[Support] Move LLD parallel algorithms to LLVM.

[zturner]

Summary

NFC, this is just a straight code move along with header / namespace fixups.

[chandlerc]

Maybe I'm stuck with an old diff or something?

lld/COFF/Writer.cpp

748 ↗

(On Diff #98385)

I find the fact that for_each is now getting looked up via ADL on the llvm::parallel::par type ... disturbing.

llvm/include/llvm/Support/TaskGroup.h

51–53 ↗

(On Diff #98385)

I thought all of this was being moved to be implementation details of the algorithms?

Also, all the code appears to still be LLD style?

[zturner]

lld/COFF/Writer.cpp

748 ↗

(On Diff #98385)

I can certainly qualify the function call, but I don't know of a way to disable ADL if that's what you're suggesting.

llvm/include/llvm/Support/TaskGroup.h

51–53 ↗

(On Diff #98385)

The executor stuff was all moved to an anonymous namespace inside of TaskGroup.cpp. TaskGroup cannot enjoy the same level of hiding because it needs to be usable from within a header file. Namely, Parallel.h. I can, however, move it to a detail namespace inside of Parallel.h if you prefer that.

also: Yes rui pointed out the names in a previous patch and I forgot to update them. All of them have been fixed except for the few instances in this file.

[chandlerc]

lld/COFF/Writer.cpp

748 ↗

(On Diff #98385)

It's "easy": foo(a, b, c) uses ADL and (foo)(a, b, c) doesn't use ADL. ;]

I'm not suggesting using (for_each) though. That seems a bit crazy.

Writing llvm::foo will also disable ADL. ADL only fires for unqualified names. But then writing the code gets super ugly:

llvm::for_each(llvm::parallel::par, ...);

Yuck. I guess I mildly prefer that? The standard sure chose an API that ends up requiring a lot of characters to type.

What do you or Rui think? I don't have strong feelings about which way you go here, so would generally defer to your or Rui's judgement.

llvm/include/llvm/Support/TaskGroup.h

51–53 ↗

(On Diff #98385)

I see, thanks.

I think a detail namespace is a touch cleaner. I'm happy for you to do that in this patch, same with any name cleanup.

[zturner]

• Changed naming conventions on the leftover LLD-style variables.
• Deleted TaskGroup.h and TaskGroup.cpp and moved corresponding code to Parallel.h and Parallel.cpp inside of the detail namespace.

[zturner]

lld/COFF/Writer.cpp

748 ↗

(On Diff #98385)

It's worse than that, because it's going to be llvm::parallel::for_each(llvm::parallel::par, ...). We could nuke the intermediate parallel namespace and go with llvm::for_each(llvm::par, ...), or we could do using llvm::parallel at the top of a file. The namespace is sufficiently scoped that bringing it in shouldn't be too onerous.

Personally I'm fine with any approach, including the ADL-based approach. That said, I don't think there's anything we can do to the API itself to alleviate this, so perhaps we don't need to do anything. Just leave it up to the users.

[zturner]

Rebased after renaming variables on the LLD side.

[ruiu]

lld/COFF/Writer.cpp

748 ↗

(On Diff #98385)

Because we have using namespace llvm at beginning of this file, you don't need to qualify it with llvm::, no?

[chandlerc]

LGTM on the LLVM side. Feel free to submit when you and Rui are happy with the usage pattern for this in LLD.

llvm/include/llvm/Support/Parallel.h

67 ↗

(On Diff #98491)

closing namespace comment?

[ruiu]

I'm happy as long as you remove unnecessary llvm:: specifier. LGTM.

lld/trunk/COFF/ICF.cpp

 //
 // See ELF/ICF.cpp for the details about the algortihm.
 //
 //===----------------------------------------------------------------------===//
 
 #include "Chunks.h"
 #include "Error.h"
 #include "Symbols.h"
-#include "lld/Core/Parallel.h"
 #include "llvm/ADT/Hashing.h"
 #include "llvm/Support/Debug.h"
+#include "llvm/Support/Parallel.h"
 #include "llvm/Support/raw_ostream.h"
 #include <algorithm>
 #include <atomic>
 #include <vector>
 
 using namespace llvm;
 
 namespace lld {

lld/trunk/COFF/MapFile.cpp

 //===----------------------------------------------------------------------===//
 
 #include "MapFile.h"
 #include "Error.h"
 #include "SymbolTable.h"
 #include "Symbols.h"
 #include "Writer.h"
 
-#include "lld/Core/Parallel.h"
+#include "llvm/Support/Parallel.h"
 #include "llvm/Support/raw_ostream.h"
 
 using namespace llvm;
 using namespace llvm::object;
 
 using namespace lld;
 using namespace lld::coff;
 

lld/trunk/COFF/Writer.cpp

 #include "DLL.h"
 #include "Error.h"
 #include "InputFiles.h"
 #include "MapFile.h"
 #include "Memory.h"
 #include "PDB.h"
 #include "SymbolTable.h"
 #include "Symbols.h"
-#include "lld/Core/Parallel.h"
 #include "llvm/ADT/DenseMap.h"
 #include "llvm/ADT/STLExtras.h"
 #include "llvm/ADT/StringSwitch.h"
 #include "llvm/Support/Debug.h"
 #include "llvm/Support/Endian.h"
 #include "llvm/Support/FileOutputBuffer.h"
+#include "llvm/Support/Parallel.h"
 #include "llvm/Support/RandomNumberGenerator.h"
 #include "llvm/Support/raw_ostream.h"
 #include <algorithm>
 #include <cstdio>
 #include <map>
 #include <memory>
 #include <utility>
 

lld/trunk/ELF/Threads.h

 //
 //===----------------------------------------------------------------------===//
 
 #ifndef LLD_ELF_THREADS_H
 #define LLD_ELF_THREADS_H
 
 #include "Config.h"
 
-#include "lld/Core/Parallel.h"
+#include "llvm/Support/Parallel.h"
 #include <functional>
 
 namespace lld {
 namespace elf {
 
 template <class IterTy, class FuncTy>
 void parallelForEach(IterTy Begin, IterTy End, FuncTy Fn) {
   if (Config->Threads)
-    for_each(parallel::par, Begin, End, Fn);
+    for_each(llvm::parallel::par, Begin, End, Fn);
   else
-    for_each(parallel::seq, Begin, End, Fn);
+    for_each(llvm::parallel::seq, Begin, End, Fn);
 }
 
 inline void parallelForEachN(size_t Begin, size_t End,
                              std::function<void(size_t)> Fn) {
   if (Config->Threads)
-    for_each_n(parallel::par, Begin, End, Fn);
+    for_each_n(llvm::parallel::par, Begin, End, Fn);
   else
-    for_each_n(parallel::seq, Begin, End, Fn);
+    for_each_n(llvm::parallel::seq, Begin, End, Fn);
 }
 }
 }
 
 #endif

lld/trunk/include/lld/Core/Parallel.h

-//===- lld/Core/Parallel.h - Parallel utilities ---------------------------===//
-//
-//                             The LLVM Linker
-//
-// This file is distributed under the University of Illinois Open Source
-// License. See LICENSE.TXT for details.
-//
-//===----------------------------------------------------------------------===//
-
-#ifndef LLD_CORE_PARALLEL_H
-#define LLD_CORE_PARALLEL_H
-
-#include "lld/Core/LLVM.h"
-#include "lld/Core/TaskGroup.h"
-#include "llvm/ADT/STLExtras.h"
-#include "llvm/Config/llvm-config.h"
-#include "llvm/Support/MathExtras.h"
-
-#include <algorithm>
-
-#if defined(_MSC_VER) && LLVM_ENABLE_THREADS
-#include <concrt.h>
-#include <ppl.h>
-#endif
-
-namespace lld {
-
-namespace parallel {
-struct sequential_execution_policy {};
-struct parallel_execution_policy {};
-
-template <typename T>
-struct is_execution_policy
-    : public std::integral_constant<
-          bool, llvm::is_one_of<T, sequential_execution_policy,
-                                parallel_execution_policy>::value> {};
-
-constexpr sequential_execution_policy seq{};
-constexpr parallel_execution_policy par{};
-
-namespace detail {
-
-#if LLVM_ENABLE_THREADS
-
-#if defined(_MSC_VER)
-template <class RandomAccessIterator, class Comparator>
-void parallel_sort(RandomAccessIterator Start, RandomAccessIterator End,
-                   const Comparator &Comp) {
-  concurrency::parallel_sort(Start, End, Comp);
-}
-template <class IterTy, class FuncTy>
-void parallel_for_each(IterTy Begin, IterTy End, FuncTy Fn) {
-  concurrency::parallel_for_each(Begin, End, Fn);
-}
-
-template <class IndexTy, class FuncTy>
-void parallel_for_each_n(IndexTy Begin, IndexTy End, FuncTy Fn) {
-  concurrency::parallel_for(Begin, End, Fn);
-}
-
-#else
-const ptrdiff_t MinParallelSize = 1024;
-
-/// \brief Inclusive median.
-template <class RandomAccessIterator, class Comparator>
-RandomAccessIterator medianOf3(RandomAccessIterator Start,
-                               RandomAccessIterator End,
-                               const Comparator &Comp) {
-  RandomAccessIterator Mid = Start + (std::distance(Start, End) / 2);
-  return Comp(*Start, *(End - 1))
-             ? (Comp(*Mid, *(End - 1)) ? (Comp(*Start, *Mid) ? Mid : Start)
-                                       : End - 1)
-             : (Comp(*Mid, *Start) ? (Comp(*(End - 1), *Mid) ? Mid : End - 1)
-                                   : Start);
-}
-
-template <class RandomAccessIterator, class Comparator>
-void parallel_quick_sort(RandomAccessIterator Start, RandomAccessIterator End,
-                         const Comparator &Comp, TaskGroup &TG, size_t Depth) {
-  // Do a sequential sort for small inputs.
-  if (std::distance(Start, End) < detail::MinParallelSize || Depth == 0) {
-    std::sort(Start, End, Comp);
-    return;
-  }
-
-  // Partition.
-  auto Pivot = medianOf3(Start, End, Comp);
-  // Move Pivot to End.
-  std::swap(*(End - 1), *Pivot);
-  Pivot = std::partition(Start, End - 1, [&Comp, End](decltype(*Start) V) {
-    return Comp(V, *(End - 1));
-  });
-  // Move Pivot to middle of partition.
-  std::swap(*Pivot, *(End - 1));
-
-  // Recurse.
-  TG.spawn([=, &Comp, &TG] {
-    parallel_quick_sort(Start, Pivot, Comp, TG, Depth - 1);
-  });
-  parallel_quick_sort(Pivot + 1, End, Comp, TG, Depth - 1);
-}
-
-template <class RandomAccessIterator, class Comparator>
-void parallel_sort(RandomAccessIterator Start, RandomAccessIterator End,
-                   const Comparator &Comp) {
-  TaskGroup TG;
-  parallel_quick_sort(Start, End, Comp, TG,
-                      llvm::Log2_64(std::distance(Start, End)) + 1);
-}
-
-template <class IterTy, class FuncTy>
-void parallel_for_each(IterTy Begin, IterTy End, FuncTy Fn) {
-  // TaskGroup has a relatively high overhead, so we want to reduce
-  // the number of spawn() calls. We'll create up to 1024 tasks here.
-  // (Note that 1024 is an arbitrary number. This code probably needs
-  // improving to take the number of available cores into account.)
-  ptrdiff_t TaskSize = std::distance(Begin, End) / 1024;
-  if (TaskSize == 0)
-    TaskSize = 1;
-
-  TaskGroup TG;
-  while (TaskSize <= std::distance(Begin, End)) {
-    TG.spawn([=, &Fn] { std::for_each(Begin, Begin + TaskSize, Fn); });
-    Begin += TaskSize;
-  }
-  TG.spawn([=, &Fn] { std::for_each(Begin, End, Fn); });
-}
-
-template <class IndexTy, class FuncTy>
-void parallel_for_each_n(IndexTy Begin, IndexTy End, FuncTy Fn) {
-  ptrdiff_t TaskSize = (End - Begin) / 1024;
-  if (TaskSize == 0)
-    TaskSize = 1;
-
-  TaskGroup TG;
-  IndexTy I = Begin;
-  for (; I + TaskSize < End; I += TaskSize) {
-    TG.spawn([=, &Fn] {
-      for (IndexTy J = I, E = I + TaskSize; J != E; ++J)
-        Fn(J);
-    });
-  }
-  TG.spawn([=, &Fn] {
-    for (IndexTy J = I; J < End; ++J)
-      Fn(J);
-  });
-}
-
-#endif
-
-#endif
-
-template <typename Iter>
-using DefComparator =
-    std::less<typename std::iterator_traits<Iter>::value_type>;
-
-} // namespace detail
-
-// sequential algorithm implementations.
-template <class Policy, class RandomAccessIterator,
-          class Comparator = detail::DefComparator<RandomAccessIterator>>
-void sort(Policy policy, RandomAccessIterator Start, RandomAccessIterator End,
-          const Comparator &Comp = Comparator()) {
-  static_assert(is_execution_policy<Policy>::value,
-                "Invalid execution policy!");
-  std::sort(Start, End, Comp);
-}
-
-template <class Policy, class IterTy, class FuncTy>
-void for_each(Policy policy, IterTy Begin, IterTy End, FuncTy Fn) {
-  static_assert(is_execution_policy<Policy>::value,
-                "Invalid execution policy!");
-  std::for_each(Begin, End, Fn);
-}
-
-template <class Policy, class IndexTy, class FuncTy>
-void for_each_n(Policy policy, IndexTy Begin, IndexTy End, FuncTy Fn) {
-  static_assert(is_execution_policy<Policy>::value,
-                "Invalid execution policy!");
-  for (IndexTy I = Begin; I != End; ++I)
-    Fn(I);
-}
-
-// Parallel algorithm implementations, only available when LLVM_ENABLE_THREADS
-// is true.
-#if LLVM_ENABLE_THREADS
-template <class RandomAccessIterator,
-          class Comparator = detail::DefComparator<RandomAccessIterator>>
-void sort(parallel_execution_policy policy, RandomAccessIterator Start,
-          RandomAccessIterator End, const Comparator &Comp = Comparator()) {
-  detail::parallel_sort(Start, End, Comp);
-}
-
-template <class IterTy, class FuncTy>
-void for_each(parallel_execution_policy policy, IterTy Begin, IterTy End,
-              FuncTy Fn) {
-  detail::parallel_for_each(Begin, End, Fn);
-}
-
-template <class IndexTy, class FuncTy>
-void for_each_n(parallel_execution_policy policy, IndexTy Begin, IndexTy End,
-                FuncTy Fn) {
-  detail::parallel_for_each_n(Begin, End, Fn);
-}
-#endif
-
-} // namespace parallel
-} // End namespace lld
-
-#endif // LLD_CORE_PARALLEL_H

lld/trunk/include/lld/Core/TaskGroup.h

-//===- lld/Core/TaskGroup.h - Task Group ----------------------------------===//
-//
-//                             The LLVM Linker
-//
-// This file is distributed under the University of Illinois Open Source
-// License. See LICENSE.TXT for details.
-//
-//===----------------------------------------------------------------------===//
-
-#ifndef LLD_CORE_TASKGROUP_H
-#define LLD_CORE_TASKGROUP_H
-
-#include "lld/Core/LLVM.h"
-
-#include <condition_variable>
-#include <functional>
-#include <mutex>
-
-namespace lld {
-/// \brief Allows one or more threads to wait on a potentially unknown number of
-///   events.
-///
-/// A latch starts at \p count. inc() increments this, and dec() decrements it.
-/// All calls to sync() will block while the count is not 0.
-///
-/// Calling dec() on a Latch with a count of 0 has undefined behaivor.
-class Latch {
-  uint32_t Count;
-  mutable std::mutex Mutex;
-  mutable std::condition_variable Cond;
-
-public:
-  explicit Latch(uint32_t count = 0) : Count(count) {}
-  ~Latch() { sync(); }
-
-  void inc() {
-    std::unique_lock<std::mutex> lock(Mutex);
-    ++Count;
-  }
-
-  void dec() {
-    std::unique_lock<std::mutex> lock(Mutex);
-    if (--Count == 0)
-      Cond.notify_all();
-  }
-
-  void sync() const {
-    std::unique_lock<std::mutex> lock(Mutex);
-    Cond.wait(lock, [&] { return Count == 0; });
-  }
-};
-
-/// \brief Allows launching a number of tasks and waiting for them to finish
-///   either explicitly via sync() or implicitly on destruction.
-class TaskGroup {
-  Latch L;
-
-public:
-  void spawn(std::function<void()> F);
-
-  void sync() const { L.sync(); }
-};
-}
-
-#endif

lld/trunk/lib/Core/CMakeLists.txt

 if(NOT LLD_BUILT_STANDALONE)
   set(tablegen_deps intrinsics_gen)
 endif()
 
 add_lld_library(lldCore
   DefinedAtom.cpp
   Error.cpp
   File.cpp
   LinkingContext.cpp
   Reader.cpp
   Reproduce.cpp
   Resolver.cpp
   SymbolTable.cpp
   TargetOptionsCommandFlags.cpp
-  TaskGroup.cpp
   Writer.cpp
 
   ADDITIONAL_HEADER_DIRS
   ${LLD_INCLUDE_DIR}/lld/Core
 
   LINK_COMPONENTS
     MC
     Support
   
   LINK_LIBS
   ${LLVM_PTHREAD_LIB}
 
   DEPENDS
   ${tablegen_deps}
   )

lld/trunk/lib/Core/TaskGroup.cpp

-//===- lld/Core/TaskGroup.cpp - Task Group --------------------------------===//
-//
-//                             The LLVM Linker
-//
-// This file is distributed under the University of Illinois Open Source
-// License. See LICENSE.TXT for details.
-//
-//===----------------------------------------------------------------------===//
-
-#include "lld/Core/TaskGroup.h"
-#include "llvm/Config/llvm-config.h"
-
-#include <atomic>
-#include <stack>
-#include <thread>
-
-#if defined(_MSC_VER) && LLVM_ENABLE_THREADS
-#include <concrt.h>
-#include <ppl.h>
-#endif
-
-using namespace lld;
-
-namespace {
-
-/// \brief An abstract class that takes closures and runs them asynchronously.
-class Executor {
-public:
-  virtual ~Executor() = default;
-  virtual void add(std::function<void()> func) = 0;
-
-  static Executor *getDefaultExecutor();
-};
-
-#if !LLVM_ENABLE_THREADS
-class SyncExecutor : public Executor {
-public:
-  virtual void add(std::function<void()> F) { F(); }
-};
-
-Executor *Executor::getDefaultExecutor() {
-  static SyncExecutor Exec;
-  return &Exec;
-}
-
-#elif defined(_MSC_VER)
-/// \brief An Executor that runs tasks via ConcRT.
-class ConcRTExecutor : public Executor {
-  struct Taskish {
-    Taskish(std::function<void()> Task) : Task(Task) {}
-
-    std::function<void()> Task;
-
-    static void run(void *P) {
-      Taskish *Self = static_cast<Taskish *>(P);
-      Self->Task();
-      concurrency::Free(Self);
-    }
-  };
-
-public:
-  virtual void add(std::function<void()> F) {
-    Concurrency::CurrentScheduler::ScheduleTask(
-        Taskish::run, new (concurrency::Alloc(sizeof(Taskish))) Taskish(F));
-  }
-};
-
-Executor *Executor::getDefaultExecutor() {
-  static ConcRTExecutor exec;
-  return &exec;
-}
-
-#else
-/// \brief An implementation of an Executor that runs closures on a thread pool
-///   in filo order.
-class ThreadPoolExecutor : public Executor {
-public:
-  explicit ThreadPoolExecutor(
-      unsigned ThreadCount = std::thread::hardware_concurrency())
-      : Done(ThreadCount) {
-    // Spawn all but one of the threads in another thread as spawning threads
-    // can take a while.
-    std::thread([&, ThreadCount] {
-      for (size_t i = 1; i < ThreadCount; ++i) {
-        std::thread([=] { work(); }).detach();
-      }
-      work();
-    }).detach();
-  }
-
-  ~ThreadPoolExecutor() override {
-    std::unique_lock<std::mutex> Lock(Mutex);
-    Stop = true;
-    Lock.unlock();
-    Cond.notify_all();
-    // Wait for ~Latch.
-  }
-
-  void add(std::function<void()> F) override {
-    std::unique_lock<std::mutex> Lock(Mutex);
-    WorkStack.push(F);
-    Lock.unlock();
-    Cond.notify_one();
-  }
-
-private:
-  void work() {
-    while (true) {
-      std::unique_lock<std::mutex> Lock(Mutex);
-      Cond.wait(Lock, [&] { return Stop || !WorkStack.empty(); });
-      if (Stop)
-        break;
-      auto Task = WorkStack.top();
-      WorkStack.pop();
-      Lock.unlock();
-      Task();
-    }
-    Done.dec();
-  }
-
-  std::atomic<bool> Stop{false};
-  std::stack<std::function<void()>> WorkStack;
-  std::mutex Mutex;
-  std::condition_variable Cond;
-  Latch Done;
-};
-
-Executor *Executor::getDefaultExecutor() {
-  static ThreadPoolExecutor exec;
-  return &exec;
-}
-#endif
-}
-
-void TaskGroup::spawn(std::function<void()> F) {
-  L.inc();
-  Executor::getDefaultExecutor()->add([&, F] {
-    F();
-    L.dec();
-  });
-}

lld/trunk/lib/ReaderWriter/MachO/LayoutPass.cpp

 //===-- ReaderWriter/MachO/LayoutPass.cpp - Layout atoms ------------------===//
 //
 //                             The LLVM Linker
 //
 // This file is distributed under the University of Illinois Open Source
 // License. See LICENSE.TXT for details.
 //
 //===----------------------------------------------------------------------===//
 
 #include "LayoutPass.h"
 #include "lld/Core/Instrumentation.h"
-#include "lld/Core/Parallel.h"
 #include "lld/Core/PassManager.h"
 #include "lld/ReaderWriter/MachOLinkingContext.h"
 #include "llvm/ADT/STLExtras.h"
 #include "llvm/ADT/Twine.h"
 #include "llvm/Support/Debug.h"
+#include "llvm/Support/Parallel.h"
 #include <algorithm>
 #include <set>
 #include <utility>
 
 using namespace lld;
 
 #define DEBUG_TYPE "LayoutPass"
 
llvm::Error LayoutPass::perform(SimpleFile &mergedFile) {
   buildOrdinalOverrideMap(atomRange);
 
   DEBUG({
     llvm::dbgs() << "unsorted atoms:\n";
     printDefinedAtoms(atomRange);
   });
 
   std::vector<LayoutPass::SortKey> vec = decorate(atomRange);
-  sort(parallel::par, vec.begin(), vec.end(),
+  sort(llvm::parallel::par, vec.begin(), vec.end(),
        [&](const LayoutPass::SortKey &l, const LayoutPass::SortKey &r) -> bool {
          return compareAtoms(l, r, _customSorter);
        });
   DEBUG(checkTransitivity(vec, _customSorter));
   undecorate(atomRange, vec);
 
   DEBUG({
     llvm::dbgs() << "sorted atoms:\n";

lld/trunk/unittests/CMakeLists.txt

 add_custom_target(LLDUnitTests)
 set_target_properties(LLDUnitTests PROPERTIES FOLDER "lld tests")
 
 set(CMAKE_BUILD_WITH_INSTALL_RPATH OFF)
 
 # add_lld_unittest(test_dirname file1.cpp file2.cpp)
 #
 # Will compile the list of files together and link against lld
 # Produces a binary named 'basename(test_dirname)'.
 function(add_lld_unittest test_dirname)
   add_unittest(LLDUnitTests ${test_dirname} ${ARGN})
   target_link_libraries(${test_dirname} ${LLVM_COMMON_LIBS})
 endfunction()
 
-add_subdirectory(CoreTests)
 add_subdirectory(DriverTests)
 add_subdirectory(MachOTests)

llvm/trunk/include/llvm/Support/Parallel.h

+//===- llvm/Support/Parallel.h - Parallel algorithms ----------------------===//
+//
+//                     The LLVM Compiler Infrastructure
+//
+// This file is distributed under the University of Illinois Open Source
+// License. See LICENSE.TXT for details.
+//
+//===----------------------------------------------------------------------===//
+
+#ifndef LLVM_SUPPORT_PARALLEL_H
+#define LLVM_SUPPORT_PARALLEL_H
+
+#include "llvm/ADT/STLExtras.h"
+#include "llvm/Config/llvm-config.h"
+#include "llvm/Support/MathExtras.h"
+
+#include <algorithm>
+#include <condition_variable>
+#include <functional>
+#include <mutex>
+
+#if defined(_MSC_VER) && LLVM_ENABLE_THREADS
+#pragma warning(push)
+#pragma warning(disable : 4530)
+#include <concrt.h>
+#include <ppl.h>
+#pragma warning(pop)
+#endif
+
+namespace llvm {
+
+namespace detail {
+class Latch {
+  uint32_t Count;
+  mutable std::mutex Mutex;
+  mutable std::condition_variable Cond;
+
+public:
+  explicit Latch(uint32_t count = 0) : Count(Count) {}
+  ~Latch() { sync(); }
+
+  void inc() {
+    std::unique_lock<std::mutex> lock(Mutex);
+    ++Count;
+  }
+
+  void dec() {
+    std::unique_lock<std::mutex> lock(Mutex);
+    if (--Count == 0)
+      Cond.notify_all();
+  }
+
+  void sync() const {
+    std::unique_lock<std::mutex> lock(Mutex);
+    Cond.wait(lock, [&] { return Count == 0; });
+  }
+};
+
+class TaskGroup {
+  Latch L;
+
+public:
+  void spawn(std::function<void()> f);
+
+  void sync() const { L.sync(); }
+};
+}
+
+namespace parallel {
+struct sequential_execution_policy {};
+struct parallel_execution_policy {};
+
+template <typename T>
+struct is_execution_policy
+    : public std::integral_constant<
+          bool, llvm::is_one_of<T, sequential_execution_policy,
+                                parallel_execution_policy>::value> {};
+
+constexpr sequential_execution_policy seq{};
+constexpr parallel_execution_policy par{};
+
+namespace detail {
+
+#if LLVM_ENABLE_THREADS
+
+#if defined(_MSC_VER)
+template <class RandomAccessIterator, class Comparator>
+void parallel_sort(RandomAccessIterator Start, RandomAccessIterator End,
+                   const Comparator &Comp) {
+  concurrency::parallel_sort(Start, End, Comp);
+}
+template <class IterTy, class FuncTy>
+void parallel_for_each(IterTy Begin, IterTy End, FuncTy Fn) {
+  concurrency::parallel_for_each(Begin, End, Fn);
+}
+
+template <class IndexTy, class FuncTy>
+void parallel_for_each_n(IndexTy Begin, IndexTy End, FuncTy Fn) {
+  concurrency::parallel_for(Begin, End, Fn);
+}
+
+#else
+const ptrdiff_t MinParallelSize = 1024;
+
+/// \brief Inclusive median.
+template <class RandomAccessIterator, class Comparator>
+RandomAccessIterator medianOf3(RandomAccessIterator Start,
+                               RandomAccessIterator End,
+                               const Comparator &Comp) {
+  RandomAccessIterator Mid = Start + (std::distance(Start, End) / 2);
+  return Comp(*Start, *(End - 1))
+             ? (Comp(*Mid, *(End - 1)) ? (Comp(*Start, *Mid) ? Mid : Start)
+                                       : End - 1)
+             : (Comp(*Mid, *Start) ? (Comp(*(End - 1), *Mid) ? Mid : End - 1)
+                                   : Start);
+}
+
+template <class RandomAccessIterator, class Comparator>
+void parallel_quick_sort(RandomAccessIterator Start, RandomAccessIterator End,
+                         const Comparator &Comp, TaskGroup &TG, size_t Depth) {
+  // Do a sequential sort for small inputs.
+  if (std::distance(Start, End) < detail::MinParallelSize || Depth == 0) {
+    std::sort(Start, End, Comp);
+    return;
+  }
+
+  // Partition.
+  auto Pivot = medianOf3(Start, End, Comp);
+  // Move Pivot to End.
+  std::swap(*(End - 1), *Pivot);
+  Pivot = std::partition(Start, End - 1, [&Comp, End](decltype(*Start) V) {
+    return Comp(V, *(End - 1));
+  });
+  // Move Pivot to middle of partition.
+  std::swap(*Pivot, *(End - 1));
+
+  // Recurse.
+  TG.spawn([=, &Comp, &TG] {
+    parallel_quick_sort(Start, Pivot, Comp, TG, Depth - 1);
+  });
+  parallel_quick_sort(Pivot + 1, End, Comp, TG, Depth - 1);
+}
+
+template <class RandomAccessIterator, class Comparator>
+void parallel_sort(RandomAccessIterator Start, RandomAccessIterator End,
+                   const Comparator &Comp) {
+  TaskGroup TG;
+  parallel_quick_sort(Start, End, Comp, TG,
+                      llvm::Log2_64(std::distance(Start, End)) + 1);
+}
+
+template <class IterTy, class FuncTy>
+void parallel_for_each(IterTy Begin, IterTy End, FuncTy Fn) {
+  // TaskGroup has a relatively high overhead, so we want to reduce
+  // the number of spawn() calls. We'll create up to 1024 tasks here.
+  // (Note that 1024 is an arbitrary number. This code probably needs
+  // improving to take the number of available cores into account.)
+  ptrdiff_t TaskSize = std::distance(Begin, End) / 1024;
+  if (TaskSize == 0)
+    TaskSize = 1;
+
+  TaskGroup TG;
+  while (TaskSize <= std::distance(Begin, End)) {
+    TG.spawn([=, &Fn] { std::for_each(Begin, Begin + TaskSize, Fn); });
+    Begin += TaskSize;
+  }
+  TG.spawn([=, &Fn] { std::for_each(Begin, End, Fn); });
+}
+
+template <class IndexTy, class FuncTy>
+void parallel_for_each_n(IndexTy Begin, IndexTy End, FuncTy Fn) {
+  ptrdiff_t TaskSize = (End - Begin) / 1024;
+  if (TaskSize == 0)
+    TaskSize = 1;
+
+  TaskGroup TG;
+  IndexTy I = Begin;
+  for (; I + TaskSize < End; I += TaskSize) {
+    TG.spawn([=, &Fn] {
+      for (IndexTy J = I, E = I + TaskSize; J != E; ++J)
+        Fn(J);
+    });
+  }
+  TG.spawn([=, &Fn] {
+    for (IndexTy J = I; J < End; ++J)
+      Fn(J);
+  });
+}
+
+#endif
+
+#endif
+
+template <typename Iter>
+using DefComparator =
+    std::less<typename std::iterator_traits<Iter>::value_type>;
+
+} // namespace detail
+
+// sequential algorithm implementations.
+template <class Policy, class RandomAccessIterator,
+          class Comparator = detail::DefComparator<RandomAccessIterator>>
+void sort(Policy policy, RandomAccessIterator Start, RandomAccessIterator End,
+          const Comparator &Comp = Comparator()) {
+  static_assert(is_execution_policy<Policy>::value,
+                "Invalid execution policy!");
+  std::sort(Start, End, Comp);
+}
+
+template <class Policy, class IterTy, class FuncTy>
+void for_each(Policy policy, IterTy Begin, IterTy End, FuncTy Fn) {
+  static_assert(is_execution_policy<Policy>::value,
+                "Invalid execution policy!");
+  std::for_each(Begin, End, Fn);
+}
+
+template <class Policy, class IndexTy, class FuncTy>
+void for_each_n(Policy policy, IndexTy Begin, IndexTy End, FuncTy Fn) {
+  static_assert(is_execution_policy<Policy>::value,
+                "Invalid execution policy!");
+  for (IndexTy I = Begin; I != End; ++I)
+    Fn(I);
+}
+
+// Parallel algorithm implementations, only available when LLVM_ENABLE_THREADS
+// is true.
+#if LLVM_ENABLE_THREADS
+template <class RandomAccessIterator,
+          class Comparator = detail::DefComparator<RandomAccessIterator>>
+void sort(parallel_execution_policy policy, RandomAccessIterator Start,
+          RandomAccessIterator End, const Comparator &Comp = Comparator()) {
+  detail::parallel_sort(Start, End, Comp);
+}
+
+template <class IterTy, class FuncTy>
+void for_each(parallel_execution_policy policy, IterTy Begin, IterTy End,
+              FuncTy Fn) {
+  detail::parallel_for_each(Begin, End, Fn);
+}
+
+template <class IndexTy, class FuncTy>
+void for_each_n(parallel_execution_policy policy, IndexTy Begin, IndexTy End,
+                FuncTy Fn) {
+  detail::parallel_for_each_n(Begin, End, Fn);
+}
+#endif
+
+} // namespace parallel
+} // namespace llvm
+
+#endif // LLVM_SUPPORT_PARALLEL_H

llvm/trunk/lib/Support/CMakeLists.txt

add_llvm_library(LLVMSupport
   LockFileManager.cpp
   LowLevelType.cpp
   ManagedStatic.cpp
   MathExtras.cpp
   MemoryBuffer.cpp
   MD5.cpp
   NativeFormatting.cpp
   Options.cpp
+  Parallel.cpp
   PluginLoader.cpp
   PrettyStackTrace.cpp
   RandomNumberGenerator.cpp
   Regex.cpp
   ScaledNumber.cpp
   ScopedPrinter.cpp
   SHA1.cpp
   SmallPtrSet.cpp

llvm/trunk/lib/Support/Parallel.cpp

+//===- llvm/Support/Parallel.cpp - Parallel algorithms --------------------===//
+//
+//                     The LLVM Compiler Infrastructure
+//
+// This file is distributed under the University of Illinois Open Source
+// License. See LICENSE.TXT for details.
+//
+//===----------------------------------------------------------------------===//
+
+#include "llvm/Support/Parallel.h"
+#include "llvm/Config/llvm-config.h"
+
+#include <atomic>
+#include <stack>
+#include <thread>
+
+using namespace llvm;
+
+namespace {
+
+/// \brief An abstract class that takes closures and runs them asynchronously.
+class Executor {
+public:
+  virtual ~Executor() = default;
+  virtual void add(std::function<void()> func) = 0;
+
+  static Executor *getDefaultExecutor();
+};
+
+#if !LLVM_ENABLE_THREADS
+class SyncExecutor : public Executor {
+public:
+  virtual void add(std::function<void()> F) { F(); }
+};
+
+Executor *Executor::getDefaultExecutor() {
+  static SyncExecutor Exec;
+  return &Exec;
+}
+
+#elif defined(_MSC_VER)
+/// \brief An Executor that runs tasks via ConcRT.
+class ConcRTExecutor : public Executor {
+  struct Taskish {
+    Taskish(std::function<void()> Task) : Task(Task) {}
+
+    std::function<void()> Task;
+
+    static void run(void *P) {
+      Taskish *Self = static_cast<Taskish *>(P);
+      Self->Task();
+      concurrency::Free(Self);
+    }
+  };
+
+public:
+  virtual void add(std::function<void()> F) {
+    Concurrency::CurrentScheduler::ScheduleTask(
+        Taskish::run, new (concurrency::Alloc(sizeof(Taskish))) Taskish(F));
+  }
+};
+
+Executor *Executor::getDefaultExecutor() {
+  static ConcRTExecutor exec;
+  return &exec;
+}
+
+#else
+/// \brief An implementation of an Executor that runs closures on a thread pool
+///   in filo order.
+class ThreadPoolExecutor : public Executor {
+public:
+  explicit ThreadPoolExecutor(
+      unsigned ThreadCount = std::thread::hardware_concurrency())
+      : Done(ThreadCount) {
+    // Spawn all but one of the threads in another thread as spawning threads
+    // can take a while.
+    std::thread([&, ThreadCount] {
+      for (size_t i = 1; i < ThreadCount; ++i) {
+        std::thread([=] { work(); }).detach();
+      }
+      work();
+    }).detach();
+  }
+
+  ~ThreadPoolExecutor() override {
+    std::unique_lock<std::mutex> Lock(Mutex);
+    Stop = true;
+    Lock.unlock();
+    Cond.notify_all();
+    // Wait for ~Latch.
+  }
+
+  void add(std::function<void()> F) override {
+    std::unique_lock<std::mutex> Lock(Mutex);
+    WorkStack.push(F);
+    Lock.unlock();
+    Cond.notify_one();
+  }
+
+private:
+  void work() {
+    while (true) {
+      std::unique_lock<std::mutex> Lock(Mutex);
+      Cond.wait(Lock, [&] { return Stop || !WorkStack.empty(); });
+      if (Stop)
+        break;
+      auto Task = WorkStack.top();
+      WorkStack.pop();
+      Lock.unlock();
+      Task();
+    }
+    Done.dec();
+  }
+
+  std::atomic<bool> Stop{false};
+  std::stack<std::function<void()>> WorkStack;
+  std::mutex Mutex;
+  std::condition_variable Cond;
+  Latch Done;
+};
+
+Executor *Executor::getDefaultExecutor() {
+  static ThreadPoolExecutor exec;
+  return &exec;
+}
+#endif
+}
+
+void detail::TaskGroup::spawn(std::function<void()> F) {
+  L.inc();
+  Executor::getDefaultExecutor()->add([&, F] {
+    F();
+    L.dec();
+  });
+}

llvm/trunk/unittests/Support/CMakeLists.txt

add_llvm_unittest(SupportTests
   LineIteratorTest.cpp
   LockFileManagerTest.cpp
   MD5Test.cpp
   ManagedStatic.cpp
   MathExtrasTest.cpp
   MemoryBufferTest.cpp
   MemoryTest.cpp
   NativeFormatTests.cpp
+  ParallelTest.cpp
   Path.cpp
   ProcessTest.cpp
   ProgramTest.cpp
   RegexTest.cpp
   ReplaceFileTest.cpp
   ScaledNumberTest.cpp
   SourceMgrTest.cpp
   SpecialCaseListTest.cpp

llvm/trunk/unittests/Support/ParallelTest.cpp

+//===- llvm/unittest/Support/ParallelTest.cpp -----------------------------===//
+//
+//                     The LLVM Compiler Infrastructure
+//
+// This file is distributed under the University of Illinois Open Source
+// License. See LICENSE.TXT for details.
+//
+//===----------------------------------------------------------------------===//
+///
+/// \file
+/// \brief Parallel.h unit tests.
+///
+//===----------------------------------------------------------------------===//
+
+#include "llvm/Support/Parallel.h"
+#include "gtest/gtest.h"
+#include <array>
+#include <random>
+
+uint32_t array[1024 * 1024];
+
+using namespace llvm;
+
+TEST(Parallel, sort) {
+  std::mt19937 randEngine;
+  std::uniform_int_distribution<uint32_t> dist;
+
+  for (auto &i : array)
+    i = dist(randEngine);
+
+  sort(parallel::par, std::begin(array), std::end(array));
+  ASSERT_TRUE(std::is_sorted(std::begin(array), std::end(array)));
+}
+
+TEST(Parallel, parallel_for) {
+  // We need to test the case with a TaskSize > 1. We are white-box testing
+  // here. The TaskSize is calculated as (End - Begin) / 1024 at the time of
+  // writing.
+  uint32_t range[2050];
+  std::fill(range, range + 2050, 1);
+  for_each_n(parallel::par, 0, 2049, [&range](size_t I) { ++range[I]; });
+
+  uint32_t expected[2049];
+  std::fill(expected, expected + 2049, 2);
+  ASSERT_TRUE(std::equal(range, range + 2049, expected));
+  // Check that we don't write past the end of the requested range.
+  ASSERT_EQ(range[2049], 1u);
+}