gh-128563: Move lltrace into the frame struct

[Fidget-Spinner]

• Issue: A new tail-calling interpreter for significantly better interpreter performance #128563

[markshannon]

The change to the size of the frame can be avoided by using a bitfield.

I'm curious why the atomic loads and stores are needed.

[bedevere-app]

When you're done making the requested changes, leave the comment: I have made the requested changes; please review again.

[Fidget-Spinner]

I have made the requested changes; please review again.

[bedevere-app]

Thanks for making the requested changes!

@markshannon: please review the changes made to this pull request.

[markshannon]

Why 4?
You only need one bit (although it is best to use uint8_t as the signedness of char is implementation defined.

My mistake, lltrace is not boolean.

Since char might be signed, it is best to make this a uint8_t to avoid overflow for PYTHON_LLTRACE=9

[markshannon]

One small issue, otherwise looks good.

lltrace should be unsigned (uint8_t), and maybe bigger, to avoid overflow. visited only needs one bit, so lltrace could use the rest.
For consistency, make visited a uint8_t in the non-debug build as well.

[markshannon]

Looks good.

[chris-eibl]

Sorry for being late here. These lines seem to be unused, since the PR did not have to touch them?

cpython/Python/ceval.c

Lines 1066 to 1071 in 05d12ee

	#ifdef Py_DEBUG
	#define DPRINTF(level, ...) \
	if (lltrace >= (level)) { printf(__VA_ARGS__); }
	#else
	#define DPRINTF(level, ...)
	#endif

[Fidget-Spinner]

@chris-eibl it seems so. If you would like to, could you please open a PR to either update to fix them or remove them? Thanks!

[chris-eibl]

Yeah, but since this is gonna be my first PR, I'll have to work myself through the devguide. Most probably not before the weekend ...