Swift: rework file redirection

The hash map mechanism that was already in use for reading swiftmodule
files on macOS is now in use also on Linux. The output replacing
mechanism has been also reworked so that:
* frontend module emission modes have the remapping done directly in
  the internal frontend options instead of painstakingly modifying input
  flags (this requires a patch on the swift headers though)
* object emission mode is silenced to be just a type checking pass,
  thus producing no output files
* all other passes but some debugging and version related ones become
  noops

The open file read redirection uses a global weak pointer instance to
maximize robustness in the face of possibly multi-threaded calls to open
happening while `main` is exiting. Possibly overkill, but better safe
than sorry.

Author: redsun82

swift/extractor/BUILD.bazel

@@ -9,6 +9,7 @@ swift_cc_binary(
     ]),
     visibility = ["//swift:__pkg__"],
     deps = [
+        "//swift/extractor/config",
         "//swift/extractor/infra",
         "//swift/extractor/invocation",
         "//swift/extractor/remapping",

swift/extractor/SwiftExtractor.h

@@ -1,6 +1,6 @@
 #pragma once
 
-#include "swift/extractor/SwiftExtractorConfiguration.h"
+#include "swift/extractor/config/SwiftExtractorConfiguration.h"
 #include <swift/AST/SourceFile.h>
 #include <swift/Frontend/Frontend.h>
 #include <memory>

swift/extractor/TargetTrapFile.cpp

@@ -11,11 +11,6 @@ std::optional<TargetFile> createTargetTrapFile(const SwiftExtractorConfiguration
     for (const auto& opt : configuration.frontendOptions) {
       *ret << "  " << std::quoted(opt) << " \\\n";
     }
-    *ret << "\n*/\n"
-            "/* swift-frontend-args:\n";
-    for (const auto& opt : configuration.patchedFrontendOptions) {
-      *ret << "  " << std::quoted(opt) << " \\\n";
-    }
     *ret << "\n*/\n";
   }
   return ret;

swift/extractor/TargetTrapFile.h

@@ -1,7 +1,7 @@
 #pragma once
 
 #include "swift/extractor/infra/file/TargetFile.h"
-#include "swift/extractor/SwiftExtractorConfiguration.h"
+#include "swift/extractor/config/SwiftExtractorConfiguration.h"
 
 namespace codeql {
 

swift/extractor/config/BUILD.bazel

@@ -0,0 +1,8 @@
+load("//swift:rules.bzl", "swift_cc_library")
+
+swift_cc_library(
+    name = "config",
+    srcs = glob(["*.cpp"]),
+    hdrs = glob(["*.h"]),
+    visibility = ["//swift:__subpackages__"],
+)

swift/extractor/SwiftExtractorConfiguration.h renamed to swift/extractor/config/SwiftExtractorConfiguration.h

@@ -17,9 +17,7 @@ struct SwiftExtractorConfiguration {
   std::filesystem::path scratchDir;
 
   // The original arguments passed to the extractor. Used for debugging.
-  std::vector<std::string> frontendOptions;
-  // The patched arguments passed to the swift::performFrontend/ Used for debugging.
-  std::vector<std::string> patchedFrontendOptions;
+  std::vector<const char*> frontendOptions;
 
   // A temporary directory that contains TRAP files before they are moved into their final
   // destination.

swift/extractor/main.cpp

@@ -12,13 +12,58 @@
 
 #include "swift/extractor/SwiftExtractor.h"
 #include "swift/extractor/TargetTrapFile.h"
-#include "swift/extractor/remapping/SwiftOutputRewrite.h"
-#include "swift/extractor/remapping/SwiftOpenInterception.h"
+#include "swift/extractor/remapping/SwiftFileInterception.h"
 #include "swift/extractor/invocation/SwiftDiagnosticsConsumer.h"
 #include "swift/extractor/trap/TrapDomain.h"
 
 using namespace std::string_literals;
 
+static void lockOutputSwiftModuleTraps(const codeql::SwiftExtractorConfiguration& config,
+                                       const swift::CompilerInstance& compiler) {
+  std::filesystem::path output = compiler.getInvocation().getOutputFilename();
+  if (output.extension() == ".swiftmodule") {
+    if (auto target = codeql::createTargetTrapFile(config, output)) {
+      *target << "// trap file deliberately empty\n"
+                 "// this swiftmodule was created during the build, so its entities must have"
+                 " been extracted directly from source files";
+    }
+  }
+}
+
+static void modifyFrontendOptions(swift::FrontendOptions& options) {
+  using Action = swift::FrontendOptions::ActionType;
+  switch (options.RequestedAction) {
+    case Action::EmitModuleOnly:
+    case Action::MergeModules:
+    case Action::CompileModuleFromInterface:
+      // for module emission actions, we redirect the output to our internal artifact storage
+      {
+        swift::SupplementaryOutputPaths paths;
+        paths.ModuleOutputPath =
+            codeql::redirect(options.InputsAndOutputs.getSingleOutputFilename()).string();
+        options.InputsAndOutputs.setMainAndSupplementaryOutputs(std::vector{paths.ModuleOutputPath},
+                                                                std::vector{paths});
+        return;
+      }
+    case Action::EmitObject:
+      // for object emission, we do a type check pass instead, muting output but getting the sema
+      // phase to run in order to extract everything
+      options.RequestedAction = Action::Typecheck;
+      return;
+    case Action::PrintVersion:
+    case Action::DumpAST:
+    case Action::PrintAST:
+    case Action::PrintASTDecl:
+      // these actions are nice to have on the extractor for debugging, so we preserve them. Also,
+      // version printing is used by CI to match up the correct compiler version
+      return;
+    default:
+      // otherwise, do nothing (the closest action to doing nothing is printing the version)
+      options.RequestedAction = Action::PrintVersion;
+      break;
+  }
+}
+
 // This is part of the swiftFrontendTool interface, we hook into the
 // compilation pipeline and extract files after the Swift frontend performed
 // semantic analysis
@@ -28,8 +73,13 @@ class Observer : public swift::FrontendObserver {
                     codeql::SwiftDiagnosticsConsumer& diagConsumer)
       : config{config}, diagConsumer{diagConsumer} {}
 
+  void parsedArgs(swift::CompilerInvocation& invocation) override {
+    modifyFrontendOptions(invocation.getFrontendOptions());
+  }
+
   void configuredCompiler(swift::CompilerInstance& instance) override {
     instance.addDiagnosticConsumer(&diagConsumer);
+    lockOutputSwiftModuleTraps(config, instance);
   }
 
   void performedSemanticAnalysis(swift::CompilerInstance& compiler) override {
@@ -48,19 +98,6 @@ static std::string getenv_or(const char* envvar, const std::string& def) {
   return def;
 }
 
-static void lockOutputSwiftModuleTraps(const codeql::SwiftExtractorConfiguration& config,
-                                       const codeql::PathRemapping& remapping) {
-  for (const auto& [oldPath, newPath] : remapping) {
-    if (oldPath.extension() == ".swiftmodule") {
-      if (auto target = codeql::createTargetTrapFile(config, oldPath)) {
-        *target << "// trap file deliberately empty\n"
-                   "// this swiftmodule was created during the build, so its entities must have"
-                   " been extracted directly from source files";
-      }
-    }
-  }
-}
-
 static bool checkRunUnderFilter(int argc, char* const* argv) {
   auto runUnderFilter = getenv("CODEQL_EXTRACTOR_SWIFT_RUN_UNDER_FILTER");
   if (runUnderFilter == nullptr) {
@@ -106,7 +143,7 @@ static void checkWhetherToRunUnderTool(int argc, char* const* argv) {
 
 // Creates a target file that should store per-invocation info, e.g. compilation args,
 // compilations, diagnostics, etc.
-codeql::TargetFile invocationTargetFile(codeql::SwiftExtractorConfiguration& configuration) {
+codeql::TargetFile invocationTargetFile(const codeql::SwiftExtractorConfiguration& configuration) {
   auto timestamp = std::chrono::system_clock::now().time_since_epoch().count();
   auto filename = std::to_string(timestamp) + '-' + std::to_string(getpid());
   auto target = std::filesystem::path("invocations") / std::filesystem::path(filename);
@@ -118,6 +155,15 @@ codeql::TargetFile invocationTargetFile(codeql::SwiftExtractorConfiguration& con
   return std::move(maybeFile.value());
 }
 
+codeql::SwiftExtractorConfiguration configure(int argc, char** argv) {
+  codeql::SwiftExtractorConfiguration configuration{};
+  configuration.trapDir = getenv_or("CODEQL_EXTRACTOR_SWIFT_TRAP_DIR", ".");
+  configuration.sourceArchiveDir = getenv_or("CODEQL_EXTRACTOR_SWIFT_SOURCE_ARCHIVE_DIR", ".");
+  configuration.scratchDir = getenv_or("CODEQL_EXTRACTOR_SWIFT_SCRATCH_DIR", ".");
+  configuration.frontendOptions.assign(argv + 1, argv + argc);
+  return configuration;
+}
+
 int main(int argc, char** argv) {
   checkWhetherToRunUnderTool(argc, argv);
 
@@ -130,36 +176,16 @@ int main(int argc, char** argv) {
   PROGRAM_START(argc, argv);
   INITIALIZE_LLVM();
 
-  codeql::SwiftExtractorConfiguration configuration{};
-  configuration.trapDir = getenv_or("CODEQL_EXTRACTOR_SWIFT_TRAP_DIR", ".");
-  configuration.sourceArchiveDir = getenv_or("CODEQL_EXTRACTOR_SWIFT_SOURCE_ARCHIVE_DIR", ".");
-  configuration.scratchDir = getenv_or("CODEQL_EXTRACTOR_SWIFT_SCRATCH_DIR", ".");
-
-  codeql::initRemapping(configuration.getTempArtifactDir());
-
-  configuration.frontendOptions.reserve(argc - 1);
-  for (int i = 1; i < argc; i++) {
-    configuration.frontendOptions.push_back(argv[i]);
-  }
-  configuration.patchedFrontendOptions = configuration.frontendOptions;
-
-  auto remapping = codeql::rewriteOutputsInPlace(configuration.getTempArtifactDir(),
-                                                 configuration.patchedFrontendOptions);
-  codeql::ensureDirectoriesForNewPathsExist(remapping);
-  lockOutputSwiftModuleTraps(configuration, remapping);
+  const auto configuration = configure(argc, argv);
 
-  std::vector<const char*> args;
-  for (auto& arg : configuration.patchedFrontendOptions) {
-    args.push_back(arg.c_str());
-  }
+  auto openInterception = codeql::setupFileInterception(configuration.getTempArtifactDir());
 
   auto invocationTrapFile = invocationTargetFile(configuration);
   codeql::TrapDomain invocationDomain(invocationTrapFile);
   codeql::SwiftDiagnosticsConsumer diagConsumer(invocationDomain);
   Observer observer(configuration, diagConsumer);
-  int frontend_rc = swift::performFrontend(args, "swift-extractor", (void*)main, &observer);
-
-  codeql::finalizeRemapping(remapping);
+  int frontend_rc = swift::performFrontend(configuration.frontendOptions, "swift-extractor",
+                                           (void*)main, &observer);
 
   return frontend_rc;
 }

swift/extractor/remapping/BUILD.bazel

@@ -2,23 +2,11 @@ load("//swift:rules.bzl", "swift_cc_library")
 
 swift_cc_library(
     name = "remapping",
-    srcs = ["SwiftOutputRewrite.cpp"] + select({
-        "@platforms//os:linux": [
-            "SwiftOpenInterception.Linux.cpp",
-        ],
-        "@platforms//os:macos": [
-            "SwiftOpenInterception.macOS.cpp",
-        ],
-    }),
+    srcs = glob(["*.cpp"]),
     hdrs = glob(["*.h"]),
     visibility = ["//swift:__subpackages__"],
     deps = [
-        "//swift/third_party/swift-llvm-support",
         "//swift/extractor/infra/file",
-    ] + select({
-        "@platforms//os:linux": [],
-        "@platforms//os:macos": [
-            "@fishhook//:fishhook",
-        ],
-    }),
+        "//swift/third_party/swift-llvm-support",
+    ],
 )