ret = PTR_ERR(data->hwmon_dev);
> + goto out_remove_group;
> + }
> +
> + return 0;
> +
> +out_remove_group:
> + sysfs_remove_group(&client->dev.kobj, &smm665_group);
> +out_unregister:
> + i2c_unregister_device(data->cmdreg);
> +out_kfree:
> + kfree(data);
> +out_return:
> + return ret;
> +}
> +
> +static int smm665_remove(struct i2c_client *client)
> +{
> + struct smm665_data *data = i2c_get_clientdata(client);
> +
> + i2c_unregister_device(data->cmdreg);
> + hwmon_device_unregister(data->hwmon_dev);
> + sysfs_remove_group(&client->dev.kobj, &smm665_group);
> +
> + kfree(data);
> +
> + return 0;
> +}
> +
> +static const struct i2c_device_id smm665_id[] = {
> + {"smm465", smm465},
> + {"smm665", smm665},
> + {"smm665c", smm665c},
> + {"smm764", smm764},
> + {"smm766", smm766},
> + {}
> +};
> +
> +MODULE_DEVICE_TABLE(i2c, smm665_id);
> +
> +/* This is the driver that will be inserted */
> +static struct i2c_driver smm665_driver = {
> + .driver = {
> + .name = "smm665",
> + },
> + .probe = smm665_probe,
> + .remove = smm665_remove,
> + .id_table = smm665_id,
> +};
> +
> +static int __init smm665_init(void)
> +{
> + return i2c_add_driver(&smm665_driver);
> +}
> +
> +static void __exit smm665_exit(void)
> +{
> + i2c_del_driver(&smm665_driver);
> +}
> +
> +MODULE_AUTHOR("Guenter Roeck");
> +MODULE_DESCRIPTION("SMM665 driver");
> +MODULE_LICENSE("GPL");
> +
> +module_init(smm665_init);
> +module_exit(smm665_exit);
> --
> 1.7.0.87.g0901d
>

2010-06-24 23:32:44

by Ira W. Snyder

[permalink] [raw]
Subject: Re: [PATCH v3 3/4] hwmon: Update sysfs-interface to include new attributes

On Thu, Jun 24, 2010 at 03:01:00PM -0700, Guenter Roeck wrote:
> Added _lcrit and _crit to voltage attributes.
> Added _lcrit to temperature attributes.
>
> Signed-off-by: Guenter Roeck <[email protected]>
> ---
> Documentation/hwmon/sysfs-interface | 21 ++++++++++++++++++++-
> 1 files changed, 20 insertions(+), 1 deletions(-)
>
> diff --git a/Documentation/hwmon/sysfs-interface b/Documentation/hwmon/sysfs-interface
> index d4e2917..7f60c0d 100644
> --- a/Documentation/hwmon/sysfs-interface
> +++ b/Documentation/hwmon/sysfs-interface
> @@ -107,10 +107,24 @@ in[0-*]_min Voltage min value.
> Unit: millivolt
> RW
>
> +in[0-*]_lcrit Voltage critical min value.
> + Unit: millivolt
> + RW
> + If voltage drops to or below this limit, the system may
> + take drastic action such as power down or reset. At the very
> + least, it should report a fault.
> +
> in[0-*]_max Voltage max value.
> Unit: millivolt
> RW
>
> +in[0-*]_crit Voltage critical max value.
> + Unit: millivolt
> + RW
> + If voltage reaches or exceeds this limit, the system may
> + take drastic action such as power down or reset. At the very
> + least, it should report a fault.
> +

The critical voltages API looks good to me.

I have a hwmon chip driver for the TI INA209 (which Jean NAK'd) that
could use these limits. I'll submit it again soon with changes to use
this new set of sysfs attributes.

It would be great to see a patch to libsensors and the sensors program
to allow it to display these limits too.

Thanks,
Ira

> in[0-*]_input Voltage input value.
> Unit: millivolt
> RO
> @@ -284,11 +298,16 @@ temp[1-*]_input Temperature input value.
> Unit: millidegree Celsius
> RO
>
> -temp[1-*]_crit Temperature critical value, typically greater than
> +temp[1-*]_crit Temperature critical max value, typically greater than
> corresponding temp_max values.
> Unit: millidegree Celsius
> RW
>
> +temp[1-*]_lcrit Temperature critical min value, typically lower than
> + corresponding temp_min values.
> + Unit: millidegree Celsius
> + RW
> +
> temp[1-*]_crit_hyst
> Temperature hysteresis value for critical limit.
> Unit: millidegree Celsius
> --
> 1.7.0.87.g0901d
>

2010-06-24 23:42:39

by Guenter Roeck

[permalink] [raw]
Subject: Re: [PATCH v3 3/4] hwmon: Update sysfs-interface to include new attributes

On Thu, Jun 24, 2010 at 07:32:40PM -0400, Ira W. Snyder wrote:
> On Thu, Jun 24, 2010 at 03:01:00PM -0700, Guenter Roeck wrote:
> > Added _lcrit and _crit to voltage attributes.
> > Added _lcrit to temperature attributes.
> >
> > Signed-off-by: Guenter Roeck <[email protected]>
> > ---
> > Documentation/hwmon/sysfs-interface | 21 ++++++++++++++++++++-
> > 1 files changed, 20 insertions(+), 1 deletions(-)
> >
> > diff --git a/Documentation/hwmon/sysfs-interface b/Documentation/hwmon/sysfs-interface
> > index d4e2917..7f60c0d 100644
> > --- a/Documentation/hwmon/sysfs-interface
> > +++ b/Documentation/hwmon/sysfs-interface
> > @@ -107,10 +107,24 @@ in[0-*]_min Voltage min value.
> > Unit: millivolt
> > RW
> >
> > +in[0-*]_lcrit Voltage critical min value.
> > + Unit: millivolt
> > + RW
> > + If voltage drops to or below this limit, the system may
> > + take drastic action such as power down or reset. At the very
> > + least, it should report a fault.
> > +
> > in[0-*]_max Voltage max value.
> > Unit: millivolt
> > RW
> >
> > +in[0-*]_crit Voltage critical max value.
> > + Unit: millivolt
> > + RW
> > + If voltage reaches or exceeds this limit, the system may
> > + take drastic action such as power down or reset. At the very
> > + least, it should report a fault.
> > +
>
> The critical voltages API looks good to me.
>
> I have a hwmon chip driver for the TI INA209 (which Jean NAK'd) that
> could use these limits. I'll submit it again soon with changes to use
> this new set of sysfs attributes.
>
> It would be great to see a patch to libsensors and the sensors program
> to allow it to display these limits too.
>
I am working on it. I'll send the patch to Jean as soon as the kernel changes
have been approved.

Guenter

> Thanks,
> Ira
>
> > in[0-*]_input Voltage input value.
> > Unit: millivolt
> > RO
> > @@ -284,11 +298,16 @@ temp[1-*]_input Temperature input value.
> > Unit: millidegree Celsius
> > RO
> >
> > -temp[1-*]_crit Temperature critical value, typically greater than
> > +temp[1-*]_crit Temperature critical max value, typically greater than
> > corresponding temp_max values.
> > Unit: millidegree Celsius
> > RW
> >
> > +temp[1-*]_lcrit Temperature critical min value, typically lower than
> > + corresponding temp_min values.
> > + Unit: millidegree Celsius
> > + RW
> > +
> > temp[1-*]_crit_hyst
> > Temperature hysteresis value for critical limit.
> > Unit: millidegree Celsius
> > --
> > 1.7.0.87.g0901d
> >

2010-06-25 08:05:48

by Jean Delvare

[permalink] [raw]
Subject: Re: [PATCH v3 3/4] hwmon: Update sysfs-interface to include new attributes

Hi Guenter,

On Thu, 24 Jun 2010 15:01:00 -0700, Guenter Roeck wrote:
> Added _lcrit and _crit to voltage attributes.
> Added _lcrit to temperature attributes.
>
> Signed-off-by: Guenter Roeck <[email protected]>
> ---
> Documentation/hwmon/sysfs-interface | 21 ++++++++++++++++++++-
> 1 files changed, 20 insertions(+), 1 deletions(-)
>
> diff --git a/Documentation/hwmon/sysfs-interface b/Documentation/hwmon/sysfs-interface
> index d4e2917..7f60c0d 100644
> --- a/Documentation/hwmon/sysfs-interface
> +++ b/Documentation/hwmon/sysfs-interface
> @@ -107,10 +107,24 @@ in[0-*]_min Voltage min value.
> Unit: millivolt
> RW
>
> +in[0-*]_lcrit Voltage critical min value.
> + Unit: millivolt
> + RW
> + If voltage drops to or below this limit, the system may
> + take drastic action such as power down or reset. At the very
> + least, it should report a fault.
> +
> in[0-*]_max Voltage max value.
> Unit: millivolt
> RW
>
> +in[0-*]_crit Voltage critical max value.
> + Unit: millivolt
> + RW
> + If voltage reaches or exceeds this limit, the system may
> + take drastic action such as power down or reset. At the very
> + least, it should report a fault.
> +
> in[0-*]_input Voltage input value.
> Unit: millivolt
> RO
> @@ -284,11 +298,16 @@ temp[1-*]_input Temperature input value.
> Unit: millidegree Celsius
> RO
>
> -temp[1-*]_crit Temperature critical value, typically greater than
> +temp[1-*]_crit Temperature critical max value, typically greater than
> corresponding temp_max values.
> Unit: millidegree Celsius
> RW
>
> +temp[1-*]_lcrit Temperature critical min value, typically lower than
> + corresponding temp_min values.
> + Unit: millidegree Celsius
> + RW
> +

I'd rather have you insert that one _after_ temp[1-*]_crit_hyst, as
temp[1-*]_crit_hyst is rightly related to temp[1-*]_crit.

> temp[1-*]_crit_hyst
> Temperature hysteresis value for critical limit.
> Unit: millidegree Celsius

Other than this minor issue, which I have addresses myself, this patch
looks very good to me and I've applied it.

Thanks,
--
Jean Delvare

2010-06-25 08:09:53

by Jean Delvare

[permalink] [raw]
Subject: Re: [PATCH v3 4/4] hwmon: Update MAINTAINERS for smm665 driver

On Thu, 24 Jun 2010 15:01:01 -0700, Guenter Roeck wrote:
> Signed-off-by: Guenter Roeck <[email protected]>
> ---
> MAINTAINERS | 7 +++++++
> 1 files changed, 7 insertions(+), 0 deletions(-)
>
> diff --git a/MAINTAINERS b/MAINTAINERS
> index 6d119c9..2e8334d 100644
> --- a/MAINTAINERS
> +++ b/MAINTAINERS
> @@ -5205,6 +5205,13 @@ M: Nicolas Pitre <[email protected]>
> S: Odd Fixes
> F: drivers/net/smc91x.*
>
> +SMM665 HARDWARE MONITOR DRIVER
> +M: "Guenter Roeck" <[email protected]>

Quotes aren't needed. They are present below for Mark only because of
the dot in his name, which needs quoting.

> +L: [email protected]
> +S: Maintained
> +F: Documentation/hwmon/smm665
> +F: drivers/hwmon/smm665.c
> +
> SMSC47B397 HARDWARE MONITOR DRIVER
> M: "Mark M. Hoffman" <[email protected]>
> L: [email protected]

Other than this, this patch looks good to me.

Acked-by: Jean Delvare <[email protected]>

--
Jean Delvare

2010-06-27 10:21:33

by Mark Brown

[permalink] [raw]
Subject: Re: [PATCH v3 1/4] hwmon: Driver for SMM665 Six-Channel Active DC Output Controller/Monitor

On Thu, Jun 24, 2010 at 03:00:58PM -0700, Guenter Roeck wrote:
> This driver adds support for the monitoring features of the Summit
> Microelectronics SMM665 Six-Channel Active DC Output Controller/Monitor.

A bit late to the game here but this looks like the chip has some
regulator control functionality as well as monitoring functionality (and
the product page on the Summit web site suggests so also). This means
that when fully supported in software the driver would cross multiple
subsystems so it might make sense to start off with a MFD rather than
direct I2C control?

If the non-monitoring functionality can't be controlled from software
this isn't an issue.

2010-06-27 15:12:08

by Guenter Roeck

[permalink] [raw]
Subject: Re: [PATCH v3 1/4] hwmon: Driver for SMM665 Six-Channel Active DC Output Controller/Monitor

On Sun, Jun 27, 2010 at 06:20:59AM -0400, Mark Brown wrote:
> On Thu, Jun 24, 2010 at 03:00:58PM -0700, Guenter Roeck wrote:
> > This driver adds support for the monitoring features of the Summit
> > Microelectronics SMM665 Six-Channel Active DC Output Controller/Monitor.
>
> A bit late to the game here but this looks like the chip has some
> regulator control functionality as well as monitoring functionality (and
> the product page on the Summit web site suggests so also). This means
> that when fully supported in software the driver would cross multiple
> subsystems so it might make sense to start off with a MFD rather than
> direct I2C control?
>
> If the non-monitoring functionality can't be controlled from software
> this isn't an issue.

I thought about that when I started working on the driver, but concluded that
it does not really make sense.

The chip is commonly used to control all supply voltages on a board.
Changing those voltages is not a good idea. For that reason, the chip can be
set into read-only mode, where changing the voltages is no longer possible
after initial programming.

While it is theoretically possible that someone might use the device to control
not only fixed but also dynamic voltages, I think that is highly unlikely, given
the risk involved in blowing up the board. Thus, moving the driver to mfd would
effectively serve no real purpose other than to cause confusion and add unnecessary
complexity.

Guenter

2010-06-27 19:54:12

by Mark Brown

[permalink] [raw]
Subject: Re: [PATCH v3 1/4] hwmon: Driver for SMM665 Six-Channel Active DC Output Controller/Monitor

On 27 Jun 2010, at 16:10, Guenter Roeck wrote:

> On Sun, Jun 27, 2010 at 06:20:59AM -0400, Mark Brown wrote:

>> A bit late to the game here but this looks like the chip has some
>> regulator control functionality as well as monitoring functionality (and
>> the product page on the Summit web site suggests so also). This means
>> that when fully supported in software the driver would cross multiple
>> subsystems so it might make sense to start off with a MFD rather than
>> direct I2C control?
>>
>> If the non-monitoring functionality can't be controlled from software
>> this isn't an issue.
>
> I thought about that when I started working on the driver, but concluded that
> it does not really make sense.
>
> The chip is commonly used to control all supply voltages on a board.
> Changing those voltages is not a good idea. For that reason, the chip can be

There's rather a lot of systems out there doing DVFS or using things like MMC
cards which would disagree with the idea that it's a bad idea to change supply
voltages at runtime. Even for fixed voltage supplies enabling and disabling at
runtime is useful to save power. I think it's fair to say that the overall trend is
towards more dynamic power management.

> set into read-only mode, where changing the voltages is no longer possible
> after initial programming.

This tends to be done through paranoia more than anything else - people get
very worried about things like accidental writes to their PMICs.

> While it is theoretically possible that someone might use the device to control
> not only fixed but also dynamic voltages, I think that is highly unlikely, given
> the risk involved in blowing up the board. Thus, moving the driver to mfd would
> effectively serve no real purpose other than to cause confusion and add unnecessary

Pretty much any current generation CPU can use dynamic voltage configuration
for DVFS.-

2010-06-27 22:08:43

by Guenter Roeck

[permalink] [raw]
Subject: Re: [PATCH v3 1/4] hwmon: Driver for SMM665 Six-Channel Active DC Output Controller/Monitor

On Sun, Jun 27, 2010 at 03:54:06PM -0400, Mark Brown wrote:
> On 27 Jun 2010, at 16:10, Guenter Roeck wrote:
>
> > On Sun, Jun 27, 2010 at 06:20:59AM -0400, Mark Brown wrote:
>
> >> A bit late to the game here but this looks like the chip has some
> >> regulator control functionality as well as monitoring functionality (and
> >> the product page on the Summit web site suggests so also). This means
> >> that when fully supported in software the driver would cross multiple
> >> subsystems so it might make sense to start off with a MFD rather than
> >> direct I2C control?
> >>
> >> If the non-monitoring functionality can't be controlled from software
> >> this isn't an issue.
> >
> > I thought about that when I started working on the driver, but concluded that
> > it does not really make sense.
> >
> > The chip is commonly used to control all supply voltages on a board.
> > Changing those voltages is not a good idea. For that reason, the chip can be
>
> There's rather a lot of systems out there doing DVFS or using things like MMC
> cards which would disagree with the idea that it's a bad idea to change supply
> voltages at runtime. Even for fixed voltage supplies enabling and disabling at
> runtime is useful to save power. I think it's fair to say that the overall trend is
> towards more dynamic power management.
>
> > set into read-only mode, where changing the voltages is no longer possible
> > after initial programming.
>
> This tends to be done through paranoia more than anything else - people get
> very worried about things like accidental writes to their PMICs.
>
Ok, I'll accept that.

Even then, I think it would make more sense to stick with the current approach.
If someone wants to extend the driver at some point in the future to support
regulator functionality, it would be straightforward to create the necessary
mfd framework for the chip at that time. After all, the hwmon driver and with it
most of the current code would still be required.

> > While it is theoretically possible that someone might use the device to control
> > not only fixed but also dynamic voltages, I think that is highly unlikely, given
> > the risk involved in blowing up the board. Thus, moving the driver to mfd would
> > effectively serve no real purpose other than to cause confusion and add unnecessary
>
> Pretty much any current generation CPU can use dynamic voltage configuration
> for DVFS.