AsmPrinter: Emit the DwarfStringPool offset directly when possible

Change `DwarfStringPool` to calculate byte offsets on-the-fly, and
update `DwarfUnit::getLocalString()` to use a `DIEInteger` instead of a
`DIEDelta` when Dwarf doesn't use relocations (i.e., Mach-O).  This
eliminates another call to `EmitLabelDifference()`, and drops memory
usage from 865 MB down to 861 MB, around 0.5%.

(I'm looking at `llc` memory usage on `verify-uselistorder.lto.opt.bc`;
see r236629 for details.)

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@238114 91177308-0d34-0410-b5e6-96231b3b80d8

lib/CodeGen/AsmPrinter/DwarfStringPool.cpp

@@ -12,12 +12,16 @@
 
 using namespace llvm;
 
-std::pair<MCSymbol *, unsigned> &DwarfStringPool::getEntry(AsmPrinter &Asm,
-                                                           StringRef Str) {
-  std::pair<MCSymbol *, unsigned> &Entry = Pool[Str];
-  if (!Entry.first) {
-    Entry.second = Pool.size() - 1;
-    Entry.first = Asm.createTempSymbol(Prefix);
+DwarfStringPool::EntryTy &DwarfStringPool::getEntry(AsmPrinter &Asm,
+                                                    StringRef Str) {
+  auto &Entry = Pool[Str];
+  if (!Entry.Symbol) {
+    Entry.Index = Pool.size() - 1;
+    Entry.Offset = NumBytes;
+    Entry.Symbol = Asm.createTempSymbol(Prefix);
+
+    NumBytes += Str.size() + 1;
+    assert(NumBytes > Entry.Offset && "Unexpected overflow");
   }
   return Entry;
 }
@@ -32,17 +36,18 @@ void DwarfStringPool::emit(AsmPrinter &Asm, MCSection *StrSection,
 
   // Get all of the string pool entries and put them in an array by their ID so
   // we can sort them.
-  SmallVector<const StringMapEntry<std::pair<MCSymbol *, unsigned>> *, 64>
-  Entries(Pool.size());
+  SmallVector<const StringMapEntry<EntryTy> *, 64> Entries(Pool.size());
 
   for (const auto &E : Pool)
-    Entries[E.getValue().second] = &E;
+    Entries[E.getValue().Index] = &E;
 
   for (const auto &Entry : Entries) {
     // Emit a label for reference from debug information entries.
-    Asm.OutStreamer->EmitLabel(Entry->getValue().first);
+    Asm.OutStreamer->EmitLabel(Entry->getValue().Symbol);
 
     // Emit the string itself with a terminating null byte.
+    Asm.OutStreamer->AddComment("string offset=" +
+                                Twine(Entry->getValue().Offset));
     Asm.OutStreamer->EmitBytes(
         StringRef(Entry->getKeyData(), Entry->getKeyLength() + 1));
   }
@@ -50,11 +55,8 @@ void DwarfStringPool::emit(AsmPrinter &Asm, MCSection *StrSection,
   // If we've got an offset section go ahead and emit that now as well.
   if (OffsetSection) {
     Asm.OutStreamer->SwitchSection(OffsetSection);
-    unsigned offset = 0;
     unsigned size = 4; // FIXME: DWARF64 is 8.
-    for (const auto &Entry : Entries) {
-      Asm.OutStreamer->EmitIntValue(offset, size);
-      offset += Entry->getKeyLength() + 1;
-    }
+    for (const auto &Entry : Entries)
+      Asm.OutStreamer->EmitIntValue(Entry->getValue().Offset, size);
   }
 }

lib/CodeGen/AsmPrinter/DwarfStringPool.h

@@ -25,8 +25,14 @@ class StringRef;
 // A String->Symbol mapping of strings used by indirect
 // references.
 class DwarfStringPool {
-  StringMap<std::pair<MCSymbol *, unsigned>, BumpPtrAllocator &> Pool;
+  struct EntryTy {
+    MCSymbol *Symbol;
+    unsigned Offset;
+    unsigned Index;
+  };
+  StringMap<EntryTy, BumpPtrAllocator &> Pool;
   StringRef Prefix;
+  unsigned NumBytes = 0;
 
 public:
   DwarfStringPool(BumpPtrAllocator &A, AsmPrinter &Asm, StringRef Prefix)
@@ -38,19 +44,24 @@ public:
   /// \brief Returns an entry into the string pool with the given
   /// string text.
   MCSymbol *getSymbol(AsmPrinter &Asm, StringRef Str) {
-    return getEntry(Asm, Str).first;
+    return getEntry(Asm, Str).Symbol;
+  }
+
+  /// Get a byte offset into the string pool with the given text.
+  unsigned getOffset(AsmPrinter &Asm, StringRef Str) {
+    return getEntry(Asm, Str).Offset;
   }
 
   /// \brief Returns the index into the string pool with the given
   /// string text.
   unsigned getIndex(AsmPrinter &Asm, StringRef Str) {
-    return getEntry(Asm, Str).second;
+    return getEntry(Asm, Str).Index;
   }
 
   bool empty() const { return Pool.empty(); }
 
 private:
-  std::pair<MCSymbol *, unsigned> &getEntry(AsmPrinter &Asm, StringRef Str);
+  EntryTy &getEntry(AsmPrinter &Asm, StringRef Str);
 };
 }
 #endif

lib/CodeGen/AsmPrinter/DwarfUnit.cpp

@@ -239,14 +239,13 @@ void DwarfUnit::addIndexedString(DIE &Die, dwarf::Attribute Attribute,
 
 void DwarfUnit::addLocalString(DIE &Die, dwarf::Attribute Attribute,
                                StringRef String) {
-  const TargetLoweringObjectFile &TLOF = Asm->getObjFileLowering();
-  MCSymbol *Symb = DU->getStringPool().getSymbol(*Asm, String);
   DIEValue *Value;
   if (Asm->MAI->doesDwarfUseRelocationsAcrossSections())
-    Value = new (DIEValueAllocator) DIELabel(Symb);
+    Value = new (DIEValueAllocator)
+        DIELabel(DU->getStringPool().getSymbol(*Asm, String));
   else
     Value = new (DIEValueAllocator)
-        DIEDelta(Symb, TLOF.getDwarfStrSection()->getBeginSymbol());
+        DIEInteger(DU->getStringPool().getOffset(*Asm, String));
   DIEValue *Str = new (DIEValueAllocator) DIEString(Value, String);
   Die.addValue(Attribute, dwarf::DW_FORM_strp, Str);
 }

test/DebugInfo/2010-06-29-InlinedFnLocalVar.ll

@@ -3,7 +3,7 @@
 
 ; CHECK: section_info
 ; CHECK:	DW_TAG_structure_type
-; CHECK-NEXT:	info_string
+; CHECK-NEXT:	DW_AT_name
 
 
 @i = common global i32 0                          ; <i32*> [#uses=2]

test/DebugInfo/X86/stringpool.ll

@@ -18,7 +18,7 @@
 ; LINUX: .section .debug_str,"MS",@progbits,1
 ; LINUX: yyyy
 ; DARWIN: .section __DWARF,__debug_str,regular,debug
-; DARWIN: yyyy
+; DARWIN: .asciz "yyyy" ## string offset=[[YYYY:[0-9]+]]
 
 ; Verify that we refer to 'yyyy' with a relocation.
 ; LINUX:      .long   .Linfo_string3          # DW_AT_name
@@ -30,9 +30,8 @@
 ; LINUX-NEXT: .byte   3
 ; LINUX-NEXT: .quad   yyyy
 
-; Verify that we refer to 'yyyy' without a relocation.
-; DARWIN: Lset[[ID:[0-9]+]] = Linfo_string3-Linfo_string ## DW_AT_name
-; DARWIN-NEXT:        .long   Lset[[ID]]
+; Verify that we refer to 'yyyy' with a direct offset.
+; DARWIN:             .long   [[YYYY]]
 ; DARWIN-NEXT:        .long   {{[0-9]+}}              ## DW_AT_type
 ; DARWIN-NEXT:                                        ## DW_AT_external
 ; DARWIN-NEXT:        .byte   1                       ## DW_AT_decl_file