fitzgen opened Issue #2695:
I originally punted on implementing this by hand and just gave all the type constraints to Z3, which was easy and is usually fast, but occasionally Z3 gets confused and hangs. Also the fuzzers have discovered inputs where Z3 hangs, which leads to noise in our fuzzing results.
We should rewrite Peepmatic's type checking to use the traditional approach to unifying type constraints via union-find.
fitzgen labeled Issue #2695:
I originally punted on implementing this by hand and just gave all the type constraints to Z3, which was easy and is usually fast, but occasionally Z3 gets confused and hangs. Also the fuzzers have discovered inputs where Z3 hangs, which leads to noise in our fuzzing results.
We should rewrite Peepmatic's type checking to use the traditional approach to unifying type constraints via union-find.
github-actions[bot] commented on Issue #2695:
Subscribe to Label Action
cc @fitzgen
<details>
This issue or pull request has been labeled: "cranelift:area:peepmatic"Thus the following users have been cc'd because of the following labels:
- fitzgen: cranelift:area:peepmatic
To subscribe or unsubscribe from this label, edit the <code>.github/subscribe-to-label.json</code> configuration file.
Learn more.
</details>
fitzgen commented on Issue #2695:
When we do this, we should re-introduce the fuzz targets removed in https://github.com/bytecodealliance/wasmtime/pull/2696
fitzgen closed Issue #2695:
I originally punted on implementing this by hand and just gave all the type constraints to Z3, which was easy and is usually fast, but occasionally Z3 gets confused and hangs. Also the fuzzers have discovered inputs where Z3 hangs, which leads to noise in our fuzzing results.
We should rewrite Peepmatic's type checking to use the traditional approach to unifying type constraints via union-find.
lqd commented on Issue #2695:
@fitzgen this was probably fixed by mistake automatically by github because of the "when we fix" this issue text in that PR ? :)
fitzgen reopened Issue #2695:
I originally punted on implementing this by hand and just gave all the type constraints to Z3, which was easy and is usually fast, but occasionally Z3 gets confused and hangs. Also the fuzzers have discovered inputs where Z3 hangs, which leads to noise in our fuzzing results.
We should rewrite Peepmatic's type checking to use the traditional approach to unifying type constraints via union-find.
fitzgen commented on Issue #2695:
Ah yes, thanks for the heads up!
Last updated: Dec 23 2024 at 12:05 UTC