[DO NOT MERGE] LMC: Mavlink ULog rate control support #612

This would prevent client to stop working if agent restarts in middle of connection initiation. If entity creation fails the client stop connecting and waiting for next ping response from agent before next connection attempt.

Mavlink start supports '-i <x>' flag to read IP address from MAV_<x>_REMOTE_IPn params.

-T is the command line argument for LD, it should not be in the script name

The type of main (entry point) is now relevant, as the entry point is no longer explicitly casted to main_t

For some reason, the type of the linker defined symbols was changed. The reasoning was "they typically represent pointers" which is absolutely not true at all. They typically define memory locations, but they can be whatever. It is just a way to provide symbols from the linker to the executable, which is done via relocation symbols.

- Allocate a separate wait semaphore for each process (one in PROTECTED_BUILD). Allocate the semaphore in "REGISTER" ioctl, free it in "UNREGISTER" if process exits - Duplicate the user side call events to kernel memory to make them accessible by kernel at all times. Keep track of the allocated items in "freelist" queue - Replace fixed size callout queue with an "inflight" queue Signed-off-by: Jukka Laitinen <[email protected]>

…versions These give errors on some compilers Signed-off-by: Jukka Laitinen <[email protected]>

…tion

microdds_client dds topic generator to support CamelCase topic names

Mavlink msg generator (pymavlink) does not handle mavlink2 message extensions properly. Receiver does not zero-fill the receiver buffer, so due to the extension zero suppression featyure the last "suppressed" fields may contain garbage data. HITL_SENSOR message last field is sensor_id, so it is explicitely set zero to avoid issues.

dds_topics.yaml has support for adding optional rate parameter for each publication topics e.g. Limit SensorCombined topic pub rate to 10 Hz: - topic: /fmu/out/sensor_combined type: px4_msgs::msg::SensorCombined rate: 10

Not usable from user space

The ioctl parameter type was wrong, need to give the handle as well.

The shutdown hook is run when the system is shut down, what we want is to call hrt_stop() when the process exits -> use atexit().

- Pop the entry from any queue when fiddling around with it, this prevents two threads accessing the same data via the queue - When looking for any item in any queue, the search should be performed with BOTH user entry, and the handle provided. This is because user entries are ambiquous (user virtual addresses) and thus the kernel handle (which is unique) is needed - Fix HRT_CANCEL on the kernel side - Fix HRT_UNREGISTER on the kernel side - Remove potential double booking in callout_queue and callout_inflight. - Do not free items from the freelist, as they do not belong to any thread specifically (as they have been freed) - Remove kernel_entry from the hrt_call structure, the same information can be done by comparing user entry + handle

The only thing needed from there was the definition of struct builtin_s. The builtin functions are not meant to be used this way anyhow, it is supposed to be a way to inject user-space code into the kernel (the builtin list is _copied_ into kernel memory). The board_ioctl module is is kernel-space only. Also, remove the usage of the "FAR" keyword, we use processors from this millennium that do not use memory segmentation into 64K chunks... The real reason for this change is that CONFIG_BUILD_KERNEL does not allow setting CONFIG_BUILTIN (makes sense, injecting user code into the kernel is not feasible _at all_ with address environments).

Allow starting a process with its full path i.e. strip the path from argv[0] and use the base name to find the entry point.

Do not link the soft float math library, it is not needed for anything and it confuses the linker which does relocations to the FP instructions

Signed-off-by: Jukka Laitinen <[email protected]>

…gnal Sigaction lets us specify that only the wanted signal is handled with the callback Signed-off-by: Jukka Laitinen <[email protected]>

Signed-off-by: Jukka Laitinen <[email protected]>

In protected NuttX targets, uORB resides in both kernel- and userspace. However, mmap is not a kernel space API, and must not be directly used from kernel space, if MMU mappings are in use. Why? Because when the kernel side uORB implementation calls mmap, where would the mapping go ? Obviously it would be mapped to the user shared memory virtual addresses, which is a no-no. Fix this by implementing a very dirty, temporary workaround for kernel.

datawriters were not recreated in case of reconnection to agent. Pub&Sub instances are cleared after connection lost is detected, so they are forced to be recreated when connection is regained.

- Add/Modify UAVCAN module parameters to allow finer behavior control - Expand UAVCAN mem pool block size

…c node server initialization

…rspace MTIME register is mapped by MMU directly to user space; thre is no need to do the heavy IOCTL for it Signed-off-by: Jukka Laitinen <[email protected]>

Instead of using a const value for USRIO_START, which fails when dynamic linking is used (i.e. CONFIG_BUILD_KERNEL), obtain the location of MTIME via syscall to the kernel side HRT.

…ed with different algorithms This can be used e.g. for bootloader/table of contents creation or in crypto backend code Signed-off-by: Jukka Laitinen <[email protected]>

…orithms Signed-off-by: Jukka Laitinen <[email protected]>

…oesn't exist This is only to resolve linking issues with bootloaders which don't use the random Signed-off-by: Jukka Laitinen <[email protected]>

… use the PEM for signing in cryptotools.py This harmonizes the generation and usage of keys. All keys can be simply generated as PEM files with openssl, and public key parts can be injected in DER format. Signed-off-by: Jukka Laitinen <[email protected]>

In case where keystore backend is real HW, it may need deinitializing in case it is used both in the bootloader and in PX4. Add the interface function for that. Signed-off-by: Jukka Laitinen <[email protected]>

Signed-off-by: Jukka Laitinen <[email protected]>

Makes it a bit simpler to select correct linker script(s) for each build type.

…ices This is more or less hardcoded module, which can be used as basis for more advanced functionality later. Signed-off-by: Jukka Laitinen <[email protected]>

Signed-off-by: Jukka Laitinen <[email protected]>

…st in rcS This is done only because rpnet is currently slow to start. Revert this change if/when the rpnet startup issues are fixed Signed-off-by: Jukka Laitinen <[email protected]>

Instead of starting the worker when a process is created, start it when a call to a hrt service happens. This drops ~30 unnecessary threads from the system, as most user processes use hrt only to obtain the system time.

… name for "sdcard" in board configuration Signed-off-by: Jukka Laitinen <[email protected]>

Usage: Define CONFIG_UXRCE_DDS_TOPICS_YAML="<name of new yaml file>" into board specific .px4board config file. Create/copy dds topics yaml file into board dir src/ directory. e.g. boards/px4/fmu5x/src/dds_topics_low_bw.yaml Build log shows the used yaml file name: "microdds_client: use dds topics yaml: 'xxxxx'"

Signed-off-by: Jukka Laitinen <[email protected]>

assert() does not work in px4 because it makes a release build that defines NDEBUG for all source files, which in turn disables assert() tests.

Signed-off-by: Jukka Laitinen <[email protected]>

The code that checks for CDC ACM connection needs to be an actual process (module) as: - It cannot run inside the NuttX kernel, it must be capable of starting mavlink - It cannot run inside "the userspace" either, as with CONFIG_BUILD_KERNEL there are several "userspaces" / processes. Do it by: - Forklifting sercon_main() and serdis_main() from libapps, the reason is that with kernel build the symbols are NOT packed into libapps and thus are inaccessible. - Create wrapper for builtin_exec() that can do file exec via posix_spawn()

On UAVCAN drone, sometimes, due to randomness in installation, the motor spinning direction is reverted thus require reversed current and voltage from ESC. This is needed so that only the absolute current value is passed to PX4 system, preventing trigger of esc current < 0 warning

…n ones Some architectures might not define long double (128-bits) but use the 64-bit double for such arithmetics instead. This fixes a build error due to missing name##l symbol.

This allows the driver to start also on 400kHz bus, and is according to the component spec Signed-off-by: Jukka Laitinen <[email protected]>

Signed-off-by: Jukka Laitinen <[email protected]>

ADS1115 is first commanded to start sampling and after 1/4 of the sample period the single channel sample is read out. However, when using ScheduleOnInterval this fails if the sample reading is blocked for long enough. The resulting _average_ sample rate is still correct (80Hz) but the time between triggering the new sample on the ADC and reading it out can be anything. If the call to ADS1115::RunImpl() happens too soon after starting the new sample, reading it will obviously fail. Fix this by scheduling the sample reading _after_ the sampling has been triggered on the ADC. This ensures the sample is ready the next time when ADS1115::RunImpl() is executed.

Some modules might want it, removes link error for uavcan: libuavcan/libuavcan/include/uavcan/node/global_data_type_registry.hpp:215: more undefined references to `__dso_handle'

sem_wait() can be interrupted if the task receives a signal, however the blockinglist implementation depends on blocking until the semaphore can be obtained.

The userspace work queue is not started automatically by the system, so start it when dispatching the shutdown command. It would be best if there was an init function for shutdown.cpp but this "lazy" init is just fine too.

Geo needs libm, which is automatically added by prebuild_targets

- uORB depends on uorb_msgs (get_orb_meta) - px4_platform_init() should not call param_init, if param module is disabled - px4_protected_layers re-organize the link_library commands into one - reboot does not need to explicitly add px4_platform, it gets it anyway

Move user libraries from the "start/end-group" clause to a proper target_link_libraries so CMake knows the dependency correctly.

Create a kernel version of the C++ user library. This is necessary as otherwise some calls will be directed to the user space.

This is phase 0.1 for enabling PX4 in NuttX kernel mode, this builds the simplest possible configuration i.e. NuttX + nsh + px4 userspace init, but no userspace processes yet. One result is a bootable ROMFS which is placed into the kernel binary and mounted into /sbin by the board initialization logic. The kernel knows the mount point and starts the user space from /sbin/init, which is the nsh shell. A second result are the NuttX apps, placed and installed into /bin The next step is to start building the modules into processes

This commit adds support to build px4 modules into individual processes. This is done by using a common main.cpp file, which calls the user space init function and then the module entry point. The entry point is provided by a command line CDEF -Dentry=module_main The processes are installed into /bin and this is mounted as from a ROMFS The dependencies are still a huge mess, and will be fixed later...

The environment is cleared for px4 processes, but the initial process needs to have the environment intact. Without this patch the environment is wiped for nsh (the init process) via: px4_platform_init -> px4::WorkQueueManagerStart(); -> px4_task_spawn_cmd() -> clearenv() nsh _NEEDS_ the environment to be intact in kernel mode.

…ules The proxy process "launch_kmod" is used to start the kernel side modules, to keep backwards compatibility with nsh builtin apps, use symbolic links to achive the same functionality. rc.sysinit is used for this, it is perfect for setting up low level infrastructure needed by the init process, as it is loaded _before_ the rcS / init.d script.

malloc/realloc/free/mallinfo etc don't work correctly when executing NuttX kernel code, as they are used for user heap. For kernel modules, use the kernel versions instead.

Also, add mechanism to turn the daemon off, for simple apps that don't need it. One example is launch_kmod.

Add the startup to rc.sysinit, so the module daemons are all running before calling the actual init.d/rcS script

Do not start wq:manager for every process during init, start it only when it is needed. This drops ~20 useless threads from the system.

The problem is that prebuild_targets pulls in nuttx_xx (user version) unconditionally, for both user and kernel. This in turn pulls in nuttx_c (user version) which causes a link issue: NuttX/nuttx/libs/libc/libc.a(pthread_attr_init.o): in function `.LANCHOR0': NuttX/nuttx/include/nuttx/pthread.h:118: multiple definition of `g_default_pthread_attr'; NuttX/nuttx/sched/libsched.a(pthread_create.o):/home/ville/wpc_toolchain/workspace/px4-firmware/platforms/nuttx/NuttX/nuttx/include/nuttx/pthread.h:118: first defined here This temporarily patches the issue, the real fix is to remove nuttx_xx from prebuild_targets (which causes a ton of new issues to resolve...)

… logic Bind the daemon and task_create/delete to the correct files

Disable all code regarding hardfaults (ARM specific trap) when they are not used / needed. Fixes build issue with NuttX upstream: In file included from /src/modules/logger/log_writer_file.cpp:46: /src/lib/systemlib/hardfault_log.h:144:3: error: conflicting declaration 'typedef struct stack_t stack_t' 144 | } stack_t;

Fixes build error: arm-none-eabi/bin/ld: NuttX/nuttx/drivers/libdrivers.a(pipe_common.o): in function `pipecommon_close': nuttx/drivers/pipes/pipe_common.c:381: undefined reference to `circbuf_uninit'

use vehicle torque and thrust setpoint for rover control

Must wait for romfs.copy.stamp, otherwise the init script might get wiped.

- Subscribe to poll uorbs instead of sleeping px4_usleep to allow more accurate timing. - Change rate control to be controlled by uorb subscribe instead of custom rate reduction in microdds_client.

To support several ads1115 modules publishing to different adc_report instances Signed-off-by: Jukka Laitinen <[email protected]>

Signed-off-by: Jukka Laitinen <[email protected]>

These are simple and stateless applications -> NO_DAEMON saves 1 process and socket from being created for nothing.

Events have a global, system-wide sequence number, which must be handled atomically, (fetching and incrementing the sequence AND sending the event to uORB must be atomic). Currently in FLAT mode, only one instance of this sequence number exists, so it is OK to have it in px4_platform. However, in PROTECTED mode px4_platform is instantiated both in kernel- and user spaces, which makes two instances of this sequence number, which causes problems in the mavlink event handling logic. When mavlink receives and handles events, it expects that: - The sequence numbers arrive in order (seq n is followed by n+1 etc) - It increments by 1 - There is only one instance of the sequence number In PROTECTED mode this is violated, as the kernel and user sequence numbers run freely on their own. This patch fixes the issue by moving the event backend to the kernel and by providing user access to it via ioctl.

In kernel mode there might be quite a lot of waiters, so increase the callback list size substantially. Leave the other modes unaffected.

This is a (temporary) regression mitigation after NuttX 12.0 release, where priority inheritance (pi) was disabled by default. This changes the default behavior, as previously pi was enabled by default. Patch it by enabling it within the px4_sem_init() call.

Signed-off-by: Jukka Laitinen <[email protected]>

The head/tail semaphores are not used as lock but rather as resource counters and thus relate more as signaling semaphores. Disable PI for them. I run my code with CONFIG_DEBUG_ASSERTIONS=y and the kernel panics due to the semaphore having no holder, disabling PI fixes this.

WorkItemSingleShot::_sem is a signaling semaphore, disable PI for it. Set CONFIG_DEBUG_ASSERTIONS=y and the kernel panics due to the semaphore having no holder, disabling PI fixes this.

The pointer types should be compatible, still, for OCD purposes change them.

When the semaphore is released and re-instantiated, do proper cleanup in g_sem_pool.reserve(). Also remove GlobalSemPool and GlobalLock::set and replace them with free() which just destroys the mutex.

NuttX kernel now handles mappings into kernel memory, so the 1 page mapping limitation is unnecessary.

If px4_mmap fails, do not proceed and cause random system crash later, instead produce meaningful error messages from uorb_start().

Temporary change, fix when semaphores work properly in kernel mode

Remove _cb_handle before calling unregister_callback, to ensure no one uses the old stale handle before it is cleared.

…ing cb There seems to be a race condition where the callback thread is signaled but the sub is not registered when the orb_callback thread executes. This patches such errors, but the root cause of the race condition is not known. Regardless, this sanity check is not wrong.

Signed-off-by: Jukka Laitinen <[email protected]>

Keep this target building still Signed-off-by: Jukka Laitinen <[email protected]>

The "px4 bootloader" is no longer supported for these Signed-off-by: Jukka Laitinen <[email protected]>

…on and smooth path following

…i airframe

There is no one using the position_setpoint_triplet messages in MC side currently and because no users there, it causes SROS ACL errors about uxrce-agent has no permissions to write to topic.

So we can cherry-pick changes for nuttx-apps. This change maybe worth reverting at some point (make nuttx-apps point to some release again).

Kernel mode issues are fixed, mostly CPU usage, almost all of these are patches from upstream.

…nto device_boot_info To avoid clashing variable names in px4 Signed-off-by: Jukka Laitinen <[email protected]>

Signed-off-by: Jukka Laitinen <[email protected]>

Under OSX, user group (staff) is in the low tens (20) and conflicts with system groups of Ubuntu. '-o' allows to create group with same GID but different name.

Add a generic px4 exec function to spawn processes either via: - builtin list if configuration does not use file apps - posix_spawn if configuration uses file apps

… build Move the mavlink shell to its own process and launch it via px4_exec. The task is very simple and does not need anything from the MavlinkShell class, it is defined as a static function anyway in order to be passed as pointer-to-function to px4_task_spawn. The shell application is now compiled as its own module, and added to the builtin list or list of module binaries, and can / is started via px4_exec instead.

Icicle boards is not really ssrc HW. In addition, we want to add all "ssrc" boards via a separate submodule. So, move icicle away from "ssrc" Signed-off-by: Jukka Laitinen <[email protected]>

If SIGNING_ARGS environment variable is defined, build script uses its value as a signing key location. Also a new submodule saluki-sec-scripts is added. If SIGNING_ARGS is not defined test keys are used under Tools/saluki-sec-scripts/test_keys/ saluki-sec-scripts contains also signing tools and second set of keys (custom_keys), which are used in FPGA secure boot builds. Signed-off-by: Tero Salminen <[email protected]>

Signed-off-by: Jukka Laitinen <[email protected]>

also remove some old unused workflows: - .github/workflows/tiiuae-gzebo-data.yaml - .github/workflows/tiiuae-pixhawk-artifact-publish.yaml

These can be used to transfer version information accross the system (ver, mavlink...) without link dependency to version libraries Also this allows easy publishing the autopilot versions to ROS2 over uXRCE-DDS Signed-off-by: Jukka Laitinen <[email protected]>

Signed-off-by: Jukka Laitinen <[email protected]>

#if __PX4_NUTTX does nothing, change it to #if defined()

There is a potential problem with sub->call() and sub->[un-]register(), when the callback is executing, if someone drops the callback registration, it will result in a resource leak and a crash. This does not happen in flat mode, as the uORB::DeviceNode::lock protects both, but when a callback thread is used, there is a race condition which is fixed by this patch.

Signed-off-by: Jukka Laitinen <[email protected]>

Makes analyzing core dumps a whole lot easier

This is needed if using authenticated encryption/decryption (e.g. AES-GCM) Signed-off-by: Jukka Laitinen <[email protected]>

Add renew_nonce and and decrypt_data interfaces, and also ioctl's for memory protected builds. Stub these functions in sw_crypto driver. the functions are currently not used but a stub is needed in order to build properly. Add implementation for signature_check in PX4Crypto class, there is already an implementation in the backend, but the function is missing from the PX4Crypto interface. Signed-off-by: Jukka Laitinen <[email protected]>

Signed-off-by: Jukka Laitinen <[email protected]>

Combination of multiple dependabot updates: build(deps): bump scipy build(deps): bump actions/upload-artifact from 2.pre.preview to 3.1.3 build(deps): bump docker/setup-buildx-action from 2 to 3 build(deps): bump stefanzweifel/git-auto-commit-action from 4 to 5 build(deps): bump codecov/codecov-action from 1 to 3 build(deps): bump actions/cache from 2 to 3 Signed-off-by: dependabot[bot] <[email protected]>

Signed-off-by: Jukka Laitinen <[email protected]>

Use different size input / output buffers for testing, add some preliminary RSA test code Signed-off-by: Jukka Laitinen <[email protected]>

Signed-off-by: Jukka Laitinen <[email protected]>

The size input argument for monocypher crypto_xchacha20_ctr should be the plaintext message length. The promise of the interface is, that the call to encrypt_data updates the ciphertext message length after the call succeeds. The crypto should check that the output buffer length (cipher length) is large enough to contain the encrypted data. Fix these issues; these have gone unnoticed for a long time since the interface has been only used by logger, and passing the same size for both in and out. Signed-off-by: Jukka Laitinen <[email protected]>

Signed-off-by: Jukka Laitinen <[email protected]>

Add a systemcmd to check R&D feature flags for the board Signed-off-by: Jukka Laitinen <[email protected]>

… allowed by R&D flags Signed-off-by: Jukka Laitinen <[email protected]>

Place them inside the board configuration instead Signed-off-by: Jukka Laitinen <[email protected]>

Temporary disable tamper protection Signed-off-by: Jukka Laitinen <[email protected]>

Signed-off-by: Jukka Laitinen <[email protected]>

…_LEAVE macros These macros were only used to protect write / copy functions. Use them also for callback registration. Signed-off-by: Jukka Laitinen <[email protected]>

…erate function Signed-off-by: Silvan Fuhrer <[email protected]>

Signed-off-by: Silvan Fuhrer <[email protected]>

clear dataman file and init initial value on creation

Rework of dataman

A version update is needed since the dataman is showing errors if data doesn't exist or if it is wrongly stored. This will force default data to be initialized.

tests dataman will trigger errors and run_nsh_cmd.py is not written to detect only when a test fails. This is properly tested with run_tests.py

…anges This avoids the need to regularly access dataman for checking.

Avoids unnecessary dataman accesses

…dataman' As this test is expected to contain errors.

This can happen e.g. when the item is cleared.

Instead of also when geofence/safe points updated. This prevents reporting multiple times.

…sion is not valid, use the state from the last mission validation.

…e dataman state. If at the last powercycle one mission was uploaded, the counter in dataman was 1. On the next powercycle the mavlink mission counter was reset to zero and on first mission upload updated to 1 again. Other modules check, if the mission was changed based on the counter, like the mission.cpp loaded the mission counter from the dataman. On a new mission, the comparison of the counters failed, because both were the same value even if the mission was completely different.

Fetch changes to enable compiler optimized memcpy et al

Signed-off-by: Jukka Laitinen <[email protected]>

… generation This manifests only with the new uORB library, which always returns 0 updates available when data is not valid Signed-off-by: Jukka Laitinen <[email protected]>

With the latest NuttX, for some reason printfs don't match the regexps as before (probably due to \n character in the end) Signed-off-by: Jukka Laitinen <[email protected]>

Signed-off-by: Jukka Laitinen <[email protected]>

- build custom signed px4 - package fpga binaries to px4 container - bootloader is now included in fpga bitstream so no need to package it anymore - build from added tag: v1.14.0-*

In protected builds, put all SubscriptionCallbacks into process specific lists, and in the callback thread execute them in a loop. Set the priority of the callback threads to the priority of nav_and_controllers+1, this should be always enough. It is currently very difficult to find the optimal value. Signed-off-by: Jukka Laitinen <[email protected]>

Signed-off-by: Jukka Laitinen <[email protected]>

This fixes the error: etc/init.d-posix/rcS: 39: [: Illegal number: Signed-off-by: Julian Oes <[email protected]>

This fixes the error: /px4_sitl_etc/init.d-posix/rcS.sim_udp: 26: [: =: unexpected operator

CA_ROTOR and PWM_MAIN_FUNC params added to init.d-posix/4400 airframe to make actuator outputs working for sitl

Bumps [docker/login-action](https://github.com/docker/login-action) from 2 to 3. - [Release notes](https://github.com/docker/login-action/releases) - [Commits](docker/login-action@v2...v3) --- updated-dependencies: - dependency-name: docker/login-action dependency-type: direct:production update-type: version-update:semver-major ... Signed-off-by: dependabot[bot] <[email protected]>

Pull latest changes + get rid of merge commits

Signed-off-by: Jukka Laitinen <[email protected]>

…ates not set If update not called, it may lead to infinite loop in case check_for_updates not set and there are new messages in ulog_stream_acked uorb.

Signed-off-by: Jukka Laitinen <[email protected]>

An EEPROM behind and i2c media is not 100% failure proof. But any errors in e.g. mtd_params reading can cause catastrophic failures in flight. Mitigate any possible errors in px4 mtd_params reading by adding crc code to every page on EEPROM In addition: - Read every block into a temporary buffer before copying to clients buffer; this fixes some issues with memory protected builds when writing directly to user buffers from i2c driver interrupt handler - If a read request fails, don't return any partially filled buffer to user Signed-off-by: Jukka Laitinen <[email protected]>

Like __dso_handle, this symbol must NOT be pulled from the user libc. The __cxa_atexitatexit() functions in kernel are no-ops as we don't support shared library destructors.

…_for_updates not set" This reverts commit f3a8744. Updating and sometimes ignoring ulog_stream_acked reduces the load, but may cause log to be corrupted due to missing definition data. Need to revert the patch to avoid broken log files.

Commits on Dec 8, 2023

Update pf crypto submodule

jnippula committed Dec 8, 2023

Configuration menu

View commit details

Copy full SHA for 53fb2b0

Browse repository at this point

Copy the full SHA

53fb2b0 View commit details

Browse the repository at this point in the history

Commits on Dec 11, 2023

HITL: mavlink instance start with params

jnippula committed Dec 11, 2023

Configuration menu

View commit details

Copy full SHA for a32204f

Browse repository at this point

Copy the full SHA

a32204f View commit details

Browse the repository at this point in the history

Commits on Jan 10, 2024

Update nuttx submodule

pussuw committed Jan 10, 2024

Configuration menu

View commit details

Copy full SHA for a84ac3c

Browse repository at this point

Copy the full SHA

a84ac3c View commit details

Browse the repository at this point in the history

Commits on Jan 11, 2024

nuttx/toc: Add image config message to ToC

pussuw committed Jan 11, 2024

Configuration menu

View commit details

Copy full SHA for 3eb5581

Browse repository at this point

Copy the full SHA

3eb5581 View commit details

Browse the repository at this point in the history

Commits on Jan 22, 2024

Updated FPGA to sha-4230961

kordets committed Jan 22, 2024

Configuration menu

View commit details

Copy full SHA for d783993

Browse repository at this point

Copy the full SHA

d783993 View commit details

Browse the repository at this point in the history

Commits on Jan 23, 2024

Update saluki-v2 submodule

pussuw committed Jan 23, 2024

Configuration menu

View commit details

Copy full SHA for c1ba103

Browse repository at this point

Copy the full SHA

c1ba103 View commit details

Browse the repository at this point in the history

[DO NOT MERGE] LMC: Mavlink ULog rate control support #612

Are you sure you want to change the base?

[DO NOT MERGE] LMC: Mavlink ULog rate control support #612

Commits on Oct 26, 2023

Commits on Oct 27, 2023

Commits on Oct 30, 2023

Commits on Oct 31, 2023

Commits on Nov 1, 2023

Commits on Nov 2, 2023

Commits on Nov 6, 2023

Commits on Nov 7, 2023

Commits on Nov 8, 2023

Commits on Nov 9, 2023

Commits on Nov 14, 2023

Commits on Nov 15, 2023

Commits on Nov 21, 2023

Commits on Nov 23, 2023

Commits on Nov 27, 2023

Commits on Nov 29, 2023

Commits on Nov 30, 2023

Commits on Dec 1, 2023

Commits on Dec 5, 2023

Commits on Dec 8, 2023

Commits on Dec 11, 2023

Commits on Dec 12, 2023

Commits on Dec 13, 2023

Commits on Dec 14, 2023

Commits on Dec 19, 2023

Commits on Dec 20, 2023

Commits on Dec 21, 2023

Commits on Jan 9, 2024

Commits on Jan 10, 2024

Commits on Jan 11, 2024

Commits on Jan 12, 2024

Commits on Jan 15, 2024

Commits on Jan 16, 2024

Commits on Jan 18, 2024

Commits on Jan 19, 2024

Commits on Jan 22, 2024

Commits on Jan 23, 2024

Commits on Jan 25, 2024

Commits on Jan 30, 2024