Skip to content

Conversation

@ZzEeKkAa
Copy link

Intel SVML (Short Vector Math Library) requires custom calling conventions, which are different from system ones (especially on Windows). This patch adds SVML-specific calling conventions to LLVM, including enums, parsing/printing and tablegen. The actual usage in vectorizer will be done in follow up patches.

@github-actions
Copy link

github-actions bot commented Sep 30, 2023

✅ With the latest revision this PR passed the C/C++ code formatter.

@ZzEeKkAa ZzEeKkAa force-pushed the feature/svml_calling_convention branch from e40e69d to 6551e80 Compare September 30, 2023 14:13
@h-vetinari
Copy link
Contributor

@llvm/pr-subscribers-backend-x86
This patch has been lying around unreviewed for a while, and is the first step of a larger effort for upstreaming SVML support. Could someone take a look please?

@RKSimon RKSimon requested review from RKSimon and phoebewang January 5, 2024 06:48
@RKSimon
Copy link
Collaborator

RKSimon commented Jan 5, 2024

Please can you rebase so that it cleanly merges with trunk

Copy link
Collaborator

@RKSimon RKSimon left a comment

Choose a reason for hiding this comment

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

Needs rebasing before review

@h-vetinari
Copy link
Contributor

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment