* [BK PATCH] i2c driver changes for 2.5.72
@ 2005-05-19 6:24 ` Greg KH
0 siblings, 0 replies; 32+ messages in thread
From: Greg KH @ 2003-06-18 18:24 UTC (permalink / raw)
To: torvalds; +Cc: linux-kernel, sensors
Hi,
Here are some i2c driver changes for 2.5.72. These consist of some
more bug fixes, and the addition of a lm78 i2c driver.
Please pull from: bk://kernel.bkbits.net/gregkh/linux/i2c-2.5
thanks,
greg k-h
drivers/i2c/busses/i2c-ali15x3.c | 18
drivers/i2c/busses/i2c-i801.c | 10
drivers/i2c/chips/Kconfig | 14
drivers/i2c/chips/Makefile | 1
drivers/i2c/chips/adm1021.c | 18
drivers/i2c/chips/lm78.c | 895 +++++++++++++++++++++++++++++++++++++++
drivers/i2c/chips/lm85.c | 41 +
drivers/i2c/chips/w83781d.c | 86 ++-
8 files changed, 1019 insertions(+), 64 deletions(-)
-----
<margitsw:t-online.de>:
o I2C: Sensors patch for adm1021
o I2C: lm85 fixups
<mhoffman:lightlink.com>:
o I2C: w83781d bugfix
o i2c: Add lm78 sensor chip support
Greg Kroah-Hartman:
o I2C: fix resource leak in i2c-ali15x3.c
o I2C: add lm78 chip to Makefile
Martin Schlemmer:
o I2C: fix for previous W83627THF sensor chip patch
o I2C: ICH5 SMBus and W83627THF additions
^ permalink raw reply [flat|nested] 32+ messages in thread
* [PATCH] i2c driver changes for 2.5.72
2005-05-19 6:24 ` Greg KH
@ 2005-05-19 6:24 ` Greg KH
-1 siblings, 0 replies; 32+ messages in thread
From: Greg KH @ 2003-06-18 18:25 UTC (permalink / raw)
To: linux-kernel, sensors
ChangeSet 1.1318.3.1, 2003/06/16 11:31:30-07:00, mhoffman@lightlink.com
[PATCH] i2c: Add lm78 sensor chip support
This patch vs. 2.5.70 adds support for LM78, LM78-J, and LM79 sensors
chips based on lm_sensors project CVS. This works on one of my boards.
I want to draw attention to something I did with this driver by
comparing it to it87.c in 2.5.70:
> #define IT87_INIT_TEMP_HIGH_1 600
> #define IT87_INIT_TEMP_LOW_1 200
The hardware uses degrees C, and sysfs uses degrees C * 1000. But
these #defines are apparently in units of degrees C * 10. This
arbitrary intermediate representation bugs me. And given the new 2.5
sysfs standard, it's unnecessary.
In this patch for lm78, I rewrote the conversion routines in terms
of the sysfs units - getting rid of the intermediate nonsense. If
there are no objections, I'm going to start passing patches to do
this to the other sensor chip drivers in 2.5 as well. It would be
nice to get some help with this too... especially since I don't
have all that hardware at hand to test the results.
drivers/i2c/chips/Kconfig | 14
drivers/i2c/chips/lm78.c | 895 ++++++++++++++++++++++++++++++++++++++++++++++
2 files changed, 909 insertions(+)
diff -Nru a/drivers/i2c/chips/Kconfig b/drivers/i2c/chips/Kconfig
--- a/drivers/i2c/chips/Kconfig Wed Jun 18 11:19:49 2003
+++ b/drivers/i2c/chips/Kconfig Wed Jun 18 11:19:49 2003
@@ -62,6 +62,20 @@
in the lm_sensors package, which you can download at
http://www.lm-sensors.nu
+config SENSORS_LM78
+ tristate " National Semiconductors LM78 and compatibles"
+ depends on I2C && EXPERIMENTAL
+ help
+ If you say yes here you get support for National Semiconductor LM78,
+ LM78-J and LM79. This can also be built as a module which can be
+ inserted and removed while the kernel is running.
+
+ The module will be called lm78.
+
+ You will also need the latest user-space utilties: you can find them
+ in the lm_sensors package, which you can download at
+ http://www.lm-sensors.nu
+
config SENSORS_VIA686A
tristate " VIA686A"
depends on I2C && EXPERIMENTAL
diff -Nru a/drivers/i2c/chips/lm78.c b/drivers/i2c/chips/lm78.c
--- /dev/null Wed Dec 31 16:00:00 1969
+++ b/drivers/i2c/chips/lm78.c Wed Jun 18 11:19:49 2003
@@ -0,0 +1,895 @@
+/*
+ lm78.c - Part of lm_sensors, Linux kernel modules for hardware
+ monitoring
+ Copyright (c) 1998, 1999 Frodo Looijaard <frodol@dds.nl>
+
+ 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., 675 Mass Ave, Cambridge, MA 02139, USA.
+*/
+
+#include <linux/module.h>
+#include <linux/init.h>
+#include <linux/slab.h>
+#include <linux/i2c.h>
+#include <linux/i2c-sensor.h>
+#include <asm/io.h>
+
+/* Addresses to scan */
+static unsigned short normal_i2c[] = { I2C_CLIENT_END };
+static unsigned short normal_i2c_range[] = { 0x20, 0x2f, I2C_CLIENT_END };
+static unsigned int normal_isa[] = { 0x0290, I2C_CLIENT_ISA_END };
+static unsigned int normal_isa_range[] = { I2C_CLIENT_ISA_END };
+
+/* Insmod parameters */
+SENSORS_INSMOD_3(lm78, lm78j, lm79);
+
+/* Many LM78 constants specified below */
+
+/* Length of ISA address segment */
+#define LM78_EXTENT 8
+
+/* Where are the ISA address/data registers relative to the base address */
+#define LM78_ADDR_REG_OFFSET 5
+#define LM78_DATA_REG_OFFSET 6
+
+/* The LM78 registers */
+#define LM78_REG_IN_MAX(nr) (0x2b + (nr) * 2)
+#define LM78_REG_IN_MIN(nr) (0x2c + (nr) * 2)
+#define LM78_REG_IN(nr) (0x20 + (nr))
+
+#define LM78_REG_FAN_MIN(nr) (0x3b + (nr))
+#define LM78_REG_FAN(nr) (0x28 + (nr))
+
+#define LM78_REG_TEMP 0x27
+#define LM78_REG_TEMP_OVER 0x39
+#define LM78_REG_TEMP_HYST 0x3a
+
+#define LM78_REG_ALARM1 0x41
+#define LM78_REG_ALARM2 0x42
+
+#define LM78_REG_VID_FANDIV 0x47
+
+#define LM78_REG_CONFIG 0x40
+#define LM78_REG_CHIPID 0x49
+#define LM78_REG_I2C_ADDR 0x48
+
+
+/* Conversions. Rounding and limit checking is only done on the TO_REG
+ variants. */
+
+/* IN: mV, (0V to 4.08V)
+ REG: 16mV/bit */
+static inline u8 IN_TO_REG(unsigned long val)
+{
+ unsigned long nval = SENSORS_LIMIT(val, 0, 4080);
+ return (nval + 8) / 16;
+}
+#define IN_FROM_REG(val) ((val) * 16)
+
+static inline u8 FAN_TO_REG(long rpm, int div)
+{
+ if (rpm == 0)
+ return 255;
+ rpm = SENSORS_LIMIT(rpm, 1, 1000000);
+ return SENSORS_LIMIT((1350000 + rpm * div / 2) / (rpm * div), 1, 254);
+}
+
+static inline int FAN_FROM_REG(u8 val, int div)
+{
+ return val==0 ? -1 : val==255 ? 0 : 1350000/(val*div);
+}
+
+/* TEMP: mC (-128C to +127C)
+ REG: 1C/bit, two's complement */
+static inline u8 TEMP_TO_REG(int val)
+{
+ int nval = SENSORS_LIMIT(val, -128000, 127000) ;
+ return nval<0 ? (nval-500)/1000+0x100 : (nval+500)/1000;
+}
+
+static inline int TEMP_FROM_REG(u8 val)
+{
+ return (val>=0x80 ? val-0x100 : val) * 1000;
+}
+
+/* VID: mV
+ REG: (see doc/vid) */
+static inline int VID_FROM_REG(u8 val)
+{
+ return val==0x1f ? 0 : val>=0x10 ? 5100-val*100 : 2050-val*50;
+}
+
+/* ALARMS: chip-specific bitmask
+ REG: (same) */
+#define ALARMS_FROM_REG(val) (val)
+
+/* FAN DIV: 1, 2, 4, or 8 (defaults to 2)
+ REG: 0, 1, 2, or 3 (respectively) (defaults to 1) */
+static inline u8 DIV_TO_REG(int val)
+{
+ return val==8 ? 3 : val==4 ? 2 : val==1 ? 0 : 1;
+}
+#define DIV_FROM_REG(val) (1 << (val))
+
+/* Initial limits. To keep them sane, we use the 'standard' translation as
+ specified in the LM78 sheet. Use the config file to set better limits. */
+#define LM78_INIT_IN_0(vid) ((vid)==3500 ? 2800 : (vid))
+#define LM78_INIT_IN_1(vid) ((vid)==3500 ? 2800 : (vid))
+#define LM78_INIT_IN_2 3300
+#define LM78_INIT_IN_3 (((5000) * 100)/168)
+#define LM78_INIT_IN_4 (((12000) * 10)/38)
+#define LM78_INIT_IN_5 (((-12000) * -604)/2100)
+#define LM78_INIT_IN_6 (((-5000) * -604)/909)
+
+#define LM78_INIT_IN_PERCENTAGE 10
+
+#define LM78_INIT_IN_MIN_0(vid) (LM78_INIT_IN_0(vid) - \
+ LM78_INIT_IN_0(vid) * LM78_INIT_IN_PERCENTAGE / 100)
+#define LM78_INIT_IN_MAX_0(vid) (LM78_INIT_IN_0(vid) + \
+ LM78_INIT_IN_0(vid) * LM78_INIT_IN_PERCENTAGE / 100)
+#define LM78_INIT_IN_MIN_1(vid) (LM78_INIT_IN_1(vid) - \
+ LM78_INIT_IN_1(vid) * LM78_INIT_IN_PERCENTAGE / 100)
+#define LM78_INIT_IN_MAX_1(vid) (LM78_INIT_IN_1(vid) + \
+ LM78_INIT_IN_1(vid) * LM78_INIT_IN_PERCENTAGE / 100)
+
+#define LM78_INIT_IN_MIN_2 \
+ (LM78_INIT_IN_2 - LM78_INIT_IN_2 * LM78_INIT_IN_PERCENTAGE / 100)
+#define LM78_INIT_IN_MAX_2 \
+ (LM78_INIT_IN_2 + LM78_INIT_IN_2 * LM78_INIT_IN_PERCENTAGE / 100)
+#define LM78_INIT_IN_MIN_3 \
+ (LM78_INIT_IN_3 - LM78_INIT_IN_3 * LM78_INIT_IN_PERCENTAGE / 100)
+#define LM78_INIT_IN_MAX_3 \
+ (LM78_INIT_IN_3 + LM78_INIT_IN_3 * LM78_INIT_IN_PERCENTAGE / 100)
+#define LM78_INIT_IN_MIN_4 \
+ (LM78_INIT_IN_4 - LM78_INIT_IN_4 * LM78_INIT_IN_PERCENTAGE / 100)
+#define LM78_INIT_IN_MAX_4 \
+ (LM78_INIT_IN_4 + LM78_INIT_IN_4 * LM78_INIT_IN_PERCENTAGE / 100)
+#define LM78_INIT_IN_MIN_5 \
+ (LM78_INIT_IN_5 - LM78_INIT_IN_5 * LM78_INIT_IN_PERCENTAGE / 100)
+#define LM78_INIT_IN_MAX_5 \
+ (LM78_INIT_IN_5 + LM78_INIT_IN_5 * LM78_INIT_IN_PERCENTAGE / 100)
+#define LM78_INIT_IN_MIN_6 \
+ (LM78_INIT_IN_6 - LM78_INIT_IN_6 * LM78_INIT_IN_PERCENTAGE / 100)
+#define LM78_INIT_IN_MAX_6 \
+ (LM78_INIT_IN_6 + LM78_INIT_IN_6 * LM78_INIT_IN_PERCENTAGE / 100)
+
+#define LM78_INIT_FAN_MIN_1 3000
+#define LM78_INIT_FAN_MIN_2 3000
+#define LM78_INIT_FAN_MIN_3 3000
+
+#define LM78_INIT_TEMP_OVER 60000
+#define LM78_INIT_TEMP_HYST 50000
+
+/* There are some complications in a module like this. First off, LM78 chips
+ may be both present on the SMBus and the ISA bus, and we have to handle
+ those cases separately at some places. Second, there might be several
+ LM78 chips available (well, actually, that is probably never done; but
+ it is a clean illustration of how to handle a case like that). Finally,
+ a specific chip may be attached to *both* ISA and SMBus, and we would
+ not like to detect it double. Fortunately, in the case of the LM78 at
+ least, a register tells us what SMBus address we are on, so that helps
+ a bit - except if there could be more than one SMBus. Groan. No solution
+ for this yet. */
+
+/* This module may seem overly long and complicated. In fact, it is not so
+ bad. Quite a lot of bookkeeping is done. A real driver can often cut
+ some corners. */
+
+/* For each registered LM78, we need to keep some data in memory. That
+ data is pointed to by lm78_list[NR]->data. The structure itself is
+ dynamically allocated, at the same time when a new lm78 client is
+ allocated. */
+struct lm78_data {
+ struct semaphore lock;
+ enum chips type;
+
+ struct semaphore update_lock;
+ char valid; /* !=0 if following fields are valid */
+ unsigned long last_updated; /* In jiffies */
+
+ u8 in[7]; /* Register value */
+ u8 in_max[7]; /* Register value */
+ u8 in_min[7]; /* Register value */
+ u8 fan[3]; /* Register value */
+ u8 fan_min[3]; /* Register value */
+ u8 temp; /* Register value */
+ u8 temp_over; /* Register value */
+ u8 temp_hyst; /* Register value */
+ u8 fan_div[3]; /* Register encoding, shifted right */
+ u8 vid; /* Register encoding, combined */
+ u16 alarms; /* Register encoding, combined */
+};
+
+
+static int lm78_attach_adapter(struct i2c_adapter *adapter);
+static int lm78_detect(struct i2c_adapter *adapter, int address, int kind);
+static int lm78_detach_client(struct i2c_client *client);
+
+static int lm78_read_value(struct i2c_client *client, u8 register);
+static int lm78_write_value(struct i2c_client *client, u8 register, u8 value);
+static void lm78_update_client(struct i2c_client *client);
+static void lm78_init_client(struct i2c_client *client);
+
+
+static struct i2c_driver lm78_driver = {
+ .owner = THIS_MODULE,
+ .name = "lm78",
+ .id = I2C_DRIVERID_LM78,
+ .flags = I2C_DF_NOTIFY,
+ .attach_adapter = lm78_attach_adapter,
+ .detach_client = lm78_detach_client,
+};
+
+/* 7 Voltages */
+static ssize_t show_in(struct device *dev, char *buf, int nr)
+{
+ struct i2c_client *client = to_i2c_client(dev);
+ struct lm78_data *data = i2c_get_clientdata(client);
+ lm78_update_client(client);
+ return sprintf(buf, "%d\n", IN_FROM_REG(data->in[nr]));
+}
+
+static ssize_t show_in_min(struct device *dev, char *buf, int nr)
+{
+ struct i2c_client *client = to_i2c_client(dev);
+ struct lm78_data *data = i2c_get_clientdata(client);
+ lm78_update_client(client);
+ return sprintf(buf, "%d\n", IN_FROM_REG(data->in_min[nr]));
+}
+
+static ssize_t show_in_max(struct device *dev, char *buf, int nr)
+{
+ struct i2c_client *client = to_i2c_client(dev);
+ struct lm78_data *data = i2c_get_clientdata(client);
+ lm78_update_client(client);
+ return sprintf(buf, "%d\n", IN_FROM_REG(data->in_max[nr]));
+}
+
+static ssize_t set_in_min(struct device *dev, const char *buf,
+ size_t count, int nr)
+{
+ struct i2c_client *client = to_i2c_client(dev);
+ struct lm78_data *data = i2c_get_clientdata(client);
+ unsigned long val = simple_strtoul(buf, NULL, 10);
+ data->in_min[nr] = IN_TO_REG(val);
+ lm78_write_value(client, LM78_REG_IN_MIN(nr), data->in_min[nr]);
+ return count;
+}
+
+static ssize_t set_in_max(struct device *dev, const char *buf,
+ size_t count, int nr)
+{
+ struct i2c_client *client = to_i2c_client(dev);
+ struct lm78_data *data = i2c_get_clientdata(client);
+ unsigned long val = simple_strtoul(buf, NULL, 10);
+ data->in_max[nr] = IN_TO_REG(val);
+ lm78_write_value(client, LM78_REG_IN_MAX(nr), data->in_max[nr]);
+ return count;
+}
+
+#define show_in_offset(offset) \
+static ssize_t \
+ show_in##offset (struct device *dev, char *buf) \
+{ \
+ return show_in(dev, buf, 0x##offset); \
+} \
+static DEVICE_ATTR(in_input##offset, S_IRUGO, \
+ show_in##offset, NULL) \
+static ssize_t \
+ show_in##offset##_min (struct device *dev, char *buf) \
+{ \
+ return show_in_min(dev, buf, 0x##offset); \
+} \
+static ssize_t \
+ show_in##offset##_max (struct device *dev, char *buf) \
+{ \
+ return show_in_max(dev, buf, 0x##offset); \
+} \
+static ssize_t set_in##offset##_min (struct device *dev, \
+ const char *buf, size_t count) \
+{ \
+ return set_in_min(dev, buf, count, 0x##offset); \
+} \
+static ssize_t set_in##offset##_max (struct device *dev, \
+ const char *buf, size_t count) \
+{ \
+ return set_in_max(dev, buf, count, 0x##offset); \
+} \
+static DEVICE_ATTR(in_min##offset, S_IRUGO | S_IWUSR, \
+ show_in##offset##_min, set_in##offset##_min) \
+static DEVICE_ATTR(in_max##offset, S_IRUGO | S_IWUSR, \
+ show_in##offset##_max, set_in##offset##_max)
+
+show_in_offset(0);
+show_in_offset(1);
+show_in_offset(2);
+show_in_offset(3);
+show_in_offset(4);
+show_in_offset(5);
+show_in_offset(6);
+
+/* Temperature */
+static ssize_t show_temp(struct device *dev, char *buf)
+{
+ struct i2c_client *client = to_i2c_client(dev);
+ struct lm78_data *data = i2c_get_clientdata(client);
+ lm78_update_client(client);
+ return sprintf(buf, "%d\n", TEMP_FROM_REG(data->temp));
+}
+
+static ssize_t show_temp_over(struct device *dev, char *buf)
+{
+ struct i2c_client *client = to_i2c_client(dev);
+ struct lm78_data *data = i2c_get_clientdata(client);
+ lm78_update_client(client);
+ return sprintf(buf, "%d\n", TEMP_FROM_REG(data->temp_over));
+}
+
+static ssize_t set_temp_over(struct device *dev, const char *buf, size_t count)
+{
+ struct i2c_client *client = to_i2c_client(dev);
+ struct lm78_data *data = i2c_get_clientdata(client);
+ long val = simple_strtol(buf, NULL, 10);
+ data->temp_over = TEMP_TO_REG(val);
+ lm78_write_value(client, LM78_REG_TEMP_OVER, data->temp_over);
+ return count;
+}
+
+static ssize_t show_temp_hyst(struct device *dev, char *buf)
+{
+ struct i2c_client *client = to_i2c_client(dev);
+ struct lm78_data *data = i2c_get_clientdata(client);
+ lm78_update_client(client);
+ return sprintf(buf, "%d\n", TEMP_FROM_REG(data->temp_hyst));
+}
+
+static ssize_t set_temp_hyst(struct device *dev, const char *buf, size_t count)
+{
+ struct i2c_client *client = to_i2c_client(dev);
+ struct lm78_data *data = i2c_get_clientdata(client);
+ long val = simple_strtol(buf, NULL, 10);
+ data->temp_hyst = TEMP_TO_REG(val);
+ lm78_write_value(client, LM78_REG_TEMP_HYST, data->temp_hyst);
+ return count;
+}
+
+static DEVICE_ATTR(temp_input, S_IRUGO, show_temp, NULL)
+static DEVICE_ATTR(temp_max, S_IRUGO | S_IWUSR,
+ show_temp_over, set_temp_over)
+static DEVICE_ATTR(temp_min, S_IRUGO | S_IWUSR,
+ show_temp_hyst, set_temp_hyst)
+
+/* 3 Fans */
+static ssize_t show_fan(struct device *dev, char *buf, int nr)
+{
+ struct i2c_client *client = to_i2c_client(dev);
+ struct lm78_data *data = i2c_get_clientdata(client);
+ lm78_update_client(client);
+ return sprintf(buf, "%d\n", FAN_FROM_REG(data->fan[nr],
+ DIV_FROM_REG(data->fan_div[nr])) );
+}
+
+static ssize_t show_fan_min(struct device *dev, char *buf, int nr)
+{
+ struct i2c_client *client = to_i2c_client(dev);
+ struct lm78_data *data = i2c_get_clientdata(client);
+ lm78_update_client(client);
+ return sprintf(buf,"%d\n", FAN_FROM_REG(data->fan_min[nr],
+ DIV_FROM_REG(data->fan_div[nr])) );
+}
+
+static ssize_t set_fan_min(struct device *dev, const char *buf,
+ size_t count, int nr)
+{
+ struct i2c_client *client = to_i2c_client(dev);
+ struct lm78_data *data = i2c_get_clientdata(client);
+ unsigned long val = simple_strtoul(buf, NULL, 10);
+ data->fan_min[nr] = FAN_TO_REG(val, DIV_FROM_REG(data->fan_div[nr]));
+ lm78_write_value(client, LM78_REG_FAN_MIN(nr), data->fan_min[nr]);
+ return count;
+}
+
+static ssize_t show_fan_div(struct device *dev, char *buf, int nr)
+{
+ struct i2c_client *client = to_i2c_client(dev);
+ struct lm78_data *data = i2c_get_clientdata(client);
+ lm78_update_client(client);
+ return sprintf(buf, "%d\n", DIV_FROM_REG(data->fan_div[nr]) );
+}
+
+/* Note: we save and restore the fan minimum here, because its value is
+ determined in part by the fan divisor. This follows the principle of
+ least suprise; the user doesn't expect the fan minimum to change just
+ because the divisor changed. */
+static ssize_t set_fan_div(struct device *dev, const char *buf,
+ size_t count, int nr)
+{
+ struct i2c_client *client = to_i2c_client(dev);
+ struct lm78_data *data = i2c_get_clientdata(client);
+ unsigned long min = FAN_FROM_REG(data->fan_min[nr],
+ DIV_FROM_REG(data->fan_div[nr]));
+ unsigned long val = simple_strtoul(buf, NULL, 10);
+ int reg = lm78_read_value(client, LM78_REG_VID_FANDIV);
+ data->fan_div[nr] = DIV_TO_REG(val);
+ switch (nr) {
+ case 0:
+ reg = (reg & 0xcf) | (data->fan_div[nr] << 4);
+ break;
+ case 1:
+ reg = (reg & 0x3f) | (data->fan_div[nr] << 6);
+ break;
+ }
+ lm78_write_value(client, LM78_REG_VID_FANDIV, reg);
+ data->fan_min[nr] =
+ FAN_TO_REG(min, DIV_FROM_REG(data->fan_div[nr]));
+ lm78_write_value(client, LM78_REG_FAN_MIN(nr), data->fan_min[nr]);
+ return count;
+}
+
+#define show_fan_offset(offset) \
+static ssize_t show_fan_##offset (struct device *dev, char *buf) \
+{ \
+ return show_fan(dev, buf, 0x##offset - 1); \
+} \
+static ssize_t show_fan_##offset##_min (struct device *dev, char *buf) \
+{ \
+ return show_fan_min(dev, buf, 0x##offset - 1); \
+} \
+static ssize_t show_fan_##offset##_div (struct device *dev, char *buf) \
+{ \
+ return show_fan_div(dev, buf, 0x##offset - 1); \
+} \
+static ssize_t set_fan_##offset##_min (struct device *dev, \
+ const char *buf, size_t count) \
+{ \
+ return set_fan_min(dev, buf, count, 0x##offset - 1); \
+} \
+static DEVICE_ATTR(fan_input##offset, S_IRUGO, show_fan_##offset, NULL) \
+static DEVICE_ATTR(fan_min##offset, S_IRUGO | S_IWUSR, \
+ show_fan_##offset##_min, set_fan_##offset##_min)
+
+static ssize_t set_fan_1_div(struct device *dev, const char *buf,
+ size_t count)
+{
+ return set_fan_div(dev, buf, count, 0) ;
+}
+
+static ssize_t set_fan_2_div(struct device *dev, const char *buf,
+ size_t count)
+{
+ return set_fan_div(dev, buf, count, 1) ;
+}
+
+show_fan_offset(1);
+show_fan_offset(2);
+show_fan_offset(3);
+
+/* Fan 3 divisor is locked in H/W */
+static DEVICE_ATTR(fan_div1, S_IRUGO | S_IWUSR,
+ show_fan_1_div, set_fan_1_div)
+static DEVICE_ATTR(fan_div2, S_IRUGO | S_IWUSR,
+ show_fan_2_div, set_fan_2_div)
+static DEVICE_ATTR(fan_div3, S_IRUGO, show_fan_3_div, NULL)
+
+/* VID */
+static ssize_t show_vid(struct device *dev, char *buf)
+{
+ struct i2c_client *client = to_i2c_client(dev);
+ struct lm78_data *data = i2c_get_clientdata(client);
+ lm78_update_client(client);
+ return sprintf(buf, "%d\n", VID_FROM_REG(data->vid));
+}
+static DEVICE_ATTR(vid, S_IRUGO, show_vid, NULL);
+
+/* Alarms */
+static ssize_t show_alarms(struct device *dev, char *buf)
+{
+ struct i2c_client *client = to_i2c_client(dev);
+ struct lm78_data *data = i2c_get_clientdata(client);
+ lm78_update_client(client);
+ return sprintf(buf, "%d\n", ALARMS_FROM_REG(data->alarms));
+}
+static DEVICE_ATTR(alarms, S_IRUGO, show_alarms, NULL);
+
+/* This function is called when:
+ * lm78_driver is inserted (when this module is loaded), for each
+ available adapter
+ * when a new adapter is inserted (and lm78_driver is still present) */
+static int lm78_attach_adapter(struct i2c_adapter *adapter)
+{
+ if (!(adapter->class & I2C_ADAP_CLASS_SMBUS))
+ return 0;
+ return i2c_detect(adapter, &addr_data, lm78_detect);
+}
+
+/* This function is called by i2c_detect */
+int lm78_detect(struct i2c_adapter *adapter, int address, int kind)
+{
+ int i, err;
+ struct i2c_client *new_client;
+ struct lm78_data *data;
+ const char *client_name = "";
+ int is_isa = i2c_is_isa_adapter(adapter);
+
+ if (!is_isa &&
+ !i2c_check_functionality(adapter, I2C_FUNC_SMBUS_BYTE_DATA)) {
+ err = -ENODEV;
+ goto ERROR0;
+ }
+
+ /* Reserve the ISA region */
+ if (is_isa)
+ if (!request_region(address, LM78_EXTENT, "lm78")) {
+ err = -EBUSY;
+ goto ERROR0;
+ }
+
+ /* Probe whether there is anything available on this address. Already
+ done for SMBus clients */
+ if (kind < 0) {
+ if (is_isa) {
+
+#define REALLY_SLOW_IO
+ /* We need the timeouts for at least some LM78-like
+ chips. But only if we read 'undefined' registers. */
+ i = inb_p(address + 1);
+ if (inb_p(address + 2) != i) {
+ err = -ENODEV;
+ goto ERROR1;
+ }
+ if (inb_p(address + 3) != i) {
+ err = -ENODEV;
+ goto ERROR1;
+ }
+ if (inb_p(address + 7) != i) {
+ err = -ENODEV;
+ goto ERROR1;
+ }
+#undef REALLY_SLOW_IO
+
+ /* Let's just hope nothing breaks here */
+ i = inb_p(address + 5) & 0x7f;
+ outb_p(~i & 0x7f, address + 5);
+ if ((inb_p(address + 5) & 0x7f) != (~i & 0x7f)) {
+ outb_p(i, address + 5);
+ err = -ENODEV;
+ goto ERROR1;
+ }
+ }
+ }
+
+ /* OK. For now, we presume we have a valid client. We now create the
+ client structure, even though we cannot fill it completely yet.
+ But it allows us to access lm78_{read,write}_value. */
+
+ if (!(new_client = kmalloc((sizeof(struct i2c_client)) +
+ sizeof(struct lm78_data),
+ GFP_KERNEL))) {
+ err = -ENOMEM;
+ goto ERROR1;
+ }
+ memset(new_client, 0, sizeof(struct i2c_client) +
+ sizeof(struct lm78_data));
+
+ data = (struct lm78_data *) (new_client + 1);
+ if (is_isa)
+ init_MUTEX(&data->lock);
+ i2c_set_clientdata(new_client, data);
+ new_client->addr = address;
+ new_client->adapter = adapter;
+ new_client->driver = &lm78_driver;
+ new_client->flags = 0;
+
+ /* Now, we do the remaining detection. */
+ if (kind < 0) {
+ if (lm78_read_value(new_client, LM78_REG_CONFIG) & 0x80) {
+ err = -ENODEV;
+ goto ERROR2;
+ }
+ if (!is_isa && (lm78_read_value(
+ new_client, LM78_REG_I2C_ADDR) != address)) {
+ err = -ENODEV;
+ goto ERROR2;
+ }
+ }
+
+ /* Determine the chip type. */
+ if (kind <= 0) {
+ i = lm78_read_value(new_client, LM78_REG_CHIPID);
+ if (i == 0x00 || i == 0x20)
+ kind = lm78;
+ else if (i == 0x40)
+ kind = lm78j;
+ else if ((i & 0xfe) == 0xc0)
+ kind = lm79;
+ else {
+ if (kind == 0)
+ printk(KERN_WARNING "lm78.o: Ignoring 'force' "
+ "parameter for unknown chip at "
+ "adapter %d, address 0x%02x\n",
+ i2c_adapter_id(adapter), address);
+ err = -ENODEV;
+ goto ERROR2;
+ }
+ }
+
+ if (kind == lm78) {
+ client_name = "LM78 chip";
+ } else if (kind == lm78j) {
+ client_name = "LM78-J chip";
+ } else if (kind == lm79) {
+ client_name = "LM79 chip";
+ } else {
+ dev_dbg(&adapter->dev, "Internal error: unknown kind (%d)?!?",
+ kind);
+ err = -ENODEV;
+ goto ERROR2;
+ }
+
+ /* Fill in the remaining client fields and put into the global list */
+ strlcpy(new_client->dev.name, client_name, DEVICE_NAME_SIZE);
+ data->type = kind;
+
+ data->valid = 0;
+ init_MUTEX(&data->update_lock);
+
+ /* Tell the I2C layer a new client has arrived */
+ if ((err = i2c_attach_client(new_client)))
+ goto ERROR2;
+
+ /* register sysfs hooks */
+ device_create_file(&new_client->dev, &dev_attr_in_input0);
+ device_create_file(&new_client->dev, &dev_attr_in_min0);
+ device_create_file(&new_client->dev, &dev_attr_in_max0);
+ device_create_file(&new_client->dev, &dev_attr_in_input1);
+ device_create_file(&new_client->dev, &dev_attr_in_min1);
+ device_create_file(&new_client->dev, &dev_attr_in_max1);
+ device_create_file(&new_client->dev, &dev_attr_in_input2);
+ device_create_file(&new_client->dev, &dev_attr_in_min2);
+ device_create_file(&new_client->dev, &dev_attr_in_max2);
+ device_create_file(&new_client->dev, &dev_attr_in_input3);
+ device_create_file(&new_client->dev, &dev_attr_in_min3);
+ device_create_file(&new_client->dev, &dev_attr_in_max3);
+ device_create_file(&new_client->dev, &dev_attr_in_input4);
+ device_create_file(&new_client->dev, &dev_attr_in_min4);
+ device_create_file(&new_client->dev, &dev_attr_in_max4);
+ device_create_file(&new_client->dev, &dev_attr_in_input5);
+ device_create_file(&new_client->dev, &dev_attr_in_min5);
+ device_create_file(&new_client->dev, &dev_attr_in_max5);
+ device_create_file(&new_client->dev, &dev_attr_in_input6);
+ device_create_file(&new_client->dev, &dev_attr_in_min6);
+ device_create_file(&new_client->dev, &dev_attr_in_max6);
+ device_create_file(&new_client->dev, &dev_attr_temp_input);
+ device_create_file(&new_client->dev, &dev_attr_temp_min);
+ device_create_file(&new_client->dev, &dev_attr_temp_max);
+ device_create_file(&new_client->dev, &dev_attr_fan_input1);
+ device_create_file(&new_client->dev, &dev_attr_fan_min1);
+ device_create_file(&new_client->dev, &dev_attr_fan_div1);
+ device_create_file(&new_client->dev, &dev_attr_fan_input2);
+ device_create_file(&new_client->dev, &dev_attr_fan_min2);
+ device_create_file(&new_client->dev, &dev_attr_fan_div2);
+ device_create_file(&new_client->dev, &dev_attr_fan_input3);
+ device_create_file(&new_client->dev, &dev_attr_fan_min3);
+ device_create_file(&new_client->dev, &dev_attr_fan_div3);
+ device_create_file(&new_client->dev, &dev_attr_alarms);
+ device_create_file(&new_client->dev, &dev_attr_vid);
+
+ /* Initialize the LM78 chip */
+ lm78_init_client(new_client);
+ return 0;
+
+ERROR2:
+ kfree(new_client);
+ERROR1:
+ if (is_isa)
+ release_region(address, LM78_EXTENT);
+ERROR0:
+ return err;
+}
+
+static int lm78_detach_client(struct i2c_client *client)
+{
+ int err;
+
+ /* release ISA region first */
+ if(i2c_is_isa_client(client))
+ release_region(client->addr, LM78_EXTENT);
+
+ /* now it's safe to scrap the rest */
+ if ((err = i2c_detach_client(client))) {
+ dev_err(&client->dev,
+ "Client deregistration failed, client not detached.\n");
+ return err;
+ }
+
+ kfree(client);
+
+ return 0;
+}
+
+/* The SMBus locks itself, but ISA access must be locked explicitely!
+ We don't want to lock the whole ISA bus, so we lock each client
+ separately.
+ We ignore the LM78 BUSY flag at this moment - it could lead to deadlocks,
+ would slow down the LM78 access and should not be necessary.
+ There are some ugly typecasts here, but the good new is - they should
+ nowhere else be necessary! */
+static int lm78_read_value(struct i2c_client *client, u8 reg)
+{
+ int res;
+ if (i2c_is_isa_client(client)) {
+ struct lm78_data *data = i2c_get_clientdata(client);
+ down(&data->lock);
+ outb_p(reg, client->addr + LM78_ADDR_REG_OFFSET);
+ res = inb_p(client->addr + LM78_DATA_REG_OFFSET);
+ up(&data->lock);
+ return res;
+ } else
+ return i2c_smbus_read_byte_data(client, reg);
+}
+
+/* The SMBus locks itself, but ISA access muse be locked explicitely!
+ We don't want to lock the whole ISA bus, so we lock each client
+ separately.
+ We ignore the LM78 BUSY flag at this moment - it could lead to deadlocks,
+ would slow down the LM78 access and should not be necessary.
+ There are some ugly typecasts here, but the good new is - they should
+ nowhere else be necessary! */
+static int lm78_write_value(struct i2c_client *client, u8 reg, u8 value)
+{
+ if (i2c_is_isa_client(client)) {
+ struct lm78_data *data = i2c_get_clientdata(client);
+ down(&data->lock);
+ outb_p(reg, client->addr + LM78_ADDR_REG_OFFSET);
+ outb_p(value, client->addr + LM78_DATA_REG_OFFSET);
+ up(&data->lock);
+ return 0;
+ } else
+ return i2c_smbus_write_byte_data(client, reg, value);
+}
+
+/* Called when we have found a new LM78. It should set limits, etc. */
+static void lm78_init_client(struct i2c_client *client)
+{
+ struct lm78_data *data = i2c_get_clientdata(client);
+ int vid;
+
+ /* Reset all except Watchdog values and last conversion values
+ This sets fan-divs to 2, among others */
+ lm78_write_value(client, LM78_REG_CONFIG, 0x80);
+
+ vid = lm78_read_value(client, LM78_REG_VID_FANDIV) & 0x0f;
+ if (data->type == lm79)
+ vid |=
+ (lm78_read_value(client, LM78_REG_CHIPID) & 0x01) << 4;
+ else
+ vid |= 0x10;
+ vid = VID_FROM_REG(vid);
+
+ lm78_write_value(client, LM78_REG_IN_MIN(0),
+ IN_TO_REG(LM78_INIT_IN_MIN_0(vid)));
+ lm78_write_value(client, LM78_REG_IN_MAX(0),
+ IN_TO_REG(LM78_INIT_IN_MAX_0(vid)));
+ lm78_write_value(client, LM78_REG_IN_MIN(1),
+ IN_TO_REG(LM78_INIT_IN_MIN_1(vid)));
+ lm78_write_value(client, LM78_REG_IN_MAX(1),
+ IN_TO_REG(LM78_INIT_IN_MAX_1(vid)));
+ lm78_write_value(client, LM78_REG_IN_MIN(2),
+ IN_TO_REG(LM78_INIT_IN_MIN_2));
+ lm78_write_value(client, LM78_REG_IN_MAX(2),
+ IN_TO_REG(LM78_INIT_IN_MAX_2));
+ lm78_write_value(client, LM78_REG_IN_MIN(3),
+ IN_TO_REG(LM78_INIT_IN_MIN_3));
+ lm78_write_value(client, LM78_REG_IN_MAX(3),
+ IN_TO_REG(LM78_INIT_IN_MAX_3));
+ lm78_write_value(client, LM78_REG_IN_MIN(4),
+ IN_TO_REG(LM78_INIT_IN_MIN_4));
+ lm78_write_value(client, LM78_REG_IN_MAX(4),
+ IN_TO_REG(LM78_INIT_IN_MAX_4));
+ lm78_write_value(client, LM78_REG_IN_MIN(5),
+ IN_TO_REG(LM78_INIT_IN_MIN_5));
+ lm78_write_value(client, LM78_REG_IN_MAX(5),
+ IN_TO_REG(LM78_INIT_IN_MAX_5));
+ lm78_write_value(client, LM78_REG_IN_MIN(6),
+ IN_TO_REG(LM78_INIT_IN_MIN_6));
+ lm78_write_value(client, LM78_REG_IN_MAX(6),
+ IN_TO_REG(LM78_INIT_IN_MAX_6));
+ lm78_write_value(client, LM78_REG_FAN_MIN(0),
+ FAN_TO_REG(LM78_INIT_FAN_MIN_1, 2));
+ lm78_write_value(client, LM78_REG_FAN_MIN(1),
+ FAN_TO_REG(LM78_INIT_FAN_MIN_2, 2));
+ lm78_write_value(client, LM78_REG_FAN_MIN(2),
+ FAN_TO_REG(LM78_INIT_FAN_MIN_3, 2));
+ lm78_write_value(client, LM78_REG_TEMP_OVER,
+ TEMP_TO_REG(LM78_INIT_TEMP_OVER));
+ lm78_write_value(client, LM78_REG_TEMP_HYST,
+ TEMP_TO_REG(LM78_INIT_TEMP_HYST));
+
+ /* Start monitoring */
+ lm78_write_value(client, LM78_REG_CONFIG,
+ (lm78_read_value(client, LM78_REG_CONFIG) & 0xf7)
+ | 0x01);
+
+}
+
+static void lm78_update_client(struct i2c_client *client)
+{
+ struct lm78_data *data = i2c_get_clientdata(client);
+ int i;
+
+ down(&data->update_lock);
+
+ if ((jiffies - data->last_updated > HZ + HZ / 2) ||
+ (jiffies < data->last_updated) || !data->valid) {
+
+ dev_dbg(&client->dev, "Starting lm78 update\n");
+
+ for (i = 0; i <= 6; i++) {
+ data->in[i] =
+ lm78_read_value(client, LM78_REG_IN(i));
+ data->in_min[i] =
+ lm78_read_value(client, LM78_REG_IN_MIN(i));
+ data->in_max[i] =
+ lm78_read_value(client, LM78_REG_IN_MAX(i));
+ }
+ for (i = 0; i < 3; i++) {
+ data->fan[i] =
+ lm78_read_value(client, LM78_REG_FAN(i));
+ data->fan_min[i] =
+ lm78_read_value(client, LM78_REG_FAN_MIN(i));
+ }
+ data->temp = lm78_read_value(client, LM78_REG_TEMP);
+ data->temp_over =
+ lm78_read_value(client, LM78_REG_TEMP_OVER);
+ data->temp_hyst =
+ lm78_read_value(client, LM78_REG_TEMP_HYST);
+ i = lm78_read_value(client, LM78_REG_VID_FANDIV);
+ data->vid = i & 0x0f;
+ if (data->type == lm79)
+ data->vid |=
+ (lm78_read_value(client, LM78_REG_CHIPID) &
+ 0x01) << 4;
+ else
+ data->vid |= 0x10;
+ data->fan_div[0] = (i >> 4) & 0x03;
+ data->fan_div[1] = i >> 6;
+ data->alarms = lm78_read_value(client, LM78_REG_ALARM1) +
+ (lm78_read_value(client, LM78_REG_ALARM2) << 8);
+ data->last_updated = jiffies;
+ data->valid = 1;
+
+ data->fan_div[2] = 1;
+ }
+
+ up(&data->update_lock);
+}
+
+static int __init sm_lm78_init(void)
+{
+ return i2c_add_driver(&lm78_driver);
+}
+
+static void __exit sm_lm78_exit(void)
+{
+ i2c_del_driver(&lm78_driver);
+}
+
+
+
+MODULE_AUTHOR("Frodo Looijaard <frodol@dds.nl>");
+MODULE_DESCRIPTION("LM78, LM78-J and LM79 driver");
+MODULE_LICENSE("GPL");
+
+module_init(sm_lm78_init);
+module_exit(sm_lm78_exit);
^ permalink raw reply [flat|nested] 32+ messages in thread
* Re: [PATCH] i2c driver changes for 2.5.72
2005-05-19 6:24 ` Greg KH
@ 2005-05-19 6:24 ` Greg KH
-1 siblings, 0 replies; 32+ messages in thread
From: Greg KH @ 2003-06-18 18:25 UTC (permalink / raw)
To: linux-kernel, sensors
ChangeSet 1.1318.3.2, 2003/06/16 11:31:43-07:00, margitsw@t-online.de
[PATCH] I2C: lm85 fixups
OK Here's the patch which :
1) Fixes the race conditions
2) Correctly reports the temps :-)
3) Removes a bit of gunk in the defines which I forgot
drivers/i2c/chips/lm85.c | 41 ++++++++++++++++++++++++++++-------------
1 files changed, 28 insertions(+), 13 deletions(-)
diff -Nru a/drivers/i2c/chips/lm85.c b/drivers/i2c/chips/lm85.c
--- a/drivers/i2c/chips/lm85.c Wed Jun 18 11:19:44 2003
+++ b/drivers/i2c/chips/lm85.c Wed Jun 18 11:19:44 2003
@@ -148,20 +148,17 @@
#define SCALE(val,from,to) (((val)*(to) + ((from)/2))/(from))
#define INS_TO_REG(n,val) (SENSORS_LIMIT(SCALE(val,lm85_scaling[n],192),0,255))
#define INSEXT_FROM_REG(n,val,ext) (SCALE((val)*4 + (ext),192*4,lm85_scaling[n]))
-/*
#define INS_FROM_REG(n,val) (INSEXT_FROM_REG(n,val,0))
-*/
-#define INS_FROM_REG(n,val) ( ( (val*4*lm85_scaling[n]) + (192*4/2) ) / (192*4) )
/* FAN speed is measured using 90kHz clock */
#define FAN_TO_REG(val) (SENSORS_LIMIT( (val)<=0?0: 5400000/(val),0,65534))
#define FAN_FROM_REG(val) ((val)==0?-1:(val)==0xffff?0:5400000/(val))
-/* Temperature is reported in .01 degC increments */
-#define TEMP_TO_REG(val) (SENSORS_LIMIT(((val)+50)/100,-127,127))
-#define TEMPEXT_FROM_REG(val,ext) ((val)*100 + (ext)*25)
+/* Temperature is reported in .001 degC increments */
+#define TEMP_TO_REG(val) (SENSORS_LIMIT(((val)+500)/1000,-127,127))
+#define TEMPEXT_FROM_REG(val,ext) ((val)*1000 + (ext)*250)
#define TEMP_FROM_REG(val) (TEMPEXT_FROM_REG(val,0))
-#define EXTTEMP_TO_REG(val) (SENSORS_LIMIT((val)/25,-127,127))
+#define EXTTEMP_TO_REG(val) (SENSORS_LIMIT((val)/250,-127,127))
#define PWM_TO_REG(val) (SENSORS_LIMIT(val,0,255))
#define PWM_FROM_REG(val) (val)
@@ -437,10 +434,13 @@
{
struct i2c_client *client = to_i2c_client(dev);
struct lm85_data *data = i2c_get_clientdata(client);
+ int val;
- int val = simple_strtol(buf, NULL, 10);
+ down(&data->update_lock);
+ val = simple_strtol(buf, NULL, 10);
data->fan_min[nr] = FAN_TO_REG(val);
lm85_write_value(client, LM85_REG_FAN_MIN(nr), data->fan_min[nr]);
+ up(&data->update_lock);
return count;
}
@@ -528,10 +528,13 @@
{
struct i2c_client *client = to_i2c_client(dev);
struct lm85_data *data = i2c_get_clientdata(client);
+ int val;
- int val = simple_strtol(buf, NULL, 10);
+ down(&data->update_lock);
+ val = simple_strtol(buf, NULL, 10);
data->pwm[nr] = PWM_TO_REG(val);
lm85_write_value(client, LM85_REG_PWM(nr), data->pwm[nr]);
+ up(&data->update_lock);
return count;
}
static ssize_t show_pwm_enable(struct device *dev, char *buf, int nr)
@@ -590,10 +593,13 @@
{
struct i2c_client *client = to_i2c_client(dev);
struct lm85_data *data = i2c_get_clientdata(client);
+ int val;
- int val = simple_strtol(buf, NULL, 10);
+ down(&data->update_lock);
+ val = simple_strtol(buf, NULL, 10);
data->in_min[nr] = INS_TO_REG(nr, val);
lm85_write_value(client, LM85_REG_IN_MIN(nr), data->in_min[nr]);
+ up(&data->update_lock);
return count;
}
static ssize_t show_in_max(struct device *dev, char *buf, int nr)
@@ -609,10 +615,13 @@
{
struct i2c_client *client = to_i2c_client(dev);
struct lm85_data *data = i2c_get_clientdata(client);
+ int val;
- int val = simple_strtol(buf, NULL, 10);
+ down(&data->update_lock);
+ val = simple_strtol(buf, NULL, 10);
data->in_max[nr] = INS_TO_REG(nr, val);
lm85_write_value(client, LM85_REG_IN_MAX(nr), data->in_max[nr]);
+ up(&data->update_lock);
return count;
}
#define show_in_reg(offset) \
@@ -673,10 +682,13 @@
{
struct i2c_client *client = to_i2c_client(dev);
struct lm85_data *data = i2c_get_clientdata(client);
+ int val;
- int val = simple_strtol(buf, NULL, 10);
+ down(&data->update_lock);
+ val = simple_strtol(buf, NULL, 10);
data->temp_min[nr] = TEMP_TO_REG(val);
lm85_write_value(client, LM85_REG_TEMP_MIN(nr), data->temp_min[nr]);
+ up(&data->update_lock);
return count;
}
static ssize_t show_temp_max(struct device *dev, char *buf, int nr)
@@ -692,10 +704,13 @@
{
struct i2c_client *client = to_i2c_client(dev);
struct lm85_data *data = i2c_get_clientdata(client);
+ int val;
- int val = simple_strtol(buf, NULL, 10);
+ down(&data->update_lock);
+ val = simple_strtol(buf, NULL, 10);
data->temp_max[nr] = TEMP_TO_REG(val);
lm85_write_value(client, LM85_REG_TEMP_MAX(nr), data->temp_max[nr]);
+ up(&data->update_lock);
return count;
}
#define show_temp_reg(offset) \
^ permalink raw reply [flat|nested] 32+ messages in thread
* Re: [PATCH] i2c driver changes for 2.5.72
2005-05-19 6:24 ` Greg KH
@ 2005-05-19 6:24 ` Greg KH
-1 siblings, 0 replies; 32+ messages in thread
From: Greg KH @ 2003-06-18 18:25 UTC (permalink / raw)
To: linux-kernel, sensors
ChangeSet 1.1318.3.3, 2003/06/16 11:31:55-07:00, mhoffman@lightlink.com
[PATCH] I2C: w83781d bugfix
My first patch was naive; the patch below solves the problem by
letting w83781d_detach_client remove the three clients (1 * primary
+ 2 * subclients) independently. It's a noisy patch because I had
to change the way the subclients were kmalloc'ed - sorry. The meat
is around line 1422. This patch works for me... comments?
drivers/i2c/chips/w83781d.c | 76 +++++++++++++++++++++++---------------------
1 files changed, 40 insertions(+), 36 deletions(-)
diff -Nru a/drivers/i2c/chips/w83781d.c b/drivers/i2c/chips/w83781d.c
--- a/drivers/i2c/chips/w83781d.c Wed Jun 18 11:19:39 2003
+++ b/drivers/i2c/chips/w83781d.c Wed Jun 18 11:19:39 2003
@@ -299,8 +299,8 @@
char valid; /* !=0 if following fields are valid */
unsigned long last_updated; /* In jiffies */
- struct i2c_client *lm75; /* for secondary I2C addresses */
- /* pointer to array of 2 subclients */
+ struct i2c_client *lm75[2]; /* for secondary I2C addresses */
+ /* array of 2 pointers to subclients */
u8 in[9]; /* Register value - 8 & 9 for 782D only */
u8 in_max[9]; /* Register value - 8 & 9 for 782D only */
@@ -1043,12 +1043,12 @@
const char *client_name;
struct w83781d_data *data = i2c_get_clientdata(new_client);
- if (!(data->lm75 = kmalloc(2 * sizeof (struct i2c_client),
- GFP_KERNEL))) {
+ data->lm75[0] = kmalloc(sizeof(struct i2c_client), GFP_KERNEL);
+ if (!(data->lm75[0])) {
err = -ENOMEM;
goto ERROR_SC_0;
}
- memset(data->lm75, 0x00, 2 * sizeof (struct i2c_client));
+ memset(data->lm75[0], 0x00, sizeof (struct i2c_client));
id = i2c_adapter_id(adapter);
@@ -1066,25 +1066,33 @@
w83781d_write_value(new_client, W83781D_REG_I2C_SUBADDR,
(force_subclients[2] & 0x07) |
((force_subclients[3] & 0x07) << 4));
- data->lm75[0].addr = force_subclients[2];
+ data->lm75[0]->addr = force_subclients[2];
} else {
val1 = w83781d_read_value(new_client, W83781D_REG_I2C_SUBADDR);
- data->lm75[0].addr = 0x48 + (val1 & 0x07);
+ data->lm75[0]->addr = 0x48 + (val1 & 0x07);
}
if (kind != w83783s) {
+
+ data->lm75[1] = kmalloc(sizeof(struct i2c_client), GFP_KERNEL);
+ if (!(data->lm75[1])) {
+ err = -ENOMEM;
+ goto ERROR_SC_1;
+ }
+ memset(data->lm75[1], 0x0, sizeof(struct i2c_client));
+
if (force_subclients[0] == id &&
force_subclients[1] == address) {
- data->lm75[1].addr = force_subclients[3];
+ data->lm75[1]->addr = force_subclients[3];
} else {
- data->lm75[1].addr = 0x48 + ((val1 >> 4) & 0x07);
+ data->lm75[1]->addr = 0x48 + ((val1 >> 4) & 0x07);
}
- if (data->lm75[0].addr == data->lm75[1].addr) {
+ if (data->lm75[0]->addr == data->lm75[1]->addr) {
dev_err(&new_client->dev,
"Duplicate addresses 0x%x for subclients.\n",
- data->lm75[0].addr);
+ data->lm75[0]->addr);
err = -EBUSY;
- goto ERROR_SC_1;
+ goto ERROR_SC_2;
}
}
@@ -1103,19 +1111,19 @@
for (i = 0; i <= 1; i++) {
/* store all data in w83781d */
- i2c_set_clientdata(&data->lm75[i], NULL);
- data->lm75[i].adapter = adapter;
- data->lm75[i].driver = &w83781d_driver;
- data->lm75[i].flags = 0;
- strlcpy(data->lm75[i].dev.name, client_name,
+ i2c_set_clientdata(data->lm75[i], NULL);
+ data->lm75[i]->adapter = adapter;
+ data->lm75[i]->driver = &w83781d_driver;
+ data->lm75[i]->flags = 0;
+ strlcpy(data->lm75[i]->dev.name, client_name,
DEVICE_NAME_SIZE);
- if ((err = i2c_attach_client(&(data->lm75[i])))) {
+ if ((err = i2c_attach_client(data->lm75[i]))) {
dev_err(&new_client->dev, "Subclient %d "
"registration at address 0x%x "
- "failed.\n", i, data->lm75[i].addr);
+ "failed.\n", i, data->lm75[i]->addr);
if (i == 1)
- goto ERROR_SC_2;
- goto ERROR_SC_1;
+ goto ERROR_SC_3;
+ goto ERROR_SC_2;
}
if (kind == w83783s)
break;
@@ -1124,10 +1132,14 @@
return 0;
/* Undo inits in case of errors */
+ERROR_SC_3:
+ i2c_detach_client(data->lm75[0]);
ERROR_SC_2:
- i2c_detach_client(&(data->lm75[0]));
+ if (NULL != data->lm75[1])
+ kfree(data->lm75[1]);
ERROR_SC_1:
- kfree(data->lm75);
+ if (NULL != data->lm75[0])
+ kfree(data->lm75[0]);
ERROR_SC_0:
return err;
}
@@ -1326,7 +1338,8 @@
kind, new_client)))
goto ERROR3;
} else {
- data->lm75 = NULL;
+ data->lm75[0] = NULL;
+ data->lm75[1] = NULL;
}
device_create_file_in(new_client, 0);
@@ -1409,20 +1422,11 @@
static int
w83781d_detach_client(struct i2c_client *client)
{
- struct w83781d_data *data = i2c_get_clientdata(client);
int err;
- /* release ISA region or I2C subclients first */
- if (i2c_is_isa_client(client)) {
+ if (i2c_is_isa_client(client))
release_region(client->addr, W83781D_EXTENT);
- } else {
- i2c_detach_client(&data->lm75[0]);
- if (data->type != w83783s)
- i2c_detach_client(&data->lm75[1]);
- kfree(data->lm75);
- }
- /* now it's safe to scrap the rest */
if ((err = i2c_detach_client(client))) {
dev_err(&client->dev,
"Client deregistration failed, client not detached.\n");
@@ -1484,7 +1488,7 @@
res = i2c_smbus_read_byte_data(client, reg & 0xff);
} else {
/* switch to subclient */
- cl = &data->lm75[bank - 1];
+ cl = data->lm75[bank - 1];
/* convert from ISA to LM75 I2C addresses */
switch (reg & 0xff) {
case 0x50: /* TEMP */
@@ -1555,7 +1559,7 @@
value & 0xff);
} else {
/* switch to subclient */
- cl = &data->lm75[bank - 1];
+ cl = data->lm75[bank - 1];
/* convert from ISA to LM75 I2C addresses */
switch (reg & 0xff) {
case 0x52: /* CONFIG */
^ permalink raw reply [flat|nested] 32+ messages in thread
* Re: [PATCH] i2c driver changes for 2.5.72
2005-05-19 6:24 ` Greg KH
@ 2005-05-19 6:24 ` Greg KH
-1 siblings, 0 replies; 32+ messages in thread
From: Greg KH @ 2003-06-18 18:25 UTC (permalink / raw)
To: linux-kernel, sensors
ChangeSet 1.1318.3.4, 2003/06/16 11:32:06-07:00, azarah@gentoo.org
[PATCH] I2C: ICH5 SMBus and W83627THF additions
I have been trying to get the W83627THF chip working on this
board. It is an Asus P4C800 with Intel 875p chipset and a
W83627THF connected via the SMBus.
There are no data sheet for the W83627THF as far as I can see,
but supposidly it is a W83627HF with advance Fan control, etc.
I have applied attached patches, and tried various other things
to get it to work, but no avail. The SMBus on the ICH5 seems to
work, and it seems to detect the sensor chip just fine, but it
do not seem to read any values from the W83627THF.
drivers/i2c/busses/i2c-i801.c | 10 +++++++++-
drivers/i2c/chips/w83781d.c | 2 +-
2 files changed, 10 insertions(+), 2 deletions(-)
diff -Nru a/drivers/i2c/busses/i2c-i801.c b/drivers/i2c/busses/i2c-i801.c
--- a/drivers/i2c/busses/i2c-i801.c Wed Jun 18 11:19:35 2003
+++ b/drivers/i2c/busses/i2c-i801.c Wed Jun 18 11:19:35 2003
@@ -27,6 +27,7 @@
82801BA 2443
82801CA/CAM 2483
82801DB 24C3 (HW PEC supported, 32 byte buffer not supported)
+ 82801EB 24D3 (HW PEC supported, 32 byte buffer not supported)
This driver supports several versions of Intel's I/O Controller Hubs (ICH).
For SMBus support, they are similar to the PIIX4 and are part
@@ -121,7 +122,8 @@
return -ENODEV;
I801_dev = dev;
- if (dev->device == PCI_DEVICE_ID_INTEL_82801DB_3)
+ if ((dev->device == PCI_DEVICE_ID_INTEL_82801DB_3) ||
+ (dev->device == PCI_DEVICE_ID_INTEL_82801EB_3))
isich4 = 1;
else
isich4 = 0;
@@ -584,6 +586,12 @@
.device = PCI_DEVICE_ID_INTEL_82801DB_3,
.subvendor = PCI_ANY_ID,
.subdevice = PCI_ANY_ID,
+ },
+ {
+ .vendor = PCI_VENDOR_ID_INTEL,
+ .device = PCI_DEVICE_ID_INTEL_82801EB_3,
+ .subvendor = PCI_ANY_ID,
+ .subdevice = PCI_ANY_ID,
},
{ 0, }
};
diff -Nru a/drivers/i2c/chips/w83781d.c b/drivers/i2c/chips/w83781d.c
--- a/drivers/i2c/chips/w83781d.c Wed Jun 18 11:19:35 2003
+++ b/drivers/i2c/chips/w83781d.c Wed Jun 18 11:19:35 2003
@@ -1285,7 +1285,7 @@
kind = w83782d;
else if (val1 == 0x40 && vendid == winbond && !is_isa)
kind = w83783s;
- else if (val1 == 0x20 && vendid == winbond)
+ else if ((val1 == 0x20 || val1 == 0x72) && vendid == winbond)
kind = w83627hf;
else if (val1 == 0x30 && vendid == asus && !is_isa)
kind = as99127f;
^ permalink raw reply [flat|nested] 32+ messages in thread
* Re: [PATCH] i2c driver changes for 2.5.72
2005-05-19 6:24 ` Greg KH
@ 2005-05-19 6:24 ` Greg KH
-1 siblings, 0 replies; 32+ messages in thread
From: Greg KH @ 2003-06-18 18:25 UTC (permalink / raw)
To: linux-kernel, sensors
ChangeSet 1.1318.3.5, 2003/06/16 11:32:48-07:00, greg@kroah.com
[PATCH] I2C: add lm78 chip to Makefile
drivers/i2c/chips/Makefile | 1 +
1 files changed, 1 insertion(+)
diff -Nru a/drivers/i2c/chips/Makefile b/drivers/i2c/chips/Makefile
--- a/drivers/i2c/chips/Makefile Wed Jun 18 11:19:30 2003
+++ b/drivers/i2c/chips/Makefile Wed Jun 18 11:19:30 2003
@@ -5,6 +5,7 @@
obj-$(CONFIG_SENSORS_ADM1021) += adm1021.o
obj-$(CONFIG_SENSORS_IT87) += it87.o
obj-$(CONFIG_SENSORS_LM75) += lm75.o
+obj-$(CONFIG_SENSORS_LM78) += lm78.o
obj-$(CONFIG_SENSORS_LM85) += lm85.o
obj-$(CONFIG_SENSORS_VIA686A) += via686a.o
obj-$(CONFIG_SENSORS_W83781D) += w83781d.o
^ permalink raw reply [flat|nested] 32+ messages in thread
* Re: [PATCH] i2c driver changes for 2.5.72
2005-05-19 6:24 ` Greg KH
@ 2005-05-19 6:24 ` Greg KH
-1 siblings, 0 replies; 32+ messages in thread
From: Greg KH @ 2003-06-18 18:25 UTC (permalink / raw)
To: linux-kernel, sensors
ChangeSet 1.1318.3.7, 2003/06/17 15:34:14-07:00, azarah@gentoo.org
[PATCH] I2C: fix for previous W83627THF sensor chip patch
Ok, I was wrong in assuming that the W83627THF was on the I2C bus.
It is on the ISA bus, id 0x90 (thanks to Alex Van Kaam author of
MBM who corrected my assumption).
drivers/i2c/chips/w83781d.c | 8 ++++++--
1 files changed, 6 insertions(+), 2 deletions(-)
diff -Nru a/drivers/i2c/chips/w83781d.c b/drivers/i2c/chips/w83781d.c
--- a/drivers/i2c/chips/w83781d.c Wed Jun 18 11:19:21 2003
+++ b/drivers/i2c/chips/w83781d.c Wed Jun 18 11:19:21 2003
@@ -28,6 +28,7 @@
asb100 "bach" (type_name = as99127f) 0x30 0x0694 yes no
w83781d 7 3 0 3 0x10 0x5ca3 yes yes
w83627hf 9 3 2 3 0x20 0x5ca3 yes yes(LPC)
+ w83627thf 9 3 2 3 0x90 0x5ca3 no yes(LPC)
w83782d 9 3 2-4 3 0x30 0x5ca3 yes yes
w83783s 5-6 3 2 1-2 0x40 0x5ca3 yes no
w83697hf 8 2 2 2 0x60 0x5ca3 no yes(LPC)
@@ -1285,7 +1286,7 @@
kind = w83782d;
else if (val1 == 0x40 && vendid == winbond && !is_isa)
kind = w83783s;
- else if ((val1 == 0x20 || val1 == 0x72) && vendid == winbond)
+ else if ((val1 == 0x20 || val1 == 0x90) && vendid == winbond)
kind = w83627hf;
else if (val1 == 0x30 && vendid == asus && !is_isa)
kind = as99127f;
@@ -1309,7 +1310,10 @@
} else if (kind == w83783s) {
client_name = "W83783S chip";
} else if (kind == w83627hf) {
- client_name = "W83627HF chip";
+ if (val1 == 0x90)
+ client_name = "W83627THF chip";
+ else
+ client_name = "W83627HF chip";
} else if (kind == as99127f) {
client_name = "AS99127F chip";
} else if (kind == w83697hf) {
^ permalink raw reply [flat|nested] 32+ messages in thread
* Re: [PATCH] i2c driver changes for 2.5.72
2005-05-19 6:24 ` Greg KH
@ 2005-05-19 6:24 ` Greg KH
-1 siblings, 0 replies; 32+ messages in thread
From: Greg KH @ 2003-06-18 18:25 UTC (permalink / raw)
To: linux-kernel, sensors
ChangeSet 1.1318.3.8, 2003/06/17 15:37:40-07:00, greg@kroah.com
I2C: fix resource leak in i2c-ali15x3.c
drivers/i2c/busses/i2c-ali15x3.c | 18 +++++++++++-------
1 files changed, 11 insertions(+), 7 deletions(-)
diff -Nru a/drivers/i2c/busses/i2c-ali15x3.c b/drivers/i2c/busses/i2c-ali15x3.c
--- a/drivers/i2c/busses/i2c-ali15x3.c Wed Jun 18 11:19:16 2003
+++ b/drivers/i2c/busses/i2c-ali15x3.c Wed Jun 18 11:19:16 2003
@@ -177,17 +177,18 @@
if(force_addr) {
dev_info(&ALI15X3_dev->dev, "forcing ISA address 0x%04X\n",
ali15x3_smba);
- if (PCIBIOS_SUCCESSFUL !=
- pci_write_config_word(ALI15X3_dev, SMBBA, ali15x3_smba))
- return -ENODEV;
- if (PCIBIOS_SUCCESSFUL !=
- pci_read_config_word(ALI15X3_dev, SMBBA, &a))
- return -ENODEV;
+ if (PCIBIOS_SUCCESSFUL != pci_write_config_word(ALI15X3_dev,
+ SMBBA,
+ ali15x3_smba))
+ goto error;
+ if (PCIBIOS_SUCCESSFUL != pci_read_config_word(ALI15X3_dev,
+ SMBBA, &a))
+ goto error;
if ((a & ~(ALI15X3_SMB_IOSIZE - 1)) != ali15x3_smba) {
/* make sure it works */
dev_err(&ALI15X3_dev->dev,
"force address failed - not supported?\n");
- return -ENODEV;
+ goto error;
}
}
/* check if whole device is enabled */
@@ -219,6 +220,9 @@
dev_dbg(&ALI15X3_dev->dev, "iALI15X3_smba = 0x%X\n", ali15x3_smba);
return 0;
+error:
+ release_region(ali15x3_smba, ALI15X3_SMB_IOSIZE);
+ return -ENODEV;
}
/* Internally used pause function */
^ permalink raw reply [flat|nested] 32+ messages in thread
* Re: [PATCH] i2c driver changes for 2.5.72
2005-05-19 6:24 ` Greg KH
@ 2005-05-19 6:24 ` Greg KH
-1 siblings, 0 replies; 32+ messages in thread
From: Greg KH @ 2003-06-18 18:25 UTC (permalink / raw)
To: linux-kernel, sensors
ChangeSet 1.1318.3.6, 2003/06/16 11:40:37-07:00, margitsw@t-online.de
[PATCH] I2C: Sensors patch for adm1021
Patch for adm1021
This corrects temp reporting and a major error whereby
"alarms" and "die_code" were being put though the "TEMP" macro.
Compiled but don't have the hardware to test.
drivers/i2c/chips/adm1021.c | 18 ++++++++++++++----
1 files changed, 14 insertions(+), 4 deletions(-)
diff -Nru a/drivers/i2c/chips/adm1021.c b/drivers/i2c/chips/adm1021.c
--- a/drivers/i2c/chips/adm1021.c Wed Jun 18 11:19:25 2003
+++ b/drivers/i2c/chips/adm1021.c Wed Jun 18 11:19:25 2003
@@ -88,8 +88,8 @@
these macros are called: arguments may be evaluated more than once.
Fixing this is just not worth it. */
/* Conversions note: 1021 uses normal integer signed-byte format*/
-#define TEMP_FROM_REG(val) (val > 127 ? val-256 : val)
-#define TEMP_TO_REG(val) (SENSORS_LIMIT((val < 0 ? val+256 : val),0,255))
+#define TEMP_FROM_REG(val) (val > 127 ? (val-256)*1000 : val*1000)
+#define TEMP_TO_REG(val) (SENSORS_LIMIT((val < 0 ? (val/1000)+256 : val/1000),0,255))
/* Initial values */
@@ -172,8 +172,18 @@
show(remote_temp_max);
show(remote_temp_hyst);
show(remote_temp_input);
-show(alarms);
-show(die_code);
+
+#define show2(value) \
+static ssize_t show_##value(struct device *dev, char *buf) \
+{ \
+ struct i2c_client *client = to_i2c_client(dev); \
+ struct adm1021_data *data = i2c_get_clientdata(client); \
+ \
+ adm1021_update_client(client); \
+ return sprintf(buf, "%d\n", data->value); \
+}
+show2(alarms);
+show2(die_code);
#define set(value, reg) \
static ssize_t set_##value(struct device *dev, const char *buf, size_t count) \
^ permalink raw reply [flat|nested] 32+ messages in thread
* [PATCH] i2c driver changes for 2.5.72
@ 2005-05-19 6:24 ` Greg KH
0 siblings, 0 replies; 32+ messages in thread
From: Greg KH @ 2005-05-19 6:24 UTC (permalink / raw)
To: linux-kernel, sensors
ChangeSet 1.1318.3.1, 2003/06/16 11:31:30-07:00, mhoffman@lightlink.com
[PATCH] i2c: Add lm78 sensor chip support
This patch vs. 2.5.70 adds support for LM78, LM78-J, and LM79 sensors
chips based on lm_sensors project CVS. This works on one of my boards.
I want to draw attention to something I did with this driver by
comparing it to it87.c in 2.5.70:
> #define IT87_INIT_TEMP_HIGH_1 600
> #define IT87_INIT_TEMP_LOW_1 200
The hardware uses degrees C, and sysfs uses degrees C * 1000. But
these #defines are apparently in units of degrees C * 10. This
arbitrary intermediate representation bugs me. And given the new 2.5
sysfs standard, it's unnecessary.
In this patch for lm78, I rewrote the conversion routines in terms
of the sysfs units - getting rid of the intermediate nonsense. If
there are no objections, I'm going to start passing patches to do
this to the other sensor chip drivers in 2.5 as well. It would be
nice to get some help with this too... especially since I don't
have all that hardware at hand to test the results.
drivers/i2c/chips/Kconfig | 14
drivers/i2c/chips/lm78.c | 895 ++++++++++++++++++++++++++++++++++++++++++++++
2 files changed, 909 insertions(+)
diff -Nru a/drivers/i2c/chips/Kconfig b/drivers/i2c/chips/Kconfig
--- a/drivers/i2c/chips/Kconfig Wed Jun 18 11:19:49 2003
+++ b/drivers/i2c/chips/Kconfig Wed Jun 18 11:19:49 2003
@@ -62,6 +62,20 @@
in the lm_sensors package, which you can download at
http://www.lm-sensors.nu
+config SENSORS_LM78
+ tristate " National Semiconductors LM78 and compatibles"
+ depends on I2C && EXPERIMENTAL
+ help
+ If you say yes here you get support for National Semiconductor LM78,
+ LM78-J and LM79. This can also be built as a module which can be
+ inserted and removed while the kernel is running.
+
+ The module will be called lm78.
+
+ You will also need the latest user-space utilties: you can find them
+ in the lm_sensors package, which you can download at
+ http://www.lm-sensors.nu
+
config SENSORS_VIA686A
tristate " VIA686A"
depends on I2C && EXPERIMENTAL
diff -Nru a/drivers/i2c/chips/lm78.c b/drivers/i2c/chips/lm78.c
--- /dev/null Wed Dec 31 16:00:00 1969
+++ b/drivers/i2c/chips/lm78.c Wed Jun 18 11:19:49 2003
@@ -0,0 +1,895 @@
+/*
+ lm78.c - Part of lm_sensors, Linux kernel modules for hardware
+ monitoring
+ Copyright (c) 1998, 1999 Frodo Looijaard <frodol@dds.nl>
+
+ 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., 675 Mass Ave, Cambridge, MA 02139, USA.
+*/
+
+#include <linux/module.h>
+#include <linux/init.h>
+#include <linux/slab.h>
+#include <linux/i2c.h>
+#include <linux/i2c-sensor.h>
+#include <asm/io.h>
+
+/* Addresses to scan */
+static unsigned short normal_i2c[] = { I2C_CLIENT_END };
+static unsigned short normal_i2c_range[] = { 0x20, 0x2f, I2C_CLIENT_END };
+static unsigned int normal_isa[] = { 0x0290, I2C_CLIENT_ISA_END };
+static unsigned int normal_isa_range[] = { I2C_CLIENT_ISA_END };
+
+/* Insmod parameters */
+SENSORS_INSMOD_3(lm78, lm78j, lm79);
+
+/* Many LM78 constants specified below */
+
+/* Length of ISA address segment */
+#define LM78_EXTENT 8
+
+/* Where are the ISA address/data registers relative to the base address */
+#define LM78_ADDR_REG_OFFSET 5
+#define LM78_DATA_REG_OFFSET 6
+
+/* The LM78 registers */
+#define LM78_REG_IN_MAX(nr) (0x2b + (nr) * 2)
+#define LM78_REG_IN_MIN(nr) (0x2c + (nr) * 2)
+#define LM78_REG_IN(nr) (0x20 + (nr))
+
+#define LM78_REG_FAN_MIN(nr) (0x3b + (nr))
+#define LM78_REG_FAN(nr) (0x28 + (nr))
+
+#define LM78_REG_TEMP 0x27
+#define LM78_REG_TEMP_OVER 0x39
+#define LM78_REG_TEMP_HYST 0x3a
+
+#define LM78_REG_ALARM1 0x41
+#define LM78_REG_ALARM2 0x42
+
+#define LM78_REG_VID_FANDIV 0x47
+
+#define LM78_REG_CONFIG 0x40
+#define LM78_REG_CHIPID 0x49
+#define LM78_REG_I2C_ADDR 0x48
+
+
+/* Conversions. Rounding and limit checking is only done on the TO_REG
+ variants. */
+
+/* IN: mV, (0V to 4.08V)
+ REG: 16mV/bit */
+static inline u8 IN_TO_REG(unsigned long val)
+{
+ unsigned long nval = SENSORS_LIMIT(val, 0, 4080);
+ return (nval + 8) / 16;
+}
+#define IN_FROM_REG(val) ((val) * 16)
+
+static inline u8 FAN_TO_REG(long rpm, int div)
+{
+ if (rpm = 0)
+ return 255;
+ rpm = SENSORS_LIMIT(rpm, 1, 1000000);
+ return SENSORS_LIMIT((1350000 + rpm * div / 2) / (rpm * div), 1, 254);
+}
+
+static inline int FAN_FROM_REG(u8 val, int div)
+{
+ return val=0 ? -1 : val=255 ? 0 : 1350000/(val*div);
+}
+
+/* TEMP: mC (-128C to +127C)
+ REG: 1C/bit, two's complement */
+static inline u8 TEMP_TO_REG(int val)
+{
+ int nval = SENSORS_LIMIT(val, -128000, 127000) ;
+ return nval<0 ? (nval-500)/1000+0x100 : (nval+500)/1000;
+}
+
+static inline int TEMP_FROM_REG(u8 val)
+{
+ return (val>=0x80 ? val-0x100 : val) * 1000;
+}
+
+/* VID: mV
+ REG: (see doc/vid) */
+static inline int VID_FROM_REG(u8 val)
+{
+ return val=0x1f ? 0 : val>=0x10 ? 5100-val*100 : 2050-val*50;
+}
+
+/* ALARMS: chip-specific bitmask
+ REG: (same) */
+#define ALARMS_FROM_REG(val) (val)
+
+/* FAN DIV: 1, 2, 4, or 8 (defaults to 2)
+ REG: 0, 1, 2, or 3 (respectively) (defaults to 1) */
+static inline u8 DIV_TO_REG(int val)
+{
+ return val=8 ? 3 : val=4 ? 2 : val=1 ? 0 : 1;
+}
+#define DIV_FROM_REG(val) (1 << (val))
+
+/* Initial limits. To keep them sane, we use the 'standard' translation as
+ specified in the LM78 sheet. Use the config file to set better limits. */
+#define LM78_INIT_IN_0(vid) ((vid)=3500 ? 2800 : (vid))
+#define LM78_INIT_IN_1(vid) ((vid)=3500 ? 2800 : (vid))
+#define LM78_INIT_IN_2 3300
+#define LM78_INIT_IN_3 (((5000) * 100)/168)
+#define LM78_INIT_IN_4 (((12000) * 10)/38)
+#define LM78_INIT_IN_5 (((-12000) * -604)/2100)
+#define LM78_INIT_IN_6 (((-5000) * -604)/909)
+
+#define LM78_INIT_IN_PERCENTAGE 10
+
+#define LM78_INIT_IN_MIN_0(vid) (LM78_INIT_IN_0(vid) - \
+ LM78_INIT_IN_0(vid) * LM78_INIT_IN_PERCENTAGE / 100)
+#define LM78_INIT_IN_MAX_0(vid) (LM78_INIT_IN_0(vid) + \
+ LM78_INIT_IN_0(vid) * LM78_INIT_IN_PERCENTAGE / 100)
+#define LM78_INIT_IN_MIN_1(vid) (LM78_INIT_IN_1(vid) - \
+ LM78_INIT_IN_1(vid) * LM78_INIT_IN_PERCENTAGE / 100)
+#define LM78_INIT_IN_MAX_1(vid) (LM78_INIT_IN_1(vid) + \
+ LM78_INIT_IN_1(vid) * LM78_INIT_IN_PERCENTAGE / 100)
+
+#define LM78_INIT_IN_MIN_2 \
+ (LM78_INIT_IN_2 - LM78_INIT_IN_2 * LM78_INIT_IN_PERCENTAGE / 100)
+#define LM78_INIT_IN_MAX_2 \
+ (LM78_INIT_IN_2 + LM78_INIT_IN_2 * LM78_INIT_IN_PERCENTAGE / 100)
+#define LM78_INIT_IN_MIN_3 \
+ (LM78_INIT_IN_3 - LM78_INIT_IN_3 * LM78_INIT_IN_PERCENTAGE / 100)
+#define LM78_INIT_IN_MAX_3 \
+ (LM78_INIT_IN_3 + LM78_INIT_IN_3 * LM78_INIT_IN_PERCENTAGE / 100)
+#define LM78_INIT_IN_MIN_4 \
+ (LM78_INIT_IN_4 - LM78_INIT_IN_4 * LM78_INIT_IN_PERCENTAGE / 100)
+#define LM78_INIT_IN_MAX_4 \
+ (LM78_INIT_IN_4 + LM78_INIT_IN_4 * LM78_INIT_IN_PERCENTAGE / 100)
+#define LM78_INIT_IN_MIN_5 \
+ (LM78_INIT_IN_5 - LM78_INIT_IN_5 * LM78_INIT_IN_PERCENTAGE / 100)
+#define LM78_INIT_IN_MAX_5 \
+ (LM78_INIT_IN_5 + LM78_INIT_IN_5 * LM78_INIT_IN_PERCENTAGE / 100)
+#define LM78_INIT_IN_MIN_6 \
+ (LM78_INIT_IN_6 - LM78_INIT_IN_6 * LM78_INIT_IN_PERCENTAGE / 100)
+#define LM78_INIT_IN_MAX_6 \
+ (LM78_INIT_IN_6 + LM78_INIT_IN_6 * LM78_INIT_IN_PERCENTAGE / 100)
+
+#define LM78_INIT_FAN_MIN_1 3000
+#define LM78_INIT_FAN_MIN_2 3000
+#define LM78_INIT_FAN_MIN_3 3000
+
+#define LM78_INIT_TEMP_OVER 60000
+#define LM78_INIT_TEMP_HYST 50000
+
+/* There are some complications in a module like this. First off, LM78 chips
+ may be both present on the SMBus and the ISA bus, and we have to handle
+ those cases separately at some places. Second, there might be several
+ LM78 chips available (well, actually, that is probably never done; but
+ it is a clean illustration of how to handle a case like that). Finally,
+ a specific chip may be attached to *both* ISA and SMBus, and we would
+ not like to detect it double. Fortunately, in the case of the LM78 at
+ least, a register tells us what SMBus address we are on, so that helps
+ a bit - except if there could be more than one SMBus. Groan. No solution
+ for this yet. */
+
+/* This module may seem overly long and complicated. In fact, it is not so
+ bad. Quite a lot of bookkeeping is done. A real driver can often cut
+ some corners. */
+
+/* For each registered LM78, we need to keep some data in memory. That
+ data is pointed to by lm78_list[NR]->data. The structure itself is
+ dynamically allocated, at the same time when a new lm78 client is
+ allocated. */
+struct lm78_data {
+ struct semaphore lock;
+ enum chips type;
+
+ struct semaphore update_lock;
+ char valid; /* !=0 if following fields are valid */
+ unsigned long last_updated; /* In jiffies */
+
+ u8 in[7]; /* Register value */
+ u8 in_max[7]; /* Register value */
+ u8 in_min[7]; /* Register value */
+ u8 fan[3]; /* Register value */
+ u8 fan_min[3]; /* Register value */
+ u8 temp; /* Register value */
+ u8 temp_over; /* Register value */
+ u8 temp_hyst; /* Register value */
+ u8 fan_div[3]; /* Register encoding, shifted right */
+ u8 vid; /* Register encoding, combined */
+ u16 alarms; /* Register encoding, combined */
+};
+
+
+static int lm78_attach_adapter(struct i2c_adapter *adapter);
+static int lm78_detect(struct i2c_adapter *adapter, int address, int kind);
+static int lm78_detach_client(struct i2c_client *client);
+
+static int lm78_read_value(struct i2c_client *client, u8 register);
+static int lm78_write_value(struct i2c_client *client, u8 register, u8 value);
+static void lm78_update_client(struct i2c_client *client);
+static void lm78_init_client(struct i2c_client *client);
+
+
+static struct i2c_driver lm78_driver = {
+ .owner = THIS_MODULE,
+ .name = "lm78",
+ .id = I2C_DRIVERID_LM78,
+ .flags = I2C_DF_NOTIFY,
+ .attach_adapter = lm78_attach_adapter,
+ .detach_client = lm78_detach_client,
+};
+
+/* 7 Voltages */
+static ssize_t show_in(struct device *dev, char *buf, int nr)
+{
+ struct i2c_client *client = to_i2c_client(dev);
+ struct lm78_data *data = i2c_get_clientdata(client);
+ lm78_update_client(client);
+ return sprintf(buf, "%d\n", IN_FROM_REG(data->in[nr]));
+}
+
+static ssize_t show_in_min(struct device *dev, char *buf, int nr)
+{
+ struct i2c_client *client = to_i2c_client(dev);
+ struct lm78_data *data = i2c_get_clientdata(client);
+ lm78_update_client(client);
+ return sprintf(buf, "%d\n", IN_FROM_REG(data->in_min[nr]));
+}
+
+static ssize_t show_in_max(struct device *dev, char *buf, int nr)
+{
+ struct i2c_client *client = to_i2c_client(dev);
+ struct lm78_data *data = i2c_get_clientdata(client);
+ lm78_update_client(client);
+ return sprintf(buf, "%d\n", IN_FROM_REG(data->in_max[nr]));
+}
+
+static ssize_t set_in_min(struct device *dev, const char *buf,
+ size_t count, int nr)
+{
+ struct i2c_client *client = to_i2c_client(dev);
+ struct lm78_data *data = i2c_get_clientdata(client);
+ unsigned long val = simple_strtoul(buf, NULL, 10);
+ data->in_min[nr] = IN_TO_REG(val);
+ lm78_write_value(client, LM78_REG_IN_MIN(nr), data->in_min[nr]);
+ return count;
+}
+
+static ssize_t set_in_max(struct device *dev, const char *buf,
+ size_t count, int nr)
+{
+ struct i2c_client *client = to_i2c_client(dev);
+ struct lm78_data *data = i2c_get_clientdata(client);
+ unsigned long val = simple_strtoul(buf, NULL, 10);
+ data->in_max[nr] = IN_TO_REG(val);
+ lm78_write_value(client, LM78_REG_IN_MAX(nr), data->in_max[nr]);
+ return count;
+}
+
+#define show_in_offset(offset) \
+static ssize_t \
+ show_in##offset (struct device *dev, char *buf) \
+{ \
+ return show_in(dev, buf, 0x##offset); \
+} \
+static DEVICE_ATTR(in_input##offset, S_IRUGO, \
+ show_in##offset, NULL) \
+static ssize_t \
+ show_in##offset##_min (struct device *dev, char *buf) \
+{ \
+ return show_in_min(dev, buf, 0x##offset); \
+} \
+static ssize_t \
+ show_in##offset##_max (struct device *dev, char *buf) \
+{ \
+ return show_in_max(dev, buf, 0x##offset); \
+} \
+static ssize_t set_in##offset##_min (struct device *dev, \
+ const char *buf, size_t count) \
+{ \
+ return set_in_min(dev, buf, count, 0x##offset); \
+} \
+static ssize_t set_in##offset##_max (struct device *dev, \
+ const char *buf, size_t count) \
+{ \
+ return set_in_max(dev, buf, count, 0x##offset); \
+} \
+static DEVICE_ATTR(in_min##offset, S_IRUGO | S_IWUSR, \
+ show_in##offset##_min, set_in##offset##_min) \
+static DEVICE_ATTR(in_max##offset, S_IRUGO | S_IWUSR, \
+ show_in##offset##_max, set_in##offset##_max)
+
+show_in_offset(0);
+show_in_offset(1);
+show_in_offset(2);
+show_in_offset(3);
+show_in_offset(4);
+show_in_offset(5);
+show_in_offset(6);
+
+/* Temperature */
+static ssize_t show_temp(struct device *dev, char *buf)
+{
+ struct i2c_client *client = to_i2c_client(dev);
+ struct lm78_data *data = i2c_get_clientdata(client);
+ lm78_update_client(client);
+ return sprintf(buf, "%d\n", TEMP_FROM_REG(data->temp));
+}
+
+static ssize_t show_temp_over(struct device *dev, char *buf)
+{
+ struct i2c_client *client = to_i2c_client(dev);
+ struct lm78_data *data = i2c_get_clientdata(client);
+ lm78_update_client(client);
+ return sprintf(buf, "%d\n", TEMP_FROM_REG(data->temp_over));
+}
+
+static ssize_t set_temp_over(struct device *dev, const char *buf, size_t count)
+{
+ struct i2c_client *client = to_i2c_client(dev);
+ struct lm78_data *data = i2c_get_clientdata(client);
+ long val = simple_strtol(buf, NULL, 10);
+ data->temp_over = TEMP_TO_REG(val);
+ lm78_write_value(client, LM78_REG_TEMP_OVER, data->temp_over);
+ return count;
+}
+
+static ssize_t show_temp_hyst(struct device *dev, char *buf)
+{
+ struct i2c_client *client = to_i2c_client(dev);
+ struct lm78_data *data = i2c_get_clientdata(client);
+ lm78_update_client(client);
+ return sprintf(buf, "%d\n", TEMP_FROM_REG(data->temp_hyst));
+}
+
+static ssize_t set_temp_hyst(struct device *dev, const char *buf, size_t count)
+{
+ struct i2c_client *client = to_i2c_client(dev);
+ struct lm78_data *data = i2c_get_clientdata(client);
+ long val = simple_strtol(buf, NULL, 10);
+ data->temp_hyst = TEMP_TO_REG(val);
+ lm78_write_value(client, LM78_REG_TEMP_HYST, data->temp_hyst);
+ return count;
+}
+
+static DEVICE_ATTR(temp_input, S_IRUGO, show_temp, NULL)
+static DEVICE_ATTR(temp_max, S_IRUGO | S_IWUSR,
+ show_temp_over, set_temp_over)
+static DEVICE_ATTR(temp_min, S_IRUGO | S_IWUSR,
+ show_temp_hyst, set_temp_hyst)
+
+/* 3 Fans */
+static ssize_t show_fan(struct device *dev, char *buf, int nr)
+{
+ struct i2c_client *client = to_i2c_client(dev);
+ struct lm78_data *data = i2c_get_clientdata(client);
+ lm78_update_client(client);
+ return sprintf(buf, "%d\n", FAN_FROM_REG(data->fan[nr],
+ DIV_FROM_REG(data->fan_div[nr])) );
+}
+
+static ssize_t show_fan_min(struct device *dev, char *buf, int nr)
+{
+ struct i2c_client *client = to_i2c_client(dev);
+ struct lm78_data *data = i2c_get_clientdata(client);
+ lm78_update_client(client);
+ return sprintf(buf,"%d\n", FAN_FROM_REG(data->fan_min[nr],
+ DIV_FROM_REG(data->fan_div[nr])) );
+}
+
+static ssize_t set_fan_min(struct device *dev, const char *buf,
+ size_t count, int nr)
+{
+ struct i2c_client *client = to_i2c_client(dev);
+ struct lm78_data *data = i2c_get_clientdata(client);
+ unsigned long val = simple_strtoul(buf, NULL, 10);
+ data->fan_min[nr] = FAN_TO_REG(val, DIV_FROM_REG(data->fan_div[nr]));
+ lm78_write_value(client, LM78_REG_FAN_MIN(nr), data->fan_min[nr]);
+ return count;
+}
+
+static ssize_t show_fan_div(struct device *dev, char *buf, int nr)
+{
+ struct i2c_client *client = to_i2c_client(dev);
+ struct lm78_data *data = i2c_get_clientdata(client);
+ lm78_update_client(client);
+ return sprintf(buf, "%d\n", DIV_FROM_REG(data->fan_div[nr]) );
+}
+
+/* Note: we save and restore the fan minimum here, because its value is
+ determined in part by the fan divisor. This follows the principle of
+ least suprise; the user doesn't expect the fan minimum to change just
+ because the divisor changed. */
+static ssize_t set_fan_div(struct device *dev, const char *buf,
+ size_t count, int nr)
+{
+ struct i2c_client *client = to_i2c_client(dev);
+ struct lm78_data *data = i2c_get_clientdata(client);
+ unsigned long min = FAN_FROM_REG(data->fan_min[nr],
+ DIV_FROM_REG(data->fan_div[nr]));
+ unsigned long val = simple_strtoul(buf, NULL, 10);
+ int reg = lm78_read_value(client, LM78_REG_VID_FANDIV);
+ data->fan_div[nr] = DIV_TO_REG(val);
+ switch (nr) {
+ case 0:
+ reg = (reg & 0xcf) | (data->fan_div[nr] << 4);
+ break;
+ case 1:
+ reg = (reg & 0x3f) | (data->fan_div[nr] << 6);
+ break;
+ }
+ lm78_write_value(client, LM78_REG_VID_FANDIV, reg);
+ data->fan_min[nr] + FAN_TO_REG(min, DIV_FROM_REG(data->fan_div[nr]));
+ lm78_write_value(client, LM78_REG_FAN_MIN(nr), data->fan_min[nr]);
+ return count;
+}
+
+#define show_fan_offset(offset) \
+static ssize_t show_fan_##offset (struct device *dev, char *buf) \
+{ \
+ return show_fan(dev, buf, 0x##offset - 1); \
+} \
+static ssize_t show_fan_##offset##_min (struct device *dev, char *buf) \
+{ \
+ return show_fan_min(dev, buf, 0x##offset - 1); \
+} \
+static ssize_t show_fan_##offset##_div (struct device *dev, char *buf) \
+{ \
+ return show_fan_div(dev, buf, 0x##offset - 1); \
+} \
+static ssize_t set_fan_##offset##_min (struct device *dev, \
+ const char *buf, size_t count) \
+{ \
+ return set_fan_min(dev, buf, count, 0x##offset - 1); \
+} \
+static DEVICE_ATTR(fan_input##offset, S_IRUGO, show_fan_##offset, NULL) \
+static DEVICE_ATTR(fan_min##offset, S_IRUGO | S_IWUSR, \
+ show_fan_##offset##_min, set_fan_##offset##_min)
+
+static ssize_t set_fan_1_div(struct device *dev, const char *buf,
+ size_t count)
+{
+ return set_fan_div(dev, buf, count, 0) ;
+}
+
+static ssize_t set_fan_2_div(struct device *dev, const char *buf,
+ size_t count)
+{
+ return set_fan_div(dev, buf, count, 1) ;
+}
+
+show_fan_offset(1);
+show_fan_offset(2);
+show_fan_offset(3);
+
+/* Fan 3 divisor is locked in H/W */
+static DEVICE_ATTR(fan_div1, S_IRUGO | S_IWUSR,
+ show_fan_1_div, set_fan_1_div)
+static DEVICE_ATTR(fan_div2, S_IRUGO | S_IWUSR,
+ show_fan_2_div, set_fan_2_div)
+static DEVICE_ATTR(fan_div3, S_IRUGO, show_fan_3_div, NULL)
+
+/* VID */
+static ssize_t show_vid(struct device *dev, char *buf)
+{
+ struct i2c_client *client = to_i2c_client(dev);
+ struct lm78_data *data = i2c_get_clientdata(client);
+ lm78_update_client(client);
+ return sprintf(buf, "%d\n", VID_FROM_REG(data->vid));
+}
+static DEVICE_ATTR(vid, S_IRUGO, show_vid, NULL);
+
+/* Alarms */
+static ssize_t show_alarms(struct device *dev, char *buf)
+{
+ struct i2c_client *client = to_i2c_client(dev);
+ struct lm78_data *data = i2c_get_clientdata(client);
+ lm78_update_client(client);
+ return sprintf(buf, "%d\n", ALARMS_FROM_REG(data->alarms));
+}
+static DEVICE_ATTR(alarms, S_IRUGO, show_alarms, NULL);
+
+/* This function is called when:
+ * lm78_driver is inserted (when this module is loaded), for each
+ available adapter
+ * when a new adapter is inserted (and lm78_driver is still present) */
+static int lm78_attach_adapter(struct i2c_adapter *adapter)
+{
+ if (!(adapter->class & I2C_ADAP_CLASS_SMBUS))
+ return 0;
+ return i2c_detect(adapter, &addr_data, lm78_detect);
+}
+
+/* This function is called by i2c_detect */
+int lm78_detect(struct i2c_adapter *adapter, int address, int kind)
+{
+ int i, err;
+ struct i2c_client *new_client;
+ struct lm78_data *data;
+ const char *client_name = "";
+ int is_isa = i2c_is_isa_adapter(adapter);
+
+ if (!is_isa &&
+ !i2c_check_functionality(adapter, I2C_FUNC_SMBUS_BYTE_DATA)) {
+ err = -ENODEV;
+ goto ERROR0;
+ }
+
+ /* Reserve the ISA region */
+ if (is_isa)
+ if (!request_region(address, LM78_EXTENT, "lm78")) {
+ err = -EBUSY;
+ goto ERROR0;
+ }
+
+ /* Probe whether there is anything available on this address. Already
+ done for SMBus clients */
+ if (kind < 0) {
+ if (is_isa) {
+
+#define REALLY_SLOW_IO
+ /* We need the timeouts for at least some LM78-like
+ chips. But only if we read 'undefined' registers. */
+ i = inb_p(address + 1);
+ if (inb_p(address + 2) != i) {
+ err = -ENODEV;
+ goto ERROR1;
+ }
+ if (inb_p(address + 3) != i) {
+ err = -ENODEV;
+ goto ERROR1;
+ }
+ if (inb_p(address + 7) != i) {
+ err = -ENODEV;
+ goto ERROR1;
+ }
+#undef REALLY_SLOW_IO
+
+ /* Let's just hope nothing breaks here */
+ i = inb_p(address + 5) & 0x7f;
+ outb_p(~i & 0x7f, address + 5);
+ if ((inb_p(address + 5) & 0x7f) != (~i & 0x7f)) {
+ outb_p(i, address + 5);
+ err = -ENODEV;
+ goto ERROR1;
+ }
+ }
+ }
+
+ /* OK. For now, we presume we have a valid client. We now create the
+ client structure, even though we cannot fill it completely yet.
+ But it allows us to access lm78_{read,write}_value. */
+
+ if (!(new_client = kmalloc((sizeof(struct i2c_client)) +
+ sizeof(struct lm78_data),
+ GFP_KERNEL))) {
+ err = -ENOMEM;
+ goto ERROR1;
+ }
+ memset(new_client, 0, sizeof(struct i2c_client) +
+ sizeof(struct lm78_data));
+
+ data = (struct lm78_data *) (new_client + 1);
+ if (is_isa)
+ init_MUTEX(&data->lock);
+ i2c_set_clientdata(new_client, data);
+ new_client->addr = address;
+ new_client->adapter = adapter;
+ new_client->driver = &lm78_driver;
+ new_client->flags = 0;
+
+ /* Now, we do the remaining detection. */
+ if (kind < 0) {
+ if (lm78_read_value(new_client, LM78_REG_CONFIG) & 0x80) {
+ err = -ENODEV;
+ goto ERROR2;
+ }
+ if (!is_isa && (lm78_read_value(
+ new_client, LM78_REG_I2C_ADDR) != address)) {
+ err = -ENODEV;
+ goto ERROR2;
+ }
+ }
+
+ /* Determine the chip type. */
+ if (kind <= 0) {
+ i = lm78_read_value(new_client, LM78_REG_CHIPID);
+ if (i = 0x00 || i = 0x20)
+ kind = lm78;
+ else if (i = 0x40)
+ kind = lm78j;
+ else if ((i & 0xfe) = 0xc0)
+ kind = lm79;
+ else {
+ if (kind = 0)
+ printk(KERN_WARNING "lm78.o: Ignoring 'force' "
+ "parameter for unknown chip at "
+ "adapter %d, address 0x%02x\n",
+ i2c_adapter_id(adapter), address);
+ err = -ENODEV;
+ goto ERROR2;
+ }
+ }
+
+ if (kind = lm78) {
+ client_name = "LM78 chip";
+ } else if (kind = lm78j) {
+ client_name = "LM78-J chip";
+ } else if (kind = lm79) {
+ client_name = "LM79 chip";
+ } else {
+ dev_dbg(&adapter->dev, "Internal error: unknown kind (%d)?!?",
+ kind);
+ err = -ENODEV;
+ goto ERROR2;
+ }
+
+ /* Fill in the remaining client fields and put into the global list */
+ strlcpy(new_client->dev.name, client_name, DEVICE_NAME_SIZE);
+ data->type = kind;
+
+ data->valid = 0;
+ init_MUTEX(&data->update_lock);
+
+ /* Tell the I2C layer a new client has arrived */
+ if ((err = i2c_attach_client(new_client)))
+ goto ERROR2;
+
+ /* register sysfs hooks */
+ device_create_file(&new_client->dev, &dev_attr_in_input0);
+ device_create_file(&new_client->dev, &dev_attr_in_min0);
+ device_create_file(&new_client->dev, &dev_attr_in_max0);
+ device_create_file(&new_client->dev, &dev_attr_in_input1);
+ device_create_file(&new_client->dev, &dev_attr_in_min1);
+ device_create_file(&new_client->dev, &dev_attr_in_max1);
+ device_create_file(&new_client->dev, &dev_attr_in_input2);
+ device_create_file(&new_client->dev, &dev_attr_in_min2);
+ device_create_file(&new_client->dev, &dev_attr_in_max2);
+ device_create_file(&new_client->dev, &dev_attr_in_input3);
+ device_create_file(&new_client->dev, &dev_attr_in_min3);
+ device_create_file(&new_client->dev, &dev_attr_in_max3);
+ device_create_file(&new_client->dev, &dev_attr_in_input4);
+ device_create_file(&new_client->dev, &dev_attr_in_min4);
+ device_create_file(&new_client->dev, &dev_attr_in_max4);
+ device_create_file(&new_client->dev, &dev_attr_in_input5);
+ device_create_file(&new_client->dev, &dev_attr_in_min5);
+ device_create_file(&new_client->dev, &dev_attr_in_max5);
+ device_create_file(&new_client->dev, &dev_attr_in_input6);
+ device_create_file(&new_client->dev, &dev_attr_in_min6);
+ device_create_file(&new_client->dev, &dev_attr_in_max6);
+ device_create_file(&new_client->dev, &dev_attr_temp_input);
+ device_create_file(&new_client->dev, &dev_attr_temp_min);
+ device_create_file(&new_client->dev, &dev_attr_temp_max);
+ device_create_file(&new_client->dev, &dev_attr_fan_input1);
+ device_create_file(&new_client->dev, &dev_attr_fan_min1);
+ device_create_file(&new_client->dev, &dev_attr_fan_div1);
+ device_create_file(&new_client->dev, &dev_attr_fan_input2);
+ device_create_file(&new_client->dev, &dev_attr_fan_min2);
+ device_create_file(&new_client->dev, &dev_attr_fan_div2);
+ device_create_file(&new_client->dev, &dev_attr_fan_input3);
+ device_create_file(&new_client->dev, &dev_attr_fan_min3);
+ device_create_file(&new_client->dev, &dev_attr_fan_div3);
+ device_create_file(&new_client->dev, &dev_attr_alarms);
+ device_create_file(&new_client->dev, &dev_attr_vid);
+
+ /* Initialize the LM78 chip */
+ lm78_init_client(new_client);
+ return 0;
+
+ERROR2:
+ kfree(new_client);
+ERROR1:
+ if (is_isa)
+ release_region(address, LM78_EXTENT);
+ERROR0:
+ return err;
+}
+
+static int lm78_detach_client(struct i2c_client *client)
+{
+ int err;
+
+ /* release ISA region first */
+ if(i2c_is_isa_client(client))
+ release_region(client->addr, LM78_EXTENT);
+
+ /* now it's safe to scrap the rest */
+ if ((err = i2c_detach_client(client))) {
+ dev_err(&client->dev,
+ "Client deregistration failed, client not detached.\n");
+ return err;
+ }
+
+ kfree(client);
+
+ return 0;
+}
+
+/* The SMBus locks itself, but ISA access must be locked explicitely!
+ We don't want to lock the whole ISA bus, so we lock each client
+ separately.
+ We ignore the LM78 BUSY flag at this moment - it could lead to deadlocks,
+ would slow down the LM78 access and should not be necessary.
+ There are some ugly typecasts here, but the good new is - they should
+ nowhere else be necessary! */
+static int lm78_read_value(struct i2c_client *client, u8 reg)
+{
+ int res;
+ if (i2c_is_isa_client(client)) {
+ struct lm78_data *data = i2c_get_clientdata(client);
+ down(&data->lock);
+ outb_p(reg, client->addr + LM78_ADDR_REG_OFFSET);
+ res = inb_p(client->addr + LM78_DATA_REG_OFFSET);
+ up(&data->lock);
+ return res;
+ } else
+ return i2c_smbus_read_byte_data(client, reg);
+}
+
+/* The SMBus locks itself, but ISA access muse be locked explicitely!
+ We don't want to lock the whole ISA bus, so we lock each client
+ separately.
+ We ignore the LM78 BUSY flag at this moment - it could lead to deadlocks,
+ would slow down the LM78 access and should not be necessary.
+ There are some ugly typecasts here, but the good new is - they should
+ nowhere else be necessary! */
+static int lm78_write_value(struct i2c_client *client, u8 reg, u8 value)
+{
+ if (i2c_is_isa_client(client)) {
+ struct lm78_data *data = i2c_get_clientdata(client);
+ down(&data->lock);
+ outb_p(reg, client->addr + LM78_ADDR_REG_OFFSET);
+ outb_p(value, client->addr + LM78_DATA_REG_OFFSET);
+ up(&data->lock);
+ return 0;
+ } else
+ return i2c_smbus_write_byte_data(client, reg, value);
+}
+
+/* Called when we have found a new LM78. It should set limits, etc. */
+static void lm78_init_client(struct i2c_client *client)
+{
+ struct lm78_data *data = i2c_get_clientdata(client);
+ int vid;
+
+ /* Reset all except Watchdog values and last conversion values
+ This sets fan-divs to 2, among others */
+ lm78_write_value(client, LM78_REG_CONFIG, 0x80);
+
+ vid = lm78_read_value(client, LM78_REG_VID_FANDIV) & 0x0f;
+ if (data->type = lm79)
+ vid |+ (lm78_read_value(client, LM78_REG_CHIPID) & 0x01) << 4;
+ else
+ vid |= 0x10;
+ vid = VID_FROM_REG(vid);
+
+ lm78_write_value(client, LM78_REG_IN_MIN(0),
+ IN_TO_REG(LM78_INIT_IN_MIN_0(vid)));
+ lm78_write_value(client, LM78_REG_IN_MAX(0),
+ IN_TO_REG(LM78_INIT_IN_MAX_0(vid)));
+ lm78_write_value(client, LM78_REG_IN_MIN(1),
+ IN_TO_REG(LM78_INIT_IN_MIN_1(vid)));
+ lm78_write_value(client, LM78_REG_IN_MAX(1),
+ IN_TO_REG(LM78_INIT_IN_MAX_1(vid)));
+ lm78_write_value(client, LM78_REG_IN_MIN(2),
+ IN_TO_REG(LM78_INIT_IN_MIN_2));
+ lm78_write_value(client, LM78_REG_IN_MAX(2),
+ IN_TO_REG(LM78_INIT_IN_MAX_2));
+ lm78_write_value(client, LM78_REG_IN_MIN(3),
+ IN_TO_REG(LM78_INIT_IN_MIN_3));
+ lm78_write_value(client, LM78_REG_IN_MAX(3),
+ IN_TO_REG(LM78_INIT_IN_MAX_3));
+ lm78_write_value(client, LM78_REG_IN_MIN(4),
+ IN_TO_REG(LM78_INIT_IN_MIN_4));
+ lm78_write_value(client, LM78_REG_IN_MAX(4),
+ IN_TO_REG(LM78_INIT_IN_MAX_4));
+ lm78_write_value(client, LM78_REG_IN_MIN(5),
+ IN_TO_REG(LM78_INIT_IN_MIN_5));
+ lm78_write_value(client, LM78_REG_IN_MAX(5),
+ IN_TO_REG(LM78_INIT_IN_MAX_5));
+ lm78_write_value(client, LM78_REG_IN_MIN(6),
+ IN_TO_REG(LM78_INIT_IN_MIN_6));
+ lm78_write_value(client, LM78_REG_IN_MAX(6),
+ IN_TO_REG(LM78_INIT_IN_MAX_6));
+ lm78_write_value(client, LM78_REG_FAN_MIN(0),
+ FAN_TO_REG(LM78_INIT_FAN_MIN_1, 2));
+ lm78_write_value(client, LM78_REG_FAN_MIN(1),
+ FAN_TO_REG(LM78_INIT_FAN_MIN_2, 2));
+ lm78_write_value(client, LM78_REG_FAN_MIN(2),
+ FAN_TO_REG(LM78_INIT_FAN_MIN_3, 2));
+ lm78_write_value(client, LM78_REG_TEMP_OVER,
+ TEMP_TO_REG(LM78_INIT_TEMP_OVER));
+ lm78_write_value(client, LM78_REG_TEMP_HYST,
+ TEMP_TO_REG(LM78_INIT_TEMP_HYST));
+
+ /* Start monitoring */
+ lm78_write_value(client, LM78_REG_CONFIG,
+ (lm78_read_value(client, LM78_REG_CONFIG) & 0xf7)
+ | 0x01);
+
+}
+
+static void lm78_update_client(struct i2c_client *client)
+{
+ struct lm78_data *data = i2c_get_clientdata(client);
+ int i;
+
+ down(&data->update_lock);
+
+ if ((jiffies - data->last_updated > HZ + HZ / 2) ||
+ (jiffies < data->last_updated) || !data->valid) {
+
+ dev_dbg(&client->dev, "Starting lm78 update\n");
+
+ for (i = 0; i <= 6; i++) {
+ data->in[i] + lm78_read_value(client, LM78_REG_IN(i));
+ data->in_min[i] + lm78_read_value(client, LM78_REG_IN_MIN(i));
+ data->in_max[i] + lm78_read_value(client, LM78_REG_IN_MAX(i));
+ }
+ for (i = 0; i < 3; i++) {
+ data->fan[i] + lm78_read_value(client, LM78_REG_FAN(i));
+ data->fan_min[i] + lm78_read_value(client, LM78_REG_FAN_MIN(i));
+ }
+ data->temp = lm78_read_value(client, LM78_REG_TEMP);
+ data->temp_over + lm78_read_value(client, LM78_REG_TEMP_OVER);
+ data->temp_hyst + lm78_read_value(client, LM78_REG_TEMP_HYST);
+ i = lm78_read_value(client, LM78_REG_VID_FANDIV);
+ data->vid = i & 0x0f;
+ if (data->type = lm79)
+ data->vid |+ (lm78_read_value(client, LM78_REG_CHIPID) &
+ 0x01) << 4;
+ else
+ data->vid |= 0x10;
+ data->fan_div[0] = (i >> 4) & 0x03;
+ data->fan_div[1] = i >> 6;
+ data->alarms = lm78_read_value(client, LM78_REG_ALARM1) +
+ (lm78_read_value(client, LM78_REG_ALARM2) << 8);
+ data->last_updated = jiffies;
+ data->valid = 1;
+
+ data->fan_div[2] = 1;
+ }
+
+ up(&data->update_lock);
+}
+
+static int __init sm_lm78_init(void)
+{
+ return i2c_add_driver(&lm78_driver);
+}
+
+static void __exit sm_lm78_exit(void)
+{
+ i2c_del_driver(&lm78_driver);
+}
+
+
+
+MODULE_AUTHOR("Frodo Looijaard <frodol@dds.nl>");
+MODULE_DESCRIPTION("LM78, LM78-J and LM79 driver");
+MODULE_LICENSE("GPL");
+
+module_init(sm_lm78_init);
+module_exit(sm_lm78_exit);
^ permalink raw reply [flat|nested] 32+ messages in thread
* [PATCH] i2c driver changes for 2.5.72
@ 2005-05-19 6:24 ` Greg KH
0 siblings, 0 replies; 32+ messages in thread
From: Greg KH @ 2005-05-19 6:24 UTC (permalink / raw)
To: linux-kernel, sensors
ChangeSet 1.1318.3.3, 2003/06/16 11:31:55-07:00, mhoffman@lightlink.com
[PATCH] I2C: w83781d bugfix
My first patch was naive; the patch below solves the problem by
letting w83781d_detach_client remove the three clients (1 * primary
+ 2 * subclients) independently. It's a noisy patch because I had
to change the way the subclients were kmalloc'ed - sorry. The meat
is around line 1422. This patch works for me... comments?
drivers/i2c/chips/w83781d.c | 76 +++++++++++++++++++++++---------------------
1 files changed, 40 insertions(+), 36 deletions(-)
diff -Nru a/drivers/i2c/chips/w83781d.c b/drivers/i2c/chips/w83781d.c
--- a/drivers/i2c/chips/w83781d.c Wed Jun 18 11:19:39 2003
+++ b/drivers/i2c/chips/w83781d.c Wed Jun 18 11:19:39 2003
@@ -299,8 +299,8 @@
char valid; /* !=0 if following fields are valid */
unsigned long last_updated; /* In jiffies */
- struct i2c_client *lm75; /* for secondary I2C addresses */
- /* pointer to array of 2 subclients */
+ struct i2c_client *lm75[2]; /* for secondary I2C addresses */
+ /* array of 2 pointers to subclients */
u8 in[9]; /* Register value - 8 & 9 for 782D only */
u8 in_max[9]; /* Register value - 8 & 9 for 782D only */
@@ -1043,12 +1043,12 @@
const char *client_name;
struct w83781d_data *data = i2c_get_clientdata(new_client);
- if (!(data->lm75 = kmalloc(2 * sizeof (struct i2c_client),
- GFP_KERNEL))) {
+ data->lm75[0] = kmalloc(sizeof(struct i2c_client), GFP_KERNEL);
+ if (!(data->lm75[0])) {
err = -ENOMEM;
goto ERROR_SC_0;
}
- memset(data->lm75, 0x00, 2 * sizeof (struct i2c_client));
+ memset(data->lm75[0], 0x00, sizeof (struct i2c_client));
id = i2c_adapter_id(adapter);
@@ -1066,25 +1066,33 @@
w83781d_write_value(new_client, W83781D_REG_I2C_SUBADDR,
(force_subclients[2] & 0x07) |
((force_subclients[3] & 0x07) << 4));
- data->lm75[0].addr = force_subclients[2];
+ data->lm75[0]->addr = force_subclients[2];
} else {
val1 = w83781d_read_value(new_client, W83781D_REG_I2C_SUBADDR);
- data->lm75[0].addr = 0x48 + (val1 & 0x07);
+ data->lm75[0]->addr = 0x48 + (val1 & 0x07);
}
if (kind != w83783s) {
+
+ data->lm75[1] = kmalloc(sizeof(struct i2c_client), GFP_KERNEL);
+ if (!(data->lm75[1])) {
+ err = -ENOMEM;
+ goto ERROR_SC_1;
+ }
+ memset(data->lm75[1], 0x0, sizeof(struct i2c_client));
+
if (force_subclients[0] = id &&
force_subclients[1] = address) {
- data->lm75[1].addr = force_subclients[3];
+ data->lm75[1]->addr = force_subclients[3];
} else {
- data->lm75[1].addr = 0x48 + ((val1 >> 4) & 0x07);
+ data->lm75[1]->addr = 0x48 + ((val1 >> 4) & 0x07);
}
- if (data->lm75[0].addr = data->lm75[1].addr) {
+ if (data->lm75[0]->addr = data->lm75[1]->addr) {
dev_err(&new_client->dev,
"Duplicate addresses 0x%x for subclients.\n",
- data->lm75[0].addr);
+ data->lm75[0]->addr);
err = -EBUSY;
- goto ERROR_SC_1;
+ goto ERROR_SC_2;
}
}
@@ -1103,19 +1111,19 @@
for (i = 0; i <= 1; i++) {
/* store all data in w83781d */
- i2c_set_clientdata(&data->lm75[i], NULL);
- data->lm75[i].adapter = adapter;
- data->lm75[i].driver = &w83781d_driver;
- data->lm75[i].flags = 0;
- strlcpy(data->lm75[i].dev.name, client_name,
+ i2c_set_clientdata(data->lm75[i], NULL);
+ data->lm75[i]->adapter = adapter;
+ data->lm75[i]->driver = &w83781d_driver;
+ data->lm75[i]->flags = 0;
+ strlcpy(data->lm75[i]->dev.name, client_name,
DEVICE_NAME_SIZE);
- if ((err = i2c_attach_client(&(data->lm75[i])))) {
+ if ((err = i2c_attach_client(data->lm75[i]))) {
dev_err(&new_client->dev, "Subclient %d "
"registration at address 0x%x "
- "failed.\n", i, data->lm75[i].addr);
+ "failed.\n", i, data->lm75[i]->addr);
if (i = 1)
- goto ERROR_SC_2;
- goto ERROR_SC_1;
+ goto ERROR_SC_3;
+ goto ERROR_SC_2;
}
if (kind = w83783s)
break;
@@ -1124,10 +1132,14 @@
return 0;
/* Undo inits in case of errors */
+ERROR_SC_3:
+ i2c_detach_client(data->lm75[0]);
ERROR_SC_2:
- i2c_detach_client(&(data->lm75[0]));
+ if (NULL != data->lm75[1])
+ kfree(data->lm75[1]);
ERROR_SC_1:
- kfree(data->lm75);
+ if (NULL != data->lm75[0])
+ kfree(data->lm75[0]);
ERROR_SC_0:
return err;
}
@@ -1326,7 +1338,8 @@
kind, new_client)))
goto ERROR3;
} else {
- data->lm75 = NULL;
+ data->lm75[0] = NULL;
+ data->lm75[1] = NULL;
}
device_create_file_in(new_client, 0);
@@ -1409,20 +1422,11 @@
static int
w83781d_detach_client(struct i2c_client *client)
{
- struct w83781d_data *data = i2c_get_clientdata(client);
int err;
- /* release ISA region or I2C subclients first */
- if (i2c_is_isa_client(client)) {
+ if (i2c_is_isa_client(client))
release_region(client->addr, W83781D_EXTENT);
- } else {
- i2c_detach_client(&data->lm75[0]);
- if (data->type != w83783s)
- i2c_detach_client(&data->lm75[1]);
- kfree(data->lm75);
- }
- /* now it's safe to scrap the rest */
if ((err = i2c_detach_client(client))) {
dev_err(&client->dev,
"Client deregistration failed, client not detached.\n");
@@ -1484,7 +1488,7 @@
res = i2c_smbus_read_byte_data(client, reg & 0xff);
} else {
/* switch to subclient */
- cl = &data->lm75[bank - 1];
+ cl = data->lm75[bank - 1];
/* convert from ISA to LM75 I2C addresses */
switch (reg & 0xff) {
case 0x50: /* TEMP */
@@ -1555,7 +1559,7 @@
value & 0xff);
} else {
/* switch to subclient */
- cl = &data->lm75[bank - 1];
+ cl = data->lm75[bank - 1];
/* convert from ISA to LM75 I2C addresses */
switch (reg & 0xff) {
case 0x52: /* CONFIG */
^ permalink raw reply [flat|nested] 32+ messages in thread
* [PATCH] i2c driver changes for 2.5.72
@ 2005-05-19 6:24 ` Greg KH
0 siblings, 0 replies; 32+ messages in thread
From: Greg KH @ 2005-05-19 6:24 UTC (permalink / raw)
To: linux-kernel, sensors
ChangeSet 1.1318.3.2, 2003/06/16 11:31:43-07:00, margitsw@t-online.de
[PATCH] I2C: lm85 fixups
OK Here's the patch which :
1) Fixes the race conditions
2) Correctly reports the temps :-)
3) Removes a bit of gunk in the defines which I forgot
drivers/i2c/chips/lm85.c | 41 ++++++++++++++++++++++++++++-------------
1 files changed, 28 insertions(+), 13 deletions(-)
diff -Nru a/drivers/i2c/chips/lm85.c b/drivers/i2c/chips/lm85.c
--- a/drivers/i2c/chips/lm85.c Wed Jun 18 11:19:44 2003
+++ b/drivers/i2c/chips/lm85.c Wed Jun 18 11:19:44 2003
@@ -148,20 +148,17 @@
#define SCALE(val,from,to) (((val)*(to) + ((from)/2))/(from))
#define INS_TO_REG(n,val) (SENSORS_LIMIT(SCALE(val,lm85_scaling[n],192),0,255))
#define INSEXT_FROM_REG(n,val,ext) (SCALE((val)*4 + (ext),192*4,lm85_scaling[n]))
-/*
#define INS_FROM_REG(n,val) (INSEXT_FROM_REG(n,val,0))
-*/
-#define INS_FROM_REG(n,val) ( ( (val*4*lm85_scaling[n]) + (192*4/2) ) / (192*4) )
/* FAN speed is measured using 90kHz clock */
#define FAN_TO_REG(val) (SENSORS_LIMIT( (val)<=0?0: 5400000/(val),0,65534))
#define FAN_FROM_REG(val) ((val)=0?-1:(val)=0xffff?0:5400000/(val))
-/* Temperature is reported in .01 degC increments */
-#define TEMP_TO_REG(val) (SENSORS_LIMIT(((val)+50)/100,-127,127))
-#define TEMPEXT_FROM_REG(val,ext) ((val)*100 + (ext)*25)
+/* Temperature is reported in .001 degC increments */
+#define TEMP_TO_REG(val) (SENSORS_LIMIT(((val)+500)/1000,-127,127))
+#define TEMPEXT_FROM_REG(val,ext) ((val)*1000 + (ext)*250)
#define TEMP_FROM_REG(val) (TEMPEXT_FROM_REG(val,0))
-#define EXTTEMP_TO_REG(val) (SENSORS_LIMIT((val)/25,-127,127))
+#define EXTTEMP_TO_REG(val) (SENSORS_LIMIT((val)/250,-127,127))
#define PWM_TO_REG(val) (SENSORS_LIMIT(val,0,255))
#define PWM_FROM_REG(val) (val)
@@ -437,10 +434,13 @@
{
struct i2c_client *client = to_i2c_client(dev);
struct lm85_data *data = i2c_get_clientdata(client);
+ int val;
- int val = simple_strtol(buf, NULL, 10);
+ down(&data->update_lock);
+ val = simple_strtol(buf, NULL, 10);
data->fan_min[nr] = FAN_TO_REG(val);
lm85_write_value(client, LM85_REG_FAN_MIN(nr), data->fan_min[nr]);
+ up(&data->update_lock);
return count;
}
@@ -528,10 +528,13 @@
{
struct i2c_client *client = to_i2c_client(dev);
struct lm85_data *data = i2c_get_clientdata(client);
+ int val;
- int val = simple_strtol(buf, NULL, 10);
+ down(&data->update_lock);
+ val = simple_strtol(buf, NULL, 10);
data->pwm[nr] = PWM_TO_REG(val);
lm85_write_value(client, LM85_REG_PWM(nr), data->pwm[nr]);
+ up(&data->update_lock);
return count;
}
static ssize_t show_pwm_enable(struct device *dev, char *buf, int nr)
@@ -590,10 +593,13 @@
{
struct i2c_client *client = to_i2c_client(dev);
struct lm85_data *data = i2c_get_clientdata(client);
+ int val;
- int val = simple_strtol(buf, NULL, 10);
+ down(&data->update_lock);
+ val = simple_strtol(buf, NULL, 10);
data->in_min[nr] = INS_TO_REG(nr, val);
lm85_write_value(client, LM85_REG_IN_MIN(nr), data->in_min[nr]);
+ up(&data->update_lock);
return count;
}
static ssize_t show_in_max(struct device *dev, char *buf, int nr)
@@ -609,10 +615,13 @@
{
struct i2c_client *client = to_i2c_client(dev);
struct lm85_data *data = i2c_get_clientdata(client);
+ int val;
- int val = simple_strtol(buf, NULL, 10);
+ down(&data->update_lock);
+ val = simple_strtol(buf, NULL, 10);
data->in_max[nr] = INS_TO_REG(nr, val);
lm85_write_value(client, LM85_REG_IN_MAX(nr), data->in_max[nr]);
+ up(&data->update_lock);
return count;
}
#define show_in_reg(offset) \
@@ -673,10 +682,13 @@
{
struct i2c_client *client = to_i2c_client(dev);
struct lm85_data *data = i2c_get_clientdata(client);
+ int val;
- int val = simple_strtol(buf, NULL, 10);
+ down(&data->update_lock);
+ val = simple_strtol(buf, NULL, 10);
data->temp_min[nr] = TEMP_TO_REG(val);
lm85_write_value(client, LM85_REG_TEMP_MIN(nr), data->temp_min[nr]);
+ up(&data->update_lock);
return count;
}
static ssize_t show_temp_max(struct device *dev, char *buf, int nr)
@@ -692,10 +704,13 @@
{
struct i2c_client *client = to_i2c_client(dev);
struct lm85_data *data = i2c_get_clientdata(client);
+ int val;
- int val = simple_strtol(buf, NULL, 10);
+ down(&data->update_lock);
+ val = simple_strtol(buf, NULL, 10);
data->temp_max[nr] = TEMP_TO_REG(val);
lm85_write_value(client, LM85_REG_TEMP_MAX(nr), data->temp_max[nr]);
+ up(&data->update_lock);
return count;
}
#define show_temp_reg(offset) \
^ permalink raw reply [flat|nested] 32+ messages in thread
* [PATCH] i2c driver changes for 2.5.72
@ 2005-05-19 6:24 ` Greg KH
0 siblings, 0 replies; 32+ messages in thread
From: Greg KH @ 2005-05-19 6:24 UTC (permalink / raw)
To: linux-kernel, sensors
ChangeSet 1.1318.3.5, 2003/06/16 11:32:48-07:00, greg@kroah.com
[PATCH] I2C: add lm78 chip to Makefile
drivers/i2c/chips/Makefile | 1 +
1 files changed, 1 insertion(+)
diff -Nru a/drivers/i2c/chips/Makefile b/drivers/i2c/chips/Makefile
--- a/drivers/i2c/chips/Makefile Wed Jun 18 11:19:30 2003
+++ b/drivers/i2c/chips/Makefile Wed Jun 18 11:19:30 2003
@@ -5,6 +5,7 @@
obj-$(CONFIG_SENSORS_ADM1021) += adm1021.o
obj-$(CONFIG_SENSORS_IT87) += it87.o
obj-$(CONFIG_SENSORS_LM75) += lm75.o
+obj-$(CONFIG_SENSORS_LM78) += lm78.o
obj-$(CONFIG_SENSORS_LM85) += lm85.o
obj-$(CONFIG_SENSORS_VIA686A) += via686a.o
obj-$(CONFIG_SENSORS_W83781D) += w83781d.o
^ permalink raw reply [flat|nested] 32+ messages in thread
* [PATCH] i2c driver changes for 2.5.72
@ 2005-05-19 6:24 ` Greg KH
0 siblings, 0 replies; 32+ messages in thread
From: Greg KH @ 2005-05-19 6:24 UTC (permalink / raw)
To: linux-kernel, sensors
ChangeSet 1.1318.3.4, 2003/06/16 11:32:06-07:00, azarah@gentoo.org
[PATCH] I2C: ICH5 SMBus and W83627THF additions
I have been trying to get the W83627THF chip working on this
board. It is an Asus P4C800 with Intel 875p chipset and a
W83627THF connected via the SMBus.
There are no data sheet for the W83627THF as far as I can see,
but supposidly it is a W83627HF with advance Fan control, etc.
I have applied attached patches, and tried various other things
to get it to work, but no avail. The SMBus on the ICH5 seems to
work, and it seems to detect the sensor chip just fine, but it
do not seem to read any values from the W83627THF.
drivers/i2c/busses/i2c-i801.c | 10 +++++++++-
drivers/i2c/chips/w83781d.c | 2 +-
2 files changed, 10 insertions(+), 2 deletions(-)
diff -Nru a/drivers/i2c/busses/i2c-i801.c b/drivers/i2c/busses/i2c-i801.c
--- a/drivers/i2c/busses/i2c-i801.c Wed Jun 18 11:19:35 2003
+++ b/drivers/i2c/busses/i2c-i801.c Wed Jun 18 11:19:35 2003
@@ -27,6 +27,7 @@
82801BA 2443
82801CA/CAM 2483
82801DB 24C3 (HW PEC supported, 32 byte buffer not supported)
+ 82801EB 24D3 (HW PEC supported, 32 byte buffer not supported)
This driver supports several versions of Intel's I/O Controller Hubs (ICH).
For SMBus support, they are similar to the PIIX4 and are part
@@ -121,7 +122,8 @@
return -ENODEV;
I801_dev = dev;
- if (dev->device = PCI_DEVICE_ID_INTEL_82801DB_3)
+ if ((dev->device = PCI_DEVICE_ID_INTEL_82801DB_3) ||
+ (dev->device = PCI_DEVICE_ID_INTEL_82801EB_3))
isich4 = 1;
else
isich4 = 0;
@@ -584,6 +586,12 @@
.device = PCI_DEVICE_ID_INTEL_82801DB_3,
.subvendor = PCI_ANY_ID,
.subdevice = PCI_ANY_ID,
+ },
+ {
+ .vendor = PCI_VENDOR_ID_INTEL,
+ .device = PCI_DEVICE_ID_INTEL_82801EB_3,
+ .subvendor = PCI_ANY_ID,
+ .subdevice = PCI_ANY_ID,
},
{ 0, }
};
diff -Nru a/drivers/i2c/chips/w83781d.c b/drivers/i2c/chips/w83781d.c
--- a/drivers/i2c/chips/w83781d.c Wed Jun 18 11:19:35 2003
+++ b/drivers/i2c/chips/w83781d.c Wed Jun 18 11:19:35 2003
@@ -1285,7 +1285,7 @@
kind = w83782d;
else if (val1 = 0x40 && vendid = winbond && !is_isa)
kind = w83783s;
- else if (val1 = 0x20 && vendid = winbond)
+ else if ((val1 = 0x20 || val1 = 0x72) && vendid = winbond)
kind = w83627hf;
else if (val1 = 0x30 && vendid = asus && !is_isa)
kind = as99127f;
^ permalink raw reply [flat|nested] 32+ messages in thread
* [PATCH] i2c driver changes for 2.5.72
@ 2005-05-19 6:24 ` Greg KH
0 siblings, 0 replies; 32+ messages in thread
From: Greg KH @ 2005-05-19 6:24 UTC (permalink / raw)
To: linux-kernel, sensors
ChangeSet 1.1318.3.6, 2003/06/16 11:40:37-07:00, margitsw@t-online.de
[PATCH] I2C: Sensors patch for adm1021
Patch for adm1021
This corrects temp reporting and a major error whereby
"alarms" and "die_code" were being put though the "TEMP" macro.
Compiled but don't have the hardware to test.
drivers/i2c/chips/adm1021.c | 18 ++++++++++++++----
1 files changed, 14 insertions(+), 4 deletions(-)
diff -Nru a/drivers/i2c/chips/adm1021.c b/drivers/i2c/chips/adm1021.c
--- a/drivers/i2c/chips/adm1021.c Wed Jun 18 11:19:25 2003
+++ b/drivers/i2c/chips/adm1021.c Wed Jun 18 11:19:25 2003
@@ -88,8 +88,8 @@
these macros are called: arguments may be evaluated more than once.
Fixing this is just not worth it. */
/* Conversions note: 1021 uses normal integer signed-byte format*/
-#define TEMP_FROM_REG(val) (val > 127 ? val-256 : val)
-#define TEMP_TO_REG(val) (SENSORS_LIMIT((val < 0 ? val+256 : val),0,255))
+#define TEMP_FROM_REG(val) (val > 127 ? (val-256)*1000 : val*1000)
+#define TEMP_TO_REG(val) (SENSORS_LIMIT((val < 0 ? (val/1000)+256 : val/1000),0,255))
/* Initial values */
@@ -172,8 +172,18 @@
show(remote_temp_max);
show(remote_temp_hyst);
show(remote_temp_input);
-show(alarms);
-show(die_code);
+
+#define show2(value) \
+static ssize_t show_##value(struct device *dev, char *buf) \
+{ \
+ struct i2c_client *client = to_i2c_client(dev); \
+ struct adm1021_data *data = i2c_get_clientdata(client); \
+ \
+ adm1021_update_client(client); \
+ return sprintf(buf, "%d\n", data->value); \
+}
+show2(alarms);
+show2(die_code);
#define set(value, reg) \
static ssize_t set_##value(struct device *dev, const char *buf, size_t count) \
^ permalink raw reply [flat|nested] 32+ messages in thread
* [PATCH] i2c driver changes for 2.5.72
@ 2005-05-19 6:24 ` Greg KH
0 siblings, 0 replies; 32+ messages in thread
From: Greg KH @ 2005-05-19 6:24 UTC (permalink / raw)
To: linux-kernel, sensors
ChangeSet 1.1318.3.7, 2003/06/17 15:34:14-07:00, azarah@gentoo.org
[PATCH] I2C: fix for previous W83627THF sensor chip patch
Ok, I was wrong in assuming that the W83627THF was on the I2C bus.
It is on the ISA bus, id 0x90 (thanks to Alex Van Kaam author of
MBM who corrected my assumption).
drivers/i2c/chips/w83781d.c | 8 ++++++--
1 files changed, 6 insertions(+), 2 deletions(-)
diff -Nru a/drivers/i2c/chips/w83781d.c b/drivers/i2c/chips/w83781d.c
--- a/drivers/i2c/chips/w83781d.c Wed Jun 18 11:19:21 2003
+++ b/drivers/i2c/chips/w83781d.c Wed Jun 18 11:19:21 2003
@@ -28,6 +28,7 @@
asb100 "bach" (type_name = as99127f) 0x30 0x0694 yes no
w83781d 7 3 0 3 0x10 0x5ca3 yes yes
w83627hf 9 3 2 3 0x20 0x5ca3 yes yes(LPC)
+ w83627thf 9 3 2 3 0x90 0x5ca3 no yes(LPC)
w83782d 9 3 2-4 3 0x30 0x5ca3 yes yes
w83783s 5-6 3 2 1-2 0x40 0x5ca3 yes no
w83697hf 8 2 2 2 0x60 0x5ca3 no yes(LPC)
@@ -1285,7 +1286,7 @@
kind = w83782d;
else if (val1 = 0x40 && vendid = winbond && !is_isa)
kind = w83783s;
- else if ((val1 = 0x20 || val1 = 0x72) && vendid = winbond)
+ else if ((val1 = 0x20 || val1 = 0x90) && vendid = winbond)
kind = w83627hf;
else if (val1 = 0x30 && vendid = asus && !is_isa)
kind = as99127f;
@@ -1309,7 +1310,10 @@
} else if (kind = w83783s) {
client_name = "W83783S chip";
} else if (kind = w83627hf) {
- client_name = "W83627HF chip";
+ if (val1 = 0x90)
+ client_name = "W83627THF chip";
+ else
+ client_name = "W83627HF chip";
} else if (kind = as99127f) {
client_name = "AS99127F chip";
} else if (kind = w83697hf) {
^ permalink raw reply [flat|nested] 32+ messages in thread
* [PATCH] i2c driver changes for 2.5.72
@ 2005-05-19 6:24 ` Greg KH
0 siblings, 0 replies; 32+ messages in thread
From: Greg KH @ 2005-05-19 6:24 UTC (permalink / raw)
To: linux-kernel, sensors
ChangeSet 1.1318.3.8, 2003/06/17 15:37:40-07:00, greg@kroah.com
I2C: fix resource leak in i2c-ali15x3.c
drivers/i2c/busses/i2c-ali15x3.c | 18 +++++++++++-------
1 files changed, 11 insertions(+), 7 deletions(-)
diff -Nru a/drivers/i2c/busses/i2c-ali15x3.c b/drivers/i2c/busses/i2c-ali15x3.c
--- a/drivers/i2c/busses/i2c-ali15x3.c Wed Jun 18 11:19:16 2003
+++ b/drivers/i2c/busses/i2c-ali15x3.c Wed Jun 18 11:19:16 2003
@@ -177,17 +177,18 @@
if(force_addr) {
dev_info(&ALI15X3_dev->dev, "forcing ISA address 0x%04X\n",
ali15x3_smba);
- if (PCIBIOS_SUCCESSFUL !- pci_write_config_word(ALI15X3_dev, SMBBA, ali15x3_smba))
- return -ENODEV;
- if (PCIBIOS_SUCCESSFUL !- pci_read_config_word(ALI15X3_dev, SMBBA, &a))
- return -ENODEV;
+ if (PCIBIOS_SUCCESSFUL != pci_write_config_word(ALI15X3_dev,
+ SMBBA,
+ ali15x3_smba))
+ goto error;
+ if (PCIBIOS_SUCCESSFUL != pci_read_config_word(ALI15X3_dev,
+ SMBBA, &a))
+ goto error;
if ((a & ~(ALI15X3_SMB_IOSIZE - 1)) != ali15x3_smba) {
/* make sure it works */
dev_err(&ALI15X3_dev->dev,
"force address failed - not supported?\n");
- return -ENODEV;
+ goto error;
}
}
/* check if whole device is enabled */
@@ -219,6 +220,9 @@
dev_dbg(&ALI15X3_dev->dev, "iALI15X3_smba = 0x%X\n", ali15x3_smba);
return 0;
+error:
+ release_region(ali15x3_smba, ALI15X3_SMB_IOSIZE);
+ return -ENODEV;
}
/* Internally used pause function */
^ permalink raw reply [flat|nested] 32+ messages in thread
* [PATCH] i2c driver changes for 2.5.72
2005-05-19 6:24 ` Greg KH
` (3 preceding siblings ...)
(?)
@ 2005-05-19 6:24 ` Mark M. Hoffman
-1 siblings, 0 replies; 32+ messages in thread
From: Mark M. Hoffman @ 2005-05-19 6:24 UTC (permalink / raw)
To: lm-sensors
* Greg KH <greg@kroah.com> [2003-06-18 11:25:07 -0700]:
> ChangeSet 1.1318.3.2, 2003/06/16 11:31:43-07:00, margitsw@t-online.de
>
> [PATCH] I2C: lm85 fixups
>
> OK Here's the patch which :
> 1) Fixes the race conditions
<cut>
> @@ -437,10 +434,13 @@
> {
> struct i2c_client *client = to_i2c_client(dev);
> struct lm85_data *data = i2c_get_clientdata(client);
> + int val;
>
> - int val = simple_strtol(buf, NULL, 10);
> + down(&data->update_lock);
> + val = simple_strtol(buf, NULL, 10);
> data->fan_min[nr] = FAN_TO_REG(val);
> lm85_write_value(client, LM85_REG_FAN_MIN(nr), data->fan_min[nr]);
> + up(&data->update_lock);
> return count;
> }
Ugh. Looks like this sort of fix is needed in every single chip
driver in 2.5. And the CVS chip drivers have the same problem.
So... are we going to tackle this for the sensors 2.8.0 release?
Regards,
--
Mark M. Hoffman
mhoffman@lightlink.com
^ permalink raw reply [flat|nested] 32+ messages in thread
* [PATCH] i2c driver changes for 2.5.72
2005-05-19 6:24 ` Greg KH
(?)
(?)
@ 2005-05-19 6:24 ` Greg KH
-1 siblings, 0 replies; 32+ messages in thread
From: Greg KH @ 2005-05-19 6:24 UTC (permalink / raw)
To: lm-sensors
On Wed, Jun 18, 2003 at 10:25:57PM -0400, Mark M. Hoffman wrote:
> * Greg KH <greg@kroah.com> [2003-06-18 11:25:07 -0700]:
> > ChangeSet 1.1318.3.2, 2003/06/16 11:31:43-07:00, margitsw@t-online.de
> >
> > [PATCH] I2C: lm85 fixups
> >
> > OK Here's the patch which :
> > 1) Fixes the race conditions
>
> <cut>
>
> > @@ -437,10 +434,13 @@
> > {
> > struct i2c_client *client = to_i2c_client(dev);
> > struct lm85_data *data = i2c_get_clientdata(client);
> > + int val;
> >
> > - int val = simple_strtol(buf, NULL, 10);
> > + down(&data->update_lock);
> > + val = simple_strtol(buf, NULL, 10);
> > data->fan_min[nr] = FAN_TO_REG(val);
> > lm85_write_value(client, LM85_REG_FAN_MIN(nr), data->fan_min[nr]);
> > + up(&data->update_lock);
> > return count;
> > }
>
> Ugh. Looks like this sort of fix is needed in every single chip
> driver in 2.5. And the CVS chip drivers have the same problem.
Hm, wait, what are we trying to protect with this lock? The data->
values? Or access to the hardware? Or something else?
thanks,
greg k-h
^ permalink raw reply [flat|nested] 32+ messages in thread
* [PATCH] i2c driver changes for 2.5.72
2005-05-19 6:24 ` Greg KH
` (2 preceding siblings ...)
(?)
@ 2005-05-19 6:24 ` Philip Pokorny
-1 siblings, 0 replies; 32+ messages in thread
From: Philip Pokorny @ 2005-05-19 6:24 UTC (permalink / raw)
To: lm-sensors
The data->foo and device register need to be in sync. So I suppose both...
At first I thought that if I always updated the chip first and then the
data->foo cached value, I'd be OK, but it seemed like there were too
many exceptions where values were broken apart, or recombined...
The primary thing I was trying to avoid was one CPU writing a value
while another CPU was reading a value (and therefore calling
update_client) and the result being that the data->foo cached value was
different from actual device register.
:v)
Greg KH wrote:
> On Wed, Jun 18, 2003 at 10:25:57PM -0400, Mark M. Hoffman wrote:
>
>>* Greg KH <greg@kroah.com> [2003-06-18 11:25:07 -0700]:
>>
>>>ChangeSet 1.1318.3.2, 2003/06/16 11:31:43-07:00, margitsw@t-online.de
>>>
>>>[PATCH] I2C: lm85 fixups
>>>
>>>OK Here's the patch which :
>>>1) Fixes the race conditions
>>
>><cut>
>>
>>>@@ -437,10 +434,13 @@
>>> {
>>> struct i2c_client *client = to_i2c_client(dev);
>>> struct lm85_data *data = i2c_get_clientdata(client);
>>>+ int val;
>>>
>>>- int val = simple_strtol(buf, NULL, 10);
>>>+ down(&data->update_lock);
>>>+ val = simple_strtol(buf, NULL, 10);
>>> data->fan_min[nr] = FAN_TO_REG(val);
>>> lm85_write_value(client, LM85_REG_FAN_MIN(nr), data->fan_min[nr]);
>>>+ up(&data->update_lock);
>>> return count;
>>> }
>>
>>Ugh. Looks like this sort of fix is needed in every single chip
>>driver in 2.5. And the CVS chip drivers have the same problem.
>
>
> Hm, wait, what are we trying to protect with this lock? The data->
> values? Or access to the hardware? Or something else?
>
> thanks,
>
> greg k-h
>
--
Philip Pokorny, Director of Engineering
Tel: 415-358-2635 Fax: 415-358-2646 Toll Free: 888-PENGUIN
PENGUIN COMPUTING, INC.
www.penguincomputing.com
^ permalink raw reply [flat|nested] 32+ messages in thread
* [PATCH] i2c driver changes for 2.5.72
2005-05-19 6:24 ` Greg KH
` (4 preceding siblings ...)
(?)
@ 2005-05-19 6:24 ` Mark M. Hoffman
-1 siblings, 0 replies; 32+ messages in thread
From: Mark M. Hoffman @ 2005-05-19 6:24 UTC (permalink / raw)
To: lm-sensors
* Philip Pokorny <ppokorny@penguincomputing.com> [2003-06-19 18:45:36 -0700]:
> The data->foo and device register need to be in sync. So I suppose both...
The device registers are protected by the I2C bus locking or ISA bus
locking in xxxx_write_value. The race (in the original) that I see
looked like this:
process a in show_bar() process b in set_foo()
data->foo = baz;
/* changes data->foo */
xxxx_update_client();
/* wrong, doesn't = baz anymore */
xxxx_write_value(..., data->foo);
But now that I look again, why not this?
{
struct i2c_client *client = to_i2c_client(dev);
u8 reg ;
int val = simple_strtol(buf, NULL, 10);
reg = FAN_TO_REG(val);
lm85_write_value(client, LM85_REG_FAN_MIN(nr), reg);
return count;
}
Now put the `data->fan_min[nr] = value;` part into xxxx_write_value()
(along with the proper locking) and be done with it; less code this
way and harder to accidentally break. Actually, the more I look at
it, the more I like this idea (he writes after deleting 50 lines of
his first attempt.) Comments?
> At first I thought that if I always updated the chip first and then the
> data->foo cached value, I'd be OK, but it seemed like there were too
> many exceptions where values were broken apart, or recombined...
>
> The primary thing I was trying to avoid was one CPU writing a value
> while another CPU was reading a value (and therefore calling
> update_client) and the result being that the data->foo cached value was
> different from actual device register.
>
> :v)
>
> Greg KH wrote:
> >On Wed, Jun 18, 2003 at 10:25:57PM -0400, Mark M. Hoffman wrote:
> >
> >>* Greg KH <greg@kroah.com> [2003-06-18 11:25:07 -0700]:
> >>
> >>>ChangeSet 1.1318.3.2, 2003/06/16 11:31:43-07:00, margitsw@t-online.de
> >>>
> >>>[PATCH] I2C: lm85 fixups
> >>>
> >>>OK Here's the patch which :
> >>>1) Fixes the race conditions
> >>
> >><cut>
> >>
> >>>@@ -437,10 +434,13 @@
> >>>{
> >>> struct i2c_client *client = to_i2c_client(dev);
> >>> struct lm85_data *data = i2c_get_clientdata(client);
> >>>+ int val;
> >>>
> >>>- int val = simple_strtol(buf, NULL, 10);
> >>>+ down(&data->update_lock);
> >>>+ val = simple_strtol(buf, NULL, 10);
> >>> data->fan_min[nr] = FAN_TO_REG(val);
> >>> lm85_write_value(client, LM85_REG_FAN_MIN(nr), data->fan_min[nr]);
> >>>+ up(&data->update_lock);
> >>> return count;
> >>>}
> >>
> >>Ugh. Looks like this sort of fix is needed in every single chip
> >>driver in 2.5. And the CVS chip drivers have the same problem.
> >
> >
> >Hm, wait, what are we trying to protect with this lock? The data->
> >values? Or access to the hardware? Or something else?
> >
--
Mark M. Hoffman
mhoffman@lightlink.com
^ permalink raw reply [flat|nested] 32+ messages in thread
* [BK PATCH] i2c driver changes for 2.5.72
@ 2005-05-19 6:24 ` Greg KH
0 siblings, 0 replies; 32+ messages in thread
From: Greg KH @ 2005-05-19 6:24 UTC (permalink / raw)
To: torvalds; +Cc: linux-kernel, sensors
Hi,
Here are some i2c driver changes for 2.5.72. These consist of some
more bug fixes, and the addition of a lm78 i2c driver.
Please pull from: bk://kernel.bkbits.net/gregkh/linux/i2c-2.5
thanks,
greg k-h
drivers/i2c/busses/i2c-ali15x3.c | 18
drivers/i2c/busses/i2c-i801.c | 10
drivers/i2c/chips/Kconfig | 14
drivers/i2c/chips/Makefile | 1
drivers/i2c/chips/adm1021.c | 18
drivers/i2c/chips/lm78.c | 895 +++++++++++++++++++++++++++++++++++++++
drivers/i2c/chips/lm85.c | 41 +
drivers/i2c/chips/w83781d.c | 86 ++-
8 files changed, 1019 insertions(+), 64 deletions(-)
-----
<margitsw:t-online.de>:
o I2C: Sensors patch for adm1021
o I2C: lm85 fixups
<mhoffman:lightlink.com>:
o I2C: w83781d bugfix
o i2c: Add lm78 sensor chip support
Greg Kroah-Hartman:
o I2C: fix resource leak in i2c-ali15x3.c
o I2C: add lm78 chip to Makefile
Martin Schlemmer:
o I2C: fix for previous W83627THF sensor chip patch
o I2C: ICH5 SMBus and W83627THF additions
^ permalink raw reply [flat|nested] 32+ messages in thread
* [PATCH] i2c driver changes for 2.5.72
2005-05-19 6:24 ` Greg KH
` (11 preceding siblings ...)
(?)
@ 2005-05-19 6:24 ` Mark M. Hoffman
-1 siblings, 0 replies; 32+ messages in thread
From: Mark M. Hoffman @ 2005-05-19 6:24 UTC (permalink / raw)
To: lm-sensors
* Philip Pokorny <ppokorny@penguincomputing.com> [2003-06-19 20:35:00 -0700]:
> > Now put the `data->fan_min[nr] = value;` part into xxxx_write_value()
> > (along with the proper locking) and be done with it; less code this
>
> Are you saying that the assignment 'data->fan_min[nr] = value' would
> happen inside the xxxx_write_value function?
>
> If so, how does xxxx_write_value know which data member to store into?
> Are you going to pass a pointer to data->fan_min[nr] as an argument to
> write_value? What about type and size issues. Some values are u8,
> others are s8, and u16.
Ack. I was just about to modify lm78.c as an example when I realized
the same thing just as your message came in.
> The other issue I see is that of partially split values. Perhaps I'm
> the only one that's done this, but some registers contain bits for
> several different sensors. (fan divisors for example...) I have been
> breaking those appart and putting them back together as necessary for
> the job at hand. In fact fan_divisors are a excellent example since
> they *have* to be split apart to make the fan RPM calculations easy.
Yeah, this is common to many chip drivers.
> What if we hoist the down/up out of update_client instead? Is there
> someplace higher up that we could serialize access to all the show/set
> functions? Is that a worse idea?
I don't know (to all three questions, more or less). I'm going to step
back a bit and look at what we're trying to accomplish with this structure
cache of the register values in the first place.
Right, the way these chips typically work... they stop sampling while they
service read or write requests. We want to allow updates no more often
than a defined frequency. So is there a better way to accomplish that?
> :v)
>
> Mark M. Hoffman wrote:
> >* Philip Pokorny <ppokorny@penguincomputing.com> [2003-06-19 18:45:36
> >-0700]:
> >
> >>The data->foo and device register need to be in sync. So I suppose
> >>both...
> >
> >
> >The device registers are protected by the I2C bus locking or ISA bus
> >locking in xxxx_write_value. The race (in the original) that I see
> >looked like this:
> >
> >process a in show_bar() process b in set_foo()
> >
> > data->foo = baz;
> >
> >/* changes data->foo */
> >xxxx_update_client();
> >
> > /* wrong, doesn't = baz anymore */
> > xxxx_write_value(..., data->foo);
> >
> >But now that I look again, why not this?
> >
> >{
> > struct i2c_client *client = to_i2c_client(dev);
> > u8 reg ;
> > int val = simple_strtol(buf, NULL, 10);
> >
> > reg = FAN_TO_REG(val);
> > lm85_write_value(client, LM85_REG_FAN_MIN(nr), reg);
> > return count;
> >}
> >
> >Now put the `data->fan_min[nr] = value;` part into xxxx_write_value()
> >(along with the proper locking) and be done with it; less code this
> >way and harder to accidentally break. Actually, the more I look at
> >it, the more I like this idea (he writes after deleting 50 lines of
> >his first attempt.) Comments?
> >
> >
> >>At first I thought that if I always updated the chip first and then the
> >>data->foo cached value, I'd be OK, but it seemed like there were too
> >>many exceptions where values were broken apart, or recombined...
> >>
> >>The primary thing I was trying to avoid was one CPU writing a value
> >>while another CPU was reading a value (and therefore calling
> >>update_client) and the result being that the data->foo cached value was
> >>different from actual device register.
> >>
> >>:v)
> >>
Regards,
--
Mark M. Hoffman
mhoffman@lightlink.com
^ permalink raw reply [flat|nested] 32+ messages in thread
* [PATCH] i2c driver changes for 2.5.72
2005-05-19 6:24 ` Greg KH
` (6 preceding siblings ...)
(?)
@ 2005-05-19 6:24 ` Mark M. Hoffman
-1 siblings, 0 replies; 32+ messages in thread
From: Mark M. Hoffman @ 2005-05-19 6:24 UTC (permalink / raw)
To: lm-sensors
* Mark M. Hoffman <mhoffman@lightlink.com> [2003-06-20 00:04:19 -0400]:
> * Philip Pokorny <ppokorny@penguincomputing.com> [2003-06-19 20:35:00 -0700]:
> > > Now put the `data->fan_min[nr] = value;` part into xxxx_write_value()
> > > (along with the proper locking) and be done with it; less code this
> >
> > Are you saying that the assignment 'data->fan_min[nr] = value' would
> > happen inside the xxxx_write_value function?
> >
> > If so, how does xxxx_write_value know which data member to store into?
> > Are you going to pass a pointer to data->fan_min[nr] as an argument to
> > write_value? What about type and size issues. Some values are u8,
> > others are s8, and u16.
>
> Ack. I was just about to modify lm78.c as an example when I realized
> the same thing just as your message came in.
>
> > The other issue I see is that of partially split values. Perhaps I'm
> > the only one that's done this, but some registers contain bits for
> > several different sensors. (fan divisors for example...) I have been
> > breaking those appart and putting them back together as necessary for
> > the job at hand. In fact fan_divisors are a excellent example since
> > they *have* to be split apart to make the fan RPM calculations easy.
>
> Yeah, this is common to many chip drivers.
>
> > What if we hoist the down/up out of update_client instead? Is there
> > someplace higher up that we could serialize access to all the show/set
> > functions? Is that a worse idea?
>
> I don't know (to all three questions, more or less). I'm going to step
> back a bit and look at what we're trying to accomplish with this structure
> cache of the register values in the first place.
>
> Right, the way these chips typically work... they stop sampling while they
> service read or write requests. We want to allow updates no more often
> than a defined frequency. So is there a better way to accomplish that?
Sorry to reply to myself - attempt #3:
Remove this line from all set_foo():
data->foo = FOO_TO_REG(...);
And add this to xxxx_write_value():
data->valid = 0;
Hmmm, probably still need data->update_lock in xxxx_write_value(), but
that should take care of it no?
Regards,
--
Mark M. Hoffman
mhoffman@lightlink.com
^ permalink raw reply [flat|nested] 32+ messages in thread
* [PATCH] i2c driver changes for 2.5.72
2005-05-19 6:24 ` Greg KH
` (5 preceding siblings ...)
(?)
@ 2005-05-19 6:24 ` Mark M. Hoffman
-1 siblings, 0 replies; 32+ messages in thread
From: Mark M. Hoffman @ 2005-05-19 6:24 UTC (permalink / raw)
To: lm-sensors
* Mark M. Hoffman <mhoffman@lightlink.com> [2003-06-20 00:14:07 -0400]:
>
> Remove this line from all set_foo():
>
> data->foo = FOO_TO_REG(...);
>
> And add this to xxxx_write_value():
>
> data->valid = 0;
>
> Hmmm, probably still need data->update_lock in xxxx_write_value(), but
> that should take care of it no?
*Sigh* no. There would still be races around the set_* functions that
read the fan divisor. OK Phil, I'm done trying to improve on your
patch. ;)
Regards,
--
Mark M. Hoffman
mhoffman@lightlink.com
^ permalink raw reply [flat|nested] 32+ messages in thread
* [PATCH] i2c driver changes for 2.5.72
2005-05-19 6:24 ` Greg KH
` (12 preceding siblings ...)
(?)
@ 2005-05-19 6:24 ` Greg KH
-1 siblings, 0 replies; 32+ messages in thread
From: Greg KH @ 2005-05-19 6:24 UTC (permalink / raw)
To: lm-sensors
On Thu, Jun 19, 2003 at 06:45:36PM -0700, Philip Pokorny wrote:
> The data->foo and device register need to be in sync. So I suppose both...
>
> At first I thought that if I always updated the chip first and then the
> data->foo cached value, I'd be OK, but it seemed like there were too
> many exceptions where values were broken apart, or recombined...
>
> The primary thing I was trying to avoid was one CPU writing a value
> while another CPU was reading a value (and therefore calling
> update_client) and the result being that the data->foo cached value was
> different from actual device register.
But if that happens, then the worse thing that could happen is we would
show "stale" data in the resulting value.
Hm, so I can understand if we need to protect access to the hardware
when we read two values in a row, like this:
res = i2c_smbus_read_byte_data(client, reg) & 0xff ;
res |= i2c_smbus_read_byte_data(client, reg+1) << 8 ;
or the same thing when writing.
But what about this simpler patch, that pushes the lock down to the
hardware reads and writes for multiple registers. And it doesn't try to
mirror the write in the "data" structure, as it's always read with the
latest values for every read.
Does it look ok? It has the added benefit of making the .c and .o files
smaller :)
thanks,
greg k-h
# I2C: clean up lm85's locking code.
diff -Nru a/drivers/i2c/chips/lm85.c b/drivers/i2c/chips/lm85.c
--- a/drivers/i2c/chips/lm85.c Mon Jun 23 16:39:23 2003
+++ b/drivers/i2c/chips/lm85.c Mon Jun 23 16:39:23 2003
@@ -433,14 +433,11 @@
size_t count, int nr)
{
struct i2c_client *client = to_i2c_client(dev);
- struct lm85_data *data = i2c_get_clientdata(client);
int val;
- down(&data->update_lock);
val = simple_strtol(buf, NULL, 10);
- data->fan_min[nr] = FAN_TO_REG(val);
- lm85_write_value(client, LM85_REG_FAN_MIN(nr), data->fan_min[nr]);
- up(&data->update_lock);
+ val = FAN_TO_REG(val);
+ lm85_write_value(client, LM85_REG_FAN_MIN(nr), val);
return count;
}
@@ -527,14 +524,11 @@
size_t count, int nr)
{
struct i2c_client *client = to_i2c_client(dev);
- struct lm85_data *data = i2c_get_clientdata(client);
int val;
- down(&data->update_lock);
val = simple_strtol(buf, NULL, 10);
- data->pwm[nr] = PWM_TO_REG(val);
- lm85_write_value(client, LM85_REG_PWM(nr), data->pwm[nr]);
- up(&data->update_lock);
+ val = PWM_TO_REG(val);
+ lm85_write_value(client, LM85_REG_PWM(nr), val);
return count;
}
static ssize_t show_pwm_enable(struct device *dev, char *buf, int nr)
@@ -592,14 +586,11 @@
size_t count, int nr)
{
struct i2c_client *client = to_i2c_client(dev);
- struct lm85_data *data = i2c_get_clientdata(client);
int val;
- down(&data->update_lock);
val = simple_strtol(buf, NULL, 10);
- data->in_min[nr] = INS_TO_REG(nr, val);
- lm85_write_value(client, LM85_REG_IN_MIN(nr), data->in_min[nr]);
- up(&data->update_lock);
+ val = INS_TO_REG(nr, val);
+ lm85_write_value(client, LM85_REG_IN_MIN(nr), val);
return count;
}
static ssize_t show_in_max(struct device *dev, char *buf, int nr)
@@ -614,14 +605,11 @@
size_t count, int nr)
{
struct i2c_client *client = to_i2c_client(dev);
- struct lm85_data *data = i2c_get_clientdata(client);
int val;
- down(&data->update_lock);
val = simple_strtol(buf, NULL, 10);
- data->in_max[nr] = INS_TO_REG(nr, val);
- lm85_write_value(client, LM85_REG_IN_MAX(nr), data->in_max[nr]);
- up(&data->update_lock);
+ val = INS_TO_REG(nr, val);
+ lm85_write_value(client, LM85_REG_IN_MAX(nr), val);
return count;
}
#define show_in_reg(offset) \
@@ -681,14 +669,11 @@
size_t count, int nr)
{
struct i2c_client *client = to_i2c_client(dev);
- struct lm85_data *data = i2c_get_clientdata(client);
int val;
- down(&data->update_lock);
val = simple_strtol(buf, NULL, 10);
- data->temp_min[nr] = TEMP_TO_REG(val);
- lm85_write_value(client, LM85_REG_TEMP_MIN(nr), data->temp_min[nr]);
- up(&data->update_lock);
+ val = TEMP_TO_REG(val);
+ lm85_write_value(client, LM85_REG_TEMP_MIN(nr), val);
return count;
}
static ssize_t show_temp_max(struct device *dev, char *buf, int nr)
@@ -703,14 +688,11 @@
size_t count, int nr)
{
struct i2c_client *client = to_i2c_client(dev);
- struct lm85_data *data = i2c_get_clientdata(client);
int val;
- down(&data->update_lock);
val = simple_strtol(buf, NULL, 10);
- data->temp_max[nr] = TEMP_TO_REG(val);
- lm85_write_value(client, LM85_REG_TEMP_MAX(nr), data->temp_max[nr]);
- up(&data->update_lock);
+ val = TEMP_TO_REG(val);
+ lm85_write_value(client, LM85_REG_TEMP_MAX(nr), val);
return count;
}
#define show_temp_reg(offset) \
@@ -955,8 +937,12 @@
int lm85_read_value(struct i2c_client *client, u8 reg)
{
+ struct lm85_data *data = i2c_get_clientdata(client);
int res;
+ /* serialize access to the hardware */
+ down(&data->update_lock);
+
/* What size location is it? */
switch( reg ) {
case LM85_REG_FAN(0) : /* Read WORD data */
@@ -980,14 +966,19 @@
res = i2c_smbus_read_byte_data(client, reg);
break ;
}
+ up(&data->update_lock);
return res ;
}
int lm85_write_value(struct i2c_client *client, u8 reg, int value)
{
+ struct lm85_data *data = i2c_get_clientdata(client);
int res ;
+ /* serialize access to the hardware */
+ down(&data->update_lock);
+
switch( reg ) {
case LM85_REG_FAN(0) : /* Write WORD data */
case LM85_REG_FAN(1) :
@@ -1009,6 +1000,7 @@
res = i2c_smbus_write_byte_data(client, reg, value);
break ;
}
+ up(&data->update_lock);
return res ;
}
@@ -1070,8 +1062,6 @@
struct lm85_data *data = i2c_get_clientdata(client);
int i;
- down(&data->update_lock);
-
if ( !data->valid ||
(jiffies - data->last_reading > LM85_DATA_INTERVAL ) ) {
/* Things that change quickly */
@@ -1209,8 +1199,6 @@
}; /* last_config */
data->valid = 1;
-
- up(&data->update_lock);
}
^ permalink raw reply [flat|nested] 32+ messages in thread
* [PATCH] i2c driver changes for 2.5.72
2005-05-19 6:24 ` Greg KH
` (9 preceding siblings ...)
(?)
@ 2005-05-19 6:24 ` Philip Pokorny
-1 siblings, 0 replies; 32+ messages in thread
From: Philip Pokorny @ 2005-05-19 6:24 UTC (permalink / raw)
To: lm-sensors
I don't think so...
In addition to the two-byte values, you've also got related values like
the extended precision bits for the A/D on the ADM1027. If two
processes both trigger an update_client() at the same time, bad things
would happen. I think the potential problem is much worse than just
"stale" data, but even "stale" data is bad.
Also, the reads are not always coming from the registers. (Or did I
miss a piece of your patch that changed update_client() and
LM85_DATA_TIMEOUT) update_client() only reads the registers after the
jiffies "timer" expires.
For the LM85, the _config_ data (like min/max values, etc.) is re-read
*much* less often than the actual readings. I expected that the cache
would be kept up to date on writes so that the re-read would be largely
unnecessary and only a backup. With the auto-fan control, there are
significant amounts of config data that would slow down access to the
readings we really care about.
Come to think of it, we can't really be sure that the cached values
won't be changed by another CPU running update_client while we're
accessing the cache. That would suggest that the down/up pair needs to
be pulled up out of update_client into the individual sysctl or set/show
functions.
Even if your version is smaller, I think it's likely to be slower or
broken in some cases...
:v)
Greg KH wrote:
> On Thu, Jun 19, 2003 at 06:45:36PM -0700, Philip Pokorny wrote:
>
>>The data->foo and device register need to be in sync. So I suppose both...
>>
>>At first I thought that if I always updated the chip first and then the
>>data->foo cached value, I'd be OK, but it seemed like there were too
>>many exceptions where values were broken apart, or recombined...
>>
>>The primary thing I was trying to avoid was one CPU writing a value
>>while another CPU was reading a value (and therefore calling
>>update_client) and the result being that the data->foo cached value was
>>different from actual device register.
>
>
> But if that happens, then the worse thing that could happen is we would
> show "stale" data in the resulting value.
>
> Hm, so I can understand if we need to protect access to the hardware
> when we read two values in a row, like this:
> res = i2c_smbus_read_byte_data(client, reg) & 0xff ;
> res |= i2c_smbus_read_byte_data(client, reg+1) << 8 ;
>
> or the same thing when writing.
>
> But what about this simpler patch, that pushes the lock down to the
> hardware reads and writes for multiple registers. And it doesn't try to
> mirror the write in the "data" structure, as it's always read with the
> latest values for every read.
>
> Does it look ok? It has the added benefit of making the .c and .o files
> smaller :)
>
> thanks,
>
> greg k-h
>
>
> # I2C: clean up lm85's locking code.
>
> diff -Nru a/drivers/i2c/chips/lm85.c b/drivers/i2c/chips/lm85.c
> --- a/drivers/i2c/chips/lm85.c Mon Jun 23 16:39:23 2003
> +++ b/drivers/i2c/chips/lm85.c Mon Jun 23 16:39:23 2003
> @@ -433,14 +433,11 @@
> size_t count, int nr)
> {
> struct i2c_client *client = to_i2c_client(dev);
> - struct lm85_data *data = i2c_get_clientdata(client);
> int val;
>
> - down(&data->update_lock);
> val = simple_strtol(buf, NULL, 10);
> - data->fan_min[nr] = FAN_TO_REG(val);
> - lm85_write_value(client, LM85_REG_FAN_MIN(nr), data->fan_min[nr]);
> - up(&data->update_lock);
> + val = FAN_TO_REG(val);
> + lm85_write_value(client, LM85_REG_FAN_MIN(nr), val);
> return count;
> }
>
> @@ -527,14 +524,11 @@
> size_t count, int nr)
> {
> struct i2c_client *client = to_i2c_client(dev);
> - struct lm85_data *data = i2c_get_clientdata(client);
> int val;
>
> - down(&data->update_lock);
> val = simple_strtol(buf, NULL, 10);
> - data->pwm[nr] = PWM_TO_REG(val);
> - lm85_write_value(client, LM85_REG_PWM(nr), data->pwm[nr]);
> - up(&data->update_lock);
> + val = PWM_TO_REG(val);
> + lm85_write_value(client, LM85_REG_PWM(nr), val);
> return count;
> }
> [ snip ] \
> @@ -955,8 +937,12 @@
>
> int lm85_read_value(struct i2c_client *client, u8 reg)
> {
> + struct lm85_data *data = i2c_get_clientdata(client);
> int res;
>
> + /* serialize access to the hardware */
> + down(&data->update_lock);
> +
> /* What size location is it? */
> switch( reg ) {
> case LM85_REG_FAN(0) : /* Read WORD data */
> @@ -980,14 +966,19 @@
> res = i2c_smbus_read_byte_data(client, reg);
> break ;
> }
> + up(&data->update_lock);
>
> return res ;
> }
>
> int lm85_write_value(struct i2c_client *client, u8 reg, int value)
> {
> + struct lm85_data *data = i2c_get_clientdata(client);
> int res ;
>
> + /* serialize access to the hardware */
> + down(&data->update_lock);
> +
> switch( reg ) {
> case LM85_REG_FAN(0) : /* Write WORD data */
> case LM85_REG_FAN(1) :
> @@ -1009,6 +1000,7 @@
> res = i2c_smbus_write_byte_data(client, reg, value);
> break ;
> }
> + up(&data->update_lock);
>
> return res ;
> }
> @@ -1070,8 +1062,6 @@
> struct lm85_data *data = i2c_get_clientdata(client);
> int i;
>
> - down(&data->update_lock);
> -
> if ( !data->valid ||
> (jiffies - data->last_reading > LM85_DATA_INTERVAL ) ) {
> /* Things that change quickly */
> @@ -1209,8 +1199,6 @@
> }; /* last_config */
>
> data->valid = 1;
> -
> - up(&data->update_lock);
> }
>
>
--
Philip Pokorny, Director of Engineering
Tel: 415-358-2635 Fax: 415-358-2646 Toll Free: 888-PENGUIN
PENGUIN COMPUTING, INC.
www.penguincomputing.com
^ permalink raw reply [flat|nested] 32+ messages in thread
* [PATCH] i2c driver changes for 2.5.72
2005-05-19 6:24 ` Greg KH
` (13 preceding siblings ...)
(?)
@ 2005-05-19 6:24 ` Mark M. Hoffman
-1 siblings, 0 replies; 32+ messages in thread
From: Mark M. Hoffman @ 2005-05-19 6:24 UTC (permalink / raw)
To: lm-sensors
* Greg KH <greg@kroah.com> [2003-06-23 16:41:14 -0700]:
> On Thu, Jun 19, 2003 at 06:45:36PM -0700, Philip Pokorny wrote:
> > The data->foo and device register need to be in sync. So I suppose both...
> >
> > At first I thought that if I always updated the chip first and then the
> > data->foo cached value, I'd be OK, but it seemed like there were too
> > many exceptions where values were broken apart, or recombined...
> >
> > The primary thing I was trying to avoid was one CPU writing a value
> > while another CPU was reading a value (and therefore calling
> > update_client) and the result being that the data->foo cached value was
> > different from actual device register.
>
> But if that happens, then the worse thing that could happen is we would
> show "stale" data in the resulting value.
Nope - as it was originally written (before Philip's most recent patch),
we could lose the value we meant to write.
> Hm, so I can understand if we need to protect access to the hardware
> when we read two values in a row, like this:
> res = i2c_smbus_read_byte_data(client, reg) & 0xff ;
> res |= i2c_smbus_read_byte_data(client, reg+1) << 8 ;
>
> or the same thing when writing.
Good point. The I2C core provides I2C/SMBus locking to prevent trashing
the bus itself, but that doesn't help the above. Which brings about
another question... why wasn't i2c_smbus_read_word_data() used in that
case? Philip?
> But what about this simpler patch, that pushes the lock down to the
> hardware reads and writes for multiple registers. And it doesn't try to
> mirror the write in the "data" structure, as it's always read with the
> latest values for every read.
It looks like this scheme would actually work for lm85.c But, for most
chip drivers with fan divisors there would still be a race. E.g. look at
set_fan_min() and set_fan_div() in lm78.c; the meaning of data->fan_min
depends on data->fan_div so those two must be coherent. That locking can't
be pushed further down.
That lm85 doesn't use fan divisors is the exception rather than the rule.
I mean, I don't hate this patch... but it doesn't tell us anything about
how to fix the rest of them. I still think Philip's original brute force
locking ;) is necessary there.
> Does it look ok? It has the added benefit of making the .c and .o files
> smaller :)
> int lm85_write_value(struct i2c_client *client, u8 reg, int value)
> {
> + struct lm85_data *data = i2c_get_clientdata(client);
> int res ;
>
> + /* serialize access to the hardware */
> + down(&data->update_lock);
> +
> switch( reg ) {
> case LM85_REG_FAN(0) : /* Write WORD data */
> case LM85_REG_FAN(1) :
> @@ -1009,6 +1000,7 @@
> res = i2c_smbus_write_byte_data(client, reg, value);
> break ;
> }
> + up(&data->update_lock);
>
> return res ;
> }
I would add "data->valid = 0;" somewhere before the up(). Then you're
guaranteed never to read stale data after a write.
> @@ -1070,8 +1062,6 @@
> struct lm85_data *data = i2c_get_clientdata(client);
> int i;
>
> - down(&data->update_lock);
> -
> if ( !data->valid ||
> (jiffies - data->last_reading > LM85_DATA_INTERVAL ) ) {
> /* Things that change quickly */
> @@ -1209,8 +1199,6 @@
> }; /* last_config */
>
> data->valid = 1;
> -
> - up(&data->update_lock);
> }
We only want to service one update_client() call per some interval.
I think this lock needs to stay.
And one last off-topic nit: data->fan_ppr is set but not used?
Regards,
--
Mark M. Hoffman
mhoffman@lightlink.com
^ permalink raw reply [flat|nested] 32+ messages in thread
* [PATCH] i2c driver changes for 2.5.72
2005-05-19 6:24 ` Greg KH
` (10 preceding siblings ...)
(?)
@ 2005-05-19 6:24 ` Mark M. Hoffman
-1 siblings, 0 replies; 32+ messages in thread
From: Mark M. Hoffman @ 2005-05-19 6:24 UTC (permalink / raw)
To: lm-sensors
* Mark M. Hoffman <mhoffman@lightlink.com> [2003-06-23 23:32:39 -0400]:
(replying to himself again...)
>
> > @@ -1070,8 +1062,6 @@
> > struct lm85_data *data = i2c_get_clientdata(client);
> > int i;
> >
> > - down(&data->update_lock);
> > -
> > if ( !data->valid ||
> > (jiffies - data->last_reading > LM85_DATA_INTERVAL ) ) {
> > /* Things that change quickly */
> > @@ -1209,8 +1199,6 @@
> > }; /* last_config */
> >
> > data->valid = 1;
> > -
> > - up(&data->update_lock);
> > }
>
> We only want to service one update_client() call per some interval.
> I think this lock needs to stay.
If this one stays then the one you added in lm85_read_value() must go.
Anyway, the latter causes the update_lock to be used before it's
initialized - lines 779-780 in lm85_detect().
Regards,
--
Mark M. Hoffman
mhoffman@lightlink.com
^ permalink raw reply [flat|nested] 32+ messages in thread
* [PATCH] i2c driver changes for 2.5.72
2005-05-19 6:24 ` Greg KH
` (7 preceding siblings ...)
(?)
@ 2005-05-19 6:24 ` Philip Pokorny
-1 siblings, 0 replies; 32+ messages in thread
From: Philip Pokorny @ 2005-05-19 6:24 UTC (permalink / raw)
To: lm-sensors
A few comments below...
:v)
Mark M. Hoffman wrote:
> * Greg KH <greg@kroah.com> [2003-06-23 16:41:14 -0700]:
>
>>On Thu, Jun 19, 2003 at 06:45:36PM -0700, Philip Pokorny wrote:
>>
>>>The data->foo and device register need to be in sync. So I suppose both...
>>>
>>>At first I thought that if I always updated the chip first and then the
>>>data->foo cached value, I'd be OK, but it seemed like there were too
>>>many exceptions where values were broken apart, or recombined...
>>>
>>>The primary thing I was trying to avoid was one CPU writing a value
>>>while another CPU was reading a value (and therefore calling
>>>update_client) and the result being that the data->foo cached value was
>>>different from actual device register.
>>
>>But if that happens, then the worse thing that could happen is we would
>>show "stale" data in the resulting value.
>
>
> Nope - as it was originally written (before Philip's most recent patch),
> we could lose the value we meant to write.
>
>
>>Hm, so I can understand if we need to protect access to the hardware
>>when we read two values in a row, like this:
>> res = i2c_smbus_read_byte_data(client, reg) & 0xff ;
>> res |= i2c_smbus_read_byte_data(client, reg+1) << 8 ;
>>
>>or the same thing when writing.
>
>
> Good point. The I2C core provides I2C/SMBus locking to prevent trashing
> the bus itself, but that doesn't help the above. Which brings about
> another question... why wasn't i2c_smbus_read_word_data() used in that
> case? Philip?
I actually tried that, but it didn't work. My original code had a
#define that would let you try and re-enable it, but I ditched it when I
didn't see any other drivers using read_word variants.
I should have spoken up that read_word seemed to be broken at the time.
But as I write this, I think I read in the chip data sheet that it
doesn't support SMBus word reads... (Go figure...) That was probably
the real reason I pulled the read_word code...
>>But what about this simpler patch, that pushes the lock down to the
>>hardware reads and writes for multiple registers. And it doesn't try to
>>mirror the write in the "data" structure, as it's always read with the
>>latest values for every read.
>
>
> It looks like this scheme would actually work for lm85.c But, for most
> chip drivers with fan divisors there would still be a race. E.g. look at
> set_fan_min() and set_fan_div() in lm78.c; the meaning of data->fan_min
> depends on data->fan_div so those two must be coherent. That locking can't
> be pushed further down.
>
> That lm85 doesn't use fan divisors is the exception rather than the rule.
> I mean, I don't hate this patch... but it doesn't tell us anything about
> how to fix the rest of them. I still think Philip's original brute force
> locking ;) is necessary there.
>
>
>>Does it look ok? It has the added benefit of making the .c and .o files
>>smaller :)
>
>
>> int lm85_write_value(struct i2c_client *client, u8 reg, int value)
>> {
>>+ struct lm85_data *data = i2c_get_clientdata(client);
>> int res ;
>>
>>+ /* serialize access to the hardware */
>>+ down(&data->update_lock);
>>+
>> switch( reg ) {
>> case LM85_REG_FAN(0) : /* Write WORD data */
>> case LM85_REG_FAN(1) :
>>@@ -1009,6 +1000,7 @@
>> res = i2c_smbus_write_byte_data(client, reg, value);
>> break ;
>> }
>>+ up(&data->update_lock);
>>
>> return res ;
>> }
>
>
> I would add "data->valid = 0;" somewhere before the up(). Then you're
> guaranteed never to read stale data after a write.
>
>
>>@@ -1070,8 +1062,6 @@
>> struct lm85_data *data = i2c_get_clientdata(client);
>> int i;
>>
>>- down(&data->update_lock);
>>-
>> if ( !data->valid ||
>> (jiffies - data->last_reading > LM85_DATA_INTERVAL ) ) {
>> /* Things that change quickly */
>>@@ -1209,8 +1199,6 @@
>> }; /* last_config */
>>
>> data->valid = 1;
>>-
>>- up(&data->update_lock);
>> }
>
>
> We only want to service one update_client() call per some interval.
> I think this lock needs to stay.
>
> And one last off-topic nit: data->fan_ppr is set but not used?
The fan_ppr features is only on the ADM1027 variant. Perhaps Margit
didn't migrate that code? Otherwise, there are calls to PPR_FROM_REG()
in the cvs code...
:v)
--
Philip Pokorny, Director of Engineering
Tel: 415-358-2635 Fax: 415-358-2646 Toll Free: 888-PENGUIN
PENGUIN COMPUTING, INC.
www.penguincomputing.com
^ permalink raw reply [flat|nested] 32+ messages in thread
* [PATCH] i2c driver changes for 2.5.72
2005-05-19 6:24 ` Greg KH
` (8 preceding siblings ...)
(?)
@ 2005-05-19 6:24 ` Mark M. Hoffman
-1 siblings, 0 replies; 32+ messages in thread
From: Mark M. Hoffman @ 2005-05-19 6:24 UTC (permalink / raw)
To: lm-sensors
* Philip Pokorny <ppokorny@penguincomputing.com> [2003-06-23 20:57:36 -0700]:
> Mark M. Hoffman wrote:
> >* Greg KH <greg@kroah.com> [2003-06-23 16:41:14 -0700]:
> >>Hm, so I can understand if we need to protect access to the hardware
> >>when we read two values in a row, like this:
> >> res = i2c_smbus_read_byte_data(client, reg) & 0xff ;
> >> res |= i2c_smbus_read_byte_data(client, reg+1) << 8 ;
> >>
> >>or the same thing when writing.
> >
> >
> >Good point. The I2C core provides I2C/SMBus locking to prevent trashing
> >the bus itself, but that doesn't help the above. Which brings about
> >another question... why wasn't i2c_smbus_read_word_data() used in that
> >case? Philip?
>
> I actually tried that, but it didn't work. My original code had a
> #define that would let you try and re-enable it, but I ditched it when I
> didn't see any other drivers using read_word variants.
>
> I should have spoken up that read_word seemed to be broken at the time.
>
> But as I write this, I think I read in the chip data sheet that it
> doesn't support SMBus word reads... (Go figure...) That was probably
> the real reason I pulled the read_word code...
It's just as well... maybe some bus drivers don't support it either. It
doesn't hurt to use lowest common denominator SMBus routines if possible.
> >And one last off-topic nit: data->fan_ppr is set but not used?
>
> The fan_ppr features is only on the ADM1027 variant. Perhaps Margit
> didn't migrate that code? Otherwise, there are calls to PPR_FROM_REG()
> in the cvs code...
Maybe it was left out because there's no mention of what to do with it
in sysfs? I have no suggestion for that; I don't know what PPR is.
Also btw, data->lock is never used. There could be others.
Regards,
--
Mark M. Hoffman
mhoffman@lightlink.com
^ permalink raw reply [flat|nested] 32+ messages in thread
* [PATCH] i2c driver changes for 2.5.72
2005-05-19 6:24 ` Greg KH
` (14 preceding siblings ...)
(?)
@ 2005-05-19 6:24 ` Mark D. Studebaker
-1 siblings, 0 replies; 32+ messages in thread
From: Mark D. Studebaker @ 2005-05-19 6:24 UTC (permalink / raw)
To: lm-sensors
I'll chime in here to agree.
You can't generally think of read_word_data as getting one byte and the next byte.
If a chip supports it at all it may return a "word" (2 bytes)
from a single address (see Winbond chips vendor ID at address 0x4f).
eeproms do it right by doing "autoincrement" to the next byte.
You can use 'i2cdump z 0xzz w' to experiment with
word reads and see how a chip behaves....
Mark M. Hoffman wrote:
> * Philip Pokorny <ppokorny@penguincomputing.com> [2003-06-23 20:57:36 -0700]:
>
>>Mark M. Hoffman wrote:
>>
>>>* Greg KH <greg@kroah.com> [2003-06-23 16:41:14 -0700]:
>
>
>>>>Hm, so I can understand if we need to protect access to the hardware
>>>>when we read two values in a row, like this:
>>>> res = i2c_smbus_read_byte_data(client, reg) & 0xff ;
>>>> res |= i2c_smbus_read_byte_data(client, reg+1) << 8 ;
>>>>
>>>>or the same thing when writing.
>>>
>>>
>>>Good point. The I2C core provides I2C/SMBus locking to prevent trashing
>>>the bus itself, but that doesn't help the above. Which brings about
>>>another question... why wasn't i2c_smbus_read_word_data() used in that
>>>case? Philip?
>>
>>I actually tried that, but it didn't work. My original code had a
>>#define that would let you try and re-enable it, but I ditched it when I
>>didn't see any other drivers using read_word variants.
>>
>>I should have spoken up that read_word seemed to be broken at the time.
>>
>>But as I write this, I think I read in the chip data sheet that it
>>doesn't support SMBus word reads... (Go figure...) That was probably
>>the real reason I pulled the read_word code...
>
>
> It's just as well... maybe some bus drivers don't support it either. It
> doesn't hurt to use lowest common denominator SMBus routines if possible.
>
>
^ permalink raw reply [flat|nested] 32+ messages in thread
end of thread, other threads:[~2005-05-19 6:24 UTC | newest]
Thread overview: 32+ messages (download: mbox.gz follow: Atom feed
-- links below jump to the message on this page --
2003-06-18 18:24 [BK PATCH] i2c driver changes for 2.5.72 Greg KH
2005-05-19 6:24 ` Greg KH
2003-06-18 18:25 ` [PATCH] " Greg KH
2005-05-19 6:24 ` Greg KH
2003-06-18 18:25 ` Greg KH
2005-05-19 6:24 ` Greg KH
2003-06-18 18:25 ` Greg KH
2005-05-19 6:24 ` Greg KH
2003-06-18 18:25 ` Greg KH
2005-05-19 6:24 ` Greg KH
2003-06-18 18:25 ` Greg KH
2005-05-19 6:24 ` Greg KH
2003-06-18 18:25 ` Greg KH
2005-05-19 6:24 ` Greg KH
2003-06-18 18:25 ` Greg KH
2005-05-19 6:24 ` Greg KH
2003-06-18 18:25 ` Greg KH
2005-05-19 6:24 ` Greg KH
2005-05-19 6:24 ` Greg KH
2005-05-19 6:24 ` Philip Pokorny
2005-05-19 6:24 ` Mark M. Hoffman
2005-05-19 6:24 ` Mark M. Hoffman
2005-05-19 6:24 ` Mark M. Hoffman
2005-05-19 6:24 ` Mark M. Hoffman
2005-05-19 6:24 ` Philip Pokorny
2005-05-19 6:24 ` Mark M. Hoffman
2005-05-19 6:24 ` Philip Pokorny
2005-05-19 6:24 ` Mark M. Hoffman
2005-05-19 6:24 ` Mark M. Hoffman
2005-05-19 6:24 ` Greg KH
2005-05-19 6:24 ` Mark M. Hoffman
2005-05-19 6:24 ` Mark D. Studebaker
This is an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.