Skip to content
This repository has been archived by the owner on Jan 23, 2023. It is now read-only.

Explode generic Intel hardware intrinsic on all the supported types #20310

Merged
merged 2 commits into from
Oct 10, 2018

Conversation

fiigii
Copy link

@fiigii fiigii commented Oct 8, 2018

close https://github.com/dotnet/coreclr/issues/20057

This PR removes all the generic Intel hardware intrinsic and explodes them on all the supported types. The generic helpers (e.g., GetLowerHalf<T>) are not handled in this PR, we will move them into Vector128/256<T> classes.

@fiigii fiigii force-pushed the nogeneric branch 2 times, most recently from e04b7d9 to 8af0651 Compare October 9, 2018 00:20
@jkotas
Copy link
Member

jkotas commented Oct 9, 2018

There is a bunch of new warnings like shared\System\Runtime\Intrinsics\X86\Avx2.cs(742,40): warning CS0108: 'Avx2.ExtractVector128(Vector256<sbyte>, byte)' hides inherited member 'Avx.ExtractVector128(Vector256<sbyte>, byte)'. Use the new keyword if hiding was intended. [D:\j\workspace\x64_checked_w---43686a05\src\System.Private.CoreLib\System.Private.CoreLib.csproj]. Could you please fix those?

@fiigii
Copy link
Author

fiigii commented Oct 9, 2018

@jkotas Thank you so much! Will fix it tomorrow.

@fiigii fiigii force-pushed the nogeneric branch 2 times, most recently from be8ae71 to 207fa64 Compare October 9, 2018 19:08
@fiigii
Copy link
Author

fiigii commented Oct 9, 2018

@fiigii
Copy link
Author

fiigii commented Oct 9, 2018

@dotnet-bot test Windows_NT x64 Checked CoreFX Tests
@dotnet-bot test Ubuntu arm Cross Checked Innerloop Build and Test

@CarolEidt
Copy link

@fiigii - you removed a bunch of related tests. Will they be replaced once this flows through corefx?

@fiigii
Copy link
Author

fiigii commented Oct 10, 2018

Yes, I will rewrite/re-enable them later.

@CarolEidt
Copy link

@fiigii - Great, thanks!

Sign up for free to subscribe to this conversation on GitHub. Already have an account? Sign in.
Labels
None yet
Projects
None yet
Development

Successfully merging this pull request may close these issues.

Remove generic Intel hardware intrinsic and explode them on all the supported types
4 participants