Closed sorawee closed 3 years ago
Nice, thanks! I'll want to run the drracket test suites to see if any of the tests that look at stacks change (and if not, maybe add something that checks to make sure that stack-checkpoint.rkt doesn't appear on the stack, maybe?).
None of the DrRacket test suites seem to detect this change, so I've pushed the commit along with a test case in the file.
@rfindler FYI: GitHub shows me that https://github.com/racket/drracket/commit/171379b470a99bf9ec8d985028ee7993b9371c49 is an empty commit.
Oops! I've tried again. Sorry about that; I'm not sure what I did.
The current code has an off-by-one error in a sense. It evaluates ccm at a wrong context, causing the remaining stack after cutting the checkpoint to be unclean (notice the line
call-with-stack-checkpoint
in the below "unpatched" screenshots). This PR fixes the issue.Racket CS (patched):
Racket CS (unpatched)
Racket BC (patched)
Racket BC (unpatched)