Re-enable SIMD tests and use new shuffle builtin #9350
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
The new builtin solves the problem of LLVM changing user-supplied shuffle masks during optimization. The emitted masks will now always be identical to the masks given in the source. Depends on https://reviews.llvm.org/D66983.
kripken
approved these changes
Aug 30, 2019
|
We decided to go a different route in LLVM. |
@tlively , could you elaborate the new route? thanks. |
|
@huningxin See the discussion on the LLVM patch linked above. We decided to teach LLVM about which shuffles are low-cost to prevent it from combining two good shuffles into a bad shuffle. I haven't implemented this yet because I've been focused on getting the tools up to date with the latest new instructions in the spec proposal, but it is on to-do list. |
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Outdated suggestions cannot be applied.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.
The new builtin solves the problem of LLVM changing user-supplied
shuffle masks during optimization. The emitted masks will now always
be identical to the masks given in the source. Depends on
https://reviews.llvm.org/D66983.