* Enhance GPS search failure handling backoff logic
* Potential fix for pull request finding
Co-authored-by: Copilot Autofix powered by AI <175728472+Copilot@users.noreply.github.com>
* Remove stray submodule gitlink for .claude worktree
A 160000 (gitlink) entry for .claude/worktrees/naughty-payne-60fdb7
pointing at f2923590bc was accidentally committed in 9db15780f. The
path isn't a real submodule — it's a Claude Code agent worktree that
shouldn't be tracked.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
---------
Co-authored-by: Copilot Autofix powered by AI <175728472+Copilot@users.noreply.github.com>
Co-authored-by: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
* Fix TransmitHistory to improve epoch handling
* Enable epoch handling in unit tests
* Improve comments and test handling for epoch persistence in TransmitHistory
* Add boot-relative timestamp handling and unit tests for TransmitHistory
* loadFromDisk should handle legacy entries and clean up old v1 files after migration
* Revert "loadFromDisk should handle legacy entries and clean up old v1 files after migration"
This reverts commit eb7e5c7acfa4ac077fe50980be752e4b42a739b8.
* Add NodeInfoModule integration for RTC quality changes and trigger immediate checks
* Update test conditions for RTC quality checks
* simplify the observer pattern, since all the called functions are const getters.
* use arduino macro over std: for numerical values and refactor local variables in drawScrollbar()
* oh, so Cppcheck actually complained about const pointers not being const.
* slowly getting out of ifdef hell
* fix inkHUD warnings as well
* last 2 check warnings
* git checks should fail on low defects from now on
Trying this to see if anything bad happen if I were to replace most raw pointers with unique_ptr.
I didn't used std::make_unique since it is only supported in C++14 and onwards but until we update esp32 to arduino 3.x the ESP32 xtensa chips use a C++11 std.
* Power off control pin on Thinknode m5 during deepsleep
* Apply suggestion from @Copilot
Co-authored-by: Copilot <175728472+Copilot@users.noreply.github.com>
* Finish HAS_RTC cleanup
* Add RTC for Thinknode M5
* Don't double-init Wire
* Specify the RTC chip directly rather than use SensorRtcHelper.
Saves a bit of flash, and avoid mis-detection
---------
Co-authored-by: Copilot <175728472+Copilot@users.noreply.github.com>
Support softsleep by defining PIN_GPS_STANDBY on CDTop CD-PA1010D.
This differs from existing MTK GPS e.g. L76K, where pulling PIN_GPS_STANDBY (WAKE-UP pin) low is not sufficient to put the GPS module in standby.
An additional `$PMTK225,4*2F` must be sent to enter "Backup Mode", which is exited by bringing PIN_GPS_STANDBY (WAKE-UP pin) high.
Refer to datasheet[0] §1.9.3 "Backup Mode".
0: https://cdn-learn.adafruit.com/assets/assets/000/084/295/original/CD_PA1010D_Datasheet_v.03.pdf
Signed-off-by: Andrew Yong <me@ndoo.sg>
* Preliminary Thinknode M4 Support
* oops
* Fix RF switch TX configuration
* trunk'd
* GPS fix for M4
* Battery handling and LED for M4
* Trunk
* Drop debug warnings
* Make Red LED notification
* Merge cleanup
* Make white LEDs flash during charge
---------
Co-authored-by: Jonathan Bennett <jbennett@incomsystems.biz>
I thought git would be smart enough to understand all the whitespace changes but even with all the flags I know to make it ignore theses it still blows up if there are identical changes on both sides.
I have a solution but it require creating a new commit at the merge base for each conflicting PR and merging it into develop.
I don't think blowing up all PRs is worth for now, maybe if we can coordinate this for V3 let's say.
This reverts commit 0d11331d18.
* Properly format timestamp in log message
* Better formatting of GPS_DEBUG logging in gps probe
* Reset GPS after serial speed change, and look for magic string to identify chip
* Add UC6580 to boot message detection, for Heltec Tracker
* Add L76K detect from boot string, for Heltec-v4
* Slightly more useful GPS debugging
* Back out detection of L76K via startup messages.
* Ignore PIN_GPS_RESET = -1 and rename passive_detect array.
---------
Co-authored-by: Tom Fifield <tom@tomfifield.net>
8 months ago, when this was added to the code, the GPS probe code
was still a little flaky.
Particularly after #6114 and #6116 were added, reliability improved
for all devices as we were sending fewer calls on the bus.
Today, the T1000E is the only Meshtastic device that regularly takes
2, 3, or 4 attempts to be detected via the probe code.
Removing these lines, on my T1000E, results in the AG3335 being
detected immediately.
Co-authored-by: Jonathan Bennett <jbennett@incomsystems.biz>
* Wait until after GPS lock hold before updating position, if we can.
After the recent patch, we hold lock for a bit before updating the position.
The positions that come in after the hold are genuinely better positions
than what we've been doing before. However, they only come 20 seconds
after we've got lock.
Previously, we would update the local position as soon as we got a lock as well
as at the end of the hold, since a hold was not always possible.
With this patch, if the settings allow, we should skip that first local position update.
Fixes https://github.com/meshtastic/firmware/issues/8029
* Fix falling edge handling.
* spelling
Co-authored-by: Copilot <175728472+Copilot@users.noreply.github.com>
* Congeal lock handling
* Add named constants
* define unit to avoid confusion
Co-authored-by: Copilot <175728472+Copilot@users.noreply.github.com>
* ifdef, not if.
Co-authored-by: Copilot <175728472+Copilot@users.noreply.github.com>
* Add handling for when we first turn on.
* Don't run if not active
* Reset fixhold
* Logic fixes
* Add path for ACTIVE--> IDLE --> ACTIVE
Previously we only covered HARDSLEEP --> ACTIVE.
* Change hold time to gps_update_interval - 10s
* Update comment
* Add extra buffer to avoid re-starting hold
---------
Co-authored-by: Copilot <175728472+Copilot@users.noreply.github.com>
Co-authored-by: Ben Meadors <benmmeadors@gmail.com>
* Do slightly better at threading the search for GPS hardware
* Formatting and minor logic fix
* Remove now-spam GPS Log messages
---------
Co-authored-by: Ben Meadors <benmmeadors@gmail.com>
* Add a new GPS model CM121.
* Add CM121 to Unicore.
* Trunk fixes, remove unneded NMEA lines
---------
Co-authored-by: Tom Fifield <tom@tomfifield.net>
* Start portduino_config refactor
* refactor GPIOs to new portduino_config
* More portduino_config work
* More conversion to portduino_config
* Finish portduino_config transition
* trunk
* yaml output work
* Simplify the GPIO config
* Trunk
* Fix overflow of time value
* Revert "Fix overflow of time value"
This reverts commit 084796920179e80a7500d36c25fd4d82b3ef4214.
* That got boogered up
* Scale probe buffer size based on current baud rate
* Throttle bad time validation logging and fix time comparison logic
* Remove comment
* Missed the other instances
* Copy pasta
In 2.7.7 / 2.7.8 we introduced some new checks for time accuracy.
In combination, these result in a spamming of the logs when a bad time is found
When the GPS is active, we're calling the GPS thread every 0.2secs.
So this log could be printed 4,500 times in a no-lock scenario :)
Reserve this experience for developers using GPS_DEBUG.
Fixes https://github.com/meshtastic/firmware/issues/7896
Further to https://github.com/meshtastic/firmware/pull/7897 ,
there was another log line that was triggering indiscriminantly on
GPS_INTERVAL_THRESHOLD .
Rather than logging a bad time 4000 times, let's just log one good time
when it is set.
* Fix GPS that hard code 2080 as the start time.
Some GPS chips, such as the AG3335 in T1000e and L96 have a hardcoded
time of 2080-01-05 when they start up.
To fix that in a way that seems permanent, let's ignore times that
are more than 40 years since the firmware was built. We should followup
in late 2039 to see if any changes are needed.
Reported-By: @b8b8
* Update src/gps/RTC.cpp
Co-authored-by: Copilot <175728472+Copilot@users.noreply.github.com>
* Put FORTY_YEARS in header and use in both places.
* Restore Ben's nicer log lines.
---------
Co-authored-by: Ben Meadors <benmmeadors@gmail.com>
Co-authored-by: Copilot <175728472+Copilot@users.noreply.github.com>
* Hold for >20s after GPS lock
GPS chips are designed to stay locked for a while to download some data and save it.
This data is important for speeding up future locks, and making them higher quality.
Our present configuration could make every lock perform similar to first lock.
This patch sets a hold of between 20s and 10% of the lock search time after lock
is acquired. This should allow the GPS to finish its work before we turn it off.
Fixes https://github.com/meshtastic/firmware/issues/7466
* Remove T1000E-specific GPS holds
The new code does the same thing, for all devices.
* Fix publishing settings
* Cleanups, removing unused variables.
* ifdef log line with GPS_DEBUG
* fixQual is not a bool.
Further to https://github.com/meshtastic/firmware/pull/7772 ,
we discovered that some RTCs have hard-coded start times well in the
past.
This patch gives RTCs the same treatment as GPS - if the time is
earlier than BUILD_EPOCH, we don't use it.
Fixes#7771Fixes#7750
There are certain GPS chips that have a hard-coded time in firmware
that they will return before lock. We set our own hard-coded time,
BUILD_EPOCH, that should be newer and use the comparison to not set
a bad time.
In https://github.com/meshtastic/firmware/pull/7261 we introduced
the RTCSetResult and improved it in https://github.com/meshtastic/firmware/pull/7375 .
However, the original try-fix left logic in GPS.cpp that could
still result in broadcasting the bad time.
Further, as part of our fix we cleared the GPS buffer if we didn't
get a good time. The mesh was hurting at the time, so this was a reasonable
approach. However, given time tends to come in when we're trying to get
early lock, this had the potential side effect of throwing away valuable
information to get position lock.
This change reverses the clearBuffer and changes the logic so if time
is not set it will not be broadcast.
Fixes https://github.com/meshtastic/firmware/issues/7771
Fixes https://github.com/meshtastic/firmware/issues/7750
TinyGPS Fix Quality has this information:
0 - fix not available,
1 - GPS fix,
2 - Differential GPS fix (values above 2 are 2.3 features)
3 = PPS fix
4 = Real Time Kinematic
5 = Float RTK
6 = estimated (dead reckoning)
7 = Manual input mode
8 = Simulation mode
the previous Airoha code would allow quality >0 , which includes
estimated positions. These wouldn't be passed through to the mesh
due to other checks, but would affect the Airoha GPS_FIX_HOLD_TIME
calculations.
Changes the calculation to 1 >= quality <=5 .
As determined by @b8b8 , enabling NAVIC meant the more modern L5 GPS
signal was not used (L1 GPS is always available).
NAVIC, India's GNSS, probably provides the best coverage in India and
the neighbouring region. However, outside of NAVIC's coverage area, L5
GPS is highly desirable.
This patch amends the AG3335-family GPS configuration to enable L5 GPS
coverage by default. If the Lora region is set to India or Nepal,
NAVIC will be enabled instead.
