Read cached icons completely, validate the RGB565 format and exact payload size, and discard corrupt cache entries so a later launch can retry acquisition.
Key the local cache by package name and version so a catalog update cannot silently reuse an older icon.
Assisted-by: OpenAI Codex:gpt-5.6-sol
Signed-off-by: aviralgarg05 <gargaviral99@gmail.com>
Keep the list behavior documentation consistent with the 20-pixel threshold used by the tested implementation.
Assisted-by: OpenAI Codex:gpt-5.6-sol
Signed-off-by: aviralgarg05 <gargaviral99@gmail.com>
Use nxstore-owned framebuffer and input configuration symbols instead of relying on an unrelated LVGL demo configuration.
Load the manifest for the installed version before launching it, validate that it matches the installed database entry, and pass its recorded arguments to posix_spawn(). Check spawn-attribute setup errors as well.
Add the shared supervisor-bar height header to the nxstore change itself so the branch builds independently and framebuffer applications can follow the required reserved-strip contract.
Assisted-by: Codex:gpt-5
Signed-off-by: aviralgarg05 <gargaviral99@gmail.com>
Remove the dependency on an ESP-specific DHCP global. Retry catalog synchronization for a bounded period only while the network stack reports readiness-related errors, allowing the same frontend to work with Wi-Fi, Ethernet, and other boards.
Keep the LVGL timer serviced between attempts so the interface remains responsive while the network comes up.
Assisted-by: Codex:gpt-5
Signed-off-by: aviralgarg05 <gargaviral99@gmail.com>
Review/hardening findings from the companion nxpkg PRs:
- Own framebuffer/input device paths instead of borrowing
CONFIG_EXAMPLES_LVGLDEMO_FBDEVPATH/INPUT_DEVPATH from an unrelated
example app: new CONFIG_SYSTEM_NXSTORE_FBDEVPATH/INPUT_DEVPATH
Kconfig string options (defaulting to /dev/fb0 and /dev/input0).
- g_index moves from a static struct to a heap-allocated
FAR struct pkg_index_s * (pkg_zalloc()), with a "Not enough memory to
load the catalog." UI fallback if the allocation fails.
- nxstore_launch() now loads the specific installed version's manifest
via pkg_metadata_load_manifest_path() instead of combining a
rollback-selected version with whatever the current catalog entry
happens to describe for that name - those can disagree after a
rollback if the catalog has since moved on. Also passes through the
installed manifest's launch_args/launch_argc (previously always
argv[1] = NULL).
- title_text buffer sized PKG_NAME_MAX + PKG_VERSION_MAX + 5 instead of
a fixed 96, fixing a real compiler truncation warning.
- README.txt moved to the companion NuttX documentation PR rather than
shipping user-facing documentation as an in-tree README.
Bugs found bringing this up on real hardware:
- LV_EVENT_LONG_PRESSED could fire for a touch that was actually
driving a scroll of the app list: if a drag starts slowly enough
that the long-press timer (400ms) elapses before the finger crosses
the scroll-lock distance, LVGL hasn't committed the gesture to
scrolling yet and still delivers the long-press event, silently
uninstalling whatever card the touch happened to land on. Ignore the
long-press if this input device is currently attributed to scrolling
any object (lv_indev_get_scroll_obj()).
- Tapping an installed-app card to launch it, while a different
install was in progress elsewhere in the list (or another app was
already running), was allowed through unconditionally - install_worker()
auto-launches its own package once done, and letting a second launch
through independently meant whichever one finished last silently
overwrote g_running, leaving the other process alive, unsupervised,
and drawing into the same shared framebuffer with no way to close it
from this UI again. Both the direct-launch and fresh-install paths in
install_btn_event_cb() now refuse (with a toast) if g_running.active
or g_active.manifest indicate another app is already running or about
to be.
- nxstore_is_installed() only checked the package *name*, not which
version was actually on disk - an older installed version showed as
plain "Installed" identically to a current one, and tapping it
silently launched the stale payload with no update indication or
action. Add nxstore_is_up_to_date() (compares the installed version
against the catalog's latest manifest) and a third status_bar color
(in addition to not-installed/up-to-date) plus a "Update available -
tap to update" subtitle for the mismatch case; tapping such a card
now goes through the install path (which fetches and auto-launches
the newly installed version) instead of the direct-launch path.
- lv_indev_set_scroll_limit() was set to 255 (copied from examples/
lvgldemo/lvgldemo.c, whose own touch-drift mitigation this file
reused), requiring a nearly-full-screen drag before a touch was even
recognized as a scroll gesture. Unlike that demo, this screen's app
list is scrolled constantly, and a threshold that large read as
broken/laggy scrolling rather than drift protection. Lowered to 20,
enough to reject typical touch-driver jitter while still recognizing
a real scroll almost immediately; momentum stays off (scroll_throw
0), which is what the dual-launch/scroll-lock fixes above actually
depend on, not the gesture-start threshold.
Also adds nxstore_load_icon(): best-effort loads and caches a
package's optional icon (manifest->icon) as a raw RGB565 image LVGL
can render with no decoder (this board has no PNG/JPEG decode
capability), falling back to the existing colored-circle-plus-glyph
rendering on any failure (no icon set, download failed, corrupt/
oversized file) so a bad icon never blocks a package from being
listed or installed.
Assisted-by: Claude:claude-sonnet-5
Signed-off-by: aviralgarg05 <gargaviral99@gmail.com>
Add nxstore, an LVGL-based frontend for nxpkg (system/nxpkg): lists
packages from the local index, installs/launches the selected one,
and supervises whatever it hands the screen to.
Card-based app list (populate_app_list()): each row shows name,
version, and description/status, install/launch state reflected via
icon glyph and color (LV_SYMBOL_DOWNLOAD/PLAY, accent/success color),
and a sliding-segment progress bar during install (no real byte-level
progress is available from pkg_install(), so this reads as
'actively working' without fabricating a percentage). Explicit
LV_STATE_PRESSED styling on every tappable row/button, since no LVGL
theme is loaded and a tap would otherwise give no visual feedback at
all.
Supervisor screen (build_run_screen()/nxstore_enter_running_screen()):
a launched app (e.g. a game that owns /dev/fb0 directly, not just
another LVGL client) gets a dedicated screen with a name label and a
Close button, confined to the border region the launched app's own
scaled/centered framebuffer output never draws into, so switching
back to it doesn't fight over pixels with whatever the app already
put in the framebuffer.
Close/reap handling (close_running_app_event_cb()/
nxstore_poll_running_app()): sends SIGTERM and polls waitpid(WNOHANG)
for the launched pid, but explicitly also treats waitpid() returning
ECHILD as 'already gone' rather than 'still running' - both the
close-button handler and the passive per-loop poll independently race
to reap the same child, so whichever one loses that race must not
spin forever waiting for a wait() that can now never succeed. Requires
the target app to install its own SIGTERM handler to exit cleanly
(this is why there is no generic force-kill fallback here: an
earlier version of this code called task_delete() when SIGTERM wasn't
reaped quickly enough, which was found on real hardware to hang the
entire board - not just the one task - when it landed mid
framebuffer/heap access on this flat-memory build).
Toast notifications (nxstore_toast()) provide a transient, unmissable
confirmation for install/uninstall/launch outcomes and app-closed
events, additive to the durable per-row subtitle text rather than a
replacement for it.
nxstore's own boot-time catalog sync waits (bounded, 15s) on
g_wifi_dhcp_ret before attempting a network fetch, since Wi-Fi
association completing doesn't imply DHCP has - an HTTP fetch
attempted in that window fails with -ENETUNREACH even though the
link itself is already up, indistinguishable from being genuinely
offline without this wait.
Assisted-by: Claude:claude-sonnet-5
Signed-off-by: aviralgarg05 <gargaviral99@gmail.com>
Record a per-boot token and owner PID in every package, installed-database, and synchronization lock. A contender now keeps a lock held while its recorded owner task is alive, regardless of unreliable FAT modification timestamps.
Reclaim locks from exited owners or earlier boots immediately, while retaining timestamp-based migration handling for legacy empty lock files. This prevents a just-created live lock from being mistaken for a decades-old stale file on targets whose mounted filesystem clock does not match CLOCK_REALTIME.
Assisted-by: OpenAI Codex:gpt-5.6-sol
Signed-off-by: aviralgarg05 <gargaviral99@gmail.com>
Shorten the installed-database lock description and centralize pkg_sync() cleanup, as requested in review. Replace repeated protocol literals with named constants and preserve errno before cleanup calls can overwrite it.
Store the synchronized catalog source in the catalog itself so the catalog and its relative-artifact base are committed atomically. Continue reading the former sidecar format for upgrade compatibility, and normalize array-form catalogs before adding the private source field.
Compare numeric version prefixes without strtol() overflow and retain lexical comparison of suffixes.
Assisted-by: Codex:gpt-5
Signed-off-by: aviralgarg05 <gargaviral99@gmail.com>
- pkg_sync(): index.jsn and repo.url were updated as two independent
atomic writes with nothing serializing the pair against a second,
concurrent pkg_sync() call - each write stayed internally
consistent, but the pair didn't, so one sync's index could end up on
disk next to a different sync's source URL. Add a dedicated sync
lock (pkg_repo_acquire_sync_lock(), mirroring the existing installed-
db lock's blocking-retry-with-stale-reclaim pattern) around the
whole read-fetch-write sequence. Also renew the lock's mtime as data
actually arrives (pkg_repo_sink(), via a new renew_lock_path field
threaded through pkg_acquire_source()) rather than only stamping it
once at acquire time - a lock acquired once and then measured
against a fixed 10-minute staleness window could otherwise be
reclaimed mid-download on a large-enough file over a slow-enough
link, even though the download was still genuinely in progress.
pkg_reclaim_stale_lock() (renamed from pkg_install_reclaim_stale_lock,
made public) is shared between both lock kinds rather than
duplicated.
- pkg_install_prune_oldest_version(): deleted the pruned version's
on-disk payload directory before the updated installed database was
even durably saved. If pkg_metadata_save_installed() subsequently
failed, the payload was already gone but the last successfully-saved
instpkg.jsn could still list that version as installed. The victim
version is now handed back to the caller (threaded through
pkg_install_add_version()/pkg_install_update_installed()) so
pkg_install() can defer the actual directory removal until after the
save succeeds.
- pkg_metadata_version_token_cmp(): two version tokens with equal
numeric prefixes (e.g. "1a" and "1b", both parsing as 1) compared as
equal instead of falling back to a lexical comparison of what
follows the number, contradicting this function's own documented
behavior and silently treating genuinely different versions as the
same one. Compare the non-numeric remainder lexically when the
numeric prefixes match instead of falling through.
Assisted-by: Claude:claude-sonnet-5
Signed-off-by: aviralgarg05 <gargaviral99@gmail.com>
Fix real correctness/security gaps found during review, on top of the
network sync and CLI completion work:
- pkg_install(): commit the installed database before writing the
current/previous pointer files, and before advancing transaction
state past ACTIVATED. Those are recovery bookkeeping and convenience
mirrors respectively - once the installed database itself is
durably committed, a failure to refresh either one must not trigger
the failure-cleanup path, which could otherwise delete a payload the
database now legitimately points at. Also stop treating an existing
version directory as newly created when reinstalling an
already-installed version, so a failed reinstall can't delete a
working install.
- pkg_uninstall()/pkg_rollback(): acquire the per-package install lock
in addition to the installed-db lock, drop the entry from the
authoritative database first, and only then remove payload files -
a crash between those two steps can leave orphaned files, which are
reclaimable, but never leaves the database pointing at a payload
that's already gone.
- pkg_sync(): stage each sync to a PID-qualified temp filename instead
of a single fixed name, so concurrent sync calls can no longer race
on the same staging file. Return real negative errno codes instead
of EXIT_SUCCESS/EXIT_FAILURE, matching the rest of this API; pkg_main.c
maps that back to a process exit code at the CLI boundary.
- pkg_metadata_parse_installed_entry(): validate that name/current/
previous/every entry in versions[] are safe path components, and that
current (and previous, if set) actually appear in versions[] - a
corrupted or tampered installed-packages database can no longer
reference a nonexistent version or smuggle a path-traversal sequence
through a field this code already trusted implicitly.
- pkg_store_write_all()/pkg_repo_sink(): treat a zero-byte write() as
-EIO instead of looping on it silently.
- Removed the malloc()-falls-back-to-kmm_malloc() logic in pkg_malloc/
pkg_zalloc/pkg_realloc/pkg_free entirely - it required tracking which
allocator owned a given pointer via kmm_heapmember(), which is
specific to this target's flat, single-heap memory model and not a
sound general application API. These are now plain wrappers around
malloc/calloc/realloc/free.
- Added pkg_metadata_load_manifest_path(), so a caller (e.g. the
nxstore GUI frontend, launching an installed package) can load the
manifest actually recorded for a specific installed version, instead
of combining a rollback-selected version with whatever the current
catalog happens to describe for that package name - those can
disagree after a rollback if the catalog has since moved on.
- Storage root default changed from /tmp/nxpkg to /var/lib/nxpkg,
following the conventional persistent application-data location
instead of a path whose own name suggests non-persistent storage. A
board without persistent storage mounted at /var, or that wants a
different location (e.g. an SD card), still overrides this via
CONFIG_SYSTEM_NXPKG_ROOT as before.
- Moved system/nxpkg/README.txt into the companion NuttX documentation
PR rather than shipping user-facing documentation as an in-tree
README.
Assisted-by: Claude:claude-sonnet-5
Signed-off-by: aviralgarg05 <gargaviral99@gmail.com>
Fill in most of what the initial nxpkg slice deferred: network sync
and artifact acquisition over plain HTTP (verified via SHA-256), an
install rewrite that supports network sources and reclaims state left
by an interrupted previous install, and wiring update/remove/rollback/
available into the CLI.
Harden the package path against untrusted input along the way: bounded
memory-safety helpers and explicit size limits throughout, storage
read/write hardened against partially-written files, path-traversal
rejection in manifest name/version before they reach the filesystem,
and a fix for a lost-update race on the shared installed-packages
database (two concurrent installs of different packages could
otherwise silently clobber each other's recorded state).
Add an optional manifest icon field for the nxstore GUI frontend to
consume, raise PKG_INDEX_MAX now that the in-memory index is
heap-allocated, and document the repository layout and local server
setup in system/nxpkg/README.txt.
Also fixes a real build break: pkg_runtime_compat() unconditionally
referenced CONFIG_ARCH_BOARD, a Kconfig string symbol with no default
clause under ARCH_BOARD_CUSTOM, so it is left entirely undefined
rather than defined-but-empty on a custom board - falls back to
CONFIG_ARCH_BOARD_CUSTOM_NAME instead. Routes pkg_error()/pkg_info()
through syslog rather than stdio, since neither is visible to a
supervisor with no attached console.
Assisted-by: Claude:claude-sonnet-5
Signed-off-by: aviralgarg05 <gargaviral99@gmail.com>
Add the "curl" NSH command: a command-line HTTP client built on top of
the netutils webclient library. It implements a subset of the real curl
options: GET and POST (and other methods via -X), custom request headers
(-H), a raw request body (-d, including -d @file), multipart/form-data
file uploads (-F name=@file), saving the response body to a file (-o)
and verbose output (-v). HTTP only (no HTTPS).
Signed-off-by: Jorge Guzman <jorge.gzm@gmail.com>
Parse the filedump path token into the existing PATH_MAX-sized buffer
before reading the optional offset and size arguments. This bounds the
write without constructing a scanf format string at runtime.
Signed-off-by: Old-Ding <35417409+Old-Ding@users.noreply.github.com>
Parse command filename tokens into PATH_MAX-sized buffers before
reading optional raw audio parameters. This bounds nxplayer and
nxrecorder input without runtime-built scanf formats.
Signed-off-by: Old-Ding <35417409+Old-Ding@users.noreply.github.com>
The RFB ServerInit message carries the desktop name as a length-prefixed field. vncviewer caps the copied name to fit conn->name, but it must still consume the remaining bytes from the socket when the advertised name is longer than the local buffer.
Drain the unused suffix so the next RFB message is read from the correct boundary. Reuse the same discard helper for other skipped RFB payloads.
Signed-off-by: Old-Ding <ai.neo.ae86@gmail.com>
Check the service argv array bound before reading the current entry in the debug dump loop. A full argument array may not have an in-array NULL terminator, so the old condition could read one entry past the array while CONFIG_SYSTEM_NXINIT_DEBUG is enabled.
Signed-off-by: Old-Ding <35417409+Old-Ding@users.noreply.github.com>
Keep the default zero CPU value when the procfs load file cannot be read, and trim only the newline that was actually present. This avoids parsing uninitialized stack data in fillcpu and avoids writing before the showinfo CPU buffer when fgets returns no data.
Signed-off-by: Old-Ding <35417409+Old-Ding@users.noreply.github.com>
The O_RDOK/O_WROK aliases have been removed from fcntl.h. Replace
all remaining O_RDOK usage with O_RDONLY in the apps repository.
Signed-off-by: Xiang Xiao <xiaoxiang@xiaomi.com>
Add the missing module metadata for the executable ELF helpers used by the package fixture flow so the generated artifacts describe their target and type consistently.
Also fix the existing embedlog spelling issue that is picked up by the current apps check, keeping this branch clean under CI.
Signed-off-by: aviralgarg05 <gargaviral99@gmail.com>
Add the initial nxpkg command, metadata and store handling, the local install/list path, and the repository export helper.
Keep the current flow scoped to local artifacts and target-qualified repository entries so it remains usable as an incremental MVP while follow-up features land separately.
Signed-off-by: aviralgarg05 <gargaviral99@gmail.com>
Use fopencookie() to attach the popen fd and shell pid to the returned FILE stream instead of copying FILE into the popen container.
Keep the upstream dpopen()/dpclose() implementation as the process and descriptor backend, and make pclose() close the cookie-backed stream directly.
Fixes#2937.
Signed-off-by: Nightt <87569709+nightt5879@users.noreply.github.com>
Introduce a new CONFIG_UORB_FORMAT Kconfig option to control whether
uORB format strings are compiled in. UORB_LISTENER, UORB_GENERATOR,
and DEBUG_UORB all select UORB_FORMAT automatically, so format strings
are included when any of these features are enabled.
This replaces the previous approach of guarding format strings with
CONFIG_DEBUG_UORB, which prevented uorb_listener from displaying
sensor data when debug output was disabled.
Signed-off-by: hanzhijian <hanzhijian@zepp.com>
listener_update() only prints topic data when delta_generation is
non-zero (i.e., new data arrived since last check). In listener_top,
the first call adds objects to the list, and subsequent calls only
print if new data was published between iterations. This results in
listener_top -T showing only the header with no topic rows.
Fix by always printing the current topic state in listener_update,
setting frequency to 0 when no new data arrives. This ensures
listener_top displays all topics every iteration.
Fixesapache/nuttx-apps#3202
Signed-off-by: hanzj <hanzhijian@zepp.com>
Fix argument parsing in init_parse_arguments() to properly handle
multiple quoted arguments like 'echo "arg1" "arg2"' by skipping
quote characters after processing them.
Signed-off-by: fangpeina <fangpeina@xiaomi.com>
The init process has blocked all signals, spawned services would
inherit that mask. This could cause services to miss important
signals like SIGTERM during graceful shutdown.
Signed-off-by: wangjianyu3 <wangjianyu3@xiaomi.com>
Pending all signals(SIGCHLD) when ppoll() is not invoked to
avoid race conditions.
Case reproduction
Set examples/hello as a service that exits immediately after startup.
```init.rc
on boot
start hello
service hello hello
restart_period 0
```
Log - without this patch:
# Service hello only restarts about 100 times, ppoll is not woken up
# after the hello process with PID 119 exits.
[ 4.391274] [ 2] [ 0] init_main: service 'hello' pid 118 exited status 0
[ 4.401423] [ 2] [ 0] init_main: started service 'hello' pid 119
Log - with this patch:
# ppoll() can still be woken up normally after tens of thousands of
# restarts of service hello in stress test.
[ 268.447747] [ 2] [ 0] init_main: service 'hello' pid 34503 exited status 0
Signed-off-by: wangjianyu3 <wangjianyu3@xiaomi.com>
Any string ending with whitespace passed to init_parse_arguments()
could cause the parser to advance past the string boundary and read
unintended memory content.
- " echo "A" \0& echo "B" should be parsed
as a command with two argvs instand of five.
- "command arg " may lead to uncertain results.
Signed-off-by: fangpeina <fangpeina@xiaomi.com>
reap_process() referenced an undeclared identifier 'wtatus' on
the WIFSIGNALED branch (typo of 'wstatus'). Some toolchains then
flagged a -Wmaybe-uninitialized on the surrounding wstatus use.
Correct the typo so WIFSIGNALED/WTERMSIG operate on the actual
wstatus value returned by waitpid().
Signed-off-by: fangpeina <fangpeina@xiaomi.com>
action.c uses clock_gettime(CLOCK_MONOTONIC, ...) but did not
pull in <nuttx/clock.h> directly, which fails to build on
configurations where the header is not transitively included.
Add the missing #include.
Signed-off-by: fangpeina <fangpeina@xiaomi.com>
When ETC_ROMFS is disabled to reduce the bin size, we can provide the init.rc
file via a pseudo-file in the boards/vendor directory. For example:
- CONFIG_ETC_ROMFS=n
- CONFIG_PSEUDOFS_FILE=y
- CONFIG_DISABLE_PSEUDOFS_OPERATIONS=n
```C
FAR const char *init_rc =
"on init\n"
" start console\n";
"service console sh\n"
" restart_period 100\n";
int fd = open("/etc/init.d/init.rc", O_WRONLY | O_CREAT);
/* ... */
ssize_t n = write(fd, init_rc, strlen(init_rc) + 1);
/* ... */
close(fd);
```
The last character '\0' in the file content will be treated as a new line,
and the number of parsed parameters will be zero (abnormal, there should be
at least one keyword).
Signed-off-by: wangjianyu3 <wangjianyu3@xiaomi.com>
Empty lines in init.rc caused parsing to fail with -EINVAL because init_parse_arguments() returns 0 for empty strings, triggering the 'argc < 1' error path in init_action_parse() and accessing uninitialized argv[0] in init_service_parse().
Fix by skipping empty lines and lines containing only whitespace before attempting to match section keywords or calling parser callbacks.
Fixesapache/nuttx-apps#3513
Signed-off-by: hanzj <hanzhijian@zepp.com>
When NSH_LIBRARY is not available, dpopen()/popen() can still execute
commands by splitting the command string by whitespace and calling
posix_spawnp() directly. Shell syntax (pipes, redirects, globbing)
is not supported in this mode.
Add CONFIG_SYSTEM_POPEN_MAXARGUMENTS (default 7) to control the
argv array size for the no-shell path.
Remove the hard dependency on NSH_LIBRARY from SYSTEM_POPEN so the
feature can be used in minimal configurations without a shell.
Signed-off-by: cuiziwei <cuiziwei@xiaomi.com>
Add dpopen()/dpclose() as the descriptor-based counterpart of
popen()/pclose(), analogous to how dprintf() relates to fprintf().
dpopen() returns a raw file descriptor instead of a FILE stream,
avoiding the stdio.h dependency for callers that only need an fd.
Refactor popen() as a thin wrapper: dpopen() + fdopen() + FILE
container. All pipe creation and process spawning logic now lives
in dpopen.c.
Also remove the hard dependency on NSH_LIBRARY from SYSTEM_POPEN.
When NSH is available, commands are executed through sh -c with full
shell syntax support. When NSH is not available, commands are split
by whitespace and executed directly via posix_spawnp().
Add CONFIG_SYSTEM_POPEN_MAXARGUMENTS (default 7) to control the
argv array size for the no-shell path.
Signed-off-by: cuiziwei <cuiziwei@xiaomi.com>
Micro XRCE-DDS expects the application to supply open/close/write/read
callbacks for the wire transport. Add a single dispatcher that
registers the selected backend via rmw_uros_set_custom_transport(),
plus two backends:
transport/microros_transport_udp.c
BSD-socket UDP backend. Opens an AF_INET/SOCK_DGRAM socket,
resolves CONFIG_MICROROS_AGENT_IP/PORT, connect()s it, and uses
send/recv. The socket fd is stashed in uxrCustomTransport->args
so no module-level state is needed.
transport/microros_transport_serial.c
termios serial backend. Opens CONFIG_MICROROS_SERIAL_DEVICE with
O_RDWR|O_NOCTTY|O_CLOEXEC, switches to raw 8N1 at
CONFIG_MICROROS_SERIAL_BAUD, and uses poll/read/write. Fd is
again stashed in uxrCustomTransport->args.
Both backends are mutually exclusive at compile time via Kconfig;
the dispatcher selects which set of callbacks to register. Wired
into both the legacy Make build (CSRCS in system/microros/Makefile)
and the CMake build (separate microros_transport static library
with the transport directory only exposed to its INTERFACE).
Signed-off-by: Arjav Patel <arjav1528@gmail.com>
libmicroros is installed by colcon under include/<pkg>/<pkg>/file.h
for ament-packaged headers (e.g. rcl, rmw, rosidl) and the flat
include/<pkg>/file.h for a few others (e.g. rclc). A single
-I include/ therefore resolves <rclc/rclc.h> but not <rcl/rcl.h>.
Enumerate every immediate subdirectory of include/ in addition to
include/ itself so both layouts resolve.
Switch the library hand-off from LDLIBS to EXTRA_LIBS. The sim
target's final link pulls EXTRA_LIBS, not LDLIBS, so the previous
form left rcl/rclc/rcutils symbols unresolved.
Signed-off-by: Arjav Patel <arjav1528@gmail.com>
micro_ros_lib/toolchain.cmake.in initialised C and CXX flags with
-D'__attribute__(x)=' to silence GCC attributes the upstream ament
build is not aware of. The strip also removed __gnu_inline__ and
__always_inline__ from NuttX's libc string.h inlines, so each
micro-ROS translation unit emitted external definitions of strcmp,
strcpy, strlen and friends. Sim final link then failed with ~50
multiple-definition errors.
Drop the define from both flag init lines so the attributes survive
and the inlines collapse as intended.
Signed-off-by: Arjav Patel <arjav1528@gmail.com>
Fix a missing space in the Kconfig help text for SYSTEM_LZF:
'Enable theLZF' → 'Enable the LZF'.
Signed-off-by: Zepp-Hanzj <Zepp-Hanzj@users.noreply.github.com>
Signed-off-by: hanzj <hanzjian@zepp.com>
Use configured key, value, and filename limits while loading and saving settings storage data.
The text backend now builds backup filenames with a sized buffer and bounded formatting, and both text and binary loading reject keys or string values that are not terminated within their configured field sizes. This completes #3109 without changing the storage formats.
Signed-off-by: Nightt <87569709+nightt5879@users.noreply.github.com>
Use strnlen() for public key, value, and storage path length checks so user-provided settings strings are validated against the configured maximum sizes before they are scanned.
Use bounded key comparisons and strlcpy() for fixed-size settings fields. This addresses part of #3109 without changing the settings API or storage formats.
Signed-off-by: Nightt <87569709+nightt5879@users.noreply.github.com>
Without debug enabled, the code would not compile due to checking
`WIFEXITED(wstatus)` when `wstatus` was uninitialized.
Signed-off-by: Matteo Golin <matteo.golin@gmail.com>
The init.rc file path is now configurable to allow users to choose where
to put the startup script. This is useful for devices that mount
external media to a special directory like `/sd`.
Signed-off-by: Matteo Golin <matteo.golin@gmail.com>
Would not compile due to typo in macro describing service name length.
Change ensures that:
* We do not malloc an extra `len` bytes since this is already allocated
as part of the struct
* The name string always has a null terminating byte
Signed-off-by: Matteo Golin <matteo.golin@gmail.com>
FIx
