* check if all frames are laggy
* apply tolerance
Co-authored-by: Kurt Nistelberger <kurt.nistelberger@gmail.com>
old-commit-hash: 4ec1c7e614a2ca3fbf38f302a82b87d112d06b02
* add mypy config matching precommit
* use local mypy, add files to config
* excludes too
* fix config
* pylint is sad now... did it get updated?
* fix typing hints
* ignore
* this should be a regexp
* mypy doesn't like Deque despite inheriting MutableSequence
* more excludes
* Revert "pylint is sad now... did it get updated?"
This reverts commit 250c632f18ecb3d33ffb931e15425f9314a0964b.
old-commit-hash: 4e310b807fb1ccdde4b2078468fab46e5324dfa2
* ev has different scales
* fix initial gradient
* fix highlight weirdness
* try smooth set of gains
* delay
* add gain idx
* oops
* set different min dc
Co-authored-by: Comma Device <device@comma.ai>
old-commit-hash: c7a0f23b45603fb8e7718234f92284854866924e
* camerad: log image sensor in camera states
* for all cams
* bump cereal
* revert that
Co-authored-by: Comma Device <device@comma.ai>
old-commit-hash: 7352e6d9409dcec6719717dcb805b97565648533
* change LSM6DS3TR(-c) gyroscope and accelerometer to interrupt
* add pthread for linking
* add interrupt collector thread to fetch in parallel to non interrupt based sensors
* change get_event interface to return true on successful read
* fetch sensor interrupts via gpiochip
* avoid sending empty messages (interrupt only, non interupt magn can leave a gap in the orphan block)
* add verifier script to sensor interrupts (sensor_data_to_hist.py)
* add/update sensord testsweet (test_sensord.py)
* add poll timed out check
* unexport interrupt gpio pins
* gpiochip on both edges, but skip falling edge if rising edge is detected, this is handled in the sensor as the status flag is checked if new data is availble
* add test to sensord to verify 100Hz interrupt frequency
* add sensor shutdown/low power mode functionality on sensord exit
* relax test, will be readded in the splitup PR
Co-authored-by: Kurt Nistelberger <kurt.nistelberger@gmail.com>
old-commit-hash: 84a3c355e5fdb7bb2f50847f925e894eecfa8918
* start porting OS04C10
* something
* is the pll why this doesn't work
* understand plls, still doesn't work
* work
* weee something
* working?
* timing
* encoderd works with less cameras
* 1928
* flip image
* color balance
* adjust black level
* extra height maybe
* cleanups
* clean up sensor i2c
* sensors
* approx 20fps
* remove mirror
* 50ms, no fsin yet
* remove imx390 support
* some attempt at AE
* linearize the response
* OX changes (#25697)
* organize
* wb
* best effort but probably messed up by artifacts
* multicam sync
Co-authored-by: Comma Device <device@comma.ai>
* clean up
* more
* fix multiexposure
* start porting OS04C10
* something
* is the pll why this doesn't work
* understand plls, still doesn't work
* work
* weee something
* working?
* timing
* encoderd works with less cameras
* 1928
* flip image
* color balance
* adjust black level
* extra height maybe
* cleanups
* clean up sensor i2c
* sensors
* approx 20fps
* remove mirror
* 50ms, no fsin yet
* remove imx390 support
* some attempt at AE
* linearize the response
* OX changes (#25697)
* organize
* wb
* best effort but probably messed up by artifacts
* multicam sync
Co-authored-by: Comma Device <device@comma.ai>
* clean up
* more
* fix multiexposure
* unbroken AE
* oops
* sort of good AE
* both should work 1
* open then init
* fix debayer
* rename
* fix test build
Co-authored-by: Comma Device <device@comma.ai>
Co-authored-by: ZwX1616 <zwx1616@gmail.com>
Co-authored-by: Adeeb Shihadeh <adeebshihadeh@gmail.com>
old-commit-hash: aea1a0d5e684daf7682703174233390b0cf79c83
* also read params to determine if we want manual gains
* check for larger than 0
Co-authored-by: Comma Device <device@comma.ai>
old-commit-hash: 7ecc0409bd767392074a37cdd0b7e44ed9953296
* first pass at usb power mode cleanup
* fix build
* a sneaky one
* little more
* fix build
* bump pnada
* remove that
* power monitoring cleanup
* fix tests
* bump submodules
old-commit-hash: 28cb1897cb8be80b35b1ce7573066fb36b01d2d7
