Skip to content

Conversation

@artagnon
Copy link
Contributor

@artagnon artagnon commented Dec 3, 2024

The configuration option is useful for returning the exact fp from FloatType::fromFloat, ignoring NaN, and is required by Minotaur [https://arxiv.org/abs/2306.00229] to synthesize correctly. use_exact_fp is turned off by default, and should pose no correctness issue when turned on.

The configuration option is useful for returning the exact fp from
FloatType::fromFloat, ignoring NaN, and is required by Minotaur
[https://arxiv.org/abs/2306.00229] to synthesize correctly. use_exact_fp
is turned off by default, and should pose no correctness issue when
turned on.
@artagnon
Copy link
Contributor Author

artagnon commented Dec 3, 2024

I have checked that when this is applied along with #1128, all tests in Minotaur pass.

@nunoplopes
Copy link
Member

This is not correct. Verification done by disabling the whole NaN stuff cannot be relied upon.
Either Minotaur assumes that or it uses an over-approximating semantics, such as using UFs.

@nunoplopes nunoplopes closed this Apr 21, 2025
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