Pull Request

变更摘要

该 PR 对 compiler/graph/build/memory/block_mem_assigner.cc 中 BlockMemAssigner::GetOutAndWorkSpaceMem 函数做性能优化：通过将 batch_all_memory_size[batch_label] 的查找结果缓存为局部引用 batch_mem，消除函数内两处重复的 operator[] 容器查找；同时将遍历输出锚点指针的循环变量由按值拷贝改为常量引用，规避 shared_ptr 按值拷贝带来的原子引用计数开销。该改动为纯重构，不改变既有功能逻辑。

主要改动

• 缓存 batch_mem 引用: 在函数入口处新增 auto &batch_mem = batch_all_memory_size[batch_label];，将后续两处 batch_all_memory_size[batch_label] 的重复 operator[] 查找替换为对 batch_mem 的直接使用（batch_mem.emplace_back(size) 与 batch_mem.insert(batch_mem.cend(), temp.cbegin(), temp.cend())），消除重复哈希查找开销。
• 遍历改为常量引用: 将 for (auto out_anchor : n->GetAllOutDataAnchorsPtr()) 改为 for (const auto &out_anchor : ...)，避免循环中对 shared_ptr 按值拷贝产生的原子引用计数增减开销。

🤖 本次改动均为数据 / 生成 / 二进制文件（如测试用例、锁文件、媒体资源），没有可审查的代码逻辑，已跳过 AI 代码审查。

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.
For more, you also can visit HICANN.

PR Approval Progress

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

Module Approval Details

💡 Tip:

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

CLA Signature Pass

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

/lgtm

/approve