build-details.json should be installed to different locations for nondebug/debug builds and have distinct base_interpreter values
还没有人认领这个 Issue。
- 主要语言
- Python
- 星标
- 77.2k
- 派生
- 36k
- PR 合并指标
- PR 指标待抓取
描述
Bug report
Bug description:
In Fedora, we build Python in 4 variants: debug, optimized, freethreading-debug and freethreading. For all of them builds the base_interpreter value stored in the file build-details.json is the same:
"base_interpreter": "/usr/bin/python3.14",
I suspect it should reflect the build type and contain the applicable suffix(es): d, t, td.
Because of this line, when we run Python's test suite on an installed python3-debug, the test fails, as you can see e.g. here: https://artifacts.dev.testing-farm.io/0ac8ab69-ed5f-4602-9c01-8c13293aa9b6/
test_base_interpreter (test.test_build_details.CPythonBuildDetailsTests.test_base_interpreter) ... FAIL
======================================================================
FAIL: test_base_interpreter (test.test_build_details.CPythonBuildDetailsTests.test_base_interpreter)
----------------------------------------------------------------------
Traceback (most recent call last):
File "/usr/lib64/python3.14/test/test_build_details.py", line 124, in test_base_interpreter
self.assertEqual(os.path.realpath(value), os.path.realpath(sys.executable))
~~~~~~~~~~~~~~~~^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
AssertionError: '/usr/bin/python3.14' != '/usr/bin/python3.14d'
- /usr/bin/python3.14
+ /usr/bin/python3.14d
? +
Another issue is that in both cases, the debug and non-debug files are installed to the same location, resulting in a single rewritten file. In case of our sequential build of 4 Pythons, we end up with just two files. Should they be stored in distinct locations for each build instead?
CPython versions tested on:
3.14
Operating systems tested on:
Linux
Linked PRs
- gh-142256
- gh-142269
- gh-150098
- gh-150414
贡献指南
从这里开始
- 先读完整个 Issue,再读项目的贡献指南。
- 在 Issue 下留言说明你要接手 —— 这能避免两个人做同样的事。
- Fork 仓库,在一个分支上完成修改。
- 提交 Pull Request,并在描述里引用这个 Issue 编号。
调研方向
从 test/test_build_details.py 开始,重点查看 test_base_interpreter,并复现 issue 中描述的已安装 debug-versus-nondebug 失败。跟踪四种变体的 build-details.json 是如何生成和安装的;完成的标准是每个已安装文件都具有适用的 base_interpreter 值和不同的位置,并且测试通过。
由索引模型根据 Issue 内容生成。
评估
- 技术栈
- python
- 领域
- build-system
- Issue 类型
- 缺陷
- 难度
- 4/5
- 预计耗时
- 3-5 天
- 活跃度
- 停滞
- 描述清晰度
- 基本清楚
- 新手友好度
- 25/100