diff options
| author | Daniel Neilson <dneilson@azul.com> | 2018-05-01 15:35:08 +0000 |
|---|---|---|
| committer | Daniel Neilson <dneilson@azul.com> | 2018-05-01 15:35:08 +0000 |
| commit | 9e4bbe801a79030934627ceafca79c75f45eb58a (patch) | |
| tree | 007da645eaf7badfa0b4afe87e1c79c05347a22f /llvm/test/Transforms/LoopVectorize/interleaved-accesses.ll | |
| parent | 969d63ea8c5820d1325ad27205fc34c4f5030a54 (diff) | |
| download | bcm5719-llvm-9e4bbe801a79030934627ceafca79c75f45eb58a.tar.gz bcm5719-llvm-9e4bbe801a79030934627ceafca79c75f45eb58a.zip | |
[LV] Preserve inbounds on created GEPs
Summary:
This is a fix for PR23997.
The loop vectorizer is not preserving the inbounds property of GEPs that it creates.
This is inhibiting some optimizations. This patch preserves the inbounds property in
the case where a load/store is being fed by an inbounds GEP.
Reviewers: mkuper, javed.absar, hsaito
Reviewed By: hsaito
Subscribers: dcaballe, hsaito, llvm-commits
Differential Revision: https://reviews.llvm.org/D46191
llvm-svn: 331269
Diffstat (limited to 'llvm/test/Transforms/LoopVectorize/interleaved-accesses.ll')
| -rw-r--r-- | llvm/test/Transforms/LoopVectorize/interleaved-accesses.ll | 4 |
1 files changed, 2 insertions, 2 deletions
diff --git a/llvm/test/Transforms/LoopVectorize/interleaved-accesses.ll b/llvm/test/Transforms/LoopVectorize/interleaved-accesses.ll index 5d730685f3e..2d8e9afbb91 100644 --- a/llvm/test/Transforms/LoopVectorize/interleaved-accesses.ll +++ b/llvm/test/Transforms/LoopVectorize/interleaved-accesses.ll @@ -245,7 +245,7 @@ for.body: ; preds = %for.body, %entry ; CHECK-LABEL: @test_reversed_load2_store2( ; CHECK: %[[G0:.+]] = getelementptr inbounds %struct.ST2, %struct.ST2* %A, i64 %offset.idx, i32 0 -; CHECK: %[[G1:.+]] = getelementptr i32, i32* %[[G0]], i64 -6 +; CHECK: %[[G1:.+]] = getelementptr inbounds i32, i32* %[[G0]], i64 -6 ; CHECK: %[[B0:.+]] = bitcast i32* %[[G1]] to <8 x i32>* ; CHECK: %wide.vec = load <8 x i32>, <8 x i32>* %[[B0]], align 4 ; CHECK: shufflevector <8 x i32> %wide.vec, <8 x i32> undef, <4 x i32> <i32 0, i32 2, i32 4, i32 6> @@ -255,7 +255,7 @@ for.body: ; preds = %for.body, %entry ; CHECK: add nsw <4 x i32> ; CHECK: sub nsw <4 x i32> ; CHECK: %[[G2:.+]] = getelementptr inbounds %struct.ST2, %struct.ST2* %B, i64 %offset.idx, i32 1 -; CHECK: %[[G3:.+]] = getelementptr i32, i32* %[[G2]], i64 -7 +; CHECK: %[[G3:.+]] = getelementptr inbounds i32, i32* %[[G2]], i64 -7 ; CHECK: %[[B1:.+]] = bitcast i32* %[[G3]] to <8 x i32>* ; CHECK: shufflevector <4 x i32> {{.*}}, <4 x i32> <i32 3, i32 2, i32 1, i32 0> ; CHECK: shufflevector <4 x i32> {{.*}}, <4 x i32> <i32 3, i32 2, i32 1, i32 0> |
