Apply transformation

Author: leewei05

llvm/lib/Transforms/InstCombine/InstCombineCompares.cpp

@@ -11,6 +11,7 @@
 //===----------------------------------------------------------------------===//
 
 #include "InstCombineInternal.h"
+#include "llvm/ADT/APInt.h"
 #include "llvm/ADT/APSInt.h"
 #include "llvm/ADT/ScopeExit.h"
 #include "llvm/ADT/SetVector.h"
@@ -24,6 +25,7 @@
 #include "llvm/IR/ConstantRange.h"
 #include "llvm/IR/DataLayout.h"
 #include "llvm/IR/InstrTypes.h"
+#include "llvm/IR/Instructions.h"
 #include "llvm/IR/IntrinsicInst.h"
 #include "llvm/IR/PatternMatch.h"
 #include "llvm/Support/KnownBits.h"
@@ -7687,6 +7689,34 @@ Instruction *InstCombinerImpl::visitICmpInst(ICmpInst &I) {
   if (Instruction *Res = foldReductionIdiom(I, Builder, DL))
     return Res;
 
+  {
+    Value *A;
+    const APInt *C1, *C2;
+    ICmpInst::Predicate Pred = I.getPredicate();
+    if (Pred == ICmpInst::ICMP_EQ || Pred == ICmpInst::ICMP_NE) {
+      // sext(a) & c1 == c2 --> a & c3 == trunc(c2)
+      // sext(a) & c1 != c2 --> a & c3 != trunc(c2)
+      if (match(Op0, m_And(m_SExtLike(m_Value(A)), m_APInt(C1))) &&
+          match(Op1, m_APInt(C2))) {
+        Type *InputTy = A->getType();
+        unsigned InputBitWidth = InputTy->getScalarSizeInBits();
+        // c2 must be non-negative at the bitwidth of a.
+        if (C2->ult(APInt::getOneBitSet(C2->getBitWidth(), InputBitWidth))) {
+          // Check if there are 1s in C1 high bits of size InputBitWidth.
+          bool HasLeading1s =
+              C1->uge(APInt::getOneBitSet(C1->getBitWidth(), InputBitWidth));
+          APInt TruncC1 = C1->trunc(InputBitWidth);
+          if (HasLeading1s)
+            TruncC1.setBit(InputBitWidth - 1);
+          Value *AndInst = Builder.CreateAnd(A, TruncC1);
+          return new ICmpInst(
+              Pred, AndInst,
+              ConstantInt::get(InputTy, C2->trunc(InputBitWidth)));
+        }
+      }
+    }
+  }
+
   return Changed ? &I : nullptr;
 }
 

llvm/test/Transforms/InstCombine/sext-and.ll

@@ -6,9 +6,8 @@ declare void @use(i8)
 define i1 @fold_sext_to_and(i8 %x) {
 ; CHECK-LABEL: define i1 @fold_sext_to_and(
 ; CHECK-SAME: i8 [[X:%.*]]) {
-; CHECK-NEXT:    [[TMP1:%.*]] = sext i8 [[X]] to i32
-; CHECK-NEXT:    [[TMP2:%.*]] = and i32 [[TMP1]], -2147483647
-; CHECK-NEXT:    [[TMP3:%.*]] = icmp eq i32 [[TMP2]], 1
+; CHECK-NEXT:    [[TMP1:%.*]] = and i8 [[X]], -127
+; CHECK-NEXT:    [[TMP3:%.*]] = icmp eq i8 [[TMP1]], 1
 ; CHECK-NEXT:    ret i1 [[TMP3]]
 ;
   %1 = sext i8 %x to i32
@@ -20,9 +19,8 @@ define i1 @fold_sext_to_and(i8 %x) {
 define i1 @fold_sext_to_and1(i8 %x) {
 ; CHECK-LABEL: define i1 @fold_sext_to_and1(
 ; CHECK-SAME: i8 [[X:%.*]]) {
-; CHECK-NEXT:    [[TMP1:%.*]] = sext i8 [[X]] to i32
-; CHECK-NEXT:    [[TMP2:%.*]] = and i32 [[TMP1]], -2147483647
-; CHECK-NEXT:    [[TMP3:%.*]] = icmp ne i32 [[TMP2]], 1
+; CHECK-NEXT:    [[TMP1:%.*]] = and i8 [[X]], -127
+; CHECK-NEXT:    [[TMP3:%.*]] = icmp ne i8 [[TMP1]], 1
 ; CHECK-NEXT:    ret i1 [[TMP3]]
 ;
   %1 = sext i8 %x to i32
@@ -36,8 +34,8 @@ define i1 @fold_sext_to_and_multi_use(i8 %x) {
 ; CHECK-SAME: i8 [[X:%.*]]) {
 ; CHECK-NEXT:    [[TMP1:%.*]] = sext i8 [[X]] to i32
 ; CHECK-NEXT:    call void @use(i32 [[TMP1]])
-; CHECK-NEXT:    [[TMP2:%.*]] = and i32 [[TMP1]], -2147483647
-; CHECK-NEXT:    [[TMP3:%.*]] = icmp eq i32 [[TMP2]], 1
+; CHECK-NEXT:    [[TMP2:%.*]] = and i8 [[X]], -127
+; CHECK-NEXT:    [[TMP3:%.*]] = icmp eq i8 [[TMP2]], 1
 ; CHECK-NEXT:    ret i1 [[TMP3]]
 ;
   %1 = sext i8 %x to i32
@@ -52,8 +50,8 @@ define i1 @fold_sext_to_and_multi_use1(i8 %x) {
 ; CHECK-SAME: i8 [[X:%.*]]) {
 ; CHECK-NEXT:    [[TMP1:%.*]] = sext i8 [[X]] to i32
 ; CHECK-NEXT:    call void @use(i32 [[TMP1]])
-; CHECK-NEXT:    [[TMP2:%.*]] = and i32 [[TMP1]], -2147483647
-; CHECK-NEXT:    [[TMP3:%.*]] = icmp ne i32 [[TMP2]], 1
+; CHECK-NEXT:    [[TMP2:%.*]] = and i8 [[X]], -127
+; CHECK-NEXT:    [[TMP3:%.*]] = icmp ne i8 [[TMP2]], 1
 ; CHECK-NEXT:    ret i1 [[TMP3]]
 ;
   %1 = sext i8 %x to i32