aboutsummaryrefslogtreecommitdiffstats
path: root/Documentation
diff options
context:
space:
mode:
authorR.Marek@sh.cvut.cz <R.Marek@sh.cvut.cz>2005-05-26 08:42:19 -0400
committerGreg Kroah-Hartman <gregkh@suse.de>2005-06-22 00:52:02 -0400
commit7f15b66468b7003d5241e352a007e73be5519b20 (patch)
treec9333e14baae06a831c8515d9e1e24808826053e /Documentation
parent2bf34a1ca9d570dd4fab4d95c4de82d873ecf718 (diff)
[PATCH] I2C: documentation update 2/3
This patch adds missing documentation for system health monitoring chips. I would like to thank all people, who helped me with this project. Signed-off-by: Rudolf Marek <r.marek@sh.cvut.cz> Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
Diffstat (limited to 'Documentation')
-rw-r--r--Documentation/i2c/chips/adm1021111
-rw-r--r--Documentation/i2c/chips/adm102551
-rw-r--r--Documentation/i2c/chips/adm102693
-rw-r--r--Documentation/i2c/chips/adm103135
-rw-r--r--Documentation/i2c/chips/asb10072
-rw-r--r--Documentation/i2c/chips/ds1621108
-rw-r--r--Documentation/i2c/chips/eeprom96
-rw-r--r--Documentation/i2c/chips/fscher169
-rw-r--r--Documentation/i2c/chips/gl518sm74
-rw-r--r--Documentation/i2c/chips/it8796
-rw-r--r--Documentation/i2c/chips/lm6357
-rw-r--r--Documentation/i2c/chips/lm7565
-rw-r--r--Documentation/i2c/chips/lm7722
-rw-r--r--Documentation/i2c/chips/lm7882
-rw-r--r--Documentation/i2c/chips/lm8056
-rw-r--r--Documentation/i2c/chips/lm8376
-rw-r--r--Documentation/i2c/chips/lm85221
-rw-r--r--Documentation/i2c/chips/lm8773
-rw-r--r--Documentation/i2c/chips/lm90121
-rw-r--r--Documentation/i2c/chips/lm9237
-rw-r--r--Documentation/i2c/chips/max161929
-rw-r--r--Documentation/i2c/chips/pc87360189
-rw-r--r--Documentation/i2c/chips/pcf857469
-rw-r--r--Documentation/i2c/chips/pcf859190
-rw-r--r--Documentation/i2c/chips/sis5595106
-rw-r--r--Documentation/i2c/chips/smsc47b39722
-rw-r--r--Documentation/i2c/chips/smsc47m152
-rw-r--r--Documentation/i2c/chips/via686a65
-rw-r--r--Documentation/i2c/chips/w83627hf66
-rw-r--r--Documentation/i2c/chips/w83781d412
-rw-r--r--Documentation/i2c/chips/w83l785ts39
31 files changed, 2849 insertions, 5 deletions
diff --git a/Documentation/i2c/chips/adm1021 b/Documentation/i2c/chips/adm1021
new file mode 100644
index 000000000000..03d02bfb3df1
--- /dev/null
+++ b/Documentation/i2c/chips/adm1021
@@ -0,0 +1,111 @@
1Kernel driver adm1021
2=====================
3
4Supported chips:
5 * Analog Devices ADM1021
6 Prefix: 'adm1021'
7 Addresses scanned: I2C 0x18 - 0x1a, 0x29 - 0x2b, 0x4c - 0x4e
8 Datasheet: Publicly available at the Analog Devices website
9 * Analog Devices ADM1021A/ADM1023
10 Prefix: 'adm1023'
11 Addresses scanned: I2C 0x18 - 0x1a, 0x29 - 0x2b, 0x4c - 0x4e
12 Datasheet: Publicly available at the Analog Devices website
13 * Genesys Logic GL523SM
14 Prefix: 'gl523sm'
15 Addresses scanned: I2C 0x18 - 0x1a, 0x29 - 0x2b, 0x4c - 0x4e
16 Datasheet:
17 * Intel Xeon Processor
18 Prefix: - any other - may require 'force_adm1021' parameter
19 Addresses scanned: none
20 Datasheet: Publicly available at Intel website
21 * Maxim MAX1617
22 Prefix: 'max1617'
23 Addresses scanned: I2C 0x18 - 0x1a, 0x29 - 0x2b, 0x4c - 0x4e
24 Datasheet: Publicly available at the Maxim website
25 * Maxim MAX1617A
26 Prefix: 'max1617a'
27 Addresses scanned: I2C 0x18 - 0x1a, 0x29 - 0x2b, 0x4c - 0x4e
28 Datasheet: Publicly available at the Maxim website
29 * National Semiconductor LM84
30 Prefix: 'lm84'
31 Addresses scanned: I2C 0x18 - 0x1a, 0x29 - 0x2b, 0x4c - 0x4e
32 Datasheet: Publicly available at the National Semiconductor website
33 * Philips NE1617
34 Prefix: 'max1617' (probably detected as a max1617)
35 Addresses scanned: I2C 0x18 - 0x1a, 0x29 - 0x2b, 0x4c - 0x4e
36 Datasheet: Publicly available at the Philips website
37 * Philips NE1617A
38 Prefix: 'max1617' (probably detected as a max1617)
39 Addresses scanned: I2C 0x18 - 0x1a, 0x29 - 0x2b, 0x4c - 0x4e
40 Datasheet: Publicly available at the Philips website
41 * TI THMC10
42 Prefix: 'thmc10'
43 Addresses scanned: I2C 0x18 - 0x1a, 0x29 - 0x2b, 0x4c - 0x4e
44 Datasheet: Publicly available at the TI website
45 * Onsemi MC1066
46 Prefix: 'mc1066'
47 Addresses scanned: I2C 0x18 - 0x1a, 0x29 - 0x2b, 0x4c - 0x4e
48 Datasheet: Publicly available at the Onsemi website
49
50
51Authors:
52 Frodo Looijaard <frodol@dds.nl>,
53 Philip Edelbrock <phil@netroedge.com>
54
55Module Parameters
56-----------------
57
58* read_only: int
59 Don't set any values, read only mode
60
61
62Description
63-----------
64
65The chips supported by this driver are very similar. The Maxim MAX1617 is
66the oldest; it has the problem that it is not very well detectable. The
67MAX1617A solves that. The ADM1021 is a straight clone of the MAX1617A.
68Ditto for the THMC10. From here on, we will refer to all these chips as
69ADM1021-clones.
70
71The ADM1021 and MAX1617A reports a die code, which is a sort of revision
72code. This can help us pinpoint problems; it is not very useful
73otherwise.
74
75ADM1021-clones implement two temperature sensors. One of them is internal,
76and measures the temperature of the chip itself; the other is external and
77is realised in the form of a transistor-like device. A special alarm
78indicates whether the remote sensor is connected.
79
80Each sensor has its own low and high limits. When they are crossed, the
81corresponding alarm is set and remains on as long as the temperature stays
82out of range. Temperatures are measured in degrees Celsius. Measurements
83are possible between -65 and +127 degrees, with a resolution of one degree.
84
85If an alarm triggers, it will remain triggered until the hardware register
86is read at least once. This means that the cause for the alarm may already
87have disappeared!
88
89This driver only updates its values each 1.5 seconds; reading it more often
90will do no harm, but will return 'old' values. It is possible to make
91ADM1021-clones do faster measurements, but there is really no good reason
92for that.
93
94Xeon support
95------------
96
97Some Xeon processors have real max1617, adm1021, or compatible chips
98within them, with two temperature sensors.
99
100Other Xeons have chips with only one sensor.
101
102If you have a Xeon, and the adm1021 module loads, and both temperatures
103appear valid, then things are good.
104
105If the adm1021 module doesn't load, you should try this:
106 modprobe adm1021 force_adm1021=BUS,ADDRESS
107 ADDRESS can only be 0x18, 0x1a, 0x29, 0x2b, 0x4c, or 0x4e.
108
109If you have dual Xeons you may have appear to have two separate
110adm1021-compatible chips, or two single-temperature sensors, at distinct
111addresses.
diff --git a/Documentation/i2c/chips/adm1025 b/Documentation/i2c/chips/adm1025
new file mode 100644
index 000000000000..39d2b781b5d6
--- /dev/null
+++ b/Documentation/i2c/chips/adm1025
@@ -0,0 +1,51 @@
1Kernel driver adm1025
2=====================
3
4Supported chips:
5 * Analog Devices ADM1025, ADM1025A
6 Prefix: 'adm1025'
7 Addresses scanned: I2C 0x2c - 0x2e
8 Datasheet: Publicly available at the Analog Devices website
9 * Philips NE1619
10 Prefix: 'ne1619'
11 Addresses scanned: I2C 0x2c - 0x2d
12 Datasheet: Publicly available at the Philips website
13
14The NE1619 presents some differences with the original ADM1025:
15 * Only two possible addresses (0x2c - 0x2d).
16 * No temperature offset register, but we don't use it anyway.
17 * No INT mode for pin 16. We don't play with it anyway.
18
19Authors:
20 Chen-Yuan Wu <gwu@esoft.com>,
21 Jean Delvare <khali@linux-fr.org>
22
23Description
24-----------
25
26(This is from Analog Devices.) The ADM1025 is a complete system hardware
27monitor for microprocessor-based systems, providing measurement and limit
28comparison of various system parameters. Five voltage measurement inputs
29are provided, for monitoring +2.5V, +3.3V, +5V and +12V power supplies and
30the processor core voltage. The ADM1025 can monitor a sixth power-supply
31voltage by measuring its own VCC. One input (two pins) is dedicated to a
32remote temperature-sensing diode and an on-chip temperature sensor allows
33ambient temperature to be monitored.
34
35One specificity of this chip is that the pin 11 can be hardwired in two
36different manners. It can act as the +12V power-supply voltage analog
37input, or as the a fifth digital entry for the VID reading (bit 4). It's
38kind of strange since both are useful, and the reason for designing the
39chip that way is obscure at least to me. The bit 5 of the configuration
40register can be used to define how the chip is hardwired. Please note that
41it is not a choice you have to make as the user. The choice was already
42made by your motherboard's maker. If the configuration bit isn't set
43properly, you'll have a wrong +12V reading or a wrong VID reading. The way
44the driver handles that is to preserve this bit through the initialization
45process, assuming that the BIOS set it up properly beforehand. If it turns
46out not to be true in some cases, we'll provide a module parameter to force
47modes.
48
49This driver also supports the ADM1025A, which differs from the ADM1025
50only in that it has "open-drain VID inputs while the ADM1025 has on-chip
51100k pull-ups on the VID inputs". It doesn't make any difference for us.
diff --git a/Documentation/i2c/chips/adm1026 b/Documentation/i2c/chips/adm1026
new file mode 100644
index 000000000000..473c689d7924
--- /dev/null
+++ b/Documentation/i2c/chips/adm1026
@@ -0,0 +1,93 @@
1Kernel driver adm1026
2=====================
3
4Supported chips:
5 * Analog Devices ADM1026
6 Prefix: 'adm1026'
7 Addresses scanned: I2C 0x2c, 0x2d, 0x2e
8 Datasheet: Publicly available at the Analog Devices website
9 http://www.analog.com/en/prod/0,,766_825_ADM1026,00.html
10
11Authors:
12 Philip Pokorny <ppokorny@penguincomputing.com> for Penguin Computing
13 Justin Thiessen <jthiessen@penguincomputing.com>
14
15Module Parameters
16-----------------
17
18* gpio_input: int array (min = 1, max = 17)
19 List of GPIO pins (0-16) to program as inputs
20* gpio_output: int array (min = 1, max = 17)
21 List of GPIO pins (0-16) to program as outputs
22* gpio_inverted: int array (min = 1, max = 17)
23 List of GPIO pins (0-16) to program as inverted
24* gpio_normal: int array (min = 1, max = 17)
25 List of GPIO pins (0-16) to program as normal/non-inverted
26* gpio_fan: int array (min = 1, max = 8)
27 List of GPIO pins (0-7) to program as fan tachs
28
29
30Description
31-----------
32
33This driver implements support for the Analog Devices ADM1026. Analog
34Devices calls it a "complete thermal system management controller."
35
36The ADM1026 implements three (3) temperature sensors, 17 voltage sensors,
3716 general purpose digital I/O lines, eight (8) fan speed sensors (8-bit),
38an analog output and a PWM output along with limit, alarm and mask bits for
39all of the above. There is even 8k bytes of EEPROM memory on chip.
40
41Temperatures are measured in degrees Celsius. There are two external
42sensor inputs and one internal sensor. Each sensor has a high and low
43limit. If the limit is exceeded, an interrupt (#SMBALERT) can be
44generated. The interrupts can be masked. In addition, there are over-temp
45limits for each sensor. If this limit is exceeded, the #THERM output will
46be asserted. The current temperature and limits have a resolution of 1
47degree.
48
49Fan rotation speeds are reported in RPM (rotations per minute) but measured
50in counts of a 22.5kHz internal clock. Each fan has a high limit which
51corresponds to a minimum fan speed. If the limit is exceeded, an interrupt
52can be generated. Each fan can be programmed to divide the reference clock
53by 1, 2, 4 or 8. Not all RPM values can accurately be represented, so some
54rounding is done. With a divider of 8, the slowest measurable speed of a
55two pulse per revolution fan is 661 RPM.
56
57There are 17 voltage sensors. An alarm is triggered if the voltage has
58crossed a programmable minimum or maximum limit. Note that minimum in this
59case always means 'closest to zero'; this is important for negative voltage
60measurements. Several inputs have integrated attenuators so they can measure
61higher voltages directly. 3.3V, 5V, 12V, -12V and battery voltage all have
62dedicated inputs. There are several inputs scaled to 0-3V full-scale range
63for SCSI terminator power. The remaining inputs are not scaled and have
64a 0-2.5V full-scale range. A 2.5V or 1.82V reference voltage is provided
65for negative voltage measurements.
66
67If an alarm triggers, it will remain triggered until the hardware register
68is read at least once. This means that the cause for the alarm may already
69have disappeared! Note that in the current implementation, all hardware
70registers are read whenever any data is read (unless it is less than 2.0
71seconds since the last update). This means that you can easily miss
72once-only alarms.
73
74The ADM1026 measures continuously. Analog inputs are measured about 4
75times a second. Fan speed measurement time depends on fan speed and
76divisor. It can take as long as 1.5 seconds to measure all fan speeds.
77
78The ADM1026 has the ability to automatically control fan speed based on the
79temperature sensor inputs. Both the PWM output and the DAC output can be
80used to control fan speed. Usually only one of these two outputs will be
81used. Write the minimum PWM or DAC value to the appropriate control
82register. Then set the low temperature limit in the tmin values for each
83temperature sensor. The range of control is fixed at 20 °C, and the
84largest difference between current and tmin of the temperature sensors sets
85the control output. See the datasheet for several example circuits for
86controlling fan speed with the PWM and DAC outputs. The fan speed sensors
87do not have PWM compensation, so it is probably best to control the fan
88voltage from the power lead rather than on the ground lead.
89
90The datasheet shows an example application with VID signals attached to
91GPIO lines. Unfortunately, the chip may not be connected to the VID lines
92in this way. The driver assumes that the chips *is* connected this way to
93get a VID voltage.
diff --git a/Documentation/i2c/chips/adm1031 b/Documentation/i2c/chips/adm1031
new file mode 100644
index 000000000000..130a38382b98
--- /dev/null
+++ b/Documentation/i2c/chips/adm1031
@@ -0,0 +1,35 @@
1Kernel driver adm1031
2=====================
3
4Supported chips:
5 * Analog Devices ADM1030
6 Prefix: 'adm1030'
7 Addresses scanned: I2C 0x2c to 0x2e
8 Datasheet: Publicly available at the Analog Devices website
9 http://products.analog.com/products/info.asp?product=ADM1030
10
11 * Analog Devices ADM1031
12 Prefix: 'adm1031'
13 Addresses scanned: I2C 0x2c to 0x2e
14 Datasheet: Publicly available at the Analog Devices website
15 http://products.analog.com/products/info.asp?product=ADM1031
16
17Authors:
18 Alexandre d'Alton <alex@alexdalton.org>
19 Jean Delvare <khali@linux-fr.org>
20
21Description
22-----------
23
24The ADM1030 and ADM1031 are digital temperature sensors and fan controllers.
25They sense their own temperature as well as the temperature of up to one
26(ADM1030) or two (ADM1031) external diodes.
27
28All temperature values are given in degrees Celsius. Resolution is 0.5
29degree for the local temperature, 0.125 degree for the remote temperatures.
30
31Each temperature channel has its own high and low limits, plus a critical
32limit.
33
34The ADM1030 monitors a single fan speed, while the ADM1031 monitors up to
35two. Each fan channel has its own low speed limit.
diff --git a/Documentation/i2c/chips/asb100 b/Documentation/i2c/chips/asb100
new file mode 100644
index 000000000000..ab7365e139be
--- /dev/null
+++ b/Documentation/i2c/chips/asb100
@@ -0,0 +1,72 @@
1Kernel driver asb100
2====================
3
4Supported Chips:
5 * Asus ASB100 and ASB100-A "Bach"
6 Prefix: 'asb100'
7 Addresses scanned: I2C 0x2d
8 Datasheet: none released
9
10Author: Mark M. Hoffman <mhoffman@lightlink.com>
11
12Description
13-----------
14
15This driver implements support for the Asus ASB100 and ASB100-A "Bach".
16These are custom ASICs available only on Asus mainboards. Asus refuses to
17supply a datasheet for these chips. Thanks go to many people who helped
18investigate their hardware, including:
19
20Vitaly V. Bursov
21Alexander van Kaam (author of MBM for Windows)
22Bertrik Sikken
23
24The ASB100 implements seven voltage sensors, three fan rotation speed
25sensors, four temperature sensors, VID lines and alarms. In addition to
26these, the ASB100-A also implements a single PWM controller for fans 2 and
273 (i.e. one setting controls both.) If you have a plain ASB100, the PWM
28controller will simply not work (or maybe it will for you... it doesn't for
29me).
30
31Temperatures are measured and reported in degrees Celsius.
32
33Fan speeds are reported in RPM (rotations per minute). An alarm is
34triggered if the rotation speed has dropped below a programmable limit.
35
36Voltage sensors (also known as IN sensors) report values in volts.
37
38The VID lines encode the core voltage value: the voltage level your
39processor should work with. This is hardcoded by the mainboard and/or
40processor itself. It is a value in volts.
41
42Alarms: (TODO question marks indicate may or may not work)
43
440x0001 => in0 (?)
450x0002 => in1 (?)
460x0004 => in2
470x0008 => in3
480x0010 => temp1 (1)
490x0020 => temp2
500x0040 => fan1
510x0080 => fan2
520x0100 => in4
530x0200 => in5 (?) (2)
540x0400 => in6 (?) (2)
550x0800 => fan3
560x1000 => chassis switch
570x2000 => temp3
58
59Alarm Notes:
60
61(1) This alarm will only trigger if the hysteresis value is 127C.
62I.e. it behaves the same as w83781d.
63
64(2) The min and max registers for these values appear to
65be read-only or otherwise stuck at 0x00.
66
67TODO:
68* Experiment with fan divisors > 8.
69* Experiment with temp. sensor types.
70* Are there really 13 voltage inputs? Probably not...
71* Cleanups, no doubt...
72
diff --git a/Documentation/i2c/chips/ds1621 b/Documentation/i2c/chips/ds1621
new file mode 100644
index 000000000000..1fee6f1e6bc5
--- /dev/null
+++ b/Documentation/i2c/chips/ds1621
@@ -0,0 +1,108 @@
1Kernel driver ds1621
2====================
3
4Supported chips:
5 * Dallas Semiconductor DS1621
6 Prefix: 'ds1621'
7 Addresses scanned: I2C 0x48 - 0x4f
8 Datasheet: Publicly available at the Dallas Semiconductor website
9 http://www.dalsemi.com/
10 * Dallas Semiconductor DS1625
11 Prefix: 'ds1621'
12 Addresses scanned: I2C 0x48 - 0x4f
13 Datasheet: Publicly available at the Dallas Semiconductor website
14 http://www.dalsemi.com/
15
16Authors:
17 Christian W. Zuckschwerdt <zany@triq.net>
18 valuable contributions by Jan M. Sendler <sendler@sendler.de>
19 ported to 2.6 by Aurelien Jarno <aurelien@aurel32.net>
20 with the help of Jean Delvare <khali@linux-fr.org>
21
22Module Parameters
23------------------
24
25* polarity int
26 Output's polarity: 0 = active high, 1 = active low
27
28Description
29-----------
30
31The DS1621 is a (one instance) digital thermometer and thermostat. It has
32both high and low temperature limits which can be user defined (i.e.
33programmed into non-volatile on-chip registers). Temperature range is -55
34degree Celsius to +125 in 0.5 increments. You may convert this into a
35Fahrenheit range of -67 to +257 degrees with 0.9 steps. If polarity
36parameter is not provided, original value is used.
37
38As for the thermostat, behavior can also be programmed using the polarity
39toggle. On the one hand ("heater"), the thermostat output of the chip,
40Tout, will trigger when the low limit temperature is met or underrun and
41stays high until the high limit is met or exceeded. On the other hand
42("cooler"), vice versa. That way "heater" equals "active low", whereas
43"conditioner" equals "active high". Please note that the DS1621 data sheet
44is somewhat misleading in this point since setting the polarity bit does
45not simply invert Tout.
46
47A second thing is that, during extensive testing, Tout showed a tolerance
48of up to +/- 0.5 degrees even when compared against precise temperature
49readings. Be sure to have a high vs. low temperature limit gap of al least
501.0 degree Celsius to avoid Tout "bouncing", though!
51
52As for alarms, you can read the alarm status of the DS1621 via the 'alarms'
53/sys file interface. The result consists mainly of bit 6 and 5 of the
54configuration register of the chip; bit 6 (0x40 or 64) is the high alarm
55bit and bit 5 (0x20 or 32) the low one. These bits are set when the high or
56low limits are met or exceeded and are reset by the module as soon as the
57respective temperature ranges are left.
58
59The alarm registers are in no way suitable to find out about the actual
60status of Tout. They will only tell you about its history, whether or not
61any of the limits have ever been met or exceeded since last power-up or
62reset. Be aware: When testing, it showed that the status of Tout can change
63with neither of the alarms set.
64
65Temperature conversion of the DS1621 takes up to 1000ms; internal access to
66non-volatile registers may last for 10ms or below.
67
68High Accuracy Temperature Reading
69---------------------------------
70
71As said before, the temperature issued via the 9-bit i2c-bus data is
72somewhat arbitrary. Internally, the temperature conversion is of a
73different kind that is explained (not so...) well in the DS1621 data sheet.
74To cut the long story short: Inside the DS1621 there are two oscillators,
75both of them biassed by a temperature coefficient.
76
77Higher resolution of the temperature reading can be achieved using the
78internal projection, which means taking account of REG_COUNT and REG_SLOPE
79(the driver manages them):
80
81Taken from Dallas Semiconductors App Note 068: 'Increasing Temperature
82Resolution on the DS1620' and App Note 105: 'High Resolution Temperature
83Measurement with Dallas Direct-to-Digital Temperature Sensors'
84
85- Read the 9-bit temperature and strip the LSB (Truncate the .5 degs)
86- The resulting value is TEMP_READ.
87- Then, read REG_COUNT.
88- And then, REG_SLOPE.
89
90 TEMP = TEMP_READ - 0.25 + ((REG_SLOPE - REG_COUNT) / REG_SLOPE)
91
92Note that this is what the DONE bit in the DS1621 configuration register is
93good for: Internally, one temperature conversion takes up to 1000ms. Before
94that conversion is complete you will not be able to read valid things out
95of REG_COUNT and REG_SLOPE. The DONE bit, as you may have guessed by now,
96tells you whether the conversion is complete ("done", in plain English) and
97thus, whether the values you read are good or not.
98
99The DS1621 has two modes of operation: "Continuous" conversion, which can
100be understood as the default stand-alone mode where the chip gets the
101temperature and controls external devices via its Tout pin or tells other
102i2c's about it if they care. The other mode is called "1SHOT", that means
103that it only figures out about the temperature when it is explicitly told
104to do so; this can be seen as power saving mode.
105
106Now if you want to read REG_COUNT and REG_SLOPE, you have to either stop
107the continuous conversions until the contents of these registers are valid,
108or, in 1SHOT mode, you have to have one conversion made.
diff --git a/Documentation/i2c/chips/eeprom b/Documentation/i2c/chips/eeprom
new file mode 100644
index 000000000000..f7e8104b5764
--- /dev/null
+++ b/Documentation/i2c/chips/eeprom
@@ -0,0 +1,96 @@
1Kernel driver eeprom
2====================
3
4Supported chips:
5 * Any EEPROM chip in the designated address range
6 Prefix: 'eeprom'
7 Addresses scanned: I2C 0x50 - 0x57
8 Datasheets: Publicly available from:
9 Atmel (www.atmel.com),
10 Catalyst (www.catsemi.com),
11 Fairchild (www.fairchildsemi.com),
12 Microchip (www.microchip.com),
13 Philips (www.semiconductor.philips.com),
14 Rohm (www.rohm.com),
15 ST (www.st.com),
16 Xicor (www.xicor.com),
17 and others.
18
19 Chip Size (bits) Address
20 24C01 1K 0x50 (shadows at 0x51 - 0x57)
21 24C01A 1K 0x50 - 0x57 (Typical device on DIMMs)
22 24C02 2K 0x50 - 0x57
23 24C04 4K 0x50, 0x52, 0x54, 0x56
24 (additional data at 0x51, 0x53, 0x55, 0x57)
25 24C08 8K 0x50, 0x54 (additional data at 0x51, 0x52,
26 0x53, 0x55, 0x56, 0x57)
27 24C16 16K 0x50 (additional data at 0x51 - 0x57)
28 Sony 2K 0x57
29
30 Atmel 34C02B 2K 0x50 - 0x57, SW write protect at 0x30-37
31 Catalyst 34FC02 2K 0x50 - 0x57, SW write protect at 0x30-37
32 Catalyst 34RC02 2K 0x50 - 0x57, SW write protect at 0x30-37
33 Fairchild 34W02 2K 0x50 - 0x57, SW write protect at 0x30-37
34 Microchip 24AA52 2K 0x50 - 0x57, SW write protect at 0x30-37
35 ST M34C02 2K 0x50 - 0x57, SW write protect at 0x30-37
36
37
38Authors:
39 Frodo Looijaard <frodol@dds.nl>,
40 Philip Edelbrock <phil@netroedge.com>,
41 Jean Delvare <khali@linux-fr.org>,
42 Greg Kroah-Hartman <greg@kroah.com>,
43 IBM Corp.
44
45Description
46-----------
47
48This is a simple EEPROM module meant to enable reading the first 256 bytes
49of an EEPROM (on a SDRAM DIMM for example). However, it will access serial
50EEPROMs on any I2C adapter. The supported devices are generically called
5124Cxx, and are listed above; however the numbering for these
52industry-standard devices may vary by manufacturer.
53
54This module was a programming exercise to get used to the new project
55organization laid out by Frodo, but it should be at least completely
56effective for decoding the contents of EEPROMs on DIMMs.
57
58DIMMS will typically contain a 24C01A or 24C02, or the 34C02 variants.
59The other devices will not be found on a DIMM because they respond to more
60than one address.
61
62DDC Monitors may contain any device. Often a 24C01, which responds to all 8
63addresses, is found.
64
65Recent Sony Vaio laptops have an EEPROM at 0x57. We couldn't get the
66specification, so it is guess work and far from being complete.
67
68The Microchip 24AA52/24LCS52, ST M34C02, and others support an additional
69software write protect register at 0x30 - 0x37 (0x20 less than the memory
70location). The chip responds to "write quick" detection at this address but
71does not respond to byte reads. If this register is present, the lower 128
72bytes of the memory array are not write protected. Any byte data write to
73this address will write protect the memory array permanently, and the
74device will no longer respond at the 0x30-37 address. The eeprom driver
75does not support this register.
76
77Lacking functionality:
78
79* Full support for larger devices (24C04, 24C08, 24C16). These are not
80typically found on a PC. These devices will appear as separate devices at
81multiple addresses.
82
83* Support for really large devices (24C32, 24C64, 24C128, 24C256, 24C512).
84These devices require two-byte address fields and are not supported.
85
86* Enable Writing. Again, no technical reason why not, but making it easy
87to change the contents of the EEPROMs (on DIMMs anyway) also makes it easy
88to disable the DIMMs (potentially preventing the computer from booting)
89until the values are restored somehow.
90
91Use:
92
93After inserting the module (and any other required SMBus/i2c modules), you
94should have some EEPROM directories in /sys/bus/i2c/devices/* of names such
95as "0-0050". Inside each of these is a series of files, the eeprom file
96contains the binary data from EEPROM.
diff --git a/Documentation/i2c/chips/fscher b/Documentation/i2c/chips/fscher
new file mode 100644
index 000000000000..64031659aff3
--- /dev/null
+++ b/Documentation/i2c/chips/fscher
@@ -0,0 +1,169 @@
1Kernel driver fscher
2====================
3
4Supported chips:
5 * Fujitsu-Siemens Hermes chip
6 Prefix: 'fscher'
7 Addresses scanned: I2C 0x73
8
9Authors:
10 Reinhard Nissl <rnissl@gmx.de> based on work
11 from Hermann Jung <hej@odn.de>,
12 Frodo Looijaard <frodol@dds.nl>,
13 Philip Edelbrock <phil@netroedge.com>
14
15Description
16-----------
17
18This driver implements support for the Fujitsu-Siemens Hermes chip. It is
19described in the 'Register Set Specification BMC Hermes based Systemboard'
20from Fujitsu-Siemens.
21
22The Hermes chip implements a hardware-based system management, e.g. for
23controlling fan speed and core voltage. There is also a watchdog counter on
24the chip which can trigger an alarm and even shut the system down.
25
26The chip provides three temperature values (CPU, motherboard and
27auxiliary), three voltage values (+12V, +5V and battery) and three fans
28(power supply, CPU and auxiliary).
29
30Temperatures are measured in degrees Celsius. The resolution is 1 degree.
31
32Fan rotation speeds are reported in RPM (rotations per minute). The value
33can be divided by a programmable divider (1, 2 or 4) which is stored on
34the chip.
35
36Voltage sensors (also known as "in" sensors) report their values in volts.
37
38All values are reported as final values from the driver. There is no need
39for further calculations.
40
41
42Detailed description
43--------------------
44
45Below you'll find a single line description of all the bit values. With
46this information, you're able to decode e. g. alarms, wdog, etc. To make
47use of the watchdog, you'll need to set the watchdog time and enable the
48watchdog. After that it is necessary to restart the watchdog time within
49the specified period of time, or a system reset will occur.
50
51* revision
52 READING & 0xff = 0x??: HERMES revision identification
53
54* alarms
55 READING & 0x80 = 0x80: CPU throttling active
56 READING & 0x80 = 0x00: CPU running at full speed
57
58 READING & 0x10 = 0x10: software event (see control:1)
59 READING & 0x10 = 0x00: no software event
60
61 READING & 0x08 = 0x08: watchdog event (see wdog:2)
62 READING & 0x08 = 0x00: no watchdog event
63
64 READING & 0x02 = 0x02: thermal event (see temp*:1)
65 READING & 0x02 = 0x00: no thermal event
66
67 READING & 0x01 = 0x01: fan event (see fan*:1)
68 READING & 0x01 = 0x00: no fan event
69
70 READING & 0x13 ! 0x00: ALERT LED is flashing
71
72* control
73 READING & 0x01 = 0x01: software event
74 READING & 0x01 = 0x00: no software event
75
76 WRITING & 0x01 = 0x01: set software event
77 WRITING & 0x01 = 0x00: clear software event
78
79* watchdog_control
80 READING & 0x80 = 0x80: power off on watchdog event while thermal event
81 READING & 0x80 = 0x00: watchdog power off disabled (just system reset enabled)
82
83 READING & 0x40 = 0x40: watchdog timebase 60 seconds (see also wdog:1)
84 READING & 0x40 = 0x00: watchdog timebase 2 seconds
85
86 READING & 0x10 = 0x10: watchdog enabled
87 READING & 0x10 = 0x00: watchdog disabled
88
89 WRITING & 0x80 = 0x80: enable "power off on watchdog event while thermal event"
90 WRITING & 0x80 = 0x00: disable "power off on watchdog event while thermal event"
91
92 WRITING & 0x40 = 0x40: set watchdog timebase to 60 seconds
93 WRITING & 0x40 = 0x00: set watchdog timebase to 2 seconds
94
95 WRITING & 0x20 = 0x20: disable watchdog
96
97 WRITING & 0x10 = 0x10: enable watchdog / restart watchdog time
98
99* watchdog_state
100 READING & 0x02 = 0x02: watchdog system reset occurred
101 READING & 0x02 = 0x00: no watchdog system reset occurred
102
103 WRITING & 0x02 = 0x02: clear watchdog event
104
105* watchdog_preset
106 READING & 0xff = 0x??: configured watch dog time in units (see wdog:3 0x40)
107
108 WRITING & 0xff = 0x??: configure watch dog time in units
109
110* in* (0: +5V, 1: +12V, 2: onboard 3V battery)
111 READING: actual voltage value
112
113* temp*_status (1: CPU sensor, 2: onboard sensor, 3: auxiliary sensor)
114 READING & 0x02 = 0x02: thermal event (overtemperature)
115 READING & 0x02 = 0x00: no thermal event
116
117 READING & 0x01 = 0x01: sensor is working
118 READING & 0x01 = 0x00: sensor is faulty
119
120 WRITING & 0x02 = 0x02: clear thermal event
121
122* temp*_input (1: CPU sensor, 2: onboard sensor, 3: auxiliary sensor)
123 READING: actual temperature value
124
125* fan*_status (1: power supply fan, 2: CPU fan, 3: auxiliary fan)
126 READING & 0x04 = 0x04: fan event (fan fault)
127 READING & 0x04 = 0x00: no fan event
128
129 WRITING & 0x04 = 0x04: clear fan event
130
131* fan*_div (1: power supply fan, 2: CPU fan, 3: auxiliary fan)
132 Divisors 2,4 and 8 are supported, both for reading and writing
133
134* fan*_pwm (1: power supply fan, 2: CPU fan, 3: auxiliary fan)
135 READING & 0xff = 0x00: fan may be switched off
136 READING & 0xff = 0x01: fan must run at least at minimum speed (supply: 6V)
137 READING & 0xff = 0xff: fan must run at maximum speed (supply: 12V)
138 READING & 0xff = 0x??: fan must run at least at given speed (supply: 6V..12V)
139
140 WRITING & 0xff = 0x00: fan may be switched off
141 WRITING & 0xff = 0x01: fan must run at least at minimum speed (supply: 6V)
142 WRITING & 0xff = 0xff: fan must run at maximum speed (supply: 12V)
143 WRITING & 0xff = 0x??: fan must run at least at given speed (supply: 6V..12V)
144
145* fan*_input (1: power supply fan, 2: CPU fan, 3: auxiliary fan)
146 READING: actual RPM value
147
148
149Limitations
150-----------
151
152* Measuring fan speed
153It seems that the chip counts "ripples" (typical fans produce 2 ripples per
154rotation while VERAX fans produce 18) in a 9-bit register. This register is
155read out every second, then the ripple prescaler (2, 4 or 8) is applied and
156the result is stored in the 8 bit output register. Due to the limitation of
157the counting register to 9 bits, it is impossible to measure a VERAX fan
158properly (even with a prescaler of 8). At its maximum speed of 3500 RPM the
159fan produces 1080 ripples per second which causes the counting register to
160overflow twice, leading to only 186 RPM.
161
162* Measuring input voltages
163in2 ("battery") reports the voltage of the onboard lithium battery and not
164+3.3V from the power supply.
165
166* Undocumented features
167Fujitsu-Siemens Computers has not documented all features of the chip so
168far. Their software, System Guard, shows that there are a still some
169features which cannot be controlled by this implementation.
diff --git a/Documentation/i2c/chips/gl518sm b/Documentation/i2c/chips/gl518sm
new file mode 100644
index 000000000000..ce0881883bca
--- /dev/null
+++ b/Documentation/i2c/chips/gl518sm
@@ -0,0 +1,74 @@
1Kernel driver gl518sm
2=====================
3
4Supported chips:
5 * Genesys Logic GL518SM release 0x00
6 Prefix: 'gl518sm'
7 Addresses scanned: I2C 0x2c and 0x2d
8 Datasheet: http://www.genesyslogic.com/pdf
9 * Genesys Logic GL518SM release 0x80
10 Prefix: 'gl518sm'
11 Addresses scanned: I2C 0x2c and 0x2d
12 Datasheet: http://www.genesyslogic.com/pdf
13
14Authors:
15 Frodo Looijaard <frodol@dds.nl>,
16 Kyösti Mälkki <kmalkki@cc.hut.fi>
17 Hong-Gunn Chew <hglinux@gunnet.org>
18 Jean Delvare <khali@linux-fr.org>
19
20Description
21-----------
22
23IMPORTANT:
24
25For the revision 0x00 chip, the in0, in1, and in2 values (+5V, +3V,
26and +12V) CANNOT be read. This is a limitation of the chip, not the driver.
27
28This driver supports the Genesys Logic GL518SM chip. There are at least
29two revision of this chip, which we call revision 0x00 and 0x80. Revision
300x80 chips support the reading of all voltages and revision 0x00 only
31for VIN3.
32
33The GL518SM implements one temperature sensor, two fan rotation speed
34sensors, and four voltage sensors. It can report alarms through the
35computer speakers.
36
37Temperatures are measured in degrees Celsius. An alarm goes off while the
38temperature is above the over temperature limit, and has not yet dropped
39below the hysteresis limit. The alarm always reflects the current
40situation. Measurements are guaranteed between -10 degrees and +110
41degrees, with a accuracy of +/-3 degrees.
42
43Rotation speeds are reported in RPM (rotations per minute). An alarm is
44triggered if the rotation speed has dropped below a programmable limit. In
45case when you have selected to turn fan1 off, no fan1 alarm is triggered.
46
47Fan readings can be divided by a programmable divider (1, 2, 4 or 8) to
48give the readings more range or accuracy. Not all RPM values can
49accurately be represented, so some rounding is done. With a divider
50of 2, the lowest representable value is around 1900 RPM.
51
52Voltage sensors (also known as VIN sensors) report their values in volts.
53An alarm is triggered if the voltage has crossed a programmable minimum or
54maximum limit. Note that minimum in this case always means 'closest to
55zero'; this is important for negative voltage measurements. The VDD input
56measures voltages between 0.000 and 5.865 volt, with a resolution of 0.023
57volt. The other inputs measure voltages between 0.000 and 4.845 volt, with
58a resolution of 0.019 volt. Note that revision 0x00 chips do not support
59reading the current voltage of any input except for VIN3; limit setting and
60alarms work fine, though.
61
62When an alarm is triggered, you can be warned by a beeping signal through your
63computer speaker. It is possible to enable all beeping globally, or only the
64beeping for some alarms.
65
66If an alarm triggers, it will remain triggered until the hardware register
67is read at least once (except for temperature alarms). This means that the
68cause for the alarm may already have disappeared! Note that in the current
69implementation, all hardware registers are read whenever any data is read
70(unless it is less than 1.5 seconds since the last update). This means that
71you can easily miss once-only alarms.
72
73The GL518SM only updates its values each 1.5 seconds; reading it more often
74will do no harm, but will return 'old' values.
diff --git a/Documentation/i2c/chips/it87 b/Documentation/i2c/chips/it87
new file mode 100644
index 000000000000..0d0195040d88
--- /dev/null
+++ b/Documentation/i2c/chips/it87
@@ -0,0 +1,96 @@
1Kernel driver it87
2==================
3
4Supported chips:
5 * IT8705F
6 Prefix: 'it87'
7 Addresses scanned: from Super I/O config space, or default ISA 0x290 (8 I/O ports)
8 Datasheet: Publicly available at the ITE website
9 http://www.ite.com.tw/
10 * IT8712F
11 Prefix: 'it8712'
12 Addresses scanned: I2C 0x28 - 0x2f
13 from Super I/O config space, or default ISA 0x290 (8 I/O ports)
14 Datasheet: Publicly available at the ITE website
15 http://www.ite.com.tw/
16 * SiS950 [clone of IT8705F]
17 Prefix: 'sis950'
18 Addresses scanned: from Super I/O config space, or default ISA 0x290 (8 I/O ports)
19 Datasheet: No longer be available
20
21Author: Christophe Gauthron <chrisg@0-in.com>
22
23
24Module Parameters
25-----------------
26
27* update_vbat: int
28
29 0 if vbat should report power on value, 1 if vbat should be updated after
30 each read. Default is 0. On some boards the battery voltage is provided
31 by either the battery or the onboard power supply. Only the first reading
32 at power on will be the actual battery voltage (which the chip does
33 automatically). On other boards the battery voltage is always fed to
34 the chip so can be read at any time. Excessive reading may decrease
35 battery life but no information is given in the datasheet.
36
37* fix_pwm_polarity int
38
39 Force PWM polarity to active high (DANGEROUS). Some chips are
40 misconfigured by BIOS - PWM values would be inverted. This option tries
41 to fix this. Please contact your BIOS manufacturer and ask him for fix.
42
43Description
44-----------
45
46This driver implements support for the IT8705F, IT8712F and SiS950 chips.
47
48This driver also supports IT8712F, which adds SMBus access, and a VID
49input, used to report the Vcore voltage of the Pentium processor.
50The IT8712F additionally features VID inputs.
51
52These chips are 'Super I/O chips', supporting floppy disks, infrared ports,
53joysticks and other miscellaneous stuff. For hardware monitoring, they
54include an 'environment controller' with 3 temperature sensors, 3 fan
55rotation speed sensors, 8 voltage sensors, and associated alarms.
56
57Temperatures are measured in degrees Celsius. An alarm is triggered once
58when the Overtemperature Shutdown limit is crossed.
59
60Fan rotation speeds are reported in RPM (rotations per minute). An alarm is
61triggered if the rotation speed has dropped below a programmable limit. Fan
62readings can be divided by a programmable divider (1, 2, 4 or 8) to give the
63readings more range or accuracy. Not all RPM values can accurately be
64represented, so some rounding is done. With a divider of 2, the lowest
65representable value is around 2600 RPM.
66
67Voltage sensors (also known as IN sensors) report their values in volts. An
68alarm is triggered if the voltage has crossed a programmable minimum or
69maximum limit. Note that minimum in this case always means 'closest to
70zero'; this is important for negative voltage measurements. All voltage
71inputs can measure voltages between 0 and 4.08 volts, with a resolution of
720.016 volt. The battery voltage in8 does not have limit registers.
73
74The VID lines (IT8712F only) encode the core voltage value: the voltage
75level your processor should work with. This is hardcoded by the mainboard
76and/or processor itself. It is a value in volts.
77
78If an alarm triggers, it will remain triggered until the hardware register
79is read at least once. This means that the cause for the alarm may already
80have disappeared! Note that in the current implementation, all hardware
81registers are read whenever any data is read (unless it is less than 1.5
82seconds since the last update). This means that you can easily miss
83once-only alarms.
84
85The IT87xx only updates its values each 1.5 seconds; reading it more often
86will do no harm, but will return 'old' values.
87
88To change sensor N to a thermistor, 'echo 2 > tempN_type' where N is 1, 2,
89or 3. To change sensor N to a thermal diode, 'echo 3 > tempN_type'.
90Give 0 for unused sensor. Any other value is invalid. To configure this at
91startup, consult lm_sensors's /etc/sensors.conf. (2 = thermistor;
923 = thermal diode)
93
94The fan speed control features are limited to manual PWM mode. Automatic
95"Smart Guardian" mode control handling is not implemented. However
96if you want to go for "manual mode" just write 1 to pwmN_enable.
diff --git a/Documentation/i2c/chips/lm63 b/Documentation/i2c/chips/lm63
new file mode 100644
index 000000000000..31660bf97979
--- /dev/null
+++ b/Documentation/i2c/chips/lm63
@@ -0,0 +1,57 @@
1Kernel driver lm63
2==================
3
4Supported chips:
5 * National Semiconductor LM63
6 Prefix: 'lm63'
7 Addresses scanned: I2C 0x4c
8 Datasheet: Publicly available at the National Semiconductor website
9 http://www.national.com/pf/LM/LM63.html
10
11Author: Jean Delvare <khali@linux-fr.org>
12
13Thanks go to Tyan and especially Alex Buckingham for setting up a remote
14access to their S4882 test platform for this driver.
15 http://www.tyan.com/
16
17Description
18-----------
19
20The LM63 is a digital temperature sensor with integrated fan monitoring
21and control.
22
23The LM63 is basically an LM86 with fan speed monitoring and control
24capabilities added. It misses some of the LM86 features though:
25 - No low limit for local temperature.
26 - No critical limit for local temperature.
27 - Critical limit for remote temperature can be changed only once. We
28 will consider that the critical limit is read-only.
29
30The datasheet isn't very clear about what the tachometer reading is.
31
32An explanation from National Semiconductor: The two lower bits of the read
33value have to be masked out. The value is still 16 bit in width.
34
35All temperature values are given in degrees Celsius. Resolution is 1.0
36degree for the local temperature, 0.125 degree for the remote temperature.
37
38The fan speed is measured using a tachometer. Contrary to most chips which
39store the value in an 8-bit register and have a selectable clock divider
40to make sure that the result will fit in the register, the LM63 uses 16-bit
41value for measuring the speed of the fan. It can measure fan speeds down to
4283 RPM, at least in theory.
43
44Note that the pin used for fan monitoring is shared with an alert out
45function. Depending on how the board designer wanted to use the chip, fan
46speed monitoring will or will not be possible. The proper chip configuration
47is left to the BIOS, and the driver will blindly trust it.
48
49A PWM output can be used to control the speed of the fan. The LM63 has two
50PWM modes: manual and automatic. Automatic mode is not fully implemented yet
51(you cannot define your custom PWM/temperature curve), and mode change isn't
52supported either.
53
54The lm63 driver will not update its values more frequently than every
55second; reading them more often will do no harm, but will return 'old'
56values.
57
diff --git a/Documentation/i2c/chips/lm75 b/Documentation/i2c/chips/lm75
new file mode 100644
index 000000000000..8e6356fe05d7
--- /dev/null
+++ b/Documentation/i2c/chips/lm75
@@ -0,0 +1,65 @@
1Kernel driver lm75
2==================
3
4Supported chips:
5 * National Semiconductor LM75
6 Prefix: 'lm75'
7 Addresses scanned: I2C 0x48 - 0x4f
8 Datasheet: Publicly available at the National Semiconductor website
9 http://www.national.com/
10 * Dallas Semiconductor DS75
11 Prefix: 'lm75'
12 Addresses scanned: I2C 0x48 - 0x4f
13 Datasheet: Publicly available at the Dallas Semiconductor website
14 http://www.maxim-ic.com/
15 * Dallas Semiconductor DS1775
16 Prefix: 'lm75'
17 Addresses scanned: I2C 0x48 - 0x4f
18 Datasheet: Publicly available at the Dallas Semiconductor website
19 http://www.maxim-ic.com/
20 * Maxim MAX6625, MAX6626
21 Prefix: 'lm75'
22 Addresses scanned: I2C 0x48 - 0x4b
23 Datasheet: Publicly available at the Maxim website
24 http://www.maxim-ic.com/
25 * Microchip (TelCom) TCN75
26 Prefix: 'lm75'
27 Addresses scanned: I2C 0x48 - 0x4f
28 Datasheet: Publicly available at the Microchip website
29 http://www.microchip.com/
30
31Author: Frodo Looijaard <frodol@dds.nl>
32
33Description
34-----------
35
36The LM75 implements one temperature sensor. Limits can be set through the
37Overtemperature Shutdown register and Hysteresis register. Each value can be
38set and read to half-degree accuracy.
39An alarm is issued (usually to a connected LM78) when the temperature
40gets higher then the Overtemperature Shutdown value; it stays on until
41the temperature falls below the Hysteresis value.
42All temperatures are in degrees Celsius, and are guaranteed within a
43range of -55 to +125 degrees.
44
45The LM75 only updates its values each 1.5 seconds; reading it more often
46will do no harm, but will return 'old' values.
47
48The LM75 is usually used in combination with LM78-like chips, to measure
49the temperature of the processor(s).
50
51The DS75, DS1775, MAX6625, and MAX6626 are supported as well.
52They are not distinguished from an LM75. While most of these chips
53have three additional bits of accuracy (12 vs. 9 for the LM75),
54the additional bits are not supported. Not only that, but these chips will
55not be detected if not in 9-bit precision mode (use the force parameter if
56needed).
57
58The TCN75 is supported as well, and is not distinguished from an LM75.
59
60The LM75 is essentially an industry standard; there may be other
61LM75 clones not listed here, with or without various enhancements,
62that are supported.
63
64The LM77 is not supported, contrary to what we pretended for a long time.
65Both chips are simply not compatible, value encoding differs.
diff --git a/Documentation/i2c/chips/lm77 b/Documentation/i2c/chips/lm77
new file mode 100644
index 000000000000..57c3a46d6370
--- /dev/null
+++ b/Documentation/i2c/chips/lm77
@@ -0,0 +1,22 @@
1Kernel driver lm77
2==================
3
4Supported chips:
5 * National Semiconductor LM77
6 Prefix: 'lm77'
7 Addresses scanned: I2C 0x48 - 0x4b
8 Datasheet: Publicly available at the National Semiconductor website
9 http://www.national.com/
10
11Author: Andras BALI <drewie@freemail.hu>
12
13Description
14-----------
15
16The LM77 implements one temperature sensor. The temperature
17sensor incorporates a band-gap type temperature sensor,
1810-bit ADC, and a digital comparator with user-programmable upper
19and lower limit values.
20
21Limits can be set through the Overtemperature Shutdown register and
22Hysteresis register.
diff --git a/Documentation/i2c/chips/lm78 b/Documentation/i2c/chips/lm78
new file mode 100644
index 000000000000..357086ed7f64
--- /dev/null
+++ b/Documentation/i2c/chips/lm78
@@ -0,0 +1,82 @@
1Kernel driver lm78
2==================
3
4Supported chips:
5 * National Semiconductor LM78
6 Prefix: 'lm78'
7 Addresses scanned: I2C 0x20 - 0x2f, ISA 0x290 (8 I/O ports)
8 Datasheet: Publicly available at the National Semiconductor website
9 http://www.national.com/
10 * National Semiconductor LM78-J
11 Prefix: 'lm78-j'
12 Addresses scanned: I2C 0x20 - 0x2f, ISA 0x290 (8 I/O ports)
13 Datasheet: Publicly available at the National Semiconductor website
14 http://www.national.com/
15 * National Semiconductor LM79
16 Prefix: 'lm79'
17 Addresses scanned: I2C 0x20 - 0x2f, ISA 0x290 (8 I/O ports)
18 Datasheet: Publicly available at the National Semiconductor website
19 http://www.national.com/
20
21Author: Frodo Looijaard <frodol@dds.nl>
22
23Description
24-----------
25
26This driver implements support for the National Semiconductor LM78, LM78-J
27and LM79. They are described as 'Microprocessor System Hardware Monitors'.
28
29There is almost no difference between the three supported chips. Functionally,
30the LM78 and LM78-J are exactly identical. The LM79 has one more VID line,
31which is used to report the lower voltages newer Pentium processors use.
32From here on, LM7* means either of these three types.
33
34The LM7* implements one temperature sensor, three fan rotation speed sensors,
35seven voltage sensors, VID lines, alarms, and some miscellaneous stuff.
36
37Temperatures are measured in degrees Celsius. An alarm is triggered once
38when the Overtemperature Shutdown limit is crossed; it is triggered again
39as soon as it drops below the Hysteresis value. A more useful behavior
40can be found by setting the Hysteresis value to +127 degrees Celsius; in
41this case, alarms are issued during all the time when the actual temperature
42is above the Overtemperature Shutdown value. Measurements are guaranteed
43between -55 and +125 degrees, with a resolution of 1 degree.
44
45Fan rotation speeds are reported in RPM (rotations per minute). An alarm is
46triggered if the rotation speed has dropped below a programmable limit. Fan
47readings can be divided by a programmable divider (1, 2, 4 or 8) to give
48the readings more range or accuracy. Not all RPM values can accurately be
49represented, so some rounding is done. With a divider of 2, the lowest
50representable value is around 2600 RPM.
51
52Voltage sensors (also known as IN sensors) report their values in volts.
53An alarm is triggered if the voltage has crossed a programmable minimum
54or maximum limit. Note that minimum in this case always means 'closest to
55zero'; this is important for negative voltage measurements. All voltage
56inputs can measure voltages between 0 and 4.08 volts, with a resolution
57of 0.016 volt.
58
59The VID lines encode the core voltage value: the voltage level your processor
60should work with. This is hardcoded by the mainboard and/or processor itself.
61It is a value in volts. When it is unconnected, you will often find the
62value 3.50 V here.
63
64In addition to the alarms described above, there are a couple of additional
65ones. There is a BTI alarm, which gets triggered when an external chip has
66crossed its limits. Usually, this is connected to all LM75 chips; if at
67least one crosses its limits, this bit gets set. The CHAS alarm triggers
68if your computer case is open. The FIFO alarms should never trigger; it
69indicates an internal error. The SMI_IN alarm indicates some other chip
70has triggered an SMI interrupt. As we do not use SMI interrupts at all,
71this condition usually indicates there is a problem with some other
72device.
73
74If an alarm triggers, it will remain triggered until the hardware register
75is read at least once. This means that the cause for the alarm may
76already have disappeared! Note that in the current implementation, all
77hardware registers are read whenever any data is read (unless it is less
78than 1.5 seconds since the last update). This means that you can easily
79miss once-only alarms.
80
81The LM7* only updates its values each 1.5 seconds; reading it more often
82will do no harm, but will return 'old' values.
diff --git a/Documentation/i2c/chips/lm80 b/Documentation/i2c/chips/lm80
new file mode 100644
index 000000000000..cb5b407ba3e6
--- /dev/null
+++ b/Documentation/i2c/chips/lm80
@@ -0,0 +1,56 @@
1Kernel driver lm80
2==================
3
4Supported chips:
5 * National Semiconductor LM80
6 Prefix: 'lm80'
7 Addresses scanned: I2C 0x28 - 0x2f
8 Datasheet: Publicly available at the National Semiconductor website
9 http://www.national.com/
10
11Authors:
12 Frodo Looijaard <frodol@dds.nl>,
13 Philip Edelbrock <phil@netroedge.com>
14
15Description
16-----------
17
18This driver implements support for the National Semiconductor LM80.
19It is described as a 'Serial Interface ACPI-Compatible Microprocessor
20System Hardware Monitor'.
21
22The LM80 implements one temperature sensor, two fan rotation speed sensors,
23seven voltage sensors, alarms, and some miscellaneous stuff.
24
25Temperatures are measured in degrees Celsius. There are two sets of limits
26which operate independently. When the HOT Temperature Limit is crossed,
27this will cause an alarm that will be reasserted until the temperature
28drops below the HOT Hysteresis. The Overtemperature Shutdown (OS) limits
29should work in the same way (but this must be checked; the datasheet
30is unclear about this). Measurements are guaranteed between -55 and
31+125 degrees. The current temperature measurement has a resolution of
320.0625 degrees; the limits have a resolution of 1 degree.
33
34Fan rotation speeds are reported in RPM (rotations per minute). An alarm is
35triggered if the rotation speed has dropped below a programmable limit. Fan
36readings can be divided by a programmable divider (1, 2, 4 or 8) to give
37the readings more range or accuracy. Not all RPM values can accurately be
38represented, so some rounding is done. With a divider of 2, the lowest
39representable value is around 2600 RPM.
40
41Voltage sensors (also known as IN sensors) report their values in volts.
42An alarm is triggered if the voltage has crossed a programmable minimum
43or maximum limit. Note that minimum in this case always means 'closest to
44zero'; this is important for negative voltage measurements. All voltage
45inputs can measure voltages between 0 and 2.55 volts, with a resolution
46of 0.01 volt.
47
48If an alarm triggers, it will remain triggered until the hardware register
49is read at least once. This means that the cause for the alarm may
50already have disappeared! Note that in the current implementation, all
51hardware registers are read whenever any data is read (unless it is less
52than 2.0 seconds since the last update). This means that you can easily
53miss once-only alarms.
54
55The LM80 only updates its values each 1.5 seconds; reading it more often
56will do no harm, but will return 'old' values.
diff --git a/Documentation/i2c/chips/lm83 b/Documentation/i2c/chips/lm83
new file mode 100644
index 000000000000..061d9ed8ff43
--- /dev/null
+++ b/Documentation/i2c/chips/lm83
@@ -0,0 +1,76 @@
1Kernel driver lm83
2==================
3
4Supported chips:
5 * National Semiconductor LM83
6 Prefix: 'lm83'
7 Addresses scanned: I2C 0x18 - 0x1a, 0x29 - 0x2b, 0x4c - 0x4e
8 Datasheet: Publicly available at the National Semiconductor website
9 http://www.national.com/pf/LM/LM83.html
10
11
12Author: Jean Delvare <khali@linux-fr.org>
13
14Description
15-----------
16
17The LM83 is a digital temperature sensor. It senses its own temperature as
18well as the temperature of up to three external diodes. It is compatible
19with many other devices such as the LM84 and all other ADM1021 clones.
20The main difference between the LM83 and the LM84 in that the later can
21only sense the temperature of one external diode.
22
23Using the adm1021 driver for a LM83 should work, but only two temperatures
24will be reported instead of four.
25
26The LM83 is only found on a handful of motherboards. Both a confirmed
27list and an unconfirmed list follow. If you can confirm or infirm the
28fact that any of these motherboards do actually have an LM83, please
29contact us. Note that the LM90 can easily be misdetected as a LM83.
30
31Confirmed motherboards:
32 SBS P014
33
34Unconfirmed motherboards:
35 Gigabyte GA-8IK1100
36 Iwill MPX2
37 Soltek SL-75DRV5
38
39The driver has been successfully tested by Magnus Forsström, who I'd
40like to thank here. More testers will be of course welcome.
41
42The fact that the LM83 is only scarcely used can be easily explained.
43Most motherboards come with more than just temperature sensors for
44health monitoring. They also have voltage and fan rotation speed
45sensors. This means that temperature-only chips are usually used as
46secondary chips coupled with another chip such as an IT8705F or similar
47chip, which provides more features. Since systems usually need three
48temperature sensors (motherboard, processor, power supply) and primary
49chips provide some temperature sensors, the secondary chip, if needed,
50won't have to handle more than two temperatures. Thus, ADM1021 clones
51are sufficient, and there is no need for a four temperatures sensor
52chip such as the LM83. The only case where using an LM83 would make
53sense is on SMP systems, such as the above-mentioned Iwill MPX2,
54because you want an additional temperature sensor for each additional
55CPU.
56
57On the SBS P014, this is different, since the LM83 is the only hardware
58monitoring chipset. One temperature sensor is used for the motherboard
59(actually measuring the LM83's own temperature), one is used for the
60CPU. The two other sensors must be used to measure the temperature of
61two other points of the motherboard. We suspect these points to be the
62north and south bridges, but this couldn't be confirmed.
63
64All temperature values are given in degrees Celsius. Local temperature
65is given within a range of 0 to +85 degrees. Remote temperatures are
66given within a range of 0 to +125 degrees. Resolution is 1.0 degree,
67accuracy is guaranteed to 3.0 degrees (see the datasheet for more
68details).
69
70Each sensor has its own high limit, but the critical limit is common to
71all four sensors. There is no hysteresis mechanism as found on most
72recent temperature sensors.
73
74The lm83 driver will not update its values more frequently than every
75other second; reading them more often will do no harm, but will return
76'old' values.
diff --git a/Documentation/i2c/chips/lm85 b/Documentation/i2c/chips/lm85
new file mode 100644
index 000000000000..9549237530cf
--- /dev/null
+++ b/Documentation/i2c/chips/lm85
@@ -0,0 +1,221 @@
1Kernel driver lm85
2==================
3
4Supported chips:
5 * National Semiconductor LM85 (B and C versions)
6 Prefix: 'lm85'
7 Addresses scanned: I2C 0x2c, 0x2d, 0x2e
8 Datasheet: http://www.national.com/pf/LM/LM85.html
9 * Analog Devices ADM1027
10 Prefix: 'adm1027'
11 Addresses scanned: I2C 0x2c, 0x2d, 0x2e
12 Datasheet: http://www.analog.com/en/prod/0,,766_825_ADM1027,00.html
13 * Analog Devices ADT7463
14 Prefix: 'adt7463'
15 Addresses scanned: I2C 0x2c, 0x2d, 0x2e
16 Datasheet: http://www.analog.com/en/prod/0,,766_825_ADT7463,00.html
17 * SMSC EMC6D100, SMSC EMC6D101
18 Prefix: 'emc6d100'
19 Addresses scanned: I2C 0x2c, 0x2d, 0x2e
20 Datasheet: http://www.smsc.com/main/tools/discontinued/6d100.pdf
21 * SMSC EMC6D102
22 Prefix: 'emc6d102'
23 Addresses scanned: I2C 0x2c, 0x2d, 0x2e
24 Datasheet: http://www.smsc.com/main/catalog/emc6d102.html
25
26Authors:
27 Philip Pokorny <ppokorny@penguincomputing.com>,
28 Frodo Looijaard <frodol@dds.nl>,
29 Richard Barrington <rich_b_nz@clear.net.nz>,
30 Margit Schubert-While <margitsw@t-online.de>,
31 Justin Thiessen <jthiessen@penguincomputing.com>
32
33Description
34-----------
35
36This driver implements support for the National Semiconductor LM85 and
37compatible chips including the Analog Devices ADM1027, ADT7463 and
38SMSC EMC6D10x chips family.
39
40The LM85 uses the 2-wire interface compatible with the SMBUS 2.0
41specification. Using an analog to digital converter it measures three (3)
42temperatures and five (5) voltages. It has four (4) 16-bit counters for
43measuring fan speed. Five (5) digital inputs are provided for sampling the
44VID signals from the processor to the VRM. Lastly, there are three (3) PWM
45outputs that can be used to control fan speed.
46
47The voltage inputs have internal scaling resistors so that the following
48voltage can be measured without external resistors:
49
50 2.5V, 3.3V, 5V, 12V, and CPU core voltage (2.25V)
51
52The temperatures measured are one internal diode, and two remote diodes.
53Remote 1 is generally the CPU temperature. These inputs are designed to
54measure a thermal diode like the one in a Pentium 4 processor in a socket
55423 or socket 478 package. They can also measure temperature using a
56transistor like the 2N3904.
57
58A sophisticated control system for the PWM outputs is designed into the
59LM85 that allows fan speed to be adjusted automatically based on any of the
60three temperature sensors. Each PWM output is individually adjustable and
61programmable. Once configured, the LM85 will adjust the PWM outputs in
62response to the measured temperatures without further host intervention.
63This feature can also be disabled for manual control of the PWM's.
64
65Each of the measured inputs (voltage, temperature, fan speed) has
66corresponding high/low limit values. The LM85 will signal an ALARM if any
67measured value exceeds either limit.
68
69The LM85 samples all inputs continuously. The lm85 driver will not read
70the registers more often than once a second. Further, configuration data is
71only read once each 5 minutes. There is twice as much config data as
72measurements, so this would seem to be a worthwhile optimization.
73
74Special Features
75----------------
76
77The LM85 has four fan speed monitoring modes. The ADM1027 has only two.
78Both have special circuitry to compensate for PWM interactions with the
79TACH signal from the fans. The ADM1027 can be configured to measure the
80speed of a two wire fan, but the input conditioning circuitry is different
81for 3-wire and 2-wire mode. For this reason, the 2-wire fan modes are not
82exposed to user control. The BIOS should initialize them to the correct
83mode. If you've designed your own ADM1027, you'll have to modify the
84init_client function and add an insmod parameter to set this up.
85
86To smooth the response of fans to changes in temperature, the LM85 has an
87optional filter for smoothing temperatures. The ADM1027 has the same
88config option but uses it to rate limit the changes to fan speed instead.
89
90The ADM1027 and ADT7463 have a 10-bit ADC and can therefore measure
91temperatures with 0.25 degC resolution. They also provide an offset to the
92temperature readings that is automatically applied during measurement.
93This offset can be used to zero out any errors due to traces and placement.
94The documentation says that the offset is in 0.25 degC steps, but in
95initial testing of the ADM1027 it was 1.00 degC steps. Analog Devices has
96confirmed this "bug". The ADT7463 is reported to work as described in the
97documentation. The current lm85 driver does not show the offset register.
98
99The ADT7463 has a THERM asserted counter. This counter has a 22.76ms
100resolution and a range of 5.8 seconds. The driver implements a 32-bit
101accumulator of the counter value to extend the range to over a year. The
102counter will stay at it's max value until read.
103
104See the vendor datasheets for more information. There is application note
105from National (AN-1260) with some additional information about the LM85.
106The Analog Devices datasheet is very detailed and describes a procedure for
107determining an optimal configuration for the automatic PWM control.
108
109The SMSC EMC6D100 & EMC6D101 monitor external voltages, temperatures, and
110fan speeds. They use this monitoring capability to alert the system to out
111of limit conditions and can automatically control the speeds of multiple
112fans in a PC or embedded system. The EMC6D101, available in a 24-pin SSOP
113package, and the EMC6D100, available in a 28-pin SSOP package, are designed
114to be register compatible. The EMC6D100 offers all the features of the
115EMC6D101 plus additional voltage monitoring and system control features.
116Unfortunately it is not possible to distinguish between the package
117versions on register level so these additional voltage inputs may read
118zero. The EMC6D102 features addtional ADC bits thus extending precision
119of voltage and temperature channels.
120
121
122Hardware Configurations
123-----------------------
124
125The LM85 can be jumpered for 3 different SMBus addresses. There are
126no other hardware configuration options for the LM85.
127
128The lm85 driver detects both LM85B and LM85C revisions of the chip. See the
129datasheet for a complete description of the differences. Other than
130identifying the chip, the driver behaves no differently with regard to
131these two chips. The LM85B is recommended for new designs.
132
133The ADM1027 and ADT7463 chips have an optional SMBALERT output that can be
134used to signal the chipset in case a limit is exceeded or the temperature
135sensors fail. Individual sensor interrupts can be masked so they won't
136trigger SMBALERT. The SMBALERT output if configured replaces one of the other
137functions (PWM2 or IN0). This functionality is not implemented in current
138driver.
139
140The ADT7463 also has an optional THERM output/input which can be connected
141to the processor PROC_HOT output. If available, the autofan control
142dynamic Tmin feature can be enabled to keep the system temperature within
143spec (just?!) with the least possible fan noise.
144
145Configuration Notes
146-------------------
147
148Besides standard interfaces driver adds following:
149
150* Temperatures and Zones
151
152Each temperature sensor is associated with a Zone. There are three
153sensors and therefore three zones (# 1, 2 and 3). Each zone has the following
154temperature configuration points:
155
156* temp#_auto_temp_off - temperature below which fans should be off or spinning very low.
157* temp#_auto_temp_min - temperature over which fans start to spin.
158* temp#_auto_temp_max - temperature when fans spin at full speed.
159* temp#_auto_temp_crit - temperature when all fans will run full speed.
160
161* PWM Control
162
163There are three PWM outputs. The LM85 datasheet suggests that the
164pwm3 output control both fan3 and fan4. Each PWM can be individually
165configured and assigned to a zone for it's control value. Each PWM can be
166configured individually according to the following options.
167
168* pwm#_auto_pwm_min - this specifies the PWM value for temp#_auto_temp_off
169 temperature. (PWM value from 0 to 255)
170
171* pwm#_auto_pwm_freq - select base frequency of PWM output. You can select
172 in range of 10.0 to 94.0 Hz in .1 Hz units.
173 (Values 100 to 940).
174
175The pwm#_auto_pwm_freq can be set to one of the following 8 values. Setting the
176frequency to a value not on this list, will result in the next higher frequency
177being selected. The actual device frequency may vary slightly from this
178specification as designed by the manufacturer. Consult the datasheet for more
179details. (PWM Frequency values: 100, 150, 230, 300, 380, 470, 620, 940)
180
181* pwm#_auto_pwm_minctl - this flags selects for temp#_auto_temp_off temperature
182 the bahaviour of fans. Write 1 to let fans spinning at
183 pwm#_auto_pwm_min or write 0 to let them off.
184
185NOTE: It has been reported that there is a bug in the LM85 that causes the flag
186to be associated with the zones not the PWMs. This contradicts all the
187published documentation. Setting pwm#_min_ctl in this case actually affects all
188PWMs controlled by zone '#'.
189
190* PWM Controlling Zone selection
191
192* pwm#_auto_channels - controls zone that is associated with PWM
193
194Configuration choices:
195
196 Value Meaning
197 ------ ------------------------------------------------
198 1 Controlled by Zone 1
199 2 Controlled by Zone 2
200 3 Controlled by Zone 3
201 23 Controlled by higher temp of Zone 2 or 3
202 123 Controlled by highest temp of Zone 1, 2 or 3
203 0 PWM always 0% (off)
204 -1 PWM always 100% (full on)
205 -2 Manual control (write to 'pwm#' to set)
206
207The National LM85's have two vendor specific configuration
208features. Tach. mode and Spinup Control. For more details on these,
209see the LM85 datasheet or Application Note AN-1260.
210
211The Analog Devices ADM1027 has several vendor specific enhancements.
212The number of pulses-per-rev of the fans can be set, Tach monitoring
213can be optimized for PWM operation, and an offset can be applied to
214the temperatures to compensate for systemic errors in the
215measurements.
216
217In addition to the ADM1027 features, the ADT7463 also has Tmin control
218and THERM asserted counts. Automatic Tmin control acts to adjust the
219Tmin value to maintain the measured temperature sensor at a specified
220temperature. There isn't much documentation on this feature in the
221ADT7463 data sheet. This is not supported by current driver.
diff --git a/Documentation/i2c/chips/lm87 b/Documentation/i2c/chips/lm87
new file mode 100644
index 000000000000..c952c57f0e11
--- /dev/null
+++ b/Documentation/i2c/chips/lm87
@@ -0,0 +1,73 @@
1Kernel driver lm87
2==================
3
4Supported chips:
5 * National Semiconductor LM87
6 Prefix: 'lm87'
7 Addresses scanned: I2C 0x2c - 0x2f
8 Datasheet: http://www.national.com/pf/LM/LM87.html
9
10Authors:
11 Frodo Looijaard <frodol@dds.nl>,
12 Philip Edelbrock <phil@netroedge.com>,
13 Mark Studebaker <mdsxyz123@yahoo.com>,
14 Stephen Rousset <stephen.rousset@rocketlogix.com>,
15 Dan Eaton <dan.eaton@rocketlogix.com>,
16 Jean Delvare <khali@linux-fr.org>,
17 Original 2.6 port Jeff Oliver
18
19Description
20-----------
21
22This driver implements support for the National Semiconductor LM87.
23
24The LM87 implements up to three temperature sensors, up to two fan
25rotation speed sensors, up to seven voltage sensors, alarms, and some
26miscellaneous stuff.
27
28Temperatures are measured in degrees Celsius. Each input has a high
29and low alarm settings. A high limit produces an alarm when the value
30goes above it, and an alarm is also produced when the value goes below
31the low limit.
32
33Fan rotation speeds are reported in RPM (rotations per minute). An alarm is
34triggered if the rotation speed has dropped below a programmable limit. Fan
35readings can be divided by a programmable divider (1, 2, 4 or 8) to give
36the readings more range or accuracy. Not all RPM values can accurately be
37represented, so some rounding is done. With a divider of 2, the lowest
38representable value is around 2600 RPM.
39
40Voltage sensors (also known as IN sensors) report their values in
41volts. An alarm is triggered if the voltage has crossed a programmable
42minimum or maximum limit. Note that minimum in this case always means
43'closest to zero'; this is important for negative voltage measurements.
44
45If an alarm triggers, it will remain triggered until the hardware register
46is read at least once. This means that the cause for the alarm may
47already have disappeared! Note that in the current implementation, all
48hardware registers are read whenever any data is read (unless it is less
49than 1.0 seconds since the last update). This means that you can easily
50miss once-only alarms.
51
52The lm87 driver only updates its values each 1.0 seconds; reading it more
53often will do no harm, but will return 'old' values.
54
55
56Hardware Configurations
57-----------------------
58
59The LM87 has four pins which can serve one of two possible functions,
60depending on the hardware configuration.
61
62Some functions share pins, so not all functions are available at the same
63time. Which are depends on the hardware setup. This driver assumes that
64the BIOS configured the chip correctly. In that respect, it differs from
65the original driver (from lm_sensors for Linux 2.4), which would force the
66LM87 to an arbitrary, compile-time chosen mode, regardless of the actual
67chipset wiring.
68
69For reference, here is the list of exclusive functions:
70 - in0+in5 (default) or temp3
71 - fan1 (default) or in6
72 - fan2 (default) or in7
73 - VID lines (default) or IRQ lines (not handled by this driver)
diff --git a/Documentation/i2c/chips/lm90 b/Documentation/i2c/chips/lm90
new file mode 100644
index 000000000000..2c4cf39471f4
--- /dev/null
+++ b/Documentation/i2c/chips/lm90
@@ -0,0 +1,121 @@
1Kernel driver lm90
2==================
3
4Supported chips:
5 * National Semiconductor LM90
6 Prefix: 'lm90'
7 Addresses scanned: I2C 0x4c
8 Datasheet: Publicly available at the National Semiconductor website
9 http://www.national.com/pf/LM/LM90.html
10 * National Semiconductor LM89
11 Prefix: 'lm99'
12 Addresses scanned: I2C 0x4c and 0x4d
13 Datasheet: Publicly available at the National Semiconductor website
14 http://www.national.com/pf/LM/LM89.html
15 * National Semiconductor LM99
16 Prefix: 'lm99'
17 Addresses scanned: I2C 0x4c and 0x4d
18 Datasheet: Publicly available at the National Semiconductor website
19 http://www.national.com/pf/LM/LM99.html
20 * National Semiconductor LM86
21 Prefix: 'lm86'
22 Addresses scanned: I2C 0x4c
23 Datasheet: Publicly available at the National Semiconductor website
24 http://www.national.com/pf/LM/LM86.html
25 * Analog Devices ADM1032
26 Prefix: 'adm1032'
27 Addresses scanned: I2C 0x4c
28 Datasheet: Publicly available at the Analog Devices website
29 http://products.analog.com/products/info.asp?product=ADM1032
30 * Analog Devices ADT7461
31 Prefix: 'adt7461'
32 Addresses scanned: I2C 0x4c
33 Datasheet: Publicly available at the Analog Devices website
34 http://products.analog.com/products/info.asp?product=ADT7461
35 Note: Only if in ADM1032 compatibility mode
36 * Maxim MAX6657
37 Prefix: 'max6657'
38 Addresses scanned: I2C 0x4c
39 Datasheet: Publicly available at the Maxim website
40 http://www.maxim-ic.com/quick_view2.cfm/qv_pk/2578
41 * Maxim MAX6658
42 Prefix: 'max6657'
43 Addresses scanned: I2C 0x4c
44 Datasheet: Publicly available at the Maxim website
45 http://www.maxim-ic.com/quick_view2.cfm/qv_pk/2578
46 * Maxim MAX6659
47 Prefix: 'max6657'
48 Addresses scanned: I2C 0x4c, 0x4d (unsupported 0x4e)
49 Datasheet: Publicly available at the Maxim website
50 http://www.maxim-ic.com/quick_view2.cfm/qv_pk/2578
51
52
53Author: Jean Delvare <khali@linux-fr.org>
54
55
56Description
57-----------
58
59The LM90 is a digital temperature sensor. It senses its own temperature as
60well as the temperature of up to one external diode. It is compatible
61with many other devices such as the LM86, the LM89, the LM99, the ADM1032,
62the MAX6657, MAX6658 and the MAX6659 all of which are supported by this driver.
63Note that there is no easy way to differentiate between the last three
64variants. The extra address and features of the MAX6659 are not supported by
65this driver. Additionally, the ADT7461 is supported if found in ADM1032
66compatibility mode.
67
68The specificity of this family of chipsets over the ADM1021/LM84
69family is that it features critical limits with hysteresis, and an
70increased resolution of the remote temperature measurement.
71
72The different chipsets of the family are not strictly identical, although
73very similar. This driver doesn't handle any specific feature for now,
74but could if there ever was a need for it. For reference, here comes a
75non-exhaustive list of specific features:
76
77LM90:
78 * Filter and alert configuration register at 0xBF.
79 * ALERT is triggered by temperatures over critical limits.
80
81LM86 and LM89:
82 * Same as LM90
83 * Better external channel accuracy
84
85LM99:
86 * Same as LM89
87 * External temperature shifted by 16 degrees down
88
89ADM1032:
90 * Consecutive alert register at 0x22.
91 * Conversion averaging.
92 * Up to 64 conversions/s.
93 * ALERT is triggered by open remote sensor.
94
95ADT7461
96 * Extended temperature range (breaks compatibility)
97 * Lower resolution for remote temperature
98
99MAX6657 and MAX6658:
100 * Remote sensor type selection
101
102MAX6659
103 * Selectable address
104 * Second critical temperature limit
105 * Remote sensor type selection
106
107All temperature values are given in degrees Celsius. Resolution
108is 1.0 degree for the local temperature, 0.125 degree for the remote
109temperature.
110
111Each sensor has its own high and low limits, plus a critical limit.
112Additionally, there is a relative hysteresis value common to both critical
113values. To make life easier to user-space applications, two absolute values
114are exported, one for each channel, but these values are of course linked.
115Only the local hysteresis can be set from user-space, and the same delta
116applies to the remote hysteresis.
117
118The lm90 driver will not update its values more frequently than every
119other second; reading them more often will do no harm, but will return
120'old' values.
121
diff --git a/Documentation/i2c/chips/lm92 b/Documentation/i2c/chips/lm92
new file mode 100644
index 000000000000..7705bfaa0708
--- /dev/null
+++ b/Documentation/i2c/chips/lm92
@@ -0,0 +1,37 @@
1Kernel driver lm92
2==================
3
4Supported chips:
5 * National Semiconductor LM92
6 Prefix: 'lm92'
7 Addresses scanned: I2C 0x48 - 0x4b
8 Datasheet: http://www.national.com/pf/LM/LM92.html
9 * National Semiconductor LM76
10 Prefix: 'lm92'
11 Addresses scanned: none, force parameter needed
12 Datasheet: http://www.national.com/pf/LM/LM76.html
13 * Maxim MAX6633/MAX6634/MAX6635
14 Prefix: 'lm92'
15 Addresses scanned: I2C 0x48 - 0x4b
16 MAX6633 with address in 0x40 - 0x47, 0x4c - 0x4f needs force parameter
17 and MAX6634 with address in 0x4c - 0x4f needs force parameter
18 Datasheet: http://www.maxim-ic.com/quick_view2.cfm/qv_pk/3074
19
20Authors:
21 Abraham van der Merwe <abraham@2d3d.co.za>
22 Jean Delvare <khali@linux-fr.org>
23
24
25Description
26-----------
27
28This driver implements support for the National Semiconductor LM92
29temperature sensor.
30
31Each LM92 temperature sensor supports a single temperature sensor. There are
32alarms for high, low, and critical thresholds. There's also an hysteresis to
33control the thresholds for resetting alarms.
34
35Support was added later for the LM76 and Maxim MAX6633/MAX6634/MAX6635,
36which are mostly compatible. They have not all been tested, so you
37may need to use the force parameter.
diff --git a/Documentation/i2c/chips/max1619 b/Documentation/i2c/chips/max1619
new file mode 100644
index 000000000000..d6f8d9cd7d7f
--- /dev/null
+++ b/Documentation/i2c/chips/max1619
@@ -0,0 +1,29 @@
1Kernel driver max1619
2=====================
3
4Supported chips:
5 * Maxim MAX1619
6 Prefix: 'max1619'
7 Addresses scanned: I2C 0x18-0x1a, 0x29-0x2b, 0x4c-0x4e
8 Datasheet: Publicly available at the Maxim website
9 http://pdfserv.maxim-ic.com/en/ds/MAX1619.pdf
10
11Authors:
12 Alexey Fisher <fishor@mail.ru>,
13 Jean Delvare <khali@linux-fr.org>
14
15Description
16-----------
17
18The MAX1619 is a digital temperature sensor. It senses its own temperature as
19well as the temperature of up to one external diode.
20
21All temperature values are given in degrees Celsius. Resolution
22is 1.0 degree for the local temperature and for the remote temperature.
23
24Only the external sensor has high and low limits.
25
26The max1619 driver will not update its values more frequently than every
27other second; reading them more often will do no harm, but will return
28'old' values.
29
diff --git a/Documentation/i2c/chips/pc87360 b/Documentation/i2c/chips/pc87360
new file mode 100644
index 000000000000..89a8fcfa78df
--- /dev/null
+++ b/Documentation/i2c/chips/pc87360
@@ -0,0 +1,189 @@
1Kernel driver pc87360
2=====================
3
4Supported chips:
5 * National Semiconductor PC87360, PC87363, PC87364, PC87365 and PC87366
6 Prefixes: 'pc87360', 'pc87363', 'pc87364', 'pc87365', 'pc87366'
7 Addresses scanned: none, address read from Super I/O config space
8 Datasheets:
9 http://www.national.com/pf/PC/PC87360.html
10 http://www.national.com/pf/PC/PC87363.html
11 http://www.national.com/pf/PC/PC87364.html
12 http://www.national.com/pf/PC/PC87365.html
13 http://www.national.com/pf/PC/PC87366.html
14
15Authors: Jean Delvare <khali@linux-fr.org>
16
17Thanks to Sandeep Mehta, Tonko de Rooy and Daniel Ceregatti for testing.
18Thanks to Rudolf Marek for helping me investigate conversion issues.
19
20
21Module Parameters
22-----------------
23
24* init int
25 Chip initialization level:
26 0: None
27 *1: Forcibly enable internal voltage and temperature channels, except in9
28 2: Forcibly enable all voltage and temperature channels, except in9
29 3: Forcibly enable all voltage and temperature channels, including in9
30
31Note that this parameter has no effect for the PC87360, PC87363 and PC87364
32chips.
33
34Also note that for the PC87366, initialization levels 2 and 3 don't enable
35all temperature channels, because some of them share pins with each other,
36so they can't be used at the same time.
37
38
39Description
40-----------
41
42The National Semiconductor PC87360 Super I/O chip contains monitoring and
43PWM control circuitry for two fans. The PC87363 chip is similar, and the
44PC87364 chip has monitoring and PWM control for a third fan.
45
46The National Semiconductor PC87365 and PC87366 Super I/O chips are complete
47hardware monitoring chipsets, not only controlling and monitoring three fans,
48but also monitoring eleven voltage inputs and two (PC87365) or up to four
49(PC87366) temperatures.
50
51 Chip #vin #fan #pwm #temp devid
52
53 PC87360 - 2 2 - 0xE1
54 PC87363 - 2 2 - 0xE8
55 PC87364 - 3 3 - 0xE4
56 PC87365 11 3 3 2 0xE5
57 PC87366 11 3 3 3-4 0xE9
58
59The driver assumes that no more than one chip is present, and one of the
60standard Super I/O addresses is used (0x2E/0x2F or 0x4E/0x4F)
61
62Fan Monitoring
63--------------
64
65Fan rotation speeds are reported in RPM (revolutions per minute). An alarm
66is triggered if the rotation speed has dropped below a programmable limit.
67A different alarm is triggered if the fan speed is too low to be measured.
68
69Fan readings are affected by a programmable clock divider, giving the
70readings more range or accuracy. Usually, users have to learn how it works,
71but this driver implements dynamic clock divider selection, so you don't
72have to care no more.
73
74For reference, here are a few values about clock dividers:
75
76 slowest accuracy highest
77 measurable around 3000 accurate
78 divider speed (RPM) RPM (RPM) speed (RPM)
79 1 1882 18 6928
80 2 941 37 4898
81 4 470 74 3464
82 8 235 150 2449
83
84For the curious, here is how the values above were computed:
85 * slowest measurable speed: clock/(255*divider)
86 * accuracy around 3000 RPM: 3000^2/clock
87 * highest accurate speed: sqrt(clock*100)
88The clock speed for the PC87360 family is 480 kHz. I arbitrarily chose 100
89RPM as the lowest acceptable accuracy.
90
91As mentioned above, you don't have to care about this no more.
92
93Note that not all RPM values can be represented, even when the best clock
94divider is selected. This is not only true for the measured speeds, but
95also for the programmable low limits, so don't be surprised if you try to
96set, say, fan1_min to 2900 and it finally reads 2909.
97
98
99Fan Control
100-----------
101
102PWM (pulse width modulation) values range from 0 to 255, with 0 meaning
103that the fan is stopped, and 255 meaning that the fan goes at full speed.
104
105Be extremely careful when changing PWM values. Low PWM values, even
106non-zero, can stop the fan, which may cause irreversible damage to your
107hardware if temperature increases too much. When changing PWM values, go
108step by step and keep an eye on temperatures.
109
110One user reported problems with PWM. Changing PWM values would break fan
111speed readings. No explanation nor fix could be found.
112
113
114Temperature Monitoring
115----------------------
116
117Temperatures are reported in degrees Celsius. Each temperature measured has
118associated low, high and overtemperature limits, each of which triggers an
119alarm when crossed.
120
121The first two temperature channels are external. The third one (PC87366
122only) is internal.
123
124The PC87366 has three additional temperature channels, based on
125thermistors (as opposed to thermal diodes for the first three temperature
126channels). For technical reasons, these channels are held by the VLM
127(voltage level monitor) logical device, not the TMS (temperature
128measurement) one. As a consequence, these temperatures are exported as
129voltages, and converted into temperatures in user-space.
130
131Note that these three additional channels share their pins with the
132external thermal diode channels, so you (physically) can't use them all at
133the same time. Although it should be possible to mix the two sensor types,
134the documents from National Semiconductor suggest that motherboard
135manufacturers should choose one type and stick to it. So you will more
136likely have either channels 1 to 3 (thermal diodes) or 3 to 6 (internal
137thermal diode, and thermistors).
138
139
140Voltage Monitoring
141------------------
142
143Voltages are reported relatively to a reference voltage, either internal or
144external. Some of them (in7:Vsb, in8:Vdd and in10:AVdd) are divided by two
145internally, you will have to compensate in sensors.conf. Others (in0 to in6)
146are likely to be divided externally. The meaning of each of these inputs as
147well as the values of the resistors used for division is left to the
148motherboard manufacturers, so you will have to document yourself and edit
149sensors.conf accordingly. National Semiconductor has a document with
150recommended resistor values for some voltages, but this still leaves much
151room for per motherboard specificities, unfortunately. Even worse,
152motherboard manufacturers don't seem to care about National Semiconductor's
153recommendations.
154
155Each voltage measured has associated low and high limits, each of which
156triggers an alarm when crossed.
157
158When available, VID inputs are used to provide the nominal CPU Core voltage.
159The driver will default to VRM 9.0, but this can be changed from user-space.
160The chipsets can handle two sets of VID inputs (on dual-CPU systems), but
161the driver will only export one for now. This may change later if there is
162a need.
163
164
165General Remarks
166---------------
167
168If an alarm triggers, it will remain triggered until the hardware register
169is read at least once. This means that the cause for the alarm may already
170have disappeared! Note that all hardware registers are read whenever any
171data is read (unless it is less than 2 seconds since the last update, in
172which case cached values are returned instead). As a consequence, when
173a once-only alarm triggers, it may take 2 seconds for it to show, and 2
174more seconds for it to disappear.
175
176Monitoring of in9 isn't enabled at lower init levels (<3) because that
177channel measures the battery voltage (Vbat). It is a known fact that
178repeatedly sampling the battery voltage reduces its lifetime. National
179Semiconductor smartly designed their chipset so that in9 is sampled only
180once every 1024 sampling cycles (that is every 34 minutes at the default
181sampling rate), so the effect is attenuated, but still present.
182
183
184Limitations
185-----------
186
187The datasheets suggests that some values (fan mins, fan dividers)
188shouldn't be changed once the monitoring has started, but we ignore that
189recommendation. We'll reconsider if it actually causes trouble.
diff --git a/Documentation/i2c/chips/pcf8574 b/Documentation/i2c/chips/pcf8574
new file mode 100644
index 000000000000..2752c8ce3167
--- /dev/null
+++ b/Documentation/i2c/chips/pcf8574
@@ -0,0 +1,69 @@
1Kernel driver pcf8574
2=====================
3
4Supported chips:
5 * Philips PCF8574
6 Prefix: 'pcf8574'
7 Addresses scanned: I2C 0x20 - 0x27
8 Datasheet: Publicly available at the Philips Semiconductors website
9 http://www.semiconductors.philips.com/pip/PCF8574P.html
10
11 * Philips PCF8574A
12 Prefix: 'pcf8574a'
13 Addresses scanned: I2C 0x38 - 0x3f
14 Datasheet: Publicly available at the Philips Semiconductors website
15 http://www.semiconductors.philips.com/pip/PCF8574P.html
16
17Authors:
18 Frodo Looijaard <frodol@dds.nl>,
19 Philip Edelbrock <phil@netroedge.com>,
20 Dan Eaton <dan.eaton@rocketlogix.com>,
21 Aurelien Jarno <aurelien@aurel32.net>,
22 Jean Delvare <khali@linux-fr.org>,
23
24
25Description
26-----------
27The PCF8574(A) is an 8-bit I/O expander for the I2C bus produced by Philips
28Semiconductors. It is designed to provide a byte I2C interface to up to 16
29separate devices (8 x PCF8574 and 8 x PCF8574A).
30
31This device consists of a quasi-bidirectional port. Each of the eight I/Os
32can be independently used as an input or output. To setup an I/O as an
33input, you have to write a 1 to the corresponding output.
34
35For more informations see the datasheet.
36
37
38Accessing PCF8574(A) via /sys interface
39-------------------------------------
40
41! Be careful !
42The PCF8574(A) is plainly impossible to detect ! Stupid chip.
43So every chip with address in the interval [20..27] and [38..3f] are
44detected as PCF8574(A). If you have other chips in this address
45range, the workaround is to load this module after the one
46for your others chips.
47
48On detection (i.e. insmod, modprobe et al.), directories are being
49created for each detected PCF8574(A):
50
51/sys/bus/i2c/devices/<0>-<1>/
52where <0> is the bus the chip was detected on (e. g. i2c-0)
53and <1> the chip address ([20..27] or [38..3f]):
54
55(example: /sys/bus/i2c/devices/1-0020/)
56
57Inside these directories, there are two files each:
58read and write (and one file with chip name).
59
60The read file is read-only. Reading gives you the current I/O input
61if the corresponding output is set as 1, otherwise the current output
62value, that is to say 0.
63
64The write file is read/write. Writing a value outputs it on the I/O
65port. Reading returns the last written value.
66
67On module initialization the chip is configured as eight inputs (all
68outputs to 1), so you can connect any circuit to the PCF8574(A) without
69being afraid of short-circuit.
diff --git a/Documentation/i2c/chips/pcf8591 b/Documentation/i2c/chips/pcf8591
new file mode 100644
index 000000000000..5628fcf4207f
--- /dev/null
+++ b/Documentation/i2c/chips/pcf8591
@@ -0,0 +1,90 @@
1Kernel driver pcf8591
2=====================
3
4Supported chips:
5 * Philips PCF8591
6 Prefix: 'pcf8591'
7 Addresses scanned: I2C 0x48 - 0x4f
8 Datasheet: Publicly available at the Philips Semiconductor website
9 http://www.semiconductors.philips.com/pip/PCF8591P.html
10
11Authors:
12 Aurelien Jarno <aurelien@aurel32.net>
13 valuable contributions by Jan M. Sendler <sendler@sendler.de>,
14 Jean Delvare <khali@linux-fr.org>
15
16
17Description
18-----------
19The PCF8591 is an 8-bit A/D and D/A converter (4 analog inputs and one
20analog output) for the I2C bus produced by Philips Semiconductors. It
21is designed to provide a byte I2C interface to up to 4 separate devices.
22
23The PCF8591 has 4 analog inputs programmable as single-ended or
24differential inputs :
25- mode 0 : four single ended inputs
26 Pins AIN0 to AIN3 are single ended inputs for channels 0 to 3
27
28- mode 1 : three differential inputs
29 Pins AIN3 is the common negative differential input
30 Pins AIN0 to AIN2 are positive differential inputs for channels 0 to 2
31
32- mode 2 : single ended and differential mixed
33 Pins AIN0 and AIN1 are single ended inputs for channels 0 and 1
34 Pins AIN2 is the positive differential input for channel 3
35 Pins AIN3 is the negative differential input for channel 3
36
37- mode 3 : two differential inputs
38 Pins AIN0 is the positive differential input for channel 0
39 Pins AIN1 is the negative differential input for channel 0
40 Pins AIN2 is the positive differential input for channel 1
41 Pins AIN3 is the negative differential input for channel 1
42
43See the datasheet for details.
44
45Module parameters
46-----------------
47
48* input_mode int
49
50 Analog input mode:
51 0 = four single ended inputs
52 1 = three differential inputs
53 2 = single ended and differential mixed
54 3 = two differential inputs
55
56
57Accessing PCF8591 via /sys interface
58-------------------------------------
59
60! Be careful !
61The PCF8591 is plainly impossible to detect ! Stupid chip.
62So every chip with address in the interval [48..4f] is
63detected as PCF8591. If you have other chips in this address
64range, the workaround is to load this module after the one
65for your others chips.
66
67On detection (i.e. insmod, modprobe et al.), directories are being
68created for each detected PCF8591:
69
70/sys/bus/devices/<0>-<1>/
71where <0> is the bus the chip was detected on (e. g. i2c-0)
72and <1> the chip address ([48..4f])
73
74Inside these directories, there are such files:
75in0, in1, in2, in3, out0_enable, out0_output, name
76
77Name contains chip name.
78
79The in0, in1, in2 and in3 files are RO. Reading gives the value of the
80corresponding channel. Depending on the current analog inputs configuration,
81files in2 and/or in3 do not exist. Values range are from 0 to 255 for single
82ended inputs and -128 to +127 for differential inputs (8-bit ADC).
83
84The out0_enable file is RW. Reading gives "1" for analog output enabled and
85"0" for analog output disabled. Writing accepts "0" and "1" accordingly.
86
87The out0_output file is RW. Writing a number between 0 and 255 (8-bit DAC), send
88the value to the digital-to-analog converter. Note that a voltage will
89only appears on AOUT pin if aout0_enable equals 1. Reading returns the last
90value written.
diff --git a/Documentation/i2c/chips/sis5595 b/Documentation/i2c/chips/sis5595
new file mode 100644
index 000000000000..b7ae36b8cdf5
--- /dev/null
+++ b/Documentation/i2c/chips/sis5595
@@ -0,0 +1,106 @@
1Kernel driver sis5595
2=====================
3
4Supported chips:
5 * Silicon Integrated Systems Corp. SiS5595 Southbridge Hardware Monitor
6 Prefix: 'sis5595'
7 Addresses scanned: ISA in PCI-space encoded address
8 Datasheet: Publicly available at the Silicon Integrated Systems Corp. site.
9
10Authors:
11 Kyösti Mälkki <kmalkki@cc.hut.fi>,
12 Mark D. Studebaker <mdsxyz123@yahoo.com>,
13 Aurelien Jarno <aurelien@aurel32.net> 2.6 port
14
15 SiS southbridge has a LM78-like chip integrated on the same IC.
16 This driver is a customized copy of lm78.c
17
18 Supports following revisions:
19 Version PCI ID PCI Revision
20 1 1039/0008 AF or less
21 2 1039/0008 B0 or greater
22
23 Note: these chips contain a 0008 device which is incompatible with the
24 5595. We recognize these by the presence of the listed
25 "blacklist" PCI ID and refuse to load.
26
27 NOT SUPPORTED PCI ID BLACKLIST PCI ID
28 540 0008 0540
29 550 0008 0550
30 5513 0008 5511
31 5581 0008 5597
32 5582 0008 5597
33 5597 0008 5597
34 630 0008 0630
35 645 0008 0645
36 730 0008 0730
37 735 0008 0735
38
39
40Module Parameters
41-----------------
42force_addr=0xaddr Set the I/O base address. Useful for boards
43 that don't set the address in the BIOS. Does not do a
44 PCI force; the device must still be present in lspci.
45 Don't use this unless the driver complains that the
46 base address is not set.
47 Example: 'modprobe sis5595 force_addr=0x290'
48
49
50Description
51-----------
52
53The SiS5595 southbridge has integrated hardware monitor functions. It also
54has an I2C bus, but this driver only supports the hardware monitor. For the
55I2C bus driver see i2c-sis5595.
56
57The SiS5595 implements zero or one temperature sensor, two fan speed
58sensors, four or five voltage sensors, and alarms.
59
60On the first version of the chip, there are four voltage sensors and one
61temperature sensor.
62
63On the second version of the chip, the temperature sensor (temp) and the
64fifth voltage sensor (in4) share a pin which is configurable, but not
65through the driver. Sorry. The driver senses the configuration of the pin,
66which was hopefully set by the BIOS.
67
68Temperatures are measured in degrees Celsius. An alarm is triggered once
69when the max is crossed; it is also triggered when it drops below the min
70value. Measurements are guaranteed between -55 and +125 degrees, with a
71resolution of 1 degree.
72
73Fan rotation speeds are reported in RPM (rotations per minute). An alarm is
74triggered if the rotation speed has dropped below a programmable limit. Fan
75readings can be divided by a programmable divider (1, 2, 4 or 8) to give
76the readings more range or accuracy. Not all RPM values can accurately be
77represented, so some rounding is done. With a divider of 2, the lowest
78representable value is around 2600 RPM.
79
80Voltage sensors (also known as IN sensors) report their values in volts. An
81alarm is triggered if the voltage has crossed a programmable minimum or
82maximum limit. Note that minimum in this case always means 'closest to
83zero'; this is important for negative voltage measurements. All voltage
84inputs can measure voltages between 0 and 4.08 volts, with a resolution of
850.016 volt.
86
87In addition to the alarms described above, there is a BTI alarm, which gets
88triggered when an external chip has crossed its limits. Usually, this is
89connected to some LM75-like chip; if at least one crosses its limits, this
90bit gets set.
91
92If an alarm triggers, it will remain triggered until the hardware register
93is read at least once. This means that the cause for the alarm may already
94have disappeared! Note that in the current implementation, all hardware
95registers are read whenever any data is read (unless it is less than 1.5
96seconds since the last update). This means that you can easily miss
97once-only alarms.
98
99The SiS5595 only updates its values each 1.5 seconds; reading it more often
100will do no harm, but will return 'old' values.
101
102Problems
103--------
104Some chips refuse to be enabled. We don't know why.
105The driver will recognize this and print a message in dmesg.
106
diff --git a/Documentation/i2c/chips/smsc47b397 b/Documentation/i2c/chips/smsc47b397
index 4737bb6109da..da9d80c96432 100644
--- a/Documentation/i2c/chips/smsc47b397
+++ b/Documentation/i2c/chips/smsc47b397
@@ -1,7 +1,19 @@
1Kernel driver smsc47b397
2========================
3
4Supported chips:
5 * SMSC LPC47B397-NC
6 Prefix: 'smsc47b397'
7 Addresses scanned: none, address read from Super I/O config space
8 Datasheet: In this file
9
10Authors: Mark M. Hoffman <mhoffman@lightlink.com>
11 Utilitek Systems, Inc.
12
1November 23, 2004 13November 23, 2004
2 14
3The following specification describes the SMSC LPC47B397-NC sensor chip 15The following specification describes the SMSC LPC47B397-NC sensor chip
4(for which there is no public datasheet available). This document was 16(for which there is no public datasheet available). This document was
5provided by Craig Kelly (In-Store Broadcast Network) and edited/corrected 17provided by Craig Kelly (In-Store Broadcast Network) and edited/corrected
6by Mark M. Hoffman <mhoffman@lightlink.com>. 18by Mark M. Hoffman <mhoffman@lightlink.com>.
7 19
@@ -10,10 +22,10 @@ by Mark M. Hoffman <mhoffman@lightlink.com>.
10Methods for detecting the HP SIO and reading the thermal data on a dc7100. 22Methods for detecting the HP SIO and reading the thermal data on a dc7100.
11 23
12The thermal information on the dc7100 is contained in the SIO Hardware Monitor 24The thermal information on the dc7100 is contained in the SIO Hardware Monitor
13(HWM). The information is accessed through an index/data pair. The index/data 25(HWM). The information is accessed through an index/data pair. The index/data
14pair is located at the HWM Base Address + 0 and the HWM Base Address + 1. The 26pair is located at the HWM Base Address + 0 and the HWM Base Address + 1. The
15HWM Base address can be obtained from Logical Device 8, registers 0x60 (MSB) 27HWM Base address can be obtained from Logical Device 8, registers 0x60 (MSB)
16and 0x61 (LSB). Currently we are using 0x480 for the HWM Base Address and 28and 0x61 (LSB). Currently we are using 0x480 for the HWM Base Address and
170x480 and 0x481 for the index/data pair. 290x480 and 0x481 for the index/data pair.
18 30
19Reading temperature information. 31Reading temperature information.
@@ -50,7 +62,7 @@ Reading the tach LSB locks the tach MSB.
50The LSB Must be read first. 62The LSB Must be read first.
51 63
52How to convert the tach reading to RPM. 64How to convert the tach reading to RPM.
53The tach reading (TCount) is given by: (Tach MSB * 256) + (Tach LSB) 65The tach reading (TCount) is given by: (Tach MSB * 256) + (Tach LSB)
54The SIO counts the number of 90kHz (11.111us) pulses per revolution. 66The SIO counts the number of 90kHz (11.111us) pulses per revolution.
55RPM = 60/(TCount * 11.111us) 67RPM = 60/(TCount * 11.111us)
56 68
diff --git a/Documentation/i2c/chips/smsc47m1 b/Documentation/i2c/chips/smsc47m1
new file mode 100644
index 000000000000..34e6478c1425
--- /dev/null
+++ b/Documentation/i2c/chips/smsc47m1
@@ -0,0 +1,52 @@
1Kernel driver smsc47m1
2======================
3
4Supported chips:
5 * SMSC LPC47B27x, LPC47M10x, LPC47M13x, LPC47M14x, LPC47M15x and LPC47M192
6 Addresses scanned: none, address read from Super I/O config space
7 Prefix: 'smsc47m1'
8 Datasheets:
9 http://www.smsc.com/main/datasheets/47b27x.pdf
10 http://www.smsc.com/main/datasheets/47m10x.pdf
11 http://www.smsc.com/main/tools/discontinued/47m13x.pdf
12 http://www.smsc.com/main/datasheets/47m14x.pdf
13 http://www.smsc.com/main/tools/discontinued/47m15x.pdf
14 http://www.smsc.com/main/datasheets/47m192.pdf
15
16Authors:
17 Mark D. Studebaker <mdsxyz123@yahoo.com>,
18 With assistance from Bruce Allen <ballen@uwm.edu>, and his
19 fan.c program: http://www.lsc-group.phys.uwm.edu/%7Eballen/driver/
20 Gabriele Gorla <gorlik@yahoo.com>,
21 Jean Delvare <khali@linux-fr.org>
22
23Description
24-----------
25
26The Standard Microsystems Corporation (SMSC) 47M1xx Super I/O chips
27contain monitoring and PWM control circuitry for two fans.
28
29The 47M15x and 47M192 chips contain a full 'hardware monitoring block'
30in addition to the fan monitoring and control. The hardware monitoring
31block is not supported by the driver.
32
33Fan rotation speeds are reported in RPM (rotations per minute). An alarm is
34triggered if the rotation speed has dropped below a programmable limit. Fan
35readings can be divided by a programmable divider (1, 2, 4 or 8) to give
36the readings more range or accuracy. Not all RPM values can accurately be
37represented, so some rounding is done. With a divider of 2, the lowest
38representable value is around 2600 RPM.
39
40PWM values are from 0 to 255.
41
42If an alarm triggers, it will remain triggered until the hardware register
43is read at least once. This means that the cause for the alarm may
44already have disappeared! Note that in the current implementation, all
45hardware registers are read whenever any data is read (unless it is less
46than 1.5 seconds since the last update). This means that you can easily
47miss once-only alarms.
48
49
50**********************
51The lm_sensors project gratefully acknowledges the support of
52Intel in the development of this driver.
diff --git a/Documentation/i2c/chips/via686a b/Documentation/i2c/chips/via686a
new file mode 100644
index 000000000000..b82014cb7c53
--- /dev/null
+++ b/Documentation/i2c/chips/via686a
@@ -0,0 +1,65 @@
1Kernel driver via686a
2=====================
3
4Supported chips:
5 * Via VT82C686A, VT82C686B Southbridge Integrated Hardware Monitor
6 Prefix: 'via686a'
7 Addresses scanned: ISA in PCI-space encoded address
8 Datasheet: On request through web form (http://www.via.com.tw/en/support/datasheets/)
9
10Authors:
11 Kyösti Mälkki <kmalkki@cc.hut.fi>,
12 Mark D. Studebaker <mdsxyz123@yahoo.com>
13 Bob Dougherty <bobd@stanford.edu>
14 (Some conversion-factor data were contributed by
15 Jonathan Teh Soon Yew <j.teh@iname.com>
16 and Alex van Kaam <darkside@chello.nl>.)
17
18Module Parameters
19-----------------
20
21force_addr=0xaddr Set the I/O base address. Useful for Asus A7V boards
22 that don't set the address in the BIOS. Does not do a
23 PCI force; the via686a must still be present in lspci.
24 Don't use this unless the driver complains that the
25 base address is not set.
26 Example: 'modprobe via686a force_addr=0x6000'
27
28Description
29-----------
30
31The driver does not distinguish between the chips and reports
32all as a 686A.
33
34The Via 686a southbridge has integrated hardware monitor functionality.
35It also has an I2C bus, but this driver only supports the hardware monitor.
36For the I2C bus driver, see <file:Documentation/i2c/busses/i2c-viapro>
37
38The Via 686a implements three temperature sensors, two fan rotation speed
39sensors, five voltage sensors and alarms.
40
41Temperatures are measured in degrees Celsius. An alarm is triggered once
42when the Overtemperature Shutdown limit is crossed; it is triggered again
43as soon as it drops below the hysteresis value.
44
45Fan rotation speeds are reported in RPM (rotations per minute). An alarm is
46triggered if the rotation speed has dropped below a programmable limit. Fan
47readings can be divided by a programmable divider (1, 2, 4 or 8) to give
48the readings more range or accuracy. Not all RPM values can accurately be
49represented, so some rounding is done. With a divider of 2, the lowest
50representable value is around 2600 RPM.
51
52Voltage sensors (also known as IN sensors) report their values in volts.
53An alarm is triggered if the voltage has crossed a programmable minimum
54or maximum limit. Voltages are internally scalled, so each voltage channel
55has a different resolution and range.
56
57If an alarm triggers, it will remain triggered until the hardware register
58is read at least once. This means that the cause for the alarm may
59already have disappeared! Note that in the current implementation, all
60hardware registers are read whenever any data is read (unless it is less
61than 1.5 seconds since the last update). This means that you can easily
62miss once-only alarms.
63
64The driver only updates its values each 1.5 seconds; reading it more often
65will do no harm, but will return 'old' values.
diff --git a/Documentation/i2c/chips/w83627hf b/Documentation/i2c/chips/w83627hf
new file mode 100644
index 000000000000..78f37c2d602e
--- /dev/null
+++ b/Documentation/i2c/chips/w83627hf
@@ -0,0 +1,66 @@
1Kernel driver w83627hf
2======================
3
4Supported chips:
5 * Winbond W83627HF (ISA accesses ONLY)
6 Prefix: 'w83627hf'
7 Addresses scanned: ISA address retrieved from Super I/O registers
8 Datasheet: http://www.winbond.com/PDF/sheet/w83627hf.pdf
9 * Winbond W83627THF
10 Prefix: 'w83627thf'
11 Addresses scanned: ISA address retrieved from Super I/O registers
12 Datasheet: http://www.winbond.com/PDF/sheet/w83627thf.pdf
13 * Winbond W83697HF
14 Prefix: 'w83697hf'
15 Addresses scanned: ISA address retrieved from Super I/O registers
16 Datasheet: http://www.winbond.com/PDF/sheet/697hf.pdf
17 * Winbond W83637HF
18 Prefix: 'w83637hf'
19 Addresses scanned: ISA address retrieved from Super I/O registers
20 Datasheet: http://www.winbond.com/PDF/sheet/w83637hf.pdf
21
22Authors:
23 Frodo Looijaard <frodol@dds.nl>,
24 Philip Edelbrock <phil@netroedge.com>,
25 Mark Studebaker <mdsxyz123@yahoo.com>,
26 Bernhard C. Schrenk <clemy@clemy.org>
27
28Module Parameters
29-----------------
30
31* force_addr: int
32 Initialize the ISA address of the sensors
33* force_i2c: int
34 Initialize the I2C address of the sensors
35* init: int
36 (default is 1)
37 Use 'init=0' to bypass initializing the chip.
38 Try this if your computer crashes when you load the module.
39
40Description
41-----------
42
43This driver implements support for ISA accesses *only* for
44the Winbond W83627HF, W83627THF, W83697HF and W83637HF Super I/O chips.
45We will refer to them collectively as Winbond chips.
46
47This driver supports ISA accesses, which should be more reliable
48than i2c accesses. Also, for Tyan boards which contain both a
49Super I/O chip and a second i2c-only Winbond chip (often a W83782D),
50using this driver will avoid i2c address conflicts and complex
51initialization that were required in the w83781d driver.
52
53If you really want i2c accesses for these Super I/O chips,
54use the w83781d driver. However this is not the preferred method
55now that this ISA driver has been developed.
56
57Technically, the w83627thf does not support a VID reading. However, it's
58possible or even likely that your mainboard maker has routed these signals
59to a specific set of general purpose IO pins (the Asus P4C800-E is one such
60board). The w83627thf driver now interprets these as VID. If the VID on
61your board doesn't work, first see doc/vid in the lm_sensors package. If
62that still doesn't help, email us at lm-sensors@lm-sensors.org.
63
64For further information on this driver see the w83781d driver
65documentation.
66
diff --git a/Documentation/i2c/chips/w83781d b/Documentation/i2c/chips/w83781d
new file mode 100644
index 000000000000..6d0e16f0a50a
--- /dev/null
+++ b/Documentation/i2c/chips/w83781d
@@ -0,0 +1,412 @@
1Kernel driver w83781d
2=====================
3
4Supported chips:
5 * Winbond W83781D
6 Prefix: 'w83781d'
7 Addresses scanned: I2C 0x20 - 0x2f, ISA 0x290 (8 I/O ports)
8 Datasheet: http://www.winbond-usa.com/products/winbond_products/pdfs/PCIC/w83781d.pdf
9 * Winbond W83782D
10 Prefix: 'w83782d'
11 Addresses scanned: I2C 0x20 - 0x2f, ISA 0x290 (8 I/O ports)
12 Datasheet: http://www.winbond.com/PDF/sheet/w83782d.pdf
13 * Winbond W83783S
14 Prefix: 'w83783s'
15 Addresses scanned: I2C 0x2d
16 Datasheet: http://www.winbond-usa.com/products/winbond_products/pdfs/PCIC/w83783s.pdf
17 * Winbond W83627HF
18 Prefix: 'w83627hf'
19 Addresses scanned: I2C 0x20 - 0x2f, ISA 0x290 (8 I/O ports)
20 Datasheet: http://www.winbond.com/PDF/sheet/w83627hf.pdf
21 * Winbond W83627THF
22 Prefix: 'w83627thf'
23 Addresses scanned: ISA address 0x290 (8 I/O ports)
24 Datasheet: http://www.winbond.com/PDF/sheet/w83627thf.pdf
25 * Winbond W83697HF
26 Prefix: 'w83697hf'
27 Addresses scanned: ISA 0x290 (8 I/O ports)
28 Datasheet: http://www.winbond-usa.com/products/winbond_products/pdfs/PCIC/w83697hf.pdf
29 * Asus AS99127F
30 Prefix: 'as99127f'
31 Addresses scanned: I2C 0x28 - 0x2f
32 Datasheet: Unavailable from Asus
33
34Authors:
35 Frodo Looijaard <frodol@dds.nl>,
36 Philip Edelbrock <phil@netroedge.com>,
37 Mark Studebaker <mdsxyz123@yahoo.com>
38
39Module parameters
40-----------------
41
42* init int
43 (default 1)
44 Use 'init=0' to bypass initializing the chip.
45 Try this if your computer crashes when you load the module.
46
47force_subclients=bus,caddr,saddr,saddr
48 This is used to force the i2c addresses for subclients of
49 a certain chip. Typical usage is `force_subclients=0,0x2d,0x4a,0x4b'
50 to force the subclients of chip 0x2d on bus 0 to i2c addresses
51 0x4a and 0x4b. This parameter is useful for certain Tyan boards.
52
53Description
54-----------
55
56This driver implements support for the Winbond W83627HF, W83627THF, W83781D,
57W83782D, W83783S, W83697HF chips, and the Asus AS99127F chips. We will refer
58to them collectively as W8378* chips.
59
60There is quite some difference between these chips, but they are similar
61enough that it was sensible to put them together in one driver.
62The W83627HF chip is assumed to be identical to the ISA W83782D.
63The Asus chips are similar to an I2C-only W83782D.
64
65Chip #vin #fanin #pwm #temp wchipid vendid i2c ISA
66as99127f 7 3 0 3 0x31 0x12c3 yes no
67as99127f rev.2 (type_name = as99127f) 0x31 0x5ca3 yes no
68w83781d 7 3 0 3 0x10-1 0x5ca3 yes yes
69w83627hf 9 3 2 3 0x21 0x5ca3 yes yes(LPC)
70w83627thf 9 3 2 3 0x90 0x5ca3 no yes(LPC)
71w83782d 9 3 2-4 3 0x30 0x5ca3 yes yes
72w83783s 5-6 3 2 1-2 0x40 0x5ca3 yes no
73w83697hf 8 2 2 2 0x60 0x5ca3 no yes(LPC)
74
75Detection of these chips can sometimes be foiled because they can be in
76an internal state that allows no clean access. If you know the address
77of the chip, use a 'force' parameter; this will put them into a more
78well-behaved state first.
79
80The W8378* implements temperature sensors (three on the W83781D and W83782D,
81two on the W83783S), three fan rotation speed sensors, voltage sensors
82(seven on the W83781D, nine on the W83782D and six on the W83783S), VID
83lines, alarms with beep warnings, and some miscellaneous stuff.
84
85Temperatures are measured in degrees Celsius. There is always one main
86temperature sensor, and one (W83783S) or two (W83781D and W83782D) other
87sensors. An alarm is triggered for the main sensor once when the
88Overtemperature Shutdown limit is crossed; it is triggered again as soon as
89it drops below the Hysteresis value. A more useful behavior
90can be found by setting the Hysteresis value to +127 degrees Celsius; in
91this case, alarms are issued during all the time when the actual temperature
92is above the Overtemperature Shutdown value. The driver sets the
93hysteresis value for temp1 to 127 at initialization.
94
95For the other temperature sensor(s), an alarm is triggered when the
96temperature gets higher then the Overtemperature Shutdown value; it stays
97on until the temperature falls below the Hysteresis value. But on the
98W83781D, there is only one alarm that functions for both other sensors!
99Temperatures are guaranteed within a range of -55 to +125 degrees. The
100main temperature sensors has a resolution of 1 degree; the other sensor(s)
101of 0.5 degree.
102
103Fan rotation speeds are reported in RPM (rotations per minute). An alarm is
104triggered if the rotation speed has dropped below a programmable limit. Fan
105readings can be divided by a programmable divider (1, 2, 4 or 8 for the
106W83781D; 1, 2, 4, 8, 16, 32, 64 or 128 for the others) to give
107the readings more range or accuracy. Not all RPM values can accurately
108be represented, so some rounding is done. With a divider of 2, the lowest
109representable value is around 2600 RPM.
110
111Voltage sensors (also known as IN sensors) report their values in volts.
112An alarm is triggered if the voltage has crossed a programmable minimum
113or maximum limit. Note that minimum in this case always means 'closest to
114zero'; this is important for negative voltage measurements. All voltage
115inputs can measure voltages between 0 and 4.08 volts, with a resolution
116of 0.016 volt.
117
118The VID lines encode the core voltage value: the voltage level your processor
119should work with. This is hardcoded by the mainboard and/or processor itself.
120It is a value in volts. When it is unconnected, you will often find the
121value 3.50 V here.
122
123The W83782D and W83783S temperature conversion machine understands about
124several kinds of temperature probes. You can program the so-called
125beta value in the sensor files. '1' is the PII/Celeron diode, '2' is the
126TN3904 transistor, and 3435 the default thermistor value. Other values
127are (not yet) supported.
128
129In addition to the alarms described above, there is a CHAS alarm on the
130chips which triggers if your computer case is open.
131
132When an alarm goes off, you can be warned by a beeping signal through
133your computer speaker. It is possible to enable all beeping globally,
134or only the beeping for some alarms.
135
136If an alarm triggers, it will remain triggered until the hardware register
137is read at least once. This means that the cause for the alarm may
138already have disappeared! Note that in the current implementation, all
139hardware registers are read whenever any data is read (unless it is less
140than 1.5 seconds since the last update). This means that you can easily
141miss once-only alarms.
142
143The chips only update values each 1.5 seconds; reading them more often
144will do no harm, but will return 'old' values.
145
146AS99127F PROBLEMS
147-----------------
148The as99127f support was developed without the benefit of a datasheet.
149In most cases it is treated as a w83781d (although revision 2 of the
150AS99127F looks more like a w83782d).
151This support will be BETA until a datasheet is released.
152One user has reported problems with fans stopping
153occasionally.
154
155Note that the individual beep bits are inverted from the other chips.
156The driver now takes care of this so that user-space applications
157don't have to know about it.
158
159Known problems:
160 - Problems with diode/thermistor settings (supported?)
161 - One user reports fans stopping under high server load.
162 - Revision 2 seems to have 2 PWM registers but we don't know
163 how to handle them. More details below.
164
165These will not be fixed unless we get a datasheet.
166If you have problems, please lobby Asus to release a datasheet.
167Unfortunately several others have without success.
168Please do not send mail to us asking for better as99127f support.
169We have done the best we can without a datasheet.
170Please do not send mail to the author or the sensors group asking for
171a datasheet or ideas on how to convince Asus. We can't help.
172
173
174NOTES:
175-----
176 783s has no in1 so that in[2-6] are compatible with the 781d/782d.
177
178 783s pin is programmable for -5V or temp1; defaults to -5V,
179 no control in driver so temp1 doesn't work.
180
181 782d and 783s datasheets differ on which is pwm1 and which is pwm2.
182 We chose to follow 782d.
183
184 782d and 783s pin is programmable for fan3 input or pwm2 output;
185 defaults to fan3 input.
186 If pwm2 is enabled (with echo 255 1 > pwm2), then
187 fan3 will report 0.
188
189 782d has pwm1-2 for ISA, pwm1-4 for i2c. (pwm3-4 share pins with
190 the ISA pins)
191
192Data sheet updates:
193------------------
194 - PWM clock registers:
195
196 000: master / 512
197 001: master / 1024
198 010: master / 2048
199 011: master / 4096
200 100: master / 8192
201
202
203Answers from Winbond tech support
204---------------------------------
205>
206> 1) In the W83781D data sheet section 7.2 last paragraph, it talks about
207> reprogramming the R-T table if the Beta of the thermistor is not
208> 3435K. The R-T table is described briefly in section 8.20.
209> What formulas do I use to program a new R-T table for a given Beta?
210>
211 We are sorry that the calculation for R-T table value is
212confidential. If you have another Beta value of thermistor, we can help
213to calculate the R-T table for you. But you should give us real R-T
214Table which can be gotten by thermistor vendor. Therefore we will calculate
215them and obtain 32-byte data, and you can fill the 32-byte data to the
216register in Bank0.CR51 of W83781D.
217
218
219> 2) In the W83782D data sheet, it mentions that pins 38, 39, and 40 are
220> programmable to be either thermistor or Pentium II diode inputs.
221> How do I program them for diode inputs? I can't find any register
222> to program these to be diode inputs.
223 --> You may program Bank0 CR[5Dh] and CR[59h] registers.
224
225 CR[5Dh] bit 1(VTIN1) bit 2(VTIN2) bit 3(VTIN3)
226
227 thermistor 0 0 0
228 diode 1 1 1
229
230
231(error) CR[59h] bit 4(VTIN1) bit 2(VTIN2) bit 3(VTIN3)
232(right) CR[59h] bit 4(VTIN1) bit 5(VTIN2) bit 6(VTIN3)
233
234 PII thermal diode 1 1 1
235 2N3904 diode 0 0 0
236
237
238Asus Clones
239-----------
240
241We have no datasheets for the Asus clones (AS99127F and ASB100 Bach).
242Here are some very useful information that were given to us by Alex Van
243Kaam about how to detect these chips, and how to read their values. He
244also gives advice for another Asus chipset, the Mozart-2 (which we
245don't support yet). Thanks Alex!
246I reworded some parts and added personal comments.
247
248# Detection:
249
250AS99127F rev.1, AS99127F rev.2 and ASB100:
251- I2C address range: 0x29 - 0x2F
252- If register 0x58 holds 0x31 then we have an Asus (either ASB100 or
253 AS99127F)
254- Which one depends on register 0x4F (manufacturer ID):
255 0x06 or 0x94: ASB100
256 0x12 or 0xC3: AS99127F rev.1
257 0x5C or 0xA3: AS99127F rev.2
258 Note that 0x5CA3 is Winbond's ID (WEC), which let us think Asus get their
259 AS99127F rev.2 direct from Winbond. The other codes mean ATT and DVC,
260 respectively. ATT could stand for Asustek something (although it would be
261 very badly chosen IMHO), I don't know what DVC could stand for. Maybe
262 these codes simply aren't meant to be decoded that way.
263
264Mozart-2:
265- I2C address: 0x77
266- If register 0x58 holds 0x56 or 0x10 then we have a Mozart-2
267- Of the Mozart there are 3 types:
268 0x58=0x56, 0x4E=0x94, 0x4F=0x36: Asus ASM58 Mozart-2
269 0x58=0x56, 0x4E=0x94, 0x4F=0x06: Asus AS2K129R Mozart-2
270 0x58=0x10, 0x4E=0x5C, 0x4F=0xA3: Asus ??? Mozart-2
271 You can handle all 3 the exact same way :)
272
273# Temperature sensors:
274
275ASB100:
276- sensor 1: register 0x27
277- sensor 2 & 3 are the 2 LM75's on the SMBus
278- sensor 4: register 0x17
279Remark: I noticed that on Intel boards sensor 2 is used for the CPU
280 and 4 is ignored/stuck, on AMD boards sensor 4 is the CPU and sensor 2 is
281 either ignored or a socket temperature.
282
283AS99127F (rev.1 and 2 alike):
284- sensor 1: register 0x27
285- sensor 2 & 3 are the 2 LM75's on the SMBus
286Remark: Register 0x5b is suspected to be temperature type selector. Bit 1
287 would control temp1, bit 3 temp2 and bit 5 temp3.
288
289Mozart-2:
290- sensor 1: register 0x27
291- sensor 2: register 0x13
292
293# Fan sensors:
294
295ASB100, AS99127F (rev.1 and 2 alike):
296- 3 fans, identical to the W83781D
297
298Mozart-2:
299- 2 fans only, 1350000/RPM/div
300- fan 1: register 0x28, divisor on register 0xA1 (bits 4-5)
301- fan 2: register 0x29, divisor on register 0xA1 (bits 6-7)
302
303# Voltages:
304
305This is where there is a difference between AS99127F rev.1 and 2.
306Remark: The difference is similar to the difference between
307 W83781D and W83782D.
308
309ASB100:
310in0=r(0x20)*0.016
311in1=r(0x21)*0.016
312in2=r(0x22)*0.016
313in3=r(0x23)*0.016*1.68
314in4=r(0x24)*0.016*3.8
315in5=r(0x25)*(-0.016)*3.97
316in6=r(0x26)*(-0.016)*1.666
317
318AS99127F rev.1:
319in0=r(0x20)*0.016
320in1=r(0x21)*0.016
321in2=r(0x22)*0.016
322in3=r(0x23)*0.016*1.68
323in4=r(0x24)*0.016*3.8
324in5=r(0x25)*(-0.016)*3.97
325in6=r(0x26)*(-0.016)*1.503
326
327AS99127F rev.2:
328in0=r(0x20)*0.016
329in1=r(0x21)*0.016
330in2=r(0x22)*0.016
331in3=r(0x23)*0.016*1.68
332in4=r(0x24)*0.016*3.8
333in5=(r(0x25)*0.016-3.6)*5.14+3.6
334in6=(r(0x26)*0.016-3.6)*3.14+3.6
335
336Mozart-2:
337in0=r(0x20)*0.016
338in1=255
339in2=r(0x22)*0.016
340in3=r(0x23)*0.016*1.68
341in4=r(0x24)*0.016*4
342in5=255
343in6=255
344
345
346# PWM
347
348Additional info about PWM on the AS99127F (may apply to other Asus
349chips as well) by Jean Delvare as of 2004-04-09:
350
351AS99127F revision 2 seems to have two PWM registers at 0x59 and 0x5A,
352and a temperature sensor type selector at 0x5B (which basically means
353that they swapped registers 0x59 and 0x5B when you compare with Winbond
354chips).
355Revision 1 of the chip also has the temperature sensor type selector at
3560x5B, but PWM registers have no effect.
357
358We don't know exactly how the temperature sensor type selection works.
359Looks like bits 1-0 are for temp1, bits 3-2 for temp2 and bits 5-4 for
360temp3, although it is possible that only the most significant bit matters
361each time. So far, values other than 0 always broke the readings.
362
363PWM registers seem to be split in two parts: bit 7 is a mode selector,
364while the other bits seem to define a value or threshold.
365
366When bit 7 is clear, bits 6-0 seem to hold a threshold value. If the value
367is below a given limit, the fan runs at low speed. If the value is above
368the limit, the fan runs at full speed. We have no clue as to what the limit
369represents. Note that there seem to be some inertia in this mode, speed
370changes may need some time to trigger. Also, an hysteresis mechanism is
371suspected since walking through all the values increasingly and then
372decreasingly led to slightly different limits.
373
374When bit 7 is set, bits 3-0 seem to hold a threshold value, while bits 6-4
375would not be significant. If the value is below a given limit, the fan runs
376at full speed, while if it is above the limit it runs at low speed (so this
377is the contrary of the other mode, in a way). Here again, we don't know
378what the limit is supposed to represent.
379
380One remarkable thing is that the fans would only have two or three
381different speeds (transitional states left apart), not a whole range as
382you usually get with PWM.
383
384As a conclusion, you can write 0x00 or 0x8F to the PWM registers to make
385fans run at low speed, and 0x7F or 0x80 to make them run at full speed.
386
387Please contact us if you can figure out how it is supposed to work. As
388long as we don't know more, the w83781d driver doesn't handle PWM on
389AS99127F chips at all.
390
391Additional info about PWM on the AS99127F rev.1 by Hector Martin:
392
393I've been fiddling around with the (in)famous 0x59 register and
394found out the following values do work as a form of coarse pwm:
395
3960x80 - seems to turn fans off after some time(1-2 minutes)... might be
397some form of auto-fan-control based on temp? hmm (Qfan? this mobo is an
398old ASUS, it isn't marketed as Qfan. Maybe some beta pre-attemp at Qfan
399that was dropped at the BIOS)
4000x81 - off
4010x82 - slightly "on-ner" than off, but my fans do not get to move. I can
402hear the high-pitched PWM sound that motors give off at too-low-pwm.
4030x83 - now they do move. Estimate about 70% speed or so.
4040x84-0x8f - full on
405
406Changing the high nibble doesn't seem to do much except the high bit
407(0x80) must be set for PWM to work, else the current pwm doesn't seem to
408change.
409
410My mobo is an ASUS A7V266-E. This behavior is similar to what I got
411with speedfan under Windows, where 0-15% would be off, 15-2x% (can't
412remember the exact value) would be 70% and higher would be full on.
diff --git a/Documentation/i2c/chips/w83l785ts b/Documentation/i2c/chips/w83l785ts
new file mode 100644
index 000000000000..1841cedc25b2
--- /dev/null
+++ b/Documentation/i2c/chips/w83l785ts
@@ -0,0 +1,39 @@
1Kernel driver w83l785ts
2=======================
3
4Supported chips:
5 * Winbond W83L785TS-S
6 Prefix: 'w83l785ts'
7 Addresses scanned: I2C 0x2e
8 Datasheet: Publicly available at the Winbond USA website
9 http://www.winbond-usa.com/products/winbond_products/pdfs/PCIC/W83L785TS-S.pdf
10
11Authors:
12 Jean Delvare <khali@linux-fr.org>
13
14Description
15-----------
16
17The W83L785TS-S is a digital temperature sensor. It senses the
18temperature of a single external diode. The high limit is
19theoretically defined as 85 or 100 degrees C through a combination
20of external resistors, so the user cannot change it. Values seen so
21far suggest that the two possible limits are actually 95 and 110
22degrees C. The datasheet is rather poor and obviously inaccurate
23on several points including this one.
24
25All temperature values are given in degrees Celsius. Resolution
26is 1.0 degree. See the datasheet for details.
27
28The w83l785ts driver will not update its values more frequently than
29every other second; reading them more often will do no harm, but will
30return 'old' values.
31
32Known Issues
33------------
34
35On some systems (Asus), the BIOS is known to interfere with the driver
36and cause read errors. The driver will retry a given number of times
37(5 by default) and then give up, returning the old value (or 0 if
38there is no old value). It seems to work well enough so that you should
39not notice anything. Thanks to James Bolt for helping test this feature.