[InstCombine] Handle commuted patterns in foldBinOpShiftWithShift

[dtcxzyw]

Closes #121775.

[llvmbot]

@llvm/pr-subscribers-llvm-transforms

Author: Yingwei Zheng (dtcxzyw)

Changes

Closes #121775.

---

Full diff: https://github.com/llvm/llvm-project/pull/122126.diff

2 Files Affected:

• (modified) llvm/lib/Transforms/InstCombine/InstructionCombining.cpp (+4-5)
• (modified) llvm/test/Transforms/InstCombine/and-xor-or.ll (+51-8)

diff --git a/llvm/lib/Transforms/InstCombine/InstructionCombining.cpp b/llvm/lib/Transforms/InstCombine/InstructionCombining.cpp
index 553435c937a70a..2fb60ef11499c7 100644
--- a/llvm/lib/Transforms/InstCombine/InstructionCombining.cpp
+++ b/llvm/lib/Transforms/InstCombine/InstructionCombining.cpp
@@ -939,12 +939,11 @@ Instruction *InstCombinerImpl::foldBinOpShiftWithShift(BinaryOperator &I) {
                m_OneUse(m_Shift(m_Value(Y), m_Value(Shift)))))
       return nullptr;
     if (!match(I.getOperand(1 - ShOpnum),
-               m_BinOp(m_Value(ShiftedX), m_Value(Mask))))
+               m_c_BinOp(m_CombineAnd(
+                             m_OneUse(m_Shift(m_Value(X), m_Specific(Shift))),
+                             m_Value(ShiftedX)),
+                         m_Value(Mask))))
       return nullptr;
-
-    if (!match(ShiftedX, m_OneUse(m_Shift(m_Value(X), m_Specific(Shift)))))
-      return nullptr;
-
     // Make sure we are matching instruction shifts and not ConstantExpr
     auto *IY = dyn_cast<Instruction>(I.getOperand(ShOpnum));
     auto *IX = dyn_cast<Instruction>(ShiftedX);
diff --git a/llvm/test/Transforms/InstCombine/and-xor-or.ll b/llvm/test/Transforms/InstCombine/and-xor-or.ll
index 5a0890e918ef0f..5a58995f6c315f 100644
--- a/llvm/test/Transforms/InstCombine/and-xor-or.ll
+++ b/llvm/test/Transforms/InstCombine/and-xor-or.ll
@@ -388,10 +388,9 @@ define i8 @xor_shl(i8 %x, i8 %y, i8 %zarg, i8 %shamt) {
 ; CHECK-LABEL: define {{[^@]+}}@xor_shl
 ; CHECK-SAME: (i8 [[X:%.*]], i8 [[Y:%.*]], i8 [[ZARG:%.*]], i8 [[SHAMT:%.*]]) {
 ; CHECK-NEXT:    [[Z:%.*]] = sdiv i8 42, [[ZARG]]
-; CHECK-NEXT:    [[SX:%.*]] = shl i8 [[X]], [[SHAMT]]
-; CHECK-NEXT:    [[SY:%.*]] = shl i8 [[Y]], [[SHAMT]]
-; CHECK-NEXT:    [[A:%.*]] = xor i8 [[Z]], [[SX]]
-; CHECK-NEXT:    [[R:%.*]] = xor i8 [[A]], [[SY]]
+; CHECK-NEXT:    [[TMP1:%.*]] = xor i8 [[X]], [[Y]]
+; CHECK-NEXT:    [[TMP2:%.*]] = shl i8 [[TMP1]], [[SHAMT]]
+; CHECK-NEXT:    [[R:%.*]] = xor i8 [[TMP2]], [[Z]]
 ; CHECK-NEXT:    ret i8 [[R]]
 ;
   %z = sdiv i8 42, %zarg ; thwart complexity-based canonicalization
@@ -406,10 +405,9 @@ define i8 @and_lshr(i8 %x, i8 %y, i8 %zarg, i8 %shamt) {
 ; CHECK-LABEL: define {{[^@]+}}@and_lshr
 ; CHECK-SAME: (i8 [[X:%.*]], i8 [[Y:%.*]], i8 [[ZARG:%.*]], i8 [[SHAMT:%.*]]) {
 ; CHECK-NEXT:    [[Z:%.*]] = sdiv i8 42, [[ZARG]]
-; CHECK-NEXT:    [[SX:%.*]] = lshr i8 [[X]], [[SHAMT]]
-; CHECK-NEXT:    [[SY:%.*]] = lshr i8 [[Y]], [[SHAMT]]
-; CHECK-NEXT:    [[A:%.*]] = and i8 [[Z]], [[SX]]
-; CHECK-NEXT:    [[R:%.*]] = and i8 [[SY]], [[A]]
+; CHECK-NEXT:    [[TMP1:%.*]] = and i8 [[X]], [[Y]]
+; CHECK-NEXT:    [[TMP2:%.*]] = lshr i8 [[TMP1]], [[SHAMT]]
+; CHECK-NEXT:    [[R:%.*]] = and i8 [[TMP2]], [[Z]]
 ; CHECK-NEXT:    ret i8 [[R]]
 ;
   %z = sdiv i8 42, %zarg ; thwart complexity-based canonicalization
@@ -435,6 +433,51 @@ define i8 @or_lshr(i8 %x, i8 %y, i8 %z, i8 %shamt) {
   ret i8 %r
 }
 
+define i8 @or_lshr_commuted1(i8 %x, i8 %y, i8 %z, i8 %shamt) {
+; CHECK-LABEL: define {{[^@]+}}@or_lshr_commuted1
+; CHECK-SAME: (i8 [[X:%.*]], i8 [[Y:%.*]], i8 [[Z:%.*]], i8 [[SHAMT:%.*]]) {
+; CHECK-NEXT:    [[TMP1:%.*]] = or i8 [[X]], [[Y]]
+; CHECK-NEXT:    [[TMP2:%.*]] = lshr i8 [[TMP1]], [[SHAMT]]
+; CHECK-NEXT:    [[R:%.*]] = or i8 [[TMP2]], [[Z]]
+; CHECK-NEXT:    ret i8 [[R]]
+;
+  %sx = lshr i8 %x, %shamt
+  %sy = lshr i8 %y, %shamt
+  %a = or i8 %z, %sx
+  %r = or i8 %sy, %a
+  ret i8 %r
+}
+
+define i8 @or_lshr_commuted2(i8 %x, i8 %y, i8 %z, i8 %shamt) {
+; CHECK-LABEL: define {{[^@]+}}@or_lshr_commuted2
+; CHECK-SAME: (i8 [[X:%.*]], i8 [[Y:%.*]], i8 [[Z:%.*]], i8 [[SHAMT:%.*]]) {
+; CHECK-NEXT:    [[TMP1:%.*]] = or i8 [[X]], [[Y]]
+; CHECK-NEXT:    [[TMP2:%.*]] = lshr i8 [[TMP1]], [[SHAMT]]
+; CHECK-NEXT:    [[R:%.*]] = or i8 [[TMP2]], [[Z]]
+; CHECK-NEXT:    ret i8 [[R]]
+;
+  %sx = lshr i8 %x, %shamt
+  %sy = lshr i8 %y, %shamt
+  %a = or i8 %z, %sx
+  %r = or i8 %a, %sy
+  ret i8 %r
+}
+
+define i8 @or_lshr_commuted3(i8 %x, i8 %y, i8 %z, i8 %shamt) {
+; CHECK-LABEL: define {{[^@]+}}@or_lshr_commuted3
+; CHECK-SAME: (i8 [[X:%.*]], i8 [[Y:%.*]], i8 [[Z:%.*]], i8 [[SHAMT:%.*]]) {
+; CHECK-NEXT:    [[TMP1:%.*]] = or i8 [[X]], [[Y]]
+; CHECK-NEXT:    [[TMP2:%.*]] = lshr i8 [[TMP1]], [[SHAMT]]
+; CHECK-NEXT:    [[R:%.*]] = or i8 [[TMP2]], [[Z]]
+; CHECK-NEXT:    ret i8 [[R]]
+;
+  %sx = lshr i8 %x, %shamt
+  %sy = lshr i8 %y, %shamt
+  %a = or i8 %sx, %z
+  %r = or i8 %a, %sy
+  ret i8 %r
+}
+
 define i8 @xor_lshr(i8 %x, i8 %y, i8 %z, i8 %shamt) {
 ; CHECK-LABEL: define {{[^@]+}}@xor_lshr
 ; CHECK-SAME: (i8 [[X:%.*]], i8 [[Y:%.*]], i8 [[Z:%.*]], i8 [[SHAMT:%.*]]) {

[nikic]

LGTM