[3.11] GH-96636: Remove all uses of NOTRACE_DISPATCH (GH-96643)

[markshannon]

Co-authored-by: Brandt Bucher [email protected]
(cherry picked from commit aa3b4cf)

• Issue: NOTRACE_DISPATCH breaks tracing #96636

[gvanrossum]

Whoops this probably should wait for @pablogsal before it gets merged.

[markshannon]

@pablogsal?

[pablogsal]

It would be great if we can quickly check that the coverage test suite passes or doesn't fail more with this PR before landing.

Otherwise, LGTM, land when ready 👍

[markshannon]

This shouldn't impact coverage.py. This is for weird code that turns on tracing in finalizers, weakrefs or asynchronously.

I'm fairly sure coverage.py doesn't do any of those things. @nedbat?

[nedbat]

Coverage.py doesn't do any of those weird things :) Also, I now have nightly build testing every day, so we'll hear soon if this broke something.

[pablogsal]

Ok, let's merge it then 👍

[neonene]

An extra DISPATCH() macro has been added accidentally, which has no harm.

cpython/Python/ceval.c

Lines 3582 to 3586 in 46f791d

	DISPATCH();
	}
	DISPATCH();
	TARGET(STORE_ATTR_ADAPTIVE) {

[brandtbucher]

Nice catch. I'll open a PR to fix.