The BLE log compression feature (CONFIG_BT_LOG_CRITICAL_ONLY ->
BLE_COMPRESSED_LOG_ENABLE) failed to build on Windows while working
correctly on Linux, due to two shell/platform-specific issues in the
compression script.
1. Module/source argument quoting. CMakeLists.txt passes the
semicolon-separated module and source lists wrapped in single quotes
("'${MODULES}'") to protect ';' from POSIX shells, which strip them.
cmd.exe does not treat single quotes as quoting characters, so on
Windows the quotes reached the script literally and
args.module.split(';') produced "'BLE_MESH" / "BLE_HOST'" instead of
the clean names. These never matched the YAML module keys, every
module was skipped ("Skipping module ... - config not found"), the
compressed sources were never generated, and the build failed. Strip
surrounding quote characters before splitting; this is a no-op on
Linux/macOS where the shell already removed them.
2. CRLF line endings. With core.autocrlf=true the IDF sources are
checked out as CRLF on Windows. The generated *_log_index.h macros
use backslash-newline line-continuation; a backslash followed by
'\r\n' is not a valid continuation in C, producing floods of syntax
errors when the header is compiled. Write generated headers with
newline='' to force LF, and normalize source content to LF right
after reading so '\r' embedded inside multi-line argument expressions
is also handled. Byte offsets stay consistent because both tree-sitter
parsing and tag replacement operate on the normalized content.
Verified by full clean builds of examples/bluetooth/esp_ble_mesh/
vendor_models/vendor_client (esp32c6, bluedroid + mesh) from both
cmd.exe and PowerShell; both produce an identical vendor_client.bin.
(cherry picked from commit aa9b565a6d)
Co-authored-by: luoxu <luoxu@espressif.com>
Move sub-Kconfig files into categorized directories (profile/mesh/,
host/, iso/audio/) and create the missing ISO/Audio compression
Kconfig with COMPRESSION + PRESERVE pairs for all 4 log levels,
matching the existing cmake tag definitions.
(cherry picked from commit c0286b25fa)
Co-authored-by: luoxu <luoxu@espressif.com>
On ESP32C2 with BT_CTRL_RUN_IN_FLASH_ONLY, the bt_default linker scheme
routes .iram1 sections to flash_text, causing BLE_LOG_IRAM_ATTR functions
to end up in flash instead of IRAM. Define a dedicated .ble_log_iram1
section for ESP32C2 and route it to iram0_bt_text unconditionally.
- Use __atomic_fetch_add for stat_mgr counters to prevent lost updates
under concurrent ISR/task access (H1)
- Use __atomic_load_n with ACQUIRE ordering for ref_count spin-loops (L1)
- Remove unnecessary BLE_LOG_IRAM_ATTR from ble_log_rt_task since it
calls flash-resident functions and cannot run during flash ops (L3)
- Add parentheses to BLE_LOG_TRANS_FREE_SPACE and BLE_LOG_MAKE_FRAME_META
macro parameters to prevent operator precedence bugs (M6)
- Reorder deinit: stop RT task before destroying peripheral driver to
prevent sending transports to a NULL dev_handle on dual-core (H2)
- Drain remaining queue items in rt_deinit to clean up transport state
- Add atomic flush_in_progress guard to prevent two concurrent
ble_log_flush() callers from deadlocking on ref_count spin-wait (H5)
- Handle scheduler-suspended and ISR context in UART redirect path
to prevent xSemaphoreTake crash during light sleep (C1)
- Check uart_driver_install return value before setting inited flag (M1)
- Always clean up SPI device handle in deinit even if acquire_bus fails (M4)
During light sleep transitions, the FreeRTOS scheduler is suspended but
xPortInIsrContext() returns false, causing xQueueSend with portMAX_DELAY
to hit a configASSERT. Add a check for taskSCHEDULER_SUSPENDED and use
a non-blocking send with rollback on queue-full to avoid resource leaks.
idf-ci discovers pyproject.toml pytest config and runs a separate
collection pass with --target all, which finds 0 target tests and
raises RuntimeError. Add to .idf_ci.toml exclude_dirs instead of
deleting tests. Restore all 227 unit tests with lint/mypy fixes.
Reflect changes from dev/ble-log-202603: ping-pong to multi-buffer
transport (4 buffers/LBM), renamed Kconfig options (LBM_TRANS_SIZE →
LBM_TRANS_BUF_SIZE with new defaults), always-enabled checksum and
enhanced statistics, UART redirection support, UHCI Out removal,
SPI Out deprecation, and updated memory estimation.
Replace the two-step acquire() + get_trans() flow with a unified
acquire_trans() that iterates all candidate pools. When a pool's
buffers are exhausted, the lock is released and the next pool is
tried, enabling load balancing across pools instead of only
resolving lock contention.
Replace __atomic_exchange_n with __atomic_load_n in
ble_log_emit_buf_util() so inflight_peak reports the all-time peak
since init rather than resetting to zero after each report.
Migrate all LBM, RT, and peripheral backend code from 2-buffer
ping-pong to 4-buffer transport with bitmask index arithmetic.
LBM changes:
- Init sets owner on each transport buffer
- get_trans/flush/dump use (idx+1) & (cnt-1) instead of !idx
- get_trans/flush use __atomic_load_n(ACQUIRE) for prph_owned
- New ble_log_write_buf_util() emits BUF_UTIL telemetry frames,
exchange peak with 0 to avoid stale baseline after bursts
RT changes:
- Queue depth uses unified BLE_LOG_TRANS_TOTAL_CNT
- rt_queue_trans uses __atomic_store_n for prph_owned, tracks
per-LBM inflight count and peak via lock-free CAS
- ISR path captures pxHigherPriorityTaskWoken and yields
Peripheral changes:
- All tx_done callbacks: decrement inflight before releasing
prph_owned with __atomic_store_n(RELEASE)
- SPI/UART send_trans error paths: decrement inflight and
release prph_owned to allow retry on next get_trans pass
- Dummy send_trans: recycle buffer immediately (pos=0, decrement
inflight, release prph_owned) since no DMA hardware to wait for
- Redir LBM upgraded to 4-buffer with owner and reset support
- All peripherals implement ble_log_prph_reset_util_counters()
Add BLE_LOG_GET_FRAME_SN() on the buffer-full loss path so the SN
counter advances even when frames are dropped. This allows the console
to detect firmware-side buffer loss via SN gaps. Bump BLE_LOG_VERSION
to reflect the semantic change.
Introduce stream write API with deferred frame encapsulation for UART
redirection on port 0. Redirected data is now properly framed (header +
payload + checksum) instead of being sent as raw ASCII, preventing frame
parser sync oscillation on the receiver.
- Add stream_seal/stream_write/stream_flush in ble_log_lbm.c
- Add BLE_LOG_SRC_REDIR source and BLE_LOG_UART_REDIR_ENABLED gate
- Simplify redir_uart_tx_chars and timer callback to use stream API
- Flush pending stream data in ble_log_prph_deinit
- Make get_trans static (no external callers after refactor)
- Move UART wrap linker flags outside CONFIG_BT_ENABLED guard
- Default UART DMA peripheral when SOC_UHCI_SUPPORTED
- Default baud rate 921600 -> 3000000
Replace embedded ble_log_enh_stat_t (packed wire struct) inside
ble_log_stat_mgr_t with flat uint32_t counters. Natural 4-byte
alignment ensures each load/store compiles to a single l32i/s32i
on Xtensa/RISC-V, making individual field access atomic without
locks.
Build the packed wire format on the stack inside a critical section
in ble_log_write_enh_stat() so the full snapshot is consistent.
Remove CONFIG_BLE_LOG_PAYLOAD_CHECKSUM_ENABLED conditional compilation
-- payload checksum is always computed over the full frame.
Remove CONFIG_BLE_LOG_XOR_CHECKSUM_ENABLED conditional -- XOR checksum
is always used; delete the sum checksum dead code path.
Remove CONFIG_BLE_LOG_ENH_STAT_ENABLED conditional -- enhanced
statistics (frame/byte counters per source) are always active.
Remove incorrect select on choice symbol
BT_LE_CONTROLLER_LOG_MODE_BLE_LOG_V2 from BLE_LOG_IS_ESP_CONTROLLER.
Move ble_log_spi_out.c and ble_log_spi_out.h into a deprecated/
subdirectory under components/bt/common/ble_log/. Extract all
BT_BLE_LOG_SPI_OUT_* Kconfig options into deprecated/Kconfig.in
with an if-block to reduce depends-on repetition, sourced from
ble_log/Kconfig.in inside a "Legacy SPI Log Output (Deprecated)"
menu. Add mutual exclusion with BLE_LOG_ENABLED so that the legacy
SPI Out and the new BLE Log Module cannot be enabled simultaneously.
Update CMakeLists.txt source path and add deprecated/include to
include dirs so existing callers are unaffected. The BLE Log
Module's SPI Master DMA peripheral transport is the replacement.
Delete ble_log_uhci_out.c and its header. Remove the UHCI Out source
from CMakeLists.txt and all BT_BLE_LOG_UHCI_OUT_* Kconfig options.
Remove per-chip BT_LE_CONTROLLER_LOG_UHCI_OUT_ENABLED Kconfig and
all UHCI Out caller references in esp32c5/c6/h2 controller bt.c.
Update ble_log_console README to reference the new BLE Log Module
UART DMA config. The functionality is superseded by the BLE Log
Module's UART DMA peripheral transport.
Add ble_log_write_hci macro that encodes upstream/downstream direction
in MSB of HCI type byte. Update all 4 callsites (NimBLE + Bluedroid)
to use the new macro. Parser reads MSB to determine direction;
backward compatible (old firmware MSB=0 defaults to "sent").
Shen Weilong
Jiang Jiang Jian
Luo Xu
Zhou Xiao
Guo Zifan
Island
Alexey Gerenkov
zhanghaipeng
Liu Linyan
Rahul Tank
Wang Mengyang
cjin
Sahil Yadav