diff --git a/llvm/lib/Target/PowerPC/PPCISelDAGToDAG.cpp b/llvm/lib/Target/PowerPC/PPCISelDAGToDAG.cpp
--- a/llvm/lib/Target/PowerPC/PPCISelDAGToDAG.cpp
+++ b/llvm/lib/Target/PowerPC/PPCISelDAGToDAG.cpp
@@ -4973,6 +4973,54 @@
     // Other cases are autogenerated.
     break;
   }
+  case ISD::MUL: {
+    SDValue Op0 = N->getOperand(0);
+    SDValue Op1 = N->getOperand(1);
+    if (Op1.getOpcode() != ISD::Constant || Op1.getValueType() != MVT::i64)
+      break;
+
+    int64_t Imm = cast<ConstantSDNode>(Op1)->getZExtValue();
+    unsigned Shift = countTrailingZeros<uint64_t>(Imm);
+    if (isInt<16>(Imm) || !Shift)
+      break;
+
+    int64_t ImmSh = Imm >> Shift;
+    if (isInt<16>(ImmSh)) {
+      uint64_t SextImm = SignExtend64(ImmSh & 0xFFFF, 16);
+      SDValue SDImm = CurDAG->getTargetConstant(SextImm, dl, MVT::i64);
+      SDNode *MulNode = CurDAG->getMachineNode(PPC::MULLI8, dl, MVT::i64,
+                                               N->getOperand(0), SDImm);
+      CurDAG->SelectNodeTo(N, PPC::RLDICR, MVT::i64, SDValue(MulNode, 0),
+                           getI32Imm(Shift, dl), getI32Imm(63 - Shift, dl));
+      return;
+    }
+    if (isPowerOf2_64(ImmSh - 1)) {
+      // The multiplicand has exactly two bits set so we don't need a multiply.
+      unsigned Shift2 = countTrailingZeros<uint64_t>(ImmSh - 1);
+      SDNode *ShNode = CurDAG->getMachineNode(PPC::RLDICR, dl, MVT::i64, Op0,
+                                              getI32Imm(Shift2, dl),
+                                              getI32Imm(63 - Shift2, dl));
+      SDNode *AddNode = CurDAG->getMachineNode(PPC::ADD8, dl, MVT::i64, Op0,
+                                               SDValue(ShNode, 0));
+      CurDAG->SelectNodeTo(N, PPC::RLDICR, MVT::i64, SDValue(AddNode, 0),
+                           getI32Imm(Shift, dl), getI32Imm(63 - Shift, dl));
+      return;
+    }
+    if (isPowerOf2_64(ImmSh + 1)) {
+      // The multiplicand is a contiguous sequence of bits. We don't need a
+      // multiply.
+      unsigned Shift2 = countTrailingZeros<uint64_t>(ImmSh + 1);
+      SDNode *ShNode = CurDAG->getMachineNode(PPC::RLDICR, dl, MVT::i64, Op0,
+                                              getI32Imm(Shift2, dl),
+                                              getI32Imm(63 - Shift2, dl));
+      SDNode *SubNode = CurDAG->getMachineNode(PPC::SUBF8, dl, MVT::i64, Op0,
+                                               SDValue(ShNode, 0));
+      CurDAG->SelectNodeTo(N, PPC::RLDICR, MVT::i64, SDValue(SubNode, 0),
+                           getI32Imm(Shift, dl), getI32Imm(63 - Shift, dl));
+      return;
+    }
+    break;
+  }
   // FIXME: Remove this once the ANDI glue bug is fixed:
   case PPCISD::ANDI_rec_1_EQ_BIT:
   case PPCISD::ANDI_rec_1_GT_BIT: {
diff --git a/llvm/test/CodeGen/PowerPC/mulli.ll b/llvm/test/CodeGen/PowerPC/mulli.ll
--- a/llvm/test/CodeGen/PowerPC/mulli.ll
+++ b/llvm/test/CodeGen/PowerPC/mulli.ll
@@ -4,9 +4,8 @@
 define i64 @test1(i64 %x) {
 ; CHECK-LABEL: test1:
 ; CHECK:       # %bb.0:
-; CHECK-NEXT:    li 4, 625
-; CHECK-NEXT:    sldi 4, 4, 36
-; CHECK-NEXT:    mulld 3, 3, 4
+; CHECK-NEXT:    mulli 3, 3, 625
+; CHECK-NEXT:    sldi 3, 3, 36
 ; CHECK-NEXT:    blr
   %y = mul i64 %x, 42949672960000
   ret i64 %y
@@ -15,9 +14,8 @@
 define i64 @test2(i64 %x) {
 ; CHECK-LABEL: test2:
 ; CHECK:       # %bb.0:
-; CHECK-NEXT:    li 4, -625
-; CHECK-NEXT:    sldi 4, 4, 36
-; CHECK-NEXT:    mulld 3, 3, 4
+; CHECK-NEXT:    mulli 3, 3, -625
+; CHECK-NEXT:    sldi 3, 3, 36
 ; CHECK-NEXT:    blr
   %y = mul i64 %x, -42949672960000
   ret i64 %y
@@ -26,9 +24,8 @@
 define i64 @test3(i64 %x) {
 ; CHECK-LABEL: test3:
 ; CHECK:       # %bb.0:
-; CHECK-NEXT:    lis 4, 74
-; CHECK-NEXT:    ori 4, 4, 16384
-; CHECK-NEXT:    mulld 3, 3, 4
+; CHECK-NEXT:    mulli 3, 3, 297
+; CHECK-NEXT:    sldi 3, 3, 14
 ; CHECK-NEXT:    blr
   %y = mul i64 %x, 4866048
   ret i64 %y
@@ -37,9 +34,8 @@
 define i64 @test4(i64 %x) {
 ; CHECK-LABEL: test4:
 ; CHECK:       # %bb.0:
-; CHECK-NEXT:    lis 4, -75
-; CHECK-NEXT:    ori 4, 4, 49152
-; CHECK-NEXT:    mulld 3, 3, 4
+; CHECK-NEXT:    mulli 3, 3, -297
+; CHECK-NEXT:    sldi 3, 3, 14
 ; CHECK-NEXT:    blr
   %y = mul i64 %x, -4866048
   ret i64 %y
@@ -48,10 +44,9 @@
 define i64 @test5(i64 %x) {
 ; CHECK-LABEL: test5:
 ; CHECK:       # %bb.0:
-; CHECK-NEXT:    lis 4, 16
-; CHECK-NEXT:    ori 4, 4, 1
-; CHECK-NEXT:    sldi 4, 4, 12
-; CHECK-NEXT:    mulld 3, 3, 4
+; CHECK-NEXT:    sldi 4, 3, 20
+; CHECK-NEXT:    add 3, 3, 4
+; CHECK-NEXT:    sldi 3, 3, 12
 ; CHECK-NEXT:    blr
   %y = mul i64 %x, 4294971392
   ret i64 %y
@@ -68,3 +63,15 @@
   %y = mul i64 %x, -4294971392
   ret i64 %y
 }
+
+define i64 @test7(i64 %x) {
+; CHECK-LABEL: test7:
+; CHECK:       # %bb.0:
+; CHECK-NEXT:    sldi 4, 3, 21
+; CHECK-NEXT:    sub 3, 4, 3
+; CHECK-NEXT:    sldi 3, 3, 13
+; CHECK-NEXT:    blr
+  %y = mul i64 %x, 17179860992
+  ret i64 %y
+}
+