Stream: git-wasmtime

Topic: wasmtime / issue #4859 Throw out fewer fuzz inputs with d...

view this post on Zulip Wasmtime GitHub notifications bot (Sep 02 2022 at 22:26):

github-actions[bot] commented on issue #4859:

Subscribe to Label Action

cc @fitzgen

<details>
This issue or pull request has been labeled: "fuzzing"

Thus the following users have been cc'd because of the following labels:

To subscribe or unsubscribe from this label, edit the <code>.github/subscribe-to-label.json</code> configuration file.

Learn more.
</details>

view this post on Zulip Wasmtime GitHub notifications bot (Sep 06 2022 at 16:50):

abrown commented on issue #4859:

Any significant change to the stats you see locally?

view this post on Zulip Wasmtime GitHub notifications bot (Sep 06 2022 at 17:41):

alexcrichton commented on issue #4859:

From oss-fuzz yesterday I see the log entry:

=== Execution rate (22116 successes / 36000 attempted modules): 61.43% ===
    wasmi: 4.16%, spec: 1.84%, wasmtime: 79.39%, v8: 14.62%
    wasm-smith: 87.01%, single-inst: 12.99%

locally after I minified my preexisting corpus and restarted from that corpus I got:

=== Execution rate (28362 successes / 36000 attempted modules): 78.78% ===
        wasmi: 8.32%, spec: 29.61%, wasmtime: 49.08%, v8: 12.99%
        wasm-smith: 74.73%, single-inst: 25.27%

so while results will likely vary it looks like this definitely helps the execution rate in addition to the spec interepreter in particular due to how rarely its conditions were able to be met. With ALLOWED_ENGINES=spec the execution rate (IIRC) went from something like 9% to 90% as well.

Last updated: Nov 22 2024 at 16:03 UTC