diff options
| author | Arnold Schwaighofer <aschwaighofer@apple.com> | 2014-01-11 20:40:34 +0000 |
|---|---|---|
| committer | Arnold Schwaighofer <aschwaighofer@apple.com> | 2014-01-11 20:40:34 +0000 |
| commit | 73c9559237d61a53c3002eff0aeeee49617b9033 (patch) | |
| tree | c9702a7760108f780073bb00881404d7d5902154 /lib | |
| parent | 810469037074bfc990db6874eac6831d34ce7ccc (diff) | |
| download | llvm-73c9559237d61a53c3002eff0aeeee49617b9033.tar.gz llvm-73c9559237d61a53c3002eff0aeeee49617b9033.tar.bz2 llvm-73c9559237d61a53c3002eff0aeeee49617b9033.tar.xz | |
LoopVectorizer: Enable strided memory accesses versioning per default
I saw no compile or execution time regressions on x86_64 -mavx -O3.
radar://13075509
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@199015 91177308-0d34-0410-b5e6-96231b3b80d8
Diffstat (limited to 'lib')
| -rw-r--r-- | lib/Transforms/Vectorize/LoopVectorize.cpp | 2 |
1 files changed, 1 insertions, 1 deletions
diff --git a/lib/Transforms/Vectorize/LoopVectorize.cpp b/lib/Transforms/Vectorize/LoopVectorize.cpp index 6a4d8b411c..3ca0f14ee3 100644 --- a/lib/Transforms/Vectorize/LoopVectorize.cpp +++ b/lib/Transforms/Vectorize/LoopVectorize.cpp @@ -126,7 +126,7 @@ TinyTripCountVectorThreshold("vectorizer-min-trip-count", cl::init(16), /// } else /// ... static cl::opt<bool> EnableMemAccessVersioning( - "enable-mem-access-versioning", cl::init(false), cl::Hidden, + "enable-mem-access-versioning", cl::init(true), cl::Hidden, cl::desc("Enable symblic stride memory access versioning")); /// We don't unroll loops with a known constant trip count below this number. |