TheRock/rocm-systems/projects/hip-tests/catch/unit/stream/hipStreamValue.cc

/*
 * Copyright (c) Advanced Micro Devices, Inc., or its affiliates.
 *
 * SPDX-License-Identifier: MIT
 */

#include <hip_test_common.hh>
#include <memory>
#include <type_traits>

constexpr unsigned int writeFlag = 0;

template <typename UIntT, typename... Args> auto waitFunc(Args... args) {
  if constexpr (std::is_same<UIntT, uint32_t>::value) {
    return hipStreamWaitValue32(args...);
  } else {
    return hipStreamWaitValue64(args...);
  }
};

template <typename UIntT, typename... Args> auto writeFunc(Args... args) {
  if constexpr (std::is_same<UIntT, uint32_t>::value) {
    return hipStreamWriteValue32(args...);
  } else {
    return hipStreamWriteValue64(args...);
  }
};

// Random predefined 32 and 64 bit values
using value32_t = std::integral_constant<uint32_t, 0x70F0F0FF>;
using value64_t = std::integral_constant<uint64_t, 0x7FFF0000FFFF0000>;
template <typename UIntT> using testValue =
    typename std::conditional<std::is_same<UIntT, uint32_t>::value, value32_t, value64_t>::type;

constexpr uint32_t DATA_INIT = 0x1234;
constexpr uint32_t DATA_UPDATE = 0X4321;

template <typename UIntT> struct TEST_WAIT {
  static_assert(std::is_same<UIntT, uint32_t>::value or std::is_same<UIntT, uint64_t>::value,
                "only implemented for 32 bit and 64 bit unsigned integers");
  unsigned int compareOp;
  UIntT mask = ~static_cast<UIntT>(0);
  UIntT waitValue;
  UIntT signalValueFail;
  UIntT signalValuePass;

  TEST_WAIT(unsigned int compareOp, UIntT waitValue, UIntT signalValueFail, UIntT signalValuePass)
      : compareOp{compareOp},
        waitValue{waitValue},
        signalValueFail{signalValueFail},
        signalValuePass{signalValuePass} {}

  TEST_WAIT(unsigned int compareOp, UIntT mask, UIntT waitValue, UIntT signalValueFail,
            UIntT signalValuePass)
      : compareOp{compareOp},
        mask{mask},
        waitValue{waitValue},
        signalValueFail{signalValueFail},
        signalValuePass{signalValuePass} {}
};

using TEST_WAIT32 = TEST_WAIT<uint32_t>;
using TEST_WAIT64 = TEST_WAIT<uint64_t>;

bool streamWaitValueSupported() {
  int device_num = 0;
  HIP_CHECK(hipGetDeviceCount(&device_num));
  for (int device_id = 0; device_id < device_num; ++device_id) {
    HIP_CHECK(hipSetDevice(device_id));
    int waitValueSupport = 0;
    auto getAttributeError = hipDeviceGetAttribute(
        &waitValueSupport, hipDeviceAttributeCanUseStreamWaitValue, device_id);
    if (getAttributeError != hipSuccess) {
      return false;
    }
    if (waitValueSupport == 1) return true;
  }
  return false;
}

// The different types of memory that can be used with hipStream[Wait|Write]
enum class PtrType { HostPtr, DevicePtr, DevicePtrToHost, Signal };

// Helper class to expose the pointer that is used with hipStream[Write|Wait]Value and also store a
// unique pointer with the deleter to simplify cleanup
// Also includes functions to update and get the value directly
template <PtrType type, typename UIntT, typename UniquePtrWithDeleter> class TestPtr {
  // This stores the memory that must be deleted, as well as the deleter
  UniquePtrWithDeleter ptrToDelete;

 public:
  // The pointer that should be used with hipStream[Write|Wait]Value
  UIntT* ptr;

  TestPtr(UIntT* ptr, UniquePtrWithDeleter ptrToDelete)
      : ptrToDelete(std::move(ptrToDelete)), ptr(ptr) {}


  // directly retrieve the value from wherever it was allocated
  UIntT getValue(size_t offset = 0) {
    if constexpr (type == PtrType::Signal || type == PtrType::HostPtr ||
                  type == PtrType::DevicePtrToHost) {
      return ptrToDelete.get()[offset];
    } else {
      static_assert(type == PtrType::DevicePtr, "Expected DevicePtr");
      UIntT value;
      HIP_CHECK(hipMemcpy(&value, ptr + offset, sizeof(UIntT), hipMemcpyDeviceToHost));
      return value;
    }
  }

  // directly set the value wherever it was allocated
  void setValue(UIntT value, size_t offset = 0) {
    if constexpr (type == PtrType::Signal || type == PtrType::DevicePtrToHost ||
                  type == PtrType::HostPtr) {
      ptrToDelete.get()[offset] = value;
    } else {
      // hipMemcpy causes deadlock, so use hipStreamWriteValue
      static_assert(type == PtrType::DevicePtr, "Expected DevicePtr");
      hipStream_t stream;
      HIP_CHECK(hipStreamCreateWithFlags(&stream, hipStreamNonBlocking));
      HIP_CHECK(writeFunc<UIntT>(stream, ptr + offset, value, writeFlag));
      HIP_CHECK(hipStreamSynchronize(stream));
      HIP_CHECK(hipStreamDestroy(stream));
    }
  }
};

// required for the static assert
template <PtrType> inline constexpr bool AMD_ACTIVE = HT_AMD == 1;

template <PtrType type, typename UIntT> auto allocMem() {
  constexpr std::size_t arraySize = 1024;
  if constexpr (type == PtrType::Signal) {
    static_assert(std::is_same<UIntT, uint64_t>::value,
                  "signal memory should only be used with 64bit memory");

    // Allocate Signal Memory
    uint64_t* signalPtr{};

    static_assert(AMD_ACTIVE<type>,
                  "nvidia backend compiler doesn't like hipExtMallocWithFlags, even in this "
                  "constexpr branch");
#if HT_AMD
    // 8 is the only acceptable size
    HIP_CHECK(
        hipExtMallocWithFlags(reinterpret_cast<void**>(&signalPtr), 8, hipMallocSignalMemory));
#endif

    // Init Memory
    *signalPtr = 0;

    auto freeStuff = [](uint64_t* sPtr) { HIP_CHECK(hipFree(sPtr)); };
    // sync ensures the memory intialization is complete before using the memory
    HIP_CHECK(hipDeviceSynchronize());
    return TestPtr<type, UIntT, std::unique_ptr<uint64_t, decltype(freeStuff)>>{
        signalPtr, std::unique_ptr<uint64_t, decltype(freeStuff)>(signalPtr, freeStuff)};
  } else if constexpr (type == PtrType::DevicePtrToHost) {
    auto hostPtr = new UIntT[arraySize];

    // Register Host Memory
    HIP_CHECK(hipHostRegister(hostPtr, sizeof(UIntT) * arraySize, 0));

    // Init memory
    std::fill(hostPtr, hostPtr + arraySize, 0);

    UIntT* devicePtr;
    // Test writing device pointer
    HIP_CHECK(hipHostGetDevicePointer(reinterpret_cast<void**>(&devicePtr), hostPtr, 0));
    auto freeStuff = [](UIntT* ptr) {
      HIP_CHECK(hipHostUnregister(ptr));
      delete[] ptr;
    };
    // sync ensures the memory intialization is complete before using the memory
    HIP_CHECK(hipDeviceSynchronize());
    return TestPtr<type, UIntT, std::unique_ptr<UIntT[], decltype(freeStuff)>>{
        devicePtr, std::unique_ptr<UIntT[], decltype(freeStuff)>(hostPtr, freeStuff)};
  } else if constexpr (type == PtrType::HostPtr) {
    auto hostPtr = new UIntT[arraySize];

    // Register Host Memory
    HIP_CHECK(hipHostRegister(hostPtr, sizeof(UIntT) * arraySize, 0));

    // Init memory
    std::fill(hostPtr, hostPtr + arraySize, 0);

    auto freeStuff = [](UIntT* ptr) {
      HIP_CHECK(hipHostUnregister(ptr));
      delete[] ptr;
    };
    // sync ensures the memory intialization is complete before using the memory
    HIP_CHECK(hipDeviceSynchronize());
    return TestPtr<type, UIntT, std::unique_ptr<UIntT[], decltype(freeStuff)>>{
        hostPtr, std::unique_ptr<UIntT[], decltype(freeStuff)>(hostPtr, freeStuff)};
  } else {
    static_assert(type == PtrType::DevicePtr, "Expected DevicePtr");
    UIntT* devicePtr;
    HIP_CHECK(hipMalloc(&devicePtr, sizeof(UIntT) * arraySize));
    HIP_CHECK(hipMemset(devicePtr, 0, sizeof(UIntT) * arraySize));
    auto freeStuff = [](UIntT* ptr) { HIP_CHECK(hipFree(ptr)); };
    // sync ensures the memory intialization is complete before using the memory
    HIP_CHECK(hipDeviceSynchronize());
    return TestPtr<type, UIntT, std::unique_ptr<UIntT, decltype(freeStuff)>>{
        devicePtr, std::unique_ptr<UIntT, decltype(freeStuff)>(devicePtr, freeStuff)};
  }
}

// allows the creation of a list of offsets while avoiding it for signal memory
template <PtrType type> constexpr auto get_offsets() {
  if constexpr (type == PtrType::Signal) {
    return std::array<size_t, 1>{0};
  } else {
    return std::array<size_t, 6>{0, 1, 2, 3, 31, 1023};
  }
}

template <typename UIntT, PtrType ptrTypeValue> struct TestParams {
  using UIntType = UIntT;
  constexpr static PtrType ptrType = ptrTypeValue;
};

#if HT_AMD
HIP_TEMPLATE_TEST_CASE(Unit_hipStreamValue_Write, (TestParams<uint32_t, PtrType::HostPtr>),
                   (TestParams<uint32_t, PtrType::DevicePtr>),
                   (TestParams<uint32_t, PtrType::DevicePtrToHost>),
                   (TestParams<uint64_t, PtrType::HostPtr>),
                   (TestParams<uint64_t, PtrType::DevicePtr>),
                   (TestParams<uint64_t, PtrType::DevicePtrToHost>),
                   (TestParams<uint64_t, PtrType::Signal>)) {
#else
HIP_TEMPLATE_TEST_CASE(Unit_hipStreamValue_Write, (TestParams<uint32_t, PtrType::HostPtr>),
                   (TestParams<uint32_t, PtrType::DevicePtr>),
                   (TestParams<uint32_t, PtrType::DevicePtrToHost>),
                   (TestParams<uint64_t, PtrType::HostPtr>),
                   (TestParams<uint64_t, PtrType::DevicePtr>),
                   (TestParams<uint64_t, PtrType::DevicePtrToHost>)) {
#endif
  if (!streamWaitValueSupported()) {
    HIP_SKIP_TEST(HipTest::SkipReason::kStreamWaitValueUnsupported);
  }

  using UIntT = typename TestType::UIntType;
  constexpr auto ptrType = TestType::ptrType;
  constexpr auto writeValue = testValue<UIntT>::value;

  hipStream_t stream{nullptr};
  HIP_CHECK(hipStreamCreate(&stream));

  const auto offsets = get_offsets<ptrType>();
  const auto offset = GENERATE_COPY(from_range(std::begin(offsets), std::end(offsets)));

  CAPTURE(offset);

  // Allocate Host Memory
  auto ptr = allocMem<ptrType, UIntT>();
  UIntT* target = ptr.ptr + offset;
  HIP_CHECK(writeFunc<UIntT>(stream, target, writeValue, writeFlag));
  HIP_CHECK(hipStreamSynchronize(stream));
  REQUIRE(ptr.getValue(offset) == writeValue);

  // Cleanup
  HIP_CHECK(hipStreamDestroy(stream));
}

template <bool isBlocking, typename UIntT, typename TestPtr>
void syncAndCheckData(hipStream_t stream, UIntT* dataPtr, TestPtr signalPtr, size_t offset,
                      TEST_WAIT<UIntT> tc, std::array<hipEvent_t, 2>& events) {
  // Ensure first part of host memory is updated
  HIP_CHECK(hipEventSynchronize(events[0]));
  REQUIRE(dataPtr[0] == DATA_UPDATE);

  if constexpr (isBlocking) {
    // Ensure second part of host memory isn't updated yet
    HIP_CHECK_ERROR(hipEventQuery(events[1]), hipErrorNotReady);
    REQUIRE(dataPtr[1] == DATA_INIT);
    // Update value to release stream
    signalPtr.setValue(tc.signalValuePass, offset);
  }

  HIP_CHECK(hipStreamSynchronize(stream));
  HIP_CHECK(hipEventQuery(events[1]));
  // Finally ensure that second part of host memory is updated
  REQUIRE(dataPtr[1] == DATA_UPDATE);
}


template <typename TestType, bool isBlocking>
void testWait(TEST_WAIT<typename TestType::UIntType> tc) {
  if (!streamWaitValueSupported()) {
    HIP_SKIP_TEST(HipTest::SkipReason::kStreamWaitValueUnsupported);
  }
  using UIntT = typename TestType::UIntType;
  constexpr auto ptrType = TestType::ptrType;
  constexpr UIntT defaultMask = ~static_cast<UIntT>(0);

  // Initialize stream
  hipStream_t stream{};
  HIP_CHECK(hipStreamCreate(&stream));

  // Allocate Host Memory
  auto dataPtr = std::make_unique<UIntT[]>(2);
  // Register Host Memory
  HIP_CHECK(hipHostRegister(dataPtr.get(), sizeof(UIntT), 0));
  HIP_CHECK(hipHostRegister(dataPtr.get() + 1, sizeof(UIntT), 0));
  std::fill(dataPtr.get(), dataPtr.get() + 2, DATA_INIT);

  std::array<hipEvent_t, 2> events;
  HIP_CHECK(hipEventCreate(&events[0]));
  HIP_CHECK(hipEventCreate(&events[1]));


  const auto offsets = get_offsets<ptrType>();
  const auto offset = GENERATE_COPY(from_range(std::begin(offsets), std::end(offsets)));

  auto waitPtr = allocMem<ptrType, UIntT>();
  UIntT* const target = waitPtr.ptr + offset;
  waitPtr.setValue(isBlocking ? tc.signalValueFail : tc.signalValuePass, offset);

  HIP_CHECK(writeFunc<UIntT>(stream, &(dataPtr.get()[0]), DATA_UPDATE, writeFlag));
  HIP_CHECK(hipEventRecord(events[0], stream));

  if (tc.mask != defaultMask) {
    HIP_CHECK(waitFunc<UIntT>(stream, target, tc.waitValue, tc.compareOp, tc.mask));
  } else {
    HIP_CHECK(waitFunc<UIntT>(stream, target, tc.waitValue, tc.compareOp));
  }

  HIP_CHECK(writeFunc<UIntT>(stream, &(dataPtr.get()[1]), DATA_UPDATE, writeFlag));

  HIP_CHECK(hipEventRecord(events[1], stream));

  syncAndCheckData<isBlocking>(stream, dataPtr.get(), std::move(waitPtr), offset, tc, events);

  // Cleanup
  HIP_CHECK(hipEventDestroy(events[0]));
  HIP_CHECK(hipEventDestroy(events[1]));
  HIP_CHECK(hipHostUnregister(dataPtr.get()));
  HIP_CHECK(hipHostUnregister(dataPtr.get() + 1));
  HIP_CHECK(hipStreamDestroy(stream));
}

// Combined blocking test case for both uint32_t and uint64_t
HIP_TEMPLATE_TEST_CASE(Unit_hipStreamValue_Wait_Blocking, uint32_t, uint64_t) {
  if (!streamWaitValueSupported()) {
    HIP_SKIP_TEST(HipTest::SkipReason::kStreamWaitValueUnsupported);
  }

  using UIntT = TestType;
  using TestWaitType = TEST_WAIT<UIntT>;

  SECTION("Mask_Gte") {
    TestWaitType testCase(hipStreamWaitValueGte, 0xF, 0x4, 0x3, 0x6);

    SECTION("HostPtr") {
      testWait<TestParams<UIntT, PtrType::HostPtr>, true>(testCase);
    }
    SECTION("DevicePtr") {
      testWait<TestParams<UIntT, PtrType::DevicePtr>, true>(testCase);
    }
    SECTION("DevicePtrToHost") {
      testWait<TestParams<UIntT, PtrType::DevicePtrToHost>, true>(testCase);
    }
#if HT_AMD
    if constexpr (std::is_same_v<UIntT, uint64_t>) {
      SECTION("Signal") {
        testWait<TestParams<UIntT, PtrType::Signal>, true>(testCase);
      }
    }
#endif
  }

  if constexpr (std::is_same_v<UIntT, uint64_t>) {
    SECTION("Mask_Gte_1") {
      TestWaitType testCase(hipStreamWaitValueGte, 0x0000FFFFFFFFFFFF, 0x000000007FFF0001,
                            0x7FFF00007FFF0000, 0x000000007FFF0001);
      SECTION("HostPtr") {
        testWait<TestParams<UIntT, PtrType::HostPtr>, true>(testCase);
      }
      SECTION("DevicePtr") {
        testWait<TestParams<UIntT, PtrType::DevicePtr>, true>(testCase);
      }
      SECTION("DevicePtrToHost") {
        testWait<TestParams<UIntT, PtrType::DevicePtrToHost>, true>(testCase);
      }
#if HT_AMD
      SECTION("Signal") {
        testWait<TestParams<UIntT, PtrType::Signal>, true>(testCase);
      }
#endif
    }
  }

  SECTION("Mask_Eq_1") {
    TestWaitType testCase = std::is_same_v<UIntT, uint32_t>
      ? TestWaitType(hipStreamWaitValueEq, 0x0000FFFF, 0x00000001, 0x0FFF0000, 0x0FFF0001)
      : TestWaitType(hipStreamWaitValueEq, 0x0000FFFFFFFFFFFF, 0x000000000FFF0001,
                     0x7FFF00000FFF0000, 0x7F0000000FFF0001);

    SECTION("HostPtr") {
      testWait<TestParams<UIntT, PtrType::HostPtr>, true>(testCase);
    }
    SECTION("DevicePtr") {
      testWait<TestParams<UIntT, PtrType::DevicePtr>, true>(testCase);
    }
    SECTION("DevicePtrToHost") {
      testWait<TestParams<UIntT, PtrType::DevicePtrToHost>, true>(testCase);
    }
#if HT_AMD
    if constexpr (std::is_same_v<UIntT, uint64_t>) {
      SECTION("Signal") {
        testWait<TestParams<UIntT, PtrType::Signal>, true>(testCase);
      }
    }
#endif
  }

  SECTION("Mask_Eq_2") {
    TestWaitType testCase(hipStreamWaitValueEq, 0xFF, 0x11, 0x25, 0x11);

    SECTION("HostPtr") {
      testWait<TestParams<UIntT, PtrType::HostPtr>, true>(testCase);
    }
    SECTION("DevicePtr") {
      testWait<TestParams<UIntT, PtrType::DevicePtr>, true>(testCase);
    }
    SECTION("DevicePtrToHost") {
      testWait<TestParams<UIntT, PtrType::DevicePtrToHost>, true>(testCase);
    }
#if HT_AMD
    if constexpr (std::is_same_v<UIntT, uint64_t>) {
      SECTION("Signal") {
        testWait<TestParams<UIntT, PtrType::Signal>, true>(testCase);
      }
    }
#endif
  }

  SECTION("Mask_And") {
    TestWaitType testCase(hipStreamWaitValueAnd, 0xFF, 0xF4A, 0xF35, 0X02);

    SECTION("HostPtr") {
      testWait<TestParams<UIntT, PtrType::HostPtr>, true>(testCase);
    }
    SECTION("DevicePtr") {
      testWait<TestParams<UIntT, PtrType::DevicePtr>, true>(testCase);
    }
    SECTION("DevicePtrToHost") {
      testWait<TestParams<UIntT, PtrType::DevicePtrToHost>, true>(testCase);
    }
#if HT_AMD
    if constexpr (std::is_same_v<UIntT, uint64_t>) {
      SECTION("Signal") {
        testWait<TestParams<UIntT, PtrType::Signal>, true>(testCase);
      }
    }
#endif
  }

  SECTION("NoMask_Eq") {
    TestWaitType testCase = std::is_same_v<UIntT, uint32_t>
      ? TestWaitType(hipStreamWaitValueEq, 0x7FFFFFFF, 0x7FFF0000, 0x7FFFFFFF)
      : TestWaitType(hipStreamWaitValueEq, 0x7FFFFFFFFFFFFFFF, 0x7FFFFFFF0FFF0000, 0x7FFFFFFFFFFFFFFF);

    SECTION("HostPtr") {
      testWait<TestParams<UIntT, PtrType::HostPtr>, true>(testCase);
    }
    SECTION("DevicePtr") {
      testWait<TestParams<UIntT, PtrType::DevicePtr>, true>(testCase);
    }
    SECTION("DevicePtrToHost") {
      testWait<TestParams<UIntT, PtrType::DevicePtrToHost>, true>(testCase);
    }
#if HT_AMD
    if constexpr (std::is_same_v<UIntT, uint64_t>) {
      SECTION("Signal") {
        testWait<TestParams<UIntT, PtrType::Signal>, true>(testCase);
      }
    }
#endif
  }

  SECTION("NoMask_Gte") {
    TestWaitType testCase = std::is_same_v<UIntT, uint32_t>
      ? TestWaitType(hipStreamWaitValueGte, 0x7FFF0001, 0x7FFF0000, 0x7FFF0010)
      : TestWaitType(hipStreamWaitValueGte, 0x7FFFFFFFFFFF0001, 0x7FFFFFFFFFFF0000, 0x7FFFFFFFFFFF0001);

    SECTION("HostPtr") {
      testWait<TestParams<UIntT, PtrType::HostPtr>, true>(testCase);
    }
    SECTION("DevicePtr") {
      testWait<TestParams<UIntT, PtrType::DevicePtr>, true>(testCase);
    }
    SECTION("DevicePtrToHost") {
      testWait<TestParams<UIntT, PtrType::DevicePtrToHost>, true>(testCase);
    }
#if HT_AMD
    if constexpr (std::is_same_v<UIntT, uint64_t>) {
      SECTION("Signal") {
        testWait<TestParams<UIntT, PtrType::Signal>, true>(testCase);
      }
    }
#endif
  }

  SECTION("NoMask_And") {
    TestWaitType testCase = std::is_same_v<UIntT, uint32_t>
      ? TestWaitType(hipStreamWaitValueAnd, 0x70F0F0F0, 0x0F0F0F0F, 0X1F0F0F0F)
      : TestWaitType(hipStreamWaitValueAnd, 0x70F0F0F0F0F0F0F0, 0x0F0F0F0F0F0F0F0F, 0X1F0F0F0F0F0F0F0F);

    SECTION("HostPtr") {
      testWait<TestParams<UIntT, PtrType::HostPtr>, true>(testCase);
    }
    SECTION("DevicePtr") {
      testWait<TestParams<UIntT, PtrType::DevicePtr>, true>(testCase);
    }
    SECTION("DevicePtrToHost") {
      testWait<TestParams<UIntT, PtrType::DevicePtrToHost>, true>(testCase);
    }
#if HT_AMD
    if constexpr (std::is_same_v<UIntT, uint64_t>) {
      SECTION("Signal") {
        testWait<TestParams<UIntT, PtrType::Signal>, true>(testCase);
      }
    }
#endif
  }

  SECTION("NoMask_Nor") {
    TestWaitType testCase = std::is_same_v<UIntT, uint32_t>
      ? TestWaitType(hipStreamWaitValueNor, 0x7AAAAAAA, 0x85555555, 0x9AAAAAAA)
      : TestWaitType(hipStreamWaitValueNor, 0x4724724747247247, 0xbddbddbdbddbddbd, 0xbddbddbdbddbddb3);

    SECTION("HostPtr") {
      testWait<TestParams<UIntT, PtrType::HostPtr>, true>(testCase);
    }
    SECTION("DevicePtr") {
      testWait<TestParams<UIntT, PtrType::DevicePtr>, true>(testCase);
    }
    SECTION("DevicePtrToHost") {
      testWait<TestParams<UIntT, PtrType::DevicePtrToHost>, true>(testCase);
    }
#if HT_AMD
    if constexpr (std::is_same_v<UIntT, uint64_t>) {
      SECTION("Signal") {
        testWait<TestParams<UIntT, PtrType::Signal>, true>(testCase);
      }
    }
#endif
  }
}

// Negative Tests
HIP_TEST_CASE(Unit_hipStreamValue_Negative_InvalidMemory) {
  if (!streamWaitValueSupported()) {
    HIP_SKIP_TEST(HipTest::SkipReason::kStreamWaitValueUnsupported);
  }

  hipStream_t stream{nullptr};
  HIP_CHECK(hipStreamCreate(&stream));
  REQUIRE(stream != nullptr);

  const auto compareOp = hipStreamWaitValueGte;
  const auto expectedError = hipErrorInvalidValue;

  // Memory pointer negative tests
  SECTION("Invalid Memory Pointer for hipStreamWriteValue32") {
    HIP_CHECK_ERROR(hipStreamWriteValue32(stream, nullptr, 0, writeFlag), expectedError);
  }
  SECTION("Invalid Memory Pointer for hipStreamWriteValue64") {
    HIP_CHECK_ERROR(hipStreamWriteValue64(stream, nullptr, 0, writeFlag), expectedError);
  }
  SECTION("Invalid Memory Pointer for hipStreamWaitValue32") {
    HIP_CHECK_ERROR(hipStreamWaitValue32(stream, nullptr, 0, compareOp), expectedError);
  }
  SECTION("Invalid Memory Pointer for hipStreamWaitValue32") {
    HIP_CHECK_ERROR(hipStreamWaitValue64(stream, nullptr, 0, compareOp), expectedError);
  }

  // Cleanup
  HIP_CHECK(hipStreamDestroy(stream));
}

// Merge the two similar negative tests
HIP_TEMPLATE_TEST_CASE(Unit_hipStreamValue_Negative_StreamAndFlag, uint32_t, uint64_t) {
  if (!streamWaitValueSupported()) {
    HIP_SKIP_TEST(HipTest::SkipReason::kStreamWaitValueUnsupported);
  }

  SECTION("Invalid Stream handle") {
    hipStream_t stream{reinterpret_cast<hipStream_t>(0xFFFF)};

    // Allocate Host Memory
    auto hostPtr = std::make_unique<TestType>();

    // Register Host Memory
    HIP_CHECK(hipHostRegister(hostPtr.get(), sizeof(TestType), 0));

    // Set dummy data
    *hostPtr = 0x0;

    const auto compareOp = hipStreamWaitValueGte;
    const auto expectedError = hipErrorContextIsDestroyed;

    SECTION("Invalid Stream handle for hipStreamWriteValue") {
      HIP_CHECK_ERROR(writeFunc<TestType>(stream, hostPtr.get(), 0, writeFlag), expectedError);
    }

    SECTION("Invalid Stream handle for hipStreamWaitValue") {
      HIP_CHECK_ERROR(waitFunc<TestType>(stream, hostPtr.get(), 0, compareOp), expectedError);
    }

    // Cleanup
    HIP_CHECK(hipHostUnregister(hostPtr.get()));
  }

  SECTION("Invalid Flag") {
    hipStream_t stream{nullptr};
    HIP_CHECK(hipStreamCreate(&stream));
    REQUIRE(stream != nullptr);

    // Allocate Host Memory
    auto hostPtr = std::make_unique<TestType>();

    // Register Host Memory
    HIP_CHECK(hipHostRegister(hostPtr.get(), sizeof(TestType), 0));

    // Set dummy data
    *hostPtr = 0x0;

    HIP_CHECK_ERROR(waitFunc<TestType>(stream, hostPtr.get(), 0, -1), hipErrorInvalidValue);

    // Cleanup
    HIP_CHECK(hipHostUnregister(hostPtr.get()));
    HIP_CHECK(hipStreamDestroy(stream));
  }
}

HIP_TEMPLATE_TEST_CASE(Unit_hipStreamWriteValue_Default, uint32_t, uint64_t) {
  if (!streamWaitValueSupported()) {
    HIP_SKIP_TEST(HipTest::SkipReason::kStreamWaitValueUnsupported);
  }

  hipStream_t stream{nullptr};
  HIP_CHECK(hipStreamCreate(&stream));
  REQUIRE(stream != nullptr);

  // Allocate Host Memory
  auto hostPtr = std::make_unique<TestType>();

  // Register Host Memory
  HIP_CHECK(hipHostRegister(hostPtr.get(), sizeof(TestType), 0));

  // Set dummy data
  *hostPtr = 0x0;

  HIP_CHECK(writeFunc<TestType>(stream, hostPtr.get(), 0, writeFlag));

  // Cleanup
  HIP_CHECK(hipHostUnregister(hostPtr.get()));
  HIP_CHECK(hipStreamDestroy(stream));
}

#if HT_AMD
TEMPLATE_TEST_CASE("Unit_hipStreamWriteValue_Increment_Default", "", uint32_t, uint64_t) {
  if (!streamWaitValueSupported()) {
    HIP_SKIP_TEST("hipStreamWriteValue not supported on this device.");
  }

  hipStream_t stream{nullptr};
  HIP_CHECK(hipStreamCreate(&stream));
  REQUIRE(stream != nullptr);

  TestType initial_value = 10;
  TestType* mem;
  HIP_CHECK(hipHostMalloc(&mem, sizeof(TestType), hipMallocSignalMemory));
  __atomic_store(mem, &initial_value, __ATOMIC_RELEASE);

  const TestType increment_value = 2;
  HIP_CHECK(writeFunc<TestType>(stream, mem, increment_value, hipExtStreamWriteValueIncrement));
  HIP_CHECK(hipStreamSynchronize(stream));

  const TestType expected_value = initial_value + increment_value;
  REQUIRE(*mem == expected_value);

  HIP_CHECK(hipStreamDestroy(stream));
  HIP_CHECK(hipHostFree(mem));
}

TEMPLATE_TEST_CASE("Unit_hipStreamWriteValue_Decrement_Default", "", uint32_t, uint64_t) {
  if (!streamWaitValueSupported()) {
    HIP_SKIP_TEST("hipStreamWriteValue not supported on this device.");
  }

  hipStream_t stream{nullptr};
  HIP_CHECK(hipStreamCreate(&stream));
  REQUIRE(stream != nullptr);

  TestType initial_value = 0;
  TestType* mem;
  HIP_CHECK(hipHostMalloc(&mem, sizeof(TestType), hipMallocSignalMemory));
  __atomic_store(mem, &initial_value, __ATOMIC_RELEASE);

  const TestType decrement_value = 10;
  HIP_CHECK(writeFunc<TestType>(stream, mem, decrement_value, hipExtStreamWriteValueDecrement));
  HIP_CHECK(hipStreamSynchronize(stream));

  const TestType expected_value = initial_value - decrement_value;
  REQUIRE(*mem == expected_value);
  HIP_CHECK(hipStreamDestroy(stream));
  HIP_CHECK(hipHostFree(mem));
}

template <typename TestType>
void testIncrementDecrementMultiStreamMultiDevice(uint32_t operationFlag) {
  if (!streamWaitValueSupported()) {
    HIP_SKIP_TEST("hipStreamWriteValue not supported on this device.");
  }

  constexpr size_t streams_per_device = 2;
  constexpr size_t iterations = 10;
  constexpr TestType increment_value = 2;
  TestType initial_value = 10;

  int deviceCount = 0;
  HIP_CHECK(hipGetDeviceCount(&deviceCount));

  std::vector<std::vector<hipStream_t>> streams(deviceCount);

  TestType* mem;
  HIP_CHECK(hipHostMalloc(&mem, sizeof(TestType), hipMallocSignalMemory));
  __atomic_store(mem, &initial_value, __ATOMIC_RELEASE);

  for (int stream_idx = 0; stream_idx < streams_per_device; ++stream_idx) {
    for (int device_idx = 0; device_idx < deviceCount; ++device_idx) {
      HIP_CHECK(hipSetDevice(device_idx));

      hipStream_t stream{nullptr};
      HIP_CHECK(hipStreamCreate(&stream));
      REQUIRE(stream != nullptr);

      streams[device_idx].push_back(stream);
    }
  }

  for (int cur_iter = 0; cur_iter < iterations; ++cur_iter) {
    for (int device_idx = 0; device_idx < deviceCount; ++device_idx) {
      HIP_CHECK(hipSetDevice(device_idx));
      for (int stream_idx = 0; stream_idx < streams_per_device; ++stream_idx) {
        HIP_CHECK(writeFunc<TestType>(streams[device_idx][stream_idx], mem, increment_value,
                                      operationFlag));
      }
    }
  }

  for (int device_idx = 0; device_idx < deviceCount; ++device_idx) {
    HIP_CHECK(hipSetDevice(device_idx));
    for (int stream_idx = 0; stream_idx < streams_per_device; ++stream_idx) {
      HIP_CHECK(hipStreamSynchronize(streams[device_idx][stream_idx]));
    }
  }

  TestType expected_value;
  switch (operationFlag) {
    case hipExtStreamWriteValueIncrement: {
      expected_value =
          initial_value + (increment_value * iterations * streams_per_device * deviceCount);
      break;
    }
    case hipExtStreamWriteValueDecrement: {
      expected_value =
          initial_value - (increment_value * iterations * streams_per_device * deviceCount);
      break;
    }
    default: {
      assert(false);
      break;
    }
  }

  REQUIRE(*mem == expected_value);

  for (int device_idx = 0; device_idx < deviceCount; ++device_idx) {
    HIP_CHECK(hipSetDevice(device_idx));
    for (int stream_idx = 0; stream_idx < streams_per_device; ++stream_idx) {
      HIP_CHECK(hipStreamDestroy(streams[device_idx][stream_idx]));
    }
  }

  HIP_CHECK(hipHostFree(mem));
}

TEMPLATE_TEST_CASE("Unit_hipStreamWriteValue_Increment_MultiStream_MultiDevice", "", uint32_t,
                   uint64_t) {
  testIncrementDecrementMultiStreamMultiDevice<TestType>(hipExtStreamWriteValueIncrement);
}


TEMPLATE_TEST_CASE("Unit_hipStreamWriteValue_Decrement_MultiStream_MultiDevice", "", uint32_t,
                   uint64_t) {
  testIncrementDecrementMultiStreamMultiDevice<TestType>(hipExtStreamWriteValueDecrement);
}
#endif

template <typename T> __global__ void add(T* a, T* b, T* c, size_t size) {
  size_t i = threadIdx.x;
  if (i < size) c[i] = a[i] + b[i];
}

HIP_TEMPLATE_TEST_CASE(Unit_hipStreamWaitValue_Default, uint32_t, uint64_t) {
  if (!streamWaitValueSupported()) {
    HIP_SKIP_TEST(HipTest::SkipReason::kStreamWaitValueUnsupported);
  }

  auto size = GENERATE(as<size_t>{}, 100, 500, 1000);
  // Create test data
  std::vector<uint32_t> a(size), b(size), c(size);
  for (size_t i = 0; i < size; i++) {
    a[i] = b[i] = (i + 1);
    c[i] = a[i] + b[i];
  }

  // Allocate IO memory to be used in kernel
  uint32_t *d_a, *d_b, *d_c;
  HIP_CHECK(hipMalloc(&d_a, sizeof(uint32_t) * size));
  HIP_CHECK(hipMalloc(&d_b, sizeof(uint32_t) * size));
  HIP_CHECK(hipMalloc(&d_c, sizeof(uint32_t) * size));

  // Create 2 separate streams. Once would be used for data copy, other will be used to execute
  // kernel
  hipStream_t dataCopyStream{nullptr};
  hipStream_t kernelExecStream{nullptr};
  HIP_CHECK(hipStreamCreate(&dataCopyStream));
  HIP_CHECK(hipStreamCreate(&kernelExecStream));
  REQUIRE(dataCopyStream != nullptr);
  REQUIRE(kernelExecStream != nullptr);

  // Create memory for streams to operate on. Kernel execution stream will wait for data stream to
  // finish copies
  auto hostPtr = std::make_unique<TestType>();
  HIP_CHECK(hipHostRegister(hostPtr.get(), sizeof(TestType), 0));
  *hostPtr = 0x0;

  // Perform copy operations on dataCopyStream
  HIP_CHECK(hipMemcpyAsync(d_a, a.data(), sizeof(uint32_t) * size, hipMemcpyHostToDevice,
                           dataCopyStream));
  HIP_CHECK(hipMemcpyAsync(d_b, b.data(), sizeof(uint32_t) * size, hipMemcpyHostToDevice,
                           dataCopyStream));
  HIP_CHECK(writeFunc<TestType>(dataCopyStream, hostPtr.get(), 1, writeFlag));

  // Perform operations on kernelExecStream
  HIP_CHECK(waitFunc<TestType>(kernelExecStream, hostPtr.get(), 1, hipStreamWaitValueGte));
  add<<<1, size, 0, kernelExecStream>>>(d_a, d_b, d_c, size);
  HIP_CHECK(hipGetLastError());

  // Synchronize both streams, copy output to host and synchronize globally
  HIP_CHECK(hipStreamSynchronize(dataCopyStream));
  HIP_CHECK(hipStreamSynchronize(kernelExecStream));

  HIP_CHECK(hipMemcpyAsync(a.data(), d_c, sizeof(uint32_t) * size, hipMemcpyDeviceToHost,
                           dataCopyStream));

  HIP_CHECK(hipDeviceSynchronize());

  // Resource cleanup and validation
  HIP_CHECK(hipFree(d_a));
  HIP_CHECK(hipFree(d_b));
  HIP_CHECK(hipFree(d_c));
  HIP_CHECK(hipHostUnregister(hostPtr.get()));
  HIP_CHECK(hipStreamDestroy(dataCopyStream));
  HIP_CHECK(hipStreamDestroy(kernelExecStream));
  REQUIRE(a == c);
}