[WebAssembly] Update WebAssemblyLowerEmscriptenEHSjLj to support separate compilation.

[sbc100]

Summary

Previously this pass would assume that it was running over the
entire program. This meant it could get away with creating the
two helper functions:

• setThrew
• setTempRet0

Instead we now assume these will be provided at link time. In
emscripten this will most likely be done via libcxxabi.

Additionally we previously created three global variable:

• THREW
• _threwValue
• __tempRet0

These are now allow assumed to be available at link time, but
we also have to handle the case where they exist in the module
being compiled (e.g. when we are doing LTO with libcxxabi).

[sbc100]

• comment

[sbc100]

-revert

[aheejin]

1. For emscripten-based EH, we don't use a separate libcxxabi; emscripten provides handful of library functions in JavaScript. Are you planning to provide them from emscripten and import them from wasm?
2. Because now this does not create those functions and global variables anymore, certain tests, such as lower-em-exceptions.ll, will fail. I guess you should update that as well.

[sbc100]

Emscripten does have at least part of libcxxabi. All I need a library file where I can add these things. Here is my provisional change: https://github.com/kripken/emscripten/compare/sjlj?expand=1

Ah yes, thanks for pointing out that test. I will run them all an fix up.

[aheejin]

lib/Target/WebAssembly/WebAssemblyLowerEmscriptenEHSjLj.cpp

72 ↗

(On Diff #155088)

Because you are gonna provide setThrew and setTempRet0 in a separate cpp file, "their values should be set in JS code" does not strictly sound correct, because that cpp file is gonna be a part of wasm as well. Could you elaborate that these functions will be provided from library that's a part of emscripten?

And for the equivalent asm.js version of JS code below ↓ , I know that's not a part of your change, but I think we can delete this by now, because we are gonna switch to the wasm as a default backend anyway. Maybe show the code in cpp form would be better.

339 ↗

(On Diff #155088)

Nit: LLVM style is GlobalVariable *V

632 ↗

(On Diff #155088)

threwValue -> __threwValue

[jgravelle-google]

lib/Target/WebAssembly/WebAssemblyLowerEmscriptenEHSjLj.cpp

487 ↗

(On Diff #155088)

Nit: unneeded braces

[sbc100]

Abandoning in favor of weak linking approach: https://reviews.llvm.org/D49263

[dschuff]

lib/Target/WebAssembly/WebAssemblyLowerEmscriptenEHSjLj.cpp

487 ↗

(On Diff #155088)

Nit: no, those braces are needed IMHO :D

[sbc100]

Re-opening since we need these symbols to be available pre-LTO.. i.e. before codegen.

[sbc100]

Rebase and fix tests

[sbc100]

• feedback

[sbc100]

lib/Target/WebAssembly/WebAssemblyLowerEmscriptenEHSjLj.cpp

487 ↗

(On Diff #155088)

At least in lld the policy is not to include them this these cases.

[sbc100]

• feedback

[sbc100]

ptal

[sbc100]

Rebase

llvm/trunk/lib/Target/WebAssembly/WebAssemblyLowerEmscriptenEHSjLj.cpp

 /// We assume setjmp-longjmp handling always run after EH handling, which means
 /// we don't expect any exception-related instructions when SjLj runs.
 /// FIXME Currently this scheme does not support indirect call of setjmp,
 /// because of the limitation of the scheme itself. fastcomp does not support it
 /// either.
 ///
 /// In detail, this pass does following things:
 ///
-/// 1) Create three global variables: __THREW__, __threwValue, and __tempRet0.
+/// 1) Assumes the existence of global variables: __THREW__, __threwValue, and
+///    __tempRet0.
 ///    __tempRet0 will be set within __cxa_find_matching_catch() function in
 ///    JS library, and __THREW__ and __threwValue will be set in invoke wrappers
 ///    in JS glue code. For what invoke wrappers are, refer to 3). These
 ///    variables are used for both exceptions and setjmp/longjmps.
 ///    __THREW__ indicates whether an exception or a longjmp occurred or not. 0
 ///    means nothing occurred, 1 means an exception occurred, and other numbers
 ///    mean a longjmp occurred. In the case of longjmp, __threwValue variable
 ///    indicates the corresponding setjmp buffer the longjmp corresponds to.
 ///    In exception handling, __tempRet0 indicates the type of an exception
 ///    caught, and in setjmp/longjmp, it means the second argument to longjmp
 ///    function.
 ///
 /// * Exception handling
 ///
-/// 2) Create setThrew and setTempRet0 functions.
-///    The global variables created in 1) will exist in wasm address space,
-///    but their values should be set in JS code, so we provide these functions
+/// 2) We assume the existence of setThrew and setTempRet0 functions at link
+///    time.
+///    The global variables in 1) will exist in wasm address space,
+///    but their values should be set in JS code, so these functions
 ///    as interfaces to JS glue code. These functions are equivalent to the
 ///    following JS functions, which actually exist in asm.js version of JS
 ///    library.
 ///
 ///    function setThrew(threw, value) {
 ///      if (__THREW__ == 0) {
 ///        __THREW__ = threw;
 ///        __threwValue = value;
class WebAssemblyLowerEmscriptenEHSjLj final : public ModulePass {
   void wrapTestSetjmp(BasicBlock *BB, Instruction *InsertPt, Value *Threw,
                       Value *SetjmpTable, Value *SetjmpTableSize, Value *&Label,
                       Value *&LongjmpResult, BasicBlock *&EndBB);
   template <typename CallOrInvoke> Function *getInvokeWrapper(CallOrInvoke *CI);
 
   bool areAllExceptionsAllowed() const { return EHWhitelistSet.empty(); }
   bool canLongjmp(Module &M, const Value *Callee) const;
 
-  void createSetThrewFunction(Module &M);
-  void createSetTempRet0Function(Module &M);
-
   void rebuildSSA(Function &F);
 
 public:
   static char ID;
 
   WebAssemblyLowerEmscriptenEHSjLj(bool EnableEH = true, bool EnableSjLj = true)
       : ModulePass(ID), EnableEH(EnableEH), EnableSjLj(EnableSjLj),
         ThrewGV(nullptr), ThrewValueGV(nullptr), TempRet0GV(nullptr),
if (const auto *F = dyn_cast<const Function>(V)) {
     if (Name == "setjmp" || Name == "longjmp")
       return false;
     return !F->doesNotThrow();
   }
   // not a function, so an indirect call - can throw, we can't tell
   return true;
 }
 
-static GlobalVariable *createGlobalVariableI32(Module &M, IRBuilder<> &IRB,
-                                               const char *Name) {
+// Get a global variable with the given name.  If it doesn't exist declare it,
+// which will generate an import and asssumes that it will exist at link time.
+static GlobalVariable *getGlobalVariableI32(Module &M, IRBuilder<> &IRB,
   if (M.getNamedGlobal(Name))
     report_fatal_error(Twine("variable name is reserved: ") + Name);
 
   return new GlobalVariable(M, IRB.getInt32Ty(), false,
-                            GlobalValue::WeakODRLinkage, IRB.getInt32(0), Name);
+                            GlobalValue::ExternalLinkage, nullptr, Name);
 }
 
 // Simple function name mangler.
 // This function simply takes LLVM's string representation of parameter types
 // and concatenate them with '_'. There are non-alphanumeric characters but llc
 // is ok with it, and we need to postprocess these names after the lowering
 // phase anyway.
 static std::string getSignature(FunctionType *FTy) {
void WebAssemblyLowerEmscriptenEHSjLj::wrapTestSetjmp(
   LabelPHI->addIncoming(IRB.getInt32(-1), ElseBB1);
 
   // Output parameter assignment
   Label = LabelPHI;
   EndBB = EndBB1;
   LongjmpResult = IRB.CreateLoad(TempRet0GV, "longjmp_result");
 }
 
-// Create setThrew function
-// function setThrew(threw, value) {
-//   if (__THREW__ == 0) {
-//     __THREW__ = threw;
-//     __threwValue = value;
-//   }
-// }
-void WebAssemblyLowerEmscriptenEHSjLj::createSetThrewFunction(Module &M) {
-  LLVMContext &C = M.getContext();
-  IRBuilder<> IRB(C);
-
-  if (M.getNamedGlobal("setThrew"))
-    report_fatal_error("setThrew already exists");
-
-  Type *Params[] = {IRB.getInt32Ty(), IRB.getInt32Ty()};
-  FunctionType *FTy = FunctionType::get(IRB.getVoidTy(), Params, false);
-  Function *F =
-      Function::Create(FTy, GlobalValue::WeakODRLinkage, "setThrew", &M);
-  Argument *Arg1 = &*(F->arg_begin());
-  Argument *Arg2 = &*std::next(F->arg_begin());
-  Arg1->setName("threw");
-  Arg2->setName("value");
-  BasicBlock *EntryBB = BasicBlock::Create(C, "entry", F);
-  BasicBlock *ThenBB = BasicBlock::Create(C, "if.then", F);
-  BasicBlock *EndBB = BasicBlock::Create(C, "if.end", F);
-
-  IRB.SetInsertPoint(EntryBB);
-  Value *Threw = IRB.CreateLoad(ThrewGV, ThrewGV->getName() + ".val");
-  Value *Cmp = IRB.CreateICmpEQ(Threw, IRB.getInt32(0), "cmp");
-  IRB.CreateCondBr(Cmp, ThenBB, EndBB);
-
-  IRB.SetInsertPoint(ThenBB);
-  IRB.CreateStore(Arg1, ThrewGV);
-  IRB.CreateStore(Arg2, ThrewValueGV);
-  IRB.CreateBr(EndBB);
-
-  IRB.SetInsertPoint(EndBB);
-  IRB.CreateRetVoid();
-}
-
-// Create setTempRet0 function
-// function setTempRet0(value) {
-//   __tempRet0 = value;
-// }
-void WebAssemblyLowerEmscriptenEHSjLj::createSetTempRet0Function(Module &M) {
-  LLVMContext &C = M.getContext();
-  IRBuilder<> IRB(C);
-
-  if (M.getNamedGlobal("setTempRet0"))
-    report_fatal_error("setTempRet0 already exists");
-  Type *Params[] = {IRB.getInt32Ty()};
-  FunctionType *FTy = FunctionType::get(IRB.getVoidTy(), Params, false);
-  Function *F =
-      Function::Create(FTy, GlobalValue::WeakODRLinkage, "setTempRet0", &M);
-  F->arg_begin()->setName("value");
-  BasicBlock *EntryBB = BasicBlock::Create(C, "entry", F);
-  IRB.SetInsertPoint(EntryBB);
-  IRB.CreateStore(&*F->arg_begin(), TempRet0GV);
-  IRB.CreateRetVoid();
-}
-
 void WebAssemblyLowerEmscriptenEHSjLj::rebuildSSA(Function &F) {
   DominatorTree &DT = getAnalysis<DominatorTreeWrapperPass>(F).getDomTree();
   DT.recalculate(F); // CFG has been changed
   SSAUpdater SSA;
   for (BasicBlock &BB : F) {
     for (Instruction &I : BB) {
       for (auto UI = I.use_begin(), UE = I.use_end(); UI != UE;) {
         Use &U = *UI;
bool WebAssemblyLowerEmscriptenEHSjLj::runOnModule(Module &M) {
   IRBuilder<> IRB(C);
 
   Function *SetjmpF = M.getFunction("setjmp");
   Function *LongjmpF = M.getFunction("longjmp");
   bool SetjmpUsed = SetjmpF && !SetjmpF->use_empty();
   bool LongjmpUsed = LongjmpF && !LongjmpF->use_empty();
   bool DoSjLj = EnableSjLj && (SetjmpUsed || LongjmpUsed);
 
-  // Create global variables __THREW__, threwValue, and __tempRet0, which are
-  // used in common for both exception handling and setjmp/longjmp handling
-  ThrewGV = createGlobalVariableI32(M, IRB, "__THREW__");
-  ThrewValueGV = createGlobalVariableI32(M, IRB, "__threwValue");
-  TempRet0GV = createGlobalVariableI32(M, IRB, "__tempRet0");
+  // Declare (or get) global variables __THREW__, __threwValue, and __tempRet0,
+  // which are used in common for both exception handling and setjmp/longjmp
+  // handling
+  ThrewGV = getGlobalVariableI32(M, IRB, "__THREW__");
+  ThrewValueGV = getGlobalVariableI32(M, IRB, "__threwValue");
+  TempRet0GV = getGlobalVariableI32(M, IRB, "__tempRet0");
 
   bool Changed = false;
 
   // Exception handling
   if (EnableEH) {
     // Register __resumeException function
     FunctionType *ResumeFTy =
         FunctionType::get(IRB.getVoidTy(), IRB.getInt8PtrTy(), false);
for (User *U : SetjmpF->users()) {
       SetjmpUsers.insert(UI->getFunction());
     }
     for (Function *F : SetjmpUsers)
       runSjLjOnFunction(*F);
   }
 
   if (!Changed) {
     // Delete unused global variables and functions
-    ThrewGV->eraseFromParent();
-    ThrewValueGV->eraseFromParent();
-    TempRet0GV->eraseFromParent();
     if (ResumeF)
       ResumeF->eraseFromParent();
     if (EHTypeIDF)
       EHTypeIDF->eraseFromParent();
     if (EmLongjmpF)
       EmLongjmpF->eraseFromParent();
     if (SaveSetjmpF)
       SaveSetjmpF->eraseFromParent();
     if (TestSetjmpF)
       TestSetjmpF->eraseFromParent();
     return false;
   }
 
-  // If we have made any changes while doing exception handling or
-  // setjmp/longjmp handling, we have to create these functions for JavaScript
-  // to call.
-  createSetThrewFunction(M);
-  createSetTempRet0Function(M);
-
   return true;
 }
 
 bool WebAssemblyLowerEmscriptenEHSjLj::runEHOnFunction(Function &F) {
   Module &M = *F.getParent();
   LLVMContext &C = F.getContext();
   IRBuilder<> IRB(C);
   bool Changed = false;

llvm/trunk/test/CodeGen/WebAssembly/lower-em-exceptions.ll

 ; RUN: opt < %s -wasm-lower-em-ehsjlj -S | FileCheck %s
 
 target datalayout = "e-m:e-p:32:32-i64:64-n32:64-S128"
 target triple = "wasm32-unknown-unknown"
 
 @_ZTIi = external constant i8*
 @_ZTIc = external constant i8*
-; CHECK-DAG: __THREW__ = weak_odr global i32 0
-; CHECK-DAG: __threwValue = weak_odr global i32 0
-; CHECK-DAG: __tempRet0 = weak_odr global i32 0
+; CHECK-DAG: __THREW__ = external global i32
+; CHECK-DAG: __threwValue = external global i32
+; CHECK-DAG: __tempRet0 = external global i32
 
 ; Test invoke instruction with clauses (try-catch block)
 define void @clause() personality i8* bitcast (i32 (...)* @__gxx_personality_v0 to i8*) {
 ; CHECK-LABEL: @clause(
 entry:
   invoke void @foo(i32 3)
           to label %invoke.cont unwind label %lpad
 ; CHECK: entry:

 declare i8* @__cxa_begin_catch(i8*)
 declare void @__cxa_end_catch()
 declare void @__cxa_call_unexpected(i8*)
 
 ; JS glue functions and invoke wrappers declaration
 ; CHECK-DAG: declare void @__resumeException(i8*)
 ; CHECK-DAG: declare void @__invoke_void_i32(void (i32)*, i32)
 ; CHECK-DAG: declare i8* @__cxa_find_matching_catch_4(i8*, i8*)
-
-; setThrew function creation
-; CHECK-LABEL: define weak_odr void @setThrew(i32 %threw, i32 %value) {
-; CHECK: entry:
-; CHECK-NEXT: %__THREW__.val = load i32, i32* @__THREW__
-; CHECK-NEXT: %cmp = icmp eq i32 %__THREW__.val, 0
-; CHECK-NEXT: br i1 %cmp, label %if.then, label %if.end
-; CHECK: if.then:
-; CHECK-NEXT: store i32 %threw, i32* @__THREW__
-; CHECK-NEXT: store i32 %value, i32* @__threwValue
-; CHECK-NEXT: br label %if.end
-; CHECK: if.end:
-; CHECK-NEXT: ret void
-; CHECK: }
-
-; setTempRet0 function creation
-; CHECK-LABEL: define weak_odr void @setTempRet0(i32 %value) {
-; CHECK: entry:
-; CHECK-NEXT: store i32 %value, i32* @__tempRet0
-; CHECK-NEXT: ret void
-; CHECK: }

llvm/trunk/test/CodeGen/WebAssembly/lower-em-sjlj.ll

 ; RUN: opt < %s -wasm-lower-em-ehsjlj -S | FileCheck %s
 
 target datalayout = "e-m:e-p:32:32-i64:64-n32:64-S128"
 target triple = "wasm32-unknown-unknown"
 
 %struct.__jmp_buf_tag = type { [6 x i32], i32, [32 x i32] }
 
 @global_var = hidden global i32 0, align 4
-; CHECK-DAG: __THREW__ = weak_odr global i32 0
-; CHECK-DAG: __threwValue = weak_odr global i32 0
-; CHECK-DAG: __tempRet0 = weak_odr global i32 0
+; CHECK-DAG: __THREW__ = external global i32
+; CHECK-DAG: __threwValue = external global i32
+; CHECK-DAG: __tempRet0 = external global i32
 
 ; Test a simple setjmp - longjmp sequence
 define hidden void @setjmp_longjmp() {
 ; CHECK-LABEL: @setjmp_longjmp
 entry:
   %buf = alloca [1 x %struct.__jmp_buf_tag], align 16
   %arraydecay = getelementptr inbounds [1 x %struct.__jmp_buf_tag], [1 x %struct.__jmp_buf_tag]* %buf, i32 0, i32 0
   %call = call i32 @setjmp(%struct.__jmp_buf_tag* %arraydecay) #0