Index: lib/CodeGen/PrologEpilogInserter.cpp
===================================================================
--- lib/CodeGen/PrologEpilogInserter.cpp
+++ lib/CodeGen/PrologEpilogInserter.cpp
@@ -630,32 +630,6 @@
       AdjustStackOffset(MFI, *I, StackGrowsDown, Offset, MaxAlign, Skew);
   }
 
-  // FIXME: Once this is working, then enable flag will change to a target
-  // check for whether the frame is large enough to want to use virtual
-  // frame index registers. Functions which don't want/need this optimization
-  // will continue to use the existing code path.
-  if (MFI->getUseLocalStackAllocationBlock()) {
-    unsigned Align = MFI->getLocalFrameMaxAlign();
-
-    // Adjust to alignment boundary.
-    Offset = RoundUpToAlignment(Offset, Align, Skew);
-
-    DEBUG(dbgs() << "Local frame base offset: " << Offset << "\n");
-
-    // Resolve offsets for objects in the local block.
-    for (unsigned i = 0, e = MFI->getLocalFrameObjectCount(); i != e; ++i) {
-      std::pair<int, int64_t> Entry = MFI->getLocalFrameObjectMap(i);
-      int64_t FIOffset = (StackGrowsDown ? -Offset : Offset) + Entry.second;
-      DEBUG(dbgs() << "alloc FI(" << Entry.first << ") at SP[" <<
-            FIOffset << "]\n");
-      MFI->setObjectOffset(Entry.first, FIOffset);
-    }
-    // Allocate the local block
-    Offset += MFI->getLocalFrameSize();
-
-    MaxAlign = std::max(Align, MaxAlign);
-  }
-
   // Make sure that the stack protector comes before the local variables on the
   // stack.
   SmallSet<int, 16> ProtectedObjs;
@@ -705,6 +679,32 @@
                           Offset, MaxAlign, Skew);
   }
 
+  // FIXME: Once this is working, then enable flag will change to a target
+  // check for whether the frame is large enough to want to use virtual
+  // frame index registers. Functions which don't want/need this optimization
+  // will continue to use the existing code path.
+  if (MFI->getUseLocalStackAllocationBlock()) {
+    unsigned Align = MFI->getLocalFrameMaxAlign();
+
+    // Adjust to alignment boundary.
+    Offset = RoundUpToAlignment(Offset, Align, Skew);
+
+    DEBUG(dbgs() << "Local frame base offset: " << Offset << "\n");
+
+    // Resolve offsets for objects in the local block.
+    for (unsigned i = 0, e = MFI->getLocalFrameObjectCount(); i != e; ++i) {
+      std::pair<int, int64_t> Entry = MFI->getLocalFrameObjectMap(i);
+      int64_t FIOffset = (StackGrowsDown ? -Offset : Offset) + Entry.second;
+      DEBUG(dbgs() << "alloc FI(" << Entry.first << ") at SP[" <<
+            FIOffset << "]\n");
+      MFI->setObjectOffset(Entry.first, FIOffset);
+    }
+    // Allocate the local block
+    Offset += MFI->getLocalFrameSize();
+
+    MaxAlign = std::max(Align, MaxAlign);
+  }
+
   // Then assign frame offsets to stack objects that are not used to spill
   // callee saved registers.
   for (unsigned i = 0, e = MFI->getObjectIndexEnd(); i != e; ++i) {
Index: test/CodeGen/AArch64/stack-guard-vaarg.ll
===================================================================
--- /dev/null
+++ test/CodeGen/AArch64/stack-guard-vaarg.ll
@@ -0,0 +1,53 @@
+; RUN: llc < %s | FileCheck %s
+target datalayout = "e-m:e-i64:64-i128:128-n32:64-S128"
+target triple = "aarch64--linux"
+
+%struct.__va_list = type { i8*, i8*, i8*, i32, i32 }
+
+; CHECK-LABEL: test
+; CHECK: ldr [[GUARD:x[0-9]+]]{{.*}}:lo12:__stack_chk_guard
+; Make sure the canary is placed relative to the frame pointer, not
+; the stack pointer.
+; CHECK: stur [[GUARD]], [x29,
+
+; Function Attrs: nounwind ssp
+define void @test(i8* %i, ...) #0 {
+entry:
+  %buf = alloca [10 x i8], align 1
+  %ap = alloca %struct.__va_list, align 8
+  %tmp = alloca %struct.__va_list, align 8
+  %0 = getelementptr inbounds [10 x i8], [10 x i8]* %buf, i64 0, i64 0
+  call void @llvm.lifetime.start(i64 10, i8* %0)
+  %1 = bitcast %struct.__va_list* %ap to i8*
+  call void @llvm.lifetime.start(i64 32, i8* %1)
+  call void @llvm.va_start(i8* %1)
+  %2 = bitcast %struct.__va_list* %tmp to i8*
+  call void @llvm.memcpy.p0i8.p0i8.i64(i8* %2, i8* %1, i64 32, i32 8, i1 false)
+  call void @baz(i8* %i, %struct.__va_list* nonnull %tmp)
+  call void @bar(i8* %0)
+  call void @llvm.va_end(i8* %1)
+  call void @llvm.lifetime.end(i64 32, i8* %1)
+  call void @llvm.lifetime.end(i64 10, i8* %0)
+  ret void
+}
+
+; Function Attrs: argmemonly nounwind
+declare void @llvm.lifetime.start(i64, i8* nocapture)
+
+; Function Attrs: nounwind
+declare void @llvm.va_start(i8*)
+
+declare void @baz(i8*, %struct.__va_list*)
+
+; Function Attrs: argmemonly nounwind
+declare void @llvm.memcpy.p0i8.p0i8.i64(i8* nocapture, i8* nocapture readonly, i64, i32, i1)
+
+declare void @bar(i8*)
+
+; Function Attrs: nounwind
+declare void @llvm.va_end(i8*)
+
+; Function Attrs: argmemonly nounwind
+declare void @llvm.lifetime.end(i64, i8* nocapture)
+
+attributes #0 = { ssp }