gh-138431: JIT Optimizer --- Fix round-tripping references for str and tuple

[Fidget-Spinner]

The problem is that when something round-trips through str or tuple, it isn't guaranteed to maintain the same reference information.

The test script deals \with some quirks of immortalization/interning in CPython.

• Issue: JIT: failed assertion in _POP_TOP_NOP calling str() on variable due to wrong reference information #138431

[markshannon]

I don't think this has anything to do with string interning, but that tuple(t) and str(s) return a new reference to the object.
What was a deferred reference becomes non-deferred when returned from a call.

[Fidget-Spinner]

I don't think this has anything to do with string interning, but that tuple(t) and str(s) return a new reference to the object. What was a deferred reference becomes non-deferred when returned from a call.

Yes I'm aware. Sorry that I didn't put it in the first sentence. What I was talking about (interning/immortalization) is a quirk of why the test script is written that way. For some reason if it's in a separate script, it stays non-immortal but interned, but if it's in the same test file, it stays immortal and interned, causing some weird behavior.

[markshannon]

Looks good

[miss-islington-app]

Thanks @Fidget-Spinner for the PR 🌮🎉.. I'm working now to backport this PR to: 3.14.
🐍🍒⛏🤖

[miss-islington-app]

Sorry, @Fidget-Spinner, I could not cleanly backport this to 3.14 due to a conflict.
Please backport using cherry_picker on command line.

cherry_picker 2402f84665df41732c94db2ccd6f90dd13479fa8 3.14