diff --git a/llvm/lib/CodeGen/SelectionDAG/SelectionDAG.cpp b/llvm/lib/CodeGen/SelectionDAG/SelectionDAG.cpp
--- a/llvm/lib/CodeGen/SelectionDAG/SelectionDAG.cpp
+++ b/llvm/lib/CodeGen/SelectionDAG/SelectionDAG.cpp
@@ -3646,6 +3646,15 @@
     // All bits are zero except the low bit.
     Known.Zero.setBitsFrom(1);
     break;
+  case ISD::ADD:
+  case ISD::SUB: {
+    SDNodeFlags Flags = Op.getNode()->getFlags();
+    Known = computeKnownBits(Op.getOperand(0), DemandedElts, Depth + 1);
+    Known2 = computeKnownBits(Op.getOperand(1), DemandedElts, Depth + 1);
+    Known = KnownBits::computeForAddSub(Op.getOpcode() == ISD::ADD,
+                                        Flags.hasNoSignedWrap(), Known, Known2);
+    break;
+  }
   case ISD::USUBO:
   case ISD::SSUBO:
   case ISD::USUBO_CARRY:
@@ -3659,7 +3668,6 @@
       break;
     }
     [[fallthrough]];
-  case ISD::SUB:
   case ISD::SUBC: {
     assert(Op.getResNo() == 0 &&
            "We only compute knownbits for the difference here.");
@@ -3687,7 +3695,6 @@
       break;
     }
     [[fallthrough]];
-  case ISD::ADD:
   case ISD::ADDC:
   case ISD::ADDE: {
     assert(Op.getResNo() == 0 && "We only compute knownbits for the sum here.");
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
@@ -2694,11 +2694,9 @@
     if (Op.getOpcode() == ISD::MUL) {
       Known = KnownBits::mul(KnownOp0, KnownOp1);
     } else { // Op.getOpcode() is either ISD::ADD or ISD::SUB.
-      // TODO: Update `computeForAddCarry` to handle the NSW flag as well so
-      //       that `Flags.hasNoSignedWrap()` can be passed through here
-      //       instead of false.
-      Known = KnownBits::computeForAddSub(Op.getOpcode() == ISD::ADD, false,
-                                          KnownOp0, KnownOp1);
+      Known = KnownBits::computeForAddSub(Op.getOpcode() == ISD::ADD,
+                                          Flags.hasNoSignedWrap(), KnownOp0,
+                                          KnownOp1);
     }
     break;
   }
diff --git a/llvm/test/CodeGen/Thumb2/mve-blockplacement.ll b/llvm/test/CodeGen/Thumb2/mve-blockplacement.ll
--- a/llvm/test/CodeGen/Thumb2/mve-blockplacement.ll
+++ b/llvm/test/CodeGen/Thumb2/mve-blockplacement.ll
@@ -366,18 +366,16 @@
 ; CHECK-NEXT:    str r0, [sp, #12] @ 4-byte Spill
 ; CHECK-NEXT:    movs r1, #4
 ; CHECK-NEXT:    strd r2, r12, [sp, #4] @ 8-byte Folded Spill
-; CHECK-NEXT:    add.w r3, r3, r4, lsr #1
 ; CHECK-NEXT:    add.w r1, r1, r4, lsr #1
-; CHECK-NEXT:    movw r4, #65532
-; CHECK-NEXT:    vdup.32 q6, r3
-; CHECK-NEXT:    movt r4, #32767
-; CHECK-NEXT:    and.w r7, r1, r4
+; CHECK-NEXT:    add.w r3, r3, r4, lsr #1
+; CHECK-NEXT:    bic r7, r1, #3
 ; CHECK-NEXT:    adr r1, .LCPI1_0
-; CHECK-NEXT:    vdup.32 q7, r3
 ; CHECK-NEXT:    vldrw.u32 q0, [r1]
 ; CHECK-NEXT:    adr r1, .LCPI1_1
 ; CHECK-NEXT:    vldrw.u32 q5, [r1]
+; CHECK-NEXT:    vdup.32 q6, r3
 ; CHECK-NEXT:    vadd.i32 q4, q0, lr
+; CHECK-NEXT:    vdup.32 q7, r3
 ; CHECK-NEXT:    b .LBB1_4
 ; CHECK-NEXT:  .LBB1_2: @ %for.body6.preheader
 ; CHECK-NEXT:    @ in Loop: Header=BB1_4 Depth=1