Skip to content

Conversation

@serge-sans-paille
Copy link
Contributor

It was only supported on Intel architectures and wasm, for no clear reason. Provide a unified experience by enforcing batch_bool instead. If someone really wants a batch of bool, then use another type to store it, e.g. uint8_t.

Fix #945

@serge-sans-paille
Copy link
Contributor Author

@JohanMabille this probably requires a companion patch in xtensor.

It was only supported on Intel architectures and wasm, for no clear
reason. Provide a unified experience by enforcing batch_bool instead. If
someone really wants a batch of bool, then use another type to store it,
e.g. uint8_t.

Fix #945
@serge-sans-paille serge-sans-paille force-pushed the feature/prohibit-batch-of-bool branch from 30cdcf6 to ead311f Compare November 12, 2023 22:01
Copy link
Contributor

@anutosh491 anutosh491 left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Looks good to me. I think we can disable batch of bools

@serge-sans-paille
Copy link
Contributor Author

I'm leaving it to @JohanMabille to choose whether we should merge this or not :-)

@JohanMabille
Copy link
Member

I'm definitely for this, but this is a backward incompatible change. Do we have other fixes we want to get in and tag a minor release before this one?

@serge-sans-paille
Copy link
Contributor Author

@JohanMabille I'd like to investigate the accuracy issue with some trigo functions, but that can wait. Feel free to merge this one and do a release

@JohanMabille JohanMabille merged commit a44922d into master Nov 29, 2023
@JohanMabille JohanMabille deleted the feature/prohibit-batch-of-bool branch November 29, 2023 09:46
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.

xtensor triggers xsimd error on arm64: <bool, xsimd::neon64> usage of batch type with unsupported type

4 participants