[InstCombine] Fold Xor with or disjoint

Implement a missing optimization to fold (A | B) ^ C to (A ^ C) ^ B

Author: AmrDeveloper

llvm/lib/Transforms/InstCombine/InstCombineAndOrXor.cpp

@@ -4706,6 +4706,15 @@ Instruction *InstCombinerImpl::visitXor(BinaryOperator &I) {
     return Xor;
 
   Value *X, *Y;
+
+  // (A | B) ^ C -> (A ^ C) ^ B
+  // C ^ (A | B) -> B ^ (A ^ C)
+  if (match(&I, m_c_Xor(m_OneUse(m_c_DisjointOr(m_Value(X), m_Value(Y))),
+                        m_Value(M)))) {
+    Value *XorAC = Builder.CreateXor(X, M);
+    return BinaryOperator::CreateXor(XorAC, Y);
+  }
+
   Constant *C1;
   if (match(Op1, m_Constant(C1))) {
     Constant *C2;

llvm/test/Transforms/InstCombine/2010-11-01-lshr-mask.ll

@@ -33,12 +33,12 @@ define i8 @foo(i8 %arg, i8 %arg1) {
 ; CHECK-NEXT:    [[T4:%.*]] = and i8 [[ARG1]], 33
 ; CHECK-NEXT:    [[T5:%.*]] = sub nsw i8 40, [[T2]]
 ; CHECK-NEXT:    [[T6:%.*]] = and i8 [[T5]], 84
-; CHECK-NEXT:    [[T7:%.*]] = or disjoint i8 [[T4]], [[T6]]
 ; CHECK-NEXT:    [[T8:%.*]] = xor i8 [[T]], [[T3]]
-; CHECK-NEXT:    [[T9:%.*]] = or disjoint i8 [[T7]], [[T8]]
 ; CHECK-NEXT:    [[TMP1:%.*]] = lshr i8 [[T8]], 2
 ; CHECK-NEXT:    [[T11:%.*]] = and i8 [[TMP1]], 32
-; CHECK-NEXT:    [[T12:%.*]] = xor i8 [[T11]], [[T9]]
+; CHECK-NEXT:    [[TMP2:%.*]] = xor i8 [[T4]], [[T11]]
+; CHECK-NEXT:    [[TMP3:%.*]] = or disjoint i8 [[TMP2]], [[T6]]
+; CHECK-NEXT:    [[T12:%.*]] = or disjoint i8 [[TMP3]], [[T8]]
 ; CHECK-NEXT:    ret i8 [[T12]]
 ;
   %t = shl i8 %arg, 7

llvm/test/Transforms/InstCombine/icmp-of-xor-x.ll

@@ -358,8 +358,8 @@ define i1 @xor_ugt_2(i8 %xx, i8 %y, i8 %z) {
 ; CHECK-LABEL: @xor_ugt_2(
 ; CHECK-NEXT:    [[X:%.*]] = add i8 [[XX:%.*]], [[Z:%.*]]
 ; CHECK-NEXT:    [[YZ:%.*]] = and i8 [[Y:%.*]], 63
-; CHECK-NEXT:    [[Y1:%.*]] = or disjoint i8 [[YZ]], 64
-; CHECK-NEXT:    [[XOR:%.*]] = xor i8 [[X]], [[Y1]]
+; CHECK-NEXT:    [[TMP1:%.*]] = xor i8 [[YZ]], [[X]]
+; CHECK-NEXT:    [[XOR:%.*]] = xor i8 [[TMP1]], 64
 ; CHECK-NEXT:    [[R:%.*]] = icmp ugt i8 [[X]], [[XOR]]
 ; CHECK-NEXT:    ret i1 [[R]]
 ;
@@ -385,8 +385,8 @@ define i1 @xor_ult(i8 %x) {
 define <2 x i1> @xor_sgt(<2 x i8> %x, <2 x i8> %y) {
 ; CHECK-LABEL: @xor_sgt(
 ; CHECK-NEXT:    [[YZ:%.*]] = and <2 x i8> [[Y:%.*]], <i8 31, i8 31>
-; CHECK-NEXT:    [[Y1:%.*]] = or disjoint <2 x i8> [[YZ]], <i8 64, i8 64>
-; CHECK-NEXT:    [[XOR:%.*]] = xor <2 x i8> [[X:%.*]], [[Y1]]
+; CHECK-NEXT:    [[TMP1:%.*]] = xor <2 x i8> [[YZ]], [[X:%.*]]
+; CHECK-NEXT:    [[XOR:%.*]] = xor <2 x i8> [[TMP1]], <i8 64, i8 64>
 ; CHECK-NEXT:    [[R:%.*]] = icmp sgt <2 x i8> [[XOR]], [[X]]
 ; CHECK-NEXT:    ret <2 x i1> [[R]]
 ;
@@ -400,8 +400,8 @@ define <2 x i1> @xor_sgt(<2 x i8> %x, <2 x i8> %y) {
 define <2 x i1> @xor_sgt_fail_no_known_msb(<2 x i8> %x, <2 x i8> %y) {
 ; CHECK-LABEL: @xor_sgt_fail_no_known_msb(
 ; CHECK-NEXT:    [[YZ:%.*]] = and <2 x i8> [[Y:%.*]], <i8 55, i8 55>
-; CHECK-NEXT:    [[Y1:%.*]] = or disjoint <2 x i8> [[YZ]], <i8 8, i8 8>
-; CHECK-NEXT:    [[XOR:%.*]] = xor <2 x i8> [[X:%.*]], [[Y1]]
+; CHECK-NEXT:    [[TMP1:%.*]] = xor <2 x i8> [[YZ]], [[X:%.*]]
+; CHECK-NEXT:    [[XOR:%.*]] = xor <2 x i8> [[TMP1]], <i8 8, i8 8>
 ; CHECK-NEXT:    [[R:%.*]] = icmp sgt <2 x i8> [[XOR]], [[X]]
 ; CHECK-NEXT:    ret <2 x i1> [[R]]
 ;

llvm/test/Transforms/InstCombine/xor.ll

@@ -1489,8 +1489,8 @@ define i4 @PR96857_xor_without_noundef(i4  %val0, i4  %val1, i4 %val2) {
 define i32 @or_disjoint_with_xor(i32 %a, i32 %b, i32 %c) {
 ; CHECK-LABEL: @or_disjoint_with_xor(
 ; CHECK-NEXT:  entry:
-; CHECK-NEXT:    [[OR:%.*]] = or disjoint i32 [[A:%.*]], [[B:%.*]]
-; CHECK-NEXT:    [[XOR:%.*]] = xor i32 [[OR]], [[C:%.*]]
+; CHECK-NEXT:    [[TMP0:%.*]] = xor i32 [[A:%.*]], [[C:%.*]]
+; CHECK-NEXT:    [[XOR:%.*]] = xor i32 [[TMP0]], [[B:%.*]]
 ; CHECK-NEXT:    ret i32 [[XOR]]
 ;
 entry:
@@ -1502,8 +1502,8 @@ entry:
 define i32 @xor_with_or_disjoint(i32 %a, i32 %b, i32 %c) {
 ; CHECK-LABEL: @xor_with_or_disjoint(
 ; CHECK-NEXT:  entry:
-; CHECK-NEXT:    [[OR:%.*]] = or disjoint i32 [[A:%.*]], [[B:%.*]]
-; CHECK-NEXT:    [[XOR:%.*]] = xor i32 [[C:%.*]], [[OR]]
+; CHECK-NEXT:    [[TMP0:%.*]] = xor i32 [[A:%.*]], [[C:%.*]]
+; CHECK-NEXT:    [[XOR:%.*]] = xor i32 [[TMP0]], [[B:%.*]]
 ; CHECK-NEXT:    ret i32 [[XOR]]
 ;
 entry:
@@ -1515,8 +1515,8 @@ entry:
 define <2 x i32> @or_disjoint_with_xor_vec(<2 x i32> %a, < 2 x i32> %b, <2 x i32> %c) {
 ; CHECK-LABEL: @or_disjoint_with_xor_vec(
 ; CHECK-NEXT:  entry:
-; CHECK-NEXT:    [[OR:%.*]] = or disjoint <2 x i32> [[A:%.*]], [[B:%.*]]
-; CHECK-NEXT:    [[XOR:%.*]] = xor <2 x i32> [[OR]], [[C:%.*]]
+; CHECK-NEXT:    [[TMP0:%.*]] = xor <2 x i32> [[A:%.*]], [[C:%.*]]
+; CHECK-NEXT:    [[XOR:%.*]] = xor <2 x i32> [[TMP0]], [[B:%.*]]
 ; CHECK-NEXT:    ret <2 x i32> [[XOR]]
 ;
 entry:
@@ -1528,8 +1528,8 @@ entry:
 define <2 x i32> @xor_with_or_disjoint_vec(<2 x i32> %a, < 2 x i32> %b, <2 x i32> %c) {
 ; CHECK-LABEL: @xor_with_or_disjoint_vec(
 ; CHECK-NEXT:  entry:
-; CHECK-NEXT:    [[OR:%.*]] = or disjoint <2 x i32> [[A:%.*]], [[B:%.*]]
-; CHECK-NEXT:    [[XOR:%.*]] = xor <2 x i32> [[C:%.*]], [[OR]]
+; CHECK-NEXT:    [[TMP0:%.*]] = xor <2 x i32> [[A:%.*]], [[C:%.*]]
+; CHECK-NEXT:    [[XOR:%.*]] = xor <2 x i32> [[TMP0]], [[B:%.*]]
 ; CHECK-NEXT:    ret <2 x i32> [[XOR]]
 ;
 entry: