[WebAssembly] Forbid use of EM_ASM with setjmp/longjmp

[quantum]

Summary

We tried to support EM_ASM with setjmp/longjmp in binaryen. But with dynamic
linking thrown into the mix, the code is no longer understandable and cannot
be maintained. We also discovered more bugs in the EM_ASM handling code.

To ensure maintainability and correctness of the binaryen code, EM_ASM will
no longer be supported with setjmp/longjmp. This is probably fine since the
support was added recently and haven't be published.

[kripken]

lgtm module the error message, let's find the best phrasing there.

llvm/lib/Target/WebAssembly/WebAssemblyLowerEmscriptenEHSjLj.cpp

1004 ↗

(On Diff #215635)

How about "alongside" instead of "with", to hint that the issue is using both in the same function?

And/or maybe replace the final "." with ", or refactor the EM_ASM to another function"?

[tlively]

llvm/lib/Target/WebAssembly/WebAssemblyLowerEmscriptenEHSjLj.cpp

550 ↗

(On Diff #215635)

Where does this list of functions come from? How do we know it's exhaustive?

554 ↗

(On Diff #215635)

Maybe use std::any_of over an array of the function names as plain strings to reduce repetition?

[quantum]

Address review comments.

llvm/lib/Target/WebAssembly/WebAssemblyLowerEmscriptenEHSjLj.cpp

550 ↗

(On Diff #215635)

Stated that the list is exhaustive and link the header where these functions came from.

554 ↗

(On Diff #215635)

This is the style used in this file, so we are doing it for consistency.

1004 ↗

(On Diff #215635)

Done.

[sbc100]

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

555

Why not just compare names here?

[quantum]

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

555

For consistency with WebAssemblyLowerEmscriptenEHSjLj::canLongjmp.

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

class WebAssemblyLowerEmscriptenEHSjLj final : public ModulePass {
   template <typename CallOrInvoke> Value *wrapInvoke(CallOrInvoke *CI);
   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;
+  bool isEmAsmCall(Module &M, const Value *Callee) const;
 
   void rebuildSSA(Function &F);
 
 public:
   static char ID;
 
   WebAssemblyLowerEmscriptenEHSjLj(bool EnableEH = true, bool EnableSjLj = true)
       : ModulePass(ID), EnableEH(EnableEH), EnableSjLj(EnableSjLj) {
if (Callee == BeginCatchF || Callee == EndCatchF ||
       Callee == AllocExceptionF || Callee == ThrowF || Callee == TerminateF ||
       Callee == GetTempRet0Func || Callee == SetTempRet0Func)
     return false;
 
   // Otherwise we don't know
   return true;
 }
 
+bool WebAssemblyLowerEmscriptenEHSjLj::isEmAsmCall(Module &M,
+                                                   const Value *Callee) const {
+  // This is an exhaustive list from Emscripten's <emscripten/em_asm.h>.
+  Function *EmAsmConstIntF = M.getFunction("emscripten_asm_const_int");
+  Function *EmAsmConstDoubleF = M.getFunction("emscripten_asm_const_double");
+  Function *EmAsmConstIntSyncMainF =
+      M.getFunction("emscripten_asm_const_int_sync_on_main_thread");
+  Function *EmAsmConstDoubleSyncMainF =
+      M.getFunction("emscripten_asm_const_double_sync_on_main_thread");
+  Function *EmAsmConstAsyncMainF =
+      M.getFunction("emscripten_asm_const_async_on_main_thread");
+
+  return Callee == EmAsmConstIntF || Callee == EmAsmConstDoubleF ||
+         Callee == EmAsmConstIntSyncMainF ||
+         Callee == EmAsmConstDoubleSyncMainF || Callee == EmAsmConstAsyncMainF;
+}
+
 // Generate testSetjmp function call seqence with preamble and postamble.
 // The code this generates is equivalent to the following JavaScript code:
 // if (%__THREW__.val != 0 & threwValue != 0) {
 //   %label = _testSetjmp(mem[%__THREW__.val], setjmpTable, setjmpTableSize);
 //   if (%label == 0)
 //     emscripten_longjmp(%__THREW__.val, threwValue);
 //   setTempRet0(threwValue);
 // } else {
for (Instruction &I : *BB) {
       assert(!isa<InvokeInst>(&I));
       auto *CI = dyn_cast<CallInst>(&I);
       if (!CI)
         continue;
 
       const Value *Callee = CI->getCalledValue();
       if (!canLongjmp(M, Callee))
         continue;
+      if (isEmAsmCall(M, Callee))
+        report_fatal_error("Cannot use EM_ASM* alongside setjmp/longjmp in " +
+                               F.getName() +
+                               ". Please consider using EM_JS, or move the "
+                               "EM_ASM into another function.",
+                           false);
 
       Value *Threw = nullptr;
       BasicBlock *Tail;
       if (Callee->getName().startswith(InvokePrefix)) {
         // If invoke wrapper has already been generated for this call in
         // previous EH phase, search for the load instruction
         // %__THREW__.val = __THREW__;
         // in postamble after the invoke wrapper call