Skip to contentSkip to conversation

v1 Intel FB-DIMM AMB temperature sensor

PATCH In review Darrick J. Wong 4 participants 16 messages Jul 13, 2007 – Oct 28, 2007 1 message not in the archive

Subsystems (2)
lore
Share & tools
lore
#0 lore raw

PATCH 1/1 v1 Intel FB-DIMM AMB temperature sensor

Hi all,

Here's a first draft of a driver to report FB-DIMM temperatures via the AMB chips on systems with the Intel 5000 series chipset. I've tested this on an IBM IntelliStation Z30, an IBM x3650 and an IBM x3550, and haven't seen any problems so far. Please let me know about any problems that you notice with this driver, and please cc me because I'm not subscribed to the list.

---
i5k_amb: hwmon driver for FB-DIMM AMB thermal sensors on Intel 5000 series chipsets

---

 drivers/hwmon/Kconfig   |   10 +
 drivers/hwmon/Makefile  |    1 
 drivers/hwmon/i5k_amb.c |  528 +++++++++++++++++++++++++++++++++++++++++++++++
 3 files changed, 539 insertions(+), 0 deletions(-)
drivers/hwmon/Kconfig +10 -0
index bb57beb..48f4332 100644
@@ -138,6 +138,16 @@ config SENSORS_K8TEMP
 	  This driver can also be built as a module.  If so, the module
 	  will be called k8temp.
 
+config SENSORS_I5K_AMB
+	tristate "FB-DIMM AMB temperature sensor on Intel 5000 series chipsets"
+	depends on I2C && EXPERIMENTAL
+	help
+	  If you say yes here you get support for FB-DIMM AMB temperature
+	  monitoring chips on systems with the Intel 5000 series chipset.
+
+	  This driver can also be built as a module. If so, the module
+	  will be called i5k_amb.
+
 config SENSORS_AMS
 	tristate "Apple Motion Sensor driver"
 	depends on PPC_PMAC && !PPC64 && INPUT && ((ADB_PMU && I2C = y) || (ADB_PMU && !I2C) || I2C) && EXPERIMENTAL
drivers/hwmon/Makefile +1 -0
index e5fc251..ab0cbb6 100644
@@ -22,6 +22,7 @@ obj-$(CONFIG_SENSORS_ADM1026)	+= adm1026.o
 obj-$(CONFIG_SENSORS_ADM1029)	+= adm1029.o
 obj-$(CONFIG_SENSORS_ADM1031)	+= adm1031.o
 obj-$(CONFIG_SENSORS_ADM9240)	+= adm9240.o
+obj-$(CONFIG_SENSORS_I5K_AMB)	+= i5k_amb.o
 obj-$(CONFIG_SENSORS_APPLESMC)	+= applesmc.o
 obj-$(CONFIG_SENSORS_AMS)	+= ams/
 obj-$(CONFIG_SENSORS_ATXP1)	+= atxp1.o
drivers/hwmon/i5k_amb.c new file +528 -0

528 lines not on this page · raw message

_______________________________________________
lm-sensors mailing list
[email protected]
http://lists.lm-sensors.org/mailman/listinfo/lm-sensors

#1 lore raw
Quoted 6 lines · Darrick J. Wong <djwong <at> us.ibm.com> writes:

Here's a first draft of a driver to report FB-DIMM temperatures via the AMB chips on systems with the Intel 5000 series chipset. I've tested this on an IBM IntelliStation Z30, an IBM x3650 and an IBM x3550, and haven't seen any problems so far. Please let me know about any problems that you notice with this driver, and please cc me because I'm not subscribed to the list.

Hello again,

I've tested this driver with kernel 2.6.20 with Tyan i5000XT motherboard
and it seems to work fine for me:

# grep . /sys/devices/platform/i5k_amb.0/temp*input
/sys/devices/platform/i5k_amb.0/temp0_input:78500
/sys/devices/platform/i5k_amb.0/temp16_input:75000
/sys/devices/platform/i5k_amb.0/temp32_input:64000
/sys/devices/platform/i5k_amb.0/temp48_input:63500

Thanks!
VZ

_______________________________________________
lm-sensors mailing list
[email protected]
http://lists.lm-sensors.org/mailman/listinfo/lm-sensors

#2 lore raw
In reply to Darrick J. Wong

[Attempting to separate the i5k_amb and adt7470 discussions.]

Quoted 8 lines · Vadim Zeitlin

On Sat, 14 Jul 2007 18:19:22 +0200 Hans de Goede <[email protected]> wrote:

HdG> > and the AMB sensors 2 and 3 were not being taken into account. Without knowing HdG> > anything about libsensors I thought it was because the file names in HdG> > /sys/devices/platform/i5k_amb.0 were named {temp0,temp16,temp32,temp48}_xxx and HdG> > not temp[1234] as usual
HdG>
HdG> Yes, thats a rather severe violation of the hwmon sysfs standard. I do not see

Oops. :)

The reason for the sparse numbering is that I wanted an easy way to map the sensor number to the memory channel/DIMM socket. Easy to do if the tempX numbers are sparse (i.e. 16 is always Ch1 DIMM0), not so easy if the identifer can change around. That said, I could create another attribute such as tempX_dimm or something that would export the real number for label matching purposes... if anyone cares. It's most likely to be the case that temp x == DIMM x anyway.

Quoted 4 lines

Hello again,

I've tested this driver with kernel 2.6.20 with Tyan i5000XT motherboard
and it seems to work fine for me:o

Excellent!

--D

_______________________________________________
lm-sensors mailing list
[email protected]
http://lists.lm-sensors.org/mailman/listinfo/lm-sensors

author Acked-by on v1 1/1
#3 lore raw
In reply to Darrick J. Wong
Quoted 1 line · Darrick J. Wong
Quoted 1 line

HdG> Yes, thats a rather severe violation of the hwmon sysfs standard. I do not see

Vadim, I've tested out your patch to adjust the tempX numbers for i5k_amb. It looks, builds, and runs ok, so:

--D

_______________________________________________
lm-sensors mailing list
[email protected]
http://lists.lm-sensors.org/mailman/listinfo/lm-sensors

#4 lore raw
In reply to Darrick J. Wong

Hi Darrick, Vadim:

Quoted 8 lines · Darrick J. Wong
Quoted 1 line · Darrick J. Wong
Quoted 1 line

HdG> Yes, thats a rather severe violation of the hwmon sysfs standard. I do not see

Vadim, I've tested out your patch to adjust the tempX numbers for i5k_amb. It looks, builds, and runs ok, so:

Um, I can't find Vadim's patch - it doesn't look like it ever hit the list.

1) Could one of you please roll the original+fix into a single patch and resend?

2) Please CC me directly on any patch that is considered ready to merge.

Thanks & regards,

Signature · 8 lines
Mark M. Hoffman
[email protected]


_______________________________________________
lm-sensors mailing list
[email protected]
http://lists.lm-sensors.org/mailman/listinfo/lm-sensors
#5 lore raw
In reply to Darrick J. Wong
Quoted 3 lines · Mark M. Hoffman

Um, I can't find Vadim's patch - it doesn't look like it ever hit the list.

1) Could one of you please roll the original+fix into a single patch and resend?

I can't combine them due to restrictions from IBM legal. However, that doesn't stop me from sending you a URL to Vadim's patch:
http://lists.lm-sensors.org/pipermail/lm-sensors/2007-July/020459.html

(The patch lacks a Signed-off-by tag, which is why I can't touch it.)

--D

_______________________________________________
lm-sensors mailing list
[email protected]
http://lists.lm-sensors.org/mailman/listinfo/lm-sensors

#6 lore raw
In reply to Darrick J. Wong

On Mon, 30 Jul 2007 12:26:09 -0700 "Darrick J. Wong" <[email protected]> wrote:

DJW> On Sun, Jul 29, 2007 at 11:38:40PM -0400, Mark M. Hoffman wrote:
DJW>
DJW> > Um, I can't find Vadim's patch - it doesn't look like it ever hit the list. DJW> >
DJW> > 1) Could one of you please roll the original+fix into a single patch and resend? DJW>
DJW> I can't combine them due to restrictions from IBM legal. However, that DJW> doesn't stop me from sending you a URL to Vadim's patch:
DJW> http://lists.lm-sensors.org/pipermail/lm-sensors/2007-July/020459.html
DJW>
DJW> (The patch lacks a Signed-off-by tag, which is why I can't touch it.)

I'm sorry, I didn't think it was worth doing it for such a trivial patch,
do you need me to resend it with the extra "Signed off by" line or is it already ok?

Thanks,
VZ

_______________________________________________
lm-sensors mailing list
[email protected]
http://lists.lm-sensors.org/mailman/listinfo/lm-sensors

#7 lore raw

PATCH 1/1 v2 of Intel FB-DIMM AMB temperature sensors

New driver to read FB-DIMM temperature sensors on systems with the Intel 5000 series chipsets, and now with all subsequent patches rolled into one.

---

 drivers/hwmon/Kconfig   |   10 +
 drivers/hwmon/Makefile  |    1 
 drivers/hwmon/i5k_amb.c |  529 +++++++++++++++++++++++++++++++++++++++++++++++
 3 files changed, 540 insertions(+), 0 deletions(-)
drivers/hwmon/Kconfig +10 -0
index 0fe3c52..00dc6fb 100644
@@ -215,6 +215,16 @@ config SENSORS_DS1621
 	  This driver can also be built as a module.  If so, the module
 	  will be called ds1621.
 
+config SENSORS_I5K_AMB
+	tristate "FB-DIMM AMB temperature sensor on Intel 5000 series chipsets"
+	depends on I2C && EXPERIMENTAL
+	help
+	  If you say yes here you get support for FB-DIMM AMB temperature
+	  monitoring chips on systems with the Intel 5000 series chipset.
+
+	  This driver can also be built as a module. If so, the module
+	  will be called i5k_amb.
+
 config SENSORS_F71805F
 	tristate "Fintek F71805F/FG and F71872F/FG"
 	depends on EXPERIMENTAL
drivers/hwmon/Makefile +1 -0
index bd75b44..a29805e 100644
@@ -35,6 +35,7 @@ obj-$(CONFIG_SENSORS_FSCPOS)	+= fscpos.o
 obj-$(CONFIG_SENSORS_GL518SM)	+= gl518sm.o
 obj-$(CONFIG_SENSORS_GL520SM)	+= gl520sm.o
 obj-$(CONFIG_SENSORS_HDAPS)	+= hdaps.o
+obj-$(CONFIG_SENSORS_I5K_AMB)	+= i5k_amb.o
 obj-$(CONFIG_SENSORS_IT87)	+= it87.o
 obj-$(CONFIG_SENSORS_K8TEMP)	+= k8temp.o
 obj-$(CONFIG_SENSORS_LM63)	+= lm63.o
drivers/hwmon/i5k_amb.c new file +516 -0

516 lines not on this page · raw message

_______________________________________________
lm-sensors mailing list
[email protected]
http://lists.lm-sensors.org/mailman/listinfo/lm-sensors

#8 lore raw

Re: [lm-sensors] [PATCH] v2 of Intel FB-DIMM AMB temperature sensors

Hi Darrick:

Quoted 17 lines · Darrick J. Wong

New driver to read FB-DIMM temperature sensors on systems with the Intel 5000 series chipsets, and now with all subsequent patches rolled into one.

---

 drivers/hwmon/Kconfig   |   10 +
 drivers/hwmon/Makefile  |    1 
 drivers/hwmon/i5k_amb.c |  529 +++++++++++++++++++++++++++++++++++++++++++++++
 3 files changed, 540 insertions(+), 0 deletions(-)
drivers/hwmon/Kconfig +0 -0
index 0fe3c52..00dc6fb 100644

drivers/hwmon/Kconfig config SENSORS_DS1621 · line 220

Quoted 6 lines
drivers/hwmon/Kconfig +3 -0
@@ -215,6 +215,16 @@ config SENSORS_DS1621
 	  This driver can also be built as a module.  If so, the module
 	  will be called ds1621.
 
+config SENSORS_I5K_AMB
+	tristate "FB-DIMM AMB temperature sensor on Intel 5000 series chipsets"
+	depends on I2C && EXPERIMENTAL

...and X86, presumably.

Quoted 58 lines
drivers/hwmon/Kconfig +7 -0
+	help
+	  If you say yes here you get support for FB-DIMM AMB temperature
+	  monitoring chips on systems with the Intel 5000 series chipset.
+
+	  This driver can also be built as a module. If so, the module
+	  will be called i5k_amb.
+
 config SENSORS_F71805F
 	tristate "Fintek F71805F/FG and F71872F/FG"
 	depends on EXPERIMENTAL
drivers/hwmon/Makefile +1 -0
index bd75b44..a29805e 100644
@@ -35,6 +35,7 @@ obj-$(CONFIG_SENSORS_FSCPOS)	+= fscpos.o
 obj-$(CONFIG_SENSORS_GL518SM)	+= gl518sm.o
 obj-$(CONFIG_SENSORS_GL520SM)	+= gl520sm.o
 obj-$(CONFIG_SENSORS_HDAPS)	+= hdaps.o
+obj-$(CONFIG_SENSORS_I5K_AMB)	+= i5k_amb.o
 obj-$(CONFIG_SENSORS_IT87)	+= it87.o
 obj-$(CONFIG_SENSORS_K8TEMP)	+= k8temp.o
 obj-$(CONFIG_SENSORS_LM63)	+= lm63.o
drivers/hwmon/i5k_amb.c new file +30 -0
new file mode 100644
index 0000000..4e5072c
@@ -0,0 +1,529 @@
+/*
+ * A hwmon driver for the Intel 5000 series chipset FB-DIMM AMB
+ * temperature sensors
+ * Copyright (C) 2007 IBM
+ *
+ * Author: Darrick J. Wong <[email protected]>
+ *
+ * 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; either version 2 of the License, or
+ * (at your option) any later version.
+ *
+ * This program is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+ * GNU General Public License for more details.
+ *
+ * You should have received a copy of the GNU General Public License
+ * along with this program; if not, write to the Free Software
+ * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA  02111-1307  USA
+ */
+
+#include <linux/module.h>
+#include <linux/jiffies.h>
+#include <linux/hwmon.h>
+#include <linux/hwmon-sysfs.h>
+#include <linux/err.h>
+#include <linux/mutex.h>
+#include <linux/delay.h>
+#include <linux/log2.h>

drivers/hwmon/i5k_amb.c line 40

Quoted 10 lines
drivers/hwmon/i5k_amb.c new file +10 -0
+#include <linux/pci.h>
+#include <linux/platform_device.h>
+
+#define DRVNAME "i5k_amb"
+
+#define PCI_VENDOR_INTEL		0x8086
+#define PCI_DEVICE_INTEL_5K_MEM		0x25F0
+#define PCI_DEVICE_INTEL_5K_FBD0	0x25F5
+#define PCI_DEVICE_INTEL_5K_FBD1	0x25F6
+

PCI_VENDOR_ID_INTEL is already defined. The rest should be added to include/linux/pci_ids.h, not here. You can do that in the same patch.

Quoted 53 lines
drivers/hwmon/i5k_amb.c new file +53 -0
+#define I5K_REG_AMB_BASE_ADDR		0x48
+#define I5K_REG_AMB_LEN_ADDR		0x50
+#define I5K_REG_CHAN0_PRESENCE_ADDR	0x64
+#define I5K_REG_CHAN1_PRESENCE_ADDR	0x66
+
+#define AMB_REG_TEMP_MIN_ADDR		0x80
+#define AMB_REG_TEMP_MID_ADDR		0x81
+#define AMB_REG_TEMP_MAX_ADDR		0x82
+#define AMB_REG_TEMP_STATUS_ADDR	0x84
+#define AMB_REG_TEMP_ADDR		0x85
+
+#define AMB_CONFIG_SIZE			2048
+#define AMB_FUNC_3_OFFSET		768
+
+#define AMB_REG_TEMP_STATUS(amb)	((amb) * AMB_CONFIG_SIZE + \
+			AMB_FUNC_3_OFFSET + AMB_REG_TEMP_STATUS_ADDR)
+#define AMB_REG_TEMP_MIN(amb)		((amb) * AMB_CONFIG_SIZE + \
+			AMB_FUNC_3_OFFSET + AMB_REG_TEMP_MIN_ADDR)
+#define AMB_REG_TEMP_MID(amb)		((amb) * AMB_CONFIG_SIZE + \
+			AMB_FUNC_3_OFFSET + AMB_REG_TEMP_MID_ADDR)
+#define AMB_REG_TEMP_MAX(amb)		((amb) * AMB_CONFIG_SIZE + \
+			AMB_FUNC_3_OFFSET + AMB_REG_TEMP_MAX_ADDR)
+#define AMB_REG_TEMP(amb)		((amb) * AMB_CONFIG_SIZE + \
+			AMB_FUNC_3_OFFSET + AMB_REG_TEMP_ADDR)
+
+#define MAX_MEM_CHANNELS		4
+#define MAX_AMBS_PER_CHANNEL		16
+#define MAX_AMBS			(MAX_MEM_CHANNELS * \
+					 MAX_AMBS_PER_CHANNEL)
+/*
+ * Ugly hack: For some reason the highest bit is set if there
+ * are _any_ DIMMs in the channel.  Attempting to read from
+ * this "high-order" AMB results in a memory bus error, so
+ * for now we'll just ignore that top bit, even though that
+ * might prevent us from seeing the 16th DIMM in the channel.
+ */
+#define REAL_MAX_AMBS_PER_CHANNEL	15
+#define KNOBS_PER_AMB			5
+
+#define AMB_NUM_FROM_REG(byte_num, bit_num)	((byte_num) * \
+			MAX_AMBS_PER_CHANNEL) + (bit_num)
+
+struct i5k_amb_data {
+	struct class_device *class_dev;
+	struct resource *resource;
+
+	unsigned long amb_base;
+	unsigned long amb_len;
+	u16 amb_present[MAX_MEM_CHANNELS];
+	void __iomem *amb_mmio;
+	struct sensor_device_attribute *attrs;
+	unsigned int num_attrs;
+};

drivers/hwmon/i5k_amb.c line 103

Quoted 10 lines
drivers/hwmon/i5k_amb.c new file +10 -0
+
+static ssize_t show_name(struct device *dev, struct device_attribute *devattr,
+			 char *buf)
+{
+	return sprintf(buf, "%s\n", DRVNAME);
+}
+
+
+static SENSOR_DEVICE_ATTR(name, S_IRUGO, show_name, NULL, 0);
+

Just DEVICE_ATTR is enough here.

drivers/hwmon/i5k_amb.c line 105

Quoted 2 lines
drivers/hwmon/i5k_amb.c new file +2 -0
+static struct platform_device *amb_pdev = NULL;
+

Initialization to 0/NULL is unnecessary.

drivers/hwmon/i5k_amb.c line 108

Quoted 3 lines
drivers/hwmon/i5k_amb.c new file +3 -0
+static u8 amb_read_byte(struct i5k_amb_data *data, unsigned long offset)
+{
+	return (*(u8*)(data->amb_mmio + offset));
	return ioread8(data->amb_mmio + offset);

drivers/hwmon/i5k_amb.c line 114

Quoted 6 lines
drivers/hwmon/i5k_amb.c new file +6 -0
+}
+
+static void amb_write_byte(struct i5k_amb_data *data, unsigned long offset,
+			   u8 val)
+{
+	(*(u8*)(data->amb_mmio + offset)) = val;
	iowrite8(val, data->amb_mmio + offset);
Quoted 21 lines
drivers/hwmon/i5k_amb.c new file +21 -0
+}
+
+static ssize_t show_amb_alarm(struct device *dev,
+			     struct device_attribute *devattr,
+			     char *buf)
+{
+	struct sensor_device_attribute *attr = to_sensor_dev_attr(devattr);
+	struct i5k_amb_data *data = dev_get_drvdata(dev);
+
+	if ( !(amb_read_byte(data, AMB_REG_TEMP_STATUS(attr->index)) & 0x20) &&
+	      (amb_read_byte(data, AMB_REG_TEMP_STATUS(attr->index)) & 0x8) )
+		return sprintf(buf, "1\n");
+	else
+		return sprintf(buf, "0\n");
+}
+
+static ssize_t store_amb_min(struct device *dev,
+			     struct device_attribute *devattr,
+			     const char *buf,
+			     size_t count)
+{

drivers/hwmon/i5k_amb.c line 145

Quoted 10 lines
drivers/hwmon/i5k_amb.c new file +10 -0
+	struct sensor_device_attribute *attr = to_sensor_dev_attr(devattr);
+	struct i5k_amb_data *data = dev_get_drvdata(dev);
+	int temp = simple_strtol(buf, NULL, 10) / 500;
+
+	if (temp > 255)
+		temp = 255;
+
+	amb_write_byte(data, AMB_REG_TEMP_MIN(attr->index), temp);
+	return count;
+}

See Documentation/hwmon/sysfs_interface: sysfs attribute writes interpretation

Summary: use long instead of int, and range check on both ends.

Quoted 6 lines
drivers/hwmon/i5k_amb.c new file +6 -0
+
+static ssize_t store_amb_mid(struct device *dev,
+			     struct device_attribute *devattr,
+			     const char *buf,
+			     size_t count)
+{

drivers/hwmon/i5k_amb.c line 161

Quoted 10 lines
drivers/hwmon/i5k_amb.c new file +10 -0
+	struct sensor_device_attribute *attr = to_sensor_dev_attr(devattr);
+	struct i5k_amb_data *data = dev_get_drvdata(dev);
+	int temp = simple_strtol(buf, NULL, 10) / 500;
+
+	if (temp > 255)
+		temp = 255;
+
+	amb_write_byte(data, AMB_REG_TEMP_MID(attr->index), temp);
+	return count;
+}

Ditto.

Quoted 6 lines
drivers/hwmon/i5k_amb.c new file +6 -0
+
+static ssize_t store_amb_max(struct device *dev,
+			     struct device_attribute *devattr,
+			     const char *buf,
+			     size_t count)
+{

drivers/hwmon/i5k_amb.c line 177

Quoted 10 lines
drivers/hwmon/i5k_amb.c new file +10 -0
+	struct sensor_device_attribute *attr = to_sensor_dev_attr(devattr);
+	struct i5k_amb_data *data = dev_get_drvdata(dev);
+	int temp = simple_strtol(buf, NULL, 10) / 500;
+
+	if (temp > 255)
+		temp = 255;
+
+	amb_write_byte(data, AMB_REG_TEMP_MAX(attr->index), temp);
+	return count;
+}

Ditto.

Quoted 67 lines
drivers/hwmon/i5k_amb.c new file +67 -0
+
+static ssize_t show_amb_min(struct device *dev,
+			     struct device_attribute *devattr,
+			     char *buf)
+{
+	struct sensor_device_attribute *attr = to_sensor_dev_attr(devattr);
+	struct i5k_amb_data *data = dev_get_drvdata(dev);
+	return sprintf(buf, "%d\n",
+		500 * amb_read_byte(data, AMB_REG_TEMP_MIN(attr->index)));
+}
+
+static ssize_t show_amb_mid(struct device *dev,
+			     struct device_attribute *devattr,
+			     char *buf)
+{
+	struct sensor_device_attribute *attr = to_sensor_dev_attr(devattr);
+	struct i5k_amb_data *data = dev_get_drvdata(dev);
+	return sprintf(buf, "%d\n",
+		500 * amb_read_byte(data, AMB_REG_TEMP_MID(attr->index)));
+}
+
+static ssize_t show_amb_max(struct device *dev,
+			     struct device_attribute *devattr,
+			     char *buf)
+{
+	struct sensor_device_attribute *attr = to_sensor_dev_attr(devattr);
+	struct i5k_amb_data *data = dev_get_drvdata(dev);
+	return sprintf(buf, "%d\n",
+		500 * amb_read_byte(data, AMB_REG_TEMP_MAX(attr->index)));
+}
+
+static ssize_t show_amb_temp(struct device *dev,
+			     struct device_attribute *devattr,
+			     char *buf)
+{
+	struct sensor_device_attribute *attr = to_sensor_dev_attr(devattr);
+	struct i5k_amb_data *data = dev_get_drvdata(dev);
+	return sprintf(buf, "%d\n",
+		500 * amb_read_byte(data, AMB_REG_TEMP(attr->index)));
+}
+
+static int __devinit i5k_amb_hwmon_init(struct platform_device *pdev)
+{
+	int i, j, k, d = 0;
+	u16 c;
+	char *n;
+	int res = 0;
+	int num_ambs = 0;
+	struct i5k_amb_data *data = platform_get_drvdata(pdev);
+
+	/* Count the number of AMBs found */
+	for (i = 0; i < MAX_MEM_CHANNELS; i++) {
+		c = data->amb_present[i];
+		for (j = 0; j < REAL_MAX_AMBS_PER_CHANNEL; j++) {
+			if (c & 0x1)
+				num_ambs++;
+			c >>= 1;
+		}
+	}
+
+	/* Set up sysfs stuff */
+	data->attrs = kzalloc(sizeof(*data->attrs) * num_ambs * KNOBS_PER_AMB,
+				GFP_KERNEL);
+	if (!data->attrs)
+		return -ENOMEM;
+	data->num_attrs = 0;
+

drivers/hwmon/i5k_amb.c line 254

Quoted 10 lines
drivers/hwmon/i5k_amb.c new file +10 -0
+	for (i = 0; i < MAX_MEM_CHANNELS; i++) {
+		c = data->amb_present[i];
+		for (j = 0; j < REAL_MAX_AMBS_PER_CHANNEL; j++, c >>= 1) {
+			k = AMB_NUM_FROM_REG(i, j);
+			if (!(c & 0x1))
+				continue;
+			d++;
+
+			/* Temperature sysfs knob */
+			n = kmalloc(13, GFP_KERNEL);

Magic constant.

drivers/hwmon/i5k_amb.c line 255

Quoted 1 line
drivers/hwmon/i5k_amb.c new file +1 -0
+			sprintf(n, "temp%d_input", d);

Is it possible, at least in theory, for d > 99?

drivers/hwmon/i5k_amb.c line 267

Quoted 12 lines
drivers/hwmon/i5k_amb.c new file +12 -0
+			data->attrs[data->num_attrs].dev_attr.attr.name = n;
+			data->attrs[data->num_attrs].dev_attr.attr.mode > +					S_IRUGO;
+			data->attrs[data->num_attrs].dev_attr.show > +					show_amb_temp;
+			data->attrs[data->num_attrs].index = k;
+			res = device_create_file(&pdev->dev,
+					&data->attrs[data->num_attrs].dev_attr);
+			if (res)
+				goto exit_remove;
+			data->num_attrs++;
+
+			/* Temperature min sysfs knob */
+			n = kmalloc(13, GFP_KERNEL);

Magic constant, and not even correct this time, etc.

Quoted 64 lines
drivers/hwmon/i5k_amb.c new file +64 -0
+			sprintf(n, "temp%d_min", d);
+			data->attrs[data->num_attrs].dev_attr.attr.name = n;
+			data->attrs[data->num_attrs].dev_attr.attr.mode > +					S_IWUSR | S_IRUGO;
+			data->attrs[data->num_attrs].dev_attr.show > +					show_amb_min;
+			data->attrs[data->num_attrs].dev_attr.store > +					store_amb_min;
+			data->attrs[data->num_attrs].index = k;
+			res = device_create_file(&pdev->dev,
+					&data->attrs[data->num_attrs].dev_attr);
+			if (res)
+				goto exit_remove;
+			data->num_attrs++;
+
+			/* Temperature mid sysfs knob */
+			n = kmalloc(13, GFP_KERNEL);
+			sprintf(n, "temp%d_mid", d);
+			data->attrs[data->num_attrs].dev_attr.attr.name = n;
+			data->attrs[data->num_attrs].dev_attr.attr.mode > +					S_IWUSR | S_IRUGO;
+			data->attrs[data->num_attrs].dev_attr.show > +					show_amb_mid;
+			data->attrs[data->num_attrs].dev_attr.store > +					store_amb_mid;
+			data->attrs[data->num_attrs].index = k;
+			res = device_create_file(&pdev->dev,
+					&data->attrs[data->num_attrs].dev_attr);
+			if (res)
+				goto exit_remove;
+			data->num_attrs++;
+
+			/* Temperature max sysfs knob */
+			n = kmalloc(13, GFP_KERNEL);
+			sprintf(n, "temp%d_max", d);
+			data->attrs[data->num_attrs].dev_attr.attr.name = n;
+			data->attrs[data->num_attrs].dev_attr.attr.mode > +					S_IWUSR | S_IRUGO;
+			data->attrs[data->num_attrs].dev_attr.show > +					show_amb_max;
+			data->attrs[data->num_attrs].dev_attr.store > +					store_amb_max;
+			data->attrs[data->num_attrs].index = k;
+			res = device_create_file(&pdev->dev,
+					&data->attrs[data->num_attrs].dev_attr);
+			if (res)
+				goto exit_remove;
+			data->num_attrs++;
+
+			/* Temperature alarm sysfs knob */
+			n = kmalloc(13, GFP_KERNEL);
+			sprintf(n, "temp%d_alarm", d);
+			data->attrs[data->num_attrs].dev_attr.attr.name = n;
+			data->attrs[data->num_attrs].dev_attr.attr.mode > +					S_IRUGO;
+			data->attrs[data->num_attrs].dev_attr.show > +					show_amb_alarm;
+			data->attrs[data->num_attrs].index = k;
+			res = device_create_file(&pdev->dev,
+					&data->attrs[data->num_attrs].dev_attr);
+			if (res)
+				goto exit_remove;
+			data->num_attrs++;
+		}
+	}
+
+	res = device_create_file(&pdev->dev, &sensor_dev_attr_name.dev_attr);
+	if (res)
+		goto exit_remove;
+
+	data->class_dev = hwmon_device_register(&pdev->dev);
+	if (IS_ERR(data->class_dev)) {
+		res = PTR_ERR(data->class_dev);
+		goto exit_remove;
+	}

drivers/hwmon/i5k_amb.c line 341

Quoted 10 lines
drivers/hwmon/i5k_amb.c new file +10 -0
+
+	return res;
+
+exit_remove:
+	device_remove_file(&pdev->dev, &sensor_dev_attr_name.dev_attr);
+	for (i = 0; i < data->num_attrs; i++) {
+		device_remove_file(&pdev->dev, &data->attrs[i].dev_attr);
+		kfree(data->attrs[i].dev_attr.attr.name);
+	}
+	kfree(data->attrs);

Hmm, would be nice if data->attrs included space for all the strings, too. Then you could kmalloc/kfree in one shot.

Quoted 21 lines
drivers/hwmon/i5k_amb.c new file +21 -0
+
+	return res;
+}
+
+static int __devinit i5k_amb_add(void)
+{
+	int res = -ENODEV;
+
+	/* only ever going to be one of these */
+	amb_pdev = platform_device_alloc(DRVNAME, 0);
+	if (!amb_pdev)
+		return -ENOMEM;
+
+	res = platform_device_add(amb_pdev);
+	if (res)
+		goto err;
+	return 0;
+
+err:
+	platform_device_put(amb_pdev);
+	return res;

drivers/hwmon/i5k_amb.c line 372

Quoted 10 lines
drivers/hwmon/i5k_amb.c new file +10 -0
+}
+
+static int __devinit i5k_amb_probe(struct platform_device *pdev)
+{
+	struct pci_dev *pcidev;
+	struct i5k_amb_data *data;
+	struct resource *reso;
+	u32 val32;
+	u16 val16;
+	int res = -ENOMEM;

-ENODEV seems to be a better default for this function.

Quoted 38 lines
drivers/hwmon/i5k_amb.c new file +38 -0
+
+	data = kzalloc(sizeof(*data), GFP_KERNEL);
+	if (!data)
+		return -ENOMEM;
+
+	/* Figure out where the AMB registers live */
+	pcidev = pci_get_device(PCI_VENDOR_INTEL, PCI_DEVICE_INTEL_5K_MEM,
+				NULL);
+	if (!pcidev)
+		goto out;
+
+	if (pci_read_config_dword(pcidev, I5K_REG_AMB_BASE_ADDR, &val32))
+		goto err_free_dev;
+	data->amb_base = val32;
+
+	if (pci_read_config_dword(pcidev, I5K_REG_AMB_LEN_ADDR, &val32))
+		goto err_free_dev;
+	data->amb_len = val32;
+
+	pci_dev_put(pcidev);
+
+	if (data->amb_len < AMB_CONFIG_SIZE * MAX_AMBS) {
+		printk(KERN_ERR DRVNAME "AMB region too small!\n");
+		goto out;
+	}
+
+
+	/* Copy the DIMM presence map for the first two channels */
+	pcidev = pci_get_device(PCI_VENDOR_INTEL, PCI_DEVICE_INTEL_5K_FBD0,
+				NULL);
+	if (!pcidev)
+		goto out;
+
+	if (pci_read_config_word(pcidev, I5K_REG_CHAN0_PRESENCE_ADDR, &val16))
+		goto err_free_dev;
+	data->amb_present[0] = val16;
+	if (pci_read_config_word(pcidev, I5K_REG_CHAN1_PRESENCE_ADDR, &val16))
+		goto err_free_dev;

drivers/hwmon/i5k_amb.c line 420

Quoted 10 lines
drivers/hwmon/i5k_amb.c new file +10 -0
+	data->amb_present[1] = val16;
+
+	pci_dev_put(pcidev);
+
+	/* Copy the DIMM presence map for the optional second two channels */
+	pcidev = pci_get_device(PCI_VENDOR_INTEL, PCI_DEVICE_INTEL_5K_FBD1,
+				NULL);
+	if (!pcidev)
+		goto setup_resources;
+

Abuse of goto. OK, maybe it's nit-picky... the accepted idiom in Linux is to use goto to unwind inits after an error. This is not an error condition, *and* this goto is easily replaced by an if statement.

Quoted 10 lines
drivers/hwmon/i5k_amb.c new file +10 -0
+	if (pci_read_config_word(pcidev, I5K_REG_CHAN0_PRESENCE_ADDR, &val16))
+		goto err_free_dev;
+	data->amb_present[2] = val16;
+	if (pci_read_config_word(pcidev, I5K_REG_CHAN1_PRESENCE_ADDR, &val16))
+		goto err_free_dev;
+	data->amb_present[3] = val16;
+
+	pci_dev_put(pcidev);
+
+setup_resources:

drivers/hwmon/i5k_amb.c line 440

Quoted 10 lines
drivers/hwmon/i5k_amb.c new file +10 -0
+	/* Set up resource regions */
+	reso = request_mem_region(data->amb_base, data->amb_len, DRVNAME);
+	if (!reso) {
+		res = -EBUSY;
+		goto out;
+	}
+	
+	data->amb_mmio = ioremap_nocache(data->amb_base, data->amb_len);
+	if (!data->amb_mmio)
+		goto err2;

-EBUSY

Quoted 7 lines
drivers/hwmon/i5k_amb.c new file +7 -0
+
+	platform_set_drvdata(pdev, data);
+
+	res = i5k_amb_hwmon_init(pdev);
+	if (res)
+		goto err;
+		

drivers/hwmon/i5k_amb.c line 457

Quoted 10 lines
drivers/hwmon/i5k_amb.c new file +10 -0
+	return res;
+
+err:
+	iounmap(data->amb_mmio);
+	platform_set_drvdata(pdev, NULL);
+err2:
+	release_mem_region(data->amb_base, data->amb_len);
+	goto out;
+err_free_dev:
+	pci_dev_put(pcidev);

Careful: some of the goto paths that come through here don't hold a pcidev.

drivers/hwmon/i5k_amb.c line 462

Quoted 5 lines
drivers/hwmon/i5k_amb.c new file +5 -0
+out:
+	kfree(data);
+	return res;
+}
+

Please rename those labels in some consistent way.

Quoted 44 lines
drivers/hwmon/i5k_amb.c new file +44 -0
+static int __devexit i5k_amb_remove(struct platform_device *pdev)
+{
+	int i;
+	struct i5k_amb_data *data = platform_get_drvdata(pdev);
+
+	hwmon_device_unregister(data->class_dev);
+	device_remove_file(&pdev->dev, &sensor_dev_attr_name.dev_attr);
+	for (i = 0; i < data->num_attrs; i++) {
+		device_remove_file(&pdev->dev, &data->attrs[i].dev_attr);
+		kfree(data->attrs[i].dev_attr.attr.name);
+	}
+	iounmap(data->amb_mmio);
+	release_mem_region(data->amb_base, data->amb_len);
+	platform_set_drvdata(pdev, NULL);
+	kfree(data);
+	return 0;
+}
+
+static struct platform_driver i5k_amb_driver = {
+	.driver = {
+		.owner = THIS_MODULE,
+		.name = DRVNAME,
+	},
+	.probe = i5k_amb_probe,
+	.remove = __devexit_p(i5k_amb_remove),
+};
+
+static int __init i5k_amb_init(void)
+{
+	int res;
+
+	res = platform_driver_register(&i5k_amb_driver);
+	if (res)
+		return res;
+
+	res = i5k_amb_add();
+	if (res)
+		platform_driver_unregister(&i5k_amb_driver);
+
+	return res;
+}
+
+static void __exit i5k_amb_exit(void)
+{

drivers/hwmon/i5k_amb.c line 516

Quoted 10 lines
drivers/hwmon/i5k_amb.c new file +10 -0
+	platform_device_unregister(amb_pdev);
+	platform_driver_unregister(&i5k_amb_driver);
+}
+
+MODULE_AUTHOR("Darrick J. Wong <[email protected]>");
+MODULE_DESCRIPTION("Intel 5000 chipset FB-DIMM AMB temperature sensor");
+MODULE_LICENSE("GPL");
+
+module_init(i5k_amb_init);
+module_exit(i5k_amb_exit);

Regards,

Signature · 8 lines
Mark M. Hoffman
[email protected]


_______________________________________________
lm-sensors mailing list
[email protected]
http://lists.lm-sensors.org/mailman/listinfo/lm-sensors
#9 lore raw

PATCH v3 1/1 i5k_amb: New memory temperature sensor driver

New driver to read FB-DIMM temperature sensors on systems with the Intel 5000 series chipsets.

---

 drivers/hwmon/Kconfig   |   10 +
 drivers/hwmon/Makefile  |    1 
 drivers/hwmon/i5k_amb.c |  548 +++++++++++++++++++++++++++++++++++++++++++++++
 include/linux/pci_ids.h |    3 
 4 files changed, 562 insertions(+), 0 deletions(-)
drivers/hwmon/Kconfig +10 -0
index b8854b9..c2acb37 100644
@@ -215,6 +215,16 @@ config SENSORS_DS1621
 	  This driver can also be built as a module.  If so, the module
 	  will be called ds1621.
 
+config SENSORS_I5K_AMB
+	tristate "FB-DIMM AMB temperature sensor on Intel 5000 series chipsets"
+	depends on PCI && EXPERIMENTAL
+	help
+	  If you say yes here you get support for FB-DIMM AMB temperature
+	  monitoring chips on systems with the Intel 5000 series chipset.
+
+	  This driver can also be built as a module. If so, the module
+	  will be called i5k_amb.
+
 config SENSORS_F71805F
 	tristate "Fintek F71805F/FG, F71806F/FG and F71872F/FG"
 	depends on EXPERIMENTAL
drivers/hwmon/Makefile +1 -0
index 2f592e2..914c48c 100644
@@ -37,6 +37,7 @@ obj-$(CONFIG_SENSORS_FSCPOS)	+= fscpos.o
 obj-$(CONFIG_SENSORS_GL518SM)	+= gl518sm.o
 obj-$(CONFIG_SENSORS_GL520SM)	+= gl520sm.o
 obj-$(CONFIG_SENSORS_HDAPS)	+= hdaps.o
+obj-$(CONFIG_SENSORS_I5K_AMB)	+= i5k_amb.o
 obj-$(CONFIG_SENSORS_IBMPEX)	+= ibmpex.o
 obj-$(CONFIG_SENSORS_IT87)	+= it87.o
 obj-$(CONFIG_SENSORS_K8TEMP)	+= k8temp.o
drivers/hwmon/i5k_amb.c new file +548 -0

548 lines not on this page · raw message

include/linux/pci_ids.h +3 -0
index 55f307f..d641b2f 100644
@@ -2241,6 +2241,9 @@
 #define PCI_DEVICE_ID_INTEL_82915G_IG	0x2582
 #define PCI_DEVICE_ID_INTEL_82915GM_HB	0x2590
 #define PCI_DEVICE_ID_INTEL_82915GM_IG	0x2592
+#define PCI_DEVICE_ID_INTEL_5000_ERR	0x25F0
+#define PCI_DEVICE_ID_INTEL_5000_FBD0	0x25F5
+#define PCI_DEVICE_ID_INTEL_5000_FBD1	0x25F6
 #define PCI_DEVICE_ID_INTEL_82945G_HB	0x2770
 #define PCI_DEVICE_ID_INTEL_82945G_IG	0x2772
 #define PCI_DEVICE_ID_INTEL_3000_HB	0x2778

_______________________________________________
lm-sensors mailing list
[email protected]
http://lists.lm-sensors.org/mailman/listinfo/lm-sensors

#10 lore raw

Re: [PATCH v3] i5k_amb: New memory temperature sensor driver

Hi:

Just a couple more things...

Quoted 127 lines · Darrick J. Wong

New driver to read FB-DIMM temperature sensors on systems with the Intel 5000 series chipsets.

---

 drivers/hwmon/Kconfig   |   10 +
 drivers/hwmon/Makefile  |    1 
 drivers/hwmon/i5k_amb.c |  548 +++++++++++++++++++++++++++++++++++++++++++++++
 include/linux/pci_ids.h |    3 
 4 files changed, 562 insertions(+), 0 deletions(-)
drivers/hwmon/Kconfig +10 -0
index b8854b9..c2acb37 100644
@@ -215,6 +215,16 @@ config SENSORS_DS1621
 	  This driver can also be built as a module.  If so, the module
 	  will be called ds1621.
 
+config SENSORS_I5K_AMB
+	tristate "FB-DIMM AMB temperature sensor on Intel 5000 series chipsets"
+	depends on PCI && EXPERIMENTAL
+	help
+	  If you say yes here you get support for FB-DIMM AMB temperature
+	  monitoring chips on systems with the Intel 5000 series chipset.
+
+	  This driver can also be built as a module. If so, the module
+	  will be called i5k_amb.
+
 config SENSORS_F71805F
 	tristate "Fintek F71805F/FG, F71806F/FG and F71872F/FG"
 	depends on EXPERIMENTAL
drivers/hwmon/Makefile +1 -0
index 2f592e2..914c48c 100644
@@ -37,6 +37,7 @@ obj-$(CONFIG_SENSORS_FSCPOS)	+= fscpos.o
 obj-$(CONFIG_SENSORS_GL518SM)	+= gl518sm.o
 obj-$(CONFIG_SENSORS_GL520SM)	+= gl520sm.o
 obj-$(CONFIG_SENSORS_HDAPS)	+= hdaps.o
+obj-$(CONFIG_SENSORS_I5K_AMB)	+= i5k_amb.o
 obj-$(CONFIG_SENSORS_IBMPEX)	+= ibmpex.o
 obj-$(CONFIG_SENSORS_IT87)	+= it87.o
 obj-$(CONFIG_SENSORS_K8TEMP)	+= k8temp.o
drivers/hwmon/i5k_amb.c new file +76 -0
new file mode 100644
index 0000000..3bde717
@@ -0,0 +1,548 @@
+/*
+ * A hwmon driver for the Intel 5000 series chipset FB-DIMM AMB
+ * temperature sensors
+ * Copyright (C) 2007 IBM
+ *
+ * Author: Darrick J. Wong <[email protected]>
+ *
+ * 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; either version 2 of the License, or
+ * (at your option) any later version.
+ *
+ * This program is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+ * GNU General Public License for more details.
+ *
+ * You should have received a copy of the GNU General Public License
+ * along with this program; if not, write to the Free Software
+ * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA  02111-1307  USA
+ */
+
+#include <linux/module.h>
+#include <linux/jiffies.h>
+#include <linux/hwmon.h>
+#include <linux/hwmon-sysfs.h>
+#include <linux/err.h>
+#include <linux/mutex.h>
+#include <linux/delay.h>
+#include <linux/log2.h>
+#include <linux/pci.h>
+#include <linux/platform_device.h>
+
+#define DRVNAME "i5k_amb"
+
+#define I5K_REG_AMB_BASE_ADDR		0x48
+#define I5K_REG_AMB_LEN_ADDR		0x50
+#define I5K_REG_CHAN0_PRESENCE_ADDR	0x64
+#define I5K_REG_CHAN1_PRESENCE_ADDR	0x66
+
+#define AMB_REG_TEMP_MIN_ADDR		0x80
+#define AMB_REG_TEMP_MID_ADDR		0x81
+#define AMB_REG_TEMP_MAX_ADDR		0x82
+#define AMB_REG_TEMP_STATUS_ADDR	0x84
+#define AMB_REG_TEMP_ADDR		0x85
+
+#define AMB_CONFIG_SIZE			2048
+#define AMB_FUNC_3_OFFSET		768
+
+#define AMB_REG_TEMP_STATUS(amb)	((amb) * AMB_CONFIG_SIZE + \
+			AMB_FUNC_3_OFFSET + AMB_REG_TEMP_STATUS_ADDR)
+#define AMB_REG_TEMP_MIN(amb)		((amb) * AMB_CONFIG_SIZE + \
+			AMB_FUNC_3_OFFSET + AMB_REG_TEMP_MIN_ADDR)
+#define AMB_REG_TEMP_MID(amb)		((amb) * AMB_CONFIG_SIZE + \
+			AMB_FUNC_3_OFFSET + AMB_REG_TEMP_MID_ADDR)
+#define AMB_REG_TEMP_MAX(amb)		((amb) * AMB_CONFIG_SIZE + \
+			AMB_FUNC_3_OFFSET + AMB_REG_TEMP_MAX_ADDR)
+#define AMB_REG_TEMP(amb)		((amb) * AMB_CONFIG_SIZE + \
+			AMB_FUNC_3_OFFSET + AMB_REG_TEMP_ADDR)
+
+#define MAX_MEM_CHANNELS		4
+#define MAX_AMBS_PER_CHANNEL		16
+#define MAX_AMBS			(MAX_MEM_CHANNELS * \
+					 MAX_AMBS_PER_CHANNEL)
+/*
+ * Ugly hack: For some reason the highest bit is set if there
+ * are _any_ DIMMs in the channel.  Attempting to read from
+ * this "high-order" AMB results in a memory bus error, so
+ * for now we'll just ignore that top bit, even though that
+ * might prevent us from seeing the 16th DIMM in the channel.
+ */
+#define REAL_MAX_AMBS_PER_CHANNEL	15
+#define KNOBS_PER_AMB			5
+
+#define AMB_NUM_FROM_REG(byte_num, bit_num)	((byte_num) * \
+			MAX_AMBS_PER_CHANNEL) + (bit_num)

drivers/hwmon/i5k_amb.c line 86

Quoted 10 lines
drivers/hwmon/i5k_amb.c new file +10 -0
+
+#define AMB_SYSFS_NAME_LEN		16
+struct i5k_device_attribute {
+	struct sensor_device_attribute s_attr;
+	char name[AMB_SYSFS_NAME_LEN];
+};
+
+struct i5k_amb_data {
+	struct device *hwmon_dev;
+	struct resource *resource;

This member is unused.

Quoted 139 lines
drivers/hwmon/i5k_amb.c new file +139 -0
+
+	unsigned long amb_base;
+	unsigned long amb_len;
+	u16 amb_present[MAX_MEM_CHANNELS];
+	void __iomem *amb_mmio;
+	struct i5k_device_attribute *attrs;
+	unsigned int num_attrs;
+};
+
+static ssize_t show_name(struct device *dev, struct device_attribute *devattr,
+			 char *buf)
+{
+	return sprintf(buf, "%s\n", DRVNAME);
+}
+
+
+static DEVICE_ATTR(name, S_IRUGO, show_name, NULL);
+
+static struct platform_device *amb_pdev;
+
+static u8 amb_read_byte(struct i5k_amb_data *data, unsigned long offset)
+{
+	return ioread8(data->amb_mmio + offset);
+}
+
+static void amb_write_byte(struct i5k_amb_data *data, unsigned long offset,
+			   u8 val)
+{
+	iowrite8(val, data->amb_mmio + offset);
+}
+
+static ssize_t show_amb_alarm(struct device *dev,
+			     struct device_attribute *devattr,
+			     char *buf)
+{
+	struct sensor_device_attribute *attr = to_sensor_dev_attr(devattr);
+	struct i5k_amb_data *data = dev_get_drvdata(dev);
+
+	if (!(amb_read_byte(data, AMB_REG_TEMP_STATUS(attr->index)) & 0x20) &&
+	     (amb_read_byte(data, AMB_REG_TEMP_STATUS(attr->index)) & 0x8))
+		return sprintf(buf, "1\n");
+	else
+		return sprintf(buf, "0\n");
+}
+
+static ssize_t store_amb_min(struct device *dev,
+			     struct device_attribute *devattr,
+			     const char *buf,
+			     size_t count)
+{
+	struct sensor_device_attribute *attr = to_sensor_dev_attr(devattr);
+	struct i5k_amb_data *data = dev_get_drvdata(dev);
+	unsigned long temp = simple_strtoul(buf, NULL, 10) / 500;
+
+	if (temp > 255)
+		temp = 255;
+
+	amb_write_byte(data, AMB_REG_TEMP_MIN(attr->index), temp);
+	return count;
+}
+
+static ssize_t store_amb_mid(struct device *dev,
+			     struct device_attribute *devattr,
+			     const char *buf,
+			     size_t count)
+{
+	struct sensor_device_attribute *attr = to_sensor_dev_attr(devattr);
+	struct i5k_amb_data *data = dev_get_drvdata(dev);
+	unsigned long temp = simple_strtoul(buf, NULL, 10) / 500;
+
+	if (temp > 255)
+		temp = 255;
+
+	amb_write_byte(data, AMB_REG_TEMP_MID(attr->index), temp);
+	return count;
+}
+
+static ssize_t store_amb_max(struct device *dev,
+			     struct device_attribute *devattr,
+			     const char *buf,
+			     size_t count)
+{
+	struct sensor_device_attribute *attr = to_sensor_dev_attr(devattr);
+	struct i5k_amb_data *data = dev_get_drvdata(dev);
+	unsigned long temp = simple_strtoul(buf, NULL, 10) / 500;
+
+	if (temp > 255)
+		temp = 255;
+
+	amb_write_byte(data, AMB_REG_TEMP_MAX(attr->index), temp);
+	return count;
+}
+
+static ssize_t show_amb_min(struct device *dev,
+			     struct device_attribute *devattr,
+			     char *buf)
+{
+	struct sensor_device_attribute *attr = to_sensor_dev_attr(devattr);
+	struct i5k_amb_data *data = dev_get_drvdata(dev);
+	return sprintf(buf, "%d\n",
+		500 * amb_read_byte(data, AMB_REG_TEMP_MIN(attr->index)));
+}
+
+static ssize_t show_amb_mid(struct device *dev,
+			     struct device_attribute *devattr,
+			     char *buf)
+{
+	struct sensor_device_attribute *attr = to_sensor_dev_attr(devattr);
+	struct i5k_amb_data *data = dev_get_drvdata(dev);
+	return sprintf(buf, "%d\n",
+		500 * amb_read_byte(data, AMB_REG_TEMP_MID(attr->index)));
+}
+
+static ssize_t show_amb_max(struct device *dev,
+			     struct device_attribute *devattr,
+			     char *buf)
+{
+	struct sensor_device_attribute *attr = to_sensor_dev_attr(devattr);
+	struct i5k_amb_data *data = dev_get_drvdata(dev);
+	return sprintf(buf, "%d\n",
+		500 * amb_read_byte(data, AMB_REG_TEMP_MAX(attr->index)));
+}
+
+static ssize_t show_amb_temp(struct device *dev,
+			     struct device_attribute *devattr,
+			     char *buf)
+{
+	struct sensor_device_attribute *attr = to_sensor_dev_attr(devattr);
+	struct i5k_amb_data *data = dev_get_drvdata(dev);
+	return sprintf(buf, "%d\n",
+		500 * amb_read_byte(data, AMB_REG_TEMP(attr->index)));
+}
+
+static int __devinit i5k_amb_hwmon_init(struct platform_device *pdev)
+{
+	int i, j, k, d = 0;
+	u16 c;
+	int res = 0;
+	int num_ambs = 0;

drivers/hwmon/i5k_amb.c line 235

Quoted 10 lines
drivers/hwmon/i5k_amb.c new file +10 -0
+	struct i5k_amb_data *data = platform_get_drvdata(pdev);
+
+	/* Count the number of AMBs found */
+	for (i = 0; i < MAX_MEM_CHANNELS; i++) {
+		c = data->amb_present[i];
+		for (j = 0; j < REAL_MAX_AMBS_PER_CHANNEL; j++) {
+			if (c & 0x1)
+				num_ambs++;
+			c >>= 1;
+		}

You're counting set bits the hard way.

		/* ignore the high-order bit, see "Ugly hack" comment above */
		num_ambs += hweight16(data->amb_present[i] & 0x7fff);

drivers/hwmon/i5k_amb.c line 247

Quoted 12 lines
drivers/hwmon/i5k_amb.c new file +12 -0
+	}
+
+	/* Set up sysfs stuff */
+	data->attrs = kzalloc(sizeof(*data->attrs) * num_ambs * KNOBS_PER_AMB,
+				GFP_KERNEL);
+	if (!data->attrs)
+		return -ENOMEM;
+	data->num_attrs = 0;
+
+	for (i = 0; i < MAX_MEM_CHANNELS; i++) {
+		c = data->amb_present[i];
+		for (j = 0; j < REAL_MAX_AMBS_PER_CHANNEL; j++, c >>= 1) {

You could use ffs() here, but whatever, it's fine that way too.

drivers/hwmon/i5k_amb.c line 258

Quoted 11 lines
drivers/hwmon/i5k_amb.c new file +11 -0
+			struct i5k_device_attribute *iattr;
+
+			k = AMB_NUM_FROM_REG(i, j);
+			if (!(c & 0x1))
+				continue;
+			d++;
+
+			/* Temperature sysfs knob */
+			iattr = data->attrs + data->num_attrs;
+			snprintf(iattr->name, AMB_SYSFS_NAME_LEN,
+				 "temp%d_input", d);

Much better, thanks.

Quoted 153 lines
drivers/hwmon/i5k_amb.c new file +153 -0
+			iattr->s_attr.dev_attr.attr.name = iattr->name;
+			iattr->s_attr.dev_attr.attr.mode = S_IRUGO;
+			iattr->s_attr.dev_attr.show = show_amb_temp;
+			iattr->s_attr.index = k;
+			res = device_create_file(&pdev->dev,
+						 &iattr->s_attr.dev_attr);
+			if (res)
+				goto exit_remove;
+			data->num_attrs++;
+
+			/* Temperature min sysfs knob */
+			iattr = data->attrs + data->num_attrs;
+			snprintf(iattr->name, AMB_SYSFS_NAME_LEN,
+				 "temp%d_min", d);
+			iattr->s_attr.dev_attr.attr.name = iattr->name;
+			iattr->s_attr.dev_attr.attr.mode = S_IWUSR | S_IRUGO;
+			iattr->s_attr.dev_attr.show = show_amb_min;
+			iattr->s_attr.dev_attr.store = store_amb_min;
+			iattr->s_attr.index = k;
+			res = device_create_file(&pdev->dev,
+						 &iattr->s_attr.dev_attr);
+			if (res)
+				goto exit_remove;
+			data->num_attrs++;
+
+			/* Temperature mid sysfs knob */
+			iattr = data->attrs + data->num_attrs;
+			snprintf(iattr->name, AMB_SYSFS_NAME_LEN,
+				 "temp%d_mid", d);
+			iattr->s_attr.dev_attr.attr.name = iattr->name;
+			iattr->s_attr.dev_attr.attr.mode = S_IWUSR | S_IRUGO;
+			iattr->s_attr.dev_attr.show = show_amb_mid;
+			iattr->s_attr.dev_attr.store = store_amb_mid;
+			iattr->s_attr.index = k;
+			res = device_create_file(&pdev->dev,
+						 &iattr->s_attr.dev_attr);
+			if (res)
+				goto exit_remove;
+			data->num_attrs++;
+
+			/* Temperature max sysfs knob */
+			iattr = data->attrs + data->num_attrs;
+			snprintf(iattr->name, AMB_SYSFS_NAME_LEN,
+				 "temp%d_max", d);
+			iattr->s_attr.dev_attr.attr.name = iattr->name;
+			iattr->s_attr.dev_attr.attr.mode = S_IWUSR | S_IRUGO;
+			iattr->s_attr.dev_attr.show = show_amb_max;
+			iattr->s_attr.dev_attr.store = store_amb_max;
+			iattr->s_attr.index = k;
+			res = device_create_file(&pdev->dev,
+						 &iattr->s_attr.dev_attr);
+			if (res)
+				goto exit_remove;
+			data->num_attrs++;
+
+			/* Temperature alarm sysfs knob */
+			iattr = data->attrs + data->num_attrs;
+			snprintf(iattr->name, AMB_SYSFS_NAME_LEN,
+				 "temp%d_alarm", d);
+			iattr->s_attr.dev_attr.attr.name = iattr->name;
+			iattr->s_attr.dev_attr.attr.mode = S_IRUGO;
+			iattr->s_attr.dev_attr.show = show_amb_alarm;
+			iattr->s_attr.index = k;
+			res = device_create_file(&pdev->dev,
+						 &iattr->s_attr.dev_attr);
+			if (res)
+				goto exit_remove;
+			data->num_attrs++;
+		}
+	}
+
+	res = device_create_file(&pdev->dev, &dev_attr_name);
+	if (res)
+		goto exit_remove;
+
+	data->hwmon_dev = hwmon_device_register(&pdev->dev);
+	if (IS_ERR(data->hwmon_dev)) {
+		res = PTR_ERR(data->hwmon_dev);
+		goto exit_remove;
+	}
+
+	return res;
+
+exit_remove:
+	device_remove_file(&pdev->dev, &dev_attr_name);
+	for (i = 0; i < data->num_attrs; i++)
+		device_remove_file(&pdev->dev, &data->attrs[i].s_attr.dev_attr);
+	kfree(data->attrs);
+
+	return res;
+}
+
+static int __devinit i5k_amb_add(void)
+{
+	int res = -ENODEV;
+
+	/* only ever going to be one of these */
+	amb_pdev = platform_device_alloc(DRVNAME, 0);
+	if (!amb_pdev)
+		return -ENOMEM;
+
+	res = platform_device_add(amb_pdev);
+	if (res)
+		goto err;
+	return 0;
+
+err:
+	platform_device_put(amb_pdev);
+	return res;
+}
+
+static int __devinit i5k_find_amb_registers(struct i5k_amb_data *data)
+{
+	struct pci_dev *pcidev;
+	u32 val32;
+	int res = -ENODEV;
+
+	/* Find AMB register memory space */
+	pcidev = pci_get_device(PCI_VENDOR_ID_INTEL,
+				PCI_DEVICE_ID_INTEL_5000_ERR,
+				NULL);
+	if (!pcidev)
+		return -ENODEV;
+
+	if (pci_read_config_dword(pcidev, I5K_REG_AMB_BASE_ADDR, &val32))
+		goto out;
+	data->amb_base = val32;
+
+	if (pci_read_config_dword(pcidev, I5K_REG_AMB_LEN_ADDR, &val32))
+		goto out;
+	data->amb_len = val32;
+
+	/* Is it big enough? */
+	if (data->amb_len < AMB_CONFIG_SIZE * MAX_AMBS) {
+		printk(KERN_ERR DRVNAME "AMB region too small!\n");
+		goto out;
+	}
+
+	res = 0;
+out:
+	pci_dev_put(pcidev);
+	return res;
+}
+
+static int __devinit i5k_channel_probe(struct i5k_amb_data *data, int which)
+{
+	struct pci_dev *pcidev;
+	unsigned long dev_id;
+	u16 val16;
+	int res = -ENODEV;
+
+	/* Two memory channels per FBD PCI device */
+	switch (which) {

drivers/hwmon/i5k_amb.c line 421

Quoted 10 lines
drivers/hwmon/i5k_amb.c new file +10 -0
+	case 0:
+		dev_id = PCI_DEVICE_ID_INTEL_5000_FBD0;
+		break;
+	case 1:
+		dev_id = PCI_DEVICE_ID_INTEL_5000_FBD1;
+		break;
+	default:
+		return -EINVAL;
+	}
+

Awkward: see suggested patch at the bottom of this message.

Quoted 132 lines
drivers/hwmon/i5k_amb.c new file +127 -0
+	/* Copy the DIMM presence map for these two channels */
+	pcidev = pci_get_device(PCI_VENDOR_ID_INTEL, dev_id, NULL);
+	if (!pcidev)
+		return -ENODEV;
+
+	if (pci_read_config_word(pcidev, I5K_REG_CHAN0_PRESENCE_ADDR, &val16))
+		goto out;
+	data->amb_present[which * 2] = val16;
+
+	if (pci_read_config_word(pcidev, I5K_REG_CHAN1_PRESENCE_ADDR, &val16))
+		goto out;
+	data->amb_present[which * 2 + 1] = val16;
+
+	res = 0;
+
+out:
+	pci_dev_put(pcidev);
+	return res;
+}
+
+static int __devinit i5k_amb_probe(struct platform_device *pdev)
+{
+	struct i5k_amb_data *data;
+	struct resource *reso;
+	int res = -ENODEV;
+
+	data = kzalloc(sizeof(*data), GFP_KERNEL);
+	if (!data)
+		return -ENOMEM;
+
+	/* Figure out where the AMB registers live */
+	res = i5k_find_amb_registers(data);
+	if (res)
+		goto err;
+
+	/* Copy the DIMM presence map for the first two channels */
+	res = i5k_channel_probe(data, 0);
+	if (res)
+		goto err;
+
+	/* Copy the DIMM presence map for the optional second two channels */
+	i5k_channel_probe(data, 1);
+
+	/* Set up resource regions */
+	reso = request_mem_region(data->amb_base, data->amb_len, DRVNAME);
+	if (!reso) {
+		res = -EBUSY;
+		goto err;
+	}
+
+	data->amb_mmio = ioremap_nocache(data->amb_base, data->amb_len);
+	if (!data->amb_mmio) {
+		res = -EBUSY;
+		goto err_map_failed;
+	}
+
+	platform_set_drvdata(pdev, data);
+
+	res = i5k_amb_hwmon_init(pdev);
+	if (res)
+		goto err_init_failed;
+
+	return res;
+
+err_init_failed:
+	iounmap(data->amb_mmio);
+	platform_set_drvdata(pdev, NULL);
+err_map_failed:
+	release_mem_region(data->amb_base, data->amb_len);
+err:
+	kfree(data);
+	return res;
+}
+
+static int __devexit i5k_amb_remove(struct platform_device *pdev)
+{
+	int i;
+	struct i5k_amb_data *data = platform_get_drvdata(pdev);
+
+	hwmon_device_unregister(data->hwmon_dev);
+	device_remove_file(&pdev->dev, &dev_attr_name);
+	for (i = 0; i < data->num_attrs; i++)
+		device_remove_file(&pdev->dev, &data->attrs[i].s_attr.dev_attr);
+	kfree(data->attrs);
+	iounmap(data->amb_mmio);
+	release_mem_region(data->amb_base, data->amb_len);
+	platform_set_drvdata(pdev, NULL);
+	kfree(data);
+	return 0;
+}
+
+static struct platform_driver i5k_amb_driver = {
+	.driver = {
+		.owner = THIS_MODULE,
+		.name = DRVNAME,
+	},
+	.probe = i5k_amb_probe,
+	.remove = __devexit_p(i5k_amb_remove),
+};
+
+static int __init i5k_amb_init(void)
+{
+	int res;
+
+	res = platform_driver_register(&i5k_amb_driver);
+	if (res)
+		return res;
+
+	res = i5k_amb_add();
+	if (res)
+		platform_driver_unregister(&i5k_amb_driver);
+
+	return res;
+}
+
+static void __exit i5k_amb_exit(void)
+{
+	platform_device_unregister(amb_pdev);
+	platform_driver_unregister(&i5k_amb_driver);
+}
+
+MODULE_AUTHOR("Darrick J. Wong <[email protected]>");
+MODULE_DESCRIPTION("Intel 5000 chipset FB-DIMM AMB temperature sensor");
+MODULE_LICENSE("GPL");
+
+module_init(i5k_amb_init);
+module_exit(i5k_amb_exit);
include/linux/pci_ids.h +0 -0
index 55f307f..d641b2f 100644

include/linux/pci_ids.h line 2249

Quoted 9 lines
include/linux/pci_ids.h +3 -0
@@ -2241,6 +2241,9 @@
 #define PCI_DEVICE_ID_INTEL_82915G_IG	0x2582
 #define PCI_DEVICE_ID_INTEL_82915GM_HB	0x2590
 #define PCI_DEVICE_ID_INTEL_82915GM_IG	0x2592
+#define PCI_DEVICE_ID_INTEL_5000_ERR	0x25F0
+#define PCI_DEVICE_ID_INTEL_5000_FBD0	0x25F5
+#define PCI_DEVICE_ID_INTEL_5000_FBD1	0x25F6
 #define PCI_DEVICE_ID_INTEL_82945G_HB	0x2770
 #define PCI_DEVICE_ID_INTEL_82945G_IG	0x2772
 #define PCI_DEVICE_ID_INTEL_3000_HB	0x2778
drivers/hwmon/i5k_amb.c +7 -18
index 3bde717..65ecc56 100644
@@ -400,25 +400,12 @@ out:
 	return res;
 }
 
-static int __devinit i5k_channel_probe(struct i5k_amb_data *data, int which)
+static int __devinit i5k_channel_probe(u16 *amb_present, unsigned long dev_id)
 {
 	struct pci_dev *pcidev;
-	unsigned long dev_id;
 	u16 val16;
 	int res = -ENODEV;
 
-	/* Two memory channels per FBD PCI device */
-	switch (which) {
-	case 0:
-		dev_id = PCI_DEVICE_ID_INTEL_5000_FBD0;
-		break;
-	case 1:
-		dev_id = PCI_DEVICE_ID_INTEL_5000_FBD1;
-		break;
-	default:
-		return -EINVAL;
-	}
-
 	/* Copy the DIMM presence map for these two channels */
 	pcidev = pci_get_device(PCI_VENDOR_ID_INTEL, dev_id, NULL);
 	if (!pcidev)
@@ -426,11 +413,11 @@ static int __devinit i5k_channel_probe(struct i5k_amb_data *data, int which)
 
 	if (pci_read_config_word(pcidev, I5K_REG_CHAN0_PRESENCE_ADDR, &val16))
 		goto out;
-	data->amb_present[which * 2] = val16;
+	amb_present[0] = val16;
 
 	if (pci_read_config_word(pcidev, I5K_REG_CHAN1_PRESENCE_ADDR, &val16))
 		goto out;
-	data->amb_present[which * 2 + 1] = val16;
+	amb_present[1] = val16;
 
 	res = 0;
 
@@ -455,12 +442,14 @@ static int __devinit i5k_amb_probe(struct platform_device *pdev)
 		goto err;
 
 	/* Copy the DIMM presence map for the first two channels */
-	res = i5k_channel_probe(data, 0);
+	res = i5k_channel_probe(&data->amb_present[0],
+			PCI_DEVICE_ID_INTEL_5000_FBD0);
 	if (res)
 		goto err;
 
 	/* Copy the DIMM presence map for the optional second two channels */
-	i5k_channel_probe(data, 1);
+	i5k_channel_probe(&data->amb_present[2], 
+			PCI_DEVICE_ID_INTEL_5000_FBD1);
 
 	/* Set up resource regions */
 	reso = request_mem_region(data->amb_base, data->amb_len, DRVNAME);

Thanks & regards,

Signature · 8 lines
Mark M. Hoffman
[email protected]


_______________________________________________
lm-sensors mailing list
[email protected]
http://lists.lm-sensors.org/mailman/listinfo/lm-sensors
#11 lore raw

PATCH v4 1/1 i5k_amb: New memory temperature sensor driver

New driver to read FB-DIMM temperature sensors on systems with the Intel 5000 series chipsets.

---

 drivers/hwmon/Kconfig   |   10 +
 drivers/hwmon/Makefile  |    1 
 drivers/hwmon/i5k_amb.c |  531 +++++++++++++++++++++++++++++++++++++++++++++++
 include/linux/pci_ids.h |    3 
 4 files changed, 545 insertions(+), 0 deletions(-)
drivers/hwmon/Kconfig +10 -0
index b8854b9..c2acb37 100644
@@ -215,6 +215,16 @@ config SENSORS_DS1621
 	  This driver can also be built as a module.  If so, the module
 	  will be called ds1621.
 
+config SENSORS_I5K_AMB
+	tristate "FB-DIMM AMB temperature sensor on Intel 5000 series chipsets"
+	depends on PCI && EXPERIMENTAL
+	help
+	  If you say yes here you get support for FB-DIMM AMB temperature
+	  monitoring chips on systems with the Intel 5000 series chipset.
+
+	  This driver can also be built as a module. If so, the module
+	  will be called i5k_amb.
+
 config SENSORS_F71805F
 	tristate "Fintek F71805F/FG, F71806F/FG and F71872F/FG"
 	depends on EXPERIMENTAL
drivers/hwmon/Makefile +1 -0
index 2f592e2..914c48c 100644
@@ -37,6 +37,7 @@ obj-$(CONFIG_SENSORS_FSCPOS)	+= fscpos.o
 obj-$(CONFIG_SENSORS_GL518SM)	+= gl518sm.o
 obj-$(CONFIG_SENSORS_GL520SM)	+= gl520sm.o
 obj-$(CONFIG_SENSORS_HDAPS)	+= hdaps.o
+obj-$(CONFIG_SENSORS_I5K_AMB)	+= i5k_amb.o
 obj-$(CONFIG_SENSORS_IBMPEX)	+= ibmpex.o
 obj-$(CONFIG_SENSORS_IT87)	+= it87.o
 obj-$(CONFIG_SENSORS_K8TEMP)	+= k8temp.o
drivers/hwmon/i5k_amb.c new file +531 -0

531 lines not on this page · raw message

include/linux/pci_ids.h +3 -0
index 55f307f..d641b2f 100644
@@ -2241,6 +2241,9 @@
 #define PCI_DEVICE_ID_INTEL_82915G_IG	0x2582
 #define PCI_DEVICE_ID_INTEL_82915GM_HB	0x2590
 #define PCI_DEVICE_ID_INTEL_82915GM_IG	0x2592
+#define PCI_DEVICE_ID_INTEL_5000_ERR	0x25F0
+#define PCI_DEVICE_ID_INTEL_5000_FBD0	0x25F5
+#define PCI_DEVICE_ID_INTEL_5000_FBD1	0x25F6
 #define PCI_DEVICE_ID_INTEL_82945G_HB	0x2770
 #define PCI_DEVICE_ID_INTEL_82945G_IG	0x2772
 #define PCI_DEVICE_ID_INTEL_3000_HB	0x2778

_______________________________________________
lm-sensors mailing list
[email protected]
http://lists.lm-sensors.org/mailman/listinfo/lm-sensors

#12 lore raw

Re: [PATCH v4] i5k_amb: New memory temperature sensor driver

Hi Darrick:

Quoted 11 lines · Darrick J. Wong

New driver to read FB-DIMM temperature sensors on systems with the Intel 5000 series chipsets.

---

 drivers/hwmon/Kconfig   |   10 +
 drivers/hwmon/Makefile  |    1 
 drivers/hwmon/i5k_amb.c |  531 +++++++++++++++++++++++++++++++++++++++++++++++
 include/linux/pci_ids.h |    3 
 4 files changed, 545 insertions(+), 0 deletions(-)

Applied to hwmon-2.6.git/testing, thanks.

Signature · 8 lines
Mark M. Hoffman
[email protected]


_______________________________________________
lm-sensors mailing list
[email protected]
http://lists.lm-sensors.org/mailman/listinfo/lm-sensors

A message that is not in the archive (sent off-list or not yet received): [email protected]

#14 lore raw

PATCH 1/1 i5k_amb: Convert macros to C functions

akpm objected to some of the macros, so convert them into functions.

---

 drivers/hwmon/i5k_amb.c |   65 +++++++++++++++++++++++++++++++----------------
 1 files changed, 43 insertions(+), 22 deletions(-)
drivers/hwmon/i5k_amb.c +43 -22
index 7fdbe81..6ac5c6f 100644
@@ -47,16 +47,35 @@
 #define AMB_CONFIG_SIZE			2048
 #define AMB_FUNC_3_OFFSET		768
 
-#define AMB_REG_TEMP_STATUS(amb)	((amb) * AMB_CONFIG_SIZE + \
-			AMB_FUNC_3_OFFSET + AMB_REG_TEMP_STATUS_ADDR)
-#define AMB_REG_TEMP_MIN(amb)		((amb) * AMB_CONFIG_SIZE + \
-			AMB_FUNC_3_OFFSET + AMB_REG_TEMP_MIN_ADDR)
-#define AMB_REG_TEMP_MID(amb)		((amb) * AMB_CONFIG_SIZE + \
-			AMB_FUNC_3_OFFSET + AMB_REG_TEMP_MID_ADDR)
-#define AMB_REG_TEMP_MAX(amb)		((amb) * AMB_CONFIG_SIZE + \
-			AMB_FUNC_3_OFFSET + AMB_REG_TEMP_MAX_ADDR)
-#define AMB_REG_TEMP(amb)		((amb) * AMB_CONFIG_SIZE + \
-			AMB_FUNC_3_OFFSET + AMB_REG_TEMP_ADDR)
+static unsigned long amb_reg_temp_status(unsigned int amb)
+{
+	return AMB_FUNC_3_OFFSET + AMB_REG_TEMP_STATUS_ADDR +
+	       AMB_CONFIG_SIZE * amb;
+}
+
+static unsigned long amb_reg_temp_min(unsigned int amb)
+{
+	return AMB_FUNC_3_OFFSET + AMB_REG_TEMP_MIN_ADDR +
+	       AMB_CONFIG_SIZE * amb;
+}
+
+static unsigned long amb_reg_temp_mid(unsigned int amb)
+{
+	return AMB_FUNC_3_OFFSET + AMB_REG_TEMP_MID_ADDR +
+	       AMB_CONFIG_SIZE * amb;
+}
+
+static unsigned long amb_reg_temp_max(unsigned int amb)
+{
+	return AMB_FUNC_3_OFFSET + AMB_REG_TEMP_MAX_ADDR +
+	       AMB_CONFIG_SIZE * amb;
+}
+
+static unsigned long amb_reg_temp(unsigned int amb)
+{
+	return AMB_FUNC_3_OFFSET + AMB_REG_TEMP_ADDR +
+	       AMB_CONFIG_SIZE * amb;
+}
 
 #define MAX_MEM_CHANNELS		4
 #define MAX_AMBS_PER_CHANNEL		16
@@ -72,8 +91,10 @@
 #define REAL_MAX_AMBS_PER_CHANNEL	15
 #define KNOBS_PER_AMB			5
 
-#define AMB_NUM_FROM_REG(byte_num, bit_num)	((byte_num) * \
-			MAX_AMBS_PER_CHANNEL) + (bit_num)
+static unsigned long amb_num_from_reg(unsigned int byte_num, unsigned int bit)
+{
+	return byte_num * MAX_AMBS_PER_CHANNEL + bit;
+}
 
 #define AMB_SYSFS_NAME_LEN		16
 struct i5k_device_attribute {
@@ -121,8 +142,8 @@ static ssize_t show_amb_alarm(struct device *dev,
 	struct sensor_device_attribute *attr = to_sensor_dev_attr(devattr);
 	struct i5k_amb_data *data = dev_get_drvdata(dev);
 
-	if (!(amb_read_byte(data, AMB_REG_TEMP_STATUS(attr->index)) & 0x20) &&
-	     (amb_read_byte(data, AMB_REG_TEMP_STATUS(attr->index)) & 0x8))
+	if (!(amb_read_byte(data, amb_reg_temp_status(attr->index)) & 0x20) &&
+	     (amb_read_byte(data, amb_reg_temp_status(attr->index)) & 0x8))
 		return sprintf(buf, "1\n");
 	else
 		return sprintf(buf, "0\n");
@@ -140,7 +161,7 @@ static ssize_t store_amb_min(struct device *dev,
 	if (temp > 255)
 		temp = 255;
 
-	amb_write_byte(data, AMB_REG_TEMP_MIN(attr->index), temp);
+	amb_write_byte(data, amb_reg_temp_min(attr->index), temp);
 	return count;
 }
 
@@ -156,7 +177,7 @@ static ssize_t store_amb_mid(struct device *dev,
 	if (temp > 255)
 		temp = 255;
 
-	amb_write_byte(data, AMB_REG_TEMP_MID(attr->index), temp);
+	amb_write_byte(data, amb_reg_temp_mid(attr->index), temp);
 	return count;
 }
 
@@ -172,7 +193,7 @@ static ssize_t store_amb_max(struct device *dev,
 	if (temp > 255)
 		temp = 255;
 
-	amb_write_byte(data, AMB_REG_TEMP_MAX(attr->index), temp);
+	amb_write_byte(data, amb_reg_temp_max(attr->index), temp);
 	return count;
 }
 
@@ -183,7 +204,7 @@ static ssize_t show_amb_min(struct device *dev,
 	struct sensor_device_attribute *attr = to_sensor_dev_attr(devattr);
 	struct i5k_amb_data *data = dev_get_drvdata(dev);
 	return sprintf(buf, "%d\n",
-		500 * amb_read_byte(data, AMB_REG_TEMP_MIN(attr->index)));
+		500 * amb_read_byte(data, amb_reg_temp_min(attr->index)));
 }
 
 static ssize_t show_amb_mid(struct device *dev,
@@ -193,7 +214,7 @@ static ssize_t show_amb_mid(struct device *dev,
 	struct sensor_device_attribute *attr = to_sensor_dev_attr(devattr);
 	struct i5k_amb_data *data = dev_get_drvdata(dev);
 	return sprintf(buf, "%d\n",
-		500 * amb_read_byte(data, AMB_REG_TEMP_MID(attr->index)));
+		500 * amb_read_byte(data, amb_reg_temp_mid(attr->index)));
 }
 
 static ssize_t show_amb_max(struct device *dev,
@@ -203,7 +224,7 @@ static ssize_t show_amb_max(struct device *dev,
 	struct sensor_device_attribute *attr = to_sensor_dev_attr(devattr);
 	struct i5k_amb_data *data = dev_get_drvdata(dev);
 	return sprintf(buf, "%d\n",
-		500 * amb_read_byte(data, AMB_REG_TEMP_MAX(attr->index)));
+		500 * amb_read_byte(data, amb_reg_temp_max(attr->index)));
 }
 
 static ssize_t show_amb_temp(struct device *dev,
@@ -213,7 +234,7 @@ static ssize_t show_amb_temp(struct device *dev,
 	struct sensor_device_attribute *attr = to_sensor_dev_attr(devattr);
 	struct i5k_amb_data *data = dev_get_drvdata(dev);
 	return sprintf(buf, "%d\n",
-		500 * amb_read_byte(data, AMB_REG_TEMP(attr->index)));
+		500 * amb_read_byte(data, amb_reg_temp(attr->index)));
 }
 
 static int __devinit i5k_amb_hwmon_init(struct platform_device *pdev)
@@ -241,7 +262,7 @@ static int __devinit i5k_amb_hwmon_init(struct platform_device *pdev)
 		for (j = 0; j < REAL_MAX_AMBS_PER_CHANNEL; j++, c >>= 1) {
 			struct i5k_device_attribute *iattr;
 
-			k = AMB_NUM_FROM_REG(i, j);
+			k = amb_num_from_reg(i, j);
 			if (!(c & 0x1))
 				continue;
 			d++;

_______________________________________________
lm-sensors mailing list
[email protected]
http://lists.lm-sensors.org/mailman/listinfo/lm-sensors

#15 lore raw

Re: [PATCH] i5k_amb: Convert macros to C functions

Hi:

Quoted 7 lines · Darrick J. Wong

akpm objected to some of the macros, so convert them into functions.

---

 drivers/hwmon/i5k_amb.c |   65 +++++++++++++++++++++++++++++++----------------
 1 files changed, 43 insertions(+), 22 deletions(-)

Applied to hwmon-2.6.git/testing, thanks.

Signature · 8 lines
Mark M. Hoffman
[email protected]


_______________________________________________
lm-sensors mailing list
[email protected]
http://lists.lm-sensors.org/mailman/listinfo/lm-sensors
#16 lore raw

Re: [lm-sensors] [PATCH] v2 of Intel FB-DIMM AMB temperature sensors

In reply to Darrick J. Wong
Quoted 5 lines · Mark M. Hoffman

drivers/hwmon/Kconfig config SENSORS_DS1621 · line 220

Quoted 3 lines
drivers/hwmon/Kconfig +3 -0
+config SENSORS_I5K_AMB
+	tristate "FB-DIMM AMB temperature sensor on Intel 5000 series chipsets"
+	depends on I2C && EXPERIMENTAL

...and X86, presumably.

Come to think of it, PCI as well.

Quoted 8 lines

drivers/hwmon/i5k_amb.c line 40

Quoted 5 lines
drivers/hwmon/i5k_amb.c new file +5 -0
+#define PCI_VENDOR_INTEL		0x8086
+#define PCI_DEVICE_INTEL_5K_MEM		0x25F0
+#define PCI_DEVICE_INTEL_5K_FBD0	0x25F5
+#define PCI_DEVICE_INTEL_5K_FBD1	0x25F6
+

PCI_VENDOR_ID_INTEL is already defined. The rest should be added to include/linux/pci_ids.h, not here. You can do that in the same patch.

Will do.

Quoted 13 lines
Quoted 1 line

+static SENSOR_DEVICE_ATTR(name, S_IRUGO, show_name, NULL, 0);

Just DEVICE_ATTR is enough here.

Quoted 1 line

+static struct platform_device *amb_pdev = NULL;

Initialization to 0/NULL is unnecessary.

drivers/hwmon/i5k_amb.c line 108

Quoted 3 lines
drivers/hwmon/i5k_amb.c new file +3 -0
+static u8 amb_read_byte(struct i5k_amb_data *data, unsigned long offset)
+{
+	return (*(u8*)(data->amb_mmio + offset));
	return ioread8(data->amb_mmio + offset);
Quoted 8 lines

drivers/hwmon/i5k_amb.c line 114

Quoted 6 lines
drivers/hwmon/i5k_amb.c new file +6 -0
+}
+
+static void amb_write_byte(struct i5k_amb_data *data, unsigned long offset,
+			   u8 val)
+{
+	(*(u8*)(data->amb_mmio + offset)) = val;
	iowrite8(val, data->amb_mmio + offset);

Ack to all of these.

Quoted 11 lines

drivers/hwmon/i5k_amb.c line 145

Quoted 8 lines
drivers/hwmon/i5k_amb.c new file +8 -0
+	int temp = simple_strtol(buf, NULL, 10) / 500;
+
+	if (temp > 255)
+		temp = 255;
+
+	amb_write_byte(data, AMB_REG_TEMP_MIN(attr->index), temp);
+	return count;
+}

See Documentation/hwmon/sysfs_interface: sysfs attribute writes interpretation

Quoted 1 line

Summary: use long instead of int, and range check on both ends.

And these.

Quoted 3 lines

drivers/hwmon/i5k_amb.c line 255

Quoted 1 line
drivers/hwmon/i5k_amb.c new file +1 -0
+			sprintf(n, "temp%d_input", d);

Is it possible, at least in theory, for d > 99?

No. The highest that d will ever get is MAX_MEM_CHANNELS * REAL_MAX_AMBS_PER_CHANNEL, which is 60. As I recall, the specs for the 5000 chipset also say that there are 4 channels with a maximum of 16 DIMMs per channel, so the most you'd ever see on a system is 64 DIMMs.

Quoted 8 lines

drivers/hwmon/i5k_amb.c line 341

Quoted 5 lines
drivers/hwmon/i5k_amb.c new file +5 -0
+	for (i = 0; i < data->num_attrs; i++) {
+		device_remove_file(&pdev->dev, &data->attrs[i].dev_attr);
+		kfree(data->attrs[i].dev_attr.attr.name);
+	}
+	kfree(data->attrs);

Hmm, would be nice if data->attrs included space for all the strings, too. Then you could kmalloc/kfree in one shot.

Good idea.

Quoted 10 lines

drivers/hwmon/i5k_amb.c line 420

Quoted 6 lines
drivers/hwmon/i5k_amb.c new file +6 -0
+	/* Copy the DIMM presence map for the optional second two channels */
+	pcidev = pci_get_device(PCI_VENDOR_INTEL, PCI_DEVICE_INTEL_5K_FBD1,
+				NULL);
+	if (!pcidev)
+		goto setup_resources;
+

Abuse of goto. OK, maybe it's nit-picky... the accepted idiom in Linux is to use goto to unwind inits after an error. This is not an error condition, *and* this goto is easily replaced by an if statement.

Sorry for the spaghetti code. :)

Thanks for the detailed review! I'll correct the problems that you pointed out in the next revision of the patch.

--D

_______________________________________________
lm-sensors mailing list
[email protected]
http://lists.lm-sensors.org/mailman/listinfo/lm-sensors