TensorRT-LLMs/cpp/tensorrt_llm/kernels/moe_utils.cuh
Yihan Wang 9df4dad3b6
[None][fix] Introduce inline namespace to avoid symbol collision (#9541)
Signed-off-by: Yihan Wang <yihwang@nvidia.com>
2025-12-12 23:32:15 +08:00

51 lines
1.4 KiB
Plaintext

/*
* SPDX-FileCopyrightText: Copyright (c) 2025 NVIDIA CORPORATION & AFFILIATES. All rights reserved.
* SPDX-License-Identifier: Apache-2.0
*
* 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.
*/
#pragma once
#include "tensorrt_llm/common/config.h"
TRTLLM_NAMESPACE_BEGIN
namespace kernels
{
// TODO Could linear search be better for small # experts
template <class T>
__device__ inline int64_t findTotalEltsLessThanTarget(T const* sorted_indices, int64_t const arr_length, T const target)
{
int64_t low = 0, high = arr_length - 1, target_location = -1;
while (low <= high)
{
int64_t mid = (low + high) / 2;
if (sorted_indices[mid] >= target)
{
high = mid - 1;
}
else
{
low = mid + 1;
target_location = mid;
}
}
return target_location + 1;
}
} // namespace kernels
TRTLLM_NAMESPACE_END