Wasmtime GitHub notifications bot (Nov 21 2023 at 18:08): Wasmtime GitHub notifications bot (Nov 21 2023 at 18:08):guybedford edited PR #7569.
Wasmtime GitHub notifications bot (Nov 21 2023 at 18:08):guybedford edited PR #7569:
Node.js mtime setting only works up to 100µs accuracy.
This updates the test to work at this level at minimum.
We could make this a configurable test option, I just wasn't sure how to set up such a conditional build feature for the test program - suggestions welcome if not reducing overall accuracy is preferred.
Wasmtime GitHub notifications bot (Nov 21 2023 at 18:32):alexcrichton submitted PR review.
Wasmtime GitHub notifications bot (Nov 21 2023 at 18:32):alexcrichton submitted PR review.
Wasmtime GitHub notifications bot (Nov 21 2023 at 18:32):alexcrichton created PR review comment:
With 100us accuracy in Node this gets rounded to no change, right? Could this -2000 be increased so the mtim is changed in node?
Wasmtime GitHub notifications bot (Nov 21 2023 at 18:32):alexcrichton created PR review comment:
Could these assertions be left as-is with a larger adjustment above that isn't arounded?
Wasmtime GitHub notifications bot (Nov 21 2023 at 21:35):guybedford submitted PR review.
Wasmtime GitHub notifications bot (Nov 21 2023 at 21:35):guybedford created PR review comment:
2000 for u64 is 2000 * 100 nanoseconds is 200 microseconds. So I think this is the right accuracy. Reducing the order by one leads to the times being equal in the check. So I think this is right.
Wasmtime GitHub notifications bot (Nov 21 2023 at 21:36):guybedford submitted PR review.
Wasmtime GitHub notifications bot (Nov 21 2023 at 21:36):guybedford created PR review comment:
Do you mean does changing
2000to say7243lead to the test still passing? If so, I've verified that is true, yes.
Wasmtime GitHub notifications bot (Nov 21 2023 at 21:42):guybedford updated PR #7569.
Wasmtime GitHub notifications bot (Nov 21 2023 at 21:45):guybedford updated PR #7569.
Last updated: Dec 23 2024 at 13:07 UTC