Replace the oneshot latency assumption with a 1ms periodic timer that
runs only across a forced dirty-partition erase, and fail if irq_count
falls below 80% of the expected count from erase duration.
Move 7 trace-related examples (app_trace_basic, app_trace_to_plot,
esp_trace_custom_library, function_tracing, gcov, sysview_tracing,
sysview_tracing_heap_log) from examples/system/ into a dedicated
examples/system/tracing/ subdirectory for better organization.
Update all path references across documentation (en + zh_CN),
build-test-rules, CI configs, Kconfig, and component READMEs.
Keep legacy ULP apps on the CMake v1 build path while buildv2 coverage
is limited to full_subproject ULP apps.
Also add the missing esp_driver_gpio dependency to the esp_pm test app
so buildv2 dependency checks see driver/rtc_io.h explicitly.
Add examples/system/ulp/lp_core/build_system_v2/ demonstrating the
cmakev2 ULP build with native IDF component support, Kconfig
integration, component manager usage, linker scripts (static and
ldgen-processed), and per-project menuconfig.
The example shows a custom ULP project (via ulp_add_project) structured
as a proper IDF-like project with a main component and
idf_build_executable(). The ULP sub-project runs its own kconfgen with
the parent's sdkconfig as SDKCONFIG_DEFAULTS, inheriting the parent's
IDF configuration while adding ULP-specific Kconfig options from its
own components.
Three components demonstrate different integration mechanisms:
- ulp_sum and ulp_math: pure cmakev2 components in the components/
subdirectory, using add_library(${COMPONENT_TARGET}) directly.
Each provides a Kconfig file with configurable options.
- ulp_clamp: an external component in the extra/ subdirectory,
declared as a local path dependency in ulp_sum/idf_component.yml.
The component manager resolves this dependency automatically.
ulp_clamp also provides a Kconfig file with a configurable
maximum clamp value (CONFIG_ULP_CLAMP_MAX_VALUE).
Linker script support is demonstrated in two ways:
- Static linker script: ulp_math registers ulp_math_sections.ld via
idf_component_set_property(LINKER_SCRIPTS), defining a custom
linker symbol (ulp_math_version).
- ldgen-processed template: ulp_math provides a linker fragment file
(ulp_math.lf) that describes a custom .ulp_math_fast section
placement. The main component registers ulp_sections.ld.in via
target_linker_script(PROCESS), which ldgen processes using the
fragment to generate the final linker script. The
ulp_math_multiply() function is placed in .ulp_math_fast using
__attribute__((section)). Both static and ldgen linker scripts
are handled by the standard idf_build_library() flow without any
ULP-specific linker script functions.
The ULP project has its own sdkconfig.defaults that overrides the
ulp_clamp default (100 instead of 255), demonstrating per-project
configuration layering. Users can append to SDKCONFIG_DEFAULTS
before include(idf.cmake) to add project-specific defaults.
A menuconfig-<app_name> target is registered in the parent build,
allowing ULP-specific configuration via:
idf.py --no-hints menuconfig-ulp_build_system_example
Signed-off-by: Frantisek Hrbata <frantisek.hrbata@espressif.com>
pytest_gcov hardcoded `__idf_<component>.dir/` as the per-component CMake
intermediate directory where each .gcda is expected to land. The directory
name is build-system specific (Build system v1 uses `__idf_<component>.dir`,
Build system v2 uses `_idf_<component>.dir`). Asserting on a single literal
made every test_gcov_uart run fail under Build system v2 even though the
chip was streaming valid coverage data and the host-side UartGcovCapture
was writing the files to the build's actual layout.
Probe `CMakeFiles/*_<component>.dir/` and prefer the candidate whose .gcno
files were emitted by the active build, so the assertion follows the build
output rather than dictating it.
ESP32-C2 and ESP32-C61 have no RSA based Secure Boot V2 support, so the
CI configs that pin the RSA signing scheme and key fail the app signing
scheme/key check on these targets.
- flash_enc_wifi_2.data_partition_verification: add C2/C61 sdkconfig
overlays that switch to the ECDSA P-256 signing scheme and key (no
force-enable needed as Secure Boot V2 itself is not enabled here).
- on_update_no_sb_rsa: disable the build on targets without
SOC_SECURE_BOOT_V2_RSA, mirroring simple_ota_example, since this
config specifically exercises the RSA scheme.
fix(mmap): fixed mmap read data wrong when flash being erased/written and cache not disabled
Closes IDFGH-14084
See merge request espressif/esp-idf!29804
Address these:
lock_example.c:162:60: error: cast to pointer from integer of different size
[-Werror=int-to-pointer-cast]
162 | xTaskCreatePinnedToCore(inc_num_mutex, NULL, 4096, (void *)thread_id, TASK_PRIO_3, NULL, tskNO_AFFINITY);
| ^
create_task_example.c:30:19: error: cast from pointer to integer of different size
[-Werror=pointer-to-int-cast]
30 | int task_id = (int)arg;
ESP_FAULT_ASSERT(C) was silently deleted by the optimizer when C is a cached
flag/status already proven by a preceding `if (!C) return/goto`: the compiler
folds C to a constant and drops all three checks, removing the fault-injection
protection with no warning.
Before:
The cache won't be disabled when XIP on psram. But during flash
erasing/programming, read data will be courrupt.
When XIP in psram is enabled, the image is not mapped to the cache so
usually there will be no flash access. The only way to read from flash
is via the driver or use mmap. The driver has protection during erasing,
while th mmap region not.
Now:
Mmap APIs provide a flag to make mmap->unmap region mutually exclusive
to flash erase/programming when XIP from psram. SPI Flash write APIs
will benefit from this. When the flag is used, no concurrent access to
mapped region will happen while writing; otherwise the cache will be
disable to avoid data corruption.
Most ESP-IDF APIs calls mmap with this flag. As for users calling
mmap-like APIs directly, they can choose whether to enable this by a
flag.
Closes https://github.com/espressif/esp-idf/issues/14897
