| Age | Commit message (Collapse) | Author |
|
根据这个issues #6857 在idle.c文件中找到一个函数缺失参数
|
|
|
|
|
|
|
|
|
|
* [utest/mm] add testcase for create/init
format codes of create/init in components/mm
* [libcpu/aarch64] fix user stack check routine
* [kservice] export API for utest
* [utest/mm] testcase for aspace_map
format & modify the files under components/mm related with aspace_map
* [lwp/user_mm] add user_map_varea for mmap feature
* [mm] rename rt_mm_fault_try_fix to rt_aspace_fault_try_fix
* [utest/mm] testcase for synchronization
* [mm] modify unmap api to improve throughput
* [utest/mm] testcases for cache and varea map
* [format] remove extra space
* [utest/mm] fix testcase problem in header
* [lwp] extend map_user_varea with a flag
* [utest/mm] testcase for lwp_map_user_varea
* [libcpu/arm/cortex-a] fix kernel space layout
* [utest/mm] adjust for armv7 arch
|
|
|
|
* Update Kconfig
* Update trap_gcc.S
* Update bsp/hifive1/drivers/SConscript
Co-authored-by: Man, Jianting (Meco) <[email protected]>
* Update SConscript
* [atomic]提交一份arm与risc-v架构下的常用原子操作函数
* 修改变量类型
* 更新rtatomic.h与atomic_port.c
* 更新rt-thread\libcpu\arm\common\atomic_port.c
* 更新include/rtatomic.h与libcpu/arm/common/SConscript
* 更新include/rtatomic.h
* 修正格式与Kconfig
* 修正格式与文件结构
* 规范文件格式与文件重命名
* 添加测试用例与CI
* 添加函数声明
* 修改virt64/SConscript 添加atomic_riscv.c
* 1.规范代码风格
* 2.添加RISC-V64原子指令支持 解决在RV64下编译器将32-bit运算结果扩展为64-bit 导致判断错误
* 添加C11标准库原子操作测试
---------
Co-authored-by: Man, Jianting (Meco) <[email protected]>
|
|
|
|
* [fix] reinsert should not suspend twice
* [mutex] fix prioceiling
* [mutex] fix prioceiling
|
|
This reverts commit 93a7cff44ccc2921f57dd05096268064bcaadd06.
|
|
|
|
|
|
|
|
|
|
|
|
rt_ssize_t can give negative error code, which follows the unix style correctly
|
|
|
|
|
|
Signed-off-by: supperthomas <[email protected]>
|
|
|
|
synchronize virtual memory system works.
adding kernel virtual memory management layer for page-based MMU enabled architecture
porting libcpu MMU codes
porting lwp memory related codes
|
|
|
|
RT_NAME_MAX为0是nano中使用,完整版中不支持const name
|
|
增加静态object 名字,用户可以根据内存实际使用情况决定使用动态还是静态。适用于资源极度受限的情况下使用。该功能在RT-Thread完整版本中不常用,主要用于RT-Thread Nano,以缩减对内存的占用。由于没有动态拼接支持,在静态名字下,空闲线程的名字在多核时将命名成相同的名字。
|
|
* [smart/aarch64] sync aarch64
|
|
* [dfs] sync cromfs
* [rt-smart]Weaken RT_USING_LWP, use RT_USING_SMART as macro configuration
* [format] fix some format issue.
|
|
* [rtdef] rename RT_WEAK attribute as rt_weak
* [rtdef] rename RT_USED attribute as rt_used
* [rtdef] rename RT_SECTION attribute as rt_section
* [rtdef] rename ALIGN attribute as rt_align
* [legacy] add RT_USED ALIGN RT_SECTION RT_WEAK as legacy support
|
|
* Synchronize the code of the rt mart branch to the master branch.
* TTY device
* Add lwP code from rt-smart
* Add vnode in DFS, but DFS will be re-write for rt-smart
* There are three libcpu for rt-smart:
* arm/cortex-a, arm/aarch64
* riscv64
Co-authored-by: Rbb666 <[email protected]>
Co-authored-by: zhkag <[email protected]>
|
|
This reverts commit 2d17fe736f79860237d8a05c57f9aabcfdab4c87.
|
|
|
|
|
|
[Kernel] Add nested mutex feature
|
|
|
|
|
|
enable zero-latency isr for critical interrupts.
|
|
|
|
|
|
|
|
|
|
The RT_ASSERT(obj != object) line is only compiled for debug mode.
But the rt_enter/exit_critical causes the compiler not to optimize
at least these 2 calls, even if it could optimize out the whole loop,
because the rt_list_entry function has no side effect, and RT_ASSET
is a no-operation in release mode. So this patch fixes this:
- no warnings anymore
- better speed in release mode
|
|
* [ulog] Fix the problem that the terminal cannot input
* [mem] Fix NULL pointer access
|
|
* [fix] Fix simulator compilation warnings
* Update bsp/simulator/rtconfig_project.h
Co-authored-by: Man, Jianting (Meco) <[email protected]>
|
|
|
|
版本控制规范:https://semver.org/
目前的版本子号不是很规范,其中RT_VERSION甚至和VS那边造成了冲突
```c
/* RT-Thread version information */
#define RT_VERSION 4L /**< major version number */
#define RT_SUBVERSION 1L /**< minor version number */
#define RT_REVISION 1L /**< revise version number */
```
参考ESP32:
```c
/** Major version number (X.x.x) */
#define ESP_ARDUINO_VERSION_MAJOR 2
/** Minor version number (x.X.x) */
#define ESP_ARDUINO_VERSION_MINOR 0
/** Patch version number (x.x.X) */
#define ESP_ARDUINO_VERSION_PATCH 3
```
建议改成:
```c
/* Major version number (X.x.x) */
#define RT_VERSION_MAJOR 4
/* Minor version number (x.X.x) */
#define RT_VERSION_MINOR 2
/* Patch version number (x.x.X) */
#define RT_VERSION_PATCH 0
```
|
|
和 RT_PRINTF_SPECIAL 做了相应处理 (#6361)
* 增加printf格式2进制和8进制,RT_KPRINTF_USING_LONGLONG 和 RT_PRINTF_SPECIAL 做了相应处理
|
|
|
|
|
|
|
|
解决 tms320f28379d bsp 启动的问题。主要改动如下。
修正context.s中汇编代码错误。在旧版的代码中,操作数为32位而汇编命令却使用了针对16位数据的命令MOV,导致程序在某些情况无法正常启动线程。
由于C28x的平台下,SP只支持16bit寻址,所以用于线程空间存放的heap以及ebss段都需要放在低16位的空间,针对这个问题修改了CMD文件。此外还增加基于CMD文件基于RAM的支持,方便调试。
新增rtdef.h中RT_SECTION,RT_USED,ALIGN和RT_WEAK的定义。旧版bsp中这些定义为空,导致INIT_EXPORT注册的函数失效。
修改程序启动代码,在程序入口直接调用原生启动代码
以上改动在LAUNCHXL-F28379D 通过了测试。
|