Move individual group name strings from the OptionTable into one big char array. Then only store offsets into it in the OptionTable. Saves about 4K from the clang binary and removes 400 relocation entries from DiagnosticIDs.o.

git-svn-id: https://llvm.org/svn/llvm-project/cfe/trunk@189568 91177308-0d34-0410-b5e6-96231b3b80d8

5 files changed, 42 insertions, 33 deletions

diff --git a/lib/Basic/DiagnosticIDs.cpp b/lib/Basic/DiagnosticIDs.cpp
index cccf418f65..71752bb38f 100644
--- a/lib/Basic/DiagnosticIDs.cpp
+++ b/lib/Basic/DiagnosticIDs.cpp
@@ -501,21 +501,22 @@ DiagnosticIDs::getDiagnosticLevel(unsigned DiagID, unsigned DiagClass,
   return Result;
 }
 
+#define GET_DIAG_ARRAYS
+#include "clang/Basic/DiagnosticGroups.inc"
+#undef GET_DIAG_ARRAYS
+
 struct clang::WarningOption {
-  const char *NameStr;
-  uint16_t NameLen;
+  uint16_t NameOffset;
   uint16_t Members;
   uint16_t SubGroups;
 
+  // String is stored with a pascal-style length byte.
   StringRef getName() const {
-    return StringRef(NameStr, NameLen);
+    return StringRef(DiagGroupNames + NameOffset + 1,
+                     DiagGroupNames[NameOffset]);
   }
 };
 
-#define GET_DIAG_ARRAYS
-#include "clang/Basic/DiagnosticGroups.inc"
-#undef GET_DIAG_ARRAYS
-
 // Second the table of options, sorted by name for fast binary lookup.
 static const WarningOption OptionTable[] = {
 #define GET_DIAG_TABLE
@@ -524,9 +525,8 @@ static const WarningOption OptionTable[] = {
 };
 static const size_t OptionTableSize = llvm::array_lengthof(OptionTable);
 
-static bool WarningOptionCompare(const WarningOption &LHS,
-                                 const WarningOption &RHS) {
-  return LHS.getName() < RHS.getName();
+static bool WarningOptionCompare(const WarningOption &LHS, StringRef RHS) {
+  return LHS.getName() < RHS;
 }
 
 /// getWarningOptionForDiag - Return the lowest-level warning option that
@@ -555,11 +555,8 @@ void DiagnosticIDs::getDiagnosticsInGroup(
 bool DiagnosticIDs::getDiagnosticsInGroup(
     StringRef Group,
     SmallVectorImpl<diag::kind> &Diags) const {
-  if (Group.size() > UINT16_MAX)
-    return true; // Option is too long to be one in the table.
-  WarningOption Key = { Group.data(), (uint16_t)Group.size(), 0, 0 };
   const WarningOption *Found =
-  std::lower_bound(OptionTable, OptionTable + OptionTableSize, Key,
+  std::lower_bound(OptionTable, OptionTable + OptionTableSize, Group,
                    WarningOptionCompare);
   if (Found == OptionTable + OptionTableSize ||
       Found->getName() != Group)
diff --git a/tools/diagtool/DiagnosticNames.cpp b/tools/diagtool/DiagnosticNames.cpp
index fa510eb6ec..34c3229c6e 100644
--- a/tools/diagtool/DiagnosticNames.cpp
+++ b/tools/diagtool/DiagnosticNames.cpp
@@ -73,6 +73,10 @@ static const GroupRecord OptionTable[] = {
 #undef GET_DIAG_TABLE
 };
 
+llvm::StringRef GroupRecord::getName() const {
+  return StringRef(DiagGroupNames + NameOffset + 1, DiagGroupNames[NameOffset]);
+}
+
 GroupRecord::subgroup_iterator GroupRecord::subgroup_begin() const {
   return DiagSubGroups + SubGroups;
 }
diff --git a/tools/diagtool/DiagnosticNames.h b/tools/diagtool/DiagnosticNames.h
index 957a1818e8..a3321fa681 100644
--- a/tools/diagtool/DiagnosticNames.h
+++ b/tools/diagtool/DiagnosticNames.h
@@ -35,14 +35,11 @@ namespace diagtool {
 
 
   struct GroupRecord {
-    const char *NameStr;
-    uint16_t NameLen;
+    uint16_t NameOffset;
     uint16_t Members;
     uint16_t SubGroups;
 
-    llvm::StringRef getName() const {
-      return llvm::StringRef(NameStr, NameLen);
-    }
+    llvm::StringRef getName() const;
 
     template<typename RecordType>
     class group_iterator {
@@ -94,8 +91,8 @@ namespace diagtool {
     diagnostics_iterator diagnostics_begin() const;
     diagnostics_iterator diagnostics_end() const;
 
-    bool operator<(const GroupRecord &Other) const {
-      return getName() < Other.getName();
+    bool operator<(llvm::StringRef Other) const {
+      return getName() < Other;
     }
   };
 
diff --git a/tools/diagtool/TreeView.cpp b/tools/diagtool/TreeView.cpp
index 10809c1d8d..fd548ef011 100644
--- a/tools/diagtool/TreeView.cpp
+++ b/tools/diagtool/TreeView.cpp
@@ -99,9 +99,8 @@ static int showGroup(llvm::raw_ostream &out, StringRef RootGroup,
     return 1;
   }
 
-  GroupRecord Key = { RootGroup.data(), (uint16_t)RootGroup.size(), 0, 0 };
   const GroupRecord *Found =
-    std::lower_bound(AllGroups.begin(), AllGroups.end(), Key);
+    std::lower_bound(AllGroups.begin(), AllGroups.end(), RootGroup);
   
   if (Found == AllGroups.end() || Found->getName() != RootGroup) {
     llvm::errs() << "No such diagnostic group exists\n";
diff --git a/utils/TableGen/ClangDiagnosticsEmitter.cpp b/utils/TableGen/ClangDiagnosticsEmitter.cpp
index 9a1cd345a7..d1bd58e1aa 100644
--- a/utils/TableGen/ClangDiagnosticsEmitter.cpp
+++ b/utils/TableGen/ClangDiagnosticsEmitter.cpp
@@ -24,6 +24,7 @@
 #include "llvm/Support/Debug.h"
 #include "llvm/TableGen/Error.h"
 #include "llvm/TableGen/Record.h"
+#include "llvm/TableGen/StringToOffsetTable.h"
 #include "llvm/TableGen/TableGenBackend.h"
 #include <algorithm>
 #include <cctype>
@@ -671,7 +672,20 @@ void EmitClangDiagGroups(RecordKeeper &Records, raw_ostream &OS) {
       OS << "-1,\n";
     }
   }
-  OS << "};\n";
+  OS << "};\n\n";
+
+  StringToOffsetTable GroupNames;
+  for (std::map<std::string, GroupInfo>::const_iterator
+         I = DiagsInGroup.begin(), E = DiagsInGroup.end(); I != E; ++I) {
+    // Store a pascal-style length byte at the beginning of the string.
+    std::string Name = char(I->first.size()) + I->first;
+    GroupNames.GetOrAddStringOffset(Name, false);
+  }
+
+  OS << "static const char DiagGroupNames[] = {\n";
+  GroupNames.EmitString(OS);
+  OS << "};\n\n";
+
   OS << "#endif // GET_DIAG_ARRAYS\n\n";
 
   // Emit the table now.
@@ -680,18 +694,16 @@ void EmitClangDiagGroups(RecordKeeper &Records, raw_ostream &OS) {
   for (std::map<std::string, GroupInfo>::const_iterator
        I = DiagsInGroup.begin(), E = DiagsInGroup.end(); I != E; ++I) {
     // Group option string.
-    OS << "  { ";
-    OS << "\"";
+    OS << "  { /* ";
     if (I->first.find_first_not_of("abcdefghijklmnopqrstuvwxyz"
                                    "ABCDEFGHIJKLMNOPQRSTUVWXYZ"
                                    "0123456789!@#$%^*-+=:?")!=std::string::npos)
       PrintFatalError("Invalid character in diagnostic group '" +
                       I->first + "'");
-    OS.write_escaped(I->first) << "\","
-                               << std::string(MaxLen-I->first.size()+1, ' ');
-    if (I->first.size() > UINT16_MAX)
-      PrintFatalError("Diagnostic group name is too long for NameLen field.");
-    OS << I->first.size() << ", ";
+    OS << I->first << " */ " << std::string(MaxLen-I->first.size(), ' ');
+    // Store a pascal-style length byte at the beginning of the string.
+    std::string Name = char(I->first.size()) + I->first;
+    OS << GroupNames.GetOrAddStringOffset(Name, false) << ", ";
 
     // Special handling for 'pedantic'.
     const bool IsPedantic = I->first == "pedantic";
@@ -707,7 +719,7 @@ void EmitClangDiagGroups(RecordKeeper &Records, raw_ostream &OS) {
         DiagArrayIndex += DiagsInPedantic.size();
       DiagArrayIndex += V.size() + 1;
     } else {
-      OS << "/* Empty */ 0, ";
+      OS << "/* Empty */     0, ";
     }
 
     // Subgroups.
@@ -720,7 +732,7 @@ void EmitClangDiagGroups(RecordKeeper &Records, raw_ostream &OS) {
         SubGroupIndex += GroupsInPedantic.size();
       SubGroupIndex += SubGroups.size() + 1;
     } else {
-      OS << "/* Empty */ 0";
+      OS << "/* Empty */         0";
     }
     OS << " },\n";
   }