I just tested random SSD1315 from aliexpress and it worked without any modifications.

This reverts commit eec5481.

Starting the 10.x branch from 9.x

…up Node States message at 1 Hz.

… GPIO pins.

…t 1Mbps!

Set Node Status to real data.

Pass __FPU_PRESENT=1U (matching the CMSIS header) instead of =1,
and route vendor include directories through SYSTEM_INCLUDE_DIRECTORIES
(-isystem) so upstream vendor headers can't trigger -Wmacro-redefined.

Also remove the INAV-specific #ifndef __FPU_PRESENT guard from
stm32h743xx.h — it is not present in the upstream ST CMSIS and is
no longer needed with the cmake fix in place. This avoids confusion
on future HAL/CMSIS updates.

cortex-m4f.cmake is shared by F4 and AT32 targets only (not H7).
The F4 CMSIS header defines __FPU_PRESENT as 1 (without U suffix),
so passing =1U from cmake triggers -Wmacro-redefined which is fatal
with -Werror. Revert to =1 to match the header value.

cortex-m7.cmake: back to __FPU_PRESENT=1 (matches F7 headers, avoids
-Wmacro-redefined on F7 targets where headers define '1' not '1U')

stm32h743xx.h: restore #ifndef guard so the header's own 1U define
is skipped when cmake has already set __FPU_PRESENT=1, avoiding
any token mismatch between cmake '1' and header '1U'

…-v1.3.3-update

Update STM32F7xx HAL to v1.3.3

SET_ALT fixes and tightening

Remove STM32F4 HAL as it is not used in the project.

…13.0-update

Update STM32H7xx HAL to v1.11.6 and CMSIS Device to v1.10.7

…sb-intercept-calc

ADSB selected the most dangerous vehicle by CPA calculation

…t tests

Validate that the packet contains exactly timerCount*2 bytes after the
count byte. Previously a truncated packet was silently accepted, applying
fewer overrides than requested and returning a plausible-but-wrong preview.

Add unit tests covering:
- pwmCalculateAssignment() save/restore invariant (hardware flags and
  timer override modes are identical before and after the simulation)
- TIMER_HW_MAX guard (out stays zero-initialised when count exceeds limit)
- QUERY payload validation (truncated, oversized, and well-formed payloads)

…-assignment-api-v2

output assignment: firmware-authoritative MSP2 READ/QUERY API

Removed commented-out timer definitions and adjusted the order of active timer definitions.

add a new target BLADE_F4

…type

fix datatype of timerCount

SSD1315 OLEDs also seem to work

Reset free PIDs if coefficients are changed via CLI or MSP