[OpenMP] Introduce target memory manager

Target memory manager is introduced in this patch which aims to manage target
memory such that they will not be freed immediately when they are not used
because the overhead of memory allocation and free is very large. For CUDA
device, cuMemFree even blocks the context switch on device which affects
concurrent kernel execution.

The memory manager can be taken as a memory pool. It divides the pool into
multiple buckets according to the size such that memory allocation/free
distributed to different buckets will not affect each other.

In this version, we use the exact-equality policy to find a free buffer. This
is an open question: will best-fit work better here? IMO, best-fit is not good
for target memory management because computation on GPU usually requires GBs of
data. Best-fit might lead to a serious waste. For example, there is a free
buffer of size 1960MB, and now we need a buffer of size 1200MB. If best-fit,
the free buffer will be returned, leading to a 760MB waste.

The allocation will happen when there is no free memory left, and the memory
free on device will take place in the following two cases:
1. The program ends. Obviously. However, there is a little problem that plugin
library is destroyed before the memory manager is destroyed, leading to a fact
that the call to target plugin will not succeed.
2. Device is out of memory when we request a new memory. The manager will walk
through all free buffers from the bucket with largest base size, pick up one
buffer, free it, and try to allocate immediately. If it succeeds, it will
return right away rather than freeing all buffers in free list.

Update:
A threshold (8KB by default) is set such that users could control what size of memory
will be managed by the manager. It can also be configured by an environment variable
`LIBOMPTARGET_MEMORY_MANAGER_THRESHOLD`.

Reviewed By: jdoerfert, ye-luo, JonChesterfield

Differential Revision: https://reviews.llvm.org/D81054

Author: shiltian

openmp/libomptarget/src/CMakeLists.txt

@@ -1,9 +1,9 @@
 ##===----------------------------------------------------------------------===##
-# 
+#
 # Part of the LLVM Project, under the Apache License v2.0 with LLVM Exceptions.
 # See https://llvm.org/LICENSE.txt for license information.
 # SPDX-License-Identifier: Apache-2.0 WITH LLVM-exception
-# 
+#
 ##===----------------------------------------------------------------------===##
 #
 # Build offloading library libomptarget.so.
@@ -16,6 +16,7 @@ set(src_files
   api.cpp
   device.cpp
   interface.cpp
+  MemoryManager.cpp
   rtl.cpp
   omptarget.cpp
 )

openmp/libomptarget/src/MemoryManager.cpp

@@ -0,0 +1,256 @@
+//===----------- MemoryManager.cpp - Target independent memory manager ----===//
+//
+// Part of the LLVM Project, under the Apache License v2.0 with LLVM Exceptions.
+// See https://llvm.org/LICENSE.txt for license information.
+// SPDX-License-Identifier: Apache-2.0 WITH LLVM-exception
+//
+//===----------------------------------------------------------------------===//
+//
+// Functionality for managing target memory.
+// It is very expensive to call alloc/free functions of target devices. The
+// MemoryManagerTy in this file is to reduce the number of invocations of those
+// functions by buffering allocated device memory. In this way, when a memory is
+// not used, it will not be freed on the device directly. The buffer is
+// organized in a number of buckets for efficient look up. A memory will go to
+// corresponding bucket based on its size. When a new memory request comes in,
+// it will first check whether there is free memory of same size. If yes,
+// returns it directly. Otherwise, allocate one on device.
+//
+// It also provides a way to opt out the memory manager. Memory
+// allocation/deallocation will only be managed if the requested size is less
+// than SizeThreshold, which can be configured via an environment variable
+// LIBOMPTARGET_MEMORY_MANAGER_THRESHOLD.
+//
+//===----------------------------------------------------------------------===//
+
+#include "MemoryManager.h"
+#include "device.h"
+#include "private.h"
+#include "rtl.h"
+
+namespace {
+constexpr const size_t BucketSize[] = {
+    0,       1U << 2, 1U << 3,  1U << 4,  1U << 5,  1U << 6, 1U << 7,
+    1U << 8, 1U << 9, 1U << 10, 1U << 11, 1U << 12, 1U << 13};
+
+constexpr const int NumBuckets = sizeof(BucketSize) / sizeof(BucketSize[0]);
+
+/// The threshold to manage memory using memory manager. If the request size is
+/// larger than \p SizeThreshold, the allocation will not be managed by the
+/// memory manager. This variable can be configured via an env \p
+/// LIBOMPTARGET_MEMORY_MANAGER_THRESHOLD. By default, the value is 8KB.
+size_t SizeThreshold = 1U << 13;
+
+/// Find the previous number that is power of 2 given a number that is not power
+/// of 2.
+size_t floorToPowerOfTwo(size_t Num) {
+  Num |= Num >> 1;
+  Num |= Num >> 2;
+  Num |= Num >> 4;
+  Num |= Num >> 8;
+  Num |= Num >> 16;
+  Num |= Num >> 32;
+  Num += 1;
+  return Num >> 1;
+}
+
+/// Find a suitable bucket
+int findBucket(size_t Size) {
+  const size_t F = floorToPowerOfTwo(Size);
+
+  DP("findBucket: Size %zu is floored to %zu.\n", Size, F);
+
+  int L = 0, H = NumBuckets - 1;
+  while (H - L > 1) {
+    int M = (L + H) >> 1;
+    if (BucketSize[M] == F)
+      return M;
+    if (BucketSize[M] > F)
+      H = M - 1;
+    else
+      L = M;
+  }
+
+  assert(L >= 0 && L < NumBuckets && "L is out of range");
+
+  DP("findBucket: Size %zu goes to bucket %d\n", Size, L);
+
+  return L;
+}
+} // namespace
+
+MemoryManagerTy::MemoryManagerTy(DeviceTy &Dev, size_t Threshold)
+    : FreeLists(NumBuckets), FreeListLocks(NumBuckets), Device(Dev) {
+  if (Threshold)
+    SizeThreshold = Threshold;
+}
+
+MemoryManagerTy::~MemoryManagerTy() {
+  // TODO: There is a little issue that target plugin is destroyed before this
+  // object, therefore the memory free will not succeed.
+  // Deallocate all memory in map
+  for (auto Itr = PtrToNodeTable.begin(); Itr != PtrToNodeTable.end(); ++Itr) {
+    assert(Itr->second.Ptr && "nullptr in map table");
+    deleteOnDevice(Itr->second.Ptr);
+  }
+}
+
+void *MemoryManagerTy::allocateOnDevice(size_t Size, void *HstPtr) const {
+  return Device.RTL->data_alloc(Device.RTLDeviceID, Size, HstPtr);
+}
+
+int MemoryManagerTy::deleteOnDevice(void *Ptr) const {
+  return Device.RTL->data_delete(Device.RTLDeviceID, Ptr);
+}
+
+void *MemoryManagerTy::freeAndAllocate(size_t Size, void *HstPtr) {
+  std::vector<void *> RemoveList;
+
+  // Deallocate all memory in FreeList
+  for (int I = 0; I < NumBuckets; ++I) {
+    FreeListTy &List = FreeLists[I];
+    std::lock_guard<std::mutex> Lock(FreeListLocks[I]);
+    if (List.empty())
+      continue;
+    for (const NodeTy &N : List) {
+      deleteOnDevice(N.Ptr);
+      RemoveList.push_back(N.Ptr);
+    }
+    FreeLists[I].clear();
+  }
+
+  // Remove all nodes in the map table which have been released
+  if (!RemoveList.empty()) {
+    std::lock_guard<std::mutex> LG(MapTableLock);
+    for (void *P : RemoveList)
+      PtrToNodeTable.erase(P);
+  }
+
+  // Try allocate memory again
+  return allocateOnDevice(Size, HstPtr);
+}
+
+void *MemoryManagerTy::allocateOrFreeAndAllocateOnDevice(size_t Size,
+                                                         void *HstPtr) {
+  void *TgtPtr = allocateOnDevice(Size, HstPtr);
+  // We cannot get memory from the device. It might be due to OOM. Let's
+  // free all memory in FreeLists and try again.
+  if (TgtPtr == nullptr) {
+    DP("Failed to get memory on device. Free all memory in FreeLists and "
+       "try again.\n");
+    TgtPtr = freeAndAllocate(Size, HstPtr);
+  }
+
+#ifdef OMPTARGET_DEBUG
+  if (TgtPtr == nullptr)
+    DP("Still cannot get memory on device probably because the device is "
+       "OOM.\n");
+#endif
+
+  return TgtPtr;
+}
+
+void *MemoryManagerTy::allocate(size_t Size, void *HstPtr) {
+  // If the size is zero, we will not bother the target device. Just return
+  // nullptr directly.
+  if (Size == 0)
+    return nullptr;
+
+  DP("MemoryManagerTy::allocate: size %zu with host pointer " DPxMOD ".\n",
+     Size, DPxPTR(HstPtr));
+
+  // If the size is greater than the threshold, allocate it directly from
+  // device.
+  if (Size > SizeThreshold) {
+    DP("%zu is greater than the threshold %zu. Allocate it directly from "
+       "device\n",
+       Size, SizeThreshold);
+    void *TgtPtr = allocateOrFreeAndAllocateOnDevice(Size, HstPtr);
+
+    DP("Got target pointer " DPxMOD ". Return directly.\n", DPxPTR(TgtPtr));
+
+    return TgtPtr;
+  }
+
+  NodeTy *NodePtr = nullptr;
+
+  // Try to get a node from FreeList
+  {
+    const int B = findBucket(Size);
+    FreeListTy &List = FreeLists[B];
+
+    NodeTy TempNode(Size, nullptr);
+    std::lock_guard<std::mutex> LG(FreeListLocks[B]);
+    FreeListTy::const_iterator Itr = List.find(TempNode);
+
+    if (Itr != List.end()) {
+      NodePtr = &Itr->get();
+      List.erase(Itr);
+    }
+  }
+
+#ifdef OMPTARGET_DEBUG
+  if (NodePtr != nullptr)
+    DP("Find one node " DPxMOD " in the bucket.\n", DPxPTR(NodePtr));
+#endif
+
+  // We cannot find a valid node in FreeLists. Let's allocate on device and
+  // create a node for it.
+  if (NodePtr == nullptr) {
+    DP("Cannot find a node in the FreeLists. Allocate on device.\n");
+    // Allocate one on device
+    void *TgtPtr = allocateOrFreeAndAllocateOnDevice(Size, HstPtr);
+
+    if (TgtPtr == nullptr)
+      return nullptr;
+
+    // Create a new node and add it into the map table
+    {
+      std::lock_guard<std::mutex> Guard(MapTableLock);
+      auto Itr = PtrToNodeTable.emplace(TgtPtr, NodeTy(Size, TgtPtr));
+      NodePtr = &Itr.first->second;
+    }
+
+    DP("Node address " DPxMOD ", target pointer " DPxMOD ", size %zu\n",
+       DPxPTR(NodePtr), DPxPTR(TgtPtr), Size);
+  }
+
+  assert(NodePtr && "NodePtr should not be nullptr at this point");
+
+  return NodePtr->Ptr;
+}
+
+int MemoryManagerTy::free(void *TgtPtr) {
+  DP("MemoryManagerTy::free: target memory " DPxMOD ".\n", DPxPTR(TgtPtr));
+
+  NodeTy *P = nullptr;
+
+  // Look it up into the table
+  {
+    std::lock_guard<std::mutex> G(MapTableLock);
+    auto Itr = PtrToNodeTable.find(TgtPtr);
+
+    // We don't remove the node from the map table because the map does not
+    // change.
+    if (Itr != PtrToNodeTable.end())
+      P = &Itr->second;
+  }
+
+  // The memory is not managed by the manager
+  if (P == nullptr) {
+    DP("Cannot find its node. Delete it on device directly.\n");
+    return deleteOnDevice(TgtPtr);
+  }
+
+  // Insert the node to the free list
+  const int B = findBucket(P->Size);
+
+  DP("Found its node " DPxMOD ". Insert it to bucket %d.\n", DPxPTR(P), B);
+
+  {
+    std::lock_guard<std::mutex> G(FreeListLocks[B]);
+    FreeLists[B].insert(*P);
+  }
+
+  return OFFLOAD_SUCCESS;
+}

openmp/libomptarget/src/MemoryManager.h

@@ -0,0 +1,95 @@
+//===----------- MemoryManager.h - Target independent memory manager ------===//
+//
+// Part of the LLVM Project, under the Apache License v2.0 with LLVM Exceptions.
+// See https://llvm.org/LICENSE.txt for license information.
+// SPDX-License-Identifier: Apache-2.0 WITH LLVM-exception
+//
+//===----------------------------------------------------------------------===//
+//
+// Declarations for target independent memory manager.
+//
+//===----------------------------------------------------------------------===//
+
+#ifndef LLVM_OPENMP_LIBOMPTARGET_SRC_MEMORYMANAGER_H
+#define LLVM_OPENMP_LIBOMPTARGET_SRC_MEMORYMANAGER_H
+
+#include <cassert>
+#include <functional>
+#include <list>
+#include <mutex>
+#include <set>
+#include <unordered_map>
+#include <vector>
+
+// Forward declaration
+struct DeviceTy;
+
+class MemoryManagerTy {
+  /// A structure stores the meta data of a target pointer
+  struct NodeTy {
+    /// Memory size
+    const size_t Size;
+    /// Target pointer
+    void *Ptr;
+
+    /// Constructor
+    NodeTy(size_t Size, void *Ptr) : Size(Size), Ptr(Ptr) {}
+  };
+
+  /// To make \p NodePtrTy ordered when they're put into \p std::multiset.
+  struct NodeCmpTy {
+    bool operator()(const NodeTy &LHS, const NodeTy &RHS) const {
+      return LHS.Size < RHS.Size;
+    }
+  };
+
+  /// A \p FreeList is a set of Nodes. We're using \p std::multiset here to make
+  /// the look up procedure more efficient.
+  using FreeListTy = std::multiset<std::reference_wrapper<NodeTy>, NodeCmpTy>;
+
+  /// A list of \p FreeListTy entries, each of which is a \p std::multiset of
+  /// Nodes whose size is less or equal to a specific bucket size.
+  std::vector<FreeListTy> FreeLists;
+  /// A list of mutex for each \p FreeListTy entry
+  std::vector<std::mutex> FreeListLocks;
+  /// A table to map from a target pointer to its node
+  std::unordered_map<void *, NodeTy> PtrToNodeTable;
+  /// The mutex for the table \p PtrToNodeTable
+  std::mutex MapTableLock;
+  /// A reference to its corresponding \p DeviceTy object
+  DeviceTy &Device;
+
+  /// Request memory from target device
+  void *allocateOnDevice(size_t Size, void *HstPtr) const;
+
+  /// Deallocate data on device
+  int deleteOnDevice(void *Ptr) const;
+
+  /// This function is called when it tries to allocate memory on device but the
+  /// device returns out of memory. It will first free all memory in the
+  /// FreeList and try to allocate again.
+  void *freeAndAllocate(size_t Size, void *HstPtr);
+
+  /// The goal is to allocate memory on the device. It first tries to allocate
+  /// directly on the device. If a \p nullptr is returned, it might be because
+  /// the device is OOM. In that case, it will free all unused memory and then
+  /// try again.
+  void *allocateOrFreeAndAllocateOnDevice(size_t Size, void *HstPtr);
+
+public:
+  /// Constructor. If \p Threshold is non-zero, then the default threshold will
+  /// be overwritten by \p Threshold.
+  MemoryManagerTy(DeviceTy &Dev, size_t Threshold = 0);
+
+  /// Destructor
+  ~MemoryManagerTy();
+
+  /// Allocate memory of size \p Size from target device. \p HstPtr is used to
+  /// assist the allocation.
+  void *allocate(size_t Size, void *HstPtr);
+
+  /// Deallocate memory pointed by \p TgtPtr
+  int free(void *TgtPtr);
+};
+
+#endif // LLVM_OPENMP_LIBOMPTARGET_SRC_MEMORYMANAGER_H