合并受阻
变更摘要
本 PR 修复 reliability/crypto/docs/aclnnCrypto.md 文档中与 issue #58 对应的三处问题，涉及接口原型名不一致、示例中 IV 缓冲区形状与宿主数据长度不匹配导致的越界读取，以及示例 CreateAclTensor 缺少长度校验。改动将该文档的"函数原型"代码块对齐到 reliability/crypto/op_host/op_api/aclnn_crypto.h 的实际声明，将示例 ivHostData 对齐到 32 字节以消除越界，并在内存分配/拷贝前补充 hostData 字节数与 shape 字节数的比较。
主要改动
- 函数原型对齐头文件声明: 文档"函数原型"代码块由
aclnnCryptoAicpuGetWorkspaceSize/aclnnCryptoAicpu改为头文件实际声明的aclnnCryptoGetWorkspaceSize/aclnnCrypto，与正文描述及调用示例一致。 - 修复示例 IV 缓冲区越界读取: 将示例中
ivHostData由 16 字节对齐到 32 字节，使其与iv_shape_data = {32}匹配，消除aclrtMemcpy按 shape 计算 32 字节而从宿主向量复制 32 字节造成的越界读取和未定义 IV 数据。 - 为示例
CreateAclTensor增加长度校验: 在aclrtMalloc/aclrtMemcpy之前增加 hostData 字节数与 shape 字节数的比较，使 shape 与 hostData 不匹配时快速失败，而非产生越界 memcpy。
🤖 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。
Hi @NUAA_wqy, welcome to submitting your first PR to ops-ras!
PR Merge Steps
1. CLA Signing
If the current PR label includes cann-cla/yes, it means you have signed the CLA and can proceed to the next step. If the label includes cann-cla/no, please sign the CLA first. If you have any questions, please refer to the FAQ.
2. CI Check
Please comment /compile to trigger the CI pipeline check. If the CI run is successful, the PR will be tagged with ci-pipeline-passed and you can proceed to the next step. If the CI run fails, the PR will be tagged with ci-pipeline-failed, please check the CI logs to fix the issues in the PR. If you have any questions, please refer to the FAQ.
3. Code Review
After CI passes, please refer to the PR Approval Progress and proactively @ the committers in the table to review the code. After approval, committers will comment /lgtm and /approve. Once the lgtm and approved labels are successfully added, the PR will be merged automatically.
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
⚠️ This PR does not yet meet the following requirements:lgtm (requires ≥ 2 person(s) per module)、approve (requires ≥ 1 person(s) per module)
Module Approval Details
| module | lgtm status | approve status |
|---|---|---|
| repo-cann/ops-ras | ❌ (0/2)(You can also ask: 陈兴宇, 柳宗谷, 冯彤, yue-ma, 周奇龙) | ❌ (0/1)(You can also ask: 陈兴宇, 唐燕峰, 於欣洁, 冯彤, yue-ma) |
💡 Tip:
- Committer can comment
/approveor/lgtm- Commenting
/approveimplies both code review (lgtm) and intent to merge (approve)
描述
reliability/crypto/docs/aclnnCrypto.md存在三处与 issue #58 对应的问题，本 MR 一并修复：aclnnCryptoAicpuGetWorkspaceSize/aclnnCryptoAicpu，而reliability/crypto/op_host/op_api/aclnn_crypto.h实际声明为aclnnCryptoGetWorkspaceSize/aclnnCrypto（无Aicpu后缀）。读者按文档原型链接会失败。本 MR 将文档原型对齐到头文件声明，与正文描述（第 35 行）及调用示例（aclnnCryptoGetWorkspaceSize/aclnnCrypto）保持一致。iv_shape_data = {32}但ivHostData仅 16 字节，CreateAclTensor按 shape 计算 32 字节并在aclrtMemcpy中从 16 字节 host 向量复制 32 字节，越界读取 16 字节未定义数据，设备侧 IV 含未定义数据，加解密结果不可信。本 MR 将ivHostData对齐到 32 字节，从根因消除越界。aclrtMalloc/aclrtMemcpy前增加 hostData 字节数与 shape 字节数的比较，使未来再次出现 shape/hostData 不匹配时快速失败，而非产生越界 memcpy。变更类型
关联的Issue
Fixes #58
测试
本地静态验证（无 NPU 环境下的可重复验证）：
reliability/crypto/op_host/op_api/aclnn_crypto.h第 38/52 行声明，确认文档原型代码块现已与头文件一致（aclnnCryptoGetWorkspaceSize/aclnnCrypto）。main中所有CreateAclTensor调用点的hostData.size()*sizeof(T)与GetShapeSize(shape)*sizeof(T)：keyHostData(16)↔key_shape_data={16}✓inputHostData(msg_len)↔input_shape_data={msg_len}✓ivHostData(32)↔iv_shape_data={32}✓（本次修复后匹配）opConfigHostData(6 × uint32)↔shape_op_cfg={6}✓outputHostData(msg_len)↔shape_output={msg_len}✓yHostData(1 × uint32)↔shape_y={1}✓tagHostData(16)↔shape_tag={GCM_TAG_SIZE=16}✓全部调用点满足新增的长度校验，不会被误拦截。
git diff确认改动仅限aclnnCrypto.md一个文件，8 行新增 / 3 行删除，无无关改动。补充的UT用例：
NA（文档示例修复，无新增 UT）
文档更新
本次即为文档修复。
合入检查