[Fix] <components>:finsh/shell.c
When using char rt_hw_console_getchar(void),
some compilers may default to returning 255 instead of the expected -1,
causing the condition if(ch < 0) at F: shell.c L:519 to fail and enter an erroneous loop.
Solution: Use the signed char type return value as rt_hw_console_getchar
Signed-off-by: Yucai Liu <1486344514@qq.com>
[Descriptions]:
1. Make these functions public to avoid code duplication and allow direct usage by other components.
2. Fix ifdef's error note.
[Root Cause]:NA
[Side Effects]:NA
This patch only updated the groups defined in documentation/0.doxygen.
Other groups will be moved into documentation/0.doxygen and updated at
that time.
Signed-off-by: Chen Wang <unicorn_wang@outlook.com>
There is a problem with the matching of case names
in the original code. Due to original code use memcmp
with len, if the input case name and the existing
case name have an inclusion relationship, for example,
if the actual case name is "gpip_irq", and run
`utest_run gpio` will also match successfully, but it's
not expected.
Modify the logic of exact matching and use strcmp instead.
Keep the original wildcard logic, that is,
`utest_run gpio*` can match both "gpio_irq" and "gpio".
Signed-off-by: Chen Wang <unicorn_wang@outlook.com>
gcc build warnings:
cast from pointer to integer of different size [-Wpointer-to-int-cast]
For rt_adc_ops.enabled, the type of second param is rt_int8_t,
original _adc_control also called with wrong type casting.
Convert first to rt_base_t, then cast to rt_int8_t to remove the
build warnings.
Signed-off-by: Chen Wang <unicorn_wang@outlook.com>
The originally defined "group_Drivers" should actually
be part of the device-driver component, so
"group_Drivers" is renamed to "group_device_driver".
In this way, the original "group_Device" and
"group_Drivers" are merged into "group_device_driver".
Adjusted the writing method of doxygen comments in the
driver code to be unified as follows:
```c
* @defgroup group_XXX XXX
* @brief xxxxxx
* @ingroup group_device_driver
```
Signed-off-by: Chen Wang <unicorn_wang@outlook.com>
To enhance the robustness of key recognition and solve the problem of errors and omissions caused by the lag of state assignment statement
增强方向键识别的鲁棒性，解决状态赋值语句滞后带来的错漏问题
- Implemented vDSO functionality for the RISC-V architecture, including the necessary source files and linker scripts.
- Introduced a new `vdso_sys.c` file for RISC-V, containing functions to handle time retrieval using the vDSO mechanism.
- Created architecture-specific linker scripts (`vdso.lds.S`) for both AArch64 and RISC-V.
- Updated the build system to support vDSO compilation for RISC-V, including necessary adjustments in `SConstruct` files.
- Refactored existing vDSO code to improve compatibility and maintainability across architectures.
- Adjusted the maximum number of PTY devices in the terminal configuration from 64 to 32 for better resource management.
- Fixed minor issues in existing code, including correcting the path for the vDSO shared library and ensuring proper function definitions.
Achieve driver framework by operating memory to simulate audio
peripheral drivers. And it could be as a draft standrad for other
drivers test framework.
Signed-off-by: 1078249029 <1078249029@qq.com>
Fix garble caused by empty implementation of earlycon series
function in driver.If driver doesn't offer earlycon_id or rt_f
dt_earlycon_id.setup function but open option, the checking
of best_earlycon_id will failed.
Signed-off-by: 1078249029 <1078249029@qq.com>
If a node request a new node or it's parent node. the request's node will probe a double time.
So we check the device object of node in ofw probe entry and ofw probe child exit.
ricky