[llvm] r351920 - hwasan: Move memory access checks into small outlined functions on aarch64.

Peter Collingbourne via llvm-commits llvm-commits at lists.llvm.org
Tue Jan 22 18:20:11 PST 2019


Author: pcc
Date: Tue Jan 22 18:20:10 2019
New Revision: 351920

URL: http://llvm.org/viewvc/llvm-project?rev=351920&view=rev
Log:
hwasan: Move memory access checks into small outlined functions on aarch64.

Each hwasan check requires emitting a small piece of code like this:
https://clang.llvm.org/docs/HardwareAssistedAddressSanitizerDesign.html#memory-accesses

The problem with this is that these code blocks typically bloat code
size significantly.

An obvious solution is to outline these blocks of code. In fact, this
has already been implemented under the -hwasan-instrument-with-calls
flag. However, as currently implemented this has a number of problems:
- The functions use the same calling convention as regular C functions.
  This means that the backend must spill all temporary registers as
  required by the platform's C calling convention, even though the
  check only needs two registers on the hot path.
- The functions take the address to be checked in a fixed register,
  which increases register pressure.
Both of these factors can diminish the code size effect and increase
the performance hit of -hwasan-instrument-with-calls.

The solution that this patch implements is to involve the aarch64
backend in outlining the checks. An intrinsic and pseudo-instruction
are created to represent a hwasan check. The pseudo-instruction
is register allocated like any other instruction, and we allow the
register allocator to select almost any register for the address to
check. A particular combination of (register selection, type of check)
triggers the creation in the backend of a function to handle the check
for specifically that pair. The resulting functions are deduplicated by
the linker. The pseudo-instruction (really the function) is specified
to preserve all registers except for the registers that the AAPCS
specifies may be clobbered by a call.

To measure the code size and performance effect of this change, I
took a number of measurements using Chromium for Android on aarch64,
comparing a browser with inlined checks (the baseline) against a
browser with outlined checks.

Code size: Size of .text decreases from 243897420 to 171619972 bytes,
or a 30% decrease.

Performance: Using Chromium's blink_perf.layout microbenchmarks I
measured a median performance regression of 6.24%.

The fact that a perf/size tradeoff is evident here suggests that
we might want to make the new behaviour conditional on -Os/-Oz.
But for now I've enabled it unconditionally, my reasoning being that
hwasan users typically expect a relatively large perf hit, and ~6%
isn't really adding much. We may want to revisit this decision in
the future, though.

I also tried experimenting with varying the number of registers
selectable by the hwasan check pseudo-instruction (which would result
in fewer variants being created), on the hypothesis that creating
fewer variants of the function would expose another perf/size tradeoff
by reducing icache pressure from the check functions at the cost of
register pressure. Although I did observe a code size increase with
fewer registers, I did not observe a strong correlation between the
number of registers and the performance of the resulting browser on the
microbenchmarks, so I conclude that we might as well use ~all registers
to get the maximum code size improvement. My results are below:

Regs | .text size | Perf hit
-----+------------+---------
~all | 171619972  | 6.24%
  16 | 171765192  | 7.03%
   8 | 172917788  | 5.82%
   4 | 177054016  | 6.89%

Differential Revision: https://reviews.llvm.org/D56954

Added:
    llvm/trunk/test/CodeGen/AArch64/hwasan-check-memaccess.ll
Modified:
    llvm/trunk/include/llvm/IR/Intrinsics.td
    llvm/trunk/lib/Target/AArch64/AArch64AsmPrinter.cpp
    llvm/trunk/lib/Target/AArch64/AArch64InstrInfo.td
    llvm/trunk/lib/Target/AArch64/AArch64RegisterBankInfo.cpp
    llvm/trunk/lib/Target/AArch64/AArch64RegisterInfo.td
    llvm/trunk/lib/Transforms/Instrumentation/HWAddressSanitizer.cpp
    llvm/trunk/test/Instrumentation/HWAddressSanitizer/alloca.ll
    llvm/trunk/test/Instrumentation/HWAddressSanitizer/atomic.ll
    llvm/trunk/test/Instrumentation/HWAddressSanitizer/basic.ll
    llvm/trunk/test/Instrumentation/HWAddressSanitizer/kernel.ll
    llvm/trunk/test/Instrumentation/HWAddressSanitizer/prologue.ll

Modified: llvm/trunk/include/llvm/IR/Intrinsics.td
URL: http://llvm.org/viewvc/llvm-project/llvm/trunk/include/llvm/IR/Intrinsics.td?rev=351920&r1=351919&r2=351920&view=diff
==============================================================================
--- llvm/trunk/include/llvm/IR/Intrinsics.td (original)
+++ llvm/trunk/include/llvm/IR/Intrinsics.td Tue Jan 22 18:20:10 2019
@@ -1048,6 +1048,9 @@ def int_icall_branch_funnel : Intrinsic<
 def int_load_relative: Intrinsic<[llvm_ptr_ty], [llvm_ptr_ty, llvm_anyint_ty],
                                  [IntrReadMem, IntrArgMemOnly]>;
 
+def int_hwasan_check_memaccess :
+  Intrinsic<[], [llvm_ptr_ty, llvm_ptr_ty, llvm_i32_ty], [IntrInaccessibleMemOnly]>;
+
 // Xray intrinsics
 //===----------------------------------------------------------------------===//
 // Custom event logging for x-ray.

Modified: llvm/trunk/lib/Target/AArch64/AArch64AsmPrinter.cpp
URL: http://llvm.org/viewvc/llvm-project/llvm/trunk/lib/Target/AArch64/AArch64AsmPrinter.cpp?rev=351920&r1=351919&r2=351920&view=diff
==============================================================================
--- llvm/trunk/lib/Target/AArch64/AArch64AsmPrinter.cpp (original)
+++ llvm/trunk/lib/Target/AArch64/AArch64AsmPrinter.cpp Tue Jan 22 18:20:10 2019
@@ -28,6 +28,7 @@
 #include "llvm/ADT/Triple.h"
 #include "llvm/ADT/Twine.h"
 #include "llvm/BinaryFormat/COFF.h"
+#include "llvm/BinaryFormat/ELF.h"
 #include "llvm/CodeGen/AsmPrinter.h"
 #include "llvm/CodeGen/MachineBasicBlock.h"
 #include "llvm/CodeGen/MachineFunction.h"
@@ -43,6 +44,7 @@
 #include "llvm/MC/MCContext.h"
 #include "llvm/MC/MCInst.h"
 #include "llvm/MC/MCInstBuilder.h"
+#include "llvm/MC/MCSectionELF.h"
 #include "llvm/MC/MCStreamer.h"
 #include "llvm/MC/MCSymbol.h"
 #include "llvm/Support/Casting.h"
@@ -95,6 +97,10 @@ public:
   void LowerPATCHABLE_FUNCTION_EXIT(const MachineInstr &MI);
   void LowerPATCHABLE_TAIL_CALL(const MachineInstr &MI);
 
+  std::map<std::pair<unsigned, uint32_t>, MCSymbol *> HwasanMemaccessSymbols;
+  void LowerHWASAN_CHECK_MEMACCESS(const MachineInstr &MI);
+  void EmitHwasanMemaccessSymbols(Module &M);
+
   void EmitSled(const MachineInstr &MI, SledKind Kind);
 
   /// tblgen'erated driver function for lowering simple MI->MC
@@ -229,7 +235,109 @@ void AArch64AsmPrinter::EmitSled(const M
   recordSled(CurSled, MI, Kind);
 }
 
+void AArch64AsmPrinter::LowerHWASAN_CHECK_MEMACCESS(const MachineInstr &MI) {
+  unsigned Reg = MI.getOperand(0).getReg();
+  uint32_t AccessInfo = MI.getOperand(1).getImm();
+  MCSymbol *&Sym = HwasanMemaccessSymbols[{Reg, AccessInfo}];
+  if (!Sym) {
+    // FIXME: Make this work on non-ELF.
+    if (!TM.getTargetTriple().isOSBinFormatELF())
+      report_fatal_error("llvm.hwasan.check.memaccess only supported on ELF");
+
+    std::string SymName = "__hwasan_check_x" + utostr(Reg - AArch64::X0) + "_" +
+                          utostr(AccessInfo);
+    Sym = OutContext.getOrCreateSymbol(SymName);
+  }
+
+  EmitToStreamer(*OutStreamer,
+                 MCInstBuilder(AArch64::BL)
+                     .addExpr(MCSymbolRefExpr::create(Sym, OutContext)));
+}
+
+void AArch64AsmPrinter::EmitHwasanMemaccessSymbols(Module &M) {
+  if (HwasanMemaccessSymbols.empty())
+    return;
+
+  const Triple &TT = TM.getTargetTriple();
+  assert(TT.isOSBinFormatELF());
+  std::unique_ptr<MCSubtargetInfo> STI(
+      TM.getTarget().createMCSubtargetInfo(TT.str(), "", ""));
+
+  MCSymbol *HwasanTagMismatchSym =
+      OutContext.getOrCreateSymbol("__hwasan_tag_mismatch");
+
+  for (auto &P : HwasanMemaccessSymbols) {
+    unsigned Reg = P.first.first;
+    uint32_t AccessInfo = P.first.second;
+    MCSymbol *Sym = P.second;
+
+    OutStreamer->SwitchSection(OutContext.getELFSection(
+        ".text.hot", ELF::SHT_PROGBITS,
+        ELF::SHF_EXECINSTR | ELF::SHF_ALLOC | ELF::SHF_GROUP, 0,
+        Sym->getName()));
+
+    OutStreamer->EmitSymbolAttribute(Sym, MCSA_ELF_TypeFunction);
+    OutStreamer->EmitSymbolAttribute(Sym, MCSA_Weak);
+    OutStreamer->EmitSymbolAttribute(Sym, MCSA_Hidden);
+    OutStreamer->EmitLabel(Sym);
+
+    OutStreamer->EmitInstruction(MCInstBuilder(AArch64::UBFMXri)
+                                     .addReg(AArch64::X16)
+                                     .addReg(Reg)
+                                     .addImm(4)
+                                     .addImm(55),
+                                 *STI);
+    OutStreamer->EmitInstruction(MCInstBuilder(AArch64::LDRBBroX)
+                                     .addReg(AArch64::W16)
+                                     .addReg(AArch64::X9)
+                                     .addReg(AArch64::X16)
+                                     .addImm(0)
+                                     .addImm(0),
+                                 *STI);
+    OutStreamer->EmitInstruction(MCInstBuilder(AArch64::UBFMXri)
+                                     .addReg(AArch64::X17)
+                                     .addReg(Reg)
+                                     .addImm(56)
+                                     .addImm(63),
+                                 *STI);
+    OutStreamer->EmitInstruction(MCInstBuilder(AArch64::SUBSWrs)
+                                     .addReg(AArch64::WZR)
+                                     .addReg(AArch64::W16)
+                                     .addReg(AArch64::W17)
+                                     .addImm(0),
+                                 *STI);
+    MCSymbol *HandleMismatchSym = OutContext.createTempSymbol();
+    OutStreamer->EmitInstruction(
+        MCInstBuilder(AArch64::Bcc)
+            .addImm(AArch64CC::NE)
+            .addExpr(MCSymbolRefExpr::create(HandleMismatchSym, OutContext)),
+        *STI);
+    OutStreamer->EmitInstruction(
+        MCInstBuilder(AArch64::RET).addReg(AArch64::LR), *STI);
+
+    OutStreamer->EmitLabel(HandleMismatchSym);
+    if (Reg != AArch64::X0)
+      OutStreamer->EmitInstruction(MCInstBuilder(AArch64::ORRXrs)
+                                       .addReg(AArch64::X0)
+                                       .addReg(AArch64::XZR)
+                                       .addReg(Reg)
+                                       .addImm(0),
+                                   *STI);
+    OutStreamer->EmitInstruction(MCInstBuilder(AArch64::MOVZXi)
+                                     .addReg(AArch64::X1)
+                                     .addImm(AccessInfo)
+                                     .addImm(0),
+                                 *STI);
+    OutStreamer->EmitInstruction(
+        MCInstBuilder(AArch64::B)
+            .addExpr(MCSymbolRefExpr::create(HwasanTagMismatchSym, OutContext)),
+        *STI);
+  }
+}
+
 void AArch64AsmPrinter::EmitEndOfAsmFile(Module &M) {
+  EmitHwasanMemaccessSymbols(M);
+
   const Triple &TT = TM.getTargetTriple();
   if (TT.isOSBinFormatMachO()) {
     // Funny Darwin hack: This flag tells the linker that no global symbols
@@ -883,6 +991,10 @@ void AArch64AsmPrinter::EmitInstruction(
     LowerPATCHABLE_TAIL_CALL(*MI);
     return;
 
+  case AArch64::HWASAN_CHECK_MEMACCESS:
+    LowerHWASAN_CHECK_MEMACCESS(*MI);
+    return;
+
   case AArch64::SEH_StackAlloc:
     TS->EmitARM64WinCFIAllocStack(MI->getOperand(0).getImm());
     return;

Modified: llvm/trunk/lib/Target/AArch64/AArch64InstrInfo.td
URL: http://llvm.org/viewvc/llvm-project/llvm/trunk/lib/Target/AArch64/AArch64InstrInfo.td?rev=351920&r1=351919&r2=351920&view=diff
==============================================================================
--- llvm/trunk/lib/Target/AArch64/AArch64InstrInfo.td (original)
+++ llvm/trunk/lib/Target/AArch64/AArch64InstrInfo.td Tue Jan 22 18:20:10 2019
@@ -763,6 +763,13 @@ def MSRpstateImm4 : MSRpstateImm0_15;
 def MOVbaseTLS : Pseudo<(outs GPR64:$dst), (ins),
                        [(set GPR64:$dst, AArch64threadpointer)]>, Sched<[WriteSys]>;
 
+let Uses = [ X9 ], Defs = [ X16, X17, LR, NZCV ] in {
+def HWASAN_CHECK_MEMACCESS : Pseudo<
+  (outs), (ins GPR64noip:$ptr, i32imm:$accessinfo),
+  [(int_hwasan_check_memaccess X9, GPR64noip:$ptr, (i32 imm:$accessinfo))]>,
+  Sched<[]>;
+}
+
 // The cycle counter PMC register is PMCCNTR_EL0.
 let Predicates = [HasPerfMon] in
 def : Pat<(readcyclecounter), (MRS 0xdce8)>;

Modified: llvm/trunk/lib/Target/AArch64/AArch64RegisterBankInfo.cpp
URL: http://llvm.org/viewvc/llvm-project/llvm/trunk/lib/Target/AArch64/AArch64RegisterBankInfo.cpp?rev=351920&r1=351919&r2=351920&view=diff
==============================================================================
--- llvm/trunk/lib/Target/AArch64/AArch64RegisterBankInfo.cpp (original)
+++ llvm/trunk/lib/Target/AArch64/AArch64RegisterBankInfo.cpp Tue Jan 22 18:20:10 2019
@@ -248,6 +248,9 @@ const RegisterBank &AArch64RegisterBankI
   case AArch64::GPR64spRegClassID:
   case AArch64::GPR64sponlyRegClassID:
   case AArch64::GPR64allRegClassID:
+  case AArch64::GPR64noipRegClassID:
+  case AArch64::GPR64common_and_GPR64noipRegClassID:
+  case AArch64::GPR64noip_and_tcGPR64RegClassID:
   case AArch64::tcGPR64RegClassID:
   case AArch64::WSeqPairsClassRegClassID:
   case AArch64::XSeqPairsClassRegClassID:

Modified: llvm/trunk/lib/Target/AArch64/AArch64RegisterInfo.td
URL: http://llvm.org/viewvc/llvm-project/llvm/trunk/lib/Target/AArch64/AArch64RegisterInfo.td?rev=351920&r1=351919&r2=351920&view=diff
==============================================================================
--- llvm/trunk/lib/Target/AArch64/AArch64RegisterInfo.td (original)
+++ llvm/trunk/lib/Target/AArch64/AArch64RegisterInfo.td Tue Jan 22 18:20:10 2019
@@ -205,6 +205,11 @@ def tcGPR64 : RegisterClass<"AArch64", [
 // BTI-protected function.
 def rtcGPR64 : RegisterClass<"AArch64", [i64], 64, (add X16, X17)>;
 
+// Register set that excludes registers that are reserved for procedure calls.
+// This is used for pseudo-instructions that are actually implemented using a
+// procedure call.
+def GPR64noip : RegisterClass<"AArch64", [i64], 64, (sub GPR64, X16, X17, LR)>;
+
 // GPR register classes for post increment amount of vector load/store that
 // has alternate printing when Rm=31 and prints a constant immediate value
 // equal to the total number of bytes transferred.

Modified: llvm/trunk/lib/Transforms/Instrumentation/HWAddressSanitizer.cpp
URL: http://llvm.org/viewvc/llvm-project/llvm/trunk/lib/Transforms/Instrumentation/HWAddressSanitizer.cpp?rev=351920&r1=351919&r2=351920&view=diff
==============================================================================
--- llvm/trunk/lib/Transforms/Instrumentation/HWAddressSanitizer.cpp (original)
+++ llvm/trunk/lib/Transforms/Instrumentation/HWAddressSanitizer.cpp Tue Jan 22 18:20:10 2019
@@ -155,6 +155,11 @@ static cl::opt<bool>
     ClInstrumentMemIntrinsics("hwasan-instrument-mem-intrinsics",
                               cl::desc("instrument memory intrinsics"),
                               cl::Hidden, cl::init(true));
+
+static cl::opt<bool> ClInlineAllChecks("hwasan-inline-all-checks",
+                                       cl::desc("inline all checks"),
+                                       cl::Hidden, cl::init(false));
+
 namespace {
 
 /// An instrumentation pass implementing detection of addressability bugs
@@ -181,8 +186,9 @@ public:
   Value *getDynamicShadowNonTls(IRBuilder<> &IRB);
 
   void untagPointerOperand(Instruction *I, Value *Addr);
-  Value *memToShadow(Value *Shadow, Type *Ty, IRBuilder<> &IRB);
-  void instrumentMemAccessInline(Value *PtrLong, bool IsWrite,
+  Value *shadowBase();
+  Value *memToShadow(Value *Shadow, IRBuilder<> &IRB);
+  void instrumentMemAccessInline(Value *Ptr, bool IsWrite,
                                  unsigned AccessSizeIndex,
                                  Instruction *InsertBefore);
   void instrumentMemIntrinsic(MemIntrinsic *MI);
@@ -252,6 +258,7 @@ private:
   Type *IntptrTy;
   Type *Int8PtrTy;
   Type *Int8Ty;
+  Type *Int32Ty;
 
   bool CompileKernel;
   bool Recover;
@@ -307,6 +314,7 @@ bool HWAddressSanitizer::doInitializatio
   IntptrTy = IRB.getIntPtrTy(DL);
   Int8PtrTy = IRB.getInt8PtrTy();
   Int8Ty = IRB.getInt8Ty();
+  Int32Ty = IRB.getInt32Ty();
 
   HwasanCtorFunction = nullptr;
   if (!CompileKernel) {
@@ -403,16 +411,16 @@ Value *HWAddressSanitizer::getDynamicSha
 
   if (Mapping.InGlobal) {
     // An empty inline asm with input reg == output reg.
-    // An opaque pointer-to-int cast, basically.
+    // An opaque no-op cast, basically.
     InlineAsm *Asm = InlineAsm::get(
-        FunctionType::get(IntptrTy, {ShadowGlobal->getType()}, false),
+        FunctionType::get(Int8PtrTy, {ShadowGlobal->getType()}, false),
         StringRef(""), StringRef("=r,0"),
         /*hasSideEffects=*/false);
     return IRB.CreateCall(Asm, {ShadowGlobal}, ".hwasan.shadow");
   } else {
     Value *GlobalDynamicAddress =
         IRB.GetInsertBlock()->getParent()->getParent()->getOrInsertGlobal(
-            kHwasanShadowMemoryDynamicAddress, IntptrTy);
+            kHwasanShadowMemoryDynamicAddress, Int8PtrTy);
     return IRB.CreateLoad(GlobalDynamicAddress);
   }
 }
@@ -505,29 +513,44 @@ void HWAddressSanitizer::untagPointerOpe
   I->setOperand(getPointerOperandIndex(I), UntaggedPtr);
 }
 
-Value *HWAddressSanitizer::memToShadow(Value *Mem, Type *Ty, IRBuilder<> &IRB) {
+Value *HWAddressSanitizer::shadowBase() {
+  if (LocalDynamicShadow)
+    return LocalDynamicShadow;
+  return ConstantExpr::getIntToPtr(ConstantInt::get(IntptrTy, Mapping.Offset),
+                                   Int8PtrTy);
+}
+
+Value *HWAddressSanitizer::memToShadow(Value *Mem, IRBuilder<> &IRB) {
   // Mem >> Scale
   Value *Shadow = IRB.CreateLShr(Mem, Mapping.Scale);
   if (Mapping.Offset == 0)
-    return Shadow;
+    return IRB.CreateIntToPtr(Shadow, Int8PtrTy);
   // (Mem >> Scale) + Offset
-  Value *ShadowBase;
-  if (LocalDynamicShadow)
-    ShadowBase = LocalDynamicShadow;
-  else
-    ShadowBase = ConstantInt::get(Ty, Mapping.Offset);
-  return IRB.CreateAdd(Shadow, ShadowBase);
+  return IRB.CreateGEP(Int8Ty, shadowBase(), Shadow);
 }
 
-void HWAddressSanitizer::instrumentMemAccessInline(Value *PtrLong, bool IsWrite,
+void HWAddressSanitizer::instrumentMemAccessInline(Value *Ptr, bool IsWrite,
                                                    unsigned AccessSizeIndex,
                                                    Instruction *InsertBefore) {
+  const int64_t AccessInfo = Recover * 0x20 + IsWrite * 0x10 + AccessSizeIndex;
   IRBuilder<> IRB(InsertBefore);
+
+  if (!ClInlineAllChecks && TargetTriple.isAArch64() &&
+      TargetTriple.isOSBinFormatELF() && !Recover) {
+    Module *M = IRB.GetInsertBlock()->getParent()->getParent();
+    Ptr = IRB.CreateBitCast(Ptr, Int8PtrTy);
+    IRB.CreateCall(
+        Intrinsic::getDeclaration(M, Intrinsic::hwasan_check_memaccess),
+        {shadowBase(), Ptr, ConstantInt::get(Int32Ty, AccessInfo)});
+    return;
+  }
+
+  Value *PtrLong = IRB.CreatePointerCast(Ptr, IntptrTy);
   Value *PtrTag = IRB.CreateTrunc(IRB.CreateLShr(PtrLong, kPointerTagShift),
                                   IRB.getInt8Ty());
   Value *AddrLong = untagPointer(IRB, PtrLong);
-  Value *ShadowLong = memToShadow(AddrLong, PtrLong->getType(), IRB);
-  Value *MemTag = IRB.CreateLoad(IRB.CreateIntToPtr(ShadowLong, Int8PtrTy));
+  Value *Shadow = memToShadow(AddrLong, IRB);
+  Value *MemTag = IRB.CreateLoad(Shadow);
   Value *TagMismatch = IRB.CreateICmpNE(PtrTag, MemTag);
 
   int matchAllTag = ClMatchAllTag.getNumOccurrences() > 0 ?
@@ -543,7 +566,6 @@ void HWAddressSanitizer::instrumentMemAc
                                 MDBuilder(*C).createBranchWeights(1, 100000));
 
   IRB.SetInsertPoint(CheckTerm);
-  const int64_t AccessInfo = Recover * 0x20 + IsWrite * 0x10 + AccessSizeIndex;
   InlineAsm *Asm;
   switch (TargetTriple.getArch()) {
     case Triple::x86_64:
@@ -609,7 +631,6 @@ bool HWAddressSanitizer::instrumentMemAc
     return false; //FIXME
 
   IRBuilder<> IRB(I);
-  Value *AddrLong = IRB.CreatePointerCast(Addr, IntptrTy);
   if (isPowerOf2_64(TypeSize) &&
       (TypeSize / 8 <= (1UL << (kNumberOfAccessSizes - 1))) &&
       (Alignment >= (1UL << Mapping.Scale) || Alignment == 0 ||
@@ -617,13 +638,14 @@ bool HWAddressSanitizer::instrumentMemAc
     size_t AccessSizeIndex = TypeSizeToSizeIndex(TypeSize);
     if (ClInstrumentWithCalls) {
       IRB.CreateCall(HwasanMemoryAccessCallback[IsWrite][AccessSizeIndex],
-                     AddrLong);
+                     IRB.CreatePointerCast(Addr, IntptrTy));
     } else {
-      instrumentMemAccessInline(AddrLong, IsWrite, AccessSizeIndex, I);
+      instrumentMemAccessInline(Addr, IsWrite, AccessSizeIndex, I);
     }
   } else {
     IRB.CreateCall(HwasanMemoryAccessCallbackSized[IsWrite],
-                   {AddrLong, ConstantInt::get(IntptrTy, TypeSize / 8)});
+                   {IRB.CreatePointerCast(Addr, IntptrTy),
+                    ConstantInt::get(IntptrTy, TypeSize / 8)});
   }
   untagPointerOperand(I, Addr);
 
@@ -654,9 +676,7 @@ bool HWAddressSanitizer::tagAlloca(IRBui
                     ConstantInt::get(IntptrTy, Size)});
   } else {
     size_t ShadowSize = Size >> Mapping.Scale;
-    Value *ShadowPtr = IRB.CreateIntToPtr(
-        memToShadow(IRB.CreatePointerCast(AI, IntptrTy), AI->getType(), IRB),
-        Int8PtrTy);
+    Value *ShadowPtr = memToShadow(IRB.CreatePointerCast(AI, IntptrTy), IRB);
     // If this memset is not inlined, it will be intercepted in the hwasan
     // runtime library. That's OK, because the interceptor skips the checks if
     // the address is in the shadow region.
@@ -883,6 +903,7 @@ Value *HWAddressSanitizer::emitPrologue(
           ThreadLongMaybeUntagged,
           ConstantInt::get(IntptrTy, (1ULL << kShadowBaseAlignment) - 1)),
       ConstantInt::get(IntptrTy, 1), "hwasan.shadow");
+  ShadowBase = IRB.CreateIntToPtr(ShadowBase, Int8PtrTy);
   return ShadowBase;
 }
 

Added: llvm/trunk/test/CodeGen/AArch64/hwasan-check-memaccess.ll
URL: http://llvm.org/viewvc/llvm-project/llvm/trunk/test/CodeGen/AArch64/hwasan-check-memaccess.ll?rev=351920&view=auto
==============================================================================
--- llvm/trunk/test/CodeGen/AArch64/hwasan-check-memaccess.ll (added)
+++ llvm/trunk/test/CodeGen/AArch64/hwasan-check-memaccess.ll Tue Jan 22 18:20:10 2019
@@ -0,0 +1,63 @@
+; RUN: llc < %s | FileCheck %s
+
+target triple = "aarch64--linux-android"
+
+define i8* @f1(i8* %x0, i8* %x1) {
+  ; CHECK: f1:
+  ; CHECK: str x30, [sp, #-16]!
+  ; CHECK-NEXT: .cfi_def_cfa_offset 16
+  ; CHECK-NEXT: .cfi_offset w30, -16
+  ; CHECK-NEXT: mov x9, x0
+  ; CHECK-NEXT: bl __hwasan_check_x1_123
+  ; CHECK-NEXT: mov x0, x1
+  ; CHECK-NEXT: ldr x30, [sp], #16
+  ; CHECK-NEXT: ret
+  call void @llvm.hwasan.check.memaccess(i8* %x0, i8* %x1, i32 123)
+  ret i8* %x1
+}
+
+define i8* @f2(i8* %x0, i8* %x1) {
+  ; CHECK: f2:
+  ; CHECK: str x30, [sp, #-16]!
+  ; CHECK-NEXT: .cfi_def_cfa_offset 16
+  ; CHECK-NEXT: .cfi_offset w30, -16
+  ; CHECK-NEXT: mov x9, x1
+  ; CHECK-NEXT: bl __hwasan_check_x0_456
+  ; CHECK-NEXT: ldr x30, [sp], #16
+  ; CHECK-NEXT: ret
+  call void @llvm.hwasan.check.memaccess(i8* %x1, i8* %x0, i32 456)
+  ret i8* %x0
+}
+
+declare void @llvm.hwasan.check.memaccess(i8*, i8*, i32)
+
+; CHECK:      .section .text.hot,"axG", at progbits,__hwasan_check_x0_456,comdat
+; CHECK-NEXT: .type __hwasan_check_x0_456, at function
+; CHECK-NEXT: .weak __hwasan_check_x0_456
+; CHECK-NEXT: .hidden __hwasan_check_x0_456
+; CHECK-NEXT: __hwasan_check_x0_456:
+; CHECK-NEXT: ubfx x16, x0, #4, #52
+; CHECK-NEXT: ldrb w16, [x9, x16]
+; CHECK-NEXT: lsr x17, x0, #56
+; CHECK-NEXT: cmp w16, w17
+; CHECK-NEXT: b.ne .Ltmp0
+; CHECK-NEXT: ret
+; CHECK-NEXT: .Ltmp0:
+; CHECK-NEXT: mov x1, #456
+; CHECK-NEXT: b __hwasan_tag_mismatch
+
+; CHECK:      .section .text.hot,"axG", at progbits,__hwasan_check_x1_123,comdat
+; CHECK-NEXT: .type __hwasan_check_x1_123, at function
+; CHECK-NEXT: .weak __hwasan_check_x1_123
+; CHECK-NEXT: .hidden __hwasan_check_x1_123
+; CHECK-NEXT: __hwasan_check_x1_123:
+; CHECK-NEXT: ubfx x16, x1, #4, #52
+; CHECK-NEXT: ldrb w16, [x9, x16]
+; CHECK-NEXT: lsr x17, x1, #56
+; CHECK-NEXT: cmp w16, w17
+; CHECK-NEXT: b.ne .Ltmp1
+; CHECK-NEXT: ret
+; CHECK-NEXT: .Ltmp1:
+; CHECK-NEXT: mov x0, x1
+; CHECK-NEXT: mov x1, #123
+; CHECK-NEXT: b __hwasan_tag_mismatch

Modified: llvm/trunk/test/Instrumentation/HWAddressSanitizer/alloca.ll
URL: http://llvm.org/viewvc/llvm-project/llvm/trunk/test/Instrumentation/HWAddressSanitizer/alloca.ll?rev=351920&r1=351919&r2=351920&view=diff
==============================================================================
--- llvm/trunk/test/Instrumentation/HWAddressSanitizer/alloca.ll (original)
+++ llvm/trunk/test/Instrumentation/HWAddressSanitizer/alloca.ll Tue Jan 22 18:20:10 2019
@@ -26,8 +26,7 @@ define void @test_alloca() sanitize_hwad
 ; CHECK: %[[X_TAG2:[^ ]*]] = trunc i64 %[[X_TAG]] to i8
 ; CHECK: %[[E:[^ ]*]] = ptrtoint i32* %[[X]] to i64
 ; CHECK: %[[F:[^ ]*]] = lshr i64 %[[E]], 4
-; DYNAMIC-SHADOW: %[[F_DYN:[^ ]*]] = add i64 %[[F]], %.hwasan.shadow
-; DYNAMIC-SHADOW: %[[X_SHADOW:[^ ]*]] = inttoptr i64 %[[F_DYN]] to i8*
+; DYNAMIC-SHADOW: %[[X_SHADOW:[^ ]*]] = getelementptr i8, i8* %.hwasan.shadow, i64 %[[F]]
 ; ZERO-BASED-SHADOW: %[[X_SHADOW:[^ ]*]] = inttoptr i64 %[[F]] to i8*
 ; CHECK: call void @llvm.memset.p0i8.i64(i8* align 1 %[[X_SHADOW]], i8 %[[X_TAG2]], i64 1, i1 false)
 ; CHECK: call void @use32(i32* nonnull %[[X_HWASAN]])
@@ -36,8 +35,7 @@ define void @test_alloca() sanitize_hwad
 ; UAR-TAGS: %[[X_TAG_UAR:[^ ]*]] = trunc i64 %[[BASE_TAG_COMPL]] to i8
 ; CHECK: %[[E2:[^ ]*]] = ptrtoint i32* %[[X]] to i64
 ; CHECK: %[[F2:[^ ]*]] = lshr i64 %[[E2]], 4
-; DYNAMIC-SHADOW: %[[F2_DYN:[^ ]*]] = add i64 %[[F2]], %.hwasan.shadow
-; DYNAMIC-SHADOW: %[[X_SHADOW2:[^ ]*]] = inttoptr i64 %[[F2_DYN]] to i8*
+; DYNAMIC-SHADOW: %[[X_SHADOW2:[^ ]*]] = getelementptr i8, i8* %.hwasan.shadow, i64 %[[F2]]
 ; ZERO-BASED-SHADOW: %[[X_SHADOW2:[^ ]*]] = inttoptr i64 %[[F2]] to i8*
 ; NO-UAR-TAGS: call void @llvm.memset.p0i8.i64(i8* align 1 %[[X_SHADOW2]], i8 0, i64 1, i1 false)
 ; UAR-TAGS: call void @llvm.memset.p0i8.i64(i8* align 1 %[[X_SHADOW2]], i8 %[[X_TAG_UAR]], i64 1, i1 false)

Modified: llvm/trunk/test/Instrumentation/HWAddressSanitizer/atomic.ll
URL: http://llvm.org/viewvc/llvm-project/llvm/trunk/test/Instrumentation/HWAddressSanitizer/atomic.ll?rev=351920&r1=351919&r2=351920&view=diff
==============================================================================
--- llvm/trunk/test/Instrumentation/HWAddressSanitizer/atomic.ll (original)
+++ llvm/trunk/test/Instrumentation/HWAddressSanitizer/atomic.ll Tue Jan 22 18:20:10 2019
@@ -7,8 +7,8 @@ target triple = "aarch64--linux-android"
 
 define void @atomicrmw(i64* %ptr) sanitize_hwaddress {
 ; CHECK-LABEL: @atomicrmw(
-; CHECK: lshr i64 %[[A:[^ ]*]], 56
-; CHECK: call void asm sideeffect "brk #2323", "{x0}"(i64 %[[A]])
+; CHECK: [[PTRI8:%[^ ]*]] = bitcast i64* %ptr to i8*
+; CHECK: call void @llvm.hwasan.check.memaccess({{.*}}, i8* [[PTRI8]], i32 19)
 ; CHECK: atomicrmw add i64* %ptr, i64 1 seq_cst
 ; CHECK: ret void
 
@@ -19,8 +19,8 @@ entry:
 
 define void @cmpxchg(i64* %ptr, i64 %compare_to, i64 %new_value) sanitize_hwaddress {
 ; CHECK-LABEL: @cmpxchg(
-; CHECK: lshr i64 %[[A:[^ ]*]], 56
-; CHECK: call void asm sideeffect "brk #2323", "{x0}"(i64 %[[A]])
+; CHECK: [[PTRI8:%[^ ]*]] = bitcast i64* %ptr to i8*
+; CHECK: call void @llvm.hwasan.check.memaccess({{.*}}, i8* [[PTRI8]], i32 19)
 ; CHECK: cmpxchg i64* %ptr, i64 %compare_to, i64 %new_value seq_cst seq_cst
 ; CHECK: ret void
 

Modified: llvm/trunk/test/Instrumentation/HWAddressSanitizer/basic.ll
URL: http://llvm.org/viewvc/llvm-project/llvm/trunk/test/Instrumentation/HWAddressSanitizer/basic.ll?rev=351920&r1=351919&r2=351920&view=diff
==============================================================================
--- llvm/trunk/test/Instrumentation/HWAddressSanitizer/basic.ll (original)
+++ llvm/trunk/test/Instrumentation/HWAddressSanitizer/basic.ll Tue Jan 22 18:20:10 2019
@@ -1,9 +1,9 @@
 ; Test basic address sanitizer instrumentation.
 ;
-; RUN: opt < %s -hwasan -hwasan-recover=0 -hwasan-with-ifunc=1 -hwasan-with-tls=0 -S | FileCheck %s --check-prefixes=CHECK,ABORT,DYNAMIC-SHADOW
-; RUN: opt < %s -hwasan -hwasan-recover=1 -hwasan-with-ifunc=1 -hwasan-with-tls=0 -S | FileCheck %s --check-prefixes=CHECK,RECOVER,DYNAMIC-SHADOW
-; RUN: opt < %s -hwasan -hwasan-recover=0 -hwasan-mapping-offset=0 -S | FileCheck %s --check-prefixes=CHECK,ABORT,ZERO-BASED-SHADOW
-; RUN: opt < %s -hwasan -hwasan-recover=1 -hwasan-mapping-offset=0 -S | FileCheck %s --check-prefixes=CHECK,RECOVER,ZERO-BASED-SHADOW
+; RUN: opt < %s -hwasan -hwasan-recover=0 -hwasan-with-ifunc=1 -hwasan-with-tls=0 -S | FileCheck %s --check-prefixes=CHECK,ABORT,ABORT-DYNAMIC-SHADOW
+; RUN: opt < %s -hwasan -hwasan-recover=1 -hwasan-with-ifunc=1 -hwasan-with-tls=0 -S | FileCheck %s --check-prefixes=CHECK,RECOVER,RECOVER-DYNAMIC-SHADOW
+; RUN: opt < %s -hwasan -hwasan-recover=0 -hwasan-mapping-offset=0 -S | FileCheck %s --check-prefixes=CHECK,ABORT,ABORT-ZERO-BASED-SHADOW
+; RUN: opt < %s -hwasan -hwasan-recover=1 -hwasan-mapping-offset=0 -S | FileCheck %s --check-prefixes=CHECK,RECOVER,RECOVER-ZERO-BASED-SHADOW
 
 ; CHECK: @llvm.global_ctors = appending global [1 x { i32, void ()*, i8* }] [{ i32, void ()*, i8* } { i32 0, void ()* @hwasan.module_ctor, i8* bitcast (void ()* @hwasan.module_ctor to i8*) }]
 ; CHECK: @__hwasan = private constant [0 x i8] zeroinitializer, section "__hwasan_frames", comdat($hwasan.module_ctor)
@@ -13,23 +13,23 @@ target triple = "aarch64--linux-android"
 
 define i8 @test_load8(i8* %a) sanitize_hwaddress {
 ; CHECK-LABEL: @test_load8(
-; CHECK: %[[A:[^ ]*]] = ptrtoint i8* %a to i64
-; CHECK: %[[B:[^ ]*]] = lshr i64 %[[A]], 56
-; CHECK: %[[PTRTAG:[^ ]*]] = trunc i64 %[[B]] to i8
-; CHECK: %[[C:[^ ]*]] = and i64 %[[A]], 72057594037927935
-; CHECK: %[[D:[^ ]*]] = lshr i64 %[[C]], 4
-; DYNAMIC-SHADOW: %[[D_DYN:[^ ]*]] = add i64 %[[D]], %.hwasan.shadow
-; DYNAMIC-SHADOW: %[[E:[^ ]*]] = inttoptr i64 %[[D_DYN]] to i8*
-; ZERO-BASED-SHADOW: %[[E:[^ ]*]] = inttoptr i64 %[[D]] to i8*
-; CHECK: %[[MEMTAG:[^ ]*]] = load i8, i8* %[[E]]
-; CHECK: %[[F:[^ ]*]] = icmp ne i8 %[[PTRTAG]], %[[MEMTAG]]
-; CHECK: br i1 %[[F]], label {{.*}}, label {{.*}}, !prof {{.*}}
+; RECOVER: %[[A:[^ ]*]] = ptrtoint i8* %a to i64
+; RECOVER: %[[B:[^ ]*]] = lshr i64 %[[A]], 56
+; RECOVER: %[[PTRTAG:[^ ]*]] = trunc i64 %[[B]] to i8
+; RECOVER: %[[C:[^ ]*]] = and i64 %[[A]], 72057594037927935
+; RECOVER: %[[D:[^ ]*]] = lshr i64 %[[C]], 4
+; RECOVER-DYNAMIC-SHADOW: %[[E:[^ ]*]] = getelementptr i8, i8* %.hwasan.shadow, i64 %4
+; RECOVER-ZERO-BASED-SHADOW: %[[E:[^ ]*]] = inttoptr i64 %[[D]] to i8*
+; RECOVER: %[[MEMTAG:[^ ]*]] = load i8, i8* %[[E]]
+; RECOVER: %[[F:[^ ]*]] = icmp ne i8 %[[PTRTAG]], %[[MEMTAG]]
+; RECOVER: br i1 %[[F]], label {{.*}}, label {{.*}}, !prof {{.*}}
 
-; ABORT: call void asm sideeffect "brk #2304", "{x0}"(i64 %[[A]])
-; ABORT: unreachable
 ; RECOVER: call void asm sideeffect "brk #2336", "{x0}"(i64 %[[A]])
 ; RECOVER: br label
 
+; ABORT-DYNAMIC-SHADOW: call void @llvm.hwasan.check.memaccess(i8* %.hwasan.shadow, i8* %a, i32 0)
+; ABORT-ZERO-BASED-SHADOW: call void @llvm.hwasan.check.memaccess(i8* null, i8* %a, i32 0)
+
 ; CHECK: %[[G:[^ ]*]] = load i8, i8* %a, align 4
 ; CHECK: ret i8 %[[G]]
 
@@ -40,23 +40,24 @@ entry:
 
 define i16 @test_load16(i16* %a) sanitize_hwaddress {
 ; CHECK-LABEL: @test_load16(
-; CHECK: %[[A:[^ ]*]] = ptrtoint i16* %a to i64
-; CHECK: %[[B:[^ ]*]] = lshr i64 %[[A]], 56
-; CHECK: %[[PTRTAG:[^ ]*]] = trunc i64 %[[B]] to i8
-; CHECK: %[[C:[^ ]*]] = and i64 %[[A]], 72057594037927935
-; CHECK: %[[D:[^ ]*]] = lshr i64 %[[C]], 4
-; DYNAMIC-SHADOW: %[[D_DYN:[^ ]*]] = add i64 %[[D]], %.hwasan.shadow
-; DYNAMIC-SHADOW: %[[E:[^ ]*]] = inttoptr i64 %[[D_DYN]] to i8*
-; ZERO-BASED-SHADOW: %[[E:[^ ]*]] = inttoptr i64 %[[D]] to i8*
-; CHECK: %[[MEMTAG:[^ ]*]] = load i8, i8* %[[E]]
-; CHECK: %[[F:[^ ]*]] = icmp ne i8 %[[PTRTAG]], %[[MEMTAG]]
-; CHECK: br i1 %[[F]], label {{.*}}, label {{.*}}, !prof {{.*}}
+; RECOVER: %[[A:[^ ]*]] = ptrtoint i16* %a to i64
+; RECOVER: %[[B:[^ ]*]] = lshr i64 %[[A]], 56
+; RECOVER: %[[PTRTAG:[^ ]*]] = trunc i64 %[[B]] to i8
+; RECOVER: %[[C:[^ ]*]] = and i64 %[[A]], 72057594037927935
+; RECOVER: %[[D:[^ ]*]] = lshr i64 %[[C]], 4
+; RECOVER-DYNAMIC-SHADOW: %[[E:[^ ]*]] = getelementptr i8, i8* %.hwasan.shadow, i64 %4
+; RECOVER-ZERO-BASED-SHADOW: %[[E:[^ ]*]] = inttoptr i64 %[[D]] to i8*
+; RECOVER: %[[MEMTAG:[^ ]*]] = load i8, i8* %[[E]]
+; RECOVER: %[[F:[^ ]*]] = icmp ne i8 %[[PTRTAG]], %[[MEMTAG]]
+; RECOVER: br i1 %[[F]], label {{.*}}, label {{.*}}, !prof {{.*}}
 
-; ABORT: call void asm sideeffect "brk #2305", "{x0}"(i64 %[[A]])
-; ABORT: unreachable
 ; RECOVER: call void asm sideeffect "brk #2337", "{x0}"(i64 %[[A]])
 ; RECOVER: br label
 
+; ABORT: %[[A:[^ ]*]] = bitcast i16* %a to i8*
+; ABORT-DYNAMIC-SHADOW: call void @llvm.hwasan.check.memaccess(i8* %.hwasan.shadow, i8* %[[A]], i32 1)
+; ABORT-ZERO-BASED-SHADOW: call void @llvm.hwasan.check.memaccess(i8* null, i8* %[[A]], i32 1)
+
 ; CHECK: %[[G:[^ ]*]] = load i16, i16* %a, align 4
 ; CHECK: ret i16 %[[G]]
 
@@ -67,23 +68,24 @@ entry:
 
 define i32 @test_load32(i32* %a) sanitize_hwaddress {
 ; CHECK-LABEL: @test_load32(
-; CHECK: %[[A:[^ ]*]] = ptrtoint i32* %a to i64
-; CHECK: %[[B:[^ ]*]] = lshr i64 %[[A]], 56
-; CHECK: %[[PTRTAG:[^ ]*]] = trunc i64 %[[B]] to i8
-; CHECK: %[[C:[^ ]*]] = and i64 %[[A]], 72057594037927935
-; CHECK: %[[D:[^ ]*]] = lshr i64 %[[C]], 4
-; DYNAMIC-SHADOW: %[[D_DYN:[^ ]*]] = add i64 %[[D]], %.hwasan.shadow
-; DYNAMIC-SHADOW: %[[E:[^ ]*]] = inttoptr i64 %[[D_DYN]] to i8*
-; ZERO-BASED-SHADOW: %[[E:[^ ]*]] = inttoptr i64 %[[D]] to i8*
-; CHECK: %[[MEMTAG:[^ ]*]] = load i8, i8* %[[E]]
-; CHECK: %[[F:[^ ]*]] = icmp ne i8 %[[PTRTAG]], %[[MEMTAG]]
-; CHECK: br i1 %[[F]], label {{.*}}, label {{.*}}, !prof {{.*}}
+; RECOVER: %[[A:[^ ]*]] = ptrtoint i32* %a to i64
+; RECOVER: %[[B:[^ ]*]] = lshr i64 %[[A]], 56
+; RECOVER: %[[PTRTAG:[^ ]*]] = trunc i64 %[[B]] to i8
+; RECOVER: %[[C:[^ ]*]] = and i64 %[[A]], 72057594037927935
+; RECOVER: %[[D:[^ ]*]] = lshr i64 %[[C]], 4
+; RECOVER-DYNAMIC-SHADOW: %[[E:[^ ]*]] = getelementptr i8, i8* %.hwasan.shadow, i64 %4
+; RECOVER-ZERO-BASED-SHADOW: %[[E:[^ ]*]] = inttoptr i64 %[[D]] to i8*
+; RECOVER: %[[MEMTAG:[^ ]*]] = load i8, i8* %[[E]]
+; RECOVER: %[[F:[^ ]*]] = icmp ne i8 %[[PTRTAG]], %[[MEMTAG]]
+; RECOVER: br i1 %[[F]], label {{.*}}, label {{.*}}, !prof {{.*}}
 
-; ABORT: call void asm sideeffect "brk #2306", "{x0}"(i64 %[[A]])
-; ABORT: unreachable
 ; RECOVER: call void asm sideeffect "brk #2338", "{x0}"(i64 %[[A]])
 ; RECOVER: br label
 
+; ABORT: %[[A:[^ ]*]] = bitcast i32* %a to i8*
+; ABORT-DYNAMIC-SHADOW: call void @llvm.hwasan.check.memaccess(i8* %.hwasan.shadow, i8* %[[A]], i32 2)
+; ABORT-ZERO-BASED-SHADOW: call void @llvm.hwasan.check.memaccess(i8* null, i8* %[[A]], i32 2)
+
 ; CHECK: %[[G:[^ ]*]] = load i32, i32* %a, align 4
 ; CHECK: ret i32 %[[G]]
 
@@ -94,23 +96,24 @@ entry:
 
 define i64 @test_load64(i64* %a) sanitize_hwaddress {
 ; CHECK-LABEL: @test_load64(
-; CHECK: %[[A:[^ ]*]] = ptrtoint i64* %a to i64
-; CHECK: %[[B:[^ ]*]] = lshr i64 %[[A]], 56
-; CHECK: %[[PTRTAG:[^ ]*]] = trunc i64 %[[B]] to i8
-; CHECK: %[[C:[^ ]*]] = and i64 %[[A]], 72057594037927935
-; CHECK: %[[D:[^ ]*]] = lshr i64 %[[C]], 4
-; DYNAMIC-SHADOW: %[[D_DYN:[^ ]*]] = add i64 %[[D]], %.hwasan.shadow
-; DYNAMIC-SHADOW: %[[E:[^ ]*]] = inttoptr i64 %[[D_DYN]] to i8*
-; ZERO-BASED-SHADOW: %[[E:[^ ]*]] = inttoptr i64 %[[D]] to i8*
-; CHECK: %[[MEMTAG:[^ ]*]] = load i8, i8* %[[E]]
-; CHECK: %[[F:[^ ]*]] = icmp ne i8 %[[PTRTAG]], %[[MEMTAG]]
-; CHECK: br i1 %[[F]], label {{.*}}, label {{.*}}, !prof {{.*}}
+; RECOVER: %[[A:[^ ]*]] = ptrtoint i64* %a to i64
+; RECOVER: %[[B:[^ ]*]] = lshr i64 %[[A]], 56
+; RECOVER: %[[PTRTAG:[^ ]*]] = trunc i64 %[[B]] to i8
+; RECOVER: %[[C:[^ ]*]] = and i64 %[[A]], 72057594037927935
+; RECOVER: %[[D:[^ ]*]] = lshr i64 %[[C]], 4
+; RECOVER-DYNAMIC-SHADOW: %[[E:[^ ]*]] = getelementptr i8, i8* %.hwasan.shadow, i64 %4
+; RECOVER-ZERO-BASED-SHADOW: %[[E:[^ ]*]] = inttoptr i64 %[[D]] to i8*
+; RECOVER: %[[MEMTAG:[^ ]*]] = load i8, i8* %[[E]]
+; RECOVER: %[[F:[^ ]*]] = icmp ne i8 %[[PTRTAG]], %[[MEMTAG]]
+; RECOVER: br i1 %[[F]], label {{.*}}, label {{.*}}, !prof {{.*}}
 
-; ABORT: call void asm sideeffect "brk #2307", "{x0}"(i64 %[[A]])
-; ABORT: unreachable
 ; RECOVER: call void asm sideeffect "brk #2339", "{x0}"(i64 %[[A]])
 ; RECOVER: br label
 
+; ABORT: %[[A:[^ ]*]] = bitcast i64* %a to i8*
+; ABORT-DYNAMIC-SHADOW: call void @llvm.hwasan.check.memaccess(i8* %.hwasan.shadow, i8* %[[A]], i32 3)
+; ABORT-ZERO-BASED-SHADOW: call void @llvm.hwasan.check.memaccess(i8* null, i8* %[[A]], i32 3)
+
 ; CHECK: %[[G:[^ ]*]] = load i64, i64* %a, align 8
 ; CHECK: ret i64 %[[G]]
 
@@ -121,23 +124,24 @@ entry:
 
 define i128 @test_load128(i128* %a) sanitize_hwaddress {
 ; CHECK-LABEL: @test_load128(
-; CHECK: %[[A:[^ ]*]] = ptrtoint i128* %a to i64
-; CHECK: %[[B:[^ ]*]] = lshr i64 %[[A]], 56
-; CHECK: %[[PTRTAG:[^ ]*]] = trunc i64 %[[B]] to i8
-; CHECK: %[[C:[^ ]*]] = and i64 %[[A]], 72057594037927935
-; CHECK: %[[D:[^ ]*]] = lshr i64 %[[C]], 4
-; DYNAMIC-SHADOW: %[[D_DYN:[^ ]*]] = add i64 %[[D]], %.hwasan.shadow
-; DYNAMIC-SHADOW: %[[E:[^ ]*]] = inttoptr i64 %[[D_DYN]] to i8*
-; ZERO-BASED-SHADOW: %[[E:[^ ]*]] = inttoptr i64 %[[D]] to i8*
-; CHECK: %[[MEMTAG:[^ ]*]] = load i8, i8* %[[E]]
-; CHECK: %[[F:[^ ]*]] = icmp ne i8 %[[PTRTAG]], %[[MEMTAG]]
-; CHECK: br i1 %[[F]], label {{.*}}, label {{.*}}, !prof {{.*}}
+; RECOVER: %[[A:[^ ]*]] = ptrtoint i128* %a to i64
+; RECOVER: %[[B:[^ ]*]] = lshr i64 %[[A]], 56
+; RECOVER: %[[PTRTAG:[^ ]*]] = trunc i64 %[[B]] to i8
+; RECOVER: %[[C:[^ ]*]] = and i64 %[[A]], 72057594037927935
+; RECOVER: %[[D:[^ ]*]] = lshr i64 %[[C]], 4
+; RECOVER-DYNAMIC-SHADOW: %[[E:[^ ]*]] = getelementptr i8, i8* %.hwasan.shadow, i64 %4
+; RECOVER-ZERO-BASED-SHADOW: %[[E:[^ ]*]] = inttoptr i64 %[[D]] to i8*
+; RECOVER: %[[MEMTAG:[^ ]*]] = load i8, i8* %[[E]]
+; RECOVER: %[[F:[^ ]*]] = icmp ne i8 %[[PTRTAG]], %[[MEMTAG]]
+; RECOVER: br i1 %[[F]], label {{.*}}, label {{.*}}, !prof {{.*}}
 
-; ABORT: call void asm sideeffect "brk #2308", "{x0}"(i64 %[[A]])
-; ABORT: unreachable
 ; RECOVER: call void asm sideeffect "brk #2340", "{x0}"(i64 %[[A]])
 ; RECOVER: br label
 
+; ABORT: %[[A:[^ ]*]] = bitcast i128* %a to i8*
+; ABORT-DYNAMIC-SHADOW: call void @llvm.hwasan.check.memaccess(i8* %.hwasan.shadow, i8* %[[A]], i32 4)
+; ABORT-ZERO-BASED-SHADOW: call void @llvm.hwasan.check.memaccess(i8* null, i8* %[[A]], i32 4)
+
 ; CHECK: %[[G:[^ ]*]] = load i128, i128* %a, align 16
 ; CHECK: ret i128 %[[G]]
 
@@ -161,23 +165,23 @@ entry:
 
 define void @test_store8(i8* %a, i8 %b) sanitize_hwaddress {
 ; CHECK-LABEL: @test_store8(
-; CHECK: %[[A:[^ ]*]] = ptrtoint i8* %a to i64
-; CHECK: %[[B:[^ ]*]] = lshr i64 %[[A]], 56
-; CHECK: %[[PTRTAG:[^ ]*]] = trunc i64 %[[B]] to i8
-; CHECK: %[[C:[^ ]*]] = and i64 %[[A]], 72057594037927935
-; CHECK: %[[D:[^ ]*]] = lshr i64 %[[C]], 4
-; DYNAMIC-SHADOW: %[[D_DYN:[^ ]*]] = add i64 %[[D]], %.hwasan.shadow
-; DYNAMIC-SHADOW: %[[E:[^ ]*]] = inttoptr i64 %[[D_DYN]] to i8*
-; ZERO-BASED-SHADOW: %[[E:[^ ]*]] = inttoptr i64 %[[D]] to i8*
-; CHECK: %[[MEMTAG:[^ ]*]] = load i8, i8* %[[E]]
-; CHECK: %[[F:[^ ]*]] = icmp ne i8 %[[PTRTAG]], %[[MEMTAG]]
-; CHECK: br i1 %[[F]], label {{.*}}, label {{.*}}, !prof {{.*}}
+; RECOVER: %[[A:[^ ]*]] = ptrtoint i8* %a to i64
+; RECOVER: %[[B:[^ ]*]] = lshr i64 %[[A]], 56
+; RECOVER: %[[PTRTAG:[^ ]*]] = trunc i64 %[[B]] to i8
+; RECOVER: %[[C:[^ ]*]] = and i64 %[[A]], 72057594037927935
+; RECOVER: %[[D:[^ ]*]] = lshr i64 %[[C]], 4
+; RECOVER-DYNAMIC-SHADOW: %[[E:[^ ]*]] = getelementptr i8, i8* %.hwasan.shadow, i64 %4
+; RECOVER-ZERO-BASED-SHADOW: %[[E:[^ ]*]] = inttoptr i64 %[[D]] to i8*
+; RECOVER: %[[MEMTAG:[^ ]*]] = load i8, i8* %[[E]]
+; RECOVER: %[[F:[^ ]*]] = icmp ne i8 %[[PTRTAG]], %[[MEMTAG]]
+; RECOVER: br i1 %[[F]], label {{.*}}, label {{.*}}, !prof {{.*}}
 
-; ABORT: call void asm sideeffect "brk #2320", "{x0}"(i64 %[[A]])
-; ABORT: unreachable
 ; RECOVER: call void asm sideeffect "brk #2352", "{x0}"(i64 %[[A]])
 ; RECOVER: br label
 
+; ABORT-DYNAMIC-SHADOW: call void @llvm.hwasan.check.memaccess(i8* %.hwasan.shadow, i8* %a, i32 16)
+; ABORT-ZERO-BASED-SHADOW: call void @llvm.hwasan.check.memaccess(i8* null, i8* %a, i32 16)
+
 ; CHECK: store i8 %b, i8* %a, align 4
 ; CHECK: ret void
 
@@ -188,23 +192,24 @@ entry:
 
 define void @test_store16(i16* %a, i16 %b) sanitize_hwaddress {
 ; CHECK-LABEL: @test_store16(
-; CHECK: %[[A:[^ ]*]] = ptrtoint i16* %a to i64
-; CHECK: %[[B:[^ ]*]] = lshr i64 %[[A]], 56
-; CHECK: %[[PTRTAG:[^ ]*]] = trunc i64 %[[B]] to i8
-; CHECK: %[[C:[^ ]*]] = and i64 %[[A]], 72057594037927935
-; CHECK: %[[D:[^ ]*]] = lshr i64 %[[C]], 4
-; DYNAMIC-SHADOW: %[[D_DYN:[^ ]*]] = add i64 %[[D]], %.hwasan.shadow
-; DYNAMIC-SHADOW: %[[E:[^ ]*]] = inttoptr i64 %[[D_DYN]] to i8*
-; ZERO-BASED-SHADOW: %[[E:[^ ]*]] = inttoptr i64 %[[D]] to i8*
-; CHECK: %[[MEMTAG:[^ ]*]] = load i8, i8* %[[E]]
-; CHECK: %[[F:[^ ]*]] = icmp ne i8 %[[PTRTAG]], %[[MEMTAG]]
-; CHECK: br i1 %[[F]], label {{.*}}, label {{.*}}, !prof {{.*}}
+; RECOVER: %[[A:[^ ]*]] = ptrtoint i16* %a to i64
+; RECOVER: %[[B:[^ ]*]] = lshr i64 %[[A]], 56
+; RECOVER: %[[PTRTAG:[^ ]*]] = trunc i64 %[[B]] to i8
+; RECOVER: %[[C:[^ ]*]] = and i64 %[[A]], 72057594037927935
+; RECOVER: %[[D:[^ ]*]] = lshr i64 %[[C]], 4
+; RECOVER-DYNAMIC-SHADOW: %[[E:[^ ]*]] = getelementptr i8, i8* %.hwasan.shadow, i64 %4
+; RECOVER-ZERO-BASED-SHADOW: %[[E:[^ ]*]] = inttoptr i64 %[[D]] to i8*
+; RECOVER: %[[MEMTAG:[^ ]*]] = load i8, i8* %[[E]]
+; RECOVER: %[[F:[^ ]*]] = icmp ne i8 %[[PTRTAG]], %[[MEMTAG]]
+; RECOVER: br i1 %[[F]], label {{.*}}, label {{.*}}, !prof {{.*}}
 
-; ABORT: call void asm sideeffect "brk #2321", "{x0}"(i64 %[[A]])
-; ABORT: unreachable
 ; RECOVER: call void asm sideeffect "brk #2353", "{x0}"(i64 %[[A]])
 ; RECOVER: br label
 
+; ABORT: %[[A:[^ ]*]] = bitcast i16* %a to i8*
+; ABORT-DYNAMIC-SHADOW: call void @llvm.hwasan.check.memaccess(i8* %.hwasan.shadow, i8* %[[A]], i32 17)
+; ABORT-ZERO-BASED-SHADOW: call void @llvm.hwasan.check.memaccess(i8* null, i8* %[[A]], i32 17)
+
 ; CHECK: store i16 %b, i16* %a, align 4
 ; CHECK: ret void
 
@@ -215,23 +220,24 @@ entry:
 
 define void @test_store32(i32* %a, i32 %b) sanitize_hwaddress {
 ; CHECK-LABEL: @test_store32(
-; CHECK: %[[A:[^ ]*]] = ptrtoint i32* %a to i64
-; CHECK: %[[B:[^ ]*]] = lshr i64 %[[A]], 56
-; CHECK: %[[PTRTAG:[^ ]*]] = trunc i64 %[[B]] to i8
-; CHECK: %[[C:[^ ]*]] = and i64 %[[A]], 72057594037927935
-; CHECK: %[[D:[^ ]*]] = lshr i64 %[[C]], 4
-; DYNAMIC-SHADOW: %[[D_DYN:[^ ]*]] = add i64 %[[D]], %.hwasan.shadow
-; DYNAMIC-SHADOW: %[[E:[^ ]*]] = inttoptr i64 %[[D_DYN]] to i8*
-; ZERO-BASED-SHADOW: %[[E:[^ ]*]] = inttoptr i64 %[[D]] to i8*
-; CHECK: %[[MEMTAG:[^ ]*]] = load i8, i8* %[[E]]
-; CHECK: %[[F:[^ ]*]] = icmp ne i8 %[[PTRTAG]], %[[MEMTAG]]
-; CHECK: br i1 %[[F]], label {{.*}}, label {{.*}}, !prof {{.*}}
+; RECOVER: %[[A:[^ ]*]] = ptrtoint i32* %a to i64
+; RECOVER: %[[B:[^ ]*]] = lshr i64 %[[A]], 56
+; RECOVER: %[[PTRTAG:[^ ]*]] = trunc i64 %[[B]] to i8
+; RECOVER: %[[C:[^ ]*]] = and i64 %[[A]], 72057594037927935
+; RECOVER: %[[D:[^ ]*]] = lshr i64 %[[C]], 4
+; RECOVER-DYNAMIC-SHADOW: %[[E:[^ ]*]] = getelementptr i8, i8* %.hwasan.shadow, i64 %4
+; RECOVER-ZERO-BASED-SHADOW: %[[E:[^ ]*]] = inttoptr i64 %[[D]] to i8*
+; RECOVER: %[[MEMTAG:[^ ]*]] = load i8, i8* %[[E]]
+; RECOVER: %[[F:[^ ]*]] = icmp ne i8 %[[PTRTAG]], %[[MEMTAG]]
+; RECOVER: br i1 %[[F]], label {{.*}}, label {{.*}}, !prof {{.*}}
 
-; ABORT: call void asm sideeffect "brk #2322", "{x0}"(i64 %[[A]])
-; ABORT: unreachable
 ; RECOVER: call void asm sideeffect "brk #2354", "{x0}"(i64 %[[A]])
 ; RECOVER: br label
 
+; ABORT: %[[A:[^ ]*]] = bitcast i32* %a to i8*
+; ABORT-DYNAMIC-SHADOW: call void @llvm.hwasan.check.memaccess(i8* %.hwasan.shadow, i8* %[[A]], i32 18)
+; ABORT-ZERO-BASED-SHADOW: call void @llvm.hwasan.check.memaccess(i8* null, i8* %[[A]], i32 18)
+
 ; CHECK: store i32 %b, i32* %a, align 4
 ; CHECK: ret void
 
@@ -242,23 +248,24 @@ entry:
 
 define void @test_store64(i64* %a, i64 %b) sanitize_hwaddress {
 ; CHECK-LABEL: @test_store64(
-; CHECK: %[[A:[^ ]*]] = ptrtoint i64* %a to i64
-; CHECK: %[[B:[^ ]*]] = lshr i64 %[[A]], 56
-; CHECK: %[[PTRTAG:[^ ]*]] = trunc i64 %[[B]] to i8
-; CHECK: %[[C:[^ ]*]] = and i64 %[[A]], 72057594037927935
-; CHECK: %[[D:[^ ]*]] = lshr i64 %[[C]], 4
-; DYNAMIC-SHADOW: %[[D_DYN:[^ ]*]] = add i64 %[[D]], %.hwasan.shadow
-; DYNAMIC-SHADOW: %[[E:[^ ]*]] = inttoptr i64 %[[D_DYN]] to i8*
-; ZERO-BASED-SHADOW: %[[E:[^ ]*]] = inttoptr i64 %[[D]] to i8*
-; CHECK: %[[MEMTAG:[^ ]*]] = load i8, i8* %[[E]]
-; CHECK: %[[F:[^ ]*]] = icmp ne i8 %[[PTRTAG]], %[[MEMTAG]]
-; CHECK: br i1 %[[F]], label {{.*}}, label {{.*}}, !prof {{.*}}
+; RECOVER: %[[A:[^ ]*]] = ptrtoint i64* %a to i64
+; RECOVER: %[[B:[^ ]*]] = lshr i64 %[[A]], 56
+; RECOVER: %[[PTRTAG:[^ ]*]] = trunc i64 %[[B]] to i8
+; RECOVER: %[[C:[^ ]*]] = and i64 %[[A]], 72057594037927935
+; RECOVER: %[[D:[^ ]*]] = lshr i64 %[[C]], 4
+; RECOVER-DYNAMIC-SHADOW: %[[E:[^ ]*]] = getelementptr i8, i8* %.hwasan.shadow, i64 %4
+; RECOVER-ZERO-BASED-SHADOW: %[[E:[^ ]*]] = inttoptr i64 %[[D]] to i8*
+; RECOVER: %[[MEMTAG:[^ ]*]] = load i8, i8* %[[E]]
+; RECOVER: %[[F:[^ ]*]] = icmp ne i8 %[[PTRTAG]], %[[MEMTAG]]
+; RECOVER: br i1 %[[F]], label {{.*}}, label {{.*}}, !prof {{.*}}
 
-; ABORT: call void asm sideeffect "brk #2323", "{x0}"(i64 %[[A]])
-; ABORT: unreachable
 ; RECOVER: call void asm sideeffect "brk #2355", "{x0}"(i64 %[[A]])
 ; RECOVER: br label
 
+; ABORT: %[[A:[^ ]*]] = bitcast i64* %a to i8*
+; ABORT-DYNAMIC-SHADOW: call void @llvm.hwasan.check.memaccess(i8* %.hwasan.shadow, i8* %[[A]], i32 19)
+; ABORT-ZERO-BASED-SHADOW: call void @llvm.hwasan.check.memaccess(i8* null, i8* %[[A]], i32 19)
+
 ; CHECK: store i64 %b, i64* %a, align 8
 ; CHECK: ret void
 
@@ -269,23 +276,24 @@ entry:
 
 define void @test_store128(i128* %a, i128 %b) sanitize_hwaddress {
 ; CHECK-LABEL: @test_store128(
-; CHECK: %[[A:[^ ]*]] = ptrtoint i128* %a to i64
-; CHECK: %[[B:[^ ]*]] = lshr i64 %[[A]], 56
-; CHECK: %[[PTRTAG:[^ ]*]] = trunc i64 %[[B]] to i8
-; CHECK: %[[C:[^ ]*]] = and i64 %[[A]], 72057594037927935
-; CHECK: %[[D:[^ ]*]] = lshr i64 %[[C]], 4
-; DYNAMIC-SHADOW: %[[D_DYN:[^ ]*]] = add i64 %[[D]], %.hwasan.shadow
-; DYNAMIC-SHADOW: %[[E:[^ ]*]] = inttoptr i64 %[[D_DYN]] to i8*
-; ZERO-BASED-SHADOW: %[[E:[^ ]*]] = inttoptr i64 %[[D]] to i8*
-; CHECK: %[[MEMTAG:[^ ]*]] = load i8, i8* %[[E]]
-; CHECK: %[[F:[^ ]*]] = icmp ne i8 %[[PTRTAG]], %[[MEMTAG]]
-; CHECK: br i1 %[[F]], label {{.*}}, label {{.*}}, !prof {{.*}}
+; RECOVER: %[[A:[^ ]*]] = ptrtoint i128* %a to i64
+; RECOVER: %[[B:[^ ]*]] = lshr i64 %[[A]], 56
+; RECOVER: %[[PTRTAG:[^ ]*]] = trunc i64 %[[B]] to i8
+; RECOVER: %[[C:[^ ]*]] = and i64 %[[A]], 72057594037927935
+; RECOVER: %[[D:[^ ]*]] = lshr i64 %[[C]], 4
+; RECOVER-DYNAMIC-SHADOW: %[[E:[^ ]*]] = getelementptr i8, i8* %.hwasan.shadow, i64 %4
+; RECOVER-ZERO-BASED-SHADOW: %[[E:[^ ]*]] = inttoptr i64 %[[D]] to i8*
+; RECOVER: %[[MEMTAG:[^ ]*]] = load i8, i8* %[[E]]
+; RECOVER: %[[F:[^ ]*]] = icmp ne i8 %[[PTRTAG]], %[[MEMTAG]]
+; RECOVER: br i1 %[[F]], label {{.*}}, label {{.*}}, !prof {{.*}}
 
-; ABORT: call void asm sideeffect "brk #2324", "{x0}"(i64 %[[A]])
-; ABORT: unreachable
 ; RECOVER: call void asm sideeffect "brk #2356", "{x0}"(i64 %[[A]])
 ; RECOVER: br label
 
+; ABORT: %[[A:[^ ]*]] = bitcast i128* %a to i8*
+; ABORT-DYNAMIC-SHADOW: call void @llvm.hwasan.check.memaccess(i8* %.hwasan.shadow, i8* %[[A]], i32 20)
+; ABORT-ZERO-BASED-SHADOW: call void @llvm.hwasan.check.memaccess(i8* null, i8* %[[A]], i32 20)
+
 ; CHECK: store i128 %b, i128* %a, align 16
 ; CHECK: ret void
 

Modified: llvm/trunk/test/Instrumentation/HWAddressSanitizer/kernel.ll
URL: http://llvm.org/viewvc/llvm-project/llvm/trunk/test/Instrumentation/HWAddressSanitizer/kernel.ll?rev=351920&r1=351919&r2=351920&view=diff
==============================================================================
--- llvm/trunk/test/Instrumentation/HWAddressSanitizer/kernel.ll (original)
+++ llvm/trunk/test/Instrumentation/HWAddressSanitizer/kernel.ll Tue Jan 22 18:20:10 2019
@@ -1,11 +1,9 @@
 ; Test KHWASan instrumentation.
 ;
-; RUN: opt < %s -hwasan -hwasan-kernel=1 -S | FileCheck %s --allow-empty --check-prefixes=INIT
-; RUN: opt < %s -hwasan -hwasan-kernel=1 -S | FileCheck %s  --check-prefixes=CHECK,NOOFFSET,MATCH-ALL
-; RUN: opt < %s -hwasan -hwasan-kernel=1 -hwasan-mapping-offset=12345678 -S | FileCheck %s  --check-prefixes=CHECK,OFFSET,MATCH-ALL
-; RUN: opt < %s -hwasan -hwasan-kernel=1 -hwasan-recover=0 -S | FileCheck %s  --check-prefixes=CHECK,NOOFFSET,ABORT,MATCH-ALL
-; RUN: opt < %s -hwasan -hwasan-kernel=1 -hwasan-recover=1 -S | FileCheck %s  --check-prefixes=CHECK,NOOFFSET,RECOVER,MATCH-ALL
-; RUN: opt < %s -hwasan -hwasan-kernel=1 -hwasan-recover=1 -hwasan-match-all-tag=-1 -S | FileCheck %s  --check-prefixes=CHECK,NOOFFSET,RECOVER,NO-MATCH-ALL
+; RUN: opt < %s -hwasan -hwasan-kernel=1 -hwasan-recover=1 -S | FileCheck %s --allow-empty --check-prefixes=INIT
+; RUN: opt < %s -hwasan -hwasan-kernel=1 -hwasan-recover=1 -S | FileCheck %s  --check-prefixes=CHECK,NOOFFSET,MATCH-ALL
+; RUN: opt < %s -hwasan -hwasan-kernel=1 -hwasan-recover=1 -hwasan-mapping-offset=12345678 -S | FileCheck %s  --check-prefixes=CHECK,OFFSET,MATCH-ALL
+; RUN: opt < %s -hwasan -hwasan-kernel=1 -hwasan-recover=1 -hwasan-match-all-tag=-1 -S | FileCheck %s  --check-prefixes=CHECK,NOOFFSET,NO-MATCH-ALL
 
 target datalayout = "e-m:e-i8:8:32-i16:16:32-i64:64-i128:128-n32:64-S128"
 target triple = "aarch64--linux-android"
@@ -20,8 +18,7 @@ define i8 @test_load(i8* %a) sanitize_hw
 
 ; NOOFFSET: %[[E:[^ ]*]] = inttoptr i64 %[[D]] to i8*
 
-; OFFSET: %[[D1:[^ ]*]] = add i64 %[[D]], 12345678
-; OFFSET: %[[E:[^ ]*]] = inttoptr i64 %[[D1]] to i8*
+; OFFSET: %[[E:[^ ]*]] = getelementptr i8, i8* inttoptr (i64 12345678 to i8*), i64 %[[D]]
 
 ; CHECK: %[[MEMTAG:[^ ]*]] = load i8, i8* %[[E]]
 ; CHECK: %[[F:[^ ]*]] = icmp ne i8 %[[PTRTAG]], %[[MEMTAG]]
@@ -32,10 +29,8 @@ define i8 @test_load(i8* %a) sanitize_hw
 
 ; NO-MATCH-ALL: br i1 %[[F]], label {{.*}}, label {{.*}}, !prof {{.*}}
 
-; ABORT: call void asm sideeffect "brk #2304", "{x0}"(i64 %[[A]])
-; ABORT: unreachable
-; RECOVER: call void asm sideeffect "brk #2336", "{x0}"(i64 %[[A]])
-; RECOVER: br label
+; CHECK: call void asm sideeffect "brk #2336", "{x0}"(i64 %[[A]])
+; CHECK: br label
 
 ; CHECK: %[[G:[^ ]*]] = load i8, i8* %a, align 4
 ; CHECK: ret i8 %[[G]]

Modified: llvm/trunk/test/Instrumentation/HWAddressSanitizer/prologue.ll
URL: http://llvm.org/viewvc/llvm-project/llvm/trunk/test/Instrumentation/HWAddressSanitizer/prologue.ll?rev=351920&r1=351919&r2=351920&view=diff
==============================================================================
--- llvm/trunk/test/Instrumentation/HWAddressSanitizer/prologue.ll (original)
+++ llvm/trunk/test/Instrumentation/HWAddressSanitizer/prologue.ll Tue Jan 22 18:20:10 2019
@@ -23,10 +23,10 @@ define i32 @test_load(i32* %a) sanitize_
 ; CHECK-LABEL: @test_load
 ; CHECK: entry:
 
-; CHECK-IFUNC:   %[[A:[^ ]*]] = call i64 asm "", "=r,0"([0 x i8]* @__hwasan_shadow)
-; CHECK-IFUNC:   add i64 %{{.*}}, %[[A]]
+; CHECK-IFUNC:   %[[A:[^ ]*]] = call i8* asm "", "=r,0"([0 x i8]* @__hwasan_shadow)
+; CHECK-IFUNC:   @llvm.hwasan.check.memaccess(i8* %[[A]]
 
-; CHECK-GLOBAL: load i64, i64* @__hwasan_shadow_memory_dynamic_address
+; CHECK-GLOBAL: load i8*, i8** @__hwasan_shadow_memory_dynamic_address
 
 ; CHECK-TLS:   %[[A:[^ ]*]] = call i8* @llvm.thread.pointer()
 ; CHECK-TLS:   %[[B:[^ ]*]] = getelementptr i8, i8* %[[A]], i32 48
@@ -54,10 +54,10 @@ define void @test_alloca() sanitize_hwad
 ; CHECK-LABEL: @test_alloca
 ; CHECK: entry:
 
-; CHECK-IFUNC:   %[[A:[^ ]*]] = call i64 asm "", "=r,0"([0 x i8]* @__hwasan_shadow)
-; CHECK-IFUNC:   add i64 %{{.*}}, %[[A]]
+; CHECK-IFUNC:   %[[A:[^ ]*]] = call i8* asm "", "=r,0"([0 x i8]* @__hwasan_shadow)
+; CHECK-IFUNC:   getelementptr i8, i8* %[[A]]
 
-; CHECK-GLOBAL: load i64, i64* @__hwasan_shadow_memory_dynamic_address
+; CHECK-GLOBAL: load i8*, i8** @__hwasan_shadow_memory_dynamic_address
 
 ; CHECK-TLS:   %[[A:[^ ]*]] = call i8* @llvm.thread.pointer()
 ; CHECK-TLS:   %[[B:[^ ]*]] = getelementptr i8, i8* %[[A]], i32 48




More information about the llvm-commits mailing list