[mach-o] Make anonymous atom out of section content before any symbol

In sections that are broken into atoms at symbols, if the first symbol in the
section is not at the start of the section, then make an anonymous atom for
the section content that is before the first symbol.

llvm-svn: 210142

2 files changed, 35 insertions, 22 deletions

diff --git a/lld/lib/ReaderWriter/MachO/MachONormalizedFileToAtoms.cpp b/lld/lib/ReaderWriter/MachO/MachONormalizedFileToAtoms.cpp
index 9be4d0ec410..e5ce190f924 100644
--- a/lld/lib/ReaderWriter/MachO/MachONormalizedFileToAtoms.cpp
+++ b/lld/lib/ReaderWriter/MachO/MachONormalizedFileToAtoms.cpp
@@ -209,31 +209,29 @@ void appendSymbolsInSection(const std::vector<Symbol> &inSymbols,
 }
 
 void atomFromSymbol(DefinedAtom::ContentType atomType, const Section &section,
-                    MachOFile &file, const Symbol &symbol,
+                    MachOFile &file, uint64_t symbolAddr, StringRef symbolName,
+                    bool symbolWeakDef, Atom::Scope symbolScope,
                     uint64_t nextSymbolAddr, bool copyRefs) {
   // Mach-O symbol table does have size in it. Instead the size is the
   // difference between this and the next symbol.
-  uint64_t size = nextSymbolAddr - symbol.value;
+  uint64_t size = nextSymbolAddr - symbolAddr;
   if (section.type == llvm::MachO::S_ZEROFILL) {
-    file.addZeroFillDefinedAtom(symbol.name, atomScope(symbol.scope),
-                                size, copyRefs);
+    file.addZeroFillDefinedAtom(symbolName, symbolScope, size, copyRefs);
   } else {
-    uint64_t offset = symbol.value - section.address;
+    uint64_t offset = symbolAddr - section.address;
     ArrayRef<uint8_t> atomContent = section.content.slice(offset, size);
-    DefinedAtom::Merge merge = DefinedAtom::mergeNo;
-    if (symbol.desc & N_WEAK_DEF)
-      merge = DefinedAtom::mergeAsWeak;
+    DefinedAtom::Merge merge = symbolWeakDef 
+                              ? DefinedAtom::mergeAsWeak : DefinedAtom::mergeNo;
     if (atomType == DefinedAtom::typeUnknown) {
       // Mach-O needs a segment and section name.  Concatentate those two
       // with a / seperator (e.g. "seg/sect") to fit into the lld model
       // of just a section name.
       std::string segSectName = section.segmentName.str()
                                 + "/" + section.sectionName.str();
-      file.addDefinedAtomInCustomSection(symbol.name, atomScope(symbol.scope),
-                                         atomType, merge, atomContent,
-                                         segSectName, true);
+      file.addDefinedAtomInCustomSection(symbolName, symbolScope, atomType,
+                                         merge, atomContent, segSectName, true);
     } else {
-      file.addDefinedAtom(symbol.name, atomScope(symbol.scope), atomType, merge,
+      file.addDefinedAtom(symbolName, symbolScope, atomType, merge,
                           atomContent, copyRefs);
     }
   }
@@ -270,20 +268,26 @@ error_code processSymboledSection(DefinedAtom::ContentType atomType,
   if (symbols.empty() && section.content.empty())
     return error_code();
 
-  if (symbols.front()->value != section.address) {
+  const uint64_t firstSymbolAddr = symbols.front()->value;
+  if (firstSymbolAddr != section.address) {
     // Section has anonymous content before first symbol.
-    // FIXME
+    atomFromSymbol(atomType, section, file, section.address, StringRef(),
+                  false, Atom::scopeTranslationUnit, firstSymbolAddr, copyRefs);
   }
 
   const Symbol *lastSym = nullptr;
+  bool lastSymIsWeakDef;
   for (const Symbol *sym : symbols) {
     if (lastSym != nullptr) {
-      atomFromSymbol(atomType, section, file, *lastSym, sym->value, copyRefs);
+      atomFromSymbol(atomType, section, file, lastSym->value, lastSym->name,
+                     lastSymIsWeakDef, atomScope(lastSym->scope), sym->value, copyRefs);
     }
     lastSym = sym;
+    lastSymIsWeakDef = (lastSym->desc & N_WEAK_DEF);
   }
   if (lastSym != nullptr) {
-    atomFromSymbol(atomType, section, file, *lastSym,
+    atomFromSymbol(atomType, section, file, lastSym->value, lastSym->name,
+                   lastSymIsWeakDef, atomScope(lastSym->scope),
                    section.address + section.content.size(), copyRefs);
   }
   return error_code();
diff --git a/lld/test/mach-o/parse-function.yaml b/lld/test/mach-o/parse-function.yaml
index 5eb721da386..19a935adb55 100644
--- a/lld/test/mach-o/parse-function.yaml
+++ b/lld/test/mach-o/parse-function.yaml
@@ -16,51 +16,60 @@ sections:
     attributes:      [ S_ATTR_PURE_INSTRUCTIONS, S_ATTR_SOME_INSTRUCTIONS ]
     alignment:       4
     address:         0x0000000000000000
-    content:         [ 0xC3, 0xC3, 0xC3, 0xC3, 0xC3 ]
+    content:         [ 0xCC, 0xC3, 0x90, 0xC3, 0x90, 0x90, 0xC3, 0x90, 
+                       0x90, 0x90, 0xC3, 0x90, 0x90, 0x90, 0x90, 0xC3 ]
 local-symbols:
   - name:            _myStatic
     type:            N_SECT
     sect:            1
-    value:           0x0000000000000004
+    value:           0x000000000000000B
 global-symbols:
   - name:            _myGlobal
     type:            N_SECT
     scope:           [ N_EXT ]
     sect:            1
-    value:           0x0000000000000000
+    value:           0x0000000000000001
   - name:            _myGlobalWeak
     type:            N_SECT
     scope:           [ N_EXT ]
     sect:            1
     desc:            [ N_WEAK_DEF ]
-    value:           0x0000000000000001
+    value:           0x0000000000000002
   - name:            _myHidden
     type:            N_SECT
     scope:           [ N_EXT, N_PEXT ]
     sect:            1
-    value:           0x0000000000000002
+    value:           0x0000000000000004
   - name:            _myHiddenWeak
     type:            N_SECT
     scope:           [ N_EXT, N_PEXT ]
     sect:            1
     desc:            [ N_WEAK_DEF ]
-    value:           0x0000000000000003
+    value:           0x0000000000000007
 ...
 
+# CHECK-NOT:  name:  
+# CHECK:      content:         [ CC ]
+
 # CHECK:      name:   _myGlobal
 # CHECK:      scope:  global
+# CHECK:      content:         [ C3 ]
 
 # CHECK:      name:   _myGlobalWeak
 # CHECK:      scope:  global
+# CHECK:      content:         [ 90, C3 ]
 # CHECK:      merge:  as-weak
 
 # CHECK:      name:   _myHidden
 # CHECK:      scope:  hidden
+# CHECK:      content:         [ 90, 90, C3 ]
 
 # CHECK:      name:   _myHiddenWeak
 # CHECK:      scope:  hidden
+# CHECK:      content:         [ 90, 90, 90, C3 ]
 # CHECK:      merge:  as-weak
 
 # CHECK:      name:   _myStatic
 # CHECK-NOT:	scope:  global
 # CHECK-NOT:	scope:  hidden
+# CHECK:      content:         [ 90, 90, 90, 90, C3 ]