# Copyright (c) Microsoft Corporation.
# Licensed under the MIT License.

"""clip_grads module in msamp.megatron package."""

import torch
from torch import inf
import amp_C
from apex.multi_tensor_apply import multi_tensor_applier


def clip_grad_norm_fp32(    # noqa: C901
    parameters, grads_for_norm, max_norm, norm_type=2, model_parallel_group=None
):
    """Clips gradient norm of an iterable of parameters whose gradients are in fp32 and fp8.

    This is adapted from torch.nn.utils.clip_grad.clip_grad_norm_ and
    added functionality to handle model parallel parameters. Note that
    the gradients are modified in place.

    Arguments:
        parameters (Iterable[Tensor] or Tensor): an iterable of Tensors or a
            single Tensor that will have gradients normalized
        grads_for_norm (Iterable[Tensor]): an iterable of Tensors or a single
            Tensor that will be used for calculating the grad norm.
        max_norm (float or int): max norm of the gradients
        norm_type (float or int): type of the used p-norm. Can be ``'inf'`` for
            infinity norm.
        model_parallel_group (group): given the nature of the distributed
            optimizer, this is passed as an argument.

    Returns:
        Total norm of the parameters (viewed as a single vector).
    """
    if isinstance(parameters, torch.Tensor):
        parameters = [parameters]
    if isinstance(grads_for_norm, torch.Tensor):
        grads_for_norm = [grads_for_norm]

    # Grads.
    grads = []
    fp8_grads = []
    for param in parameters:
        if param.grad is not None:
            if torch.is_tensor(param):
                assert param.grad.type() == 'torch.cuda.FloatTensor'
                grads.append(param.grad.detach())
            else:
                fp8_grads.append(param.grad)

    # Norm parameters.
    max_norm = float(max_norm)
    norm_type = float(norm_type)
    total_norm = 0.0

    # Calculate norm.
    if norm_type == inf:
        total_norm = max(grad.abs().max() for grad in grads_for_norm)
        total_norm_cuda = torch.cuda.FloatTensor([float(total_norm)])
        # Take max across all model-parallel GPUs.
        torch.distributed.all_reduce(total_norm_cuda, op=torch.distributed.ReduceOp.MAX, group=model_parallel_group)
        total_norm = total_norm_cuda[0].item()

    else:
        if norm_type == 2.0:
            fp8_grads_for_norm = []
            torch_grads_for_norm = []
            for grad in grads_for_norm:
                if torch.is_tensor(grad):
                    torch_grads_for_norm.append(grad)
                else:
                    fp8_grads_for_norm.append(grad)

            dummy_overflow_buf = torch.cuda.IntTensor([0])
            # Use apex's multi-tensor applier for efficiency reasons.
            # Multi-tensor applier takes a function and a list of list
            # and performs the operation on that list all in one kernel.
            if torch_grads_for_norm:
                grad_norm, _ = multi_tensor_applier(
                    amp_C.multi_tensor_l2norm,
                    dummy_overflow_buf,
                    [torch_grads_for_norm],
                    False    # no per-parameter norm
                )
            else:
                grad_norm = torch.cuda.FloatTensor([0])
            # Since we will be summing across data parallel groups,
            # we need the pow(norm-type).
            total_norm = grad_norm**norm_type

            for grad in fp8_grads_for_norm:
                grad_norm = torch.norm(grad.float(), norm_type)
                total_norm += grad_norm**norm_type
        else:
            for grad in grads_for_norm:
                grad_norm = torch.norm(grad, norm_type)
                total_norm += grad_norm**norm_type

        # Sum across all model-parallel GPUs.
        torch.distributed.all_reduce(total_norm, op=torch.distributed.ReduceOp.SUM, group=model_parallel_group)
        total_norm = total_norm.item()**(1.0 / norm_type)

    # Scale.
    clip_coeff = max_norm / (total_norm + 1.0e-6)
    if clip_coeff < 1.0:
        dummy_overflow_buf = torch.cuda.IntTensor([0])
        multi_tensor_applier(amp_C.multi_tensor_scale, dummy_overflow_buf, [grads, grads], clip_coeff)

    return total_norm