Given it is impossible for tools doing passive inspection of Python VM instances to execute code, co_linetable's exact format will be depended on just as co_lnotab was. co_lnotab was only quasi-"officially" documented in the Python docs, it's spec lives in https://github.com/python/cpython/blob/master/Objects/lnotab_notes.txt (pointed to by a couple module's docs). The lnotab format "changed" once, in 3.6, an unsigned delta was changed to signed (but I don't believe anything beyond some experiments ever actually used negatives?).
Negatives definitely happen. When I comment out the line in coverage.py that deals with negative deltas, 34 of my tests fail.
For example:
a = (
1
)
With 3.8 compiles to:
2 0 LOAD_CONST 0 (1)
1 2 STORE_NAME 0 (a)
4 LOAD_CONST 1 (None)
6 RETURN_VALUE
With an lnotab of "02 ff".
When executed, this produces these trace events:
call on line 1
line on line 2
line on line 1
return on line 1
--Ned.