Mangle NaN values returned from math functions #48616
Closed
Conversation
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
|
I think this is too harsh. We do expect bitwise equality of the result for a particular version, given equal inputs, even if we don't guarantee what that bit pattern is. |
|
Agreed. I've updated the OP to clarify |
|
This is still a reasonable idea but it has gone stale and will be easier to recreate from scratch than update. |
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
To test dependance on NaN payloads with pkgeval. Supports discussion at #48523.
This is never intended to merge. It would be a terrible semantic and also bad for performance. It's only to inform discussion by seeing what it would hypothetically break.