Change generation of 2^n values for fixed point conversions. #1261
Conversation
This file contains 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
As demonstrated by test code in simd-everywhere#1260 the behavior of pow() in non-round-to-nearest rounding modes is not exact. This causes behavior divergence from ARMv8 hardware when not using round-to-nearest. The updated forms match hardware properly across a range of values. The tests are not updated to handle rounding modes, as doing this in a cross-platform way is not trivial. However, all existing test vectors pass properly, and in more detailed testing, these changes are closer to hardware.
|
Okeday, needs a few tweaks... |
Let me know if you need help dealing with the compiler complaints :-) |
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Outdated suggestions cannot be applied.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.
As demonstrated by test code in #1260 the behavior of pow() in non-round-to-nearest rounding modes is not exact. This causes behavior divergence from ARMv8 hardware when not using round-to-nearest. The updated forms match hardware properly across a range of values. The tests are not updated to handle rounding modes, as doing this in a cross-platform way is not trivial. However, all existing test vectors pass properly, and in more detailed testing, these changes are closer to hardware.