- Gate bta_sys_event() on both 'is_reg' and 'reg[id]' to prevent
stale event delivery.
- Defer bta_sys_deregister() to the end of profile disable handlers
to ensure pending DISABLE events are processed.
- Add disabling flag to HFP AG for tracking asynchronous teardown.
The Key Manager hardware peripheral in its current form needs further
design changes before it can be offered as a production feature.
Until a revised peripheral design is available, withdraw ESP-IDF
support for it on all Key Manager capable targets.
F_SETFL was replacing the whole flags word, so fcntl(fd, F_SETFL, O_APPEND)
made F_GETFL report O_RDONLY|O_APPEND. Keep O_ACCMODE and apply only POSIX
status flags.
Sort bond entries by bond_count after in-place updates to maintain correct
eviction order, and log IRK resolving-list failures instead of failing the bond.
AVDT:
- Roll back CCB allocation when cmd/rsp queue creation fails
- Free media packet on invalid handle in AVDT_WriteReqOpt
- Zero-init timeout failure message before GETCAP callback
- Initialize lcid_tbl to 0xFF to avoid mapping to tc_tbl[0]
BTA/AVRCP:
- Use size_t for AVRC message copy buffer allocation
- Allocate before register in BTA_AvEnable
- Guard BTA_AvRegister callback when enable never completed
- Remove invalid free of inline Cover Art image_descriptor
A2DP BTC/API:
- Default g_a2dp_on_deinit to true before profile init
- Reject source audio send when A2DP is deiniting
- Add shutdown state check in btc_a2dp_sink_shutdown
- Guard A2DP source timer against freed dynamic local param
The previous commit touches tools/ci/check_type_comments.py, which brings the
file into ruff's scope for the first time in a while and surfaces two
pre-existing findings:
A004 Import `exit` is shadowing a Python builtin
F401 `typing.List` imported but unused
Import sys and call sys.exit() rather than shadowing the builtin, which is
also how master writes this file. List is used, but only inside `# type:`
comments that ruff cannot see, so mark the import instead of dropping it:
check_type_comments.py is not on tools/ci/mypy_ignore_list.txt, so mypy checks
it under disallow_untyped_defs and the name has to resolve.
No behaviour change. Verified with the pinned ruff 0.9.7 (check and
format --check) and with mypy 1.19.1, and both exit paths of the script still
behave as before.
Signed-off-by: Frantisek Hrbata <frantisek.hrbata@espressif.com>
check_type_comments.py runs ignore-listed files through a relaxed mypy pass
with --python-version 3.8, which overrides the python_version = 3.9 that
.mypy.ini already sets. mypy dropped 3.8 as a modelling target in 1.17.0 and
now rejects the flag while parsing arguments, before it reads any source:
mypy: error: argument --python-version: Python 3.8 is not supported
(must be 3.9 or higher)
mypy check failed for:
tools/ldgen/ldgen.py
The hook declares mypy without a version, so which mypy is used is decided
when the ci/images pre-commit-idf-v5.5 image is built. That image is based on
Python 3.9, where Requires-Python caps mypy at 1.19.1, and the image
currently in use carries exactly that. Any MR touching a file listed in
tools/ci/mypy_ignore_list.txt therefore fails check_pre_commit
deterministically, whatever the file contains.
Pass 3.9 instead, matching .mypy.ini and the OLDEST_PYTHON_SUPPORTED = (3, 9)
that tools/python_version_checker.py enforces on this branch. This restores
the pairing master and release/v6.0 already have, where the flag mirrors
.mypy.ini at 3.10.
release/v5.4 and release/v5.3 need no equivalent change: their pre-commit
images are based on Python 3.8, where mypy caps at 1.14.1 and the flag is
still accepted.
Signed-off-by: Frantisek Hrbata <frantisek.hrbata@espressif.com>
release/v5.5's tools/ldgen/ldgen.py predates the ruff-format layout that
pre-commit enforces, so running the hook over this file reports it as
needing reformatting because of code that no change here touches.
Apply ruff-format. That also resolves the E501 on the long
"with open(output_path, ...)" line, since the formatter wraps the call.
No behaviour change.
Signed-off-by: Frantisek Hrbata <frantisek.hrbata@espressif.com>
On some Windows systems, antivirus, endpoint-security or DLP/encryption
software intercepts short-lived toolchain processes and strips their
stdout when the build captures it through a pipe, while the same command
prints its output normally when run by hand. The tool exits successfully
but returns nothing, and each build step that reads toolchain output then
failed with a different, cryptic error far from the real cause:
- CMake configuration aborted with "Unknown arguments specified" in
components/xtensa/project_include.cmake, or "check_expected_tool_version
invoked with incorrect arguments" in components/esp_common.
- ldgen turned the empty objdump output into an opaque pyparsing
"Expected 'In archive'" traceback.
- idf_tools.py silently reported the compiler/debugger version as
"unknown", sending users into a fruitless reinstall loop.
Detect the empty result at each consumer and fail (or warn) with an
actionable message that names the likely cause and the remedy:
- tools/cmake/compiler_query.cmake: new __compiler_query() helper runs a
compiler query and fails with a clear error on empty or failed output.
It is a standalone module included by the build system utilities, so
that it is available to the esp_common and xtensa project_include.cmake
files that call it. The xtensa if() arguments are now quoted so an
empty result no longer collapses into a parse error.
- tools/ldgen/ldgen.py: _run_objdump() rejects empty objdump output, and
non-empty-but-unparsable section info is caught and re-raised as a clear
LdGenFailure instead of a raw pyparsing traceback.
- tools/idf_tools.py: empty version output now warns with the cause and
returns UNKNOWN_VERSION instead of silently reporting "unknown".
Closes https://github.com/espressif/esp-idf/issues/18727
Signed-off-by: Frantisek Hrbata <frantisek.hrbata@espressif.com>