diff options
author | Sanjay Patel <spatel@rotateright.com> | 2014-09-01 18:44:57 +0000 |
---|---|---|
committer | Sanjay Patel <spatel@rotateright.com> | 2014-09-01 18:44:57 +0000 |
commit | 5ad239e15a3cd609b86dc6ca06c2928f7a9c1aae (patch) | |
tree | 413882d4635e36fa60f23a0c987bc1e94cbe8c86 /llvm/test/Transforms/LoopVectorize/exact.ll | |
parent | 3ae91933713b20c6f6831b579811acad95560c2b (diff) | |
download | bcm5719-llvm-5ad239e15a3cd609b86dc6ca06c2928f7a9c1aae.tar.gz bcm5719-llvm-5ad239e15a3cd609b86dc6ca06c2928f7a9c1aae.zip |
Add a convenience method to copy wrapping, exact, and fast-math flags (NFC).
The loop vectorizer preserves wrapping, exact, and fast-math properties of scalar instructions.
This patch adds a convenience method to make that operation easier because we need to do this
in the loop vectorizer, SLP vectorizer, and possibly other places.
Although this is a 'no functional change' patch, I've added a testcase to verify that the exact
flag is preserved by the loop vectorizer. The wrapping and fast-math flags are already checked
in existing testcases.
Differential Revision: http://reviews.llvm.org/D5138
llvm-svn: 216886
Diffstat (limited to 'llvm/test/Transforms/LoopVectorize/exact.ll')
-rw-r--r-- | llvm/test/Transforms/LoopVectorize/exact.ll | 24 |
1 files changed, 24 insertions, 0 deletions
diff --git a/llvm/test/Transforms/LoopVectorize/exact.ll b/llvm/test/Transforms/LoopVectorize/exact.ll new file mode 100644 index 00000000000..0a8fbf33734 --- /dev/null +++ b/llvm/test/Transforms/LoopVectorize/exact.ll @@ -0,0 +1,24 @@ +; RUN: opt < %s -loop-vectorize -force-vector-width=4 -S | FileCheck %s + +target datalayout = "e-m:o-i64:64-f80:128-n8:16:32:64-S128" +target triple = "x86_64-apple-macosx10.9.0" + +; CHECK-LABEL: @lshr_exact( +; CHECK: lshr exact <4 x i32> +define void @lshr_exact(i32* %x) { +entry: + br label %for.body + +for.body: + %iv = phi i64 [ 0, %entry ], [ %iv.next, %for.body ] + %arrayidx = getelementptr inbounds i32* %x, i64 %iv + %0 = load i32* %arrayidx, align 4 + %conv1 = lshr exact i32 %0, 1 + store i32 %conv1, i32* %arrayidx, align 4 + %iv.next = add nuw nsw i64 %iv, 1 + %exitcond = icmp eq i64 %iv.next, 256 + br i1 %exitcond, label %for.end, label %for.body + +for.end: + ret void +} |