[CI/Build] Upgrade CANN to 8.2.RC1

[MengqingCao]

What this PR does / why we need it?

Upgrade CANN to 8.2.rc1

Backport: #1653

Does this PR introduce any user-facing change?

Yes, docker image will use 8.2.RC1

How was this patch tested?

CI passed

• vLLM version: v0.10.0
• vLLM main: vllm-project/vllm@7728dd7

[codecov]

Codecov Report

✅ All modified and coverable lines are covered by tests.
✅ Project coverage is 73.11%. Comparing base (17a430f) to head (5cf9058).
⚠️ Report is 597 commits behind head on main.

Additional details and impacted files

@@            Coverage Diff             @@
##             main    #1653      +/-   ##
==========================================
- Coverage   73.16%   73.11%   -0.06%     
==========================================
  Files          90       90              
  Lines        9929     9929              
==========================================
- Hits         7265     7260       -5     
- Misses       2664     2669       +5     

Flag	Coverage Δ
unittests	73.11% <ø> (-0.06%)	⬇️

Flags with carried forward coverage won't be shown. Click here to find out more.

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.

🚀 New features to boost your workflow:

• ❄️ Test Analytics: Detect flaky tests, report on failures, and find test suite problems.

[github-actions]

This pull request has conflicts, please resolve those before we can evaluate the pull request.

[Yikun]

Consdier the CANN alpha3 release bring many oom and internal error, we will not ugprade this

[1] https://github.com/vllm-project/vllm-ascend/actions/runs/16133440759/job/45525106786?pr=1653

RuntimeError: replay:build/CMakeFiles/torch_npu.dir/compiler_depend.ts:201 NPU function error: c10_npu::acl::AclmdlRIExecuteAsync(model_ri_, c10_npu::getCurrentNPUStream()), error code is 507000

[2] OOM as unexpected:
https://github.com/vllm-project/vllm-ascend/actions/runs/16133440759/job/45525106780?pr=1653

RuntimeError: NPU out of memory. Tried to allocate 98.00 MiB (NPU 0; 29.50 GiB total capacity; 1010.15 MiB already allocated; 1010.15 MiB current active; 39.48 MiB free; 1.02 GiB reserved in total by PyTorch) If reserved memory is >> allocated memory try setting max_split_size_mb to avoid fragmentation.

[Yikun]

https://github.com/vllm-project/vllm-ascend/blob/main/.github%2FDockerfile.buildwheel

should also updated, hold on this until ci infra upgrade.

[Yikun]

It seems manylinux image not ready yet, let's first upgrade cann image first