The last version is already much better than the original, mostly because it doesn't call
CriticalSectionLocker, which has a ton of overhead due to debug tracking. But this version is
another 4 instructions / 12 bytes shorter. Does as much match in 32-bit land as possible, and
avoids math operations that span 64-bits (i.e. either operate on the lower half or upper half, but
not both). The result is only 3 instructions not including the necessary 4 loads (ptr to now, now,
ptr to upper, upper), 1 store (upper), and 1 return/branch.
Disallow scaled_channel for float. I can't think of a reason to allow it, and it gets in the way
of rounding.
Add separate template param to setTable; let the compiler sort out whether assignment can happen
between TElement and VElement without forcing them to be the same at function call time.
Remove some implicit C array to pointer conversions in Map3D by adding
(), *, and & where appropriate. This allows us to remove getBinPtr.
Call interpolate3d from Map3d so there's no code duplication.
* scale map Y axis
* allow different row/col types
* scaled channel detector
* interpolation
* looks like this actually works
* tests, no manual scaling
* comment
* wiring
* static
* implementation
* comment
* unit tests happy
* guard
* guard smarter, not harder
* guard smarter not harder
* guard even smarter, not even harder
* don't need that
* big pch energy
* put back ramdisk stub
* tests are happy
* h743 nucleo
* kinetis
* I love deleting code!
* make stepper happy
Co-authored-by: Matthew Kennedy <makenne@microsoft.com>
* post cranking factor limits
* sensible defaults
* but not for tests
* add pointer version of get bin
* test via map3d
* put that back for a minute
* new impl
* fix order
* rename and fix table orientation
* flip the table to the correct orientation
* dead cleanup, rename
Co-authored-by: Matthew Kennedy <makenne@microsoft.com>
* these start themselves now
* dead
* configure ports
* start serial
* dead
* minimal pins
* simulator logging works
* why not enable lua
* speeeeeed
* start lua
* spammy print
* helper
* use it miata
* do type safe copy
* miata vvt
* etb
* default iat correction
* magic is afoot here
* hunchback
* misc
* neon
* neon
* aspire
* citroen
* neon
* missed one
* dead
* festiva
* remove andreyism
* very dead
* no need for copy of that function
* cleanup
* finish it off
* one more logger ptr
* fwd decl
* Revert "one more logger ptr"
This reverts commit a21fb0087ddc748978d716db2710800cf26e437b.
* unused
Co-authored-by: Matthew Kennedy <makenne@microsoft.com>
* minor cleanup
* enable mailboxes
* priority
* implement new logger
* more cleaning
* signature
* remove debug
* put the assertion back in
* remove debugging
* spelling
* doy
* tweaks & comments
* cleaning
* size_t
* guard more
* test build now
* needs more ram until we can remove LogginWithSTorage
* hunt for memory
* bootloader
* unused
* stub simulator
* it would help to get the signature right
* geez kinetis only has 64k ram
* more guarding because kinetis
* that's now very legal and very cool
* templatify
* s
* force null terminate
* probably fix
* const
* write a test
* disable sensor chart for now
* hide SC ui
* oh nope that breaks many things
This reverts commit c3c1bb324fb484c3d9cc44b7715c234bc0392e1e.
* missed one
* reset after scheduleLogging
* we can't print out the full message because it contains commas which break parsing
* force terminate long buffers
* let's see how much memory we have to play with
* a lot, is the answer
* real thread name
* shrink this for now before its deleted
* turn that back on
* biiig comment
Co-authored-by: Matthew Kennedy <makenne@microsoft.com>
* disable sensor chart
* type signature, guards
* we can't log the message as that confuses the parser
* dead
* this syntax was technically illegal
* more
* turn it back on
* remove config
* api
* implement
* ui
* java ui
* start cleaning
* disabled feature means don't print that it's disabled
* cli
* usages
* printMsg -> scheduleMsg
* this is alllll dead now
* no crlf please
* all hope is lost if you manage to hit this line
* tests
* tests
* ok we did actually need that part
* sim
* d
* kinetis
* it did ifdef, not if
Co-authored-by: Matthew Kennedy <makenne@microsoft.com>
* template table2d
* assertions
* oops
* tests
* test bins too
* more assert
* format
* explicitly handle and test NaN
Co-authored-by: Matthew Kennedy <makenne@microsoft.com>
* round rpm instead of truncating
* efiround is expensive
* Revert "efiround is expensive"
This reverts commit e5690f89e1b1988aacf5ced1f024d576465a7cd6.
* round is better than rintf
* testing
* it works now?!
* comment
Co-authored-by: Matthew Kennedy <makenne@microsoft.com>
* switch to CriticalSectionLocker
* that's just about all
* clean up last usage
* include hpp for sim
* need the cpp wrappers in the makefile too
* include dir
* include
* lambda
* default configs, fix build
* config, gauge, UI
* more digits
* change menu text
* fix that while I'm here
* doesn't need extra word
* todo note
* tests
* last test
* missed a config
* update java
* store target lambda
* more java
* format
* hopefully enough java for CI to pass
* add
* remove old
* remove more old
* makefile
* fix init
* patch up other stuff
* actually fix it this time
* raw output channels
* gauges, UI
* scaled channel
* fix field ordering
* remove old impl
* add cpp wrapper to os_access
* clean up os_access
* remove touchTimeCounter
* new impl
* put comment about lock free impl
* include cpp wrappers for BL
Co-authored-by: Matthew Kennedy <makenne@microsoft.com>
* s
* science
* set pin mode
* turn stuff off so it fits
* filtering maybe
* filtering actually works
* generate filter parameters internally
* shorter window
* guard behind enable flag
* use checked in filter
* add biquad reset
* tracing
* const
* exec order
* do it from a thread
* smaller buffer, comment
* configure with header
* only for proteus
* oops
* unused
* not needed
* guards
* pin config
* don't need that include
* precook filter steady state
* define sample rate
* config enable switch
* improve biquad
* cleanup
* add filtering to subscriptions
* config sensors
* comment
* doesn't need to be that fast
Co-authored-by: Matthew Kennedy <makenne@microsoft.com>
* configurable stoich ratio
* afr table always in gas afr
* default
* typo
* use defined ratio
* do it in config too
* missed a spot
* oops wrong one
Co-authored-by: Matthew Kennedy <makenne@microsoft.com>
Matthew Kennedy
rusefi
rusefillc
Scott Smith
racer-coder
GitHub build-firmware Action
Andrey G
David Holdeman