diff --git a/emscripten-version.txt b/emscripten-version.txt
index fb1348a3e91..2286388304d 100644
--- a/emscripten-version.txt
+++ b/emscripten-version.txt
@@ -1,2 +1,2 @@
-1.29.3
+1.29.4
 
diff --git a/lib/Target/JSBackend/JSBackend.cpp b/lib/Target/JSBackend/JSBackend.cpp
index bc78c3f8e27..5c468f16863 100644
--- a/lib/Target/JSBackend/JSBackend.cpp
+++ b/lib/Target/JSBackend/JSBackend.cpp
@@ -2536,7 +2536,9 @@ void JSWriter::printModuleBody() {
       } else {
         Out << ", ";
       }
-      Out << "\"_" << I->getName() << '"';
+      std::string name = I->getName();
+      sanitizeGlobal(name);
+      Out << "\"" << name << '"';
     }
   }
   Out << "],";
diff --git a/lib/Transforms/NaCl/ExpandArithWithOverflow.cpp b/lib/Transforms/NaCl/ExpandArithWithOverflow.cpp
index 0fb248668d3..0e06051628b 100644
--- a/lib/Transforms/NaCl/ExpandArithWithOverflow.cpp
+++ b/lib/Transforms/NaCl/ExpandArithWithOverflow.cpp
@@ -64,6 +64,14 @@ static uint64_t UintTypeMax(unsigned Bits) {
   return (((uint64_t) 1) << Bits) - 1;
 }
 
+static int64_t SintTypeMax(unsigned Bits) {
+  return (((int64_t) 1) << (Bits-1)) - 1;
+}
+
+static int64_t SintTypeMin(unsigned Bits) {
+  return -(((int64_t) 1) << (Bits-1));
+}
+
 static Value *CreateInsertValue(Value *StructVal, unsigned Index,
                                 Value *Field, Instruction *BasedOn) {
   SmallVector<unsigned, 1> EVIndexes;
@@ -73,12 +81,18 @@ static Value *CreateInsertValue(Value *StructVal, unsigned Index,
                        BasedOn->getName() + ".insert", BasedOn), BasedOn);
 }
 
-static bool ExpandOpForIntSize(Module *M, unsigned Bits, bool Mul) {
+static bool ExpandOpForIntSize(Module *M, unsigned Bits, bool Mul, bool Signed) {
   IntegerType *IntTy = IntegerType::get(M->getContext(), Bits);
   SmallVector<Type *, 1> Types;
   Types.push_back(IntTy);
-  Intrinsic::ID ID = (Mul ? Intrinsic::umul_with_overflow
-                          : Intrinsic::uadd_with_overflow);
+  Intrinsic::ID ID;
+  if (!Signed) {
+    ID = (Mul ? Intrinsic::umul_with_overflow
+              : Intrinsic::uadd_with_overflow);
+  } else {
+    ID = (Mul ? Intrinsic::smul_with_overflow
+              : Intrinsic::sadd_with_overflow);
+  }
   std::string Name = Intrinsic::getName(ID, Types);
   Function *Intrinsic = M->getFunction(Name);
   if (!Intrinsic)
@@ -114,7 +128,7 @@ static bool ExpandOpForIntSize(Module *M, unsigned Bits, bool Mul) {
 
     Value *ArithResult, *OverflowResult;
 
-    if (!VariableArg2) {
+    if (!VariableArg2 && !Signed) {
       ArithResult = BinaryOperator::Create(
           (Mul ? Instruction::Mul : Instruction::Add), VariableArg, ConstantArg,
           Call->getName() + ".arith", Call);
@@ -130,13 +144,40 @@ static bool ExpandOpForIntSize(Module *M, unsigned Bits, bool Mul) {
           Call->getName() + ".overflow");
     } else {
       // XXX EMSCRIPTEN: generalize this to nonconstant values, easy for addition
+      //                 also for signed values
+      assert(!Mul);
+      if (!VariableArg2) VariableArg2 = ConstantArg;
       ArithResult = BinaryOperator::Create(Instruction::Add,
           VariableArg, VariableArg2,
           Call->getName() + ".arith", Call);
-      // If x+y < x (or y), unsigned 32 addition, then an overflow occurred
-      OverflowResult = new ICmpInst(
-          Call, CmpInst::ICMP_ULT, ArithResult, VariableArg,
-          Call->getName() + ".overflow");
+      if (!Signed) {
+        // If x+y < x (or y), unsigned 32 addition, then an overflow occurred
+        OverflowResult = new ICmpInst(
+            Call, CmpInst::ICMP_ULT, ArithResult, VariableArg,
+            Call->getName() + ".overflow");
+      } else {
+        // In the signed case, we care if the sum is >127 or <-128. When looked at
+        // as an unsigned number, that is precisely when the sum is >= 128
+        Value *PositiveTemp = BinaryOperator::Create(Instruction::Add,
+            VariableArg, ConstantInt::get(IntTy, SintTypeMin(Bits)),
+            Call->getName() + ".postemp", Call);
+        Value *NegativeTemp = BinaryOperator::Create(Instruction::Add,
+            VariableArg, ConstantInt::get(IntTy, SintTypeMax(Bits)),
+            Call->getName() + ".negtemp", Call);
+        Value *PositiveCheck = new ICmpInst(
+            Call, CmpInst::ICMP_SLT, ArithResult, PositiveTemp,
+            Call->getName() + ".poscheck");
+        Value *NegativeCheck = new ICmpInst(
+            Call, CmpInst::ICMP_SGT, ArithResult, NegativeTemp,
+            Call->getName() + ".negcheck");
+        Value *IsPositive = new ICmpInst(
+            Call, CmpInst::ICMP_SGT, VariableArg, ConstantInt::get(IntTy, 0),
+            Call->getName() + ".ispos");
+        OverflowResult = SelectInst::Create(
+            IsPositive, PositiveCheck, NegativeCheck,
+            Call->getName() + ".select",
+            Call);
+      }
     }
 
     // Construct the struct result.
@@ -153,8 +194,9 @@ static bool ExpandOpForIntSize(Module *M, unsigned Bits, bool Mul) {
 
 static bool ExpandForIntSize(Module *M, unsigned Bits) {
   bool Modified = false;
-  Modified |= ExpandOpForIntSize(M, Bits, true); // Expand umul
-  Modified |= ExpandOpForIntSize(M, Bits, false); // Expand uadd
+  Modified |= ExpandOpForIntSize(M, Bits, true, false); // Expand umul
+  Modified |= ExpandOpForIntSize(M, Bits, false, false); // Expand uadd
+  Modified |= ExpandOpForIntSize(M, Bits, false, true); // Expand sadd (for ubsan) XXX EMSCRIPTEN
   return Modified;
 }