Data races with frame objects

[colesbury]

Bug report

We've fixed most of the data races with Python frame objects in:

• data races in frame inspection and tracebacks #128421

But I think we missed a few places, such as:

• repr(frame) on an frame that is executing or in the process of getting cleared
• frame.f_lasti
• Accessing variables through frame.f_locals

Additionally, I'd like to add more free threading tests for frames.

Linked PRs

• gh-144446: Fix some frame object thread-safety issues #144479
• [3.14] gh-144446: Fix some frame object thread-safety issues (gh-144479) #144546
• gh-144446: Fix thread-safety of FrameLocalsProxy on executing frames #145183
• gh-144446: Fix thread safety of gi_frame, cr_frame and ag_frame in free-threading #156037
• gh-144446: Fix _PyFrame_GetFrameObject() for test_cppext (C++) #156893
• gh-144446: Fix thread-safety of FrameLocalsProxy on executing frames #156973

[LindaSummer]

Hi @colesbury ,
I want to work on the Accessing variables through frame.f_locals sub issue.
Could I try this sub-task? 😊

[colesbury]

Thanks @LindaSummer, but I've already started on it. I'll open a draft PR with the changes. Still thinking about if this is the way I want to implement it.