diff --git a/lib/Target/AVR/AVRISelLowering.cpp b/lib/Target/AVR/AVRISelLowering.cpp
index 04f97471001..3fc8032e73e 100644
--- a/lib/Target/AVR/AVRISelLowering.cpp
+++ b/lib/Target/AVR/AVRISelLowering.cpp
@@ -137,8 +137,11 @@ AVRTargetLowering::AVRTargetLowering(AVRTargetMachine &tm)
   // Expand 16 bit multiplications.
   setOperationAction(ISD::SMUL_LOHI, MVT::i16, Expand);
   setOperationAction(ISD::UMUL_LOHI, MVT::i16, Expand);
-  setOperationAction(ISD::MULHS, MVT::i16, Expand);
-  setOperationAction(ISD::MULHU, MVT::i16, Expand);
+
+  for (MVT VT : MVT::integer_valuetypes()) {
+    setOperationAction(ISD::MULHS, VT, Expand);
+    setOperationAction(ISD::MULHU, VT, Expand);
+  }
 
   //  Runtime library functions
   {
@@ -232,6 +235,12 @@ const char *AVRTargetLowering::getTargetNodeName(unsigned Opcode) const {
   }
 }
 
+EVT AVRTargetLowering::getSetCCResultType(const DataLayout &DL, LLVMContext &,
+                                          EVT VT) const {
+  assert(!VT.isVector() && "No AVR SetCC type for vectors!");
+  return MVT::i8;
+}
+
 SDValue AVRTargetLowering::LowerShifts(SDValue Op, SelectionDAG &DAG) const {
   //:TODO: this function has to be completely rewritten to produce optimal
   // code, for now it's producing very long but correct code.
@@ -1474,9 +1483,12 @@ MachineBasicBlock *AVRTargetLowering::insertShift(MachineInstr *MI,
 }
 
 inline bool isCopyMulResult(MachineBasicBlock::iterator const &I) {
-  unsigned SrcReg = I->getOperand(1).getReg();
-  return I->getOpcode() == AVR::COPY &&
-         (SrcReg == AVR::R0 || SrcReg == AVR::R1);
+  if (I->getOpcode() == AVR::COPY) {
+    unsigned SrcReg = I->getOperand(1).getReg();
+    return (SrcReg == AVR::R0 || SrcReg == AVR::R1);
+  }
+
+  return false;
 }
 
 // The mul instructions wreak havock on our zero_reg R1. We need to clear it
diff --git a/lib/Target/AVR/AVRISelLowering.h b/lib/Target/AVR/AVRISelLowering.h
index 7355c051787..3d5a7c9d498 100644
--- a/lib/Target/AVR/AVRISelLowering.h
+++ b/lib/Target/AVR/AVRISelLowering.h
@@ -94,6 +94,9 @@ class AVRTargetLowering : public TargetLowering {
 
   bool isOffsetFoldingLegal(const GlobalAddressSDNode *GA) const override;
 
+  EVT getSetCCResultType(const DataLayout &DL, LLVMContext &Context,
+                         EVT VT) const override;
+
   MachineBasicBlock *
   EmitInstrWithCustomInserter(MachineInstr *MI,
                               MachineBasicBlock *MBB) const override;
diff --git a/test/CodeGen/AVR/issue-cannot-select-mulhs.ll b/test/CodeGen/AVR/issue-cannot-select-mulhs.ll
deleted file mode 100644
index ba4d7a3fcba..00000000000
--- a/test/CodeGen/AVR/issue-cannot-select-mulhs.ll
+++ /dev/null
@@ -1,33 +0,0 @@
-; RUN: llc < %s -march=avr | FileCheck %s
-; XFAIL: *
-
-define fastcc void @foo(i32) unnamed_addr {
-; CHECK-LABEL: foo:
-entry-block:
-  br i1 undef, label %return, label %next
-
-next:                                             ; preds = %entry-block
-  %1 = trunc i32 %0 to i8
-  br label %exit
-
-exit:                                             ; preds = %match_case1, %next
-  %result.0282 = phi i8 [ %5, %match_case1 ], [ 0, %next ]
-  %2 = icmp ult i32 undef, %0
-  br i1 %2, label %match_case0, label %return
-
-match_case0:                                      ; preds = %exit
-  %3 = tail call { i8, i1 } @llvm.smul.with.overflow.i8(i8 %result.0282, i8 %1)
-  %4 = extractvalue { i8, i1 } %3, 1
-  %brmerge = or i1 %4, undef
-  br i1 %brmerge, label %return, label %match_case1
-
-match_case1:                                      ; preds = %match_case0
-  %5 = extractvalue { i8, i1 } undef, 0
-  br label %exit
-
-return:                                           ; preds = %match_case0, %exit, %entry-block
-  ret void
-}
-
-; Function Attrs: nounwind readnone
-declare { i8, i1 } @llvm.smul.with.overflow.i8(i8, i8)
diff --git a/test/CodeGen/AVR/smul-with-overflow.ll b/test/CodeGen/AVR/smul-with-overflow.ll
new file mode 100644
index 00000000000..745e93005cc
--- /dev/null
+++ b/test/CodeGen/AVR/smul-with-overflow.ll
@@ -0,0 +1,31 @@
+; RUN: llc < %s -march=avr | FileCheck %s
+
+define i1 @signed_multiplication_did_overflow(i8, i8) unnamed_addr {
+; CHECK-LABEL: signed_multiplication_did_overflow:
+entry-block:
+  %2 = tail call { i8, i1 } @llvm.smul.with.overflow.i8(i8 %0, i8 %1)
+  %3 = extractvalue { i8, i1 } %2, 1
+  ret i1 %3
+
+; Multiply, fill the low byte with the sign of the low byte via
+; arithmetic shifting, compare it to the high byte.
+;
+; CHECK: muls   r24, r22
+; CHECK: mov    [[HIGH:r[0-9]+]], r1
+; CHECK: mov    [[LOW:r[0-9]+]], r0
+; CHECK: asr    {{.*}}[[LOW]]
+; CHECK: asr    {{.*}}[[LOW]]
+; CHECK: asr    {{.*}}[[LOW]]
+; CHECK: asr    {{.*}}[[LOW]]
+; CHECK: asr    {{.*}}[[LOW]]
+; CHECK: asr    {{.*}}[[LOW]]
+; CHECK: asr    {{.*}}[[LOW]]
+; CHECK: ldi    [[RET:r[0-9]+]], 1
+; CHECK: cp     {{.*}}[[HIGH]], {{.*}}[[LOW]]
+; CHECK: brne   [[LABEL:LBB[_0-9]+]]
+; CHECK: ldi    {{.*}}[[RET]], 0
+; CHECK: {{.*}}[[LABEL]]
+; CHECK: ret
+}
+
+declare { i8, i1 } @llvm.smul.with.overflow.i8(i8, i8)
diff --git a/test/CodeGen/AVR/umul-with-overflow.ll b/test/CodeGen/AVR/umul-with-overflow.ll
new file mode 100644
index 00000000000..aa8b10a313d
--- /dev/null
+++ b/test/CodeGen/AVR/umul-with-overflow.ll
@@ -0,0 +1,22 @@
+; RUN: llc < %s -march=avr | FileCheck %s
+
+define i1 @unsigned_multiplication_did_overflow(i8, i8) unnamed_addr {
+; CHECK-LABEL: unsigned_multiplication_did_overflow:
+entry-block:
+  %2 = tail call { i8, i1 } @llvm.umul.with.overflow.i8(i8 %0, i8 %1)
+  %3 = extractvalue { i8, i1 } %2, 1
+  ret i1 %3
+
+; Multiply, return if the high byte is zero
+;
+; CHECK: mul    r{{[0-9]+}}, r{{[0-9]+}}
+; CHECK: mov    [[HIGH:r[0-9]+]], r1
+; CHECK: ldi    [[RET:r[0-9]+]], 1
+; CHECK: cpi    {{.*}}[[HIGH]], 0
+; CHECK: brne   [[LABEL:LBB[_0-9]+]]
+; CHECK: ldi    {{.*}}[[RET]], 0
+; CHECK: {{.*}}[[LABEL]]
+; CHECK: ret
+}
+
+declare { i8, i1 } @llvm.umul.with.overflow.i8(i8, i8)