Peter Barker
8679e95fca
These are really "shouldn't happen", and only happen at startup, so impact of the memory leak is quite limited. Fixing this requires putting in a virtual destructor on the MessageWriter class - this and the calls add up to several hundred bytes of flash.
Logger Notes
Format Types
The format type specifies the amount of storage required for the entry and how the content should be interpreted.
| Char | C Type |
|---|---|
| a | int16_t[32] |
| b | int8_t |
| B | uint8_t |
| h | int16_t |
| H | uint16_t |
| i | int32_t |
| I | uint32_t |
| f | float |
| d | double |
| n | char[4] |
| N | char[16] |
| Z | char[64] |
| L | int32_t latitude/longitude (so -35.1332423 becomes -351332423) |
| M | uint8_t flight mode |
| q | int64_t |
| Q | uint64_t |
Legacy field types - do not use. These have been replaced by using the base C type and an appropriate multiplier column entry.
| Char | CType+Mult |
|---|---|
| c | int16_t * 100 |
| C | uint16_t * 100 |
| e | int32_t * 100 |
| E | uint32_t * 100 |
Units
All units here should be base units This does mean battery capacity is here as "amp*second" Please keep the names consistent with Tools/autotest/param_metadata/param.py:33
| Char | Unit Abbrev. | Description |
|---|---|---|
| '-' | "" | no units e.g. Pi |
| '?' | "UNKNOWN" | Units which haven't been worked out yet.... |
| 'A' | "A" | Ampere |
| 'd' | "deg" | of the angular variety |
| 'b' | "B" | bytes |
| 'k' | "deg/s" | degrees per second. Degrees are NOT SI |
| 'D' | "deglatitude" | degrees of latitude |
| 'e' | "deg/s/s" | degrees per second per second. Degrees are NOT SI |
| 'E' | "rad/s" | radians per second |
| 'G' | "Gauss" | Gauss is not an SI unit |
| 'h' | "degheading" | 0.? to 359.? |
| 'i' | "A.s" | Ampere second |
| 'J' | "W.s" | Joule (Watt second) |
| 'l' | "l" | litres |
| 'L' | "rad/s/s" | radians per second per second |
| 'm' | "m" | metres |
| 'n' | "m/s" | metres per second |
| 'N' | "N" | Newton |
| 'o' | "m/s/s" | metres per second per second |
| 'O' | "degC" | degrees Celsius. Not SI |
| '%' | "%" | percent |
| 'S' | "satellites" | number of satellites |
| 's' | "s" | seconds |
| 'q' | "rpm" | rounds per minute. Not SI |
| 'r' | "rad" | radians |
| 'U' | "deglongitude" | degrees of longitude |
| 'u' | "ppm" | pulses per minute |
| 'v' | "V" | Volt |
| 'P' | "Pa" | Pascal |
| 'w' | "Ohm" | Ohm |
| 'Y' | "us" | pulse width modulation in microseconds |
| 'z' | "Hz" | Hertz |
| '#' | "instance" | (e.g.)Sensor instance number |
Multipliers
This multiplier information applies to the raw value present in the log. Any adjustment implied by the format field (e.g. the "centi" in "centidegrees" is IGNORED for the purposes of scaling. Essentially "format" simply tells you the C-type, and format-type h (int16_t) is equivalent to format-type c (int16_t*100) tl;dr a GCS shouldn't/mustn't infer any scaling from the unit name
| Char | Multiplier |
|---|---|
| '-' | 0 |
| '?' | 1 |
| '2' | 1e2 |
| '1' | 1e1 |
| '0' | 1e0 |
| 'A' | 1e-1 |
| 'B' | 1e-2 |
| 'C' | 1e-3 |
| 'D' | 1e-4 |
| 'E' | 1e-5 |
| 'F' | 1e-6 |
| 'G' | 1e-7 |
| '!' | 3.6 |
| '/' | 3600 |