bugfix AS block leaks

[zccjjj]

Motivation

💡 If this PR is a Cherry Pick, the PR title needs to follow the format by adding the [Cherry-Pick] label at the very beginning and appending the original PR ID at the end. For example, [Cherry-Pick][CI] Add check trigger and logic(#5191)

💡 如若此PR是Cherry Pick，PR标题需遵循格式，在最开始加上[Cherry-Pick]标签，以及最后面加上原PR ID，例如[Cherry-Pick][CI] Add check trigger and logic(#5191)

Modifications

Usage or Command

Accuracy Tests

Checklist

• Add at least a tag in the PR title.
  • Tag list: [[FDConfig],[APIServer],[Engine], [Scheduler], [PD Disaggregation], [Executor], [Graph Optimization], [Speculative Decoding], [RL], [Models], [Quantization], [Loader], [OP], [KVCache], [DataProcessor], [BugFix], [Docs], [CI], [Optimization], [Feature], [Benchmark], [Others], [XPU], [HPU], [GCU], [DCU], [Iluvatar], [Metax]]
  • You can add new tags based on the PR content, but the semantics must be clear.
• Format your code, run pre-commit before commit.
• Add unit tests. Please write the reason in this PR if no unit tests.
• Provide accuracy results.
• If the current PR is submitting to the release branch, make sure the PR has been submitted to the develop branch, then cherry-pick it to the release branch with the [Cherry-Pick] PR tag.

[paddle-bot]

Thanks for your contribution!

[PaddlePaddle-bot]

🤖 Paddle-CI-Agent | pr_review | 2026-05-22 11:16:02

📋 Review 摘要

PR 概述：修复在 enable_prefix_caching + 分层 KV Cache 路径下，can_allocate_gpu_blocks 未通过 _get_can_schedule_prefill_threshold_block 计算预留块阈值，导致 AS block 泄漏的问题。
变更范围：fastdeploy/engine/sched/resource_manager_v1.py（调度资源管理）
影响面 Tag：[Scheduler] [KVCache]

问题

级别	文件	概述
📝 PR 规范	—	标题缺少官方 Tag；描述模板各段均为空
❓ 疑问	resource_manager_v1.py:1205	新增注释拼写错误：Warnig → Warning

📝 PR 规范检查

PR 标题 bugfix AS block leaks 使用了非官方的小写 bugfix 形式，需规范化为 [BugFix]；描述模板中 Motivation、Modifications 等段落均为占位符/空白。

标题建议（可直接复制）：

• [BugFix] Fix AS block leaks when enable_prefix_caching with hierarchical kvcache

PR 描述建议（可直接复制，必须复刻 checklist §D2 模板的完整结构）：

## Motivation
在开启 `enable_prefix_caching` 且配置了分层 KV Cache（`num_cpu_blocks > 0` 或 `kvcache_storage_backend`）时，调度器在准入新 Prefill 请求前的 GPU Block 可用性检查直接使用了原始 block 数（`need_prefill_tokens / block_size`），而未调用 `_get_can_schedule_prefill_threshold_block` 计算包含 running 请求预留块的阈值。这导致在阈值偏低时分层 Cache 会为请求分配 block 进行前缀匹配，但后续实际分配时 GPU block 不足，引发 AS block 泄漏（storage blocks leak）。

## Modifications
- `fastdeploy/engine/sched/resource_manager_v1.py`：在两处 `enable_prefix_caching + 分层 KV Cache` 路径的 `can_allocate_gpu_blocks` 检查中，改为先调用 `_get_can_schedule_prefill_threshold_block` 计算含预留块的阈值，再传入 `can_allocate_gpu_blocks`，与文件中其余调用点保持一致。同时在 `_free_blocks` 前添加注释说明潜在的 storage block leak 风险。

## Usage or Command
N/A

## Accuracy Tests
N/A

## Checklist

- [x] Add at least a tag in the PR title.
  - Tag list: [`[FDConfig]`,`[APIServer]`,`[Engine]`, `[Scheduler]`, `[PD Disaggregation]`, `[Executor]`, `[Graph Optimization]`, `[Speculative Decoding]`, `[RL]`, `[Models]`, `[Quantization]`, `[Loader]`, `[OP]`, `[KVCache]`, `[DataProcessor]`, `[BugFix]`, `[Docs]`, `[CI]`, `[Optimization]`, `[Feature]`, `[Benchmark]`, `[Others]`, `[XPU]`, `[HPU]`, `[GCU]`, `[DCU]`, `[Iluvatar]`, `[Metax]`]
  - You can add new tags based on the PR content, but the semantics must be clear.
- [ ] Format your code, run `pre-commit` before commit.
- [ ] Add unit tests. Please write the reason in this PR if no unit tests.
- [ ] Provide accuracy results.
- [ ] If the current PR is submitting to the `release` branch, make sure the PR has been submitted to the `develop` branch, then cherry-pick it to the `release` branch with the `[Cherry-Pick]` PR tag.

总体评价

修复方案逻辑正确，将两处遗漏的 can_allocate_gpu_blocks 调用统一改为通过 _get_can_schedule_prefill_threshold_block 计算阈值，与文件内其他调用点保持一致，能有效防止分层 Cache 场景下的死锁和 block 泄漏。建议补充 PR 描述和标题格式，并修正注释中的拼写错误 Warnig → Warning。

[PaddlePaddle-bot]

🤖 Paddle-CI-Agent | ci_status_monitor | 2026-05-22 15:35:45

CI报告基于以下代码生成（30分钟更新一次）:

• PR commit: 84fd8c2
• Merge base: 3ec5011 (branch: develop)
• 查看完整 Diff
• CI 详情

---

1 任务总览

✅ 所有 Required 任务均已通过，当前 CI 不阻塞合并；仍有 3 个 Optional 任务失败、1 个 Optional 任务等待中，仅供参考。

总执行（rerun次数）	总任务	✅ 通过	❌ 失败	⏳ 运行中	⏸️ 等待中	跳过
54(13)	41	37	3	0	1	0

2 任务状态汇总

2.1 Required任务 : 10/10 通过

必选任务阻塞合并，失败需优先处理。

状态	任务	耗时	根因	修复建议	日志	重跑
✅	其余 10 个必选任务通过	-	-	-	-	-

2.2 可选任务 — 27/31 通过

可选任务不阻塞合并，失败/等待仅供参考。

状态	任务	耗时	日志	重跑
❌	Run iluvatar Tests / run_iluvatar_cases	15m51s	Job	-
❌	Check PR Template	15s	Job	-
❌	Trigger Jenkins for PR	7m22s	Job	-
⏸️	CI_HPU	-	-	-
✅	其余 27 个可选任务通过	-	-	-

---

3 失败详情（仅 required）

无 required 失败任务。

[codecov-commenter]

Codecov Report

❌ Patch coverage is 50.00000% with 2 lines in your changes missing coverage. Please review.
⚠️ Please upload report for BASE (develop@3ec5011). Learn more about missing BASE report.

Files with missing lines	Patch %	Lines
fastdeploy/engine/sched/resource_manager_v1.py	50.00%	0 Missing and 2 partials ⚠️

Additional details and impacted files

@@            Coverage Diff             @@
##             develop    #7890   +/-   ##
==========================================
  Coverage           ?   63.98%           
==========================================
  Files              ?      462           
  Lines              ?    64482           
  Branches           ?     9880           
==========================================
  Hits               ?    41256           
  Misses             ?    20450           
  Partials           ?     2776           

Flag	Coverage Δ
GPU	73.04% <50.00%> (?)
XPU	16.01% <0.00%> (?)

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.