diff --git a/llvm/lib/Transforms/InstCombine/InstCombineMulDivRem.cpp b/llvm/lib/Transforms/InstCombine/InstCombineMulDivRem.cpp
--- a/llvm/lib/Transforms/InstCombine/InstCombineMulDivRem.cpp
+++ b/llvm/lib/Transforms/InstCombine/InstCombineMulDivRem.cpp
@@ -1701,87 +1701,138 @@
 // Variety of transform for (urem/srem (mul/shl X, Y), (mul/shl X, Z))
 static Instruction *simplifyIRemMulShl(BinaryOperator &I,
                                        InstCombinerImpl &IC) {
-  Value *Op0 = I.getOperand(0), *Op1 = I.getOperand(1), *X;
-  const APInt *Y, *Z;
+  Value *Op0 = I.getOperand(0), *Op1 = I.getOperand(1), *X, *Y, *Z;
   bool ShiftX = false, ShiftY = false, ShiftZ = false;
-  if ((match(Op0, m_Mul(m_Value(X), m_APInt(Y))) &&
-       match(Op1, m_c_Mul(m_Specific(X), m_APInt(Z)))) ||
-      (match(Op0, m_Mul(m_APInt(Y), m_Value(X))) &&
-       match(Op1, m_c_Mul(m_Specific(X), m_APInt(Z))))) {
+  if ((match(Op0, m_Mul(m_Value(X), m_Value(Y))) &&
+       match(Op1, m_c_Mul(m_Specific(X), m_Value(Z)))) ||
+      (match(Op0, m_Mul(m_Value(Y), m_Value(X))) &&
+       match(Op1, m_c_Mul(m_Specific(X), m_Value(Z))))) {
     // Pass
-  } else if (match(Op0, m_Shl(m_Value(X), m_APInt(Y))) &&
-             match(Op1, m_c_Mul(m_Specific(X), m_APInt(Z)))) {
+  } else if (match(Op0, m_Shl(m_Value(X), m_Value(Y))) &&
+             match(Op1, m_c_Mul(m_Specific(X), m_Value(Z)))) {
     ShiftY = true;
-  } else if (match(Op1, m_Shl(m_Value(X), m_APInt(Z))) &&
-             match(Op0, m_c_Mul(m_Specific(X), m_APInt(Y)))) {
+  } else if (match(Op1, m_Shl(m_Value(X), m_Value(Z))) &&
+             match(Op0, m_c_Mul(m_Specific(X), m_Value(Y)))) {
     ShiftZ = true;
-  } else if (match(Op0, m_Shl(m_Value(X), m_APInt(Y))) &&
-             match(Op1, m_Shl(m_Specific(X), m_APInt(Z)))) {
+  } else if (match(Op0, m_Shl(m_Value(X), m_Value(Y))) &&
+             match(Op1, m_Shl(m_Specific(X), m_Value(Z)))) {
     ShiftZ = true;
     ShiftY = true;
-  } else if (match(Op0, m_Shl(m_APInt(Y), m_Value(X))) &&
-             match(Op1, m_Shl(m_APInt(Z), m_Specific(X)))) {
+  } else if (match(Op0, m_Shl(m_Value(Y), m_Value(X))) &&
+             match(Op1, m_Shl(m_Value(Z), m_Specific(X)))) {
     ShiftX = true;
   } else {
     return nullptr;
   }
 
-  APInt AdjustedY = *Y;
-  APInt AdjustedZ = *Z;
-  // Just treat the shifts as mul, we may end up returning a mul by power
-  // of 2 but that will be cleaned up later.
-  if (ShiftY)
-    AdjustedY = APInt(AdjustedY.getBitWidth(), 1) << AdjustedY;
-  if (ShiftZ)
-    AdjustedZ = APInt(AdjustedZ.getBitWidth(), 1) << AdjustedZ;
+  OverflowingBinaryOperator *BO0 = cast<OverflowingBinaryOperator>(Op0);
+  OverflowingBinaryOperator *BO1 = cast<OverflowingBinaryOperator>(Op1);
+
+  // If X is constant 1, then we avoid both in the mul and shl case.
+  auto CX = dyn_cast<Constant>(X);
+  if (CX && CX->isOneValue())
+    return nullptr;
+
+  auto GetOperandAsConstantInt = [](Value *Op) -> ConstantInt * {
+    if (Op->getType()->isVectorTy())
+      if (auto *COp = dyn_cast<Constant>(Op)) {
+        auto *CSplat = COp->getSplatValue();
+        return CSplat ? dyn_cast<ConstantInt>(CSplat) : nullptr;
+      }
+    return dyn_cast<ConstantInt>(Op);
+  };
+
+  ConstantInt *ConstY = GetOperandAsConstantInt(Y);
+  ConstantInt *ConstZ = GetOperandAsConstantInt(Z);
 
   bool IsSRem = I.getOpcode() == Instruction::SRem;
 
-  OverflowingBinaryOperator *BO0 = cast<OverflowingBinaryOperator>(Op0);
   // TODO: We may be able to deduce more about nsw/nuw of BO0/BO1 based on Y >=
   // Z or Z >= Y.
   bool BO0HasNSW = BO0->hasNoSignedWrap();
   bool BO0HasNUW = BO0->hasNoUnsignedWrap();
-  bool BO0NoWrap = IsSRem ? BO0HasNSW : BO0HasNUW;
-
-  APInt RemYZ = IsSRem ? AdjustedY.srem(AdjustedZ) : AdjustedY.urem(AdjustedZ);
-  // (rem (mul nuw/nsw X, Y), (mul X, Z))
-  //      if (rem Y, Z) == 0
-  //          -> 0
-  if (RemYZ.isZero() && BO0NoWrap)
-    return IC.replaceInstUsesWith(I, ConstantInt::getNullValue(I.getType()));
+  bool BO1HasNSW = BO1->hasNoSignedWrap();
+  bool BO1HasNUW = BO1->hasNoUnsignedWrap();
 
   auto GetBinOpOut = [&](Value *RemSimplification) -> BinaryOperator * {
     return ShiftX ? BinaryOperator::CreateShl(RemSimplification, X)
                   : BinaryOperator::CreateMul(X, RemSimplification);
   };
 
-  OverflowingBinaryOperator *BO1 = cast<OverflowingBinaryOperator>(Op1);
-  bool BO1HasNSW = BO1->hasNoSignedWrap();
-  bool BO1HasNUW = BO1->hasNoUnsignedWrap();
-  bool BO1NoWrap = IsSRem ? BO1HasNSW : BO1HasNUW;
-  // (rem (mul X, Y), (mul nuw/nsw X, Z))
-  //      if (rem Y, Z) == Y
-  //          -> (mul nuw/nsw X, Y)
-  if (RemYZ == AdjustedY && BO1NoWrap) {
-    BinaryOperator *BO = GetBinOpOut(ConstantInt::get(I.getType(), AdjustedY));
-    // Copy any overflow flags from Op0.
-    if (IsSRem || BO0HasNSW)
+  if (ConstY && ConstZ) {
+    APInt AdjustedY = ConstY->getValue();
+    APInt AdjustedZ = ConstZ->getValue();
+
+    // Just treat the shifts as mul, we may end up returning a mul by power
+    // of 2 but that will be cleaned up later.
+    if (ShiftY)
+      AdjustedY = APInt(AdjustedY.getBitWidth(), 1) << AdjustedY;
+    if (ShiftZ)
+      AdjustedZ = APInt(AdjustedZ.getBitWidth(), 1) << AdjustedZ;
+
+    bool BO0NoWrap = IsSRem ? BO0HasNSW : BO0HasNUW;
+
+    APInt RemYZ =
+        IsSRem ? AdjustedY.srem(AdjustedZ) : AdjustedY.urem(AdjustedZ);
+    // (rem (mul nuw/nsw X, Y), (mul X, Z))
+    //      if (rem Y, Z) == 0
+    //          -> 0
+    if (RemYZ.isZero() && BO0NoWrap)
+      return IC.replaceInstUsesWith(I, ConstantInt::getNullValue(I.getType()));
+
+    bool BO1NoWrap = IsSRem ? BO1HasNSW : BO1HasNUW;
+    // (rem (mul X, Y), (mul nuw/nsw X, Z))
+    //      if (rem Y, Z) == Y
+    //          -> (mul nuw/nsw X, Y)
+    if (RemYZ == AdjustedY && BO1NoWrap) {
+      BinaryOperator *BO =
+          GetBinOpOut(ConstantInt::get(I.getType(), AdjustedY));
+      // Copy any overflow flags from Op0.
+      if (IsSRem || BO0HasNSW)
+        BO->setHasNoSignedWrap();
+      if (!IsSRem || BO0HasNUW)
+        BO->setHasNoUnsignedWrap();
+      return BO;
+    }
+
+    // (rem (mul nuw/nsw X, Y), (mul {nsw} X, Z))
+    //      if Y >= Z
+    //          -> (mul {nuw} nsw X, (rem Y, Z))
+    if (AdjustedY.uge(AdjustedZ) &&
+        (IsSRem ? (BO0HasNSW && BO1HasNSW) : BO0HasNUW)) {
+      BinaryOperator *BO = GetBinOpOut(ConstantInt::get(I.getType(), RemYZ));
       BO->setHasNoSignedWrap();
-    if (!IsSRem || BO0HasNUW)
-      BO->setHasNoUnsignedWrap();
-    return BO;
-  }
-
-  // (rem (mul nuw/nsw X, Y), (mul {nsw} X, Z))
-  //      if Y >= Z
-  //          -> (mul {nuw} nsw X, (rem Y, Z))
-  if (AdjustedY.uge(AdjustedZ) &&
-      (IsSRem ? (BO0HasNSW && BO1HasNSW) : BO0HasNUW)) {
-    BinaryOperator *BO = GetBinOpOut(ConstantInt::get(I.getType(), RemYZ));
-    BO->setHasNoSignedWrap();
-    BO->setHasNoUnsignedWrap(BO0HasNUW);
-    return BO;
+      BO->setHasNoUnsignedWrap(BO0HasNUW);
+      return BO;
+    }
+  }
+  // Check if desirable to do generic replacement.
+  // NB: It may be beneficial to do this if we have X << Z even if there are
+  // multiple uses of Op0/Op1 as it will eliminate the urem (urem of a power
+  // of 2 is converted to add/and) and urem is pretty expensive (maybe more
+  // sense in DAGCombiner).
+  if ((ConstY && ConstZ) ||
+      (Op0->hasOneUse() && Op1->hasOneUse() &&
+       (IsSRem ? (!ShiftY && !ShiftZ) : (!ShiftY || ShiftZ)))) {
+
+    // (rem (mul nuw/nsw X, Y), (mul nuw {nsw} X, Z)
+    //        -> (mul nuw/nsw X, (rem Y, Z))
+    if (IsSRem ? (BO0HasNSW && BO1HasNSW && BO1HasNUW)
+               : (BO0HasNUW && BO1HasNUW)) {
+      if (ShiftY)
+        Y = IC.Builder.CreateShl(ConstantInt::get(I.getType(), 1), Y);
+      if (ShiftZ)
+        Z = IC.Builder.CreateShl(ConstantInt::get(I.getType(), 1), Z);
+
+      BinaryOperator *BO = GetBinOpOut(IsSRem ? IC.Builder.CreateSRem(Y, Z)
+                                              : IC.Builder.CreateURem(Y, Z));
+
+      if (BO0HasNSW || BO1HasNSW)
+        BO->setHasNoSignedWrap();
+      if (!IsSRem || (BO0HasNUW && BO1HasNUW))
+        BO->setHasNoUnsignedWrap();
+      return BO;
+    }
   }
 
   return nullptr;
diff --git a/llvm/test/Transforms/InstCombine/rem-mul-shl.ll b/llvm/test/Transforms/InstCombine/rem-mul-shl.ll
--- a/llvm/test/Transforms/InstCombine/rem-mul-shl.ll
+++ b/llvm/test/Transforms/InstCombine/rem-mul-shl.ll
@@ -134,9 +134,8 @@
 
 define i8 @urem_XY_XZ_with_Y_Z_is_mul_X_RemYZ(i8 %X, i8 %Y, i8 %Z) {
 ; CHECK-LABEL: @urem_XY_XZ_with_Y_Z_is_mul_X_RemYZ(
-; CHECK-NEXT:    [[BO0:%.*]] = mul nuw i8 [[X:%.*]], [[Y:%.*]]
-; CHECK-NEXT:    [[BO1:%.*]] = mul nuw i8 [[Z:%.*]], [[X]]
-; CHECK-NEXT:    [[R:%.*]] = urem i8 [[BO0]], [[BO1]]
+; CHECK-NEXT:    [[TMP1:%.*]] = urem i8 [[Y:%.*]], [[Z:%.*]]
+; CHECK-NEXT:    [[R:%.*]] = mul nuw i8 [[TMP1]], [[X:%.*]]
 ; CHECK-NEXT:    ret i8 [[R]]
 ;
   %BO0 = mul nuw i8 %X, %Y
@@ -147,9 +146,10 @@
 
 define i8 @urem_XY_XZ_with_CX_Y_Z_is_mul_X_RemYZ(i8 %Y, i8 %Z) {
 ; CHECK-LABEL: @urem_XY_XZ_with_CX_Y_Z_is_mul_X_RemYZ(
-; CHECK-NEXT:    [[BO0:%.*]] = mul nuw i8 [[Y:%.*]], 10
-; CHECK-NEXT:    [[BO1:%.*]] = shl nuw i8 10, [[Z:%.*]]
-; CHECK-NEXT:    [[R:%.*]] = urem i8 [[BO0]], [[BO1]]
+; CHECK-NEXT:    [[NOTMASK:%.*]] = shl nsw i8 -1, [[Z:%.*]]
+; CHECK-NEXT:    [[TMP1:%.*]] = xor i8 [[NOTMASK]], -1
+; CHECK-NEXT:    [[TMP2:%.*]] = and i8 [[TMP1]], [[Y:%.*]]
+; CHECK-NEXT:    [[R:%.*]] = mul nuw i8 [[TMP2]], 10
 ; CHECK-NEXT:    ret i8 [[R]]
 ;
   %BO0 = mul nuw i8 10, %Y
@@ -160,9 +160,10 @@
 
 define i8 @urem_XY_XZ_with_Y_Z_is_mul_X_RemYZ_with_nsw_out1(i8 %X, i8 %Y, i8 %Z) {
 ; CHECK-LABEL: @urem_XY_XZ_with_Y_Z_is_mul_X_RemYZ_with_nsw_out1(
-; CHECK-NEXT:    [[BO0:%.*]] = mul nuw nsw i8 [[X:%.*]], [[Y:%.*]]
-; CHECK-NEXT:    [[BO1:%.*]] = shl nuw i8 [[X]], [[Z:%.*]]
-; CHECK-NEXT:    [[R:%.*]] = urem i8 [[BO0]], [[BO1]]
+; CHECK-NEXT:    [[NOTMASK:%.*]] = shl nsw i8 -1, [[Z:%.*]]
+; CHECK-NEXT:    [[TMP1:%.*]] = xor i8 [[NOTMASK]], -1
+; CHECK-NEXT:    [[TMP2:%.*]] = and i8 [[TMP1]], [[Y:%.*]]
+; CHECK-NEXT:    [[R:%.*]] = mul nuw nsw i8 [[TMP2]], [[X:%.*]]
 ; CHECK-NEXT:    ret i8 [[R]]
 ;
   %BO0 = mul nuw nsw i8 %X, %Y
@@ -173,9 +174,8 @@
 
 define <2 x i8> @urem_XY_XZ_with_Y_Z_is_mul_X_RemYZ_with_nsw_out2(<2 x i8> %X, <2 x i8> %Y, <2 x i8> %Z) {
 ; CHECK-LABEL: @urem_XY_XZ_with_Y_Z_is_mul_X_RemYZ_with_nsw_out2(
-; CHECK-NEXT:    [[BO0:%.*]] = shl nuw <2 x i8> [[Y:%.*]], [[X:%.*]]
-; CHECK-NEXT:    [[BO1:%.*]] = shl nuw nsw <2 x i8> [[Z:%.*]], [[X]]
-; CHECK-NEXT:    [[R:%.*]] = urem <2 x i8> [[BO0]], [[BO1]]
+; CHECK-NEXT:    [[TMP1:%.*]] = urem <2 x i8> [[Y:%.*]], [[Z:%.*]]
+; CHECK-NEXT:    [[R:%.*]] = shl nuw nsw <2 x i8> [[TMP1]], [[X:%.*]]
 ; CHECK-NEXT:    ret <2 x i8> [[R]]
 ;
   %BO0 = shl nuw <2 x i8> %Y, %X
@@ -366,9 +366,8 @@
 
 define i8 @srem_XY_XZ_with_Y_Z_is_mul_X_RemYZ(i8 %X, i8 %Y, i8 %Z) {
 ; CHECK-LABEL: @srem_XY_XZ_with_Y_Z_is_mul_X_RemYZ(
-; CHECK-NEXT:    [[BO0:%.*]] = mul nsw i8 [[Y:%.*]], [[X:%.*]]
-; CHECK-NEXT:    [[BO1:%.*]] = mul nuw nsw i8 [[X]], [[Z:%.*]]
-; CHECK-NEXT:    [[R:%.*]] = srem i8 [[BO0]], [[BO1]]
+; CHECK-NEXT:    [[TMP1:%.*]] = srem i8 [[Y:%.*]], [[Z:%.*]]
+; CHECK-NEXT:    [[R:%.*]] = mul nsw i8 [[TMP1]], [[X:%.*]]
 ; CHECK-NEXT:    ret i8 [[R]]
 ;
   %BO0 = mul nsw i8 %Y, %X
@@ -379,9 +378,8 @@
 
 define i8 @srem_XY_XZ_with_Y_Z_is_mul_X_RemYZ_with_nuw_out(i8 %X, i8 %Y, i8 %Z) {
 ; CHECK-LABEL: @srem_XY_XZ_with_Y_Z_is_mul_X_RemYZ_with_nuw_out(
-; CHECK-NEXT:    [[BO0:%.*]] = mul nuw nsw i8 [[Y:%.*]], [[X:%.*]]
-; CHECK-NEXT:    [[BO1:%.*]] = mul nuw nsw i8 [[Z:%.*]], [[X]]
-; CHECK-NEXT:    [[R:%.*]] = srem i8 [[BO0]], [[BO1]]
+; CHECK-NEXT:    [[TMP1:%.*]] = srem i8 [[Y:%.*]], [[Z:%.*]]
+; CHECK-NEXT:    [[R:%.*]] = mul nuw nsw i8 [[TMP1]], [[X:%.*]]
 ; CHECK-NEXT:    ret i8 [[R]]
 ;
   %BO0 = mul nsw nuw i8 %Y, %X
@@ -614,9 +612,8 @@
 
 define i8 @urem_shl_XX_shl_ZX(i8 %X, i8 %Z) {
 ; CHECK-LABEL: @urem_shl_XX_shl_ZX(
-; CHECK-NEXT:    [[BO0:%.*]] = shl nuw nsw i8 [[X:%.*]], [[X]]
-; CHECK-NEXT:    [[BO1:%.*]] = shl nuw nsw i8 [[Z:%.*]], [[X]]
-; CHECK-NEXT:    [[R:%.*]] = urem i8 [[BO0]], [[BO1]]
+; CHECK-NEXT:    [[TMP1:%.*]] = urem i8 [[X:%.*]], [[Z:%.*]]
+; CHECK-NEXT:    [[R:%.*]] = shl nuw nsw i8 [[TMP1]], [[X]]
 ; CHECK-NEXT:    ret i8 [[R]]
 ;
   %BO0 = shl nuw nsw i8 %X, %X
@@ -627,9 +624,8 @@
 
 define i8 @urem_shl_YX_shl_XX(i8 %X, i8 %Y) {
 ; CHECK-LABEL: @urem_shl_YX_shl_XX(
-; CHECK-NEXT:    [[BO0:%.*]] = shl nuw nsw i8 [[Y:%.*]], [[X:%.*]]
-; CHECK-NEXT:    [[BO1:%.*]] = shl nuw nsw i8 [[X]], [[X]]
-; CHECK-NEXT:    [[R:%.*]] = urem i8 [[BO0]], [[BO1]]
+; CHECK-NEXT:    [[TMP1:%.*]] = urem i8 [[Y:%.*]], [[X:%.*]]
+; CHECK-NEXT:    [[R:%.*]] = shl nuw nsw i8 [[TMP1]], [[X]]
 ; CHECK-NEXT:    ret i8 [[R]]
 ;
   %BO0 = shl nuw nsw i8 %Y, %X
@@ -640,9 +636,11 @@
 
 define i8 @urem_shl_XX_shl_XZ(i8 %X, i8 %Z) {
 ; CHECK-LABEL: @urem_shl_XX_shl_XZ(
-; CHECK-NEXT:    [[BO0:%.*]] = shl nuw nsw i8 [[X:%.*]], [[X]]
-; CHECK-NEXT:    [[BO1:%.*]] = shl nuw nsw i8 [[X]], [[Z:%.*]]
-; CHECK-NEXT:    [[R:%.*]] = urem i8 [[BO0]], [[BO1]]
+; CHECK-NEXT:    [[TMP1:%.*]] = shl nuw i8 1, [[X:%.*]]
+; CHECK-NEXT:    [[NOTMASK:%.*]] = shl nsw i8 -1, [[Z:%.*]]
+; CHECK-NEXT:    [[TMP2:%.*]] = xor i8 [[NOTMASK]], -1
+; CHECK-NEXT:    [[TMP3:%.*]] = and i8 [[TMP1]], [[TMP2]]
+; CHECK-NEXT:    [[R:%.*]] = mul nuw nsw i8 [[TMP3]], [[X]]
 ; CHECK-NEXT:    ret i8 [[R]]
 ;
   %BO0 = shl nuw nsw i8 %X, %X
@@ -653,9 +651,11 @@
 
 define i8 @urem_shl_XY_shl_XX(i8 %X, i8 %Y) {
 ; CHECK-LABEL: @urem_shl_XY_shl_XX(
-; CHECK-NEXT:    [[BO0:%.*]] = shl nuw nsw i8 [[X:%.*]], [[Y:%.*]]
-; CHECK-NEXT:    [[BO1:%.*]] = shl nuw nsw i8 [[X]], [[X]]
-; CHECK-NEXT:    [[R:%.*]] = urem i8 [[BO0]], [[BO1]]
+; CHECK-NEXT:    [[TMP1:%.*]] = shl nuw i8 1, [[Y:%.*]]
+; CHECK-NEXT:    [[NOTMASK:%.*]] = shl nsw i8 -1, [[X:%.*]]
+; CHECK-NEXT:    [[TMP2:%.*]] = xor i8 [[NOTMASK]], -1
+; CHECK-NEXT:    [[TMP3:%.*]] = and i8 [[TMP1]], [[TMP2]]
+; CHECK-NEXT:    [[R:%.*]] = mul nuw nsw i8 [[TMP3]], [[X]]
 ; CHECK-NEXT:    ret i8 [[R]]
 ;
   %BO0 = shl nuw nsw i8 %X, %Y
@@ -679,9 +679,10 @@
 
 define i8 @urem_mul_YX_shl_XX(i8 %X, i8 %Y) {
 ; CHECK-LABEL: @urem_mul_YX_shl_XX(
-; CHECK-NEXT:    [[BO0:%.*]] = mul nuw nsw i8 [[Y:%.*]], [[X:%.*]]
-; CHECK-NEXT:    [[BO1:%.*]] = shl nuw nsw i8 [[X]], [[X]]
-; CHECK-NEXT:    [[R:%.*]] = urem i8 [[BO0]], [[BO1]]
+; CHECK-NEXT:    [[NOTMASK:%.*]] = shl nsw i8 -1, [[X:%.*]]
+; CHECK-NEXT:    [[TMP1:%.*]] = xor i8 [[NOTMASK]], -1
+; CHECK-NEXT:    [[TMP2:%.*]] = and i8 [[TMP1]], [[Y:%.*]]
+; CHECK-NEXT:    [[R:%.*]] = mul nuw nsw i8 [[TMP2]], [[X]]
 ; CHECK-NEXT:    ret i8 [[R]]
 ;
   %BO0 = mul nuw nsw i8 %Y, %X
@@ -692,9 +693,10 @@
 
 define i8 @urem_mul_XX_shl_XZ(i8 %X, i8 %Z) {
 ; CHECK-LABEL: @urem_mul_XX_shl_XZ(
-; CHECK-NEXT:    [[BO0:%.*]] = mul nuw nsw i8 [[X:%.*]], [[X]]
-; CHECK-NEXT:    [[BO1:%.*]] = shl nuw nsw i8 [[X]], [[Z:%.*]]
-; CHECK-NEXT:    [[R:%.*]] = urem i8 [[BO0]], [[BO1]]
+; CHECK-NEXT:    [[NOTMASK:%.*]] = shl nsw i8 -1, [[Z:%.*]]
+; CHECK-NEXT:    [[TMP1:%.*]] = xor i8 [[NOTMASK]], -1
+; CHECK-NEXT:    [[TMP2:%.*]] = and i8 [[TMP1]], [[X:%.*]]
+; CHECK-NEXT:    [[R:%.*]] = mul nuw nsw i8 [[TMP2]], [[X]]
 ; CHECK-NEXT:    ret i8 [[R]]
 ;
   %BO0 = mul nuw nsw i8 %X, %X
@@ -705,9 +707,10 @@
 
 define i8 @urem_mul_XY_shl_XX(i8 %X, i8 %Y) {
 ; CHECK-LABEL: @urem_mul_XY_shl_XX(
-; CHECK-NEXT:    [[BO0:%.*]] = mul nuw nsw i8 [[X:%.*]], [[Y:%.*]]
-; CHECK-NEXT:    [[BO1:%.*]] = shl nuw nsw i8 [[X]], [[X]]
-; CHECK-NEXT:    [[R:%.*]] = urem i8 [[BO0]], [[BO1]]
+; CHECK-NEXT:    [[NOTMASK:%.*]] = shl nsw i8 -1, [[X:%.*]]
+; CHECK-NEXT:    [[TMP1:%.*]] = xor i8 [[NOTMASK]], -1
+; CHECK-NEXT:    [[TMP2:%.*]] = and i8 [[TMP1]], [[Y:%.*]]
+; CHECK-NEXT:    [[R:%.*]] = mul nuw nsw i8 [[TMP2]], [[X]]
 ; CHECK-NEXT:    ret i8 [[R]]
 ;
   %BO0 = mul nuw nsw i8 %X, %Y
@@ -770,9 +773,10 @@
 
 define i8 @urem_mul_XX_shl_XX(i8 %X) {
 ; CHECK-LABEL: @urem_mul_XX_shl_XX(
-; CHECK-NEXT:    [[BO0:%.*]] = mul nuw nsw i8 [[X:%.*]], [[X]]
-; CHECK-NEXT:    [[BO1:%.*]] = shl nuw nsw i8 [[X]], [[X]]
-; CHECK-NEXT:    [[R:%.*]] = urem i8 [[BO0]], [[BO1]]
+; CHECK-NEXT:    [[NOTMASK:%.*]] = shl nsw i8 -1, [[X:%.*]]
+; CHECK-NEXT:    [[TMP1:%.*]] = xor i8 [[NOTMASK]], -1
+; CHECK-NEXT:    [[TMP2:%.*]] = and i8 [[TMP1]], [[X]]
+; CHECK-NEXT:    [[R:%.*]] = mul nuw nsw i8 [[TMP2]], [[X]]
 ; CHECK-NEXT:    ret i8 [[R]]
 ;
   %BO0 = mul nuw nsw i8 %X, %X