Skip to content

Conversation

@BurntSushi
Copy link
Member

Most prominently, this reduces the threshold at which backtracking will
run. It looks like giving it much more room than this ends up making it
run slower than the Pike VM because of the zeroing that has to be done
before searching.

In the process, we get rid of the "max program size" and "max input"
distinction. What we want to optimize for is the total size of the state
table, so do that instead.

Most prominently, this reduces the threshold at which backtracking will
run. It looks like giving it much more room than this ends up making it
run slower than the Pike VM because of the zeroing that has to be done
before searching.

In the process, we get rid of the "max program size" and "max input"
distinction. What we want to optimize for is the total size of the state
table, so do that instead.
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

2 participants