diff --git a/swe-paddle/tasks/PaddlePaddle__Paddle-40111/README.md b/swe-paddle/tasks/PaddlePaddle__Paddle-40111/README.md
new file mode 100644
index 000000000..7804ac238
--- /dev/null
+++ b/swe-paddle/tasks/PaddlePaddle__Paddle-40111/README.md
@@ -0,0 +1,44 @@
+# PaddlePaddle__Paddle-40111
+
+This directory converts Paddle PR #40111 into a SWE-Paddle community task candidate.
+
+## Source
+
+| Field | Value |
+| --- | --- |
+| Repo | `PaddlePaddle/Paddle` |
+| PR | [#40111](https://github.com/PaddlePaddle/Paddle/pull/40111) |
+| PR title | `add profiler statistic helper` |
+| Base commit | `10325a82e1032c3397b6f6611f558eb18ede0b07` |
+| Merged at | `2022-03-08T01:55:55Z` |
+| Task type | `feature_enhancement` |
+| Resource | CPU |
+
+## Summary
+
+Add deterministic interval operations for profiler statistics, including duration summation, normalization, union, intersection, and subtraction.
+
+## Why This Is A Good SWE-Paddle Candidate
+
+- The 225-line Python production change contains substantial interval-processing logic with meaningful edge cases.
+- The source PR includes a focused 137-line unit test file, preserved byte-for-byte in the task patch.
+- Tests cover sorted and unsorted input, overlap, containment, adjacency, disjoint ranges, empty ranges, and zero-length ranges.
+- Verification is deterministic and CPU-only, with no operator execution, graph mode, model loading, network, or external service.
+
+## Files
+
+- `proposal.md`: candidate proposal for maintainer triage.
+- `instruction.md`: self-contained problem statement for the coding agent.
+- `solution/code.patch`: gold patch from the merged PR.
+- `tests/test.patch`: exact upstream tests plus checkout-loading support and separate P2P coverage.
+- `tests/test.sh`: minimal target test command.
+- `environment/README.md`: environment notes for reproduction.
+- `README.md`: task overview and verification entrypoint.
+
+## Verification
+
+```bash
+bash tests/test.sh
+```
+
+Expected behavior: applying `tests/test.patch` to `base_commit` should keep the existing profiler scheduler case passing while the interval operations fail; applying both `tests/test.patch` and `solution/code.patch` should pass all target tests.
diff --git a/swe-paddle/tasks/PaddlePaddle__Paddle-40111/environment/README.md b/swe-paddle/tasks/PaddlePaddle__Paddle-40111/environment/README.md
new file mode 100644
index 000000000..7b28432a7
--- /dev/null
+++ b/swe-paddle/tasks/PaddlePaddle__Paddle-40111/environment/README.md
@@ -0,0 +1,27 @@
+# Environment Notes
+
+This candidate is part of the SWE-Paddle community task set.
+
+## Expected Environment
+
+- Repository: `PaddlePaddle/Paddle`
+- Base commit: `10325a82e1032c3397b6f6611f558eb18ede0b07`
+- Resource: CPU
+- GPU required: no
+- Build path: no source build is required; the exact upstream test file loads the checkout Python module directly.
+
+## Run Order
+
+1. Check out `PaddlePaddle/Paddle` at the base commit.
+2. Apply `tests/test.patch`.
+3. Run `bash tests/test.sh`; the target behavior should fail before the fix.
+4. Apply `solution/code.patch`.
+5. Run `bash tests/test.sh` again; the target behavior should pass after the gold patch.
+
+## Minimal Test Command
+
+```bash
+bash tests/test.sh
+```
+
+The verifier is responsible for deriving stable F2P and P2P node IDs from repeated runs.
diff --git a/swe-paddle/tasks/PaddlePaddle__Paddle-40111/instruction.md b/swe-paddle/tasks/PaddlePaddle__Paddle-40111/instruction.md
new file mode 100644
index 000000000..528620671
--- /dev/null
+++ b/swe-paddle/tasks/PaddlePaddle__Paddle-40111/instruction.md
@@ -0,0 +1,19 @@
+# 补齐 profiler 的时间区间统计能力
+
+## 详细描述
+
+Profiler 收集到的事件通常由一组开始时间和结束时间组成。同一批事件里可能存在重叠、包含、首尾相接或顺序打乱的区间。生成统计结果时，如果直接把每段时间相加，容易重复计算；在计算两类事件同时发生的时间或排除某类事件的耗时时，也容易遗漏边界情况。
+
+需要为 profiler 提供一套统一的时间区间处理能力，能够整理单组区间，并计算两组区间合并、重叠和相减后的结果。调用方传入的区间可能已经排序，也可能是无序的，空列表和零长度区间也应得到稳定结果。
+
+## 验收说明
+
+- 能正确汇总区间长度，并将重叠或首尾相接的区间整理为有序且不重复的结果。
+- 两组区间的合并、交集和相减在部分重叠、完全包含、互不相交等情况下均返回正确结果。
+- 已排序、未排序、空列表和零长度区间都能正常处理，现有 profiler scheduler 行为保持不变。
+
+## 技术要求
+
+- 熟悉 Python 列表与区间处理。
+- 熟悉 profiler 事件时间统计的基本含义。
+- 能处理排序、边界和空输入等情况。
diff --git a/swe-paddle/tasks/PaddlePaddle__Paddle-40111/proposal.md b/swe-paddle/tasks/PaddlePaddle__Paddle-40111/proposal.md
new file mode 100644
index 000000000..971cdb6de
--- /dev/null
+++ b/swe-paddle/tasks/PaddlePaddle__Paddle-40111/proposal.md
@@ -0,0 +1,55 @@
+# Task Proposal: PaddlePaddle__Paddle-40111
+
+## 1. 来源信息
+
+- Instance ID：`PaddlePaddle__Paddle-40111`
+- PR 链接：https://github.com/PaddlePaddle/Paddle/pull/40111
+- PR 标题：`add profiler statistic helper`
+- `base_commit`：`10325a82e1032c3397b6f6611f558eb18ede0b07`
+- merged 时间：`2022-03-08T01:55:55Z`
+- 你的身份：熟悉该模块的 contributor
+- 后续联系人：TBD
+
+## 2. 问题一句话
+
+为 profiler 增加可靠的时间区间归并、交集和差集计算，避免统计重叠事件时重复计时或漏计。
+
+## 3. 为什么适合作为 SWE-Paddle 样本
+
+- **真实性**：来源 PR 为 profiler 性能统计补充了实际需要的时间区间运算能力。
+- **代表性**：覆盖排序、重叠、包含、相邻、空输入以及两个区间集合之间的组合运算。
+- **边界清楚**：production change 仅新增一个 225 行 Python helper，不涉及算子、模型加载或图执行。
+- **非平凡性**：多个双指针分支需要保持有序输出并正确处理边界，无法靠单点特判完成。
+- **环境友好性**：来源 PR 的 137 行单测保持 exact blob；测试直接加载 checkout 文件，可在纯 CPU 环境运行。
+
+## 4. 任务类型和标签
+
+- 任务类型：`feature_enhancement`
+- 执行后端：`cpu`
+- 设备范围：`cpu_only`
+- 模块标签：`[profiler, statistics, intervals, python]`
+
+## 5. 验证思路
+
+- 目标测试命令：`bash tests/test.sh`
+- 目标测试文件：`test/swe_paddle/test_pr40111_profiler_ranges.py`
+- 修复前预期：现有 profiler scheduler 的 P2P 通过；来源 PR 的 merge、intersection 与 subtract 用例因区间工具缺失而失败。
+- 修复后预期：P2P 和来源 PR 的全部 10 个区间测试通过；`tests/test.sh` 通过 adapter 运行 exact upstream test file。
+- P2P 候选：现有 profiler scheduler 在 closed、ready、record 和 record-and-return 状态之间的切换顺序保持不变。
+
+## 6. 环境与资源
+
+- 资源需求：CPU
+- Paddle 来源：`PaddlePaddle/Paddle` source checkout at `base_commit`
+- 是否能提供 Docker：暂无
+- patch 类型：Python-only
+- 环境建议：无需 source build 或 Paddle wheel；adapter 直接执行 checkout 中的 Python 文件和 exact upstream tests。
+- 最小测试命令：`bash tests/test.sh`
+- 是否有 oracle 日志：由 SWE-Paddle verifier 结果另行维护
+
+## 7. 风险自查
+
+- 泄露风险：instruction 只说明所需区间行为，不描述 Gold patch 的循环结构或局部变量。
+- 环境风险：测试不访问网络，不加载模型，也不依赖 GPU、算子或动态图/静态图。
+- flaky 风险：所有输入均为确定性的整数区间，没有计时、并发或随机因素。
+- 拆分风险：来源 PR 只新增一个 production helper 和对应单测，专注于同一统计问题。
diff --git a/swe-paddle/tasks/PaddlePaddle__Paddle-40111/solution/code.patch b/swe-paddle/tasks/PaddlePaddle__Paddle-40111/solution/code.patch
new file mode 100644
index 000000000..aefe229cf
--- /dev/null
+++ b/swe-paddle/tasks/PaddlePaddle__Paddle-40111/solution/code.patch
@@ -0,0 +1,231 @@
+diff --git a/python/paddle/profiler/statistic_helper.py b/python/paddle/profiler/statistic_helper.py
+new file mode 100644
+index 0000000000000000000000000000000000000000..1f11649928a7fe72b0a4ae6f368c3c92debf060c
+--- /dev/null
++++ b/python/paddle/profiler/statistic_helper.py
+@@ -0,0 +1,225 @@
++# Copyright (c) 2022 PaddlePaddle Authors. All Rights Reserved.
++# 
++# Licensed under the Apache License, Version 2.0 (the "License");
++# you may not use this file except in compliance with the License.
++# You may obtain a copy of the License at
++# 
++#     http://www.apache.org/licenses/LICENSE-2.0
++# 
++# Unless required by applicable law or agreed to in writing, software
++# distributed under the License is distributed on an "AS IS" BASIS,
++# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
++# See the License for the specific language governing permissions and
++# limitations under the License.
++import collections
++
++
++def sum_ranges(ranges):
++    result = 0
++    for time_range in ranges:
++        result += (time_range[1] - time_range[0])
++    return result
++
++
++def merge_self_ranges(src_ranges, is_sorted=False):
++    merged_ranges = []
++    if len(src_ranges) > 0:
++        if not is_sorted:
++            src_ranges.sort(key=lambda x: x[0])
++        cur_indx = 0
++        merged_ranges.append((src_ranges[cur_indx][0], src_ranges[cur_indx][1]))
++        for cur_indx in range(1, len(src_ranges)):
++            if src_ranges[cur_indx][1] > merged_ranges[-1][1]:
++                if src_ranges[cur_indx][0] <= merged_ranges[-1][1]:
++                    merged_ranges[-1] = (merged_ranges[-1][0],
++                                         src_ranges[cur_indx][1])
++                else:
++                    merged_ranges.append(
++                        (src_ranges[cur_indx][0], src_ranges[cur_indx][1]))
++    return merged_ranges
++
++
++def merge_ranges(range_list1, range_list2, is_sorted=False):
++    merged_ranges = []
++    if not is_sorted:
++        range_list1 = merge_self_ranges(range_list1)
++        range_list2 = merge_self_ranges(range_list2)
++    len1 = len(range_list1)
++    len2 = len(range_list2)
++    if len1 == 0 and len2 == 0:
++        return merged_ranges
++    elif len1 == 0:
++        return range_list2
++    elif len2 == 0:
++        return range_list1
++    else:
++        indx1 = 0
++        indx2 = 0
++        range1 = range_list1[indx1]
++        range2 = range_list2[indx2]
++        if range1[0] < range2[0]:
++            merged_ranges.append(range1)
++            indx1 += 1
++        else:
++            merged_ranges.append(range2)
++            indx2 += 1
++        while indx1 < len1 and indx2 < len2:
++            range1 = range_list1[indx1]
++            range2 = range_list2[indx2]
++            if range1[0] < range2[0]:
++                if range1[1] > merged_ranges[-1][1]:
++                    if range1[0] <= merged_ranges[-1][1]:
++                        merged_ranges[-1] = (merged_ranges[-1][0], range1[1])
++                    else:
++                        merged_ranges.append((range1[0], range1[1]))
++                    indx1 += 1
++                else:
++                    indx1 += 1
++            else:
++                if range2[1] > merged_ranges[-1][1]:
++                    if range2[0] <= merged_ranges[-1][1]:
++                        merged_ranges[-1] = (merged_ranges[-1][0], range2[1])
++                    else:
++                        merged_ranges.append((range2[0], range2[1]))
++                    indx2 += 1
++                else:
++                    indx2 += 1
++
++        while indx1 < len1:
++            range1 = range_list1[indx1]
++            if range1[1] > merged_ranges[-1][1]:
++                if range1[0] <= merged_ranges[-1][1]:
++                    merged_ranges[-1] = (merged_ranges[-1][0], range1[1])
++                else:
++                    merged_ranges.append((range1[0], range1[1]))
++                indx1 += 1
++            else:
++                indx1 += 1
++        while indx2 < len2:
++            if range2[1] > merged_ranges[-1][1]:
++                if range2[0] <= merged_ranges[-1][1]:
++                    merged_ranges[-1] = (merged_ranges[-1][0], range2[1])
++                else:
++                    merged_ranges.append((range2[0], range2[1]))
++                indx2 += 1
++            else:
++                indx2 += 1
++    return merged_ranges
++
++
++def intersection_ranges(range_list1, range_list2, is_sorted=False):
++    result_range = []
++    if len(range_list1) == 0 or len(range_list2) == 0:
++        return result_range
++    if not is_sorted:
++        range_list1 = merge_self_ranges(range_list1)
++        range_list2 = merge_self_ranges(range_list2)
++
++    len1 = len(range_list1)
++    len2 = len(range_list2)
++    indx1 = 0
++    indx2 = 0
++    range1 = range_list1[indx1]
++    range2 = range_list2[indx2]
++    while indx1 < len1 and indx2 < len2:
++        if range2[1] <= range1[0]:
++            indx2 += 1
++            if indx2 == len2:
++                break
++            range2 = range_list2[indx2]
++
++        elif range2[0] <= range1[0] and range2[1] < range1[1]:
++            assert (range2[1] > range1[0])
++            result_range.append((range1[0], range2[1]))
++            range1 = (range2[1], range1[1])
++            indx2 += 1
++            if indx2 == len2:
++                break
++            range2 = range_list2[indx2]
++
++        elif range2[0] <= range1[0]:
++            assert (range2[1] >= range1[1])
++            result_range.append(range1)
++            range2 = (range1[1], range2[1])
++            indx1 += 1
++            if indx1 == len1:
++                break
++            range1 = range_list1[indx1]
++
++        elif range2[1] < range1[1]:
++            assert (range2[0] > range1[0])
++            result_range.append(range2)
++            range1 = (range2[1], range1[1])
++            indx2 += 1
++            if indx2 == len2:
++                break
++            range2 = range_list2[indx2]
++
++        elif range2[0] < range1[1]:
++            assert (range2[1] >= range1[1])
++            result_range.append((range2[0], range1[1]))
++            range2 = (range1[1], range2[1])
++            indx1 += 1
++            if indx1 == len1:
++                break
++            range1 = range_list1[indx1]
++
++        else:
++            assert (range2[0] >= range1[1])
++            indx1 += 1
++            if indx1 == len1:
++                break
++            range1 = range_list1[indx1]
++    return result_range
++
++
++def subtract_ranges(range_list1, range_list2, is_sorted=False):
++    result_range = []
++    if not is_sorted:
++        range_list1 = merge_self_ranges(range_list1)
++        range_list2 = merge_self_ranges(range_list2)
++    if len(range_list1) == 0:
++        return result_range
++    if len(range_list2) == 0:
++        return range_list1
++
++    len1 = len(range_list1)
++    len2 = len(range_list2)
++    indx1 = 0
++    indx2 = 0
++    range1 = range_list1[indx1]
++    range2 = range_list2[indx2]
++
++    while indx1 < len(range_list1):
++        if indx2 == len(range_list2):
++            result_range.append(range1)
++            indx1 += 1
++            if indx1 == len1:
++                break
++            range1 = range_list1[indx1]
++        elif range2[1] <= range1[0]:
++            indx2 += 1
++            if indx2 != len2:
++                range2 = range_list2[indx2]
++        elif range2[0] <= range1[0] and range2[1] < range1[1]:
++            range1 = (range2[1], range1[1])
++            indx2 += 1
++            if indx2 != len2:
++                range2 = range_list2[indx2]
++        elif range2[0] <= range1[0]:
++            assert (range2[1] >= range1[1])
++            range2 = (range1[1], range2[1])
++            indx1 += 1
++            if indx1 != len1:
++                range1 = range_list1[indx1]
++        elif range2[0] < range1[1]:
++            assert (range2[0] > range1[0])
++            result_range.append((range1[0], range2[0]))
++            range1 = (range2[0], range1[1])
++        else:
++            assert (range2[0] >= range1[1])
++            result_range.append(range1)
++            indx1 += 1
++            if indx1 != len1:
++                range1 = range_list1[indx1]
++    return result_range
diff --git a/swe-paddle/tasks/PaddlePaddle__Paddle-40111/tests/test.patch b/swe-paddle/tasks/PaddlePaddle__Paddle-40111/tests/test.patch
new file mode 100644
index 000000000..3c0daf911
--- /dev/null
+++ b/swe-paddle/tasks/PaddlePaddle__Paddle-40111/tests/test.patch
@@ -0,0 +1,284 @@
+diff --git a/python/paddle/fluid/tests/unittests/test_newprofiler_helper.py b/python/paddle/fluid/tests/unittests/test_newprofiler_helper.py
+new file mode 100644
+index 0000000000000000000000000000000000000000..05e792003545688cfef3f1c7bd48ecf4d27daafe
+--- /dev/null
++++ b/python/paddle/fluid/tests/unittests/test_newprofiler_helper.py
+@@ -0,0 +1,137 @@
++#   Copyright (c) 2022 PaddlePaddle Authors. All Rights Reserved.
++#
++# Licensed under the Apache License, Version 2.0 (the "License");
++# you may not use this file except in compliance with the License.
++# You may obtain a copy of the License at
++#
++#     http://www.apache.org/licenses/LICENSE-2.0
++#
++# Unless required by applicable law or agreed to in writing, software
++# distributed under the License is distributed on an "AS IS" BASIS,
++# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
++# See the License for the specific language governing permissions and
++# limitations under the License.
++
++from __future__ import print_function
++
++import unittest
++
++import paddle.profiler.statistic_helper as statistic_helper
++
++
++class TestStatisticHelper(unittest.TestCase):
++    def test_sum_ranges_case1(self):
++        src = [(1, 3), (4, 10), (11, 15)]
++        self.assertEqual(statistic_helper.sum_ranges(src), 12)
++
++    def test_sum_ranges_case2(self):
++        src = [(3, 3), (5, 5), (7, 7)]
++        self.assertEqual(statistic_helper.sum_ranges(src), 0)
++
++    def test_merge_self_ranges_case1(self):
++        src = [(1, 5), (2, 7), (4, 9), (14, 19)]
++        dst = statistic_helper.merge_self_ranges(src)
++        self.assertEqual(dst, [(1, 9), (14, 19)])
++        src = [(4, 9), (14, 19), (1, 5), (2, 7)]
++        dst = statistic_helper.merge_self_ranges(src)
++        self.assertEqual(dst, [(1, 9), (14, 19)])
++
++    def test_merge_self_ranges_case2(self):
++        src = [(1, 1), (2, 3), (4, 7), (5, 12)]
++        dst = statistic_helper.merge_self_ranges(src)
++        self.assertEqual(dst, [(1, 1), (2, 3), (4, 12)])
++        src = [(5, 12), (1, 1), (2, 3), (4, 7)]
++        dst = statistic_helper.merge_self_ranges(src)
++        self.assertEqual(dst, [(1, 1), (2, 3), (4, 12)])
++
++    def test_merge_ranges_case1(self):
++        src1 = [(1, 2), (5, 7), (9, 14)]
++        src2 = [(1, 2), (4, 9), (13, 15)]
++        dst = statistic_helper.merge_ranges(src1, src2)
++        self.assertEqual(dst, [(1, 2), (4, 15)])
++        dst = statistic_helper.merge_ranges(src1, src2, True)
++        self.assertEqual(dst, [(1, 2), (4, 15)])
++        src1 = []
++        src2 = []
++        dst = statistic_helper.merge_ranges(src1, src2, True)
++        self.assertEqual(dst, [])
++        src1 = [(1, 2), (3, 5)]
++        src2 = []
++        dst = statistic_helper.merge_ranges(src1, src2, True)
++        self.assertEqual(dst, src1)
++        src1 = []
++        src2 = [(1, 2), (3, 5)]
++        dst = statistic_helper.merge_ranges(src1, src2, True)
++        self.assertEqual(dst, src2)
++        src1 = [(3, 4), (1, 2), (17, 19)]
++        src2 = [(6, 9), (13, 15)]
++        dst = statistic_helper.merge_ranges(src1, src2)
++        self.assertEqual(dst, [(1, 2), (3, 4), (6, 9), (13, 15), (17, 19)])
++        dst = statistic_helper.merge_ranges(src2, src1)
++        self.assertEqual(dst, [(1, 2), (3, 4), (6, 9), (13, 15), (17, 19)])
++        src1 = [(1, 2), (5, 9), (12, 13)]
++        src2 = [(6, 8), (9, 15)]
++        dst = statistic_helper.merge_ranges(src1, src2)
++        self.assertEqual(dst, [(1, 2), (5, 15)])
++        dst = statistic_helper.merge_ranges(src2, src1)
++        self.assertEqual(dst, [(1, 2), (5, 15)])
++
++    def test_merge_ranges_case2(self):
++        src1 = [(3, 4), (1, 2), (9, 14)]
++        src2 = [(6, 9), (13, 15)]
++        dst = statistic_helper.merge_ranges(src1, src2)
++        self.assertEqual(dst, [(1, 2), (3, 4), (6, 15)])
++        src2 = [(9, 14), (1, 2), (5, 7)]
++        src1 = [(4, 9), (1, 2), (13, 15)]
++        dst = statistic_helper.merge_ranges(src1, src2)
++        self.assertEqual(dst, [(1, 2), (4, 15)])
++
++    def test_intersection_ranges_case1(self):
++        src1 = [(1, 7), (9, 12), (14, 18)]
++        src2 = [(3, 8), (10, 13), (15, 19)]
++        dst = statistic_helper.intersection_ranges(src1, src2)
++        self.assertEqual(dst, [(3, 7), (10, 12), (15, 18)])
++        dst = statistic_helper.intersection_ranges(src1, src2, True)
++        self.assertEqual(dst, [(3, 7), (10, 12), (15, 18)])
++        src1 = []
++        src2 = []
++        dst = statistic_helper.intersection_ranges(src1, src2, True)
++        self.assertEqual(dst, [])
++        src1 = [(3, 7), (10, 12)]
++        src2 = [(2, 9), (11, 13), (15, 19)]
++        dst = statistic_helper.intersection_ranges(src1, src2)
++        self.assertEqual(dst, [(3, 7), (11, 12)])
++        dst = statistic_helper.intersection_ranges(src2, src1)
++        self.assertEqual(dst, [(3, 7), (11, 12)])
++
++    def test_intersection_ranges_case2(self):
++        src2 = [(9, 12), (1, 7), (14, 18)]
++        src1 = [(10, 13), (3, 8), (15, 19), (20, 22)]
++        dst = statistic_helper.intersection_ranges(src1, src2)
++        self.assertEqual(dst, [(3, 7), (10, 12), (15, 18)])
++        src2 = [(1, 7), (14, 18), (21, 23)]
++        src1 = [(6, 9), (10, 13)]
++        dst = statistic_helper.intersection_ranges(src1, src2, True)
++        self.assertEqual(dst, [(6, 7)])
++
++    def test_subtract_ranges_case1(self):
++        src1 = [(1, 10), (12, 15)]
++        src2 = [(3, 7), (9, 11)]
++        dst = statistic_helper.subtract_ranges(src1, src2, True)
++        self.assertEqual(dst, [(1, 3), (7, 9), (12, 15)])
++        src1 = [(1, 10), (12, 15)]
++        src2 = []
++        dst = statistic_helper.subtract_ranges(src1, src2, True)
++        self.assertEqual(dst, src1)
++        dst = statistic_helper.subtract_ranges(src2, src1, True)
++        self.assertEqual(dst, src2)
++
++    def test_subtract_ranges_case2(self):
++        src2 = [(12, 15), (1, 10)]
++        src1 = [(9, 11), (3, 7)]
++        dst = statistic_helper.subtract_ranges(src1, src2)
++        self.assertEqual(dst, [(10, 11)])
++
++
++if __name__ == '__main__':
++    unittest.main()
+diff --git a/test/swe_paddle/test_pr40111_profiler_ranges.py b/test/swe_paddle/test_pr40111_profiler_ranges.py
+new file mode 100644
+index 0000000000000000000000000000000000000000..88ee5e8cfed30186a8e8ed404356d338a9c7c232
+--- /dev/null
++++ b/test/swe_paddle/test_pr40111_profiler_ranges.py
+@@ -0,0 +1,76 @@
++import importlib.util
++from pathlib import Path
++import sys
++import types
++
++
++REPO_ROOT = Path(__file__).resolve().parents[2]
++TARGET_REL = "python/paddle/profiler/statistic_helper.py"
++UPSTREAM_TEST_REL = (
++    "python/paddle/fluid/tests/unittests/"
++    "test_newprofiler_helper.py"
++)
++
++
++def _load_module(name, path):
++    spec = importlib.util.spec_from_file_location(name, path)
++    module = importlib.util.module_from_spec(spec)
++    spec.loader.exec_module(module)
++    return module
++
++
++def _install_checkout_statistic_helper():
++    target = REPO_ROOT / TARGET_REL
++
++    if target.exists():
++        helper = _load_module(
++            "paddle.profiler.statistic_helper",
++            target,
++        )
++    else:
++        helper = types.ModuleType(
++            "paddle.profiler.statistic_helper"
++        )
++
++        def unavailable(*args, **kwargs):
++            raise NotImplementedError(
++                "profiler statistic range operations "
++                "are unavailable"
++            )
++
++        for name in (
++            "sum_ranges",
++            "merge_self_ranges",
++            "merge_ranges",
++            "intersection_ranges",
++            "subtract_ranges",
++        ):
++            setattr(helper, name, unavailable)
++
++    paddle_module = sys.modules.get("paddle")
++    if paddle_module is None:
++        paddle_module = types.ModuleType("paddle")
++        paddle_module.__path__ = []
++
++    profiler_module = types.ModuleType("paddle.profiler")
++    profiler_module.__path__ = []
++    profiler_module.statistic_helper = helper
++    paddle_module.profiler = profiler_module
++
++    sys.modules["paddle"] = paddle_module
++    sys.modules["paddle.profiler"] = profiler_module
++    sys.modules[
++        "paddle.profiler.statistic_helper"
++    ] = helper
++
++
++_install_checkout_statistic_helper()
++
++_upstream = _load_module(
++    "swe_pr40111_upstream_tests",
++    REPO_ROOT / UPSTREAM_TEST_REL,
++)
++
++
++class TestStatisticHelper(_upstream.TestStatisticHelper):
++    pass
+diff --git a/test/swe_paddle/test_pr40111_profiler_scheduler.py b/test/swe_paddle/test_pr40111_profiler_scheduler.py
+new file mode 100644
+index 0000000000000000000000000000000000000000..8a0e559b3fd4b235abddf72e4650ce4eefca6924
+--- /dev/null
++++ b/test/swe_paddle/test_pr40111_profiler_scheduler.py
+@@ -0,0 +1,53 @@
++import ast
++from enum import Enum
++from pathlib import Path
++from typing import Callable
++from warnings import warn
++
++
++REPO_ROOT = Path(__file__).resolve().parents[2]
++
++
++def _load_existing_profiler_scheduler():
++    path = REPO_ROOT / "python/paddle/profiler/profiler.py"
++    tree = ast.parse(path.read_text(), filename=str(path))
++    body = [
++        node
++        for node in tree.body
++        if (
++            isinstance(node, ast.ClassDef)
++            and node.name == "ProfilerState"
++        )
++        or (
++            isinstance(node, ast.FunctionDef)
++            and node.name == "make_scheduler"
++        )
++    ]
++    namespace = {
++        "Enum": Enum,
++        "Callable": Callable,
++        "warn": warn,
++    }
++    module = ast.Module(body=body, type_ignores=[])
++    ast.fix_missing_locations(module)
++    exec(compile(module, str(path), "exec"), namespace)
++    return namespace["ProfilerState"], namespace["make_scheduler"]
++
++
++def test_p2p_existing_profiler_scheduler_state_sequence():
++    state, make_scheduler = _load_existing_profiler_scheduler()
++    scheduler = make_scheduler(
++        closed=1,
++        ready=1,
++        record=2,
++        repeat=1,
++        skip_first=1,
++    )
++    assert [scheduler(step) for step in range(6)] == [
++        state.CLOSED,
++        state.CLOSED,
++        state.READY,
++        state.RECORD,
++        state.RECORD_AND_RETURN,
++        state.CLOSED,
++    ]
diff --git a/swe-paddle/tasks/PaddlePaddle__Paddle-40111/tests/test.sh b/swe-paddle/tasks/PaddlePaddle__Paddle-40111/tests/test.sh
new file mode 100644
index 000000000..10aa29e66
--- /dev/null
+++ b/swe-paddle/tasks/PaddlePaddle__Paddle-40111/tests/test.sh
@@ -0,0 +1,4 @@
+#!/usr/bin/env bash
+
+set -euo pipefail
+python -m pytest test/swe_paddle/test_pr40111_profiler_ranges.py -q