/*
 *  Copyright 2008-2013 NVIDIA Corporation
 *  Modifications Copyright© 2019-2025 Advanced Micro Devices, Inc. All rights reserved.
 *
 *  Licensed under the Apache License, Version 2.0 (the "License");
 *  you may not use this file except in compliance with the License.
 *  You may obtain a copy of the License at
 *
 *      http://www.apache.org/licenses/LICENSE-2.0
 *
 *  Unless required by applicable law or agreed to in writing, software
 *  distributed under the License is distributed on an "AS IS" BASIS,
 *  WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
 *  See the License for the specific language governing permissions and
 *  limitations under the License.
 */

#define THRUST_ENABLE_FUTURE_RAW_DATA_MEMBER

#include <thrust/detail/config.h>

// need to suppress deprecation warnings inside several thrust headers
THRUST_SUPPRESS_DEPRECATED_PUSH

#if THRUST_CPP_DIALECT >= 2017

#  include <thrust/async/copy.h>
#  include <thrust/async/reduce.h>
#  include <thrust/device_vector.h>
#  include <thrust/host_vector.h>

#  include "test_param_fixtures.hpp"
#  include "test_real_assertions.hpp"
#  include "test_utils.hpp"

TESTS_DEFINE(AsyncReduceTests, NumericalTestsParams);

template <typename T>
struct custom_plus
{
  THRUST_HOST_DEVICE T operator()(T lhs, T rhs) const
  {
    return lhs + rhs;
  }
};

#  define DEFINE_STATEFUL_ASYNC_REDUCE_INVOKER(NAME, MEMBERS, CTOR, DTOR, VALIDATE, ...) \
    template <typename T>                                                                \
    struct NAME                                                                          \
    {                                                                                    \
      MEMBERS                                                                            \
                                                                                         \
      NAME()                                                                             \
      {                                                                                  \
        CTOR                                                                             \
      }                                                                                  \
                                                                                         \
      ~NAME()                                                                            \
      {                                                                                  \
        DTOR                                                                             \
      }                                                                                  \
                                                                                         \
      template <typename Event>                                                          \
      void validate_event(Event& e)                                                      \
      {                                                                                  \
        THRUST_UNUSED_VAR(e);                                                            \
        VALIDATE                                                                         \
      }                                                                                  \
                                                                                         \
      template <typename ForwardIt, typename Sentinel>                                   \
      THRUST_HOST auto operator()(ForwardIt&& first, Sentinel&& last)                    \
        THRUST_DECLTYPE_RETURNS(::thrust::async::reduce(__VA_ARGS__))                    \
    };                                                                                   \
    /**/

#  define DEFINE_ASYNC_REDUCE_INVOKER(NAME, ...)                                                     \
    DEFINE_STATEFUL_ASYNC_REDUCE_INVOKER(                                                            \
      NAME, THRUST_PP_EMPTY(), THRUST_PP_EMPTY(), THRUST_PP_EMPTY(), THRUST_PP_EMPTY(), __VA_ARGS__) \
    /**/

#  define DEFINE_SYNC_REDUCE_INVOKER(NAME, ...)                                                                     \
    template <typename T>                                                                                           \
    struct NAME                                                                                                     \
    {                                                                                                               \
      template <typename ForwardIt, typename Sentinel>                                                              \
      THRUST_HOST auto operator()(ForwardIt&& first, Sentinel&& last) THRUST_RETURNS(::thrust::reduce(__VA_ARGS__)) \
    };                                                                                                              \
    /**/

DEFINE_ASYNC_REDUCE_INVOKER(reduce_async_invoker, THRUST_FWD(first), THRUST_FWD(last));
DEFINE_ASYNC_REDUCE_INVOKER(reduce_async_invoker_device, thrust::device, THRUST_FWD(first), THRUST_FWD(last));
DEFINE_ASYNC_REDUCE_INVOKER(reduce_async_invoker_device_allocator,
                            thrust::device(thrust::device_allocator<void>{}),
                            THRUST_FWD(first),
                            THRUST_FWD(last));
DEFINE_STATEFUL_ASYNC_REDUCE_INVOKER(
  reduce_async_invoker_device_on
  // Members.
  ,
  SPECIALIZE_DEVICE_RESOURCE_NAME(Stream_t) stream_;
  // Constructor.
  ,
  thrust::THRUST_DEVICE_BACKEND_DETAIL::throw_on_error(SPECIALIZE_DEVICE_RESOURCE_NAME(StreamCreateWithFlags)(
    &stream_, SPECIALIZE_DEVICE_RESOURCE_NAME(StreamNonBlocking)));
  // Destructor.
  ,
  thrust::THRUST_DEVICE_BACKEND_DETAIL::throw_on_error(SPECIALIZE_DEVICE_RESOURCE_NAME(StreamDestroy)(stream_));
  // `validate_event` member.
  ,
  ASSERT_EQ_QUIET(stream_, e.stream().native_handle());
  // Arguments to `thrust::async::reduce`.
  ,
  thrust::device.on(stream_),
  THRUST_FWD(first),
  THRUST_FWD(last));
DEFINE_STATEFUL_ASYNC_REDUCE_INVOKER(
  reduce_async_invoker_device_allocator_on
  // Members.
  ,
  SPECIALIZE_DEVICE_RESOURCE_NAME(Stream_t) stream_;
  // Constructor.
  ,
  thrust::THRUST_DEVICE_BACKEND_DETAIL::throw_on_error(SPECIALIZE_DEVICE_RESOURCE_NAME(StreamCreateWithFlags)(
    &stream_, SPECIALIZE_DEVICE_RESOURCE_NAME(StreamNonBlocking)));
  // Destructor.
  ,
  thrust::THRUST_DEVICE_BACKEND_DETAIL::throw_on_error(SPECIALIZE_DEVICE_RESOURCE_NAME(StreamDestroy)(stream_));
  // `validate_event` member.
  ,
  ASSERT_EQ_QUIET(stream_, e.stream().native_handle());
  // Arguments to `thrust::async::reduce`.
  ,
  thrust::device(thrust::device_allocator<void>{}).on(stream_),
  THRUST_FWD(first),
  THRUST_FWD(last));

DEFINE_SYNC_REDUCE_INVOKER(reduce_sync_invoker, THRUST_FWD(first), THRUST_FWD(last));

DEFINE_ASYNC_REDUCE_INVOKER(reduce_async_invoker_init, THRUST_FWD(first), THRUST_FWD(last), random_integer<T>());
DEFINE_ASYNC_REDUCE_INVOKER(
  reduce_async_invoker_device_init, thrust::device, THRUST_FWD(first), THRUST_FWD(last), random_integer<T>());
DEFINE_ASYNC_REDUCE_INVOKER(
  reduce_async_invoker_device_allocator_init,
  thrust::device(thrust::device_allocator<void>{}),
  THRUST_FWD(first),
  THRUST_FWD(last),
  random_integer<T>());
DEFINE_STATEFUL_ASYNC_REDUCE_INVOKER(
  reduce_async_invoker_device_on_init
  // Members.
  ,
  SPECIALIZE_DEVICE_RESOURCE_NAME(Stream_t) stream_;
  // Constructor.
  ,
  thrust::THRUST_DEVICE_BACKEND_DETAIL::throw_on_error(SPECIALIZE_DEVICE_RESOURCE_NAME(StreamCreateWithFlags)(
    &stream_, SPECIALIZE_DEVICE_RESOURCE_NAME(StreamNonBlocking)));
  // Destructor.
  ,
  thrust::THRUST_DEVICE_BACKEND_DETAIL::throw_on_error(SPECIALIZE_DEVICE_RESOURCE_NAME(StreamDestroy)(stream_));
  // `validate_event` member.
  ,
  ASSERT_EQ_QUIET(stream_, e.stream().native_handle());
  // Arguments to `thrust::async::reduce`.
  ,
  thrust::device.on(stream_),
  THRUST_FWD(first),
  THRUST_FWD(last),
  random_integer<T>());
DEFINE_STATEFUL_ASYNC_REDUCE_INVOKER(
  reduce_async_invoker_device_allocator_on_init
  // Members.
  ,
  SPECIALIZE_DEVICE_RESOURCE_NAME(Stream_t) stream_;
  // Constructor.
  ,
  thrust::THRUST_DEVICE_BACKEND_DETAIL::throw_on_error(SPECIALIZE_DEVICE_RESOURCE_NAME(StreamCreateWithFlags)(
    &stream_, SPECIALIZE_DEVICE_RESOURCE_NAME(StreamNonBlocking)));
  // Destructor.
  ,
  thrust::THRUST_DEVICE_BACKEND_DETAIL::throw_on_error(SPECIALIZE_DEVICE_RESOURCE_NAME(StreamDestroy)(stream_));
  // `validate_event` member.
  ,
  ASSERT_EQ_QUIET(stream_, e.stream().native_handle());
  // Arguments to `thrust::async::reduce`.
  ,
  thrust::device(thrust::device_allocator<void>{}).on(stream_),
  THRUST_FWD(first),
  THRUST_FWD(last),
  random_integer<T>());

DEFINE_SYNC_REDUCE_INVOKER(reduce_sync_invoker_init, THRUST_FWD(first), THRUST_FWD(last), random_integer<T>());

DEFINE_ASYNC_REDUCE_INVOKER(
  reduce_async_invoker_init_plus, THRUST_FWD(first), THRUST_FWD(last), random_integer<T>(), thrust::plus<T>());
DEFINE_ASYNC_REDUCE_INVOKER(
  reduce_async_invoker_device_init_plus,
  thrust::device,
  THRUST_FWD(first),
  THRUST_FWD(last),
  random_integer<T>(),
  thrust::plus<T>());
DEFINE_ASYNC_REDUCE_INVOKER(
  reduce_async_invoker_device_allocator_init_plus,
  thrust::device(thrust::device_allocator<void>{}),
  THRUST_FWD(first),
  THRUST_FWD(last),
  random_integer<T>(),
  thrust::plus<T>());
DEFINE_STATEFUL_ASYNC_REDUCE_INVOKER(
  reduce_async_invoker_device_on_init_plus
  // Members.
  ,
  SPECIALIZE_DEVICE_RESOURCE_NAME(Stream_t) stream_;
  // Constructor.
  ,
  thrust::THRUST_DEVICE_BACKEND_DETAIL::throw_on_error(SPECIALIZE_DEVICE_RESOURCE_NAME(StreamCreateWithFlags)(
    &stream_, SPECIALIZE_DEVICE_RESOURCE_NAME(StreamNonBlocking)));
  // Destructor.
  ,
  thrust::THRUST_DEVICE_BACKEND_DETAIL::throw_on_error(SPECIALIZE_DEVICE_RESOURCE_NAME(StreamDestroy)(stream_));
  // `validate_event` member.
  ,
  ASSERT_EQ_QUIET(stream_, e.stream().native_handle());
  // Arguments to `thrust::async::reduce`.
  ,
  thrust::device.on(stream_),
  THRUST_FWD(first),
  THRUST_FWD(last),
  random_integer<T>(),
  thrust::plus<T>());
DEFINE_STATEFUL_ASYNC_REDUCE_INVOKER(
  reduce_async_invoker_device_allocator_on_init_plus
  // Members.
  ,
  SPECIALIZE_DEVICE_RESOURCE_NAME(Stream_t) stream_;
  // Constructor.
  ,
  thrust::THRUST_DEVICE_BACKEND_DETAIL::throw_on_error(SPECIALIZE_DEVICE_RESOURCE_NAME(StreamCreateWithFlags)(
    &stream_, SPECIALIZE_DEVICE_RESOURCE_NAME(StreamNonBlocking)));
  // Destructor.
  ,
  thrust::THRUST_DEVICE_BACKEND_DETAIL::throw_on_error(SPECIALIZE_DEVICE_RESOURCE_NAME(StreamDestroy)(stream_));
  // `validate_event` member.
  ,
  ASSERT_EQ_QUIET(stream_, e.stream().native_handle());
  // Arguments to `thrust::async::reduce`.
  ,
  thrust::device(thrust::device_allocator<void>{}).on(stream_),
  THRUST_FWD(first),
  THRUST_FWD(last),
  random_integer<T>(),
  thrust::plus<T>());

DEFINE_SYNC_REDUCE_INVOKER(
  reduce_sync_invoker_init_plus, THRUST_FWD(first), THRUST_FWD(last), random_integer<T>(), thrust::plus<T>());

DEFINE_ASYNC_REDUCE_INVOKER(
  reduce_async_invoker_init_custom_plus, THRUST_FWD(first), THRUST_FWD(last), random_integer<T>(), custom_plus<T>());
DEFINE_ASYNC_REDUCE_INVOKER(
  reduce_async_invoker_device_init_custom_plus,
  thrust::device,
  THRUST_FWD(first),
  THRUST_FWD(last),
  random_integer<T>(),
  custom_plus<T>());
DEFINE_ASYNC_REDUCE_INVOKER(
  reduce_async_invoker_device_allocator_init_custom_plus,
  thrust::device(thrust::device_allocator<void>{}),
  THRUST_FWD(first),
  THRUST_FWD(last),
  random_integer<T>(),
  custom_plus<T>());
DEFINE_STATEFUL_ASYNC_REDUCE_INVOKER(
  reduce_async_invoker_device_on_init_custom_plus
  // Members.
  ,
  SPECIALIZE_DEVICE_RESOURCE_NAME(Stream_t) stream_;
  // Constructor.
  ,
  thrust::THRUST_DEVICE_BACKEND_DETAIL::throw_on_error(SPECIALIZE_DEVICE_RESOURCE_NAME(StreamCreateWithFlags)(
    &stream_, SPECIALIZE_DEVICE_RESOURCE_NAME(StreamNonBlocking)));
  // Destructor.
  ,
  thrust::THRUST_DEVICE_BACKEND_DETAIL::throw_on_error(SPECIALIZE_DEVICE_RESOURCE_NAME(StreamDestroy)(stream_));
  // `validate_event` member.
  ,
  ASSERT_EQ_QUIET(stream_, e.stream().native_handle());
  // Arguments to `thrust::async::reduce`.
  ,
  thrust::device.on(stream_),
  THRUST_FWD(first),
  THRUST_FWD(last),
  random_integer<T>(),
  custom_plus<T>());
DEFINE_STATEFUL_ASYNC_REDUCE_INVOKER(
  reduce_async_invoker_device_allocator_on_init_custom_plus
  // Members.
  ,
  SPECIALIZE_DEVICE_RESOURCE_NAME(Stream_t) stream_;
  // Constructor.
  ,
  thrust::THRUST_DEVICE_BACKEND_DETAIL::throw_on_error(SPECIALIZE_DEVICE_RESOURCE_NAME(StreamCreateWithFlags)(
    &stream_, SPECIALIZE_DEVICE_RESOURCE_NAME(StreamNonBlocking)));
  // Destructor.
  ,
  thrust::THRUST_DEVICE_BACKEND_DETAIL::throw_on_error(SPECIALIZE_DEVICE_RESOURCE_NAME(StreamDestroy)(stream_));
  // `validate_event` member.
  ,
  ASSERT_EQ_QUIET(stream_, e.stream().native_handle());
  // Arguments to `thrust::async::reduce`.
  ,
  thrust::device(thrust::device_allocator<void>{}).on(stream_),
  THRUST_FWD(first),
  THRUST_FWD(last),
  random_integer<T>(),
  custom_plus<T>());

DEFINE_SYNC_REDUCE_INVOKER(
  reduce_sync_invoker_init_custom_plus, THRUST_FWD(first), THRUST_FWD(last), random_integer<T>(), custom_plus<T>());

///////////////////////////////////////////////////////////////////////////////

template <typename T, template <typename> class AsyncReduceInvoker, template <typename> class SyncReduceInvoker>
THRUST_HOST void test_async_reduce()
{
  for (auto size : get_sizes())
  {
    SCOPED_TRACE(testing::Message() << "with size = " << size);

    thrust::host_vector<T> h0(random_integers<T>(size));
    thrust::device_vector<T> d0a(h0);
    thrust::device_vector<T> d0b(h0);
    thrust::device_vector<T> d0c(h0);
    thrust::device_vector<T> d0d(h0);

    AsyncReduceInvoker<T> invoke_async;
    SyncReduceInvoker<T> invoke_sync;

    ASSERT_EQ(h0, d0a);
    ASSERT_EQ(h0, d0b);
    ASSERT_EQ(h0, d0c);
    ASSERT_EQ(h0, d0d);

    auto f0a = invoke_async(d0a.begin(), d0a.end());
    auto f0b = invoke_async(d0b.begin(), d0b.end());
    auto f0c = invoke_async(d0c.begin(), d0c.end());
    auto f0d = invoke_async(d0d.begin(), d0d.end());

    invoke_async.validate_event(f0a);
    invoke_async.validate_event(f0b);
    invoke_async.validate_event(f0c);
    invoke_async.validate_event(f0d);

    // This potentially runs concurrently with the copies.
    auto const r0 = invoke_sync(h0.begin(), h0.end());

    decltype(f0a.extract()) r1a;
    decltype(f0b.extract()) r1b;
    decltype(f0c.extract()) r1c;
    decltype(f0d.extract()) r1d;

    test_future_value_retrieval(f0a, r1a);
    test_future_value_retrieval(f0b, r1b);
    test_future_value_retrieval(f0c, r1c);
    test_future_value_retrieval(f0d, r1d);

    ASSERT_EQ(r0, r1a);
    ASSERT_EQ(r0, r1b);
    ASSERT_EQ(r0, r1c);
    ASSERT_EQ(r0, r1d);
  }
}

TYPED_TEST(AsyncReduceTests, test_async_reduce)
{
  SCOPED_TRACE(testing::Message() << "with device_id= " << test::set_device_from_ctest());
  using T = typename TestFixture::input_type;
  test_async_reduce<T, reduce_async_invoker, reduce_sync_invoker>();
}

TYPED_TEST(AsyncReduceTests, test_async_reduce_policy)
{
  SCOPED_TRACE(testing::Message() << "with device_id= " << test::set_device_from_ctest());
  using T = typename TestFixture::input_type;
  test_async_reduce<T, reduce_async_invoker_device, reduce_sync_invoker>();
}

TYPED_TEST(AsyncReduceTests, test_async_reduce_policy_allocator)
{
  SCOPED_TRACE(testing::Message() << "with device_id= " << test::set_device_from_ctest());
  using T = typename TestFixture::input_type;
  test_async_reduce<T, reduce_async_invoker_device_allocator, reduce_sync_invoker>();
}

TYPED_TEST(AsyncReduceTests, test_async_reduce_policy_on)
{
  SCOPED_TRACE(testing::Message() << "with device_id= " << test::set_device_from_ctest());
  using T = typename TestFixture::input_type;
  test_async_reduce<T, reduce_async_invoker_device_on, reduce_sync_invoker>();
}

TYPED_TEST(AsyncReduceTests, test_async_reduce_policy_allocator_on)
{
  SCOPED_TRACE(testing::Message() << "with device_id= " << test::set_device_from_ctest());
  using T = typename TestFixture::input_type;
  test_async_reduce<T, reduce_async_invoker_device_allocator_on, reduce_sync_invoker>();
}

TYPED_TEST(AsyncReduceTests, test_async_reduce_init)
{
  SCOPED_TRACE(testing::Message() << "with device_id= " << test::set_device_from_ctest());
  using T = typename TestFixture::input_type;
  test_async_reduce<T, reduce_async_invoker_init, reduce_sync_invoker_init>();
}

TYPED_TEST(AsyncReduceTests, test_async_reduce_policy_init)
{
  SCOPED_TRACE(testing::Message() << "with device_id= " << test::set_device_from_ctest());
  using T = typename TestFixture::input_type;
  test_async_reduce<T, reduce_async_invoker_device_init, reduce_sync_invoker_init>();
}

TYPED_TEST(AsyncReduceTests, test_async_reduce_policy_allocator_init)
{
  SCOPED_TRACE(testing::Message() << "with device_id= " << test::set_device_from_ctest());
  using T = typename TestFixture::input_type;
  test_async_reduce<T, reduce_async_invoker_device_allocator_init, reduce_sync_invoker_init>();
}

TYPED_TEST(AsyncReduceTests, test_async_reduce_policy_on_init)
{
  SCOPED_TRACE(testing::Message() << "with device_id= " << test::set_device_from_ctest());
  using T = typename TestFixture::input_type;
  test_async_reduce<T, reduce_async_invoker_device_on_init, reduce_sync_invoker_init>();
}

TYPED_TEST(AsyncReduceTests, test_async_reduce_policy_allocator_on_init)
{
  SCOPED_TRACE(testing::Message() << "with device_id= " << test::set_device_from_ctest());
  using T = typename TestFixture::input_type;
  test_async_reduce<T, reduce_async_invoker_device_allocator_on_init, reduce_sync_invoker_init>();
}

TYPED_TEST(AsyncReduceTests, test_async_reduce_init_plus)
{
  SCOPED_TRACE(testing::Message() << "with device_id= " << test::set_device_from_ctest());
  using T = typename TestFixture::input_type;
  test_async_reduce<T, reduce_async_invoker_init_plus, reduce_sync_invoker_init_plus>();
}

TYPED_TEST(AsyncReduceTests, test_async_reduce_policy_init_plus)
{
  SCOPED_TRACE(testing::Message() << "with device_id= " << test::set_device_from_ctest());
  using T = typename TestFixture::input_type;
  test_async_reduce<T, reduce_async_invoker_device_init_plus, reduce_sync_invoker_init_plus>();
}

TYPED_TEST(AsyncReduceTests, test_async_reduce_policy_allocator_init_plus)
{
  SCOPED_TRACE(testing::Message() << "with device_id= " << test::set_device_from_ctest());
  using T = typename TestFixture::input_type;
  test_async_reduce<T, reduce_async_invoker_device_allocator_init_plus, reduce_sync_invoker_init_plus>();
}

TYPED_TEST(AsyncReduceTests, test_async_reduce_policy_on_init_plus)
{
  SCOPED_TRACE(testing::Message() << "with device_id= " << test::set_device_from_ctest());
  using T = typename TestFixture::input_type;
  test_async_reduce<T, reduce_async_invoker_device_on_init_plus, reduce_sync_invoker_init_plus>();
}

TYPED_TEST(AsyncReduceTests, test_async_reduce_policy_allocator_on_init_plus)
{
  SCOPED_TRACE(testing::Message() << "with device_id= " << test::set_device_from_ctest());
  using T = typename TestFixture::input_type;
  test_async_reduce<T, reduce_async_invoker_device_allocator_on_init_plus, reduce_sync_invoker_init_plus>();
}

TYPED_TEST(AsyncReduceTests, test_async_reduce_init_custom_plus)
{
  SCOPED_TRACE(testing::Message() << "with device_id= " << test::set_device_from_ctest());
  using T = typename TestFixture::input_type;
  test_async_reduce<T, reduce_async_invoker_init_custom_plus, reduce_sync_invoker_init_custom_plus>();
}

TYPED_TEST(AsyncReduceTests, test_async_reduce_policy_init_custom_plus)
{
  SCOPED_TRACE(testing::Message() << "with device_id= " << test::set_device_from_ctest());
  using T = typename TestFixture::input_type;
  test_async_reduce<T, reduce_async_invoker_device_init_custom_plus, reduce_sync_invoker_init_custom_plus>();
}

TYPED_TEST(AsyncReduceTests, test_async_reduce_policy_allocator_init_custom_plus)
{
  SCOPED_TRACE(testing::Message() << "with device_id= " << test::set_device_from_ctest());
  using T = typename TestFixture::input_type;
  test_async_reduce<T, reduce_async_invoker_device_allocator_init_custom_plus, reduce_sync_invoker_init_custom_plus>();
}

TYPED_TEST(AsyncReduceTests, test_async_reduce_policy_on_init_custom_plus)
{
  SCOPED_TRACE(testing::Message() << "with device_id= " << test::set_device_from_ctest());
  using T = typename TestFixture::input_type;
  test_async_reduce<T, reduce_async_invoker_device_on_init_custom_plus, reduce_sync_invoker_init_custom_plus>();
}

TYPED_TEST(AsyncReduceTests, test_async_reduce_policy_allocator_on_init_custom_plus)
{
  SCOPED_TRACE(testing::Message() << "with device_id= " << test::set_device_from_ctest());
  using T = typename TestFixture::input_type;
  test_async_reduce<T, reduce_async_invoker_device_allocator_on_init_custom_plus, reduce_sync_invoker_init_custom_plus>();
}

///////////////////////////////////////////////////////////////////////////////

template <typename T, template <typename> class AsyncReduceInvoker, template <typename> class SyncReduceInvoker>
THRUST_HOST void test_async_reduce_counting_iterator()
{
  constexpr std::size_t n = 15 * sizeof(T);

  ASSERT_LE(T(n), truncate_to_max_representable<T>(n));

  thrust::counting_iterator<T> first(0);
  thrust::counting_iterator<T> last(n);

  AsyncReduceInvoker<T> invoke_async;
  SyncReduceInvoker<T> invoke_sync;

  auto f0a = invoke_async(first, last);
  auto f0b = invoke_async(first, last);
  auto f0c = invoke_async(first, last);
  auto f0d = invoke_async(first, last);

  invoke_async.validate_event(f0a);
  invoke_async.validate_event(f0b);
  invoke_async.validate_event(f0c);
  invoke_async.validate_event(f0d);

  // This potentially runs concurrently with the copies.
  auto const r0 = invoke_sync(first, last);

  decltype(f0a.extract()) r1a;
  decltype(f0b.extract()) r1b;
  decltype(f0c.extract()) r1c;
  decltype(f0d.extract()) r1d;

  test_future_value_retrieval(f0a, r1a);
  test_future_value_retrieval(f0b, r1b);
  test_future_value_retrieval(f0c, r1c);
  test_future_value_retrieval(f0d, r1d);

  ASSERT_EQ(r0, r1a);
  ASSERT_EQ(r0, r1b);
  ASSERT_EQ(r0, r1c);
  ASSERT_EQ(r0, r1d);
}

TYPED_TEST(AsyncReduceTests, test_async_reduce_counting_iterator)
{
  SCOPED_TRACE(testing::Message() << "with device_id= " << test::set_device_from_ctest());
  using T = typename TestFixture::input_type;
  test_async_reduce_counting_iterator<T, reduce_async_invoker, reduce_sync_invoker>();
}

TYPED_TEST(AsyncReduceTests, test_async_reduce_policy_counting_iterator)
{
  SCOPED_TRACE(testing::Message() << "with device_id= " << test::set_device_from_ctest());
  using T = typename TestFixture::input_type;
  test_async_reduce_counting_iterator<T, reduce_async_invoker_device, reduce_sync_invoker>();
}

TYPED_TEST(AsyncReduceTests, test_async_reduce_counting_iterator_init)
{
  SCOPED_TRACE(testing::Message() << "with device_id= " << test::set_device_from_ctest());
  using T = typename TestFixture::input_type;
  test_async_reduce_counting_iterator<T, reduce_async_invoker_init, reduce_sync_invoker_init>();
}

TYPED_TEST(AsyncReduceTests, test_async_reduce_policy_counting_iterator_init)
{
  SCOPED_TRACE(testing::Message() << "with device_id= " << test::set_device_from_ctest());
  using T = typename TestFixture::input_type;
  test_async_reduce_counting_iterator<T, reduce_async_invoker_device_init, reduce_sync_invoker_init>();
}

TYPED_TEST(AsyncReduceTests, test_async_reduce_counting_iterator_init_plus)
{
  SCOPED_TRACE(testing::Message() << "with device_id= " << test::set_device_from_ctest());
  using T = typename TestFixture::input_type;
  test_async_reduce_counting_iterator<T, reduce_async_invoker_init_plus, reduce_sync_invoker_init_plus>();
}

TYPED_TEST(AsyncReduceTests, test_async_reduce_policy_counting_iterator_init_plus)
{
  SCOPED_TRACE(testing::Message() << "with device_id= " << test::set_device_from_ctest());
  using T = typename TestFixture::input_type;
  test_async_reduce_counting_iterator<T, reduce_async_invoker_device_init_plus, reduce_sync_invoker_init_plus>();
}

TYPED_TEST(AsyncReduceTests, test_async_reduce_counting_iterator_init_custom_plus)
{
  SCOPED_TRACE(testing::Message() << "with device_id= " << test::set_device_from_ctest());
  using T = typename TestFixture::input_type;
  test_async_reduce_counting_iterator<T, reduce_async_invoker_init_custom_plus, reduce_sync_invoker_init_custom_plus>();
}

TYPED_TEST(AsyncReduceTests, test_async_reduce_policy_counting_iterator_init_custom_plus)
{
  SCOPED_TRACE(testing::Message() << "with device_id= " << test::set_device_from_ctest());
  using T = typename TestFixture::input_type;
  test_async_reduce_counting_iterator<T,
                                      reduce_async_invoker_device_init_custom_plus,
                                      reduce_sync_invoker_init_custom_plus>();
}

///////////////////////////////////////////////////////////////////////////////

TYPED_TEST(AsyncReduceTests, test_async_reduce_using)
{
  SCOPED_TRACE(testing::Message() << "with device_id= " << test::set_device_from_ctest());
  using T = typename TestFixture::input_type;
  for (auto size : get_sizes())
  {
    SCOPED_TRACE(testing::Message() << "with size = " << size);

    thrust::host_vector<T> h0(random_integers<T>(size));
    thrust::device_vector<T> d0a(h0);
    thrust::device_vector<T> d0b(h0);

    ASSERT_EQ(h0, d0a);
    ASSERT_EQ(h0, d0b);

    thrust::device_future<T> f0a;
    thrust::device_future<T> f0b;

    // When you import the customization points into the global namespace,
    // they should be selected instead of the synchronous algorithms.
    {
      using namespace thrust::async;
      f0a = reduce(d0a.begin(), d0a.end());
    }
    {
      using thrust::async::reduce;
      f0b = reduce(d0b.begin(), d0b.end());
    }

    // ADL should find the synchronous algorithms.
    // This potentially runs concurrently with the copies.
    T const r0 = reduce(h0.begin(), h0.end());

    T r1a;
    T r1b;

    test_future_value_retrieval(f0a, r1a);
    test_future_value_retrieval(f0b, r1b);

    ASSERT_EQ(r0, r1a);
    ASSERT_EQ(r0, r1b);
  }
}

///////////////////////////////////////////////////////////////////////////////

TYPED_TEST(AsyncReduceTests, test_async_reduce_after)
{
  SCOPED_TRACE(testing::Message() << "with device_id= " << test::set_device_from_ctest());
  using T = typename TestFixture::input_type;
  for (auto size : get_sizes())
  {
    SCOPED_TRACE(testing::Message() << "with size = " << size);

    thrust::host_vector<T> h0(random_integers<T>(size));
    thrust::device_vector<T> d0(h0);

    ASSERT_EQ(h0, d0);

    auto f0 = thrust::async::reduce(d0.begin(), d0.end());

    ASSERT_EQ(true, f0.valid_stream());

    auto const f0_stream = f0.stream().native_handle();

    auto f1 = thrust::async::reduce(thrust::device.after(f0), d0.begin(), d0.end());

    // Verify that double consumption of a future produces an exception.
    ASSERT_THROW(auto x = thrust::async::reduce(thrust::device.after(f0), d0.begin(), d0.end());
                 THRUST_UNUSED_VAR(x), thrust::event_error);

    ASSERT_EQ_QUIET(f0_stream, f1.stream().native_handle());

    auto after_policy2 = thrust::device.after(f1);

    auto f2 = thrust::async::reduce(after_policy2, d0.begin(), d0.end());

    // Verify that double consumption of a policy produces an exception.
    ASSERT_THROW(auto x = thrust::async::reduce(after_policy2, d0.begin(), d0.end());
                 THRUST_UNUSED_VAR(x), thrust::event_error);

    ASSERT_EQ_QUIET(f0_stream, f2.stream().native_handle());

    // This potentially runs concurrently with the copies.
    T const r0 = thrust::reduce(h0.begin(), h0.end());

    T r1;

    test_future_value_retrieval(f2, r1);

    ASSERT_EQ(r0, r1);
  }
}

///////////////////////////////////////////////////////////////////////////////

TYPED_TEST(AsyncReduceTests, test_async_reduce_on_then_after)
{
  SCOPED_TRACE(testing::Message() << "with device_id= " << test::set_device_from_ctest());
  using T = typename TestFixture::input_type;
  for (auto size : get_sizes())
  {
    SCOPED_TRACE(testing::Message() << "with size = " << size);

    thrust::host_vector<T> h0(random_integers<T>(size));
    thrust::device_vector<T> d0(h0);

    ASSERT_EQ(h0, d0);

    SPECIALIZE_DEVICE_RESOURCE_NAME(Stream_t) stream;
    thrust::THRUST_DEVICE_BACKEND_DETAIL::throw_on_error(SPECIALIZE_DEVICE_RESOURCE_NAME(
      StreamCreateWithFlags)(&stream, SPECIALIZE_DEVICE_RESOURCE_NAME(StreamNonBlocking)));

    auto f0 = thrust::async::reduce(thrust::device.on(stream), d0.begin(), d0.end());

    ASSERT_EQ_QUIET(stream, f0.stream().native_handle());

    auto f1 = thrust::async::reduce(thrust::device.after(f0), d0.begin(), d0.end());

    // Verify that double consumption of a future produces an exception.
    ASSERT_THROW(auto x = thrust::async::reduce(thrust::device.after(f0), d0.begin(), d0.end());
                 THRUST_UNUSED_VAR(x), thrust::event_error);

    ASSERT_EQ_QUIET(stream, f1.stream().native_handle());

    auto after_policy2 = thrust::device.after(f1);

    auto f2 = thrust::async::reduce(after_policy2, d0.begin(), d0.end());

    // Verify that double consumption of a policy produces an exception.
    ASSERT_THROW(auto x = thrust::async::reduce(after_policy2, d0.begin(), d0.end());
                 THRUST_UNUSED_VAR(x), thrust::event_error);

    ASSERT_EQ_QUIET(stream, f2.stream().native_handle());

    // This potentially runs concurrently with the copies.
    T const r0 = thrust::reduce(h0.begin(), h0.end());

    T r1;

    test_future_value_retrieval(f2, r1);

    ASSERT_EQ(r0, r1);

    thrust::THRUST_DEVICE_BACKEND_DETAIL::throw_on_error(SPECIALIZE_DEVICE_RESOURCE_NAME(StreamDestroy)(stream));
  }
}

///////////////////////////////////////////////////////////////////////////////

TYPED_TEST(AsyncReduceTests, test_async_reduce_allocator_on_then_after)
{
  SCOPED_TRACE(testing::Message() << "with device_id= " << test::set_device_from_ctest());
  using T = typename TestFixture::input_type;
  for (auto size : get_sizes())
  {
    SCOPED_TRACE(testing::Message() << "with size = " << size);

    thrust::host_vector<T> h0(random_integers<T>(size));
    thrust::device_vector<T> d0(h0);

    ASSERT_EQ(h0, d0);

    SPECIALIZE_DEVICE_RESOURCE_NAME(Stream_t) stream0;
    thrust::THRUST_DEVICE_BACKEND_DETAIL::throw_on_error(SPECIALIZE_DEVICE_RESOURCE_NAME(
      StreamCreateWithFlags)(&stream0, SPECIALIZE_DEVICE_RESOURCE_NAME(StreamNonBlocking)));

    SPECIALIZE_DEVICE_RESOURCE_NAME(Stream_t) stream1;
    thrust::THRUST_DEVICE_BACKEND_DETAIL::throw_on_error(SPECIALIZE_DEVICE_RESOURCE_NAME(
      StreamCreateWithFlags)(&stream1, SPECIALIZE_DEVICE_RESOURCE_NAME(StreamNonBlocking)));

    auto f0 = thrust::async::reduce(thrust::device(thrust::device_allocator<void>{}).on(stream0), d0.begin(), d0.end());

    ASSERT_EQ_QUIET(stream0, f0.stream().native_handle());

    auto f1 = thrust::async::reduce(thrust::device(thrust::device_allocator<void>{}).after(f0), d0.begin(), d0.end());

    ASSERT_THROW(
      auto x = thrust::async::reduce(thrust::device(thrust::device_allocator<void>{}).after(f0), d0.begin(), d0.end());
      THRUST_UNUSED_VAR(x), thrust::event_error);

    ASSERT_EQ_QUIET(stream0, f1.stream().native_handle());

    auto f2 = thrust::async::reduce(
      thrust::device(thrust::device_allocator<void>{}).on(stream1).after(f1), d0.begin(), d0.end());

    ASSERT_THROW(auto x = thrust::async::reduce(
                   thrust::device(thrust::device_allocator<void>{}).on(stream1).after(f1), d0.begin(), d0.end());
                 THRUST_UNUSED_VAR(x), thrust::event_error);

    // KNOWN_FAILURE;
    // FIXME: The below fails because you can't combine allocator attachment,
    // `.on`, and `.after`.
    // The `#if 0` can be removed once the KNOWN_FAILURE is resolved.
#  if 0
    ASSERT_EQ_QUIET(stream1, f2.stream().native_handle());

    // This potentially runs concurrently with the copies.
    T const r0 = thrust::reduce(h0.begin(), h0.end());

    T r1;

    test_future_value_retrieval(f2, r1);

    ASSERT_EQ(r0, r1);

    thrust::THRUST_DEVICE_BACKEND_DETAIL::throw_on_error(SPECIALIZE_DEVICE_RESOURCE_NAME(StreamDestroy)(stream0));
    thrust::THRUST_DEVICE_BACKEND_DETAIL::throw_on_error(SPECIALIZE_DEVICE_RESOURCE_NAME(StreamDestroy)(stream1));
#  endif
  }
}

///////////////////////////////////////////////////////////////////////////////

TYPED_TEST(AsyncReduceTests, test_async_reduce_caching)
{
  SCOPED_TRACE(testing::Message() << "with device_id= " << test::set_device_from_ctest());
  using T = typename TestFixture::input_type;
  for (auto size : get_sizes())
  {
    SCOPED_TRACE(testing::Message() << "with size = " << size);
    constexpr std::int64_t m = 32;

    thrust::host_vector<T> h0(random_integers<T>(size));
    thrust::device_vector<T> d0(h0);

    ASSERT_EQ(h0, d0);

    T const* f0_raw_data;

    {
      // Perform one reduction to ensure there's an entry in the caching
      // allocator.
      auto f0 = thrust::async::reduce(d0.begin(), d0.end());

      TEST_EVENT_WAIT(f0);

      f0_raw_data = f0.raw_data();
    }

    for (std::int64_t i = 0; i < m; ++i)
    {
      auto f1 = thrust::async::reduce(d0.begin(), d0.end());

      ASSERT_EQ(true, f1.valid_stream());
      ASSERT_EQ(true, f1.valid_content());

      ASSERT_EQ_QUIET(f0_raw_data, f1.raw_data());

      // This potentially runs concurrently with the copies.
      T const r0 = thrust::reduce(h0.begin(), h0.end());

      T r1;
      test_future_value_retrieval(f1, r1);

      ASSERT_EQ(r0, r1);
    }
  }
}

///////////////////////////////////////////////////////////////////////////////

TYPED_TEST(AsyncReduceTests, test_async_copy_then_reduce)
{
  SCOPED_TRACE(testing::Message() << "with device_id= " << test::set_device_from_ctest());
  using T = typename TestFixture::input_type;
  for (auto size : get_sizes())
  {
    SCOPED_TRACE(testing::Message() << "with size = " << size);

    thrust::host_vector<T> h0a(random_integers<T>(size));
    thrust::host_vector<T> h0b(random_integers<T>(size));
    thrust::host_vector<T> h0c(random_integers<T>(size));
    thrust::host_vector<T> h0d(random_integers<T>(size));

    thrust::device_vector<T> d0a(size);
    thrust::device_vector<T> d0b(size);
    thrust::device_vector<T> d0c(size);
    thrust::device_vector<T> d0d(size);

    auto f0a = thrust::async::copy(h0a.begin(), h0a.end(), d0a.begin());
    auto f0b = thrust::async::copy(h0b.begin(), h0b.end(), d0b.begin());
    auto f0c = thrust::async::copy(h0c.begin(), h0c.end(), d0c.begin());
    auto f0d = thrust::async::copy(h0d.begin(), h0d.end(), d0d.begin());

    ASSERT_EQ(true, f0a.valid_stream());
    ASSERT_EQ(true, f0b.valid_stream());
    ASSERT_EQ(true, f0c.valid_stream());
    ASSERT_EQ(true, f0d.valid_stream());

    auto const f0a_stream = f0a.stream().native_handle();
    auto const f0b_stream = f0b.stream().native_handle();
    auto const f0c_stream = f0c.stream().native_handle();
    auto const f0d_stream = f0d.stream().native_handle();

    auto f1a = thrust::async::reduce(thrust::device.after(f0a), d0a.begin(), d0a.end());
    auto f1b = thrust::async::reduce(thrust::device.after(f0b), d0b.begin(), d0b.end());
    auto f1c = thrust::async::reduce(thrust::device.after(f0c), d0c.begin(), d0c.end());
    auto f1d = thrust::async::reduce(thrust::device.after(f0d), d0d.begin(), d0d.end());

    ASSERT_EQ(false, f0a.valid_stream());
    ASSERT_EQ(false, f0b.valid_stream());
    ASSERT_EQ(false, f0c.valid_stream());
    ASSERT_EQ(false, f0d.valid_stream());

    ASSERT_EQ(true, f1a.valid_stream());
    ASSERT_EQ(true, f1a.valid_content());
    ASSERT_EQ(true, f1b.valid_stream());
    ASSERT_EQ(true, f1b.valid_content());
    ASSERT_EQ(true, f1c.valid_stream());
    ASSERT_EQ(true, f1c.valid_content());
    ASSERT_EQ(true, f1d.valid_stream());
    ASSERT_EQ(true, f1d.valid_content());

    // Verify that streams were stolen.
    ASSERT_EQ_QUIET(f0a_stream, f1a.stream().native_handle());
    ASSERT_EQ_QUIET(f0b_stream, f1b.stream().native_handle());
    ASSERT_EQ_QUIET(f0c_stream, f1c.stream().native_handle());
    ASSERT_EQ_QUIET(f0d_stream, f1d.stream().native_handle());

    // This potentially runs concurrently with the copies.
    T const r0 = thrust::reduce(h0a.begin(), h0a.end());

    T r1a;
    T r1b;
    T r1c;
    T r1d;

    test_future_value_retrieval(f1a, r1a);
    test_future_value_retrieval(f1b, r1b);
    test_future_value_retrieval(f1c, r1c);
    test_future_value_retrieval(f1d, r1d);

    ASSERT_EQ(r0, r1a);
    ASSERT_EQ(r0, r1b);
    ASSERT_EQ(r0, r1c);
    ASSERT_EQ(r0, r1d);
  }
}

///////////////////////////////////////////////////////////////////////////////

// TODO: when_all from reductions.

// See also issue: https://github.com/NVIDIA/cccl/issues/1886
struct tuple_sum
{
  __device__ thrust::tuple<int, int>
  operator()(const thrust::tuple<int, int>& t1, const thrust::tuple<int, int>& t2) const
  {
    return thrust::make_tuple(thrust::get<0>(t1) + thrust::get<0>(t2), thrust::get<1>(t1) + thrust::get<1>(t2));
  }
};

TEST(AsyncReduceTests, test_async_reduce_bug1886)
{
#ifdef ADDRESS_SANITIZER_BUILD
	GTEST_SKIP() << "This test is disabled in ASAN builds due to its memory usage.";
#endif
	
  SCOPED_TRACE(testing::Message() << "with device_id= " << test::set_device_from_ctest());

  // Initialize input data
  thrust::device_vector<int> d_data1{1, 2, 3, 4, 5};
  thrust::device_vector<int> d_data2{10, 20, 30, 40, 50};

  using TupleType = thrust::tuple<int, int>;
  using IteratorType =
    thrust::zip_iterator<thrust::tuple<thrust::device_vector<int>::iterator, thrust::device_vector<int>::iterator>>;

  TupleType result;
  try
  {
    // Create zip_begin and zip_end iterators
    IteratorType zip_begin = thrust::make_zip_iterator(thrust::make_tuple(d_data1.begin(), d_data2.begin()));
    IteratorType zip_end   = thrust::make_zip_iterator(thrust::make_tuple(d_data2.end(), d_data2.end()));

    // Initialize the starting tuple
    TupleType init = thrust::make_tuple(0, 0);

    // Perform async reduce using zip_begin and zip_end
    auto future = thrust::async::reduce(thrust::device, zip_begin, zip_end, init, tuple_sum());

    // Get the result
    result = future.get();
  }
  catch (const ::std::bad_alloc& e)
  {
    (void) hipGetLastError();
    GTEST_SKIP() << "std::bad_alloc: " << e.what();
  }

  // Print the result
  std::cout << "Sum: (" << thrust::get<0>(result) << ", " << thrust::get<1>(result) << ")" << std::endl;
}

#endif

// we need to leak the suppression on clang/MSVC to suppresses warnings from the cudafe1.stub.c file
#if THRUST_HOST_COMPILER != THRUST_HOST_COMPILER_CLANG && THRUST_HOST_COMPILER != THRUST_HOST_COMPILER_MSVC
THRUST_SUPPRESS_DEPRECATED_POP
#endif // THRUST_HOST_COMPILER != THRUST_HOST_COMPILER_CLANG && THRUST_HOST_COMPILER != THRUST_HOST_COMPILER_MSVC