From 1ae1b9397e9e60438c22e240fb989ecfa39bb527 Mon Sep 17 00:00:00 2001
From: Tim Northover <tnorthover@apple.com>
Date: Mon, 30 Jan 2017 19:33:07 +0000
Subject: [PATCH] GlobalISel: translate memset & memmove.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@293541 91177308-0d34-0410-b5e6-96231b3b80d8
---
 .../llvm/CodeGen/GlobalISel/IRTranslator.h    |  4 ++-
 lib/CodeGen/GlobalISel/IRTranslator.cpp       | 34 ++++++++++++++-----
 .../AArch64/GlobalISel/arm64-irtranslator.ll  | 28 +++++++++++++++
 3 files changed, 56 insertions(+), 10 deletions(-)

diff --git a/include/llvm/CodeGen/GlobalISel/IRTranslator.h b/include/llvm/CodeGen/GlobalISel/IRTranslator.h
index 07df370ee59..25e8588c437 100644
--- a/include/llvm/CodeGen/GlobalISel/IRTranslator.h
+++ b/include/llvm/CodeGen/GlobalISel/IRTranslator.h
@@ -130,7 +130,9 @@ private:
   /// Translate an LLVM store instruction into generic IR.
   bool translateStore(const User &U, MachineIRBuilder &MIRBuilder);
 
-  bool translateMemcpy(const CallInst &CI, MachineIRBuilder &MIRBuilder);
+  /// Translate an LLVM string intrinsic (memcpy, memset, ...).
+  bool translateMemfunc(const CallInst &CI, MachineIRBuilder &MIRBuilder,
+                        unsigned Intrinsic);
 
   void getStackGuard(unsigned DstReg, MachineIRBuilder &MIRBuilder);
 
diff --git a/lib/CodeGen/GlobalISel/IRTranslator.cpp b/lib/CodeGen/GlobalISel/IRTranslator.cpp
index 58bd65ef323..3a2469dca49 100644
--- a/lib/CodeGen/GlobalISel/IRTranslator.cpp
+++ b/lib/CodeGen/GlobalISel/IRTranslator.cpp
@@ -468,13 +468,12 @@ bool IRTranslator::translateGetElementPtr(const User &U,
   return true;
 }
 
-bool IRTranslator::translateMemcpy(const CallInst &CI,
-                                   MachineIRBuilder &MIRBuilder) {
+bool IRTranslator::translateMemfunc(const CallInst &CI,
+                                    MachineIRBuilder &MIRBuilder,
+                                    unsigned ID) {
   LLT SizeTy{*CI.getArgOperand(2)->getType(), *DL};
-  if (cast<PointerType>(CI.getArgOperand(0)->getType())->getAddressSpace() !=
-          0 ||
-      cast<PointerType>(CI.getArgOperand(1)->getType())->getAddressSpace() !=
-          0 ||
+  Type *DstTy = CI.getArgOperand(0)->getType();
+  if (cast<PointerType>(DstTy)->getAddressSpace() != 0 ||
       SizeTy.getSizeInBits() != DL->getPointerSizeInBits(0))
     return false;
 
@@ -484,9 +483,24 @@ bool IRTranslator::translateMemcpy(const CallInst &CI,
     Args.emplace_back(getOrCreateVReg(*Arg), Arg->getType());
   }
 
-  MachineOperand Callee = MachineOperand::CreateES("memcpy");
+  const char *Callee;
+  switch (ID) {
+  case Intrinsic::memmove:
+  case Intrinsic::memcpy: {
+    Type *SrcTy = CI.getArgOperand(1)->getType();
+    if(cast<PointerType>(SrcTy)->getAddressSpace() != 0)
+      return false;
+    Callee = ID == Intrinsic::memcpy ? "memcpy" : "memmove";
+    break;
+  }
+  case Intrinsic::memset:
+    Callee = "memset";
+    break;
+  default:
+    return false;
+  }
 
-  return CLI->lowerCall(MIRBuilder, Callee,
+  return CLI->lowerCall(MIRBuilder, MachineOperand::CreateES(Callee),
                         CallLowering::ArgInfo(0, CI.getType()), Args);
 }
 
@@ -607,7 +621,9 @@ bool IRTranslator::translateKnownIntrinsic(const CallInst &CI, Intrinsic::ID ID,
   case Intrinsic::smul_with_overflow:
     return translateOverflowIntrinsic(CI, TargetOpcode::G_SMULO, MIRBuilder);
   case Intrinsic::memcpy:
-    return translateMemcpy(CI, MIRBuilder);
+  case Intrinsic::memmove:
+  case Intrinsic::memset:
+    return translateMemfunc(CI, MIRBuilder, ID);
   case Intrinsic::eh_typeid_for: {
     GlobalValue *GV = ExtractTypeInfo(CI.getArgOperand(0));
     unsigned Reg = getOrCreateVReg(CI);
diff --git a/test/CodeGen/AArch64/GlobalISel/arm64-irtranslator.ll b/test/CodeGen/AArch64/GlobalISel/arm64-irtranslator.ll
index 3938a3a8961..bac3017a925 100644
--- a/test/CodeGen/AArch64/GlobalISel/arm64-irtranslator.ll
+++ b/test/CodeGen/AArch64/GlobalISel/arm64-irtranslator.ll
@@ -1054,6 +1054,34 @@ define void @test_memcpy(i8* %dst, i8* %src, i64 %size) {
   ret void
 }
 
+declare void @llvm.memmove.p0i8.p0i8.i64(i8*, i8*, i64, i32 %align, i1 %volatile)
+define void @test_memmove(i8* %dst, i8* %src, i64 %size) {
+; CHECK-LABEL: name: test_memmove
+; CHECK: [[DST:%[0-9]+]](p0) = COPY %x0
+; CHECK: [[SRC:%[0-9]+]](p0) = COPY %x1
+; CHECK: [[SIZE:%[0-9]+]](s64) = COPY %x2
+; CHECK: %x0 = COPY [[DST]]
+; CHECK: %x1 = COPY [[SRC]]
+; CHECK: %x2 = COPY [[SIZE]]
+; CHECK: BL $memmove, csr_aarch64_aapcs, implicit-def %lr, implicit %sp, implicit %x0, implicit %x1, implicit %x2
+  call void @llvm.memmove.p0i8.p0i8.i64(i8* %dst, i8* %src, i64 %size, i32 1, i1 0)
+  ret void
+}
+
+declare void @llvm.memset.p0i8.i64(i8*, i8, i64, i32 %align, i1 %volatile)
+define void @test_memset(i8* %dst, i8 %val, i64 %size) {
+; CHECK-LABEL: name: test_memset
+; CHECK: [[DST:%[0-9]+]](p0) = COPY %x0
+; CHECK: [[SRC:%[0-9]+]](s8) = COPY %w1
+; CHECK: [[SIZE:%[0-9]+]](s64) = COPY %x2
+; CHECK: %x0 = COPY [[DST]]
+; CHECK: %w1 = COPY [[SRC]]
+; CHECK: %x2 = COPY [[SIZE]]
+; CHECK: BL $memset, csr_aarch64_aapcs, implicit-def %lr, implicit %sp, implicit %x0, implicit %w1, implicit %x2
+  call void @llvm.memset.p0i8.i64(i8* %dst, i8 %val, i64 %size, i32 1, i1 0)
+  ret void
+}
+
 declare i64 @llvm.objectsize.i64(i8*, i1)
 declare i32 @llvm.objectsize.i32(i8*, i1)
 define void @test_objectsize(i8* %addr0, i8* %addr1) {
-- 
2.40.0