[cfi] Avoid branch veneers in jump tables when possible.

[eugenis]

Summary

When jumptable encoding does not match target code encoding (arm vs
thumb), a veneer is inserted by the linker. We can not avoid this
in all cases, because entries within one jumptable must have the same
encoding, but we can make it less common by selecting the jumptable
encoding to match the majority of its targets.

This change only covers FullLTO, and not ThinLTO.

[pcc]

llvm/lib/Transforms/IPO/LowerTypeTests.cpp

1121

It seems arbitrary for the module architecture to be the tie breaker. With full LTO I think it will just be the same architecture as the first (or maybe last?) input file. I'd remove this part.

[eugenis]

addressed pcc's comment

[eugenis]

update the test

[pcc]

LGTM

llvm/lib/Transforms/IPO/LowerTypeTests.cpp

void lowerTypeTestCalls(
       const DenseMap<GlobalTypeMember *, uint64_t> &GlobalLayout);
   Value *lowerTypeTestCall(Metadata *TypeId, CallInst *CI,
                            const TypeIdLowering &TIL);
   void buildBitSetsFromGlobalVariables(ArrayRef<Metadata *> TypeIds,
                                        ArrayRef<GlobalTypeMember *> Globals);
   unsigned getJumpTableEntrySize();
   Type *getJumpTableEntryType();
   void createJumpTableEntry(raw_ostream &AsmOS, raw_ostream &ConstraintOS,
+                            Triple::ArchType JumpTableArch,
                             SmallVectorImpl<Value *> &AsmArgs, Function *Dest);
   void verifyTypeMDNode(GlobalObject *GO, MDNode *Type);
   void buildBitSetsFromFunctions(ArrayRef<Metadata *> TypeIds,
                                  ArrayRef<GlobalTypeMember *> Functions);
   void buildBitSetsFromFunctionsNative(ArrayRef<Metadata *> TypeIds,
                                     ArrayRef<GlobalTypeMember *> Functions);
   void buildBitSetsFromFunctionsWASM(ArrayRef<Metadata *> TypeIds,
                                      ArrayRef<GlobalTypeMember *> Functions);
unsigned LowerTypeTestsModule::getJumpTableEntrySize() {
   }
 }
 
 // Create a jump table entry for the target. This consists of an instruction
 // sequence containing a relative branch to Dest. Appends inline asm text,
 // constraints and arguments to AsmOS, ConstraintOS and AsmArgs.
 void LowerTypeTestsModule::createJumpTableEntry(
     raw_ostream &AsmOS, raw_ostream &ConstraintOS,
-    SmallVectorImpl<Value *> &AsmArgs, Function *Dest) {
+    Triple::ArchType JumpTableArch, SmallVectorImpl<Value *> &AsmArgs,
+    Function *Dest) {
   unsigned ArgIndex = AsmArgs.size();
 
-  if (Arch == Triple::x86 || Arch == Triple::x86_64) {
+  if (JumpTableArch == Triple::x86 || JumpTableArch == Triple::x86_64) {
     AsmOS << "jmp ${" << ArgIndex << ":c}@plt\n";
     AsmOS << "int3\nint3\nint3\n";
-  } else if (Arch == Triple::arm || Arch == Triple::aarch64) {
+  } else if (JumpTableArch == Triple::arm || JumpTableArch == Triple::aarch64) {
     AsmOS << "b $" << ArgIndex << "\n";
-  } else if (Arch == Triple::thumb) {
+  } else if (JumpTableArch == Triple::thumb) {
     AsmOS << "b.w $" << ArgIndex << "\n";
   } else {
     report_fatal_error("Unsupported architecture for jump tables");
   }
 
   ConstraintOS << (ArgIndex > 0 ? ",s" : "s");
   AsmArgs.push_back(Dest);
 }
void LowerTypeTestsModule::replaceWeakDeclarationWithJumpTablePtr(
   Constant *Target = ConstantExpr::getSelect(
       ConstantExpr::getICmp(CmpInst::ICMP_NE, F,
                             Constant::getNullValue(F->getType())),
       JT, Constant::getNullValue(F->getType()));
   PlaceholderFn->replaceAllUsesWith(Target);
   PlaceholderFn->eraseFromParent();
 }
 
+static bool isThumbFunction(Function *F, Triple::ArchType ModuleArch) {
+  Attribute TFAttr = F->getFnAttribute("target-features");
+  if (!TFAttr.hasAttribute(Attribute::None)) {
+    SmallVector<StringRef, 6> Features;
+    TFAttr.getValueAsString().split(Features, ',');
+    for (StringRef Feature : Features) {
+      if (Feature == "-thumb-mode")
+        return false;
+      else if (Feature == "+thumb-mode")
+        return true;
+    }
+  }
+
+  return ModuleArch == Triple::thumb;
+}
+
+// Each jump table must be either ARM or Thumb as a whole for the bit-test math
+// to work. Pick one that matches the majority of members to minimize interop
+// veneers inserted by the linker.
+static Triple::ArchType
+selectJumpTableArmEncoding(ArrayRef<GlobalTypeMember *> Functions,
+                           Triple::ArchType ModuleArch) {
+  if (ModuleArch != Triple::arm && ModuleArch != Triple::thumb)
+    return ModuleArch;
+
+  unsigned ArmCount = 0, ThumbCount = 0;
+  for (const auto GTM : Functions) {
+    if (!GTM->isDefinition()) {
+      // PLT stubs are always ARM.
+      ++ArmCount;
+      continue;
+    }
+
+    Function *F = cast<Function>(GTM->getGlobal());
+    ++(isThumbFunction(F, ModuleArch) ? ThumbCount : ArmCount);
+  }
+
+  return ArmCount > ThumbCount ? Triple::arm : Triple::thumb;
+}
+
 void LowerTypeTestsModule::createJumpTable(
     Function *F, ArrayRef<GlobalTypeMember *> Functions) {
   std::string AsmStr, ConstraintStr;
   raw_string_ostream AsmOS(AsmStr), ConstraintOS(ConstraintStr);
   SmallVector<Value *, 16> AsmArgs;
   AsmArgs.reserve(Functions.size() * 2);
 
+  Triple::ArchType JumpTableArch = selectJumpTableArmEncoding(Functions, Arch);
+
   for (unsigned I = 0; I != Functions.size(); ++I)
-    createJumpTableEntry(AsmOS, ConstraintOS, AsmArgs,
+    createJumpTableEntry(AsmOS, ConstraintOS, JumpTableArch, AsmArgs,
                          cast<Function>(Functions[I]->getGlobal()));
 
   // Try to emit the jump table at the end of the text segment.
   // Jump table must come after __cfi_check in the cross-dso mode.
   // FIXME: this magic section name seems to do the trick.
   F->setSection(ObjectFormat == Triple::MachO
                     ? "__TEXT,__text,regular,pure_instructions"
                     : ".text.cfi");
   // Align the whole table by entry size.
   F->setAlignment(getJumpTableEntrySize());
   // Skip prologue.
   // Disabled on win32 due to https://llvm.org/bugs/show_bug.cgi?id=28641#c3.
   // Luckily, this function does not get any prologue even without the
   // attribute.
   if (OS != Triple::Win32)
     F->addFnAttr(llvm::Attribute::Naked);
-  // Thumb jump table assembly needs Thumb2. The following attribute is added by
-  // Clang for -march=armv7.
-  if (Arch == Triple::thumb)
+  if (JumpTableArch == Triple::arm)
+    F->addFnAttr("target-features", "-thumb-mode");
+  if (JumpTableArch == Triple::thumb) {
+    F->addFnAttr("target-features", "+thumb-mode");
+    // Thumb jump table assembly needs Thumb2. The following attribute is added
+    // by Clang for -march=armv7.
     F->addFnAttr("target-cpu", "cortex-a8");
+  }
 
   BasicBlock *BB = BasicBlock::Create(M.getContext(), "entry", F);
   IRBuilder<> IRB(BB);
 
   SmallVector<Type *, 16> ArgTypes;
   ArgTypes.reserve(AsmArgs.size());
   for (const auto &Arg : AsmArgs)
     ArgTypes.push_back(Arg->getType());

llvm/test/Transforms/LowerTypeTests/function-arm-thumb.ll

+; RUN: opt -S -mtriple=arm-unknown-linux-gnu -lowertypetests -lowertypetests-summary-action=export -lowertypetests-read-summary=%S/Inputs/use-typeid1-typeid2.yaml -lowertypetests-write-summary=%t < %s | FileCheck %s
+
+target datalayout = "e-p:64:64"
+
+define void @f1() "target-features"="+thumb-mode" !type !0 {
+  ret void
+}
+
+define void @g1() "target-features"="-thumb-mode" !type !0 {
+  ret void
+}
+
+define void @f2() "target-features"="+thumb-mode" !type !1 {
+  ret void
+}
+
+define void @g2() "target-features"="-thumb-mode" !type !1 {
+  ret void
+}
+
+define void @h2() "target-features"="-thumb-mode" !type !1 {
+  ret void
+}
+
+!0 = !{i32 0, !"typeid1"}
+!1 = !{i32 0, !"typeid2"}
+
+; CHECK: define private void {{.*}} #[[AT:.*]] section ".text.cfi" align 4 {
+; CHECK-NEXT: entry:
+; CHECK-NEXT:  call void asm sideeffect "b.w $0\0Ab.w $1\0A", "s,s"(void ()* @f1.cfi, void ()* @g1.cfi)
+; CHECK-NEXT:  unreachable
+; CHECK-NEXT: }
+
+; CHECK: define private void {{.*}} #[[AA:.*]] section ".text.cfi" align 4 {
+; CHECK-NEXT: entry:
+; CHECK-NEXT:  call void asm sideeffect "b $0\0Ab $1\0Ab $2\0A", "s,s,s"(void ()* @f2.cfi, void ()* @g2.cfi, void ()* @h2.cfi)
+; CHECK-NEXT:  unreachable
+; CHECK-NEXT: }
+
+; CHECK-DAG: attributes #[[AA]] = { naked "target-features"="-thumb-mode" }
+; CHECK-DAG: attributes #[[AT]] = { naked "target-cpu"="cortex-a8" "target-features"="+thumb-mode" }