TBDGen: teach the compiler to take a json file indicating previous install names

Using the new linker directives $ld$previous requires the compiler to know the previous
install names for the symbols marked as removed. This patch teaches the compiler
to take a path to a Json file specifying the map between module names and previous
install names. Also, these install names can be platform-specific.

Progress towards: rdar://58281536

Author: nkcsgexi

include/swift/AST/DiagnosticsFrontend.def

@@ -260,6 +260,28 @@ WARNING(tbd_only_supported_in_whole_module,none,
         "TBD generation is only supported when the whole module can be seen",
         ())
 
+WARNING(linker_directives_choice_confusion,none,
+        "only one of -emit-ldadd-cfile-path and -module-installname-map-file can be specified;"
+        "the c file won't be generated",
+        ())
+
+ERROR(previous_installname_map_missing,none,
+      "cannot open previous install name map from %0",
+      (StringRef))
+
+ERROR(previous_installname_map_corrupted,none,
+      "previous install name map from %0 is malformed",
+      (StringRef))
+
+REMARK(default_previous_install_name, none,
+      "default previous install name for %0 is %1", (StringRef, StringRef))
+
+REMARK(platform_previous_install_name, none,
+       "previous install name for %0 in %1 is %2", (StringRef, StringRef, StringRef))
+
+ERROR(unknown_platform_name, none,
+      "unkown platform name %0", (StringRef))
+
 ERROR(symbol_in_tbd_not_in_ir,none,
       "symbol '%0' (%1) is in TBD file, but not in generated IR",
       (StringRef, StringRef))

include/swift/Option/FrontendOptions.td

@@ -658,4 +658,8 @@ def type_info_dump_filter_EQ : Joined<["-"], "type-info-dump-filter=">,
 def emit_ldadd_cfile_path
   : Separate<["-"], "emit-ldadd-cfile-path">, MetaVarName<"<path>">,
     HelpText<"Generate .c file defining symbols to add back">;
+
+def previous_module_installname_map_file
+  : Separate<["-"], "previous-module-installname-map-file">, MetaVarName<"<path>">,
+    HelpText<"Path to a Json file indicating module name to installname map for @_originallyDefinedIn">;
 } // end let Flags = [FrontendOption, NoDriverOption, HelpHidden]

include/swift/TBDGen/TBDGen.h

@@ -49,6 +49,10 @@ struct TBDGenOptions {
   /// The dylib compatibility-version to use in the generated TBD file. Defaults
   /// to empty string if not provided.
   std::string CompatibilityVersion;
+
+  /// The path to a Json file indicating the module name to install-name map
+  /// used by @_originallyDefinedIn
+  std::string ModuleInstallNameMapPath;
 };
 
 void enumeratePublicSymbols(FileUnit *module, llvm::StringSet<> &symbols,

lib/Frontend/CompilerInvocation.cpp

@@ -1018,6 +1018,9 @@ static bool ParseTBDGenArgs(TBDGenOptions &Opts, ArgList &Args,
   if (const Arg *A = Args.getLastArg(OPT_tbd_current_version)) {
     Opts.CurrentVersion = A->getValue();
   }
+  if (const Arg *A = Args.getLastArg(OPT_previous_module_installname_map_file)) {
+    Opts.ModuleInstallNameMapPath = A->getValue();
+  }
   return false;
 }
 

lib/FrontendTool/FrontendTool.cpp

@@ -1068,6 +1068,11 @@ static bool writeLdAddCFileIfNeeded(CompilerInvocation &Invocation,
                                  diag::tbd_only_supported_in_whole_module);
     return true;
   }
+  if (!Invocation.getTBDGenOptions().ModuleInstallNameMapPath.empty()) {
+    Instance.getDiags().diagnose(SourceLoc(),
+                                 diag::linker_directives_choice_confusion);
+    return true;
+  }
   auto tbdOpts = Invocation.getTBDGenOptions();
   tbdOpts.LinkerDirectivesOnly = true;
   llvm::StringSet<> ldSymbols;

lib/TBDGen/TBDGen.cpp

@@ -36,10 +36,12 @@
 #include "swift/SIL/TypeLowering.h"
 #include "clang/Basic/TargetInfo.h"
 #include "llvm/ADT/StringSet.h"
+#include "llvm/ADT/StringSwitch.h"
 #include "llvm/IR/Mangler.h"
 #include "llvm/Support/Error.h"
 #include "llvm/Support/Process.h"
 #include "llvm/Support/YAMLTraits.h"
+#include "llvm/Support/YAMLParser.h"
 #include "llvm/TextAPI/MachO/InterfaceFile.h"
 #include "llvm/TextAPI/MachO/TextAPIReader.h"
 #include "llvm/TextAPI/MachO/TextAPIWriter.h"
@@ -49,6 +51,7 @@
 using namespace swift;
 using namespace swift::irgen;
 using namespace swift::tbdgen;
+using namespace llvm::yaml;
 using StringSet = llvm::StringSet<>;
 using SymbolKind = llvm::MachO::SymbolKind;
 
@@ -79,6 +82,152 @@ static Optional<llvm::VersionTuple> getDeclMoveOSVersion(Decl *D) {
   return None;
 }
 
+enum class LinkerPlatformId: uint8_t {
+#define LD_PLATFORM(Name, Id) Name = Id,
+#include "ldPlatformKinds.def"
+};
+
+static StringRef getLinkerPlatformName(uint8_t Id) {
+  switch (Id) {
+#define LD_PLATFORM(Name, Id) case Id: return #Name;
+#include "ldPlatformKinds.def"
+  default:
+    llvm_unreachable("unrecognized platform id");
+  }
+}
+
+static Optional<uint8_t> getLinkerPlatformId(StringRef Platform) {
+  return llvm::StringSwitch<Optional<uint8_t>>(Platform)
+#define LD_PLATFORM(Name, Id) .Case(#Name, Id)
+#include "ldPlatformKinds.def"
+    .Default(None);
+}
+
+struct InstallNameStore {
+  // The default install name to use when no specific install name is specified.
+  std::string InstallName;
+  // The install name specific to the platform id. This takes precedence over
+  // the default install name.
+  std::map<uint8_t, std::string> PlatformInstallName;
+  StringRef getInstallName(LinkerPlatformId Id) const {
+    auto It = PlatformInstallName.find((uint8_t)Id);
+    if (It == PlatformInstallName.end())
+      return InstallName;
+    else
+      return It->second;
+  }
+  void remark(ASTContext &Ctx, StringRef ModuleName) const {
+    Ctx.Diags.diagnose(SourceLoc(), diag::default_previous_install_name,
+                       ModuleName, InstallName);
+    for (auto Pair: PlatformInstallName) {
+      Ctx.Diags.diagnose(SourceLoc(), diag::platform_previous_install_name,
+                         ModuleName, getLinkerPlatformName(Pair.first),
+                         Pair.second);
+    }
+  }
+};
+
+static std::string getScalaNodeText(Node *N) {
+  SmallString<32> Buffer;
+  return cast<ScalarNode>(N)->getValue(Buffer).str();
+}
+
+static std::set<int8_t> getSequenceNodePlatformList(ASTContext &Ctx, Node *N) {
+  std::set<int8_t> Results;
+  for (auto &E: *cast<SequenceNode>(N)) {
+    auto Platform = getScalaNodeText(&E);
+    auto Id = getLinkerPlatformId(Platform);
+    if (Id.hasValue()) {
+      Results.insert(*Id);
+    } else {
+      // Diagnose unrecognized platform name.
+      Ctx.Diags.diagnose(SourceLoc(), diag::unknown_platform_name, Platform);
+    }
+  }
+  return Results;
+}
+
+/// Parse an entry like this, where the "platforms" key-value pair is optional:
+///  {
+///     "module": "Foo",
+///     "platforms": ["macOS"],
+///     "install_name": "/System/MacOS"
+///  },
+static int
+parseEntry(ASTContext &Ctx,
+           Node *Node, std::map<std::string, InstallNameStore> &Stores) {
+  if (auto *SN = cast<SequenceNode>(Node)) {
+    for (auto It = SN->begin(); It != SN->end(); ++It) {
+      auto *MN = cast<MappingNode>(&*It);
+      std::string ModuleName;
+      std::string InstallName;
+      Optional<std::set<int8_t>> Platforms;
+      for (auto &Pair: *MN) {
+        auto Key = getScalaNodeText(Pair.getKey());
+        auto* Value = Pair.getValue();
+        if (Key == "module") {
+          ModuleName = getScalaNodeText(Value);
+        } else if (Key == "platforms") {
+          Platforms = getSequenceNodePlatformList(Ctx, Value);
+        } else if (Key == "install_name") {
+          InstallName = getScalaNodeText(Value);
+        } else {
+          return 1;
+        }
+      }
+      if (ModuleName.empty() || InstallName.empty())
+        return 1;
+      auto &Store = Stores.insert(std::make_pair(ModuleName,
+        InstallNameStore())).first->second;
+      if (Platforms.hasValue()) {
+        // This install name is platform-specific.
+        for (auto Id: Platforms.getValue()) {
+          Store.PlatformInstallName[Id] = InstallName;
+        }
+      } else {
+        // The install name is the default one.
+        Store.InstallName = InstallName;
+      }
+    }
+  } else {
+    return 1;
+  }
+  return 0;
+}
+
+static std::map<std::string, InstallNameStore>
+parsePreviousModuleInstallNameMap(ASTContext &Ctx, StringRef FileName) {
+  namespace yaml = llvm::yaml;
+  std::map<std::string, InstallNameStore> AllInstallNames;
+  SWIFT_DEFER {
+    for (auto Pair: AllInstallNames) {
+      Pair.second.remark(Ctx, Pair.first);
+    }
+  };
+  // Load the input file.
+  llvm::ErrorOr<std::unique_ptr<llvm::MemoryBuffer>> FileBufOrErr =
+    vfs::getFileOrSTDIN(*Ctx.SourceMgr.getFileSystem(), FileName);
+  if (!FileBufOrErr) {
+    Ctx.Diags.diagnose(SourceLoc(), diag::previous_installname_map_missing,
+                       FileName);
+    return AllInstallNames;
+  }
+  StringRef Buffer = FileBufOrErr->get()->getBuffer();
+  yaml::Stream Stream(llvm::MemoryBufferRef(Buffer, FileName),
+                      Ctx.SourceMgr.getLLVMSourceMgr());
+  for (auto DI = Stream.begin(); DI != Stream.end(); ++ DI) {
+    assert(DI != Stream.end() && "Failed to read a document");
+    yaml::Node *N = DI->getRoot();
+    assert(N && "Failed to find a root");
+    if (parseEntry(Ctx, N, AllInstallNames)) {
+      Ctx.Diags.diagnose(SourceLoc(), diag::previous_installname_map_corrupted,
+                         FileName);
+      return AllInstallNames;
+    }
+  }
+  return AllInstallNames;
+}
+
 void TBDGenVisitor::addLinkerDirectiveSymbols(StringRef name,
                                               llvm::MachO::SymbolKind kind) {
   if (kind != llvm::MachO::SymbolKind::GlobalSymbol)
@@ -733,6 +882,10 @@ static void enumeratePublicSymbolsAndWrite(ModuleDecl *M, FileUnit *singleFile,
     file.setCompatibilityVersion(*packed);
   }
 
+  if (!opts.ModuleInstallNameMapPath.empty()) {
+    parsePreviousModuleInstallNameMap(ctx, opts.ModuleInstallNameMapPath);
+  }
+
   llvm::MachO::Target target(triple);
   file.addTarget(target);
 

lib/TBDGen/ldPlatformKinds.def

@@ -0,0 +1,30 @@
+//===--- ldPlatformKinds.def - Compiler declaration metaprogramming --*- C++ -*-===//
+//
+// This source file is part of the Swift.org open source project
+//
+// Copyright (c) 2014 - 2020 Apple Inc. and the Swift project authors
+// Licensed under Apache License v2.0 with Runtime Library Exception
+//
+// See https://swift.org/LICENSE.txt for license information
+// See https://swift.org/CONTRIBUTORS.txt for the list of Swift project authors
+//
+//===----------------------------------------------------------------------===//
+//
+// This file defines platform IDs used to emit linker directives $ld$previous
+//
+//===----------------------------------------------------------------------===//
+
+#ifndef LD_PLATFORM
+#  define LD_PLATFORM(Name, Id)
+#endif
+
+LD_PLATFORM(macOS, 1)
+LD_PLATFORM(iOS, 2)
+LD_PLATFORM(tvOS, 3)
+LD_PLATFORM(watchOS, 4)
+LD_PLATFORM(macCatalyst, 6)
+LD_PLATFORM(iOS_sim, 7)
+LD_PLATFORM(tvOS_sim, 8)
+LD_PLATFORM(watchOS_sim, 9)
+
+#undef LD_PLATFORM

test/TBD/Inputs/install-name-map.json

@@ -0,0 +1,16 @@
+[
+  {
+    "module": "Foo",
+    "platforms": ["macOS"],
+    "install_name": "/System/MacOS"
+  },
+  {
+    "module": "Foo",
+    "install_name": "/System/default"
+  },
+  {
+    "module": "Foo",
+    "platforms": ["iOS", "tvOS", "watchOS"],
+    "install_name": "/System/Others"
+  },
+]

test/TBD/previous-install-name-map.swift

@@ -0,0 +1,13 @@
+// REQUIRES: VENDOR=apple
+// REQUIRES: OS=macosx
+
+// RUN: %empty-directory(%t)
+
+// RUN: %target-swift-frontend -typecheck %s -emit-tbd -emit-tbd-path %t/linker_directives.tbd -previous-module-installname-map-file %S/Inputs/install-name-map.json >& %t/remark.txt
+// RUN: %FileCheck %s < %t/remark.txt
+
+// CHECK: remark: default previous install name for Foo is /System/default
+// CHECK: remark: previous install name for Foo in macOS is /System/MacOS
+// CHECK: remark: previous install name for Foo in iOS is /System/Others
+// CHECK: remark: previous install name for Foo in tvOS is /System/Others
+// CHECK: remark: previous install name for Foo in watchOS is /System/Others