[lld][MachO]Multi-threaded i/o. Twice as fast linking a large project.

lld/MachO/Config.h

@@ -186,6 +186,7 @@ struct Configuration {
   bool interposable = false;
   bool errorForArchMismatch = false;
   bool ignoreAutoLink = false;
+  int readThreads = 0;
   // ld64 allows invalid auto link options as long as the link succeeds. LLD
   // does not, but there are cases in the wild where the invalid linker options
   // exist. This allows users to ignore the specific invalid options in the case

lld/MachO/Driver.cpp

@@ -44,6 +44,7 @@
 #include "llvm/Support/FileSystem.h"
 #include "llvm/Support/Parallel.h"
 #include "llvm/Support/Path.h"
+#include "llvm/Support/Process.h"
 #include "llvm/Support/TarWriter.h"
 #include "llvm/Support/TargetSelect.h"
 #include "llvm/Support/TimeProfiler.h"
@@ -282,11 +283,103 @@ static void saveThinArchiveToRepro(ArchiveFile const *file) {
           ": Archive::children failed: " + toString(std::move(e)));
 }
 
-static InputFile *addFile(StringRef path, LoadType loadType,
-                          bool isLazy = false, bool isExplicit = true,
-                          bool isBundleLoader = false,
-                          bool isForceHidden = false) {
-  std::optional<MemoryBufferRef> buffer = readFile(path);
+typedef struct {
+  StringRef path;
+  bool isLazy;
+  std::optional<MemoryBufferRef> buffer;
+  const char *start;
+  size_t size;
+} DeferredFile;
+typedef std::vector<DeferredFile> DeferredFiles;
+
+#ifndef _WIN32
+typedef struct {
+  DeferredFiles deferred;
+  size_t counter, total, pageSize;
+  pthread_mutex_t mutex;
+} PageInState;
+
+// Most input files have been mapped but not yet paged in.
+// This code forces the page-ins on multiple threads so
+// the process is not stalled waiting on disk buffer i/o.
+static void multiThreadedPageInBackground(PageInState *state) {
+#define MaxReadThreads 200
+  static size_t totalBytes;
+
+  pthread_t running[MaxReadThreads];
+  if (config->readThreads > MaxReadThreads)
+    config->readThreads = MaxReadThreads;
+  pthread_mutex_init(&state->mutex, nullptr);
+
+  for (int t = 0; t < config->readThreads; t++)
+    pthread_create(
+        &running[t], nullptr,
+        [](void *ptr) -> void * {
+          PageInState &state = *(PageInState *)ptr;
+          while (true) {
+            pthread_mutex_lock(&state.mutex);
+            if (state.counter >= state.deferred.size()) {
+              pthread_mutex_unlock(&state.mutex);
+              return nullptr;
+            }
+            DeferredFile &file = state.deferred[state.counter];
+            state.counter += 1;
+            pthread_mutex_unlock(&state.mutex);
+
+            const char *page = file.start, *end = page + file.size;
+            totalBytes += end - page;
+
+            int t = 0; // Reference each page to load it into memory.
+            for (; page < end; page += state.pageSize)
+              t += *page;
+            state.total += t; // Avoids the loop being optimised out.
+          }
+        },
+        state);
+
+  for (int t = 0; t < config->readThreads; t++)
+    pthread_join(running[t], nullptr);
+
+  pthread_mutex_destroy(&state->mutex);
+  if (getenv("LLD_MULTI_THREAD_PAGE"))
+    printf("multiThreadedPageIn %ld/%ld\n", totalBytes, state->deferred.size());
+}
+#endif
+
+static void multiThreadedPageIn(DeferredFiles deferred) {
+#ifndef _WIN32
+  static pthread_t running;
+  static pthread_mutex_t busy;
+
+  if (running)
+    pthread_join(running, nullptr);
+  else
+    pthread_mutex_init(&busy, nullptr);
+
+  PageInState *state =
+      new PageInState{deferred, 0, 0, llvm::sys::Process::getPageSizeEstimate(),
+                      pthread_mutex_t()};
+
+  pthread_mutex_lock(&busy);
+  pthread_create(
+      &running, nullptr,
+      [](void *ptr) -> void * {
+        PageInState *state = (PageInState *)ptr;
+        multiThreadedPageInBackground(state);
+        pthread_mutex_unlock(&busy);
+        delete state;
+        return nullptr;
+      },
+      state);
+#endif
+}
+
+static InputFile *processFile(std::optional<MemoryBufferRef> buffer,
+                              DeferredFiles *archiveContents, StringRef path,
+                              LoadType loadType, bool isLazy = false,
+                              bool isExplicit = true,
+                              bool isBundleLoader = false,
+                              bool isForceHidden = false) {
   if (!buffer)
     return nullptr;
   MemoryBufferRef mbref = *buffer;
@@ -379,6 +472,10 @@ static InputFile *addFile(StringRef path, LoadType loadType,
             continue;
           }
 
+          if (archiveContents)
+            archiveContents->push_back({path, isLazy, std::nullopt,
+                                        mb->getBuffer().data(),
+                                        mb->getBuffer().size()});
           if (!hasObjCSection(*mb))
             continue;
           if (Error e = file->fetch(c, "-ObjC"))
@@ -390,7 +487,8 @@ static InputFile *addFile(StringRef path, LoadType loadType,
                 ": Archive::children failed: " + toString(std::move(e)));
       }
     }
-    file->addLazySymbols();
+    if (!archiveContents || archiveContents->empty())
+      file->addLazySymbols();
     loadedArchives[path] = ArchiveFileInfo{file, isCommandLineLoad};
     newFile = file;
     break;
@@ -441,6 +539,23 @@ static InputFile *addFile(StringRef path, LoadType loadType,
   return newFile;
 }
 
+static InputFile *addFile(StringRef path, LoadType loadType,
+                          bool isLazy = false, bool isExplicit = true,
+                          bool isBundleLoader = false,
+                          bool isForceHidden = false) {
+  return processFile(readFile(path), nullptr, path, loadType, isLazy,
+                     isExplicit, isBundleLoader, isForceHidden);
+}
+
+static void deferFile(StringRef path, bool isLazy, DeferredFiles &deferred) {
+  std::optional<MemoryBufferRef> buffer = readFile(path);
+  if (config->readThreads)
+    deferred.push_back({path, isLazy, buffer, buffer->getBuffer().data(),
+                        buffer->getBuffer().size()});
+  else
+    processFile(buffer, nullptr, path, LoadType::CommandLine, isLazy);
+}
+
 static std::vector<StringRef> missingAutolinkWarnings;
 static void addLibrary(StringRef name, bool isNeeded, bool isWeak,
                        bool isReexport, bool isHidden, bool isExplicit,
@@ -564,13 +679,14 @@ void macho::resolveLCLinkerOptions() {
   }
 }
 
-static void addFileList(StringRef path, bool isLazy) {
+static void addFileList(StringRef path, bool isLazy,
+                        DeferredFiles &deferredFiles) {
   std::optional<MemoryBufferRef> buffer = readFile(path);
   if (!buffer)
     return;
   MemoryBufferRef mbref = *buffer;
   for (StringRef path : args::getLines(mbref))
-    addFile(rerootPath(path), LoadType::CommandLine, isLazy);
+    deferFile(rerootPath(path), isLazy, deferredFiles);
 }
 
 // We expect sub-library names of the form "libfoo", which will match a dylib
@@ -1222,14 +1338,16 @@ static void createFiles(const InputArgList &args) {
   bool isLazy = false;
   // If we've processed an opening --start-lib, without a matching --end-lib
   bool inLib = false;
+  DeferredFiles deferredFiles;
+
   for (const Arg *arg : args) {
     const Option &opt = arg->getOption();
     warnIfDeprecatedOption(opt);
     warnIfUnimplementedOption(opt);
 
     switch (opt.getID()) {
     case OPT_INPUT:
-      addFile(rerootPath(arg->getValue()), LoadType::CommandLine, isLazy);
+      deferFile(rerootPath(arg->getValue()), isLazy, deferredFiles);
       break;
     case OPT_needed_library:
       if (auto *dylibFile = dyn_cast_or_null<DylibFile>(
@@ -1249,7 +1367,7 @@ static void createFiles(const InputArgList &args) {
         dylibFile->forceWeakImport = true;
       break;
     case OPT_filelist:
-      addFileList(arg->getValue(), isLazy);
+      addFileList(arg->getValue(), isLazy, deferredFiles);
       break;
     case OPT_force_load:
       addFile(rerootPath(arg->getValue()), LoadType::CommandLineForce);
@@ -1295,6 +1413,28 @@ static void createFiles(const InputArgList &args) {
       break;
     }
   }
+
+  if (config->readThreads) {
+    multiThreadedPageIn(deferredFiles);
+
+    DeferredFiles archiveContents;
+    std::vector<ArchiveFile *> archives;
+    for (auto &file : deferredFiles)
+      if (ArchiveFile *archive = dyn_cast<ArchiveFile>(
+              processFile(file.buffer, &archiveContents, file.path,
+                          LoadType::CommandLine, file.isLazy)))
+        archives.push_back(archive);
+
+    if (!archiveContents.empty()) {
+      multiThreadedPageIn(archiveContents);
+      for (auto *archive : archives)
+        archive->addLazySymbols();
+    }
+
+    // flush threads
+    deferredFiles.clear();
+    multiThreadedPageIn(deferredFiles);
+  }
 }
 
 static void gatherInputSections() {
@@ -1687,6 +1827,14 @@ bool link(ArrayRef<const char *> argsArr, llvm::raw_ostream &stdoutOS,
     }
   }
 
+  if (auto *arg = args.getLastArg(OPT_read_threads)) {
+    StringRef v(arg->getValue());
+    unsigned threads = 0;
+    if (!llvm::to_integer(v, threads, 0) || threads < 0)
+      error(arg->getSpelling() + ": expected a positive integer, but got '" +
+            arg->getValue() + "'");
+    config->readThreads = threads;
+  }
   if (auto *arg = args.getLastArg(OPT_threads_eq)) {
     StringRef v(arg->getValue());
     unsigned threads = 0;

lld/MachO/Options.td

@@ -396,6 +396,9 @@ def dead_strip : Flag<["-"], "dead_strip">,
 def interposable : Flag<["-"], "interposable">,
     HelpText<"Indirects access to all exported symbols in an image">,
     Group<grp_opts>;
+def read_threads : Joined<["--"], "read-threads=">,
+    HelpText<"Number of threads to use paging in files.">,
+    Group<grp_lld>;
 def order_file : Separate<["-"], "order_file">,
     MetaVarName<"<file>">,
     HelpText<"Layout functions and data according to specification in <file>">,