已合并
fix: adapt Inductor compile mode for TA3.6 #45315
fix: adapt Inductor compile mode for TA3.6 #45315
已合并
luqichao创建于 20 天前
luqichao
luqichao成员
20 天前

关联 Issue:Adaptation of compilation options for TA3.6

  1. NPUKernelType.SIMT_TEMPLATE 对应的 compile_modeunstructured_in_simt 修改为 simd_simt_template,适配 TA3.6 的编译模式定义。
  2. 在 Inductor 的 TTIR 构建和 Triton 预编译路径中,为缺失或值为 Nonecompile_mode 统一使用 simd_simt_template,确保调用 triton.compile 时始终传入有效的编译模式。
likedislike
Pull Request已成功合入, 合并人@ascend-robot
(感谢 luqichao 的贡献)
luqichaoluqichao成员
20 天前 创建了 pull request,commit 64387d40
atomgit-bot
atomgit-bot
20 天前 评论:

变更摘要

本 PR 旨在适配 TA3.6 的 Inductor 编译模式:将 NPUKernelType.SIMT_TEMPLATE.compile_mode() 的返回值由 unstructured_in_simt 改为 simd_simt_template,并在 NPUCachingAutotuner 的两处 Inductor Triton 预编译路径中为该编译模式增加回退逻辑,避免配置值缺失或为 None 时直接传入 triton.compile

主要改动

  • 调整 compile_mode() 返回值:在 torch_npu/_inductor/codegen/triton_utils.py 中,NPUKernelType.SIMT_TEMPLATE.compile_mode() 由返回 unstructured_in_simt 改为返回 simd_simt_template
  • 预编译路径增加回退逻辑(一):在 torch_npu/_inductor/runtime/triton_heuristics.pyNPUCachingAutotuner 第一处 compile_meta['compile_mode'] 赋值处,将 cfg_kwargs.get('compile_mode') 改为 cfg_kwargs.get('compile_mode') or NPUKernelType.SIMT_TEMPLATE.compile_mode(),配置缺失或为 None 时回退到 simd_simt_template
  • 预编译路径增加回退逻辑(二):在同一文件的第二处 compile_meta['compile_mode'] 赋值处应用相同的回退写法,确保两条 Inductor Triton 预编译路径行为一致,None 值不会进入 triton.compile
likedislike
不准确?
atomgit-bot
atomgit-bot
20 天前 评论:

代码审查

✅ 未发现问题

likedislike
不准确?
luqichaoluqichao成员
20 天前 关联了issue:Adaptation of compilation options for TA3.6
ascend-robotascend-robot成员
20 天前 添加了label:ascend-cla/yes
ascend-robot
ascend-robot成员
20 天前 评论:

Thanks for your pull-request.
The full list of commands accepted by me can be found at here.
You can get sig-info at here.
You can self-configure the PR merge rules for this repository. For more details, please refer to Here.


PR Approval Progress

Congratulations! All modules have met the lgtm and approve requirements.

Module Approval Details

module lgtm status approve status
torch_npu/_inductor weizhan4, HinPeng (2/2) weizhan4 (1/1)

💡 Tip:

  • Committer can comment /approve or /lgtm
  • Commenting /approve implies both code review (lgtm) and intent to merge (approve)

CLA Signature Pass

luqichao, thanks for your pull request. All authors of the commits have signed the CLA. 👍

likedislike
ascend-robot
ascend-robot成员
20 天前 评论:

当前仓库存在以下 保护分支

Protected Branch Version Release
master
v2.9.0
v2.11.0
v2.10.0
v2.12.0
v2.7.1
v2.11.0-26.1.0
v2.12.0-26.1.0
v2.10.0-26.1.0
v2.9.0-26.1.0
v2.7.1-26.1.0
ci-test

评论 /sync <branch1> <branch2> ... 可将当前 PR 修改同步到其它分支(创建同步 PR):
a) 如果当前 PR 是 Open 状态,同步操作将延迟到 PR 被合并时执行
b) 如果当前 PR 已经 Merged,将立即执行同步操作

注意:

  1. /sync 命令可以指定同步到多个分支,仅最后一个 /sync 命令生效
  2. 如果创建的同步 PR 不正确,可通过向同步 PR 的源分支提交轻量级 PR 完善,或使用 /close 命令关闭
likedislike
ascend-robotascend-robot成员
20 天前 添加了label:ci-pipeline-running
ascend-robot
ascend-robot成员
20 天前 评论:

ascend docs pipeline is running...

likedislike
ascend-robotascend-robot成员
20 天前 添加了label:docs-ci-pipeline-running
ascend-robot
ascend-robot成员
20 天前 评论:

✅ 跳过 docs ci 检查,没有需要检查的文档文件

likedislike
ascend-robotascend-robot成员
20 天前 删除了label:docs-ci-pipeline-running
ascend-robotascend-robot成员
20 天前 添加了label:docs-ci-pipeline-success
ascend-robotascend-robot成员
20 天前 删除了label:ci-pipeline-running
ascend-robotascend-robot成员
20 天前 添加了label:ci-pipeline-passed
ascend-robot
ascend-robot成员
20 天前 评论:
流水线 PR-pipeline_pytorch#63783 [ commitID:401a9c71 ] 已完成
>>>代码风格自动修复执行成功(无修复内容)
阶段 任务名 状态 详情
编译构建 Build_X86 >>>
Build_ARM >>>
Build_X86_torchair 🛑 >>>
Build_ARM_torchair 🛑 >>>
patch_test 🛑 >>>
Build_X86_213 🛑 >>>
Build_ARM_213 🛑 >>>
恶意代码检查 Antipoison >>>
编码安全与规范检查 codecheck_pre-commit >>>
check_error >>>
lintrunner >>>
开源片段检查 SCA >>>
开发者测试 UT_ARM_A3_Part_01 🛑 >>>
UT_ARM_A3_Part_02 🛑 >>>
UT_ARM_A2_Part_01 >>>
UT_ARM_A2_Part_02 >>>
UT_ARM_A2_Part_03 >>>
UT_inductor_Part_01 >>>
UT_inductor_Part_02 >>>
UT_inductor_Part_03 >>>
UT_inductor_Part_04 >>>
UT_DIST_ARM_Part_01 🛑 >>>
UT_DIST_ARM_Part_02 🛑 >>>
UT_DIST_ARM_Part_03 🛑 >>>
UT_DIST_ARM_Part_04 🛑 >>>
UT_ARM_A2_Select_Part_01 >>>
UT_ARM_A2_Select_Part_02 >>>
UT_ARM_A2_Part_01_213 🛑 >>>
UT_ARM_A2_Part_02_213 🛑 >>>
UT_ARM_A2_Part_03_213 🛑 >>>
UT_inductor_Part_01_213 🛑 >>>
UT_inductor_Part_02_213 🛑 >>>
UT_inductor_Part_03_213 🛑 >>>
UT_inductor_Part_04_213 🛑 >>>
UT_DIST_ARM_Part_01_213 🛑 >>>
UT_DIST_ARM_Part_02_213 🛑 >>>
UT_DIST_ARM_Part_03_213 🛑 >>>
UT_DIST_ARM_Part_04_213 🛑 >>>
UT_ARM_A2_Select_Part_01_213 🛑 >>>
UT_ARM_A2_Select_Part_02_213 🛑 >>>
流水线 PR-pipeline_pytorch >>>
此流水线已支持下列评论快捷指令,仅PR创建者和白名单成员[wujinyuan1, huangjingwei, liangsongwei, yashi999, culechan, Dring, wuyouqi1, L1919_snow, qq_52711437, WhiteNight12, nomiz, xiu_21, ffmh, wanglijun55, hss-shuai, husichao, smallsilly, lanshaozuishuai, jimmyisme1, lzy0920232, alpha-junh, Sunshine_Youngster, wei_zhuoyi, zhangyihuiben, zyw-hw, zzzkeke, rmch, yangch0324, LucciC, AACAES, renyujin, wjlflyer, senzhen-town, pengjingyou, qsc97, limuan, yule100, xiaoqi-zhou, kuhn7, chenxingying, hanye02, zichun_ye, anyrenwei, kkjocker, wangzili121, Lu_G, yvjc, puddingfjz, HandsoemLemon, bigprestigee1, huawuyi, zhenyu10, dairenjie, du-jin-hang, zou-jieyu, adelaideliu, TrHan, wanlinan, Windwindzzz, pengqihw, kisnwang, yuheng_wang, honghao_wang, jizewei, zhangguoguang, sunyu-xuan, chenrayray, hbhu_bin, liujunzhu, c_34, LiNuoh, maoyuanpeng1, zzhongmin, zhaoyu65, bellatan, jiabaolin, zhuofanshen, wencaiwen, lu_zhuge, caoshuyang, molly12, lyx324521, LQ1206, gitcode-chenjiao, cai-weiwei1989, CHDong, ogqin, yuanlipingGit, xuqinglin1, lqz2, zouwei1, chaoluoa, paradox325, jackzhang1116, yaoyao, akh, yujiacheng, dengjie0116, Hubert11111, Shine_Ws, wslhj555, longqiand, OYtao666, JiaqingQiang, luyyyy, Kingbelial, zhanghaiyu0101, wenxp1018, yanliu-luoluo, ksun_sekiro, liyong328, wgzheng, tangky, vivi_is_coding, aoiaoisola, weixin_44494597, wangmengmengwang65667, hid57809721, qq_35468730, comeonup, C547032, gcw_m5OQChA4, yao_yao_ling_xian, cnnbwcy, szqfes_12, cora_19, cann_lilin, can, shawnylee233, fanglanyue0916, hhz0, LiNuohang, taohuoquan, Jesse, WSs_321]评论有效
  • compile、compile_inductor、compile_torchair : 运行流水线
  • retry : 重试流水线所有失败子任务
  • retry <任务名> : 仅重试指定失败子任务
  • stop : 停止流水线
likedislike
luqichaoluqichao成员
20 天前 修改了pull request 的描述
luqichaoluqichao成员
20 天前 修改了pull request 的描述
weizhan4成员
20 天前 评论:

/lgtm
/approve

likedislike
ascend-robotascend-robot成员
20 天前 添加了label:approved
Xuan Peng
20 天前 评论:

/lgtm

likedislike
ascend-robotascend-robot成员
20 天前 添加了label:lgtm
ascend-robotascend-robot成员
20 天前 合入了pull request
ascend-robot
ascend-robot成员
20 天前 评论:
流水线 pytorch_gitcode_PR_multiVersion#14505 [ commitID:401a9c71 ] 已完成
likedislike