bug#65726: 29.1.50; Crash in regexp engine

[Mattias Engdegård]

21 sep. 2023 kl. 20.08 skrev Stefan Monnier <monnier@iro.umontreal.ca>:

> In any case, I pushed a variant of the last patch to `master` and
> I think the case is now closed.  We may want to revisit it, presumably
> by really keeping track of the all visited nodes, but this seems to work
> well for now and I'm confident that it is safe (regardless of the shape
> of the bytecode, it should always terminate and it should never apply
> the optimization if it's incorrect, and with the shape of the bytecode
> we currently produce it does correctly apply the optimization in the cases
> I checked).

Good, thank you.