blender/intern/cycles/device/hip/queue.cpp

/* SPDX-FileCopyrightText: 2011-2022 Blender Foundation
 *
 * SPDX-License-Identifier: Apache-2.0 */

#ifdef WITH_HIP

#  include "device/hip/queue.h"

#  include "device/hip/device_impl.h"
#  include "device/hip/graphics_interop.h"
#  include "device/hip/kernel.h"

CCL_NAMESPACE_BEGIN

/* HIPDeviceQueue */

HIPDeviceQueue::HIPDeviceQueue(HIPDevice *device)
    : DeviceQueue(device), hip_device_(device), hip_stream_(nullptr)
{
  const HIPContextScope scope(hip_device_);
  hip_device_assert(hip_device_, hipStreamCreateWithFlags(&hip_stream_, hipStreamNonBlocking));
}

HIPDeviceQueue::~HIPDeviceQueue()
{
  const HIPContextScope scope(hip_device_);
  hipStreamDestroy(hip_stream_);
}

ConcurrentStatesParams HIPDeviceQueue::concurrent_states_params() const
{
  const int max_num_threads = hip_device_->get_num_multiprocessors() *
                              hip_device_->get_max_num_threads_per_multiprocessor();

  /* Same as CUDA. */
  ConcurrentStatesParams params;
  params.baseline = max(max_num_threads, 65536) * 16;
  params.min = params.baseline / 2;
  params.max = params.baseline * 4;
  return params;
}

void HIPDeviceQueue::get_memory_info(size_t &total, size_t &free) const
{
  hip_device_->get_device_memory_info(total, free);
}

void HIPDeviceQueue::init_execution()
{
  /* Synchronize all textures and memory copies before executing task. */
  HIPContextScope scope(hip_device_);
  hip_device_->load_image_info(nullptr);
  hip_device_assert(hip_device_, hipDeviceSynchronize());

  debug_init_execution();
}

void HIPDeviceQueue::load_image_info()
{
  HIPContextScope scope(hip_device_);
  hip_device_->load_image_info(this);
}

bool HIPDeviceQueue::enqueue(DeviceKernel kernel,
                             const int work_size,
                             const DeviceKernelArguments &args)
{
  if (hip_device_->have_error()) {
    return false;
  }

  debug_enqueue_begin(kernel, work_size);

  const HIPContextScope scope(hip_device_);

  /* Update image info in case memory moved to host. */
  if (hip_device_->load_image_info(nullptr)) {
    hip_device_assert(hip_device_, hipDeviceSynchronize());
    if (hip_device_->have_error()) {
      return false;
    }
  }

  /* Compute kernel launch parameters. */
  const HIPDeviceKernel &hip_kernel = hip_device_->kernels.get(kernel);
  const int num_threads_per_block = hip_kernel.num_threads_per_block;
  const int num_blocks = divide_up(work_size, num_threads_per_block);

  int shared_mem_bytes = 0;

  switch (kernel) {
    case DEVICE_KERNEL_INTEGRATOR_QUEUED_PATHS_ARRAY:
    case DEVICE_KERNEL_INTEGRATOR_QUEUED_SHADOW_PATHS_ARRAY:
    case DEVICE_KERNEL_INTEGRATOR_ACTIVE_PATHS_ARRAY:
    case DEVICE_KERNEL_INTEGRATOR_TERMINATED_PATHS_ARRAY:
    case DEVICE_KERNEL_INTEGRATOR_SORTED_PATHS_ARRAY:
    case DEVICE_KERNEL_INTEGRATOR_COMPACT_PATHS_ARRAY:
    case DEVICE_KERNEL_INTEGRATOR_TERMINATED_SHADOW_PATHS_ARRAY:
    case DEVICE_KERNEL_INTEGRATOR_COMPACT_SHADOW_PATHS_ARRAY:
      /* See parall_active_index.h for why this amount of shared memory is needed. */
      shared_mem_bytes = (num_threads_per_block + 1) * sizeof(int);
      break;
    default:
      break;
  }

  /* Launch kernel. */
  assert_success(hipModuleLaunchKernel(hip_kernel.function,
                                       num_blocks,
                                       1,
                                       1,
                                       num_threads_per_block,
                                       1,
                                       1,
                                       shared_mem_bytes,
                                       hip_stream_,
                                       const_cast<void **>(args.values),
                                       nullptr),
                 "enqueue");

  debug_enqueue_end();

  return !(hip_device_->have_error());
}

bool HIPDeviceQueue::synchronize()
{
  if (hip_device_->have_error()) {
    return false;
  }

  const HIPContextScope scope(hip_device_);
  assert_success(hipStreamSynchronize(hip_stream_), "synchronize");
  debug_synchronize();

  return !(hip_device_->have_error());
}

void HIPDeviceQueue::zero_to_device(device_memory &mem)
{
  assert(mem.type != MEM_IMAGE_TEXTURE);

  if (mem.memory_size() == 0) {
    return;
  }

  /* Allocate on demand. */
  if (mem.device_pointer == 0) {
    if (mem.type == MEM_GLOBAL) {
      hip_device_->global_alloc(mem);
    }
    else {
      hip_device_->mem_alloc(mem);
    }
  }

  /* Zero memory on device. */
  device_ptr d_ptr = mem.device->mem_device_ptr(mem, hip_device_);
  assert(d_ptr != 0);

  const HIPContextScope scope(hip_device_);
  assert_success(hipMemsetD8Async((hipDeviceptr_t)d_ptr, 0, mem.memory_size(), hip_stream_),
                 "zero_to_device");
}

void HIPDeviceQueue::copy_to_device(device_memory &mem)
{
  assert(mem.type != MEM_IMAGE_TEXTURE);

  if (mem.memory_size() == 0) {
    return;
  }

  /* Allocate on demand. */
  if (mem.device_pointer == 0) {
    if (mem.type == MEM_GLOBAL) {
      hip_device_->global_alloc(mem);
    }
    else {
      hip_device_->mem_alloc(mem);
    }
  }

  device_ptr d_ptr = mem.device->mem_device_ptr(mem, hip_device_);
  assert(d_ptr != 0);
  assert(mem.host_pointer != nullptr);

  /* Copy memory to device. */
  const HIPContextScope scope(hip_device_);
  assert_success(
      hipMemcpyHtoDAsync((hipDeviceptr_t)d_ptr, mem.host_pointer, mem.memory_size(), hip_stream_),
      "copy_to_device");
}

void HIPDeviceQueue::copy_from_device(device_memory &mem)
{
  assert(mem.type != MEM_GLOBAL && mem.type != MEM_IMAGE_TEXTURE);

  if (mem.memory_size() == 0) {
    return;
  }

  assert(mem.device_pointer != 0);
  assert(mem.host_pointer != nullptr);

  /* Copy memory from device. */
  const HIPContextScope scope(hip_device_);
  assert_success(
      hipMemcpyDtoHAsync(
          mem.host_pointer, (hipDeviceptr_t)mem.device_pointer, mem.memory_size(), hip_stream_),
      "copy_from_device");
}

void *HIPDeviceQueue::copy_from_device_synchronized(device_memory &mem, vector<uint8_t> &storage)
{
  if (mem.memory_size() == 0) {
    return nullptr;
  }

  storage.resize(mem.memory_size());

  device_ptr d_ptr = mem.device->mem_device_ptr(mem, hip_device_);
  assert(d_ptr != 0);

  const HIPContextScope scope(hip_device_);
  assert_success(
      hipMemcpyDtoHAsync(storage.data(), (hipDeviceptr_t)d_ptr, mem.memory_size(), hip_stream_),
      "copy_from_device_synchronized");

  synchronize();
  return storage.data();
}

void HIPDeviceQueue::assert_success(hipError_t result, const char *operation)
{
  if (result != hipSuccess) {
    const char *name = hipewErrorString(result);
    hip_device_->set_error(
        string_printf("%s in HIP queue %s (%s)", name, operation, debug_active_kernels().c_str()));
  }
}

unique_ptr<DeviceGraphicsInterop> HIPDeviceQueue::graphics_interop_create()
{
  return make_unique<HIPDeviceGraphicsInterop>(this);
}

CCL_NAMESPACE_END

#endif /* WITH_HIP */