[llvm] a0c558e - [Docs] Improve SLP code snippet
Sanjay Patel via llvm-commits
llvm-commits at lists.llvm.org
Tue Dec 10 06:32:57 PST 2019
Author: Sanjay Patel
Date: 2019-12-10T09:32:40-05:00
New Revision: a0c558ee4cc0f8cd1d30edac5f4fbdedb18eff21
URL: https://github.com/llvm/llvm-project/commit/a0c558ee4cc0f8cd1d30edac5f4fbdedb18eff21
DIFF: https://github.com/llvm/llvm-project/commit/a0c558ee4cc0f8cd1d30edac5f4fbdedb18eff21.diff
LOG: [Docs] Improve SLP code snippet
New C code snippet is more viable for SLP vectorization in most architectures.
Patch by: @lsandov1 (Leonardo Sandoval)
Differential Revision: https://reviews.llvm.org/D70866
Added:
Modified:
llvm/docs/Vectorizers.rst
Removed:
################################################################################
diff --git a/llvm/docs/Vectorizers.rst b/llvm/docs/Vectorizers.rst
index fd1525c00cac..f8477f20f037 100644
--- a/llvm/docs/Vectorizers.rst
+++ b/llvm/docs/Vectorizers.rst
@@ -418,8 +418,10 @@ into vector operations.
.. code-block:: c++
void foo(int a1, int a2, int b1, int b2, int *A) {
- A[0] = a1*(a1 + b1)/b1 + 50*b1/a1;
- A[1] = a2*(a2 + b2)/b2 + 50*b2/a2;
+ A[0] = a1*(a1 + b1);
+ A[1] = a2*(a2 + b2);
+ A[2] = a1*(a1 + b1);
+ A[3] = a2*(a2 + b2);
}
The SLP-vectorizer processes the code bottom-up, across basic blocks, in search of scalars to combine.
More information about the llvm-commits
mailing list