teach instcombine to remove allocated buffers even if there are stores, memcpy/memmove/memset, and objectsize users.

This means we can do cheap DSE for heap memory.
Nothing is done if the pointer excapes or has a load.

The churn in the tests is mostly due to objectsize, since we want to make sure we
don't delete the malloc call before evaluating the objectsize (otherwise it becomes -1/0)

llvm-svn: 159876

llvm/lib/Transforms/InstCombine/InstructionCombining.cpp

@@ -1137,12 +1137,29 @@ static bool IsOnlyNullComparedAndFreed(Value *V, SmallVectorImpl<WeakVH> &Users,
       }
     }
     if (IntrinsicInst *II = dyn_cast<IntrinsicInst>(U)) {
-      if (II->getIntrinsicID() == Intrinsic::lifetime_start ||
+      switch (II->getIntrinsicID()) {
-          II->getIntrinsicID() == Intrinsic::lifetime_end) {
+      default: return false;
+      case Intrinsic::memmove:
+      case Intrinsic::memcpy:
+      case Intrinsic::memset: {
+        MemIntrinsic *MI = cast<MemIntrinsic>(II);
+        if (MI->isVolatile() || MI->getRawDest() != V)
+          return false;
+      }
+      // fall through
+      case Intrinsic::objectsize:
+      case Intrinsic::lifetime_start:
+      case Intrinsic::lifetime_end:
         Users.push_back(II);
         continue;
       }
     }
+    if (StoreInst *SI = dyn_cast<StoreInst>(U)) {
+      if (SI->isVolatile() || SI->getPointerOperand() != V)
+        return false;
+      Users.push_back(SI);
+      continue;
+    }
     return false;
   }
   return true;
@@ -1164,6 +1181,12 @@ Instruction *InstCombiner::visitMalloc(Instruction &MI) {
                                              C->isFalseWhenEqual()));
       } else if (isa<BitCastInst>(I) || isa<GetElementPtrInst>(I)) {
         ReplaceInstUsesWith(*I, UndefValue::get(I->getType()));
+      } else if (IntrinsicInst *II = dyn_cast<IntrinsicInst>(I)) {
+        if (II->getIntrinsicID() == Intrinsic::objectsize) {
+          ConstantInt *CI = cast<ConstantInt>(II->getArgOperand(1));
+          uint64_t DontKnow = CI->isZero() ? -1ULL : 0;
+          ReplaceInstUsesWith(*I, ConstantInt::get(I->getType(), DontKnow));
+        }
       }
       EraseInstFromFunction(*I);
     }

llvm/test/Transforms/InstCombine/badmalloc.ll

@@ -16,5 +16,26 @@ define i1 @test1() {
   ret i1 %B
 
 ; CHECK: @test1
-; CHECK: ret i1 %B
+; CHECK: ret i1 false
+}
+
+; CHECK: @test2
+define noalias i8* @test2() nounwind {
+entry:
+; CHECK: @malloc
+  %A = call noalias i8* @malloc(i64 4) nounwind
+; CHECK: icmp eq
+  %tobool = icmp eq i8* %A, null
+; CHECK: br i1
+  br i1 %tobool, label %return, label %if.end
+
+if.end:
+; CHECK: store
+  store i8 7, i8* %A
+  br label %return
+
+return:
+; CHECK: phi
+  %retval.0 = phi i8* [ %A, %if.end ], [ null, %entry ]
+  ret i8* %retval.0
 }

llvm/test/Transforms/InstCombine/invoke.ll

@@ -4,6 +4,7 @@ target datalayout = "e-p:64:64:64-i1:8:8-i8:8:8-i16:16:16-i32:32:32-i64:64:64-f3
 declare i32 @__gxx_personality_v0(...)
 declare void @__cxa_call_unexpected(i8*)
 declare i64 @llvm.objectsize.i64(i8*, i1) nounwind readonly
+declare i8* @_Znwm(i64)
 
 
 ; CHECK: @f1
@@ -45,3 +46,20 @@ lpad:
   tail call void @__cxa_call_unexpected(i8* %2) noreturn nounwind
   unreachable
 }
+
+; CHECK: @f3
+define void @f3() nounwind uwtable ssp {
+; CHECK: invoke void @llvm.donothing()
+  %call = invoke noalias i8* @_Znwm(i64 13)
+          to label %invoke.cont unwind label %lpad
+
+invoke.cont:
+  ret void
+
+lpad:
+  %1 = landingpad { i8*, i32 } personality i8* bitcast (i32 (...)* @__gxx_personality_v0 to i8*)
+          filter [0 x i8*] zeroinitializer
+  %2 = extractvalue { i8*, i32 } %1, 0
+  tail call void @__cxa_call_unexpected(i8* %2) noreturn nounwind
+  unreachable
+}

llvm/test/Transforms/InstCombine/malloc-free-delete.ll

@@ -1,17 +1,17 @@
 ; RUN: opt < %s -instcombine -S | FileCheck %s
 ; PR1201
 define i32 @main(i32 %argc, i8** %argv) {
+; CHECK: @main
     %c_19 = alloca i8*
     %malloc_206 = tail call i8* @malloc(i32 mul (i32 ptrtoint (i8* getelementptr (i8* null, i32 1) to i32), i32 10))
     store i8* %malloc_206, i8** %c_19
     %tmp_207 = load i8** %c_19
     tail call void @free(i8* %tmp_207)
     ret i32 0
-; CHECK-NOT: malloc
+; CHECK-NEXT: ret i32 0
-; CHECK-NOT: free
-; CHECK: ret i32 0
 }
 
+declare noalias i8* @calloc(i32, i32) nounwind
 declare noalias i8* @malloc(i32)
 declare void @free(i8*)
 
@@ -26,13 +26,24 @@ define i1 @foo() {
 
 declare void @llvm.lifetime.start(i64, i8*)
 declare void @llvm.lifetime.end(i64, i8*)
+declare i64 @llvm.objectsize.i64(i8*, i1)
+declare void @llvm.memcpy.p0i8.p0i8.i32(i8* nocapture, i8* nocapture, i32, i32, i1) nounwind
+declare void @llvm.memmove.p0i8.p0i8.i32(i8* nocapture, i8* nocapture, i32, i32, i1) nounwind
+declare void @llvm.memset.p0i8.i32(i8*, i8, i32, i32, i1) nounwind
 
-define void @test3() {
+define void @test3(i8* %src) {
 ; CHECK: @test3
 ; CHECK-NEXT: ret void
   %a = call noalias i8* @malloc(i32 10)
   call void @llvm.lifetime.start(i64 10, i8* %a)
   call void @llvm.lifetime.end(i64 10, i8* %a)
+  %size = call i64 @llvm.objectsize.i64(i8* %a, i1 true)
+  store i8 42, i8* %a
+  call void @llvm.memcpy.p0i8.p0i8.i32(i8* %a, i8* %src, i32 32, i32 1, i1 false)
+  call void @llvm.memmove.p0i8.p0i8.i32(i8* %a, i8* %src, i32 32, i32 1, i1 false)
+  call void @llvm.memset.p0i8.i32(i8* %a, i8 5, i32 32, i32 1, i1 false)
+  %alloc2 = call noalias i8* @calloc(i32 5, i32 7) nounwind
+  %z = icmp ne i8* %alloc2, null
   ret void
 }
 
@@ -46,3 +57,37 @@ define void @test4() {
   call void @free(i8* %C)
   ret void
 }
+
+; CHECK: @test5
+define void @test5(i8* %ptr, i8** %esc) {
+; CHECK-NEXT: call i8* @malloc
+; CHECK-NEXT: call i8* @malloc
+; CHECK-NEXT: call i8* @malloc
+; CHECK-NEXT: call i8* @malloc
+; CHECK-NEXT: call i8* @malloc
+; CHECK-NEXT: call i8* @malloc
+; CHECK-NEXT: call i8* @malloc
+; CHECK-NEXT: call void @llvm.memcpy
+; CHECK-NEXT: call void @llvm.memmove
+; CHECK-NEXT: store
+; CHECK-NEXT: call void @llvm.memcpy
+; CHECK-NEXT: call void @llvm.memmove
+; CHECK-NEXT: call void @llvm.memset
+; CHECK-NEXT: store volatile
+; CHECK-NEXT: ret
+  %a = call i8* @malloc(i32 700)
+  %b = call i8* @malloc(i32 700)
+  %c = call i8* @malloc(i32 700)
+  %d = call i8* @malloc(i32 700)
+  %e = call i8* @malloc(i32 700)
+  %f = call i8* @malloc(i32 700)
+  %g = call i8* @malloc(i32 700)
+  call void @llvm.memcpy.p0i8.p0i8.i32(i8* %ptr, i8* %a, i32 32, i32 1, i1 false)
+  call void @llvm.memmove.p0i8.p0i8.i32(i8* %ptr, i8* %b, i32 32, i32 1, i1 false)
+  store i8* %c, i8** %esc
+  call void @llvm.memcpy.p0i8.p0i8.i32(i8* %d, i8* %ptr, i32 32, i32 1, i1 true)
+  call void @llvm.memmove.p0i8.p0i8.i32(i8* %e, i8* %ptr, i32 32, i32 1, i1 true)
+  call void @llvm.memset.p0i8.i32(i8* %f, i8 5, i32 32, i32 1, i1 true)
+  store volatile i8 4, i8* %g
+  ret void
+}

llvm/test/Transforms/InstCombine/objsize-64.ll

@@ -8,23 +8,25 @@ declare void @__cxa_call_unexpected(i8*)
 declare i64 @llvm.objectsize.i64(i8*, i1) nounwind readonly
 
 ; CHECK: @f1
-define i64 @f1() {
+define i64 @f1(i8 **%esc) {
   %call = call i8* @malloc(i32 4)
+  store i8* %call, i8** %esc
   %size = call i64 @llvm.objectsize.i64(i8* %call, i1 false)
-; CHECK-NEXT: ret i64 4
+; CHECK: ret i64 4
   ret i64 %size
 }
 
 
 ; CHECK: @f2
-define i64 @f2() nounwind uwtable ssp {
+define i64 @f2(i8** %esc) nounwind uwtable ssp {
 entry:
-; CHECK: invoke void @llvm.donothing()
+; CHECK: invoke noalias i8* @_Znwm(i64 13)
   %call = invoke noalias i8* @_Znwm(i64 13)
           to label %invoke.cont unwind label %lpad
 
 invoke.cont:
 ; CHECK: ret i64 13
+  store i8* %call, i8** %esc
   %0 = tail call i64 @llvm.objectsize.i64(i8* %call, i1 false)
   ret i64 %0
 

llvm/test/Transforms/InstCombine/objsize.ll

@@ -121,7 +121,7 @@ entry:
 ; rdar://7782496
 @s = external global i8*
 
-define void @test5(i32 %n) nounwind ssp {
+define i8* @test5(i32 %n) nounwind ssp {
 ; CHECK: @test5
 entry:
   %0 = tail call noalias i8* @malloc(i32 20) nounwind
@@ -130,7 +130,7 @@ entry:
 ; CHECK-NOT: @llvm.objectsize
 ; CHECK: @llvm.memcpy.p0i8.p0i8.i32(i8* %0, i8* %1, i32 10, i32 1, i1 false)
   %3 = tail call i8* @__memcpy_chk(i8* %0, i8* %2, i32 10, i32 %1) nounwind
-  ret void
+  ret i8* %0
 }
 
 define void @test6(i32 %n) nounwind ssp {
@@ -149,22 +149,24 @@ declare i8* @__memset_chk(i8*, i32, i32, i32) nounwind
 
 declare noalias i8* @malloc(i32) nounwind
 
-define i32 @test7() {
+define i32 @test7(i8** %esc) {
 ; CHECK: @test7
   %alloc = call noalias i8* @malloc(i32 48) nounwind
+  store i8* %alloc, i8** %esc
   %gep = getelementptr inbounds i8* %alloc, i32 16
   %objsize = call i32 @llvm.objectsize.i32(i8* %gep, i1 false) nounwind readonly
-; CHECK-NEXT: ret i32 32
+; CHECK: ret i32 32
   ret i32 %objsize
 }
 
 declare noalias i8* @calloc(i32, i32) nounwind
 
-define i32 @test8() {
+define i32 @test8(i8** %esc) {
 ; CHECK: @test8
   %alloc = call noalias i8* @calloc(i32 5, i32 7) nounwind
+  store i8* %alloc, i8** %esc
   %gep = getelementptr inbounds i8* %alloc, i32 5
   %objsize = call i32 @llvm.objectsize.i32(i8* %gep, i1 false) nounwind readonly
-; CHECK-NEXT: ret i32 30
+; CHECK: ret i32 30
   ret i32 %objsize
 }