Closed KitsuneAlex closed 2 years ago
Cool. thanks. Oh we use the Linux DCO now not the contributors.txt :) Please check out https://github.com/antlr/intellij-plugin-v4/pull/554/checks?check_run_id=6442612723
Wow. did this really make a diff in memory?
I am going to do some more profiling and post some results later :) Also, just saw that ^^ Thanks :)
I ammended my sign-offs, does the workflow need to be re-triggered?
I set it to pass for ease, since you signed the contribs. :) will examine later.
As i mentioned, imma keep skimming through the code while looking at VisualVM, maybe i can plug some more hotspots.
This sign-off workflow is still super confusing to me. Exuse my clumsyness xD
no prob. just make sure to use -s
on your git commits :)
These changes might be useful but real issue is antlr's algo on this I think.
For sure, i just wanna get as much fixes in while i'm at it, regardless of their size. Also i just messed something up, gonna fix that really quick.
I am sorry i have to ask again, but does the sign-off name have to match my GitHub username?
I think the email address Hass to be one associated with your GitHub account
That's weird, that is the case already.
It’s probably the previous commits that is screwing this up. There are instructions about resetting and recomittting but we can fake our way to victory if necessary on this one
I'll ammend the signoffs to these tomorrow :)
Okay, i'm closing this "failed" PR for now since i need to ammend all the signoffs anyways. I started writing an optimization starting at the ANTLR runtime level, so i'll come back to this once that's done. You'll see a new, bigger PR soon :)
This commit fixes some redundant branches, v-table lookups and an allocation hotspot caused by implicit boxing/unboxing. I'll be adding more optimizations to this until it's merged but this already dropped RAM usage a good bit and released some GC pressure.
Cheers.