Request review before merge official branch#5
Open
Apaisal wants to merge 350 commits intoEmOne:developfrom
Open
Request review before merge official branch#5Apaisal wants to merge 350 commits intoEmOne:developfrom
Apaisal wants to merge 350 commits intoEmOne:developfrom
Conversation
Some pointer arguments were updated to `const` to indicate immutability when no modification is necessary. Signed-off-by: Robin Mueller <robin.mueller.m@gmail.com> Signed-off-by: Yasushi SHOJI <yashi@spacecubics.com>
Update Waf 2.1.4.
NEW IN WAF 2.1.4
----------------
* Prevent subprocesses from disrupting the main build process
through inherited stdin
* Fix a regression that prevents displaying help from
custom commands
Signed-off-by: Yasushi SHOJI <yashi@spacecubics.com>
"abc" without NUL terminator is 3 bytes instead of 4 bytes. We don't want to send NUL terminator on CSP packet. Signed-off-by: Yasushi SHOJI <yashi@spacecubics.com>
"abc" without NUL terminator is 3 bytes instead of 4 bytes. We don't want to send NUL terminator on CSP packet. Signed-off-by: Yasushi SHOJI <yashi@spacecubics.com>
csp_if_zqmhub.c doesn't properly handle return values but depends on assert(). For now, make compiler silent by marking them by __maybe_unused. Signed-off-by: Yasushi SHOJI <yashi@spacecubics.com>
Building samples and examples with CMake has been tedious, requiring
manual specification of each sample to build. This commit introduces a
new option, `CSP_BUILD_SAMPLES`, which simplifies the process.
With this option, you can now build all samples using the following
commands:
cmake -B builddir -DCSP_BUILD_SAMPLES=1
ninja -C builddir
This enhancement streamlines the workflow for building all samples.
Signed-off-by: Yasushi SHOJI <yashi@spacecubics.com>
Teach buildall.py how to build samples and examples using `CSP_BUILD_SAMPLES`. Signed-off-by: Yasushi SHOJI <yashi@spacecubics.com>
"mps2_an385" was a old board name. With Zephyr Hardware Model v2 (aka. HWMv2), it's been renamed to mps2/an385. Signed-off-by: Yasushi SHOJI <yashi@spacecubics.com>
Add Python 3.12 for the Zephyr build test. Signed-off-by: Yasushi SHOJI <yashi@spacecubics.com>
When building libcsp with Zephyr RTOS, the compilation flags are handled by the Zephyr Build System. Adding extra flags here causes more harm than good. This commit removes these flags completely. Signed-off-by: Yasushi SHOJI <yashi@spacecubics.com>
The CSP_SYSTEM_NAME variable is introduced to specify the kernel name since CMAKE_SYSTEM_NAME does not recognize FreeRTOS. This approach ensures flexibility for adding support for new kernels in the future.
Removed an unnecessary trailing newline from the CMakeLists.txt file. No functionality has been changed. Signed-off-by: Yasushi SHOJI <yashi@spacecubics.com>
Replaced the use of CMAKE_SYSTEM_NAME with CSP_POSIX, CSP_FREERTOS, and CSP_ZEPHYR across all CMakeLists.txt files. This simplifies conditional branching by using these variables instead of relying on explicit checks for CMAKE_SYSTEM_NAME. Signed-off-by: Yasushi SHOJI <yashi@spacecubics.com>
Previously, Zephyr RTOS used "Zephyr" as the value for CMAKE_SYSTEM_NAME. To standardize behavior, CMAKE_SYSTEM_NAME is now left unchanged, defaulting to "Generic" for systems unknown to CMake. With CSP_SYSTEM_NAME introduced in recent commits, this update moves Zephyr RTOS to use CSP_SYSTEM_NAME instead. Signed-off-by: Yasushi SHOJI <yashi@spacecubics.com>
`CMAKE_SYSTEM_NAME` is typically uppercase or capitalized, but our FreeRTOS port uses lowercase. To handle this inconsistency, this commit makes `CSP_SYSTEM_NAME` checks case-insensitive by converting the value to lowercase before comparison. This change ensures compatibility with various naming conventions. Signed-off-by: Yasushi SHOJI <yashi@spacecubics.com>
Document the steps to build Python bindings using CMake. Include necessary options and additional steps for enabling the routing table and setting `PYTHONPATH` for usage. Also, add an additional headline level for improved section structure. Signed-off-by: Yasushi SHOJI <yashi@spacecubics.com>
Correct the PYTHONPATH path in the INSTALL document. The path incorrectly used `build` instead of the consistent `builddir` used throughout the document. Signed-off-by: Yasushi SHOJI <yashi@spacecubics.com>
Include the missing file `src/interfaces/csp_if_tun.c` in `wscript`. This file was introduced in commit `70e43adec06dea2` but was not added to the waf build system. Note that it was already added to CMake in commit `2f9adbfeb79ec6f`.
GitHub Actions recently updated `ubuntu-latest` from `ubuntu-22.04` to
`ubuntu-24.04`, causing documentation builds to fail with the
following error:
Exception occurred:
File "/home/runner/.local/lib/python3.12/site-packages/clang/cindex.py",
line 4179, in get_cindex_library
raise LibclangError(msg)
clang.cindex.LibclangError: /usr/lib/llvm-14/lib/libclang-14.so: cannot open
shared object file: No such file or directory. To provide a path to libclang
use Config.set_library_path() or Config.set_library_file().
The full traceback has been saved in /tmp/sphinx-err-424nf1r4.log, if you
want to report the issue to the developers.
Please also report this if it was a user error, so that a better error
message can be provided next time.
A bug report can be filed in the tracker at <https://github.com/sphinx-doc/
sphinx/issues>. Thanks!
gmake[2]: *** [CMakeFiles/docs.dir/build.make:73: html/index.html] Error 2
gmake[1]: *** [CMakeFiles/Makefile2:87: CMakeFiles/docs.dir/all] Error 2
gmake: *** [Makefile:91: all] Error 2
Error: Process completed with exit code 2.
To resolve this, the workflows are rolled back to explicitly use
`ubuntu-22.04`.
Signed-off-by: Yasushi SHOJI <yashi@spacecubics.com>
Remove MacOS from GitHub Actions workflows as it is not currently supported. Signed-off-by: Yasushi SHOJI <yashi@spacecubics.com>
Building libcsp sometimes fails with specific GCC versions due to changes in the default warning levels. This commit adds a matrix to test builds with GCC versions 10 through 14. The "exclude" strategy is used instead of "include" because, despite GitHub Actions supporting matrix expansion, I could not make it work in this case. Signed-off-by: Yasushi SHOJI <yashi@spacecubics.com>
Update datalen to use `size_t` since `zmq_msg_size()` returns `size_t` instead of `int`. Ensure strict type correctness. Signed-off-by: Yasushi SHOJI <yashi@spacecubics.com>
Replace local declarations of csp_id_strip() and csp_id_setup_rx() with `#include <csp/csp_id.h>` to use the existing declarations from the header file. Signed-off-by: Yasushi SHOJI <yashi@spacecubics.com>
Update datalen to use `size_t` since `zmq_msg_size()` returns `size_t` instead of `int`. Ensure strict type correctness. Signed-off-by: Yasushi SHOJI <yashi@spacecubics.com>
The `zmq_msg_data()` function does not return a `const` value. While using `const` is generally good practice for unmodified data, this change is a preparatory step for an upcoming commit that modifies the `rx_data`. Signed-off-by: Yasushi SHOJI <yashi@spacecubics.com>
Add `csp_id_prepend_fixup_cspv1()` and `csp_id_strip_fixup_cspv1()` to address a CSPv1 protocol issue where the ZMQ interface incorrectly uses host byte order. These functions ensure compatibility with such misbehaving interfaces. Modify `csp_id1_prepend()` and `csp_id1_strip()` to accept an optional `network_byte_order` flag for explicit byte-order control. These remain internal APIs, with the new fixup functions being their only current users. Enhance compatibility by supporting interfaces that incorrectly use host byte order in the CSPv1 protocol. Signed-off-by: Yasushi SHOJI <yashi@spacecubics.com>
Update `csp_id1_prepend()` and `csp_id1_strip()` to clearly indicate
that the second parameter (`cspv1_fixup`) is used solely for handling
the fixup byte order quirk.
Simplify the endian conversion logic by always converting to
big-endian unconditionally:
id1 = htobe32(id1_raw);
Subsequently, apply the fixup only if needed. This makes the intention
of converting to/from big-endian clearer.
No functionality should be changed.
Signed-off-by: Yasushi SHOJI <yashi@spacecubics.com>
csp_id_prepend_fixup_cspv1 and csp_id_strip_fixup_cspv1 add
unnecessary code for configurations that do not require them. Enable
these functions only when CSP_FIXUP_V1_ZMQ_LITTLE_ENDIAN is non-zero.
This change reduces code size slightly compared to previous commits:
text data bss dec hex filename
80228 2952 15016 98196 17f94 builddir/libcsp.so
79759 2936 15016 97711 17daf builddir/libcsp.so
Signed-off-by: Yasushi SHOJI <yashi@spacecubics.com>
Reintroduce the destination address as the first byte for ZMQ interface with CSPv1. The destination address byte was removed in commit `b3217c0f48d47192f2f` for CSPv2. This commit re-adds it when using CSPv1. Code changes include helper functions to handle destination address addition/removal for CSPv1. Signed-off-by: n-log-gos <nicolas.logioia@orbitalsystems.de> Signed-off-by: Yasushi SHOJI <yashi@spacecubics.com>
Address a CSPv1 protocol compatibility issue in the ZMQ interface. In libcsp v1, the ZMQ interface incorrectly used host byte-order for the CSP ID, leading to interoperability issues. Libcsp v2 corrected this by enforcing big-endian encoding for the CSP ID, regardless of the host system’s endianness. But, this mismatch prevents libcsp v2 ZMQ interfaces from communicating with nodes running libcsp v1 over CSPv1 protocol. These commit adds compatibility between libcsp v2 and nodes running libcsp v1 with CSPv1 protocol over ZMQ, ensuring consistent header handling across systems. Signed-off-by: Yasushi SHOJI <yashi@spacecubics.com>
Previously, posix instances using CAN did not accept L2 and L3 broadcast due to HW filtering
…sc (#824) Previously, posix instances using CAN did not accept L2 and L3 broadcast due to HW filtering Co-authored-by: Johan De Claville Christiansen <johan@space-inventor.com>
The issue originates in the fact that L3 data is overriding L2 data in the csp_packet header. This is resolved on the libcsp/libcsp/develop branch already, by removing the union.
We should consider actually returning an error. Instead we still just override the last available route, which seems to have been the previous intention. Co-authored-by: edvard <edvard@space-inventor.com>
When it isn't exactly 40 characters long
Added `csp_zmqhub_<add/remove>_filters()` to change promisc settings after init
This is used for multicast-like operations, similar to the original concept of CAN
From old times when the RX buffer could fluctuate more, we used to reserve two times window in the rx_queue, but after optimization we find that only 1 time is required.
These are given funky names by gcc, like csp_buffer_pool.2, which make them difficult for static analysis tools to find and put into context. Moving to file scope is only a slight scope increase which can be tolerated for the benefits of static analysis.
This might have to be 64-bit to get full nanosecond network time later. But for now, just to ensure it builds again, we add as uint32
* Remove packet copy when transmitting CSP packets Credits to @MErdbruegger for suggesting this improvement. * inlined helper function and fix a few spellings --------- Co-authored-by: Troels Jessen <troels@space-inventor.com>
Credits to @MErdbruegger for suggesting this improvement.
and fix a few spellings
Let users have access to iface iterator
This should not really be of any concern of libcsp
This is faster in most cases size is protected with if as the user could have set invalid packet length.
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
20 participants
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Review before merging from the official