Use resource_size instead of -j - #54771
Conversation
JSGette
added
changelog/no-changelog
![chatgpt-codex-connector[bot]](https://avatars.githubusercontent.com/in/1144995?s=60&v=4){kind=small}
There was a problem hiding this comment.
💡 Codex Review
Here are some automated review suggestions for this pull request.
Reviewed commit: 13aed1c609
ℹ️ About Codex in GitHub
Your team has set up Codex to review pull requests in this repo. Reviews are triggered when you
- Open a pull request for review
- Mark a draft as ready
- Comment "@codex review".
If Codex has suggestions, it will comment; otherwise it will react with 👍.
Codex can also answer questions or update the PR. Try commenting "@codex address that feedback".
| """.format( | ||
| version = PYTHON_MAJOR_MINOR, | ||
| ), | ||
| resource_size = "enormous", |
There was a problem hiding this comment.
On Linux and macOS, resource_size = "enormous" is translated by bazel/patches/rules_foreign_cc/0003-pass-j-flag-directly-on-make-command-line.patch into a -j<N> argument for every make invocation, so the remaining targets = ["install"] now executes a parallel make install. The previous configuration deliberately used separate "-j 16" and "install" targets so compilation was parallel but installation stayed serial due to CPython's known concurrent-install race; this reintroduces flaky or incomplete embedded-Python builds. Preserve a separate serial install invocation while deriving only the compilation parallelism from the resource allocation.
Useful? React with 👍 / 👎.
There was a problem hiding this comment.
looks like a valid concern, no?
There was a problem hiding this comment.
I will drop the whole PR then, krb is small anyhow
2 participants
What does this PR do?
Usage of
resource_sizeprevents action keys from being polluted by the number of threads assigned to build processes driven byrules_foreign_cc.