| 文件 | 最后提交记录 | 最后更新时间 |
|---|---|---|
【Bugfix】修复日志打印引入tqdm依赖导致锁竞争的问题 Co-authored-by: HowSir_X<zhonghao31@huawei.com> # message auto-generated for no-merge-commit merge: !1014 merge fix_log into master 【Bugfix】修复日志打印引入tqdm依赖导致锁竞争的问题 Created-by: HowSir_X Commit-by: HowSir_X Merged-by: ascend-robot Description: 【Bugfix】修复日志打印引入tqdm依赖导致锁竞争的问题 回退之前日志打印,引入tqdm的代码 2026-09-04 14:26:08 (1307277) [INFO] Successfully analyzed distributed nodes. [Graph] All tasks completed: 100%|██████████| 100% [01:01<00:00, 16.35it/s] See merge request: Ascend/msprobe!1014 | 1 个月前 | |
fix: improve config_check help message Co-authored-by: h00613304<hekunkun@huawei.com> # message auto-generated for no-merge-commit merge: !988 merge config-check-help-message-fix into master fix: improve config_check help message Created-by: kun_8 Commit-by: h00613304 Merged-by: ascend-robot Description: fix: improve config_check help message msprobe config_check –help  msprobe api_precision_compare -h  msprobe graph_visualize -h  msprobe data2db -h  msprobe parse –h  msprobe offline_dump -h  See merge request: Ascend/msprobe!988 | 1 个月前 | |
[feat]Device MD5构建工程 Co-authored-by: Tjh-UKN<2559659915@qq.com> # message auto-generated for no-merge-commit merge: !711 merge md5 into master [feat]Device MD5构建工程 Created-by: Tjh-UKN Commit-by: Tjh-UKN Merged-by: ascend-robot Description: 背景:MD5计算需要host同步,现在提供device上计算MD5的算子,加速MD5计算 自验证:https://www.yuque.com/taejohnson/tzgbbf/cbai414pbcpgr9mc See merge request: Ascend/msprobe!711 | 4 个月前 | |
【bugfix】修复mindspore低版本兼容性问题 Co-authored-by: curry808<wurui56@h-partners.com> # message auto-generated for no-merge-commit merge: !947 merge master into master 【bugfix】修复mindspore低版本兼容性问题 Created-by: curry808 Commit-by: curry808 Merged-by: ascend-robot Description: 【bugfix】修复mindspore低版本兼容性问题 See merge request: Ascend/msprobe!947 | 1 个月前 | |
【安全】delete rest owner check in offline dump and compare Co-authored-by: ylw1234<lwying007@126.com> # message auto-generated for no-merge-commit merge: !741 merge master into master 【安全】delete rest owner check in offline dump and compare Created-by: ylw1234 Commit-by: ylw1234 Merged-by: ascend-robot Description: delete rest owner check in offline dump and compare See merge request: Ascend/msprobe!741 | 4 个月前 | |
fix: correct overflow_check help message Co-authored-by: h00613304<hekunkun@huawei.com> # message auto-generated for no-merge-commit merge: !994 merge overflow-check-help-message-fix into master fix: correct overflow_check help message Created-by: kun_8 Commit-by: h00613304 Merged-by: ascend-robot Description: fix: correct overflow_check help message See merge request: Ascend/msprobe!994 | 1 个月前 | |
[bugfix]修正 aclgraph_dump/_meta.py 中 register_fake 算子命名格式(ns.op → ns::op) Co-authored-by: Tjh-UKN<2559659915@qq.com> # message auto-generated for no-merge-commit merge: !1011 merge fix/aclgraph-meta-register-fake-sep into master [bugfix]修正 aclgraph_dump/_meta.py 中 register_fake 算子命名格式(ns.op → ns::op) Created-by: Tjh-UKN Commit-by: Tjh-UKN Merged-by: ascend-robot Description: ### 概述 修复 #580 中描述的问题: torch.library.register_fake 要求算子名称使用 ns::op 双冒号格式,原代码使用 my_ns.acl_save 等点号格式,注册报错后被 except Exception: pass 静默吞掉,导致 acl_save / acl_tensor_save / acl_stat 三个算子的 fake(meta) 实现从未实际注册成功。 ### 修改内容 python/msprobe/pytorch/aclgraph_dump/_meta.py 共 3 处: - my_ns.acl_save → my_ns::acl_save - my_ns.acl_tensor_save → my_ns::acl_tensor_save - my_ns.acl_stat → my_ns::acl_stat ### 关联 Issue Fixes https://gitcode.com/Ascend/msprobe/issues/580 See merge request: Ascend/msprobe!1011 | 1 个月前 | |
【feature】统一帮助信息规范 Co-authored-by: h00613304<hekunkun@huawei.com> # message auto-generated for no-merge-commit merge: !953 merge master into master 【feature】统一帮助信息规范 Created-by: kun_8 Commit-by: h00613304 Merged-by: ascend-robot Description: 统一msprobe下命令行-h的输出格式,具体格式如下图所示  See merge request: Ascend/msprobe!953 | 1 个月前 | |
fix: improve config_check help message Co-authored-by: h00613304<hekunkun@huawei.com> # message auto-generated for no-merge-commit merge: !988 merge config-check-help-message-fix into master fix: improve config_check help message Created-by: kun_8 Commit-by: h00613304 Merged-by: ascend-robot Description: fix: improve config_check help message msprobe config_check –help  msprobe api_precision_compare -h  msprobe graph_visualize -h  msprobe data2db -h  msprobe parse –h  msprobe offline_dump -h  See merge request: Ascend/msprobe!988 | 1 个月前 | |
【bugfix】解决环境中有tensorflow环境,导入工具模块报错的问题 Co-authored-by: h00613304<hekunkun@huawei.com> # message auto-generated for no-merge-commit merge: !651 merge master into master 【bugfix】解决环境中有tensorflow环境,导入工具模块报错的问题 Created-by: kun_8 Commit-by: h00613304 Merged-by: ascend-robot Description: 执行以下代码时发生段错误: python from msprobe.pytorch import PrecisionDebugger # ← 触发崩溃 **错误输出:** Segmentation fault (core dumped) --- ## 崩溃堆栈分析 ### 调用链 Python 导入链: test_core_dump_error.py └── msprobe.pytorch └── msprobe/pytorch/monitor/data_writers.py:24 └── torch.utils.tensorboard.SummaryWriter └── tensorboard.compat.tf (懒加载) └── tensorflow └── tensorflow.python.platform.self_check.preload_check() └── _pywrap_cpu_feature_guard.InfoAboutUnusedCPUFeatures() ← 崩溃点 ### 崩溃点代码 python # tensorflow/python/platform/self_check.py:59-64 else: # Load a library that performs CPU feature guard checking. from tensorflow.python.platform import _pywrap_cpu_feature_guard _pywrap_cpu_feature_guard.InfoAboutUnusedCPUFeatures() # ← Segfault here ## 根因分析 TensorFlow 的 _pywrap_cpu_feature_guard 是一个 C++ 扩展模块,用于检测 CPU 支持的指令集特性(如 AVX, SSE, NEON 等)。 在 **ARM aarch64** 架构上,TensorFlow 2.21.0 的预编译 wheel 可能存在以下问题: 1. **CPU 特性检测代码路径未正确适配 ARM 架构** 2. **与 NPU 驱动/运行时环境存在内存冲突** 3. **缺少必要的 CPU 指令集支持检测** ### 触发路径 msprobe 使用 PyTorch 的 SummaryWriter 进行日志记录,而 PyTorch 的 tensorboard 集成会尝试导入 TensorBoard。TensorBoard 的 compat 模块会懒加载 TensorFlow 以提供兼容性支持,最终导致 TensorFlow 的初始化代码被执行。 ## 解决方案 ### 创建自动补丁文件 python """msprobe 导入补丁 - 在导入 msprobe 前加载此模块""" import sys try: import tensorboard.compat.tensorflow_stub as tensorflow_stub sys.modules['tensorflow'] = tensorflow_stub sys.modules['tensorflow.python'] = tensorflow_stub except ImportError: pass See merge request: Ascend/msprobe!651 | 4 个月前 | |
【feature】命令行version参数返回版本信息 Co-authored-by: h00613304<hekunkun@huawei.com> # message auto-generated for no-merge-commit merge: !964 merge master into master 【feature】命令行version参数返回版本信息 Created-by: kun_8 Commit-by: h00613304 Merged-by: ascend-robot Description: msprobe增加-V命令,并按照以下格式展示  See merge request: Ascend/msprobe!964 | 1 个月前 | |
[master]adump模块日落 Co-authored-by: jiangchaoj<chaojiang_j@163.com> # message auto-generated for no-merge-commit merge: !832 merge master-del_adump into master [master]adump模块日落 Created-by: jianchaoj Commit-by: jiangchaoj Merged-by: ascend-robot Description: # PR 合入模板 **注:经过自检不涉及的可标注“不涉及”或直接打勾,特殊情况请文字备注。不符合规范的 PR 不允许合入,请(后备)commit 注意。** ## 1. 修改描述 - **修改原因:** MindSpore不再支持ge_backend模式,msProbe同步移除依赖该模式得adump模块 - **修改内容:** - 删除adump源码、编译代码。修改工具安装指南 - 移除MindSpore ge_backend模式下的Dump功能,保留ms_backend模式下的Dump功能。删除配置文件中的"file_format"参数,修改对应资料 - 移除MindSpore ge_backend模式下的溢出检测功能,保留ms_backend模式下的异常Dump功能。删除配置文件中的"overflow_check" task,修改对应资料 - 移除MindSpore动态图模式下,API Kernel Dump功能。修改对应资料 ## 2. 功能验证 - 编译  - PyTorch Dump  - PyTorch overflow_check  - MindSpore 动态图Dump  - MindSpore 静态图L0 Dump  - MindSpore 静态图L2 Dump  - MindSpore overflow_check  - MindSpore 动态图L2 Dump  ## 3. 分支合并要求 同步合入master、26.1.0分支 ## 3. 代码检视 - **要求:** - 合入代码超过 200 行,需三人以上会议检视。 - 检视密度≥1个/100行。 - 检视缺陷密度未达要求需提供说明。 - 大于 1000 行代码原则上不允许合入,需进行备案。 - [x] **是否经过代码检视** - [x] **是否具备 UT 测试用例看护** (如不符合,请说明原因:____________________) ## 4. 变更知会 - [x] **资料修改** - [x] **变更通知(消息知会 + 邮件知会)** See merge request: Ascend/msprobe!832 | 3 个月前 | |
【feature】命令行version参数返回版本信息 Co-authored-by: h00613304<hekunkun@huawei.com> # message auto-generated for no-merge-commit merge: !964 merge master into master 【feature】命令行version参数返回版本信息 Created-by: kun_8 Commit-by: h00613304 Merged-by: ascend-robot Description: msprobe增加-V命令,并按照以下格式展示  See merge request: Ascend/msprobe!964 | 1 个月前 | |
【feature】命令行version参数返回版本信息 Co-authored-by: h00613304<hekunkun@huawei.com> # message auto-generated for no-merge-commit merge: !964 merge master into master 【feature】命令行version参数返回版本信息 Created-by: kun_8 Commit-by: h00613304 Merged-by: ascend-robot Description: msprobe增加-V命令,并按照以下格式展示  See merge request: Ascend/msprobe!964 | 1 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 | ||
| 4 个月前 | ||
| 1 个月前 | ||
| 4 个月前 | ||
| 1 个月前 | ||
| 1 个月前 | ||
| 1 个月前 | ||
| 1 个月前 | ||
| 4 个月前 | ||
| 1 个月前 | ||
| 3 个月前 | ||
| 1 个月前 | ||
| 1 个月前 |