From f44573ac9bfe9a6ba30851a1f1f875ac5ab41ca4 Mon Sep 17 00:00:00 2001
From: Nick Lewycky <nicholas@mxc.ca>
Date: Tue, 25 Sep 2012 22:46:21 +0000
Subject: [PATCH] Don't drop the alignment on a memcpy intrinsic when producing
 a store. This is only a missed optimization opportunity if the store is
 over-aligned, but a miscompile if the store's new type has a higher natural
 alignment than the memcpy did. Fixes PR13920!

llvm-svn: 164641
---
 lib/Transforms/Scalar/SROA.cpp    |  5 +++--
 test/Transforms/SROA/basictest.ll | 16 ++++++++++++++++
 2 files changed, 19 insertions(+), 2 deletions(-)

diff --git a/lib/Transforms/Scalar/SROA.cpp b/lib/Transforms/Scalar/SROA.cpp
index 1b3e8f9baf4..04e350c25fc 100644
--- a/lib/Transforms/Scalar/SROA.cpp
+++ b/lib/Transforms/Scalar/SROA.cpp
@@ -2272,8 +2272,9 @@ private:
                                     getName(".insert"));
     }
 
-    Value *Store = IRB.CreateStore(Src, DstPtr, II.isVolatile());
-    (void)Store;
+    StoreInst *Store = cast<StoreInst>(IRB.CreateStore(Src, DstPtr,
+                                                       II.isVolatile()));
+    Store->setAlignment(II.getAlignment());
     DEBUG(dbgs() << "          to: " << *Store << "\n");
     return !II.isVolatile();
   }
diff --git a/test/Transforms/SROA/basictest.ll b/test/Transforms/SROA/basictest.ll
index 359a56a00d5..f7a6e0e1e83 100644
--- a/test/Transforms/SROA/basictest.ll
+++ b/test/Transforms/SROA/basictest.ll
@@ -863,6 +863,7 @@ define void @test22() {
 ; CHECK-NOT: alloca
 ; CHECK: ret void
 ; PR13916
+
 entry:
   %A = alloca %test22.struct
   br i1 undef, label %if.then, label %if.end
@@ -877,3 +878,18 @@ if.end:                                           ; preds = %entry
   %tmp2 = load %test22.struct* %A
   ret void
 }
+
+define void @test23(<2 x i64> %a, i16* %b) {
+; CHECK: @test23
+; CHECK: store {{.*}}, align 2
+; CHECK: ret void
+; PR13920
+
+entry:
+  %a.addr = alloca <2 x i64>, align 16
+  store <2 x i64> %a, <2 x i64>* %a.addr, align 16
+  %0 = bitcast i16* %b to i8*
+  %1 = bitcast <2 x i64>* %a.addr to i8*
+  call void @llvm.memcpy.p0i8.p0i8.i32(i8* %0, i8* %1, i32 16, i32 2, i1 false)
+  ret void
+}