Skip to content

Conversation

@ericsnowcurrently
Copy link
Member

@ericsnowcurrently ericsnowcurrently commented Nov 20, 2017

The recent global runtime state consolidation broke pre-init usage of Py_DecodeLocale(). This patch fixes that.

https://bugs.python.org/issue32096

Copy link
Contributor

@ncoghlan ncoghlan left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

In addition to a NEWS entry, it would be preferable to add an explicit test for this to _testembed.c.

Copy link
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Doubled word "before before" in the comment.

Copy link
Member Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

fixed

@bedevere-bot
Copy link

A Python core developer has requested some changes be made to your pull request before we can consider merging it. If you could please address their requests along with any other requests in other reviews from core developers that would be appreciated.

Once you have made the requested changes, please leave a comment on this pull request containing the phrase I have made the requested changes; please review again. I will then notify any core developers who have left a review that you're ready for them to take another look at this pull request.

@ncoghlan
Copy link
Contributor

ncoghlan commented Nov 21, 2017

For the explicit test, I'd suggest a dedicated one that just does:

ptr = Py_DecodeLocale();
Py_Initialize();
Py_Finalize();
PyMem_RawFree(ptr);

We can then consider adding equivalent tests for other "safe before Py_Initialize" APIs in https://bugs.python.org/issue32086

Copy link
Contributor

@ncoghlan ncoghlan left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

This looks good to me, but I'm not entirely sure I understand the NEWS entry.

Unless the NEWS entry has typos in it, in which case it makes sense and just needs the typos fixed :)

have been crashing. This is because they rely on the raw memory allocator
having been initialized already. Before the runtime state change the
default raw allocator was initialized statically. This has now been fixed
by falling back to the defaults in PyMem_RawMalloc() and PyMem_RawFree().
Copy link
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Should the last two names here have underscore prefixes?

@ncoghlan
Copy link
Contributor

CI failure looks real though - my guess would be an allocator/deallocator mismatch due to Py_Finalize not explicitly reverting back to the default allocator.

@ericsnowcurrently
Copy link
Member Author

This PR has been superseded by PR #4532, which reverts the mem-related globals consolidation. We will revisit a long-term solution later.

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Projects

None yet

Development

Successfully merging this pull request may close these issues.

4 participants