Index: llvm/trunk/include/llvm/Transforms/Utils/SimplifyLibCalls.h
===================================================================
--- llvm/trunk/include/llvm/Transforms/Utils/SimplifyLibCalls.h
+++ llvm/trunk/include/llvm/Transforms/Utils/SimplifyLibCalls.h
@@ -138,6 +138,8 @@
   Value *optimizeStrStr(CallInst *CI, IRBuilder<> &B);
   Value *optimizeMemChr(CallInst *CI, IRBuilder<> &B);
   Value *optimizeMemCmp(CallInst *CI, IRBuilder<> &B);
+  Value *optimizeBCmp(CallInst *CI, IRBuilder<> &B);
+  Value *optimizeMemCmpBCmpCommon(CallInst *CI, IRBuilder<> &B);
   Value *optimizeMemCpy(CallInst *CI, IRBuilder<> &B);
   Value *optimizeMemMove(CallInst *CI, IRBuilder<> &B);
   Value *optimizeMemSet(CallInst *CI, IRBuilder<> &B);
Index: llvm/trunk/lib/Transforms/Utils/SimplifyLibCalls.cpp
===================================================================
--- llvm/trunk/lib/Transforms/Utils/SimplifyLibCalls.cpp
+++ llvm/trunk/lib/Transforms/Utils/SimplifyLibCalls.cpp
@@ -916,7 +916,9 @@
   return nullptr;
 }
 
-Value *LibCallSimplifier::optimizeMemCmp(CallInst *CI, IRBuilder<> &B) {
+// Most simplifications for memcmp also apply to bcmp.
+Value *LibCallSimplifier::optimizeMemCmpBCmpCommon(CallInst *CI,
+                                                   IRBuilder<> &B) {
   Value *LHS = CI->getArgOperand(0), *RHS = CI->getArgOperand(1);
   Value *Size = CI->getArgOperand(2);
 
@@ -929,16 +931,30 @@
                                                 LenC->getZExtValue(), B, DL))
       return Res;
 
+  return nullptr;
+}
+
+Value *LibCallSimplifier::optimizeMemCmp(CallInst *CI, IRBuilder<> &B) {
+  if (Value *V = optimizeMemCmpBCmpCommon(CI, B))
+    return V;
+
   // memcmp(x, y, Len) == 0 -> bcmp(x, y, Len) == 0
   // `bcmp` can be more efficient than memcmp because it only has to know that
   // there is a difference, not where it is.
   if (isOnlyUsedInZeroEqualityComparison(CI) && TLI->has(LibFunc_bcmp)) {
+    Value *LHS = CI->getArgOperand(0);
+    Value *RHS = CI->getArgOperand(1);
+    Value *Size = CI->getArgOperand(2);
     return emitBCmp(LHS, RHS, Size, B, DL, TLI);
   }
 
   return nullptr;
 }
 
+Value *LibCallSimplifier::optimizeBCmp(CallInst *CI, IRBuilder<> &B) {
+  return optimizeMemCmpBCmpCommon(CI, B);
+}
+
 Value *LibCallSimplifier::optimizeMemCpy(CallInst *CI, IRBuilder<> &B) {
   // memcpy(x, y, n) -> llvm.memcpy(align 1 x, align 1 y, n)
   B.CreateMemCpy(CI->getArgOperand(0), 1, CI->getArgOperand(1), 1,
@@ -2512,6 +2528,8 @@
       return optimizeStrStr(CI, Builder);
     case LibFunc_memchr:
       return optimizeMemChr(CI, Builder);
+    case LibFunc_bcmp:
+      return optimizeBCmp(CI, Builder);
     case LibFunc_memcmp:
       return optimizeMemCmp(CI, Builder);
     case LibFunc_memcpy:
Index: llvm/trunk/test/Transforms/InstCombine/bcmp-1.ll
===================================================================
--- llvm/trunk/test/Transforms/InstCombine/bcmp-1.ll
+++ llvm/trunk/test/Transforms/InstCombine/bcmp-1.ll
@@ -0,0 +1,144 @@
+; NOTE: Assertions have been autogenerated by utils/update_test_checks.py
+; Test that the bcmp library call simplifier works correctly.
+;
+; RUN: opt < %s -instcombine -mtriple=x86_64-unknown-linux-gnu -S | FileCheck %s
+
+target datalayout = "e-p:32:32:32-i1:8:8-i8:8:8-i16:16:16-i32:32:32-i64:32:64-f32:32:32-f64:32:64-v64:64:64-v128:128:128-a0:0:64-f80:128:128-n8:16:32:64"
+
+@foo = constant [4 x i8] c"foo\00"
+@hel = constant [4 x i8] c"hel\00"
+@hello_u = constant [8 x i8] c"hello_u\00"
+
+declare i32 @bcmp(i8*, i8*, i32)
+
+; Check bcmp(mem, mem, size) -> 0.
+
+define i32 @test_simplify1(i8* %mem, i32 %size) {
+; CHECK-LABEL: @test_simplify1(
+; CHECK-NEXT:    ret i32 0
+;
+  %ret = call i32 @bcmp(i8* %mem, i8* %mem, i32 %size)
+  ret i32 %ret
+}
+
+; Check bcmp(mem1, mem2, 0) -> 0.
+
+define i32 @test_simplify2(i8* %mem1, i8* %mem2) {
+; CHECK-LABEL: @test_simplify2(
+; CHECK-NEXT:    ret i32 0
+;
+  %ret = call i32 @bcmp(i8* %mem1, i8* %mem2, i32 0)
+  ret i32 %ret
+}
+
+;; Check bcmp(mem1, mem2, 1) -> *(unsigned char*)mem1 - *(unsigned char*)mem2.
+
+define i32 @test_simplify3(i8* %mem1, i8* %mem2) {
+; CHECK-LABEL: @test_simplify3(
+; CHECK-NEXT:    [[LHSC:%.*]] = load i8, i8* [[MEM1:%.*]], align 1
+; CHECK-NEXT:    [[LHSV:%.*]] = zext i8 [[LHSC]] to i32
+; CHECK-NEXT:    [[RHSC:%.*]] = load i8, i8* [[MEM2:%.*]], align 1
+; CHECK-NEXT:    [[RHSV:%.*]] = zext i8 [[RHSC]] to i32
+; CHECK-NEXT:    [[CHARDIFF:%.*]] = sub nsw i32 [[LHSV]], [[RHSV]]
+; CHECK-NEXT:    ret i32 [[CHARDIFF]]
+;
+  %ret = call i32 @bcmp(i8* %mem1, i8* %mem2, i32 1)
+  ret i32 %ret
+}
+
+; Check bcmp(mem1, mem2, size) -> cnst, where all arguments are constants.
+
+define i32 @test_simplify4() {
+; CHECK-LABEL: @test_simplify4(
+; CHECK-NEXT:    ret i32 0
+;
+  %mem1 = getelementptr [4 x i8], [4 x i8]* @hel, i32 0, i32 0
+  %mem2 = getelementptr [8 x i8], [8 x i8]* @hello_u, i32 0, i32 0
+  %ret = call i32 @bcmp(i8* %mem1, i8* %mem2, i32 3)
+  ret i32 %ret
+}
+
+define i32 @test_simplify5() {
+; CHECK-LABEL: @test_simplify5(
+; CHECK-NEXT:    ret i32 1
+;
+  %mem1 = getelementptr [4 x i8], [4 x i8]* @hel, i32 0, i32 0
+  %mem2 = getelementptr [4 x i8], [4 x i8]* @foo, i32 0, i32 0
+  %ret = call i32 @bcmp(i8* %mem1, i8* %mem2, i32 3)
+  ret i32 %ret
+}
+
+define i32 @test_simplify6() {
+; CHECK-LABEL: @test_simplify6(
+; CHECK-NEXT:    ret i32 -1
+;
+  %mem1 = getelementptr [4 x i8], [4 x i8]* @foo, i32 0, i32 0
+  %mem2 = getelementptr [4 x i8], [4 x i8]* @hel, i32 0, i32 0
+  %ret = call i32 @bcmp(i8* %mem1, i8* %mem2, i32 3)
+  ret i32 %ret
+}
+
+; Check bcmp(mem1, mem2, 8)==0 -> *(int64_t*)mem1 == *(int64_t*)mem2
+
+define i1 @test_simplify7(i64 %x, i64 %y) {
+; CHECK-LABEL: @test_simplify7(
+; CHECK-NEXT:    [[TMP1:%.*]] = icmp eq i64 [[X:%.*]], [[Y:%.*]]
+; CHECK-NEXT:    ret i1 [[TMP1]]
+;
+  %x.addr = alloca i64, align 8
+  %y.addr = alloca i64, align 8
+  store i64 %x, i64* %x.addr, align 8
+  store i64 %y, i64* %y.addr, align 8
+  %xptr = bitcast i64* %x.addr to i8*
+  %yptr = bitcast i64* %y.addr to i8*
+  %call = call i32 @bcmp(i8* %xptr, i8* %yptr, i32 8)
+  %cmp = icmp eq i32 %call, 0
+  ret i1 %cmp
+}
+
+; Check bcmp(mem1, mem2, 4)==0 -> *(int32_t*)mem1 == *(int32_t*)mem2
+
+define i1 @test_simplify8(i32 %x, i32 %y) {
+; CHECK-LABEL: @test_simplify8(
+; CHECK-NEXT:    [[TMP1:%.*]] = icmp eq i32 [[X:%.*]], [[Y:%.*]]
+; CHECK-NEXT:    ret i1 [[TMP1]]
+;
+  %x.addr = alloca i32, align 4
+  %y.addr = alloca i32, align 4
+  store i32 %x, i32* %x.addr, align 4
+  store i32 %y, i32* %y.addr, align 4
+  %xptr = bitcast i32* %x.addr to i8*
+  %yptr = bitcast i32* %y.addr to i8*
+  %call = call i32 @bcmp(i8* %xptr, i8* %yptr, i32 4)
+  %cmp = icmp eq i32 %call, 0
+  ret i1 %cmp
+}
+
+; Check bcmp(mem1, mem2, 2)==0 -> *(int16_t*)mem1 == *(int16_t*)mem2
+
+define i1 @test_simplify9(i16 %x, i16 %y) {
+; CHECK-LABEL: @test_simplify9(
+; CHECK-NEXT:    [[TMP1:%.*]] = icmp eq i16 [[X:%.*]], [[Y:%.*]]
+; CHECK-NEXT:    ret i1 [[TMP1]]
+;
+  %x.addr = alloca i16, align 2
+  %y.addr = alloca i16, align 2
+  store i16 %x, i16* %x.addr, align 2
+  store i16 %y, i16* %y.addr, align 2
+  %xptr = bitcast i16* %x.addr to i8*
+  %yptr = bitcast i16* %y.addr to i8*
+  %call = call i32 @bcmp(i8* %xptr, i8* %yptr, i32 2)
+  %cmp = icmp eq i32 %call, 0
+  ret i1 %cmp
+}
+
+define i1 @test_simplify10(i8* %mem1, i8* %mem2, i32 %size) {
+; CHECK-LABEL: @test_simplify10(
+; CHECK-NEXT:    [[CALL:%.*]] = call i32 @bcmp(i8* [[MEM1:%.*]], i8* [[MEM2:%.*]], i32 [[SIZE:%.*]])
+; CHECK-NEXT:    [[CMP:%.*]] = icmp eq i32 [[CALL]], 0
+; CHECK-NEXT:    ret i1 [[CMP]]
+;
+  %call = call i32 @bcmp(i8* %mem1, i8* %mem2, i32 %size)
+  %cmp = icmp eq i32 %call, 0
+  ret i1 %cmp
+}