diff --git a/llvm/include/llvm/CodeGen/TargetLowering.h b/llvm/include/llvm/CodeGen/TargetLowering.h
--- a/llvm/include/llvm/CodeGen/TargetLowering.h
+++ b/llvm/include/llvm/CodeGen/TargetLowering.h
@@ -4681,6 +4681,13 @@
       EVT SCCVT, SDValue N0, SDValue N1C, ISD::CondCode Cond,
       DAGCombinerInfo &DCI, const SDLoc &DL) const;
 
+  // Optimize SETCC comparing expanded SRL/SHL with zero by performing following
+  // transformations on SETCC's LHS:
+  // (or (or (srl X, C0), (shl Y, C1)), (srl Y, C0)) -->  (or (srl X, C0), Y)
+  // (or (or (srl Y, C0), (shl X, C1)), (shl Y, C1)) -->  (or (shl X, C1), Y)
+  SDValue optimizeSetCCOfExpandedShift(EVT SCCVT, SDValue N0, SDValue N1C, ISD::CondCode Cond,
+				       DAGCombinerInfo &DCI, const SDLoc &DL) const;
+
   SDValue prepareUREMEqFold(EVT SETCCVT, SDValue REMNode,
                             SDValue CompTargetNode, ISD::CondCode Cond,
                             DAGCombinerInfo &DCI, const SDLoc &DL,
diff --git a/llvm/lib/CodeGen/SelectionDAG/TargetLowering.cpp b/llvm/lib/CodeGen/SelectionDAG/TargetLowering.cpp
--- a/llvm/lib/CodeGen/SelectionDAG/TargetLowering.cpp
+++ b/llvm/lib/CodeGen/SelectionDAG/TargetLowering.cpp
@@ -3390,6 +3390,71 @@
   return T2;
 }
 
+// (or (or (srl X, C0), (shl Y, C1)), (srl Y, C0)) ==/!= 0
+//    -->  (or (srl X, C0), Y) ==/!= 0
+// (or (or (srl Y, C0), (shl X, C1)), (shl Y, C1)) ==/!= 0
+//    -->  (or (shl X, C1), Y) ==/!= 0
+SDValue TargetLowering::optimizeSetCCOfExpandedShift(EVT SCCVT, SDValue N0,
+                                                     SDValue N1C,
+                                                     ISD::CondCode Cond,
+                                                     DAGCombinerInfo &DCI,
+                                                     const SDLoc &DL) const {
+  assert(isConstOrConstSplat(N1C) &&
+         isConstOrConstSplat(N1C)->getAPIntValue().isZero() &&
+         "Should be a comparison with 0.");
+  assert((Cond == ISD::SETEQ || Cond == ISD::SETNE) && "Unexpected condcode");
+
+  if (N0.getOpcode() != ISD::OR || !N0.hasOneUse() ||
+      !N0.getOperand(0).hasOneUse() || !N0.getOperand(1).hasOneUse())
+    return SDValue();
+
+  auto Or = N0.getOperand(0);
+  auto Shift = N0.getOperand(1);
+  if (Or.getOpcode() != ISD::OR) {
+    if (Shift.getOpcode() != ISD::OR)
+      return SDValue();
+    std::swap(Or, Shift);
+  }
+  if (Shift.getOpcode() != ISD::SRL && Shift.getOpcode() != ISD::SHL)
+    return SDValue();
+
+  // Nested Or's SRL and SHL arms
+  auto OrSrl = Or.getOperand(0);
+  auto OrShl = Or.getOperand(1);
+  if (OrSrl.getOpcode() != ISD::SRL)
+    std::swap(OrSrl, OrShl);
+  if (OrSrl.getOpcode() != ISD::SRL || OrShl.getOpcode() != ISD::SHL)
+    return SDValue();
+
+  // Nested Or's arms shifting X and Y
+  auto OrShiftX = Shift.getOpcode() == ISD::SRL ? OrSrl : OrShl;
+  auto OrShiftY = Shift.getOpcode() == ISD::SRL ? OrShl : OrSrl;
+
+  if (OrShiftY.getOperand(0) == Shift.getOperand(0)) {
+    auto OpSizeInBits = N0.getValueType().getScalarSizeInBits();
+    auto MatchConstants = [OpSizeInBits](ConstantSDNode *LHS,
+                                         ConstantSDNode *RHS) {
+      return (LHS->getAPIntValue() + RHS->getAPIntValue()) == OpSizeInBits;
+    };
+    auto SameConstant = [](ConstantSDNode *LHS, ConstantSDNode *RHS) {
+      return LHS->getAPIntValue() == RHS->getAPIntValue();
+    };
+
+    auto C0 = OrSrl.getOperand(1);
+    auto C1 = OrShl.getOperand(1);
+    auto ShiftC = Shift.getOperand(1);
+    auto OrShiftXC = OrShiftX.getOperand(1);
+    if (ISD::matchBinaryPredicate(C0, C1, MatchConstants) &&
+        ISD::matchBinaryPredicate(OrShiftXC, ShiftC, SameConstant)) {
+      SelectionDAG &DAG = DCI.DAG;
+      auto NewOr = DAG.getNode(ISD::OR, DL, N0.getValueType(), OrShiftX,
+                               Shift.getOperand(0));
+      return DAG.getSetCC(DL, SCCVT, NewOr, N1C, Cond);
+    }
+  }
+  return SDValue();
+}
+
 /// Try to fold an equality comparison with a {add/sub/xor} binary operation as
 /// the 1st operand (N0). Callers are expected to swap the N0/N1 parameters to
 /// handle the commuted versions of these patterns.
@@ -4010,12 +4075,21 @@
     }
 
     if (Cond == ISD::SETEQ || Cond == ISD::SETNE) {
-      // (X & (C l>>/<< Y)) ==/!= 0  -->  ((X <</l>> Y) & C) ==/!= 0
-      if (C1.isZero())
+      if (C1.isZero()) {
+        // (X & (C l>>/<< Y)) ==/!= 0  -->  ((X <</l>> Y) & C) ==/!= 0
         if (SDValue CC = optimizeSetCCByHoistingAndByConstFromLogicalShift(
                 VT, N0, N1, Cond, DCI, dl))
           return CC;
 
+        // (or (or (srl X, C0), (shl Y, C1)), (srl Y, C0)) ==/!= 0
+        //    -->  (or (srl X, C0), Y) ==/!= 0
+        // (or (or (srl Y, C0), (shl X, C1)), (shl Y, C1)) ==/!= 0
+        //    -->  (or (shl X, C1), Y) ==/!= 0
+        if (SDValue CC =
+                optimizeSetCCOfExpandedShift(VT, N0, N1, Cond, DCI, dl))
+          return CC;
+      }
+
       // For all/any comparisons, replace or(x,shl(y,bw/2)) with and/or(x,y).
       // For example, when high 32-bits of i64 X are known clear:
       // all bits clear: (X | (Y<<32)) ==  0 --> (X | Y) ==  0
diff --git a/llvm/test/CodeGen/AArch64/arm64-icmp-shift-opt.ll b/llvm/test/CodeGen/AArch64/arm64-icmp-shift-opt.ll
new file mode 100644
--- /dev/null
+++ b/llvm/test/CodeGen/AArch64/arm64-icmp-shift-opt.ll
@@ -0,0 +1,139 @@
+; NOTE: Assertions have been autogenerated by utils/update_llc_test_checks.py
+; RUN: llc < %s -mtriple=arm64-eabi | FileCheck %s
+
+; Optimize expanded SRL/SHL used as an input of
+; SETCC comparing it with zero by removing rotation.
+;
+; See https://bugs.llvm.org/show_bug.cgi?id=50197
+define i128 @opt_setcc_lt_power_of_2(i128 %a) {
+; CHECK-LABEL: opt_setcc_lt_power_of_2:
+; CHECK:       // %bb.0:
+; CHECK-NEXT:  .LBB0_1: // %loop
+; CHECK-NEXT:    // =>This Inner Loop Header: Depth=1
+; CHECK-NEXT:    adds x0, x0, #1
+; CHECK-NEXT:    adcs x1, x1, xzr
+; CHECK-NEXT:    orr x8, x1, x0, lsr #60
+; CHECK-NEXT:    cbnz x8, .LBB0_1
+; CHECK-NEXT:  // %bb.2: // %exit
+; CHECK-NEXT:    ret
+  br label %loop
+
+loop:
+  %phi.a = phi i128 [ %a, %0 ], [ %inc, %loop ]
+  %inc = add i128 %phi.a, 1
+  %cmp = icmp ult i128 %inc, 1152921504606846976
+  br i1 %cmp, label %exit, label %loop
+
+exit:
+  ret i128 %inc
+}
+
+define i1 @opt_setcc_srl_eq_zero(i128 %a) {
+; CHECK-LABEL: opt_setcc_srl_eq_zero:
+; CHECK:       // %bb.0:
+; CHECK-NEXT:    orr x8, x1, x0, lsr #17
+; CHECK-NEXT:    cmp x8, #0
+; CHECK-NEXT:    cset w0, eq
+; CHECK-NEXT:    ret
+   %srl = lshr i128 %a, 17
+   %cmp = icmp eq i128 %srl, 0
+   ret i1 %cmp
+}
+
+define i1 @opt_setcc_srl_ne_zero(i128 %a) {
+; CHECK-LABEL: opt_setcc_srl_ne_zero:
+; CHECK:       // %bb.0:
+; CHECK-NEXT:    orr x8, x1, x0, lsr #17
+; CHECK-NEXT:    cmp x8, #0
+; CHECK-NEXT:    cset w0, ne
+; CHECK-NEXT:    ret
+   %srl = lshr i128 %a, 17
+   %cmp = icmp ne i128 %srl, 0
+   ret i1 %cmp
+}
+
+define i1 @opt_setcc_shl_eq_zero(i128 %a) {
+; CHECK-LABEL: opt_setcc_shl_eq_zero:
+; CHECK:       // %bb.0:
+; CHECK-NEXT:    orr x8, x0, x1, lsl #17
+; CHECK-NEXT:    cmp x8, #0
+; CHECK-NEXT:    cset w0, eq
+; CHECK-NEXT:    ret
+   %shl = shl i128 %a, 17
+   %cmp = icmp eq i128 %shl, 0
+   ret i1 %cmp
+}
+
+define i1 @opt_setcc_shl_ne_zero(i128 %a) {
+; CHECK-LABEL: opt_setcc_shl_ne_zero:
+; CHECK:       // %bb.0:
+; CHECK-NEXT:    orr x8, x0, x1, lsl #17
+; CHECK-NEXT:    cmp x8, #0
+; CHECK-NEXT:    cset w0, ne
+; CHECK-NEXT:    ret
+   %shl = shl i128 %a, 17
+   %cmp = icmp ne i128 %shl, 0
+   ret i1 %cmp
+}
+
+; Negative test: optimization should not be applied if shift has multiple users.
+define i1 @opt_setcc_shl_eq_zero_multiple_shl_users(i128 %a) {
+; CHECK-LABEL: opt_setcc_shl_eq_zero_multiple_shl_users:
+; CHECK:       // %bb.0:
+; CHECK-NEXT:    stp x30, x19, [sp, #-16]! // 16-byte Folded Spill
+; CHECK-NEXT:    .cfi_def_cfa_offset 16
+; CHECK-NEXT:    .cfi_offset w19, -8
+; CHECK-NEXT:    .cfi_offset w30, -16
+; CHECK-NEXT:    extr x1, x1, x0, #47
+; CHECK-NEXT:    lsl x0, x0, #17
+; CHECK-NEXT:    orr x8, x0, x1
+; CHECK-NEXT:    cmp x8, #0
+; CHECK-NEXT:    cset w19, eq
+; CHECK-NEXT:    bl use
+; CHECK-NEXT:    mov w0, w19
+; CHECK-NEXT:    ldp x30, x19, [sp], #16 // 16-byte Folded Reload
+; CHECK-NEXT:    ret
+   %shl = shl i128 %a, 17
+   %cmp = icmp eq i128 %shl, 0
+   call void @use(i128 %shl)
+   ret i1 %cmp
+}
+
+; Check that optimization is applied to DAG having appropriate shape
+; even if there were no actual shift's expansion.
+define i1 @opt_setcc_expanded_shl_correct_shifts(i64 %a, i64 %b) {
+; CHECK-LABEL: opt_setcc_expanded_shl_correct_shifts:
+; CHECK:       // %bb.0:
+; CHECK-NEXT:    orr x8, x1, x0, lsl #17
+; CHECK-NEXT:    cmp x8, #0
+; CHECK-NEXT:    cset w0, eq
+; CHECK-NEXT:    ret
+  %shl.a = shl i64 %a, 17
+  %srl.b = lshr i64 %b, 47
+  %or.0 = or i64 %shl.a, %srl.b
+  %shl.b = shl i64 %b, 17
+  %or.1 = or i64 %or.0, %shl.b
+  %cmp = icmp eq i64 %or.1, 0
+  ret i1 %cmp
+}
+
+; Negative test: optimization should not be applied as
+; constants used in shifts does not match.
+define i1 @opt_setcc_expanded_shl_wrong_shifts(i64 %a, i64 %b) {
+; CHECK-LABEL: opt_setcc_expanded_shl_wrong_shifts:
+; CHECK:       // %bb.0:
+; CHECK-NEXT:    extr x8, x0, x1, #47
+; CHECK-NEXT:    orr x8, x8, x1, lsl #18
+; CHECK-NEXT:    cmp x8, #0
+; CHECK-NEXT:    cset w0, eq
+; CHECK-NEXT:    ret
+  %shl.a = shl i64 %a, 17
+  %srl.b = lshr i64 %b, 47
+  %or.0 = or i64 %shl.a, %srl.b
+  %shl.b = shl i64 %b, 18
+  %or.1 = or i64 %or.0, %shl.b
+  %cmp = icmp eq i64 %or.1, 0
+  ret i1 %cmp
+}
+
+declare void @use(i128 %a);
diff --git a/llvm/test/CodeGen/ARM/arm-icmp-shift-opt.ll b/llvm/test/CodeGen/ARM/arm-icmp-shift-opt.ll
new file mode 100644
--- /dev/null
+++ b/llvm/test/CodeGen/ARM/arm-icmp-shift-opt.ll
@@ -0,0 +1,133 @@
+; NOTE: Assertions have been autogenerated by utils/update_llc_test_checks.py
+; RUN: llc -mtriple=armv7 %s -o - | FileCheck %s
+
+; Optimize expanded SRL/SHL used as an input of
+; SETCC comparing it with zero by removing rotation.
+;
+; See https://bugs.llvm.org/show_bug.cgi?id=50197
+define i64 @opt_setcc_lt_power_of_2(i64 %a) {
+; CHECK-LABEL: opt_setcc_lt_power_of_2:
+; CHECK:       @ %bb.0:
+; CHECK-NEXT:  .LBB0_1: @ %loop
+; CHECK-NEXT:    @ =>This Inner Loop Header: Depth=1
+; CHECK-NEXT:    adds r0, r0, #1
+; CHECK-NEXT:    adc r1, r1, #0
+; CHECK-NEXT:    orr r2, r1, r0, lsr #16
+; CHECK-NEXT:    cmp r2, #0
+; CHECK-NEXT:    bne .LBB0_1
+; CHECK-NEXT:  @ %bb.2: @ %exit
+; CHECK-NEXT:    bx lr
+  br label %loop
+
+loop:
+  %phi.a = phi i64 [ %a, %0 ], [ %inc, %loop ]
+  %inc = add i64 %phi.a, 1
+  %cmp = icmp ult i64 %inc, 65536
+  br i1 %cmp, label %exit, label %loop
+
+exit:
+  ret i64 %inc
+}
+
+define i1 @opt_setcc_srl_eq_zero(i64 %a) {
+; CHECK-LABEL: opt_setcc_srl_eq_zero:
+; CHECK:       @ %bb.0:
+; CHECK-NEXT:    orr r0, r1, r0, lsr #17
+; CHECK-NEXT:    clz r0, r0
+; CHECK-NEXT:    lsr r0, r0, #5
+; CHECK-NEXT:    bx lr
+   %srl = lshr i64 %a, 17
+   %cmp = icmp eq i64 %srl, 0
+   ret i1 %cmp
+}
+
+define i1 @opt_setcc_srl_ne_zero(i64 %a) {
+; CHECK-LABEL: opt_setcc_srl_ne_zero:
+; CHECK:       @ %bb.0:
+; CHECK-NEXT:    orr r0, r1, r0, lsr #17
+; CHECK-NEXT:    cmp r0, #0
+; CHECK-NEXT:    movwne r0, #1
+; CHECK-NEXT:    bx lr
+   %srl = lshr i64 %a, 17
+   %cmp = icmp ne i64 %srl, 0
+   ret i1 %cmp
+}
+
+define i1 @opt_setcc_shl_eq_zero(i64 %a) {
+; CHECK-LABEL: opt_setcc_shl_eq_zero:
+; CHECK:       @ %bb.0:
+; CHECK-NEXT:    orr r0, r0, r1, lsl #17
+; CHECK-NEXT:    clz r0, r0
+; CHECK-NEXT:    lsr r0, r0, #5
+; CHECK-NEXT:    bx lr
+   %shl = shl i64 %a, 17
+   %cmp = icmp eq i64 %shl, 0
+   ret i1 %cmp
+}
+
+define i1 @opt_setcc_shl_ne_zero(i64 %a) {
+; CHECK-LABEL: opt_setcc_shl_ne_zero:
+; CHECK:       @ %bb.0:
+; CHECK-NEXT:    orr r0, r0, r1, lsl #17
+; CHECK-NEXT:    cmp r0, #0
+; CHECK-NEXT:    movwne r0, #1
+; CHECK-NEXT:    bx lr
+   %shl = shl i64 %a, 17
+   %cmp = icmp ne i64 %shl, 0
+   ret i1 %cmp
+}
+
+; Negative test: optimization should not be applied if shift has multiple users.
+define i1 @opt_setcc_shl_eq_zero_multiple_shl_users(i64 %a) {
+; CHECK-LABEL: opt_setcc_shl_eq_zero_multiple_shl_users:
+; CHECK:       @ %bb.0:
+; CHECK-NEXT:    push {r4, r5, r11, lr}
+; CHECK-NEXT:    mov r4, r0
+; CHECK-NEXT:    lsl r0, r1, #17
+; CHECK-NEXT:    orr r5, r0, r4, lsr #15
+; CHECK-NEXT:    lsl r0, r4, #17
+; CHECK-NEXT:    mov r1, r5
+; CHECK-NEXT:    bl use
+; CHECK-NEXT:    orr r0, r5, r4, lsl #17
+; CHECK-NEXT:    clz r0, r0
+; CHECK-NEXT:    lsr r0, r0, #5
+; CHECK-NEXT:    pop {r4, r5, r11, pc}
+   %shl = shl i64 %a, 17
+   %cmp = icmp eq i64 %shl, 0
+   call void @use(i64 %shl)
+   ret i1 %cmp
+}
+
+; Check that optimization is applied to DAG having appropriate shape
+; even if there were no actual shift's expansion.
+define i1 @opt_setcc_expanded_shl_correct_shifts(i32 %a, i32 %b) {
+; CHECK-LABEL: opt_setcc_expanded_shl_correct_shifts:
+; CHECK:       @ %bb.0:
+; CHECK-NEXT:    mov r0, #0
+; CHECK-NEXT:    bx lr
+  %shl.a = shl i32 %a, 17
+  %srl.b = lshr i32 %b, 47
+  %or.0 = or i32 %shl.a, %srl.b
+  %shl.b = shl i32 %b, 17
+  %or.1 = or i32 %or.0, %shl.b
+  %cmp = icmp eq i32 %or.1, 0
+  ret i1 %cmp
+}
+
+; Negative test: optimization should not be applied as
+; constants used in shifts does not match.
+define i1 @opt_setcc_expanded_shl_wrong_shifts(i32 %a, i32 %b) {
+; CHECK-LABEL: opt_setcc_expanded_shl_wrong_shifts:
+; CHECK:       @ %bb.0:
+; CHECK-NEXT:    mov r0, #0
+; CHECK-NEXT:    bx lr
+  %shl.a = shl i32 %a, 17
+  %srl.b = lshr i32 %b, 47
+  %or.0 = or i32 %shl.a, %srl.b
+  %shl.b = shl i32 %b, 18
+  %or.1 = or i32 %or.0, %shl.b
+  %cmp = icmp eq i32 %or.1, 0
+  ret i1 %cmp
+}
+
+declare void @use(i64 %a);
diff --git a/llvm/test/CodeGen/ARM/consthoist-icmpimm.ll b/llvm/test/CodeGen/ARM/consthoist-icmpimm.ll
--- a/llvm/test/CodeGen/ARM/consthoist-icmpimm.ll
+++ b/llvm/test/CodeGen/ARM/consthoist-icmpimm.ll
@@ -630,14 +630,10 @@
 ; CHECKV7M-NEXT:    ldrd lr, r0, [sp, #8]
 ; CHECKV7M-NEXT:    beq .LBB6_2
 ; CHECKV7M-NEXT:  @ %bb.1: @ %then
-; CHECKV7M-NEXT:    lsrs r2, r2, #17
-; CHECKV7M-NEXT:    orr.w r2, r2, r3, lsl #15
-; CHECKV7M-NEXT:    orr.w r2, r2, r3, lsr #17
-; CHECKV7M-NEXT:    lsr.w r3, r12, #17
-; CHECKV7M-NEXT:    orr.w r3, r3, r1, lsl #15
+; CHECKV7M-NEXT:    orr.w r2, r3, r2, lsr #17
+; CHECKV7M-NEXT:    orr.w r1, r1, r12, lsr #17
 ; CHECKV7M-NEXT:    cmp r2, #0
 ; CHECKV7M-NEXT:    mov r2, r0
-; CHECKV7M-NEXT:    orr.w r1, r3, r1, lsr #17
 ; CHECKV7M-NEXT:    it ne
 ; CHECKV7M-NEXT:    movne r2, lr
 ; CHECKV7M-NEXT:    cmp r1, #0
@@ -646,9 +642,7 @@
 ; CHECKV7M-NEXT:    add r0, r2
 ; CHECKV7M-NEXT:    pop {r7, pc}
 ; CHECKV7M-NEXT:  .LBB6_2: @ %else
-; CHECKV7M-NEXT:    lsrs r1, r2, #17
-; CHECKV7M-NEXT:    orr.w r1, r1, r3, lsl #15
-; CHECKV7M-NEXT:    orr.w r1, r1, r3, lsr #17
+; CHECKV7M-NEXT:    orr.w r1, r3, r2, lsr #17
 ; CHECKV7M-NEXT:    cmp r1, #0
 ; CHECKV7M-NEXT:    it ne
 ; CHECKV7M-NEXT:    movne r0, lr
@@ -664,14 +658,10 @@
 ; CHECKV7A-NEXT:    lsls r4, r4, #31
 ; CHECKV7A-NEXT:    beq .LBB6_2
 ; CHECKV7A-NEXT:  @ %bb.1: @ %then
-; CHECKV7A-NEXT:    lsrs r2, r2, #17
-; CHECKV7A-NEXT:    orr.w r2, r2, r3, lsl #15
-; CHECKV7A-NEXT:    orr.w r2, r2, r3, lsr #17
-; CHECKV7A-NEXT:    lsr.w r3, r12, #17
-; CHECKV7A-NEXT:    orr.w r3, r3, r1, lsl #15
+; CHECKV7A-NEXT:    orr.w r2, r3, r2, lsr #17
+; CHECKV7A-NEXT:    orr.w r1, r1, r12, lsr #17
 ; CHECKV7A-NEXT:    cmp r2, #0
 ; CHECKV7A-NEXT:    mov r2, r0
-; CHECKV7A-NEXT:    orr.w r1, r3, r1, lsr #17
 ; CHECKV7A-NEXT:    it ne
 ; CHECKV7A-NEXT:    movne r2, lr
 ; CHECKV7A-NEXT:    cmp r1, #0
@@ -680,9 +670,7 @@
 ; CHECKV7A-NEXT:    add r0, r2
 ; CHECKV7A-NEXT:    pop {r4, pc}
 ; CHECKV7A-NEXT:  .LBB6_2: @ %else
-; CHECKV7A-NEXT:    lsrs r1, r2, #17
-; CHECKV7A-NEXT:    orr.w r1, r1, r3, lsl #15
-; CHECKV7A-NEXT:    orr.w r1, r1, r3, lsr #17
+; CHECKV7A-NEXT:    orr.w r1, r3, r2, lsr #17
 ; CHECKV7A-NEXT:    cmp r1, #0
 ; CHECKV7A-NEXT:    it ne
 ; CHECKV7A-NEXT:    movne r0, lr
diff --git a/llvm/test/CodeGen/X86/icmp-shift-opt.ll b/llvm/test/CodeGen/X86/icmp-shift-opt.ll
new file mode 100644
--- /dev/null
+++ b/llvm/test/CodeGen/X86/icmp-shift-opt.ll
@@ -0,0 +1,144 @@
+; NOTE: Assertions have been autogenerated by utils/update_llc_test_checks.py
+; RUN: llc -mtriple=x86_64-unknown-linux-gnu < %s | FileCheck %s
+
+; Optimize expanded SRL/SHL used as an input of
+; SETCC comparing it with zero by removing rotation.
+;
+; See https://bugs.llvm.org/show_bug.cgi?id=50197
+define i128 @opt_setcc_lt_power_of_2(i128 %a) {
+; CHECK-LABEL: opt_setcc_lt_power_of_2:
+; CHECK:       # %bb.0:
+; CHECK-NEXT:    movq %rsi, %rdx
+; CHECK-NEXT:    movq %rdi, %rax
+; CHECK-NEXT:    .p2align 4, 0x90
+; CHECK-NEXT:  .LBB0_1: # %loop
+; CHECK-NEXT:    # =>This Inner Loop Header: Depth=1
+; CHECK-NEXT:    addq $1, %rax
+; CHECK-NEXT:    adcq $0, %rdx
+; CHECK-NEXT:    movq %rax, %rcx
+; CHECK-NEXT:    shrq $60, %rcx
+; CHECK-NEXT:    orq %rdx, %rcx
+; CHECK-NEXT:    jne .LBB0_1
+; CHECK-NEXT:  # %bb.2: # %exit
+; CHECK-NEXT:    retq
+  br label %loop
+
+loop:
+  %phi.a = phi i128 [ %a, %0 ], [ %inc, %loop ]
+  %inc = add i128 %phi.a, 1
+  %cmp = icmp ult i128 %inc, 1152921504606846976
+  br i1 %cmp, label %exit, label %loop
+
+exit:
+  ret i128 %inc
+}
+
+define i1 @opt_setcc_srl_eq_zero(i128 %a) {
+; CHECK-LABEL: opt_setcc_srl_eq_zero:
+; CHECK:       # %bb.0:
+; CHECK-NEXT:    shrq $17, %rdi
+; CHECK-NEXT:    orq %rsi, %rdi
+; CHECK-NEXT:    sete %al
+; CHECK-NEXT:    retq
+   %srl = lshr i128 %a, 17
+   %cmp = icmp eq i128 %srl, 0
+   ret i1 %cmp
+}
+
+define i1 @opt_setcc_srl_ne_zero(i128 %a) {
+; CHECK-LABEL: opt_setcc_srl_ne_zero:
+; CHECK:       # %bb.0:
+; CHECK-NEXT:    shrq $17, %rdi
+; CHECK-NEXT:    orq %rsi, %rdi
+; CHECK-NEXT:    setne %al
+; CHECK-NEXT:    retq
+   %srl = lshr i128 %a, 17
+   %cmp = icmp ne i128 %srl, 0
+   ret i1 %cmp
+}
+
+define i1 @opt_setcc_shl_eq_zero(i128 %a) {
+; CHECK-LABEL: opt_setcc_shl_eq_zero:
+; CHECK:       # %bb.0:
+; CHECK-NEXT:    shlq $17, %rsi
+; CHECK-NEXT:    orq %rdi, %rsi
+; CHECK-NEXT:    sete %al
+; CHECK-NEXT:    retq
+   %shl = shl i128 %a, 17
+   %cmp = icmp eq i128 %shl, 0
+   ret i1 %cmp
+}
+
+define i1 @opt_setcc_shl_ne_zero(i128 %a) {
+; CHECK-LABEL: opt_setcc_shl_ne_zero:
+; CHECK:       # %bb.0:
+; CHECK-NEXT:    shlq $17, %rsi
+; CHECK-NEXT:    orq %rdi, %rsi
+; CHECK-NEXT:    setne %al
+; CHECK-NEXT:    retq
+   %shl = shl i128 %a, 17
+   %cmp = icmp ne i128 %shl, 0
+   ret i1 %cmp
+}
+
+; Negative test: optimization should not be applied if shift has multiple users.
+define i1 @opt_setcc_shl_eq_zero_multiple_shl_users(i128 %a) {
+; CHECK-LABEL: opt_setcc_shl_eq_zero_multiple_shl_users:
+; CHECK:       # %bb.0:
+; CHECK-NEXT:    pushq %rbx
+; CHECK-NEXT:    .cfi_def_cfa_offset 16
+; CHECK-NEXT:    .cfi_offset %rbx, -16
+; CHECK-NEXT:    shldq $17, %rdi, %rsi
+; CHECK-NEXT:    shlq $17, %rdi
+; CHECK-NEXT:    movq %rdi, %rax
+; CHECK-NEXT:    orq %rsi, %rax
+; CHECK-NEXT:    sete %bl
+; CHECK-NEXT:    callq use@PLT
+; CHECK-NEXT:    movl %ebx, %eax
+; CHECK-NEXT:    popq %rbx
+; CHECK-NEXT:    .cfi_def_cfa_offset 8
+; CHECK-NEXT:    retq
+   %shl = shl i128 %a, 17
+   %cmp = icmp eq i128 %shl, 0
+   call void @use(i128 %shl)
+   ret i1 %cmp
+}
+
+; Check that optimization is applied to DAG having appropriate shape
+; even if there were no actual shift's expansion.
+define i1 @opt_setcc_expanded_shl_correct_shifts(i64 %a, i64 %b) {
+; CHECK-LABEL: opt_setcc_expanded_shl_correct_shifts:
+; CHECK:       # %bb.0:
+; CHECK-NEXT:    shlq $17, %rdi
+; CHECK-NEXT:    orq %rsi, %rdi
+; CHECK-NEXT:    sete %al
+; CHECK-NEXT:    retq
+  %shl.a = shl i64 %a, 17
+  %srl.b = lshr i64 %b, 47
+  %or.0 = or i64 %shl.a, %srl.b
+  %shl.b = shl i64 %b, 17
+  %or.1 = or i64 %or.0, %shl.b
+  %cmp = icmp eq i64 %or.1, 0
+  ret i1 %cmp
+}
+
+; Negative test: optimization should not be applied as
+; constants used in shifts does not match.
+define i1 @opt_setcc_expanded_shl_wrong_shifts(i64 %a, i64 %b) {
+; CHECK-LABEL: opt_setcc_expanded_shl_wrong_shifts:
+; CHECK:       # %bb.0:
+; CHECK-NEXT:    shldq $17, %rsi, %rdi
+; CHECK-NEXT:    shlq $18, %rsi
+; CHECK-NEXT:    orq %rdi, %rsi
+; CHECK-NEXT:    sete %al
+; CHECK-NEXT:    retq
+  %shl.a = shl i64 %a, 17
+  %srl.b = lshr i64 %b, 47
+  %or.0 = or i64 %shl.a, %srl.b
+  %shl.b = shl i64 %b, 18
+  %or.1 = or i64 %or.0, %shl.b
+  %cmp = icmp eq i64 %or.1, 0
+  ret i1 %cmp
+}
+
+declare void @use(i128 %a);