[WebAssembly] Add support for weak symbols in the binary format
This also introduces the updated format for the
"linking" section which can represent extra
symbol information. See:
https://github.com/WebAssembly/tool-conventions/pull/10
Differential Revision: https://reviews.llvm.org/D34019
llvm-svn: 305769
diff --git a/llvm/lib/MC/WasmObjectWriter.cpp b/llvm/lib/MC/WasmObjectWriter.cpp
index 7ac4fb8..a4faaf0 100644
--- a/llvm/lib/MC/WasmObjectWriter.cpp
+++ b/llvm/lib/MC/WasmObjectWriter.cpp
@@ -156,9 +156,18 @@
Out << "Off=" << Offset << ", Sym=" << Symbol << ", Addend=" << Addend
<< ", Type=" << Type << ", FixupSection=" << FixupSection;
}
- void dump() const { print(errs()); }
+
+#if !defined(NDEBUG) || defined(LLVM_ENABLE_DUMP)
+ LLVM_DUMP_METHOD void dump() const { print(dbgs()); }
+#endif
};
+inline raw_ostream &operator<<(raw_ostream &OS,
+ const WasmRelocationEntry &Rel) {
+ Rel.print(OS);
+ return OS;
+}
+
class WasmObjectWriter : public MCObjectWriter {
/// Helper struct for containing some precomputed information on symbols.
struct WasmSymbolData {
@@ -229,6 +238,11 @@
void writeObject(MCAssembler &Asm, const MCAsmLayout &Layout) override;
+ void writeString(const StringRef Str) {
+ encodeULEB128(Str.size(), getStream());
+ writeBytes(Str);
+ }
+
void writeValueType(wasm::ValType Ty) {
encodeSLEB128(int32_t(Ty), getStream());
}
@@ -250,7 +264,8 @@
uint32_t NumFuncImports);
void writeCodeRelocSection();
void writeDataRelocSection(uint64_t DataSectionHeaderSize);
- void writeLinkingMetaDataSection(bool HasStackPointer,
+ void writeLinkingMetaDataSection(ArrayRef<StringRef> WeakSymbols,
+ bool HasStackPointer,
uint32_t StackPointerGlobal);
void applyRelocations(ArrayRef<WasmRelocationEntry> Relocations,
@@ -282,6 +297,7 @@
assert((Name != nullptr) == (SectionId == wasm::WASM_SEC_CUSTOM) &&
"Only custom sections can have names");
+ DEBUG(dbgs() << "startSection " << SectionId << ": " << Name << "\n");
encodeULEB128(SectionId, getStream());
Section.SizeOffset = getStream().tell();
@@ -295,8 +311,8 @@
// Custom sections in wasm also have a string identifier.
if (SectionId == wasm::WASM_SEC_CUSTOM) {
- encodeULEB128(strlen(Name), getStream());
- writeBytes(Name);
+ assert(Name);
+ writeString(StringRef(Name));
}
}
@@ -307,6 +323,7 @@
if (uint32_t(Size) != Size)
report_fatal_error("section size does not fit in a uint32_t");
+ DEBUG(dbgs() << "endSection size=" << Size << "\n");
unsigned Padding = PaddingFor5ByteULEB128(Size);
// Write the final section size to the payload_len field, which follows
@@ -411,6 +428,7 @@
unsigned Type = getRelocType(Target, Fixup);
WasmRelocationEntry Rec(FixupOffset, SymA, C, Type, &FixupSection);
+ DEBUG(dbgs() << "WasmReloc: " << Rec << "\n");
if (FixupSection.hasInstructions())
CodeRelocations.push_back(Rec);
@@ -455,7 +473,7 @@
const MCSymbolWasm *Sym = RelEntry.Symbol;
// For undefined symbols, use a hopefully invalid value.
- if (!Sym->isDefined(false))
+ if (!Sym->isDefined(/*SetUsed=*/false))
return UINT32_MAX;
MCSectionWasm &Section =
@@ -473,17 +491,23 @@
switch (RelEntry.Type) {
case wasm::R_WEBASSEMBLY_TABLE_INDEX_SLEB:
case wasm::R_WEBASSEMBLY_TABLE_INDEX_I32:
- assert(IndirectSymbolIndices.count(RelEntry.Symbol));
+ if (!IndirectSymbolIndices.count(RelEntry.Symbol))
+ report_fatal_error("symbol not found table index space:" +
+ RelEntry.Symbol->getName());
return IndirectSymbolIndices[RelEntry.Symbol];
case wasm::R_WEBASSEMBLY_FUNCTION_INDEX_LEB:
case wasm::R_WEBASSEMBLY_GLOBAL_INDEX_LEB:
case wasm::R_WEBASSEMBLY_GLOBAL_ADDR_LEB:
case wasm::R_WEBASSEMBLY_GLOBAL_ADDR_SLEB:
case wasm::R_WEBASSEMBLY_GLOBAL_ADDR_I32:
- assert(SymbolIndices.count(RelEntry.Symbol));
+ if (!SymbolIndices.count(RelEntry.Symbol))
+ report_fatal_error("symbol not found function/global index space:" +
+ RelEntry.Symbol->getName());
return SymbolIndices[RelEntry.Symbol];
case wasm::R_WEBASSEMBLY_TYPE_INDEX_LEB:
- assert(TypeIndices.count(RelEntry.Symbol));
+ if (!TypeIndices.count(RelEntry.Symbol))
+ report_fatal_error("symbol not found in type index space:" +
+ RelEntry.Symbol->getName());
return TypeIndices[RelEntry.Symbol];
default:
llvm_unreachable("invalid relocation type");
@@ -500,6 +524,7 @@
RelEntry.FixupSection->getSectionOffset() +
RelEntry.Offset;
+ DEBUG(dbgs() << "applyRelocation: " << RelEntry << "\n");
switch (RelEntry.Type) {
case wasm::R_WEBASSEMBLY_TABLE_INDEX_SLEB:
case wasm::R_WEBASSEMBLY_FUNCTION_INDEX_LEB:
@@ -577,6 +602,7 @@
endSection(Section);
}
+
void WasmObjectWriter::writeImportSection(
const SmallVector<WasmImport, 4> &Imports) {
if (Imports.empty())
@@ -587,13 +613,8 @@
encodeULEB128(Imports.size(), getStream());
for (const WasmImport &Import : Imports) {
- StringRef ModuleName = Import.ModuleName;
- encodeULEB128(ModuleName.size(), getStream());
- writeBytes(ModuleName);
-
- StringRef FieldName = Import.FieldName;
- encodeULEB128(FieldName.size(), getStream());
- writeBytes(FieldName);
+ writeString(Import.ModuleName);
+ writeString(Import.FieldName);
encodeULEB128(Import.Kind, getStream());
@@ -701,11 +722,8 @@
encodeULEB128(Exports.size(), getStream());
for (const WasmExport &Export : Exports) {
- encodeULEB128(Export.FieldName.size(), getStream());
- writeBytes(Export.FieldName);
-
+ writeString(Export.FieldName);
encodeSLEB128(Export.Kind, getStream());
-
encodeULEB128(Export.Index, getStream());
}
@@ -750,15 +768,6 @@
MCSectionWasm &FuncSection =
static_cast<MCSectionWasm &>(Func.Sym->getSection());
- if (Func.Sym->isVariable())
- report_fatal_error("weak symbols not supported yet");
-
- if (Func.Sym->getOffset() != 0)
- report_fatal_error("function sections must contain one function each");
-
- if (!Func.Sym->getSize())
- report_fatal_error("function symbols must have a size set with .size");
-
int64_t Size = 0;
if (!Func.Sym->getSize()->evaluateAsAbsolute(Size, Layout))
report_fatal_error(".size expression must be evaluatable");
@@ -819,15 +828,13 @@
for (const WasmImport &Import : Imports) {
if (Import.Kind == wasm::WASM_EXTERNAL_FUNCTION) {
encodeULEB128(Index, getStream());
- encodeULEB128(Import.FieldName.size(), getStream());
- writeBytes(Import.FieldName);
+ writeString(Import.FieldName);
++Index;
}
}
for (const WasmFunction &Func : Functions) {
encodeULEB128(Index, getStream());
- encodeULEB128(Func.Sym->getName().size(), getStream());
- writeBytes(Func.Sym->getName());
+ writeString(Func.Sym->getName());
++Index;
}
@@ -872,22 +879,37 @@
}
void WasmObjectWriter::writeLinkingMetaDataSection(
- bool HasStackPointer, uint32_t StackPointerGlobal) {
- if (!HasStackPointer)
+ ArrayRef<StringRef> WeakSymbols, bool HasStackPointer,
+ uint32_t StackPointerGlobal) {
+ if (!HasStackPointer && WeakSymbols.empty())
return;
+
SectionBookkeeping Section;
startSection(Section, wasm::WASM_SEC_CUSTOM, "linking");
+ SectionBookkeeping SubSection;
- encodeULEB128(1, getStream()); // count
+ if (HasStackPointer) {
+ startSection(SubSection, wasm::WASM_STACK_POINTER);
+ encodeULEB128(StackPointerGlobal, getStream()); // id
+ endSection(SubSection);
+ }
- encodeULEB128(wasm::WASM_STACK_POINTER, getStream()); // type
- encodeULEB128(StackPointerGlobal, getStream()); // id
+ if (WeakSymbols.size() != 0) {
+ startSection(SubSection, wasm::WASM_SYMBOL_INFO);
+ encodeULEB128(WeakSymbols.size(), getStream());
+ for (const StringRef Export: WeakSymbols) {
+ writeString(Export);
+ encodeULEB128(wasm::WASM_SYMBOL_FLAG_WEAK, getStream());
+ }
+ endSection(SubSection);
+ }
endSection(Section);
}
void WasmObjectWriter::writeObject(MCAssembler &Asm,
const MCAsmLayout &Layout) {
+ DEBUG(dbgs() << "WasmObjectWriter::writeObject\n");
MCContext &Ctx = Asm.getContext();
wasm::ValType PtrType = is64Bit() ? wasm::ValType::I64 : wasm::ValType::I32;
@@ -898,6 +920,7 @@
SmallVector<WasmGlobal, 4> Globals;
SmallVector<WasmImport, 4> Imports;
SmallVector<WasmExport, 4> Exports;
+ SmallVector<StringRef, 4> WeakSymbols;
SmallPtrSet<const MCSymbolWasm *, 4> IsAddressTaken;
unsigned NumFuncImports = 0;
unsigned NumGlobalImports = 0;
@@ -906,7 +929,7 @@
bool HasStackPointer = false;
// Populate the IsAddressTaken set.
- for (WasmRelocationEntry RelEntry : CodeRelocations) {
+ for (const WasmRelocationEntry &RelEntry : CodeRelocations) {
switch (RelEntry.Type) {
case wasm::R_WEBASSEMBLY_TABLE_INDEX_SLEB:
case wasm::R_WEBASSEMBLY_GLOBAL_ADDR_SLEB:
@@ -916,7 +939,7 @@
break;
}
}
- for (WasmRelocationEntry RelEntry : DataRelocations) {
+ for (const WasmRelocationEntry &RelEntry : DataRelocations) {
switch (RelEntry.Type) {
case wasm::R_WEBASSEMBLY_TABLE_INDEX_I32:
case wasm::R_WEBASSEMBLY_GLOBAL_ADDR_I32:
@@ -1045,14 +1068,32 @@
StackPointerGlobal = NumGlobalImports + *(const int32_t *)Contents.data();
}
- // Handle defined symbols.
+ // Handle regular defined and undefined symbols.
for (const MCSymbol &S : Asm.symbols()) {
// Ignore unnamed temporary symbols, which aren't ever exported, imported,
// or used in relocations.
if (S.isTemporary() && S.getName().empty())
continue;
+
+ // Variable references (weak references) are handled in a second pass
+ if (S.isVariable())
+ continue;
+
const auto &WS = static_cast<const MCSymbolWasm &>(S);
+ DEBUG(dbgs() << "MCSymbol: '" << S << "'"
+ << " isDefined=" << S.isDefined() << " isExternal="
+ << S.isExternal() << " isTemporary=" << S.isTemporary()
+ << " isFunction=" << WS.isFunction()
+ << " isWeak=" << WS.isWeak()
+ << " isVariable=" << WS.isVariable() << "\n");
+
+ if (WS.isWeak())
+ WeakSymbols.push_back(WS.getName());
+
unsigned Index;
+
+ //<< " function=" << S.isFunction()
+
if (WS.isFunction()) {
// Prepare the function's type, if we haven't seen it yet.
WasmFunctionType F;
@@ -1066,6 +1107,14 @@
int32_t Type = Pair.first->second;
if (WS.isDefined(/*SetUsed=*/false)) {
+ if (WS.getOffset() != 0)
+ report_fatal_error(
+ "function sections must contain one function each");
+
+ if (WS.getSize() == 0)
+ report_fatal_error(
+ "function symbols must have a size set with .size");
+
// A definition. Take the next available index.
Index = NumFuncImports + Functions.size();
@@ -1076,6 +1125,9 @@
SymbolIndices[&WS] = Index;
Functions.push_back(Func);
} else {
+ // Should be no such thing as weak undefined symbol
+ assert(!WS.isVariable());
+
// An import; the index was assigned above.
Index = SymbolIndices.find(&WS)->second;
}
@@ -1089,7 +1141,7 @@
if (WS.isTemporary() && !WS.getSize())
continue;
- if (WS.isDefined(false)) {
+ if (WS.isDefined(/*SetUsed=*/false)) {
if (WS.getOffset() != 0)
report_fatal_error("data sections must contain one variable each: " +
WS.getName());
@@ -1154,21 +1206,46 @@
}
// If the symbol is visible outside this translation unit, export it.
- if (WS.isExternal()) {
- assert(WS.isDefined(false));
+ if (WS.isExternal() && WS.isDefined(/*SetUsed=*/false)) {
WasmExport Export;
Export.FieldName = WS.getName();
Export.Index = Index;
-
if (WS.isFunction())
Export.Kind = wasm::WASM_EXTERNAL_FUNCTION;
else
Export.Kind = wasm::WASM_EXTERNAL_GLOBAL;
-
Exports.push_back(Export);
}
}
+ // Handle weak aliases
+ for (const MCSymbol &S : Asm.symbols()) {
+ if (!S.isVariable())
+ continue;
+ assert(S.isExternal());
+ assert(S.isDefined(/*SetUsed=*/false));
+
+ const auto &WS = static_cast<const MCSymbolWasm &>(S);
+
+ // Find the target symbol of this weak alias
+ const MCExpr *Expr = WS.getVariableValue();
+ auto *Inner = dyn_cast<MCSymbolRefExpr>(Expr);
+ const MCSymbolWasm *ResolvedSym = cast<MCSymbolWasm>(&Inner->getSymbol());
+ uint32_t Index = SymbolIndices.find(ResolvedSym)->second;
+ DEBUG(dbgs() << "Weak alias: '" << WS << "' -> '" << ResolvedSym << "' = " << Index << "\n");
+ SymbolIndices[&WS] = Index;
+
+ WasmExport Export;
+ Export.FieldName = WS.getName();
+ Export.Index = Index;
+ if (WS.isFunction())
+ Export.Kind = wasm::WASM_EXTERNAL_FUNCTION;
+ else
+ Export.Kind = wasm::WASM_EXTERNAL_GLOBAL;
+ WeakSymbols.push_back(Export.FieldName);
+ Exports.push_back(Export);
+ }
+
// Add types for indirect function calls.
for (const WasmRelocationEntry &Fixup : CodeRelocations) {
if (Fixup.Type != wasm::R_WEBASSEMBLY_TYPE_INDEX_LEB)
@@ -1202,7 +1279,7 @@
writeNameSection(Functions, Imports, NumFuncImports);
writeCodeRelocSection();
writeDataRelocSection(DataSectionHeaderSize);
- writeLinkingMetaDataSection(HasStackPointer, StackPointerGlobal);
+ writeLinkingMetaDataSection(WeakSymbols, HasStackPointer, StackPointerGlobal);
// TODO: Translate the .comment section to the output.
// TODO: Translate debug sections to the output.