| 文件 | 最后提交记录 | 最后更新时间 |
|---|---|---|
Merge branch 'master' of git@gitcode.com:yijobs/ability_ability_runtime_1300.git into 'master' # Conflicts: # conflict services/abilitymgr/src/interceptor/disposed_rule_interceptor.cpp | 1 个月前 | |
perf(build): enable size_optimize_config on 130 SO targets Add a new services/common:size_optimize_config (same flags as optimize_config minus the explicit -flto=full, so the toolchain's global ThinLTO mode is preserved) and apply it to the 130 ohos_shared_library targets that previously shipped without any optimization config, so unreferenced functions/data are dropped at link time and ROM shrinks without overriding the global LTO policy. -Wl,--gc-sections is an ELF-only linker flag: Apple ld rejects it, so both optimize_config and size_optimize_config now add it only when current_os != "mac", keeping the darwin SDK host build (which links dylibs with -dead_strip) working. Co-Authored-By: Agent Signed-off-by: RuiChen_01 <chenrui193@huawei.com> AI[100%] Human Fixed[0%] Human[0%] AI Adopted[100%] Co-authored-by: claude (glm-5.3) <ai@local> | 1 个月前 | |
Merge branch 'master' of git@gitcode.com:yijobs/ability_ability_runtime_1300.git into 'master' # Conflicts: # conflict services/abilitymgr/src/interceptor/disposed_rule_interceptor.cpp | 1 个月前 | |
bti Signed-off-by: Luobniz21 <luoyicong@h-partners.com> 🤖 AI[100%] 🔧 Human Fixed[0%] 🧑 Human[0%] 👌 AI Adopted[100%] Co-authored-by: opencode (glm-5.2) <ai@local> | 2 个月前 | |
补充查询子进程的接口 Signed-off-by: SKY2001 <songtianyu10@huawei.com> AI[47%] Human Fixed[0%] Human[53%] AI Adopted[100%] | 1 个月前 | |
fix_bug/add return bool in function LoadConnectServerDebuggerSo Signed-off-by: Alang <gaoyulang@huawei.com> AI[100%] Human Fixed[0%] Human[0%] AI Adopted[100%] Co-authored-by: opencode (glm-5.2) <ai@local> Change-Id: If80dd41af626c4e0311cc8acc87da3fee9d682f8 | 1 个月前 | |
bti Signed-off-by: Luobniz21 <luoyicong@h-partners.com> 🤖 AI[100%] 🔧 Human Fixed[0%] 🧑 Human[0%] 👌 AI Adopted[100%] Co-authored-by: opencode (glm-5.2) <ai@local> | 2 个月前 | |
utils_0820 Signed-off-by: Luobniz21 <luoyicong@h-partners.com> 🤖 AI[100%] 🔧 Human Fixed[0%] 🧑 Human[0%] 👌 AI Adopted[100%] Co-authored-by: opencode (glm-5.2) <ai@local> | 1 个月前 | |
!20277 merge bugfix/agent-card-multiuser into master bugfix: OTA missing 101's AgentCards Created-by: yangxuguang-huawei Commit-by: yangxuguang-huawei Merged-by: openharmony_ci Description: **IssueNo**: https://gitcode.com/openharmony/ability_ability_runtime/issues/16036 **Description**: **稳定性自检：** | 自检项 | 自检结果 | | ------------------------------------------------------------ | -------- | | 涉及跨进程调用的相关操作需要抛至主线程或加锁防止并发 | ok | | 成员变量进行赋值或创建需要排查并发 | ok| | 谨慎在lambda表达式中使用引用捕获 | ok| | 谨慎在未经拷贝的情况下使用外部传入的string、C字符串 | ok| | map\vector\list\set等stl模板类使用时需要排查并发 | ok| | 谨慎考虑加锁范围 | ok| | 在IPC通信中谨慎使用同步通信方式 | ok| | 禁止传递this指针至其他模块或线程（特别是eventhandler任务） | ok| | 禁止将外部传入的裸指针在内部直接构造智能指针 | ok| | 禁止多个独立创建的智能指针管理同一地址 | ok| | 禁止在析构函数中抛异步任务 | ok| | 禁止js对象在非js线程（例如在IPC线程）创建、使用或销毁 | ok| | 禁止在对外接口中未经判空直接使用外部传入的指针 | ok| | 禁止接口返回局部变量引用 | ok| | 禁止在信号函数中加锁 | ok| | 禁止在关键流程（SA启动、应用启动等主流程）执行耗时的操作 | ok| | 禁止将同一个cpp编译在不同的so中 | ok| **安全编码自检：** | 自检项 | 自检结果 | | -------------------------------------------------------------- | -------- | | 裸指针避免通过隐式转换构造为sptr | ok| | json对象在取值之前必须先判断类型，避免类型不匹配 | ok| | 序列化时必须对传入的数组大小进行校验，避免出现超大数组 | ok | | 避免使用未明确位宽的整型，选择使用int8_t、uint8_t等类型 | ok| | 外部传入的路径要做规范化校验，对路径中的.、..、../等特殊字符严格校验 | ok | | 指针变量、表示资源描述符的变量、bool变量必须赋初值 | ok| | readParcelable获取的对象使用前需要判空 | ok| | 分配和释放内存的函数需要成对出现 | ok| | 申请内存后异常退出前需要及时进行内存释放 | ok | | 内存申请前必须对内存大小进行合法性校验 | ok| | 内存分配后必须判断是否成功 | ok| | 禁止使用realloc、alloca函数 | ok| | 禁止打印文件路径、口令等敏感信息，如有需要，使用private修饰 | ok| | 禁止打印内存地址 | ok| | 整数之间运算时必须严格检查，确保不会出现溢出、反转、除0 | ok| | 禁止对有符号整数进行位操作符运算 | ok| | 禁止对指针进行逻辑或位运算 | ok| | 循环次数如果收外部数据控制，需要检验其合法性 | ok| | 禁止使用内存操作类危险函数，需要使用安全函数 | ok| | 谨慎使用不可重入函数 | ok| | 必须检查安全函数的返回值，并进行正确处理 | ok| | 禁止仅通过TokenType类型判断绕过权限校验 | ok| **TDD Result**: ok **XTS Result**: ok ### 是否已执行L0用例 - [x 已验证 - [ ] 不涉及。如不涉及，请写明理由 ### AI检视评分（使用本地代码检视skills扫描）: See merge request: openharmony/ability_ability_runtime!20277 | 1 个月前 | |
perf(build): enable size_optimize_config on 130 SO targets Add a new services/common:size_optimize_config (same flags as optimize_config minus the explicit -flto=full, so the toolchain's global ThinLTO mode is preserved) and apply it to the 130 ohos_shared_library targets that previously shipped without any optimization config, so unreferenced functions/data are dropped at link time and ROM shrinks without overriding the global LTO policy. -Wl,--gc-sections is an ELF-only linker flag: Apple ld rejects it, so both optimize_config and size_optimize_config now add it only when current_os != "mac", keeping the darwin SDK host build (which links dylibs with -dead_strip) working. Co-Authored-By: Agent Signed-off-by: RuiChen_01 <chenrui193@huawei.com> AI[100%] Human Fixed[0%] Human[0%] AI Adopted[100%] Co-authored-by: claude (glm-5.3) <ai@local> | 1 个月前 | |
utils_0820 Signed-off-by: Luobniz21 <luoyicong@h-partners.com> 🤖 AI[100%] 🔧 Human Fixed[0%] 🧑 Human[0%] 👌 AI Adopted[100%] Co-authored-by: opencode (glm-5.2) <ai@local> | 1 个月前 | |
bti Signed-off-by: Luobniz21 <luoyicong@h-partners.com> 🤖 AI[100%] 🔧 Human Fixed[0%] 🧑 Human[0%] 👌 AI Adopted[100%] Co-authored-by: opencode (glm-5.2) <ai@local> | 2 个月前 | |
fix(insight_intent): guard json::dump against deep-nesting stack overflow nlohmann::json::dump() recurses on nested values; super-deep input causes SIGSEGV stack overflow that try/catch cannot intercept. - Add iterative IsJsonDepthOk walker with JSON_DUMP_MAX_DEPTH (100) bound; SafeDump/SafeDumpTo reject over-depth input before dump(). Definitions live in intent_json_safe_get.cpp, compiled only into the ability_manager innerkit so symbols resolve for libabilityms.so and direct-compiling test/fuzz targets (libabilityms.map hides non-whitelisted symbols). - Switch raw .dump() calls in function_call_convert, execute_result and info_for_query to the safe wrappers; validate inputSchema depth before merge. - Add IsJsonDepthOk boundary tests (under/at/over limit). Co-Authored-By: Agent Signed-off-by: RuiChen_01 <chenrui193@huawei.com> AI[100%] Human Fixed[0%] Human[0%] AI Adopted[100%] Co-authored-by: claude (unknown) <ai@local> Co-authored-by: claude (glm-5.3) <ai@local> Co-authored-by: claude (unknown) <ai@local> Co-authored-by: claude (glm-5.3) <ai@local> Co-authored-by: claude (unknown) <ai@local> Co-authored-by: claude (glm-5.3) <ai@local> Co-authored-by: claude (unknown) <ai@local> Co-authored-by: claude (glm-5.3) <ai@local> Co-authored-by: claude (unknown) <ai@local> | 1 个月前 | |
static_0707 Signed-off-by: Luobniz21 <luoyicong@h-partners.com> | 3 个月前 | |
perf(build): enable size_optimize_config on 130 SO targets Add a new services/common:size_optimize_config (same flags as optimize_config minus the explicit -flto=full, so the toolchain's global ThinLTO mode is preserved) and apply it to the 130 ohos_shared_library targets that previously shipped without any optimization config, so unreferenced functions/data are dropped at link time and ROM shrinks without overriding the global LTO policy. -Wl,--gc-sections is an ELF-only linker flag: Apple ld rejects it, so both optimize_config and size_optimize_config now add it only when current_os != "mac", keeping the darwin SDK host build (which links dylibs with -dead_strip) working. Co-Authored-By: Agent Signed-off-by: RuiChen_01 <chenrui193@huawei.com> AI[100%] Human Fixed[0%] Human[0%] AI Adopted[100%] Co-authored-by: claude (glm-5.3) <ai@local> | 1 个月前 | |
update Signed-off-by: majian <majian59@h-partners.com> | 9 个月前 | |
perf(build): enable size_optimize_config on 130 SO targets Add a new services/common:size_optimize_config (same flags as optimize_config minus the explicit -flto=full, so the toolchain's global ThinLTO mode is preserved) and apply it to the 130 ohos_shared_library targets that previously shipped without any optimization config, so unreferenced functions/data are dropped at link time and ROM shrinks without overriding the global LTO policy. -Wl,--gc-sections is an ELF-only linker flag: Apple ld rejects it, so both optimize_config and size_optimize_config now add it only when current_os != "mac", keeping the darwin SDK host build (which links dylibs with -dead_strip) working. Co-Authored-By: Agent Signed-off-by: RuiChen_01 <chenrui193@huawei.com> AI[100%] Human Fixed[0%] Human[0%] AI Adopted[100%] Co-authored-by: claude (glm-5.3) <ai@local> | 1 个月前 | |
Add cfi/pac config Co-Authored-By: Agent Signed-off-by: wangzhen <wangzhen416@huawei.com> 🤖 AI[0%] 👌 AI Adopted[0%] 🧑 Human[100%] | 2 个月前 | |
[MM] Implement background/foreground application state notifications for ETSRuntime Issue: https://gitcode.com/openharmony/arkcompiler_runtime_core/issues/12394 Testing: static applications Co-Authored-By: Codex GPT-5.6 Sol Signed-off-by: Petrov Igor <petrov.826961@huawei.com> | 1 个月前 | |
bti Signed-off-by: Luobniz21 <luoyicong@h-partners.com> 🤖 AI[100%] 🔧 Human Fixed[0%] 🧑 Human[0%] 👌 AI Adopted[100%] Co-authored-by: opencode (glm-5.2) <ai@local> | 2 个月前 | |
appindex适配 Signed-off-by: yijobs <yisilong@h-partners.com> | 1 个月前 | |
refactor(want_agent): remove empty WantSenderStub shell PendingWantRecord now inherits IRemoteStub<IWantSender> directly. WantSenderStub exported no symbols of its own and carried no logic, so drop it along with its dedicated unittest, fuzzer and map entry. Co-Authored-By: Agent AI[100%] Human Fixed[0%] Human[0%] AI Adopted[100%] Co-authored-by: claude (glm-5.3) <ai@local> Change-Id: Ie3012a58d4bf51a52a03858974feefd0f4792e1e Signed-off-by: RuiChen_01 <chenrui193@huawei.com> | 1 个月前 | |
cameraPicker调用链上报 Signed-off-by: 朱菲墨 <zhufeimo1@huawei.com> | 11 个月前 |
| 文件 | 最后提交记录 | 最后更新时间 |
|---|---|---|
| 1 个月前 | ||
| 1 个月前 | ||
| 1 个月前 | ||
| 2 个月前 | ||
| 1 个月前 | ||
| 1 个月前 | ||
| 2 个月前 | ||
| 1 个月前 | ||
| 1 个月前 | ||
| 1 个月前 | ||
| 1 个月前 | ||
| 2 个月前 | ||
| 1 个月前 | ||
| 3 个月前 | ||
| 1 个月前 | ||
| 9 个月前 | ||
| 1 个月前 | ||
| 2 个月前 | ||
| 1 个月前 | ||
| 2 个月前 | ||
| 1 个月前 | ||
| 1 个月前 | ||
| 11 个月前 |