R.Marek@sh.cvut.cz | 7f15b66 | 2005-05-26 12:42:19 +0000 | [diff] [blame^] | 1 | Kernel driver pc87360 |
| 2 | ===================== |
| 3 | |
| 4 | Supported 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 | |
| 15 | Authors: Jean Delvare <khali@linux-fr.org> |
| 16 | |
| 17 | Thanks to Sandeep Mehta, Tonko de Rooy and Daniel Ceregatti for testing. |
| 18 | Thanks to Rudolf Marek for helping me investigate conversion issues. |
| 19 | |
| 20 | |
| 21 | Module 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 | |
| 31 | Note that this parameter has no effect for the PC87360, PC87363 and PC87364 |
| 32 | chips. |
| 33 | |
| 34 | Also note that for the PC87366, initialization levels 2 and 3 don't enable |
| 35 | all temperature channels, because some of them share pins with each other, |
| 36 | so they can't be used at the same time. |
| 37 | |
| 38 | |
| 39 | Description |
| 40 | ----------- |
| 41 | |
| 42 | The National Semiconductor PC87360 Super I/O chip contains monitoring and |
| 43 | PWM control circuitry for two fans. The PC87363 chip is similar, and the |
| 44 | PC87364 chip has monitoring and PWM control for a third fan. |
| 45 | |
| 46 | The National Semiconductor PC87365 and PC87366 Super I/O chips are complete |
| 47 | hardware monitoring chipsets, not only controlling and monitoring three fans, |
| 48 | but 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 | |
| 59 | The driver assumes that no more than one chip is present, and one of the |
| 60 | standard Super I/O addresses is used (0x2E/0x2F or 0x4E/0x4F) |
| 61 | |
| 62 | Fan Monitoring |
| 63 | -------------- |
| 64 | |
| 65 | Fan rotation speeds are reported in RPM (revolutions per minute). An alarm |
| 66 | is triggered if the rotation speed has dropped below a programmable limit. |
| 67 | A different alarm is triggered if the fan speed is too low to be measured. |
| 68 | |
| 69 | Fan readings are affected by a programmable clock divider, giving the |
| 70 | readings more range or accuracy. Usually, users have to learn how it works, |
| 71 | but this driver implements dynamic clock divider selection, so you don't |
| 72 | have to care no more. |
| 73 | |
| 74 | For 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 | |
| 84 | For 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) |
| 88 | The clock speed for the PC87360 family is 480 kHz. I arbitrarily chose 100 |
| 89 | RPM as the lowest acceptable accuracy. |
| 90 | |
| 91 | As mentioned above, you don't have to care about this no more. |
| 92 | |
| 93 | Note that not all RPM values can be represented, even when the best clock |
| 94 | divider is selected. This is not only true for the measured speeds, but |
| 95 | also for the programmable low limits, so don't be surprised if you try to |
| 96 | set, say, fan1_min to 2900 and it finally reads 2909. |
| 97 | |
| 98 | |
| 99 | Fan Control |
| 100 | ----------- |
| 101 | |
| 102 | PWM (pulse width modulation) values range from 0 to 255, with 0 meaning |
| 103 | that the fan is stopped, and 255 meaning that the fan goes at full speed. |
| 104 | |
| 105 | Be extremely careful when changing PWM values. Low PWM values, even |
| 106 | non-zero, can stop the fan, which may cause irreversible damage to your |
| 107 | hardware if temperature increases too much. When changing PWM values, go |
| 108 | step by step and keep an eye on temperatures. |
| 109 | |
| 110 | One user reported problems with PWM. Changing PWM values would break fan |
| 111 | speed readings. No explanation nor fix could be found. |
| 112 | |
| 113 | |
| 114 | Temperature Monitoring |
| 115 | ---------------------- |
| 116 | |
| 117 | Temperatures are reported in degrees Celsius. Each temperature measured has |
| 118 | associated low, high and overtemperature limits, each of which triggers an |
| 119 | alarm when crossed. |
| 120 | |
| 121 | The first two temperature channels are external. The third one (PC87366 |
| 122 | only) is internal. |
| 123 | |
| 124 | The PC87366 has three additional temperature channels, based on |
| 125 | thermistors (as opposed to thermal diodes for the first three temperature |
| 126 | channels). For technical reasons, these channels are held by the VLM |
| 127 | (voltage level monitor) logical device, not the TMS (temperature |
| 128 | measurement) one. As a consequence, these temperatures are exported as |
| 129 | voltages, and converted into temperatures in user-space. |
| 130 | |
| 131 | Note that these three additional channels share their pins with the |
| 132 | external thermal diode channels, so you (physically) can't use them all at |
| 133 | the same time. Although it should be possible to mix the two sensor types, |
| 134 | the documents from National Semiconductor suggest that motherboard |
| 135 | manufacturers should choose one type and stick to it. So you will more |
| 136 | likely have either channels 1 to 3 (thermal diodes) or 3 to 6 (internal |
| 137 | thermal diode, and thermistors). |
| 138 | |
| 139 | |
| 140 | Voltage Monitoring |
| 141 | ------------------ |
| 142 | |
| 143 | Voltages are reported relatively to a reference voltage, either internal or |
| 144 | external. Some of them (in7:Vsb, in8:Vdd and in10:AVdd) are divided by two |
| 145 | internally, you will have to compensate in sensors.conf. Others (in0 to in6) |
| 146 | are likely to be divided externally. The meaning of each of these inputs as |
| 147 | well as the values of the resistors used for division is left to the |
| 148 | motherboard manufacturers, so you will have to document yourself and edit |
| 149 | sensors.conf accordingly. National Semiconductor has a document with |
| 150 | recommended resistor values for some voltages, but this still leaves much |
| 151 | room for per motherboard specificities, unfortunately. Even worse, |
| 152 | motherboard manufacturers don't seem to care about National Semiconductor's |
| 153 | recommendations. |
| 154 | |
| 155 | Each voltage measured has associated low and high limits, each of which |
| 156 | triggers an alarm when crossed. |
| 157 | |
| 158 | When available, VID inputs are used to provide the nominal CPU Core voltage. |
| 159 | The driver will default to VRM 9.0, but this can be changed from user-space. |
| 160 | The chipsets can handle two sets of VID inputs (on dual-CPU systems), but |
| 161 | the driver will only export one for now. This may change later if there is |
| 162 | a need. |
| 163 | |
| 164 | |
| 165 | General Remarks |
| 166 | --------------- |
| 167 | |
| 168 | If an alarm triggers, it will remain triggered until the hardware register |
| 169 | is read at least once. This means that the cause for the alarm may already |
| 170 | have disappeared! Note that all hardware registers are read whenever any |
| 171 | data is read (unless it is less than 2 seconds since the last update, in |
| 172 | which case cached values are returned instead). As a consequence, when |
| 173 | a once-only alarm triggers, it may take 2 seconds for it to show, and 2 |
| 174 | more seconds for it to disappear. |
| 175 | |
| 176 | Monitoring of in9 isn't enabled at lower init levels (<3) because that |
| 177 | channel measures the battery voltage (Vbat). It is a known fact that |
| 178 | repeatedly sampling the battery voltage reduces its lifetime. National |
| 179 | Semiconductor smartly designed their chipset so that in9 is sampled only |
| 180 | once every 1024 sampling cycles (that is every 34 minutes at the default |
| 181 | sampling rate), so the effect is attenuated, but still present. |
| 182 | |
| 183 | |
| 184 | Limitations |
| 185 | ----------- |
| 186 | |
| 187 | The datasheets suggests that some values (fan mins, fan dividers) |
| 188 | shouldn't be changed once the monitoring has started, but we ignore that |
| 189 | recommendation. We'll reconsider if it actually causes trouble. |