| 文件 | 最后提交记录 | 最后更新时间 |
|---|---|---|
[refactor] Update pod entity contract Co-authored-by: zhao_lulu<zhaojiamin13@huawei.com> # message auto-generated for no-merge-commit merge: !279 merge feat-pod-update into master [refactor] Update pod entity contract Created-by: zhao_lulu Commit-by: zhao_lulu Merged-by: cann-robot Description: ## 描述 同步 hcomm 数控分离 pod 契约 v2 的数据结构与回调约定: - 调整 HcommCcuInstanceEntity 字段顺序,将 ccuInsKey 移到 ascCustom 之前,保持 ABI 字段布局一致。 - Channel 实体查询回调统一命名为 HcommCcuChannelGetEntityCallBack,并同步 ccu_channel 与 UT stub 调用。 - 移除 HcommCcuInstSpaceFreeCallBack,控制槽仅保留 Channel 实体查询、指令空间申请和指令下发。 - 同步 HcommCcuControlSlot 枚举与测试 stub,保证 asc-comm 数据面按新的 hcomm 契约消费控制槽。 ## 变更类型 - [x] ♻️ refactor: 代码重构 ## 关联的Issue https://gitcode.com/cann/hcomm/issues/1027 ## 测试 已完成的测试用例和场景: 1. clang-format --dry-run --Werror 通过。 2. git diff --check 通过。 补充的UT用例: - 同步 ccu_channel_get_stub.cc 中的 Channel 实体查询回调签名与控制槽注入。 ## 文档更新 无 ## 合入检查 - [ ] 🧐 已经详细阅读了贡献指南(CONTRIBUTING.md),并遵守了其中的所有规定,包括但不限于commit message的格式、无效commit的合并等 - [ ] 🔍 邀请 committer评论/lgtm前的必要检查 - [x] 🏷️ 标题中使用了合适的类型标签(如:[refactor]) - [x] 📄 代码修改内容已简要描述,相关文档已更新 - [x] 📝 代码注释已更新,代码遵循项目整体代码风格 - [x] 🧪 代码UT测试已更新,覆盖率已达标 - [x] 🔬 验证方法已更新到"测试"部分 - [x] 🛠️ 代码已通过静态分析工具检查,无错误 - [ ] 👥 代码检视/code review/同行评议和必要的代码串讲,确保代码质量 - [ ] ✅ 代码检视意见已处理或答复,无未处理的检视意见 - [ ] 🚀 预约 前冒烟 用例前的必要检查 - [ ] ✔️ 代码已有committer的/lgtm 和 模块committer的/lgtm评论 - [ ] 🔧 代码已通过compile,编译无错误,无告警 - [ ] 🖥️ 代码已通过基本功能本地测试或者在线测试,确保基本功能正常 - [ ] 🎯 预约 approver评论/approve,正式合入前的必要检查 - [ ] 📊 前冒烟 用例已全量通过 - [ ] 📦 新增功能已同步补充基本功能测试用例到前冒烟里 See merge request: cann/asc-comm!279 | 11 天前 | |
fix(ccu): align hcomm contract header include paths Co-authored-by: Codex<huangduo4@hisilicon.com> # message auto-generated for no-merge-commit merge: !265 merge dev1 into master fix(ccu): align hcomm contract header include paths Created-by: DragonBornHD84 Commit-by: Codex Merged-by: cann-robot Description: <!-- 感谢您的合入申请! --> ### 当前PR是否有AI参与: [x] 否 [ ] 是 __1. AI Agent 平台: __2. AI 模型: __3. Prompt上下文 : ### PR功能描述 / 为什么需要这个合入**: <!-- 本 PR 做了什么,为什么需要 / What does this PR do and why --> ### 该PR关联的issue *(格式为fixes #<issue号>, 或者resolves #<issue号>)*: fixes # ### 希望检视人员了解: ## 改动类型 / Change Type - [ ] Bug 修复 / Bug Fix - [ ] 新功能 / New Feature - [ ] 性能优化 / Performance - [ ] 代码重构 / Refactoring - [ ] 文档更新 / Documentation - [ ] 测试相关 / Test - [ ] 其它 / Other ## 测试信息 / Testing <!-- 简要测试说明或关键结果 / Brief test description or key results --> - [ ] 单元测试通过 / UT passed - [ ] 集成测试通过 / ST passed - [ ] 人工验证通过 / Manual verified ## 检查清单 / Checklist - [ ] 代码符合规范 / Code follows style guide - [ ] 测试添加并通过 / Tests added and passed - [ ] 文档已更新 / Docs updated if needed - [ ] 无硬编码敏感信息 / No secrets hardcoded - [ ] 提交信息符合规范 / Commit message follows convention See merge request: cann/asc-comm!265 | 20 天前 | |
[Bugfix] Scope cached CCU kernel handles by instance and die Co-authored-by: KaranocaVe<karanocave@qq.com> # message auto-generated for no-merge-commit merge: !234 merge fix/2026-09-27-asc-comm-api-contract-bug-hunt-02 into master [Bugfix] Scope cached CCU kernel handles by instance and die Created-by: KaranocaVe Commit-by: KaranocaVe Merged-by: cann-robot Description: # [Bugfix] Scope cached CCU kernel handles by instance and die ## Problem The host launch cache is keyed only by binary_cache_tag. When two CCU instances use the same tag, the second launch reuses the kernel handle registered for the first instance. The same issue occurs across physical dies. ## Reproduction Launch the same kernel with the same binary_cache_tag first on one CCU instance and then on a different instance, or use a different physical die. The second launch reuses the first registration instead of registering the kernel for its own instance and die. ## Expected and actual behavior - Expected: registrations are reused only when the binary tag, CCU instance, and physical die are the same. - Actual: the binary tag alone selects the cached handle, so a handle from a different instance or die is reused. ## Impact Kernel launches can use a handle registered for the wrong CCU instance or physical die when identical binary cache tags are used. ## Fix Include the CCU instance identity and physical die mask in the cache key, and cover same-instance reuse and cross-instance/die isolation in the host unit test. ## Verification The regression fails on the original implementation and passes with the fix. The asccomm_ut_ccu_host_launch suite passes all 12 tests. 关联 Issue: https://gitcode.com/cann/asc-comm/issues/48 See merge request: cann/asc-comm!234 | 13 天前 | |
add hostlaunch Co-authored-by: bluesky901<qianxinhai@huawei.com> # message auto-generated for no-merge-commit merge: !60 merge add_host_launch_br into master add hostlaunch Created-by: bluesky901 Commit-by: bluesky901 Merged-by: cann-robot Description: ### PR功能描述 / 为什么需要这个合入**: 本次提交新增 CCU Host Launch 能力,提供 asccomm_ccu_host_kernel_launch 和 asccomm_ccu_get_launch_hash_tag 接口, 支持通过直调 <<<>>> 方式拉起 CCU kernel。同时新增 ccu_direct 下 AllGather、AllGather+Add、Add+AllGather 三个样 例,以及对应 UT 和构建配置。 该能力用于打通 asc-comm 中 CCU 数据面 kernel 的 host 侧直调路径,便于业务样例直接验证 CCU 通信流程,并为后续 asc-devkit 集成、run 包独立构建和 CCU direct 调用提供基础接口与示例支撑。 ## 改动类型 / Change Type - [ ] Bug 修复 / Bug Fix - [x] 新功能 / New Feature - [ ] 性能优化 / Performance - [ ] 代码重构 / Refactoring - [ ] 文档更新 / Documentation - [ ] 测试相关 / Test - [ ] 其它 / Other ## 测试信息 / Testing <!-- 简要测试说明或关键结果 / Brief test description or key results --> - [x] 单元测试通过 / UT passed - [x] 集成测试通过 / ST passed - [x] 人工验证通过 / Manual verified See merge request: cann/asc-comm!60 | 2 个月前 |
| 文件 | 最后提交记录 | 最后更新时间 |
|---|---|---|
| 11 天前 | ||
| 20 天前 | ||
| 13 天前 | ||
| 2 个月前 |