cpp_builder_ut

ascend-robot[feat] custom hccl api with dispatch

4bf5845e创建于 3月5日历史提交

文件	最后提交记录	最后更新时间
_fault_mode_cases	!17268 Fixed the test_fault_mode.py Merge pull request !17268 from yuhaiyan/master-dev1	1 年前
_spmd	!12069 [feature]Support Functional Collectives Merge pull request !12069 from 叶子凡/master_functional_collectives	2 年前
_tensor	fix dropout Co-authored-by: DaiFu<daifu2@huawei.com> # message auto-generated for no-merge-commit merge: !27563 merge v2.7.1 into v2.7.1 fix dropout Created-by: daifu1234567 Commit-by: DaiFu Merged-by: ascend-robot Description: <!-- Thanks for sending a pull request! --> **What type of PR is this?** > Uncomment only one /kind <> line, hit enter to put that in a new line, and remove leading whitespaces from that line: > > /kind task **What does this PR do / why do we need it**: dropout没有注册sharing策略和meta注册，会导致跑dtensor和faketensor报错。考虑到底层用的native_dropout，且已适配，因此删除dropout过适配。 **Special notes for your reviewers**: 验证已跑通单脚本。 Dtensor改前：  Dtensor改后：  faketensor改前：  faketensor改后：  See merge request: Ascend/pytorch!27563	7 个月前
checkpoint	!9796 Add distributed checkpoint support Merge pull request !9796 from chuboning/master	2 年前
fsdp	AI assist developer for python dt third batch for 2.7.1 Co-authored-by: deepSeekya<chenzhihan6@huawei.com> # message auto-generated for no-merge-commit merge: !26518 merge v2.7.1 into v2.7.1 AI assist developer for python dt third batch for 2.7.1 Created-by: deepSeekya Commit-by: deepSeekya Merged-by: ascend-robot Description: 【合入说明】 AI辅助研发python dt测试用例生成。 【合入来源】 AI辅助研发需求 dt测试文件覆盖率提升。 See merge request: Ascend/pytorch!26518	10 个月前
fsdp2	!22803 add fsdp test case Merge pull request !22803 from zhangqiongwen/v2.7.1_fsdp_test_case	1 年前
pipelining	!21456 add pp ut Merge pull request !21456 from 詹昊/v2.7.1	1 年前
rpc	AI assist developer for python dt third batch for 2.7.1 Co-authored-by: deepSeekya<chenzhihan6@huawei.com> # message auto-generated for no-merge-commit merge: !26518 merge v2.7.1 into v2.7.1 AI assist developer for python dt third batch for 2.7.1 Created-by: deepSeekya Commit-by: deepSeekya Merged-by: ascend-robot Description: 【合入说明】 AI辅助研发python dt测试用例生成。 【合入来源】 AI辅助研发需求 dt测试文件覆盖率提升。 See merge request: Ascend/pytorch!26518	10 个月前
shmem	[SHMEM] support npu shmem Co-authored-by: wangchao430<wangchao430@huawei.com> # message auto-generated for no-merge-commit merge: !25901 merge v2.7.0_shmem1 into v2.7.1 [SHMEM] support npu shmem Created-by: wangchao430 Commit-by: wangchao430 Merged-by: ascend-robot Description: <!-- Thanks for sending a pull request! --> **What type of PR is this?** > Uncomment only one /kind <> line, hit enter to put that in a new line, and remove leading whitespaces from that line: > > /kind bug > /kind task > /kind feature **What does this PR do / why do we need it**: **Which issue(s) this PR fixes**: <!-- *Automatically closes linked issue when PR is merged. Usage: Fixes #<issue number>, or Fixes (paste link of issue). --> Fixes # **Special notes for your reviewers**:  See merge request: Ascend/pytorch!25901	10 个月前
tensor	update matmul sharding strategy Co-authored-by: zhangguoguang<zhangguoguang2@huawei.com> # message auto-generated for no-merge-commit merge: !31144 merge update_matmul_sharding_strategy_271 into v2.7.1 update matmul sharding strategy Created-by: zhangguoguang Commit-by: zhangguoguang Merged-by: ascend-robot Description: <!-- PR描述模板更新日期：20260203 --> # 【合入来源】 > <font color="red">**如有社区issue，请关联issue链接**</font>\ > <font color="red">**请勿携带内部流程信息（需求链接、问题单、内部issue等）**</font> - [x] 需求 - [ ] 问题单 - [ ] issue/工单 - [ ] 重构优化 - [ ] 资料更新 # 【修改方案】 补齐matmul正向和反向sharding strategy，根据输入shape，列出所有可能的sharding方式 # 【资料变更】 不涉及 # 【接口变更】 不涉及 # 【功能验证】 新增7个UT用例看护，自验通过  # 【CheckList】 > PR提交人对以下CheckList自检项进行全量自检，自检通过或不涉及，均修改 [ ] 为 [x] - [x] 代码注释完备，正确记录错误日志 - [x] 代码实现进行了返回值、空指针等校验 - [x] PR标题正确使用类型标签，如：feat、fix、refactor、docs、test等 - [x] PR持续集成流水线（CI）执行通过，代码检查无异常 See merge request: Ascend/pytorch!31144	6 个月前
watchdog	!16207 Resolve testcase randomly stuck 30mins. Merge pull request !16207 from jiangpengfei/master	1 年前
test_all_to_all.py	v2.7.1 enable internal format by default for hccl tests on A3 Co-authored-by: freezee<gengdongjie@huawei.com> # message auto-generated for no-merge-commit merge: !27493 merge v2.7.1 into v2.7.1 v2.7.1 enable internal format by default for hccl tests on A3 Created-by: freezee Commit-by: freezee Merged-by: ascend-robot Description: <!-- Thanks for sending a pull request! --> **What type of PR is this?** > Uncomment only one /kind <> line, hit enter to put that in a new line, and remove leading whitespaces from that line: > > /kind bug > /kind task > /kind feature **What does this PR do / why do we need it**: **Which issue(s) this PR fixes**: <!-- *Automatically closes linked issue when PR is merged. Usage: Fixes #<issue number>, or Fixes (paste link of issue). --> Fixes # **Special notes for your reviewers**: See merge request: Ascend/pytorch!27493	9 个月前
test_all_to_all_single.py	!12133 alltoall_base add check "Split sizes dosen't match total dim 0 size" Merge pull request !12133 from wangchao147/master_problem	2 年前
test_allgather.py	v2.7.1 enable internal format by default for hccl tests on A3 Co-authored-by: freezee<gengdongjie@huawei.com> # message auto-generated for no-merge-commit merge: !27493 merge v2.7.1 into v2.7.1 v2.7.1 enable internal format by default for hccl tests on A3 Created-by: freezee Commit-by: freezee Merged-by: ascend-robot Description: <!-- Thanks for sending a pull request! --> **What type of PR is this?** > Uncomment only one /kind <> line, hit enter to put that in a new line, and remove leading whitespaces from that line: > > /kind bug > /kind task > /kind feature **What does this PR do / why do we need it**: **Which issue(s) this PR fixes**: <!-- *Automatically closes linked issue when PR is merged. Usage: Fixes #<issue number>, or Fixes (paste link of issue). --> Fixes # **Special notes for your reviewers**: See merge request: Ascend/pytorch!27493	9 个月前
test_allgather_base.py	!17764 Fixed the failed tests. Merge pull request !17764 from yuhaiyan/master-dev2	1 年前
test_allgather_into_tensor.py	v2.7.1 enable internal format by default for hccl tests on A3 Co-authored-by: freezee<gengdongjie@huawei.com> # message auto-generated for no-merge-commit merge: !27493 merge v2.7.1 into v2.7.1 v2.7.1 enable internal format by default for hccl tests on A3 Created-by: freezee Commit-by: freezee Merged-by: ascend-robot Description: <!-- Thanks for sending a pull request! --> **What type of PR is this?** > Uncomment only one /kind <> line, hit enter to put that in a new line, and remove leading whitespaces from that line: > > /kind bug > /kind task > /kind feature **What does this PR do / why do we need it**: **Which issue(s) this PR fixes**: <!-- *Automatically closes linked issue when PR is merged. Usage: Fixes #<issue number>, or Fixes (paste link of issue). --> Fixes # **Special notes for your reviewers**: See merge request: Ascend/pytorch!27493	9 个月前
test_allreduce.py	Add PREMUL_SUM support in ProcessGroupHCCL for ReduceOp Co-authored-by: dairenjie<dairenjie@huawei.com> # message auto-generated for no-merge-commit merge: !31247 merge v2.7.1_0303 into v2.7.1 Add PREMUL_SUM support in ProcessGroupHCCL for ReduceOp Created-by: dairenjie Commit-by: dairenjie Merged-by: ascend-robot Description: <!-- PR描述模板更新日期：20260203 --> # 【合入来源】 - [x] 需求 - [ ] 问题单 - [ ] issue/工单 - [ ] 重构优化 - [ ] 资料更新 # 【修改方案】 1. 修改ProcessGroupHCCL内reduce类接口的实现，在pre函数添加input.mul_(factor)的调用（9处） 2. 修改getHcclReduceOp内处理premul_sum场景为sum # 【资料变更】 不涉及 # 【接口变更】 增加torch.distributed._make_hccl_premul_sum 和PT原生接口_make_nccl_premul_sum保持对等，已评审。 # 【功能验证】 修改/新增reduce算子premul_sum场景测试用例。 # 【CheckList】 > PR提交人对以下CheckList自检项进行全量自检，自检通过或不涉及，均修改 [ ] 为 [x] - [x] 代码注释完备，正确记录错误日志 - [x] 代码实现进行了返回值、空指针等校验 - [x] PR标题正确使用类型标签，如：feat、fix、refactor、docs、test等 - [x] PR持续集成流水线（CI）执行通过，代码检查无异常 See merge request: Ascend/pytorch!31247	6 个月前
test_api_get_hcom_name.py	!13554 feature: get_hcom_name_api support get and not init hccl Merge pull request !13554 from yangxiaorun/master	2 年前
test_api_set_hcom_name.py	!13853 update _is_support_hccl_comm_name api Merge pull request !13853 from 关龙锋/cherry-pick-1724120059	2 年前
test_avoid_record_stream.py	[MULTI_STREAM_MEMORY_REUSE] Align CUDA, and in the AVOID_RECORD_STREAM scenario, fall back to the recordStream logic. Co-authored-by: wangchao430<wangchao430@huawei.com> # message auto-generated for no-merge-commit merge: !26926 merge v2.7.0_multistream into v2.7.1 [MULTI_STREAM_MEMORY_REUSE] Align CUDA, and in the AVOID_RECORD_STREAM scenario, fall back to the recordStream logic. Created-by: wangchao430 Commit-by: wangchao430 Merged-by: ascend-robot Description: <!-- Thanks for sending a pull request! --> **What type of PR is this?** > Uncomment only one /kind <> line, hit enter to put that in a new line, and remove leading whitespaces from that line: > > /kind bug > /kind task > /kind feature **What does this PR do / why do we need it**: **Which issue(s) this PR fixes**: <!-- *Automatically closes linked issue when PR is merged. Usage: Fixes #<issue number>, or Fixes (paste link of issue). --> Fixes # **Special notes for your reviewers**: See merge request: Ascend/pytorch!26926	9 个月前
test_base_distributed.py	!7010 Rename test files: test_distributed/test_fsdp/test_rpc Merge pull request !7010 from yuhaiyan/master-dev2	2 年前
test_batch_isend_irecv.py	v2.7.1 enable internal format by default for hccl tests on A3 Co-authored-by: freezee<gengdongjie@huawei.com> # message auto-generated for no-merge-commit merge: !27493 merge v2.7.1 into v2.7.1 v2.7.1 enable internal format by default for hccl tests on A3 Created-by: freezee Commit-by: freezee Merged-by: ascend-robot Description: <!-- Thanks for sending a pull request! --> **What type of PR is this?** > Uncomment only one /kind <> line, hit enter to put that in a new line, and remove leading whitespaces from that line: > > /kind bug > /kind task > /kind feature **What does this PR do / why do we need it**: **Which issue(s) this PR fixes**: <!-- *Automatically closes linked issue when PR is merged. Usage: Fixes #<issue number>, or Fixes (paste link of issue). --> Fixes # **Special notes for your reviewers**: See merge request: Ascend/pytorch!27493	9 个月前
test_broadcast.py	v2.7.1 enable internal format by default for hccl tests on A3 Co-authored-by: freezee<gengdongjie@huawei.com> # message auto-generated for no-merge-commit merge: !27493 merge v2.7.1 into v2.7.1 v2.7.1 enable internal format by default for hccl tests on A3 Created-by: freezee Commit-by: freezee Merged-by: ascend-robot Description: <!-- Thanks for sending a pull request! --> **What type of PR is this?** > Uncomment only one /kind <> line, hit enter to put that in a new line, and remove leading whitespaces from that line: > > /kind bug > /kind task > /kind feature **What does this PR do / why do we need it**: **Which issue(s) this PR fixes**: <!-- *Automatically closes linked issue when PR is merged. Usage: Fixes #<issue number>, or Fixes (paste link of issue). --> Fixes # **Special notes for your reviewers**: See merge request: Ascend/pytorch!27493	9 个月前
test_c10d_functional_native.py	!15687 Support collectiveCoalesced. Merge pull request !15687 from jiangpengfei/master	1 年前
test_c10d_object_collectives.py	!13707 Involve with_comms in test_c10d_object_collectives.py Merge pull request !13707 from dilililiwhy/cherry-pick-1723529592	2 年前
test_coalesced_manager.py	check the size of tensors for alltoall Co-authored-by: tonglei<tonglei10@huawei.com> # message auto-generated for no-merge-commit merge: !29731 merge v2.7.1 into v2.7.1 check the size of tensors for alltoall Created-by: tongleicom Commit-by: tonglei Merged-by: ascend-robot Description: <!-- Thanks for sending a pull request! --> **What type of PR is this?** > Uncomment only one /kind <> line, hit enter to put that in a new line, and remove leading whitespaces from that line: > > /kind bug > /kind task > /kind feature **What does this PR do / why do we need it**: **Which issue(s) this PR fixes**: <!-- *Automatically closes linked issue when PR is merged. Usage: Fixes #<issue number>, or Fixes (paste link of issue). --> Fixes # **Special notes for your reviewers**: See merge request: Ascend/pytorch!29731	8 个月前
test_comm_hook.py	!7891 add ut for torch.futures Merge pull request !7891 from 邵非凡/futures	2 年前
test_data_parallel.py	!11816 bugfix: rankid can not trans to device id, use current device Merge pull request !11816 from yangxiaorun/cherry-pick-1715995071	2 年前
test_device.py	torch_npu.npu.Event inherits torch.Event, add GetPerStreamQueue in NPUGuardImpl's function Co-authored-by: zhaoyu<nanzhaogang@qq.com> # message auto-generated for no-merge-commit merge: !29915 merge v2.7.1-event_stream into v2.7.1 torch_npu.npu.Event inherits torch.Event, add GetPerStreamQueue in NPUGuardImpl's function Created-by: zhaoyu65 Commit-by: zhaoyu Merged-by: ascend-robot Description: # 【合入来源】 > (如有)请关联需求文档/issue链接 - [x] 需求 event/stream等接口对齐社区接入逻辑 - [ ] 问题单 - [ ] issue/工单 - [ ] 重构优化 - [ ] 资料更新 # 【修改方案】 1、将torch_npu.npu.Event修改为torch.Event的子类，和pytorch原生代码行为保持一致； 2、实现NPUGuardImpl::uncheckedGetDevice()； 3、NPUGuardImpl::block()、destroyEvent()支持PerStreamQueue，参考NPUEvent； 4、优化NPUGuardImpl::synchronizeEvent()； # 【资料变更】 不涉及 # 【接口变更】 不涉及 # 【功能验证】 1、使用isinstance、issubclass方法校验torch_npu.npu.Event和torch.Event的关系； 2、验证单流单queue场景是否正常； # 【CheckList】 > PR提交人对以下CheckList自检项进行全量自检，自检通过或不涉及，均修改 [ ] 为 [x] - [x] 代码注释完备，正确记录错误日志 - [x] 代码实现进行了返回值、空指针等校验 - [x] PR标题正确使用类型标签，如：feat、fix、refactor、docs、test等 - [x] PR持续集成流水线（CI）执行通过，代码检查无异常 See merge request: Ascend/pytorch!29915	7 个月前
test_device_mesh.py	refactor shutdown and destructor to match torch behavior Co-authored-by: dinghong_zou<zoudinghong3@huawei.com> # message auto-generated for no-merge-commit merge: !30623 merge v2.7.1 into v2.7.1 refactor shutdown and destructor to match torch behavior Created-by: Windwindzzz Commit-by: dinghong_zou Merged-by: ascend-robot Description: <!-- PR描述模板更新日期：20260203 --> # 【合入来源】 > (如有)请关联需求文档/issue链接 - [x] 需求 - [x] 问题单 - [ ] issue/工单 - [ ] 重构优化 - [ ] 资料更新 # 【修改方案】 > 请描述修改内容的具体实现，涉及哪些组件之间进行交互，可以用1、2、3、...进行罗列\ > 如果是需求或者重构类的PR，需要补充详细设计文档（说明上下游组件关系、时序图、类图、DFX能力等内容） 将关闭通信域从ProcessGroupHCCL类析构函数前移到shutdown()函数 去除shutdown()的参数，从而能够正确继承基类的shutdown() # 【资料变更】 > 请确认是否涉及资料变更。如涉及，需要在PR中体现，并简要说明修改内容。如不涉及，需填写“不涉及” 不涉及 # 【接口变更】 > 请确认是否涉及跨代码仓或者客户面可见的接口变更。如涉及，需要详细说明接口以及对应的变更内容，同时需要在资料中体现。如不涉及，需填写“不涉及” 不涉及 # 【功能验证】 > 说明测试场景，测试方法。如果本次测试方式与常规单元测试不同，请详细说明您的测试步骤\ > 新增/变更内容是否已新增/适配UT测试用例看护，并补充测试自验证截图 单元测试：无新增 UT 测试用例，该功能可用库上现有UT用例覆盖 回归测试：执行 torch_npu/testing 模块下的所有现有 UT 测试套件，确认本次修改未影响其他功能； 手动验证：在测试环境中导入 torch_npu.testing 模块，调用destory_process_group方法，通过日志确认通信域在shutdown()方法中关闭。 # 【CheckList】 > PR提交人对以下CheckList自检项进行全量自检，自检通过或不涉及，均修改 [ ] 为 [x] - [x] 代码注释完备，正确记录错误日志 - [x] 代码实现进行了返回值、空指针等校验 - [x] PR标题正确使用类型标签，如：feat、fix、refactor、docs、test等 - [x] PR持续集成流水线（CI）执行通过，代码检查无异常 See merge request: Ascend/pytorch!30623	6 个月前
test_distributed.py	skip test_hccl_timeout_dumps Co-authored-by: zhaoyu<nanzhaogang@qq.com> # message auto-generated for no-merge-commit merge: !26893 merge v2.7.1 into v2.7.1 skip test_hccl_timeout_dumps Created-by: zhaoyu65 Commit-by: zhaoyu Merged-by: ascend-robot Description: <!-- Thanks for sending a pull request! --> **What type of PR is this?** > Uncomment only one /kind <> line, hit enter to put that in a new line, and remove leading whitespaces from that line: > > /kind bug > /kind task > /kind feature **What does this PR do / why do we need it**: **Which issue(s) this PR fixes**: <!-- *Automatically closes linked issue when PR is merged. Usage: Fixes #<issue number>, or Fixes (paste link of issue). --> Fixes # **Special notes for your reviewers**: See merge request: Ascend/pytorch!26893	10 个月前
test_fault_mode.py	Update test_hccl_timeout Co-authored-by: yuhaiyan<yuhaiyan8@huawei.com> Co-authored-by: yuhaiyan<yuhaiyan8@h-partners.com> # message auto-generated for no-merge-commit merge: !28138 merge cherry-pick-mr-28137-1765873594786-auto into v2.7.1 Update test_hccl_timeout Created-by: yuhaiyan Commit-by: yuhaiyan Merged-by: ascend-robot Description: <!-- Thanks for sending a pull request! --> **What type of PR is this?** > Uncomment only one /kind <> line, hit enter to put that in a new line, and remove leading whitespaces from that line: > > /kind bug > /kind task > /kind feature **What does this PR do / why do we need it**: **Which issue(s) this PR fixes**: <!-- *Automatically closes linked issue when PR is merged. Usage: Fixes #<issue number>, or Fixes (paste link of issue). --> Fixes # **Special notes for your reviewers**: See merge request: Ascend/pytorch!28138	9 个月前
test_flight_recorder.py	skip test_hccl_timeout_dumps Co-authored-by: zhaoyu<nanzhaogang@qq.com> # message auto-generated for no-merge-commit merge: !26893 merge v2.7.1 into v2.7.1 skip test_hccl_timeout_dumps Created-by: zhaoyu65 Commit-by: zhaoyu Merged-by: ascend-robot Description: <!-- Thanks for sending a pull request! --> **What type of PR is this?** > Uncomment only one /kind <> line, hit enter to put that in a new line, and remove leading whitespaces from that line: > > /kind bug > /kind task > /kind feature **What does this PR do / why do we need it**: **Which issue(s) this PR fixes**: <!-- *Automatically closes linked issue when PR is merged. Usage: Fixes #<issue number>, or Fixes (paste link of issue). --> Fixes # **Special notes for your reviewers**: See merge request: Ascend/pytorch!26893	10 个月前
test_gather.py	!17048 bugfix for ddp ut Merge pull request !17048 from 邵非凡/dptest	1 年前
test_hccl_shared_buffer.py	hccl options to support hccl buffer name Co-authored-by: wanghonghao<wanghonghao9@huawei.com> # message auto-generated for no-merge-commit merge: !27549 merge 271_cclbuffer into v2.7.1 hccl options to support hccl buffer name Created-by: wanghonghao Commit-by: wanghonghao Merged-by: ascend-robot Description: <!-- Thanks for sending a pull request! --> **What type of PR is this?** /kind feature **What does this PR do / why do we need it**: hccl options to support hccl buffer name **Which issue(s) this PR fixes**: <!-- *Automatically closes linked issue when PR is merged. Usage: Fixes #<issue number>, or Fixes (paste link of issue). --> Fixes # **Special notes for your reviewers**: See merge request: Ascend/pytorch!27549	9 个月前
test_hccl_stream_id.py	!23271 set default streampool to 32 Merge pull request !23271 from SCh-zx/stream27	1 年前
test_inductor_collectives.py	!12069 [feature]Support Functional Collectives Merge pull request !12069 from 叶子凡/master_functional_collectives	2 年前
test_multi_accelerator.py	!21271 [3/N] Add multi accelerator cases Merge pull request !21271 from dilililiwhy/add_multi_accelerator_cases	1 年前
test_nslb.py	!16593 Fix Path Manager For Security Merge pull request !16593 from 周先琪/master_zxq	1 年前
test_options.py	request for hccl qos Co-authored-by: gcw_Dgfy2aKk<dujunjie2@huawei.com> # message auto-generated for no-merge-commit merge: !30439 merge Hccl_Qos_v2.7.1 into v2.7.1 request for hccl qos Created-by: gcw_Dgfy2aKk Commit-by: gcw_Dgfy2aKk Merged-by: ascend-robot Description: <!-- PR描述模板更新日期：20260203 --> # 【合入来源】 > (如有)请关联需求文档/issue链接 开放流量的QoS配置接口，并向CANN传递 - [x] 需求 # 【修改方案】 透传hccl qos参数给HCCL # 【资料变更】 不涉及 # 【接口变更】 不涉及 # 【功能验证】 执行 python test/distributed/test_options.py -k test_options_hccl_sdma_qos_wrong_type 执行结果： 执行通过， 无报错 # 【CheckList】 > PR提交人对以下CheckList自检项进行全量自检，自检通过或不涉及，均修改 [ ] 为 [x] - [x] 代码注释完备，正确记录错误日志 - [x] 代码实现进行了返回值、空指针等校验 - [x] PR标题正确使用类型标签，如：feat、fix、refactor、docs、test等 - [x] PR持续集成流水线（CI）执行通过，代码检查无异常 See merge request: Ascend/pytorch!30439	7 个月前
test_parallel_store.py	!16974 parallel tcpstore support tiered optimizer Merge pull request !16974 from wuxiaotong/cherry-pick-1734348942	1 年前
test_parallel_store_tiered_optimizer.py	!16974 parallel tcpstore support tiered optimizer Merge pull request !16974 from wuxiaotong/cherry-pick-1734348942	1 年前
test_pin_memory.py	!14369 don't change current device when pin_memory=True Merge pull request !14369 from huangyunlong/2.5pin	2 年前
test_reduce.py	Add PREMUL_SUM support in ProcessGroupHCCL for ReduceOp Co-authored-by: dairenjie<dairenjie@huawei.com> # message auto-generated for no-merge-commit merge: !31247 merge v2.7.1_0303 into v2.7.1 Add PREMUL_SUM support in ProcessGroupHCCL for ReduceOp Created-by: dairenjie Commit-by: dairenjie Merged-by: ascend-robot Description: <!-- PR描述模板更新日期：20260203 --> # 【合入来源】 - [x] 需求 - [ ] 问题单 - [ ] issue/工单 - [ ] 重构优化 - [ ] 资料更新 # 【修改方案】 1. 修改ProcessGroupHCCL内reduce类接口的实现，在pre函数添加input.mul_(factor)的调用（9处） 2. 修改getHcclReduceOp内处理premul_sum场景为sum # 【资料变更】 不涉及 # 【接口变更】 增加torch.distributed._make_hccl_premul_sum 和PT原生接口_make_nccl_premul_sum保持对等，已评审。 # 【功能验证】 修改/新增reduce算子premul_sum场景测试用例。 # 【CheckList】 > PR提交人对以下CheckList自检项进行全量自检，自检通过或不涉及，均修改 [ ] 为 [x] - [x] 代码注释完备，正确记录错误日志 - [x] 代码实现进行了返回值、空指针等校验 - [x] PR标题正确使用类型标签，如：feat、fix、refactor、docs、test等 - [x] PR持续集成流水线（CI）执行通过，代码检查无异常 See merge request: Ascend/pytorch!31247	6 个月前
test_reduce_scatter.py	Add PREMUL_SUM support in ProcessGroupHCCL for ReduceOp Co-authored-by: dairenjie<dairenjie@huawei.com> # message auto-generated for no-merge-commit merge: !31247 merge v2.7.1_0303 into v2.7.1 Add PREMUL_SUM support in ProcessGroupHCCL for ReduceOp Created-by: dairenjie Commit-by: dairenjie Merged-by: ascend-robot Description: <!-- PR描述模板更新日期：20260203 --> # 【合入来源】 - [x] 需求 - [ ] 问题单 - [ ] issue/工单 - [ ] 重构优化 - [ ] 资料更新 # 【修改方案】 1. 修改ProcessGroupHCCL内reduce类接口的实现，在pre函数添加input.mul_(factor)的调用（9处） 2. 修改getHcclReduceOp内处理premul_sum场景为sum # 【资料变更】 不涉及 # 【接口变更】 增加torch.distributed._make_hccl_premul_sum 和PT原生接口_make_nccl_premul_sum保持对等，已评审。 # 【功能验证】 修改/新增reduce算子premul_sum场景测试用例。 # 【CheckList】 > PR提交人对以下CheckList自检项进行全量自检，自检通过或不涉及，均修改 [ ] 为 [x] - [x] 代码注释完备，正确记录错误日志 - [x] 代码实现进行了返回值、空指针等校验 - [x] PR标题正确使用类型标签，如：feat、fix、refactor、docs、test等 - [x] PR持续集成流水线（CI）执行通过，代码检查无异常 See merge request: Ascend/pytorch!31247	6 个月前
test_reduce_scatter_base.py	Add PREMUL_SUM support in ProcessGroupHCCL for ReduceOp Co-authored-by: dairenjie<dairenjie@huawei.com> # message auto-generated for no-merge-commit merge: !31247 merge v2.7.1_0303 into v2.7.1 Add PREMUL_SUM support in ProcessGroupHCCL for ReduceOp Created-by: dairenjie Commit-by: dairenjie Merged-by: ascend-robot Description: <!-- PR描述模板更新日期：20260203 --> # 【合入来源】 - [x] 需求 - [ ] 问题单 - [ ] issue/工单 - [ ] 重构优化 - [ ] 资料更新 # 【修改方案】 1. 修改ProcessGroupHCCL内reduce类接口的实现，在pre函数添加input.mul_(factor)的调用（9处） 2. 修改getHcclReduceOp内处理premul_sum场景为sum # 【资料变更】 不涉及 # 【接口变更】 增加torch.distributed._make_hccl_premul_sum 和PT原生接口_make_nccl_premul_sum保持对等，已评审。 # 【功能验证】 修改/新增reduce算子premul_sum场景测试用例。 # 【CheckList】 > PR提交人对以下CheckList自检项进行全量自检，自检通过或不涉及，均修改 [ ] 为 [x] - [x] 代码注释完备，正确记录错误日志 - [x] 代码实现进行了返回值、空指针等校验 - [x] PR标题正确使用类型标签，如：feat、fix、refactor、docs、test等 - [x] PR持续集成流水线（CI）执行通过，代码检查无异常 See merge request: Ascend/pytorch!31247	6 个月前
test_reduce_scatter_tensor.py	[feat] custom hccl api with dispatch Co-authored-by: pengqi<pengqi33@huawei.com> # message auto-generated for no-merge-commit merge: !31008 merge v2.7.1 into v2.7.1 [feat] custom hccl api with dispatch Created-by: pengqihw Commit-by: pengqi Merged-by: ascend-robot Description: <!-- PR描述模板更新日期：20260203 --> 为自定义HCCL分布式API添加AsyncCollectiveTensor支持，实现自动wait触发机制。 # 【合入来源】 > <font color="red">**如有社区issue，请关联issue链接**</font>\ > <font color="red">**请勿携带内部流程信息（需求链接、问题单、内部issue等）**</font> - [x] 需求 - [ ] 问题单 - [ ] issue/工单 - [ ] 重构优化 - [ ] 资料更新 # 【修改方案】 1. 新增文件 - test/distributed/test_async_collective_tensor.py - 新增测试用例，验证三个API对AsyncCollectiveTensor的支持 - torch_npu/csrc/distributed/HcclOps.cpp - 新增HCCL操作实现 2. 核心修改 - ProcessGroupHCCL.cpp/hpp - 为以下API添加dispatch机制，支持AsyncCollectiveTensor自动触发wait： - batch_isend_irecv - reduce_scatter_tensor_uneven - all_gather_into_tensor_uneven 3. 其他修改 - test/distributed/test_reduce_scatter_tensor.py - 测试用例调整 - torch_npu_schema.json - schema更新 - Init.cpp - 初始化相关调整 # 【资料变更】 不涉及 # 【接口变更】 不涉及 # 【功能验证】 新增测试用例覆盖： - AsyncCollectiveTensor类型验证 - completed状态验证（初始False，调用后True） - 结果一致性验证（与普通Tensor结果对比）  # 【CheckList】 > PR提交人对以下CheckList自检项进行全量自检，自检通过或不涉及，均修改 [ ] 为 [x] - [ ] 代码注释完备，正确记录错误日志 - [ ] 代码实现进行了返回值、空指针等校验 - [ ] PR标题正确使用类型标签，如：feat、fix、refactor、docs、test等 - [ ] PR持续集成流水线（CI）执行通过，代码检查无异常 See merge request: Ascend/pytorch!31008	6 个月前
test_register_sharding.py	Fix UT from test_register_sharding Co-authored-by: zhangguoguang<zhangguoguang2@huawei.com> # message auto-generated for no-merge-commit merge: !29159 merge fix_UT_271 into v2.7.1 Fix UT from test_register_sharding Created-by: zhangguoguang Commit-by: zhangguoguang Merged-by: ascend-robot Description: <!-- Thanks for sending a pull request! --> **What type of PR is this?** > Uncomment only one /kind <> line, hit enter to put that in a new line, and remove leading whitespaces from that line: > > /kind bug > /kind task > /kind feature **What does this PR do / why do we need it**: **Which issue(s) this PR fixes**: <!-- *Automatically closes linked issue when PR is merged. Usage: Fixes #<issue number>, or Fixes (paste link of issue). --> Fixes # **Special notes for your reviewers**: See merge request: Ascend/pytorch!29159	8 个月前
test_rendezvous.py	AI assist developer for python DT second batch for 2.7.1 Co-authored-by: deepSeekya<chenzhihan6@huawei.com> # message auto-generated for no-merge-commit merge: !26370 merge v2.7.1 into v2.7.1 AI assist developer for python DT second batch for 2.7.1 Created-by: deepSeekya Commit-by: deepSeekya Merged-by: ascend-robot Description: 【合入说明】 AI辅助研发python dt测试用例生成。 【合入来源】 AI辅助研发需求 dt测试文件覆盖率提升。 See merge request: Ascend/pytorch!26370	10 个月前
test_run.py	AI assist developer for python DT second batch for 2.7.1 Co-authored-by: deepSeekya<chenzhihan6@huawei.com> # message auto-generated for no-merge-commit merge: !26370 merge v2.7.1 into v2.7.1 AI assist developer for python DT second batch for 2.7.1 Created-by: deepSeekya Commit-by: deepSeekya Merged-by: ascend-robot Description: 【合入说明】 AI辅助研发python dt测试用例生成。 【合入来源】 AI辅助研发需求 dt测试文件覆盖率提升。 See merge request: Ascend/pytorch!26370	10 个月前
test_scatter.py	!19882 fix reduce_scatter_with_different_shape_avg Merge pull request !19882 from zhangqiongwen/master_avg_fix	1 年前
test_send_recv.py	v2.7.1 enable internal format by default for hccl tests on A3 Co-authored-by: freezee<gengdongjie@huawei.com> # message auto-generated for no-merge-commit merge: !27493 merge v2.7.1 into v2.7.1 v2.7.1 enable internal format by default for hccl tests on A3 Created-by: freezee Commit-by: freezee Merged-by: ascend-robot Description: <!-- Thanks for sending a pull request! --> **What type of PR is this?** > Uncomment only one /kind <> line, hit enter to put that in a new line, and remove leading whitespaces from that line: > > /kind bug > /kind task > /kind feature **What does this PR do / why do we need it**: **Which issue(s) this PR fixes**: <!-- *Automatically closes linked issue when PR is merged. Usage: Fixes #<issue number>, or Fixes (paste link of issue). --> Fixes # **Special notes for your reviewers**: See merge request: Ascend/pytorch!27493	9 个月前
test_silent_check.py	!18801 SilentCheck: Add false detection test case for broadcast. Merge pull request !18801 from 王超/v2.7.0_silentbroadcast	1 年前
test_watchdog.py	!13113 Remove judgement of done in watchdog to slove the stucking problem of error exit Merge pull request !13113 from jiangpengfei/master	2 年前
test_zero_redundancy_optimizer.py	register strategy for npu custom ops Co-authored-by: jizewei<jizewei@huawei.com> # message auto-generated for no-merge-commit merge: !27074 merge v2.7.1_register_strategy into v2.7.1 register strategy for npu custom ops Created-by: jizewei Commit-by: jizewei Merged-by: ascend-robot Description: <!-- Thanks for sending a pull request! --> **What type of PR is this?** > Uncomment only one /kind <> line, hit enter to put that in a new line, and remove leading whitespaces from that line: > > /kind bug > /kind task > /kind feature /kind feature **What does this PR do / why do we need it**: register sharding strategies for npu custom ops to support DTensor: 1. npu_add_rms_norm 2. npu_moe_token_permute, npu_moe_token_permute_grad 3. npu_moe_token_unpermute, npu_moe_token_unpermute_grad 4. npu_rotary_mul, npu_rotary_mul_backward **Which issue(s) this PR fixes**: <!-- *Automatically closes linked issue when PR is merged. Usage: Fixes #<issue number>, or Fixes (paste link of issue). --> Fixes # **Special notes for your reviewers**: See merge request: Ascend/pytorch!27074	9 个月前