Hans de Goede | 3b02d33 | 2009-01-07 16:37:31 +0100 | [diff] [blame] | 1 | Kernel driver f71882fg |
| 2 | ====================== |
| 3 | |
| 4 | Supported chips: |
Jean Delvare | 5d7f77b | 2011-03-26 10:45:02 +0100 | [diff] [blame] | 5 | * Fintek F71808E |
| 6 | Prefix: 'f71808e' |
| 7 | Addresses scanned: none, address read from Super I/O config space |
| 8 | Datasheet: Not public |
Hans de Goede | 629c58b | 2011-05-25 20:43:32 +0200 | [diff] [blame] | 9 | * Fintek F71808A |
| 10 | Prefix: 'f71808a' |
| 11 | Addresses scanned: none, address read from Super I/O config space |
| 12 | Datasheet: Not public |
Hans de Goede | 09475d3 | 2009-06-15 18:39:52 +0200 | [diff] [blame] | 13 | * Fintek F71858FG |
| 14 | Prefix: 'f71858fg' |
Hans de Goede | 3b02d33 | 2009-01-07 16:37:31 +0100 | [diff] [blame] | 15 | Addresses scanned: none, address read from Super I/O config space |
| 16 | Datasheet: Available from the Fintek website |
| 17 | * Fintek F71862FG and F71863FG |
| 18 | Prefix: 'f71862fg' |
| 19 | Addresses scanned: none, address read from Super I/O config space |
| 20 | Datasheet: Available from the Fintek website |
Hans de Goede | c11bb99 | 2011-03-09 20:57:15 +0100 | [diff] [blame] | 21 | * Fintek F71869F and F71869E |
| 22 | Prefix: 'f71869' |
| 23 | Addresses scanned: none, address read from Super I/O config space |
| 24 | Datasheet: Available from the Fintek website |
Hans de Goede | 5da556e | 2011-07-03 13:32:53 +0200 | [diff] [blame^] | 25 | * Fintek F71869A |
| 26 | Prefix: 'f71869a' |
| 27 | Addresses scanned: none, address read from Super I/O config space |
| 28 | Datasheet: Not public |
Hans de Goede | 09475d3 | 2009-06-15 18:39:52 +0200 | [diff] [blame] | 29 | * Fintek F71882FG and F71883FG |
| 30 | Prefix: 'f71882fg' |
| 31 | Addresses scanned: none, address read from Super I/O config space |
| 32 | Datasheet: Available from the Fintek website |
Hans de Goede | 7669896 | 2009-12-09 20:36:01 +0100 | [diff] [blame] | 33 | * Fintek F71889FG |
| 34 | Prefix: 'f71889fg' |
| 35 | Addresses scanned: none, address read from Super I/O config space |
Hans de Goede | 3cad402 | 2011-03-09 20:57:14 +0100 | [diff] [blame] | 36 | Datasheet: Available from the Fintek website |
| 37 | * Fintek F71889ED |
| 38 | Prefix: 'f71889ed' |
| 39 | Addresses scanned: none, address read from Super I/O config space |
Hans de Goede | 7669896 | 2009-12-09 20:36:01 +0100 | [diff] [blame] | 40 | Datasheet: Should become available on the Fintek website soon |
Hans de Goede | a66c108 | 2011-03-26 10:45:02 +0100 | [diff] [blame] | 41 | * Fintek F71889A |
| 42 | Prefix: 'f71889a' |
| 43 | Addresses scanned: none, address read from Super I/O config space |
| 44 | Datasheet: Should become available on the Fintek website soon |
Hans de Goede | 3b02d33 | 2009-01-07 16:37:31 +0100 | [diff] [blame] | 45 | * Fintek F8000 |
| 46 | Prefix: 'f8000' |
| 47 | Addresses scanned: none, address read from Super I/O config space |
| 48 | Datasheet: Not public |
Jean Delvare | 5d7f77b | 2011-03-26 10:45:02 +0100 | [diff] [blame] | 49 | * Fintek F81801U |
| 50 | Prefix: 'f71889fg' |
| 51 | Addresses scanned: none, address read from Super I/O config space |
| 52 | Datasheet: Not public |
| 53 | Note: This is the 64-pin variant of the F71889FG, they have the |
| 54 | same device ID and are fully compatible as far as hardware |
| 55 | monitoring is concerned. |
Jean Delvare | 383586b | 2011-03-26 10:45:02 +0100 | [diff] [blame] | 56 | * Fintek F81865F |
| 57 | Prefix: 'f81865f' |
| 58 | Addresses scanned: none, address read from Super I/O config space |
| 59 | Datasheet: Available from the Fintek website |
Hans de Goede | 3b02d33 | 2009-01-07 16:37:31 +0100 | [diff] [blame] | 60 | |
| 61 | Author: Hans de Goede <hdegoede@redhat.com> |
| 62 | |
| 63 | |
| 64 | Description |
| 65 | ----------- |
| 66 | |
Hans de Goede | 3cad402 | 2011-03-09 20:57:14 +0100 | [diff] [blame] | 67 | Fintek F718xx/F8000 Super I/O chips include complete hardware monitoring |
| 68 | capabilities. They can monitor up to 9 voltages, 4 fans and 3 temperature |
| 69 | sensors. |
Hans de Goede | 3b02d33 | 2009-01-07 16:37:31 +0100 | [diff] [blame] | 70 | |
| 71 | These chips also have fan controlling features, using either DC or PWM, in |
| 72 | three different modes (one manual, two automatic). |
| 73 | |
| 74 | The driver assumes that no more than one chip is present, which seems |
| 75 | reasonable. |
| 76 | |
| 77 | |
| 78 | Monitoring |
| 79 | ---------- |
| 80 | |
| 81 | The Voltage, Fan and Temperature Monitoring uses the standard sysfs |
| 82 | interface as documented in sysfs-interface, without any exceptions. |
| 83 | |
| 84 | |
| 85 | Fan Control |
| 86 | ----------- |
| 87 | |
| 88 | Both PWM (pulse-width modulation) and DC fan speed control methods are |
| 89 | supported. The right one to use depends on external circuitry on the |
| 90 | motherboard, so the driver assumes that the BIOS set the method |
| 91 | properly. |
| 92 | |
Hans de Goede | 7669896 | 2009-12-09 20:36:01 +0100 | [diff] [blame] | 93 | Note that the lowest numbered temperature zone trip point corresponds to |
| 94 | to the border between the highest and one but highest temperature zones, and |
| 95 | vica versa. So the temperature zone trip points 1-4 (or 1-2) go from high temp |
| 96 | to low temp! This is how things are implemented in the IC, and the driver |
| 97 | mimicks this. |
| 98 | |
Hans de Goede | 3b02d33 | 2009-01-07 16:37:31 +0100 | [diff] [blame] | 99 | There are 2 modes to specify the speed of the fan, PWM duty cycle (or DC |
| 100 | voltage) mode, where 0-100% duty cycle (0-100% of 12V) is specified. And RPM |
| 101 | mode where the actual RPM of the fan (as measured) is controlled and the speed |
| 102 | gets specified as 0-100% of the fan#_full_speed file. |
| 103 | |
| 104 | Since both modes work in a 0-100% (mapped to 0-255) scale, there isn't a |
| 105 | whole lot of a difference when modifying fan control settings. The only |
| 106 | important difference is that in RPM mode the 0-100% controls the fan speed |
| 107 | between 0-100% of fan#_full_speed. It is assumed that if the BIOS programs |
| 108 | RPM mode, it will also set fan#_full_speed properly, if it does not then |
| 109 | fan control will not work properly, unless you set a sane fan#_full_speed |
| 110 | value yourself. |
| 111 | |
| 112 | Switching between these modes requires re-initializing a whole bunch of |
| 113 | registers, so the mode which the BIOS has set is kept. The mode is |
| 114 | printed when loading the driver. |
| 115 | |
| 116 | Three different fan control modes are supported; the mode number is written |
| 117 | to the pwm#_enable file. Note that not all modes are supported on all |
Hans de Goede | 09475d3 | 2009-06-15 18:39:52 +0200 | [diff] [blame] | 118 | chips, and some modes may only be available in RPM / PWM mode. |
Hans de Goede | 3b02d33 | 2009-01-07 16:37:31 +0100 | [diff] [blame] | 119 | Writing an unsupported mode will result in an invalid parameter error. |
| 120 | |
| 121 | * 1: Manual mode |
| 122 | You ask for a specific PWM duty cycle / DC voltage or a specific % of |
| 123 | fan#_full_speed by writing to the pwm# file. This mode is only |
Hans de Goede | 09475d3 | 2009-06-15 18:39:52 +0200 | [diff] [blame] | 124 | available on the F71858FG / F8000 if the fan channel is in RPM mode. |
Hans de Goede | 3b02d33 | 2009-01-07 16:37:31 +0100 | [diff] [blame] | 125 | |
| 126 | * 2: Normal auto mode |
| 127 | You can define a number of temperature/fan speed trip points, which % the |
| 128 | fan should run at at this temp and which temp a fan should follow using the |
| 129 | standard sysfs interface. The number and type of trip points is chip |
| 130 | depended, see which files are available in sysfs. |
| 131 | Fan/PWM channel 3 of the F8000 is always in this mode! |
| 132 | |
| 133 | * 3: Thermostat mode (Only available on the F8000 when in duty cycle mode) |
| 134 | The fan speed is regulated to keep the temp the fan is mapped to between |
| 135 | temp#_auto_point2_temp and temp#_auto_point3_temp. |
| 136 | |
Hans de Goede | 3cad402 | 2011-03-09 20:57:14 +0100 | [diff] [blame] | 137 | All of the automatic modes require that pwm1 corresponds to fan1, pwm2 to |
Hans de Goede | 3b02d33 | 2009-01-07 16:37:31 +0100 | [diff] [blame] | 138 | fan2 and pwm3 to fan3. |