[WebAssembly] Add support for data sections in the assembler.

Summary:
This is quite minimal so far, introduce them with .section,
fill them with .int8 or .asciz, end with .size

Reviewers: dschuff, sbc100, aheejin

Subscribers: jgravelle-google, sunfish, llvm-commits

Tags: #llvm

Differential Revision: https://reviews.llvm.org/D58660

llvm-svn: 355321

3 files changed, 57 insertions, 10 deletions

diff --git a/llvm/lib/MC/MCParser/WasmAsmParser.cpp b/llvm/lib/MC/MCParser/WasmAsmParser.cpp
index a8a48d1cd69..197e9052566 100644
--- a/llvm/lib/MC/MCParser/WasmAsmParser.cpp
+++ b/llvm/lib/MC/MCParser/WasmAsmParser.cpp
@@ -130,11 +130,27 @@ public:
     if (expect(AsmToken::Comma, ",") || expect(AsmToken::At, "@") ||
         expect(AsmToken::EndOfStatement, "eol"))
       return true;
-    // This is done automatically by the assembler for text sections currently,
-    // so we don't need to emit that here. This is what it would do (and may
-    // be needed later for other section types):
-    // auto WS = getContext().getWasmSection(Name, SectionKind::getText());
-    // getStreamer().SwitchSection(WS);
+    struct SectionType {
+      const char *Name;
+      SectionKind Kind;
+    };
+    static SectionType SectionTypes[] = {
+        { ".text", SectionKind::getText() },
+        { ".rodata", SectionKind::getReadOnly() },
+        { ".data", SectionKind::getData() },
+        // TODO: add more types.
+    };
+    for (size_t I = 0; I < sizeof(SectionTypes) / sizeof(SectionType); I++) {
+      if (Name.startswith(SectionTypes[I].Name)) {
+        auto WS = getContext().getWasmSection(Name, SectionTypes[I].Kind);
+        getStreamer().SwitchSection(WS);
+        return false;
+      }
+    }
+    // Not found, just ignore this section.
+    // For code in a text section WebAssemblyAsmParser automatically adds
+    // one section per function, so they're optional to be specified with
+    // this directive.
     return false;
   }
 
@@ -153,9 +169,8 @@ public:
     if (expect(AsmToken::EndOfStatement, "eol"))
       return true;
     // This is done automatically by the assembler for functions currently,
-    // so we don't need to emit that here. This is what it would do:
-    (void)Sym;
-    // getStreamer().emitELFSize(Sym, Expr);
+    // so this is only currently needed for data sections:
+    getStreamer().emitELFSize(Sym, Expr);
     return false;
   }
 
diff --git a/llvm/lib/MC/WasmObjectWriter.cpp b/llvm/lib/MC/WasmObjectWriter.cpp
index b985d4bf4d3..e82ed5898b4 100644
--- a/llvm/lib/MC/WasmObjectWriter.cpp
+++ b/llvm/lib/MC/WasmObjectWriter.cpp
@@ -1348,7 +1348,7 @@ uint64_t WasmObjectWriter::writeObject(MCAssembler &Asm,
       LLVM_DEBUG(dbgs() << "  -> function index: " << Index << "\n");
 
     } else if (WS.isData()) {
-      if (WS.isTemporary() && !WS.getSize())
+      if (!isInSymtab(WS))
         continue;
 
       if (!WS.isDefined()) {
diff --git a/llvm/lib/Target/WebAssembly/AsmParser/WebAssemblyAsmParser.cpp b/llvm/lib/Target/WebAssembly/AsmParser/WebAssemblyAsmParser.cpp
index eeccc3e6449..9d5bb3f11e0 100644
--- a/llvm/lib/Target/WebAssembly/AsmParser/WebAssemblyAsmParser.cpp
+++ b/llvm/lib/Target/WebAssembly/AsmParser/WebAssemblyAsmParser.cpp
@@ -172,6 +172,7 @@ class WebAssemblyAsmParser final : public MCTargetAsmParser {
     FunctionLocals,
     Instructions,
     EndFunction,
+    DataSection,
   } CurrentState = FileStart;
 
   // For ensuring blocks are properly nested.
@@ -552,6 +553,17 @@ public:
     return false;
   }
 
+  bool CheckDataSection() {
+    if (CurrentState != DataSection) {
+      auto WS = cast<MCSectionWasm>(getStreamer().getCurrentSection().first);
+      if (WS && WS->getKind().isText())
+        return error("data directive must occur in a data segment: ",
+                     Lexer.getTok());
+    }
+    CurrentState = DataSection;
+    return false;
+  }
+
   // This function processes wasm-specific directives streamed to
   // WebAssemblyTargetStreamer, all others go to the generic parser
   // (see WasmAsmParser).
@@ -650,6 +662,25 @@ public:
       return expect(AsmToken::EndOfStatement, "EOL");
     }
 
+    if (DirectiveID.getString() == ".int8") {
+      if (CheckDataSection()) return true;
+      int64_t V;
+      if (Parser.parseAbsoluteExpression(V))
+        return error("Cannot parse int8 constant: ", Lexer.getTok());
+      // TODO: error if value doesn't fit?
+      Out.EmitIntValue(static_cast<uint64_t>(V), 1);
+      return expect(AsmToken::EndOfStatement, "EOL");
+    }
+
+    if (DirectiveID.getString() == ".asciz") {
+      if (CheckDataSection()) return true;
+      std::string S;
+      if (Parser.parseEscapedString(S))
+        return error("Cannot parse string constant: ", Lexer.getTok());
+      Out.EmitBytes(StringRef(S.c_str(), S.length() + 1));
+      return expect(AsmToken::EndOfStatement, "EOL");
+    }
+
     return true; // We didn't process this directive.
   }
 
@@ -717,9 +748,10 @@ public:
   void onEndOfFunction() {
     // Automatically output a .size directive, so it becomes optional for the
     // user.
+    if (!LastFunctionLabel) return;
     auto TempSym = getContext().createLinkerPrivateTempSymbol();
     getStreamer().EmitLabel(TempSym);
-    auto Start = MCSymbolRefExpr::create(LastLabel, getContext());
+    auto Start = MCSymbolRefExpr::create(LastFunctionLabel, getContext());
     auto End = MCSymbolRefExpr::create(TempSym, getContext());
     auto Expr =
         MCBinaryExpr::create(MCBinaryExpr::Sub, End, Start, getContext());