Would anyone object to the removal of
mir::StatementKind::EndRegion? The only reason they have for existing now is if someone wants to inspect them with -Zdump-mir (and -Zemit-end-regions).
Wow, so much code from lexical MIR borrowck is still around.
I got foiled by Zulip for the first time :-)
I was suggesting that you just submit the PR to remove it with a taunting commit message; make them prove why it's still useful ;-)
@Matthew Jasper I think miri might use EndRegion for its modeling of Invalidating references; cc @RalfJ and @Oli
I think that's they've moved away from them with #55316 (I guess it might not be completely done on the MIRI side)
@pnkfelix @Matthew Jasper indeed,
EndRegion is no longer used by miri
okay then kill it with fire
please do :)
@Matthew Jasper "Now that lexical MIR borrowck is gone, there's no need to store Regions
unnecessarily." -- what is "lexical MIR borrowck"?
that's like Ye Olden Days borrowck
from before we made NLL mandatory
/me forgot about that phase altogether
I assume this means lexical MIR borrowck? The old AST/HIR lexical borrowck is still there, right? (Though it's days are numbered I hear.)