Reapply "[DebugInfo] Add flag to only emit referenced member functions" (#93767)

This reverts commit 02c6845,
reapplying bfabc95.

The patch was reverted due to the test failing on MacOS and Windows
where type units aren't supported. This is addressed by limiting type
unit flag/test coverage to Linux.

Complete C++ type information can be quite expensive - and there's
limited value in representing every member function, even those that
can't be called (we don't do similarly for every non-member function
anyway). So add a flag to opt out of this behavior for experimenting
with this more terse behavior.

I think Sony already does this by default, so perhaps with a change to
the defaults, Sony can migrate to this rather than a downstream patch.

This breaks current debuggers in some expected ways - but those
breakages are visible without this feature too. Consider member function
template instantiations - they can't be consistently enumerated in every
translation unit:

a.h:
```
struct t1 {
  template <int i>
  static int f1() {
    return i;
  }
};
namespace ns {
template <int i>
int f1() {
  return i;
}
}  // namespace ns
```
a.cpp:
```
void f1() {
  t1::f1<0>();
  ns::f1<0>();
}
```
b.cpp:
```
void f1();
int main() {
  f1();
  t1::f1<1>();
  ns::f1<1>();
}
```
```
(gdb) p ns::f1<0>()
$1 = 0
(gdb) p ns::f1<1>()
$2 = 1
(gdb) p t1::f1<0>()
Couldn't find method t1::f1<0>
(gdb) p t1::f1<1>()
$3 = 1
(gdb) s
f1 () at a.cpp:3
3         t1::f1<0>();
(gdb) p t1::f1<0>()
$4 = 0
(gdb) p t1::f1<1>()
Couldn't find method t1::f1<1>
(gdb)
```

(other similar non-canonical features are implicit special members
(copy/move ctor/assignment operator, default ctor) and nested types (eg:
pimpl idiom, where the nested type is declared-but-not-defined in one
TU, and defined in another TU))

lldb can't parse the template expressions above, so I'm not sure how to
test it there, but I'd guess it has similar problems. (

https://stackoverflow.com/questions/64602475/how-to-print-value-returned-by-template-member-function-in-gdb-lldb-debugging
so... I guess that's just totally not supported in lldb, how
unfortunate. And implicit special members are instantiated implicitly by
lldb, so missing those doesn't tickle the same issue)

Some very rudimentary numbers for a clang debug build:
.debug_info section size:
-g: 476MiB
-g -fdebug-types-section: 357MiB
-g -gomit-unreferenced-members: 340MiB

Though it also means a major reduction in .debug_str size,
-fdebug-types-section doesn't reduce string usage (so the first two
examples have the same .debug_str size, 247MiB), down to 175MiB.

So for total clang binary size (I don't have a quick "debug section size
reduction" on-hand): 1.45 (no type units) GiB -> 1.34 -> 1.22, so it
saves about 120MiB of binary size.

Original Differential Revision: https://reviews.llvm.org/D152017

Author: dwblaikie

clang/include/clang/Basic/DebugOptions.def

@@ -68,6 +68,8 @@ BENIGN_DEBUGOPT(NoInlineLineTables, 1, 0) ///< Whether debug info should contain
                                           ///< inline line tables.
 
 DEBUGOPT(DebugStrictDwarf, 1, 1) ///< Whether or not to use strict DWARF info.
+DEBUGOPT(DebugOmitUnreferencedMethods, 1, 0) ///< Omit unreferenced member
+					     ///< functions in type debug info.
 
 /// Control the Assignment Tracking debug info feature.
 BENIGN_ENUM_DEBUGOPT(AssignmentTrackingMode, AssignmentTrackingOpts, 2,

clang/include/clang/Driver/Options.td

@@ -4345,6 +4345,10 @@ defm strict_dwarf : BoolOption<"g", "strict-dwarf",
           "the specified version, avoiding features from later versions.">,
   NegFlag<SetFalse>, BothFlags<[], [ClangOption, CLOption, DXCOption]>>,
   Group<g_flags_Group>;
+defm omit_unreferenced_methods : BoolGOption<"omit-unreferenced-methods",
+  CodeGenOpts<"DebugOmitUnreferencedMethods">, DefaultFalse,
+  NegFlag<SetFalse>,
+  PosFlag<SetTrue, [], [CC1Option]>, BothFlags<[], [ClangOption, CLOption, DXCOption]>>;
 defm column_info : BoolOption<"g", "column-info",
   CodeGenOpts<"DebugColumnInfo">, DefaultTrue,
   NegFlag<SetFalse, [], [ClangOption, CC1Option]>,

clang/lib/CodeGen/CGDebugInfo.cpp

@@ -2836,7 +2836,7 @@ CGDebugInfo::CreateTypeDefinition(const RecordType *Ty) {
 
   // Collect data fields (including static variables and any initializers).
   CollectRecordFields(RD, DefUnit, EltTys, FwdDecl);
-  if (CXXDecl)
+  if (CXXDecl && !CGM.getCodeGenOpts().DebugOmitUnreferencedMethods)
     CollectCXXMemberFunctions(CXXDecl, DefUnit, EltTys, FwdDecl);
 
   LexicalBlockStack.pop_back();

clang/lib/Driver/ToolChains/Clang.cpp

@@ -45,6 +45,7 @@
 #include "llvm/ADT/StringExtras.h"
 #include "llvm/BinaryFormat/Magic.h"
 #include "llvm/Config/llvm-config.h"
+#include "llvm/Frontend/Debug/Options.h"
 #include "llvm/Object/ObjectFile.h"
 #include "llvm/Option/ArgList.h"
 #include "llvm/Support/CodeGen.h"
@@ -4642,6 +4643,7 @@ renderDebugOptions(const ToolChain &TC, const Driver &D, const llvm::Triple &T,
   Args.addOptInFlag(CmdArgs, options::OPT_fforce_dwarf_frame,
                     options::OPT_fno_force_dwarf_frame);
 
+  bool EnableTypeUnits = false;
   if (Args.hasFlag(options::OPT_fdebug_types_section,
                    options::OPT_fno_debug_types_section, false)) {
     if (!(T.isOSBinFormatELF() || T.isOSBinFormatWasm())) {
@@ -4652,11 +4654,24 @@ renderDebugOptions(const ToolChain &TC, const Driver &D, const llvm::Triple &T,
     } else if (checkDebugInfoOption(
                    Args.getLastArg(options::OPT_fdebug_types_section), Args, D,
                    TC)) {
+      EnableTypeUnits = true;
       CmdArgs.push_back("-mllvm");
       CmdArgs.push_back("-generate-type-units");
     }
   }
 
+  if (const Arg *A =
+          Args.getLastArg(options::OPT_gomit_unreferenced_methods,
+                          options::OPT_gno_omit_unreferenced_methods))
+    (void)checkDebugInfoOption(A, Args, D, TC);
+  if (Args.hasFlag(options::OPT_gomit_unreferenced_methods,
+                   options::OPT_gno_omit_unreferenced_methods, false) &&
+      (DebugInfoKind == llvm::codegenoptions::DebugInfoConstructor ||
+       DebugInfoKind == llvm::codegenoptions::LimitedDebugInfo) &&
+      !EnableTypeUnits) {
+    CmdArgs.push_back("-gomit-unreferenced-methods");
+  }
+
   // To avoid join/split of directory+filename, the integrated assembler prefers
   // the directory form of .file on all DWARF versions. GNU as doesn't allow the
   // form before DWARF v5.

clang/test/CodeGenCXX/debug-info-incomplete-types.cpp

@@ -0,0 +1,12 @@
+// RUN: %clang_cc1 -debug-info-kind=limited -gomit-unreferenced-methods %s -emit-llvm -o - | FileCheck %s
+
+struct t1 {
+  void f1();
+  void f2();
+};
+
+void t1::f1() { }
+
+// CHECK: distinct !DICompositeType(tag: DW_TAG_structure_type, name: "t1"
+// CHECK-SAME: elements: [[ELEMENTS:![0-9]+]]
+// CHECK: [[ELEMENTS]] = !{}

clang/test/Driver/debug-options.c

@@ -242,6 +242,12 @@
 // RUN: %clang -### -c %s 2>&1 | FileCheck -check-prefix=NORNGBSE %s
 // RUN: %clang -### -c -fdebug-ranges-base-address -fno-debug-ranges-base-address %s 2>&1 | FileCheck -check-prefix=NORNGBSE %s
 //
+// RUN: %clang -### -c -gomit-unreferenced-methods %s 2>&1 | FileCheck -check-prefix=INCTYPES %s
+// RUN: %clang -### -c %s 2>&1 | FileCheck -check-prefix=NOINCTYPES %s
+// RUN: %clang -### -c -gomit-unreferenced-methods -fdebug-types-section -target x86_64-unknown-linux %s 2>&1 \
+// RUN:        | FileCheck -check-prefix=NOINCTYPES %s
+// RUN: %clang -### -c -gomit-unreferenced-methods -fstandalone-debug %s 2>&1 | FileCheck -check-prefix=NOINCTYPES %s
+//
 // RUN: %clang -### -c -glldb %s 2>&1 | FileCheck -check-prefix=NOPUB %s
 // RUN: %clang -### -c -glldb -gno-pubnames %s 2>&1 | FileCheck -check-prefix=NOPUB %s
 //
@@ -381,6 +387,9 @@
 // RNGBSE: -fdebug-ranges-base-address
 // NORNGBSE-NOT: -fdebug-ranges-base-address
 //
+// INCTYPES: -gomit-unreferenced-methods
+// NOINCTYPES-NOT: -gomit-unreferenced-methods
+//
 // GARANGE-DAG: -generate-arange-section
 //
 // FDTS: "-mllvm" "-generate-type-units"