Skip to content
体验新版
项目
组织
正在加载...
登录
切换导航
打开侧边栏
PaddlePaddle
Paddle
提交
95d3ebc8
P
Paddle
项目概览
PaddlePaddle
/
Paddle
大约 1 年 前同步成功
通知
2298
Star
20931
Fork
5422
代码
文件
提交
分支
Tags
贡献者
分支图
Diff
Issue
1423
列表
看板
标记
里程碑
合并请求
543
Wiki
0
Wiki
分析
仓库
DevOps
项目成员
Pages
P
Paddle
项目概览
项目概览
详情
发布
仓库
仓库
文件
提交
分支
标签
贡献者
分支图
比较
Issue
1,423
Issue
1,423
列表
看板
标记
里程碑
合并请求
543
合并请求
543
Pages
分析
分析
仓库分析
DevOps
Wiki
0
Wiki
成员
成员
收起侧边栏
关闭侧边栏
动态
分支图
创建新Issue
提交
Issue看板
未验证
提交
95d3ebc8
编写于
3月 23, 2022
作者:
N
niuliling123
提交者:
GitHub
3月 23, 2022
浏览文件
操作
浏览文件
下载
电子邮件补丁
差异文件
Modified dropout Kernel with Kernel Primitive API (#40766)
上级
17b8335b
变更
3
显示空白变更内容
内联
并排
Showing
3 changed file
with
121 addition
and
157 deletion
+121
-157
paddle/fluid/operators/dropout_impl.cu.h
paddle/fluid/operators/dropout_impl.cu.h
+103
-152
paddle/phi/kernels/funcs/distribution_helper.h
paddle/phi/kernels/funcs/distribution_helper.h
+16
-2
paddle/phi/kernels/gpu/masked_select_grad_kernel.cu
paddle/phi/kernels/gpu/masked_select_grad_kernel.cu
+2
-3
未找到文件。
paddle/fluid/operators/dropout_impl.cu.h
浏览文件 @
95d3ebc8
...
@@ -35,143 +35,99 @@ limitations under the License. */
...
@@ -35,143 +35,99 @@ limitations under the License. */
#include "paddle/fluid/operators/elementwise/elementwise_op_impl.cu.h"
#include "paddle/fluid/operators/elementwise/elementwise_op_impl.cu.h"
#include "paddle/fluid/platform/aligned_vector.h"
#include "paddle/fluid/platform/aligned_vector.h"
#include "paddle/phi/backends/gpu/gpu_launch_config.h"
#include "paddle/phi/backends/gpu/gpu_launch_config.h"
#include "paddle/phi/kernels/funcs/distribution_helper.h"
#include "paddle/phi/kernels/funcs/functors.h"
#include "paddle/phi/kernels/funcs/functors.h"
namespace
paddle
{
namespace
paddle
{
namespace
operators
{
namespace
operators
{
template
<
typename
T1
,
typename
T2
=
T1
,
typename
OutT
=
T1
>
struct
DstMaskGenerator
{
const
float
dropout_prob_
;
const
bool
is_upscale_in_train_
;
using
MT
=
typename
details
::
MPTypeTrait
<
T1
>::
Type
;
MT
factor
;
HOSTDEVICE
inline
DstMaskGenerator
(
const
float
dropout_prob
,
const
bool
is_upscale_in_train
)
:
dropout_prob_
(
dropout_prob
),
is_upscale_in_train_
(
is_upscale_in_train
)
{
factor
=
static_cast
<
MT
>
(
1.0
f
/
(
1.0
f
-
dropout_prob_
));
}
template
<
typename
T
,
typename
MaskType
>
HOSTDEVICE
inline
void
operator
()(
OutT
*
dst
,
const
T1
*
src_val
,
__global__
void
RandomGenerator
(
const
size_t
n
,
uint64_t
seed
,
const
T2
*
rand
,
int
num
)
const
{
const
float
dropout_prob
,
const
T
*
src
,
static
constexpr
int
kCount
=
MaskType
*
mask
,
T
*
dst
,
phi
::
funcs
::
uniform_distribution
<
T2
>::
kReturnsCount
;
bool
is_upscale_in_train
,
uint64_t
increment
)
{
// 0 ~ kCount -1 is dist , kCount ~ 2 * kCount - 1 is mask
using
MT
=
typename
details
::
MPTypeTrait
<
T
>::
Type
;
#pragma unroll
int
idx
=
blockDim
.
x
*
blockIdx
.
x
+
threadIdx
.
x
;
for
(
int
i
=
0
;
i
<
kCount
;
i
++
)
{
#ifdef PADDLE_WITH_HIP
if
(
rand
[
i
]
<
dropout_prob_
)
{
hiprandStatePhilox4_32_10_t
state
;
dst
[
i
]
=
static_cast
<
T1
>
(
0
);
hiprand_init
(
seed
,
idx
,
increment
,
&
state
);
dst
[
i
+
kCount
]
=
dst
[
i
];
#else
curandStatePhilox4_32_10_t
state
;
curand_init
(
seed
,
idx
,
increment
,
&
state
);
#endif
MaskType
mask_val
;
T
dst_val
;
MT
factor
=
static_cast
<
MT
>
(
1.0
f
/
(
1.0
f
-
dropout_prob
));
for
(;
idx
<
n
;
idx
+=
blockDim
.
x
*
gridDim
.
x
)
{
T
src_val
=
src
[
idx
];
#ifdef PADDLE_WITH_HIP
if
(
hiprand_uniform
(
&
state
)
<
dropout_prob
)
{
#else
if
(
curand_uniform
(
&
state
)
<
dropout_prob
)
{
#endif
mask_val
=
0
;
dst_val
=
0
;
}
else
{
}
else
{
mask_val
=
1
;
dst
[
i
]
=
is_upscale_in_train_
dst_val
=
is_upscale_in_train
?
static_cast
<
T1
>
(
static_cast
<
MT
>
(
src_val
[
i
])
*
factor
)
?
static_cast
<
T
>
(
static_cast
<
MT
>
(
src_val
)
*
factor
)
:
static_cast
<
T1
>
(
src_val
[
i
]);
:
src_val
;
dst
[
i
+
kCount
]
=
static_cast
<
T1
>
(
1
)
;
}
}
mask
[
idx
]
=
mask_val
;
dst
[
idx
]
=
dst_val
;
}
}
}
}
};
template
<
typename
T
,
typename
MaskType
,
int
VecSize
>
template
<
typename
T
,
typename
MaskType
>
__global__
void
VectorizedRandomGenerator
(
const
size_t
n
,
uint64_t
seed
,
__global__
void
VectorizedRandomGenerator
(
const
size_t
n
,
uint64_t
seed
,
const
float
dropout_prob
,
const
float
dropout_prob
,
const
T
*
src
,
MaskType
*
mask
,
T
*
dst
,
const
T
*
src
,
MaskType
*
mask
,
T
*
dst
,
bool
is_upscale_in_train
,
bool
is_upscale_in_train
,
uint64_t
increment
)
{
uint64_t
increment
,
using
MT
=
typename
details
::
MPTypeTrait
<
T
>::
Type
;
size_t
main_offset
)
{
using
LoadT
=
phi
::
AlignedVector
<
T
,
VecSize
>
;
size_t
idx
=
static_cast
<
size_t
>
(
BLOCK_ID_X
*
BLOCK_NUM_X
);
using
MaskLoadT
=
phi
::
AlignedVector
<
MaskType
,
VecSize
>
;
static
constexpr
int
kCount
=
phi
::
funcs
::
uniform_distribution
<
float
>::
kReturnsCount
;
size_t
stride
=
BLOCK_NUM_X
*
GRID_NUM_X
*
kCount
;
#ifdef PADDLE_WITH_HIP
#ifdef PADDLE_WITH_HIP
int64_t
idx
=
hipBlockDim_x
*
hipBlockIdx_x
+
hipThreadIdx_x
;
hiprandStatePhilox4_32_10_t
state
;
hiprandStatePhilox4_32_10_t
state
;
hiprand_init
(
seed
,
idx
,
increment
,
&
state
);
hiprand_init
(
seed
,
idx
+
THREAD_ID_X
,
increment
,
&
state
);
using
SType
=
hiprandStatePhilox4_32_10_t
;
#else
#else
int64_t
idx
=
blockDim
.
x
*
blockIdx
.
x
+
threadIdx
.
x
;
curandStatePhilox4_32_10_t
state
;
curandStatePhilox4_32_10_t
state
;
curand_init
(
seed
,
idx
,
increment
,
&
state
);
curand_init
(
seed
,
idx
+
THREAD_ID_X
,
increment
,
&
state
);
using
SType
=
curandStatePhilox4_32_10_t
;
#endif
#endif
T
dst_mask
[
kCount
*
2
];
// 0 ~ kCount -1 : dst;kCount ~ 2 * kCount - 1: mask
MT
factor
=
static_cast
<
MT
>
(
1.0
f
/
(
1.0
f
-
dropout_prob
));
float
rands
[
kCount
];
for
(
int
i
=
idx
*
VecSize
;
i
<
n
;
i
+=
blockDim
.
x
*
gridDim
.
x
*
VecSize
)
{
MaskType
mask_result
[
kCount
];
LoadT
src_val
;
using
Rand
=
phi
::
funcs
::
uniform_distribution
<
float
>
;
phi
::
Load
<
T
,
VecSize
>
(
&
src
[
i
],
&
src_val
);
using
Cast
=
kps
::
IdentityFunctor
<
T
>
;
int
deal_size
=
BLOCK_NUM_X
*
kCount
;
#ifdef PADDLE_WITH_HIP
auto
dst_functor
=
float4
rand
=
hiprand_uniform4
(
&
state
);
DstMaskGenerator
<
T
,
float
>
(
dropout_prob
,
is_upscale_in_train
);
#else
size_t
fix
=
idx
*
kCount
;
float4
rand
=
curand_uniform4
(
&
state
);
for
(;
fix
<
main_offset
;
fix
+=
stride
)
{
#endif
kps
::
ReadData
<
T
,
kCount
,
1
,
1
,
false
>
(
&
dst_mask
[
0
],
src
+
fix
,
deal_size
);
kps
::
ElementwiseRandom
<
SType
,
float
,
kCount
,
1
,
Rand
>
(
&
rands
[
0
],
Rand
(),
LoadT
dst_val
;
&
state
);
MaskLoadT
mask_val
;
// dst
kps
::
OperatorTernary
<
T
,
float
,
T
,
DstMaskGenerator
<
T
,
float
>>
(
#pragma unroll
&
dst_mask
[
0
],
&
dst_mask
[
0
],
&
rands
[
0
],
dst_functor
,
kCount
);
for
(
int
j
=
0
;
j
<
VecSize
;
j
++
)
{
kps
::
WriteData
<
T
,
kCount
,
1
,
1
,
false
>
(
dst
+
fix
,
&
dst_mask
[
0
],
deal_size
);
if
((
&
rand
.
x
)[
j
]
<
dropout_prob
)
{
// mask
dst_val
[
j
]
=
0
;
kps
::
ElementwiseUnary
<
T
,
MaskType
,
kCount
,
1
,
1
,
Cast
>
(
mask_val
[
j
]
=
0
;
&
mask_result
[
0
],
&
dst_mask
[
kCount
],
Cast
());
}
else
{
kps
::
WriteData
<
MaskType
,
kCount
,
1
,
1
,
false
>
(
mask
+
fix
,
&
mask_result
[
0
],
dst_val
[
j
]
=
is_upscale_in_train
deal_size
);
?
static_cast
<
T
>
(
static_cast
<
MT
>
(
src_val
[
j
])
*
factor
)
:
src_val
[
j
];
mask_val
[
j
]
=
1
;
}
}
phi
::
Store
<
T
,
VecSize
>
(
dst_val
,
&
dst
[
i
]);
phi
::
Store
<
MaskType
,
VecSize
>
(
mask_val
,
&
mask
[
i
]);
}
}
template
<
typename
T
,
typename
MaskType
>
struct
CudaDropoutGradFunctor
{
using
MT
=
typename
details
::
MPTypeTrait
<
T
>::
Type
;
explicit
CudaDropoutGradFunctor
(
const
MT
factor
)
:
factor_
(
factor
)
{}
__device__
__forceinline__
T
operator
()(
const
T
dout
,
const
MaskType
mask
)
const
{
return
static_cast
<
T
>
(
static_cast
<
MT
>
(
dout
)
*
static_cast
<
MT
>
(
mask
)
*
factor_
);
}
}
int
remainder
=
n
-
fix
;
private:
if
(
remainder
>
0
)
{
MT
factor_
;
kps
::
ReadData
<
T
,
kCount
,
1
,
1
,
true
>
(
&
dst_mask
[
0
],
src
+
fix
,
remainder
);
};
kps
::
ElementwiseRandom
<
SType
,
float
,
kCount
,
1
,
Rand
>
(
&
rands
[
0
],
Rand
(),
&
state
);
template
<
typename
T
,
typename
MaskType
,
int
VecSize
>
// dst
__global__
void
DropoutGradCUDAKernel
(
kps
::
OperatorTernary
<
T
,
float
,
T
,
DstMaskGenerator
<
T
,
float
>>
(
const
T
*
dout
,
const
MaskType
*
mask
,
&
dst_mask
[
0
],
&
dst_mask
[
0
],
&
rands
[
0
],
dst_functor
,
kCount
);
const
typename
details
::
MPTypeTrait
<
T
>::
Type
factor
,
const
int64_t
size
,
kps
::
WriteData
<
T
,
kCount
,
1
,
1
,
true
>
(
dst
+
fix
,
&
dst_mask
[
0
],
remainder
);
T
*
dx
)
{
// mask
using
MT
=
typename
details
::
MPTypeTrait
<
T
>::
Type
;
kps
::
ElementwiseUnary
<
T
,
MaskType
,
kCount
,
1
,
1
,
Cast
>
(
using
LoadT
=
phi
::
AlignedVector
<
T
,
VecSize
>
;
&
mask_result
[
0
],
&
dst_mask
[
kCount
],
Cast
());
using
MaskLoadT
=
phi
::
AlignedVector
<
MaskType
,
VecSize
>
;
kps
::
WriteData
<
MaskType
,
kCount
,
1
,
1
,
true
>
(
mask
+
fix
,
&
mask_result
[
0
],
remainder
);
int64_t
idx
=
blockDim
.
x
*
blockIdx
.
x
+
threadIdx
.
x
;
for
(
int
i
=
idx
*
VecSize
;
i
<
size
;
i
+=
blockDim
.
x
*
gridDim
.
x
*
VecSize
)
{
LoadT
dout_val
;
phi
::
Load
<
T
,
VecSize
>
(
&
dout
[
i
],
&
dout_val
);
MaskLoadT
mask_val
;
phi
::
Load
<
MaskType
,
VecSize
>
(
&
mask
[
i
],
&
mask_val
);
LoadT
dx_val
;
#pragma unroll
for
(
int
j
=
0
;
j
<
VecSize
;
j
++
)
{
dx_val
[
j
]
=
static_cast
<
T
>
(
static_cast
<
MT
>
(
dout_val
[
j
])
*
static_cast
<
MT
>
(
mask_val
[
j
])
*
factor
);
}
phi
::
Store
<
T
,
VecSize
>
(
dx_val
,
&
dx
[
i
]);
}
}
}
}
...
@@ -218,42 +174,21 @@ void DropoutFwGPUKernelDriver(const phi::GPUContext& dev_ctx, bool is_test,
...
@@ -218,42 +174,21 @@ void DropoutFwGPUKernelDriver(const phi::GPUContext& dev_ctx, bool is_test,
uint64_t
seed_data
;
uint64_t
seed_data
;
uint64_t
increment
;
uint64_t
increment
;
// VectorizedRandomGenerator use curand_uniform4, so we only support
// VectorizedRandomGenerator use curand_uniform4, so we only support
// vec_size is 4;
// kVecSize is 4;
int
vec_size
=
(
phi
::
GetVectorizedSize
<
T
>
(
x_data
)
==
4
)
?
4
:
1
;
constexpr
int
kVecSize
=
phi
::
funcs
::
uniform_distribution
<
float
>::
kReturnsCount
;
auto
gpu_config
=
auto
gpu_config
=
phi
::
backends
::
gpu
::
GetGpuLaunchConfig1D
(
dev_ctx
,
x_numel
,
vec_s
ize
);
phi
::
backends
::
gpu
::
GetGpuLaunchConfig1D
(
dev_ctx
,
x_numel
,
kVecS
ize
);
auto
offset
=
auto
offset
=
((
x_numel
-
1
)
/
(
gpu_config
.
GetThreadNum
()
*
vec_size
)
+
1
)
*
vec_size
;
((
x_numel
-
1
)
/
(
gpu_config
.
GetThreadNum
()
*
kVecSize
)
+
1
)
*
kVecSize
;
GetSeedDataAndIncrement
(
dev_ctx
,
seed
,
is_fix_seed
,
seed_val
,
offset
,
GetSeedDataAndIncrement
(
dev_ctx
,
seed
,
is_fix_seed
,
seed_val
,
offset
,
&
seed_data
,
&
increment
);
&
seed_data
,
&
increment
);
size_t
main_offset
=
size
/
(
gpu_config
.
GetBlockSize
()
*
kVecSize
)
*
#ifdef __HIPCC__
(
gpu_config
.
GetBlockSize
()
*
kVecSize
);
if
(
vec_size
==
4
&&
size
%
4
==
0
)
{
VectorizedRandomGenerator
<
T
,
uint8_t
><<<
hipLaunchKernelGGL
(
gpu_config
.
GetGridSize
(),
gpu_config
.
GetBlockSize
(),
0
,
stream
>>>
(
HIP_KERNEL_NAME
(
VectorizedRandomGenerator
<
T
,
uint8_t
,
4
>
),
gpu_config
.
GetGridSize
(),
gpu_config
.
GetBlockSize
(),
0
,
stream
,
size
,
seed_data
,
dropout_prob
,
x_data
,
mask_data
,
y_data
,
upscale_in_train
,
increment
);
}
else
{
hipLaunchKernelGGL
(
HIP_KERNEL_NAME
(
RandomGenerator
<
T
,
uint8_t
>
),
gpu_config
.
GetGridSize
(),
gpu_config
.
GetBlockSize
(),
0
,
stream
,
size
,
seed_data
,
dropout_prob
,
x_data
,
mask_data
,
y_data
,
upscale_in_train
,
increment
);
}
#else
if
(
vec_size
==
4
&&
size
%
4
==
0
)
{
VectorizedRandomGenerator
<
T
,
uint8_t
,
4
><<<
gpu_config
.
block_per_grid
,
gpu_config
.
thread_per_block
,
0
,
stream
>>>
(
size
,
seed_data
,
dropout_prob
,
x_data
,
mask_data
,
y_data
,
upscale_in_train
,
increment
);
}
else
{
RandomGenerator
<
T
,
uint8_t
><<<
gpu_config
.
block_per_grid
,
gpu_config
.
thread_per_block
,
0
,
stream
>>>
(
size
,
seed_data
,
dropout_prob
,
x_data
,
mask_data
,
y_data
,
size
,
seed_data
,
dropout_prob
,
x_data
,
mask_data
,
y_data
,
upscale_in_train
,
increment
);
upscale_in_train
,
increment
,
main_offset
);
}
#endif
}
else
{
}
else
{
if
(
upscale_in_train
)
{
if
(
upscale_in_train
)
{
// todo: can y share with data with x directly?
// todo: can y share with data with x directly?
...
@@ -278,6 +213,22 @@ void DropoutFwGPUKernelDriver(const phi::GPUContext& dev_ctx, bool is_test,
...
@@ -278,6 +213,22 @@ void DropoutFwGPUKernelDriver(const phi::GPUContext& dev_ctx, bool is_test,
}
}
}
}
template
<
typename
T
,
typename
MaskType
>
struct
CudaDropoutGradFunctor
{
using
MT
=
typename
details
::
MPTypeTrait
<
T
>::
Type
;
explicit
CudaDropoutGradFunctor
(
const
MT
factor
)
:
factor_
(
factor
)
{}
__device__
__forceinline__
T
operator
()(
const
T
dout
,
const
MaskType
mask
)
const
{
return
static_cast
<
T
>
(
static_cast
<
MT
>
(
dout
)
*
static_cast
<
MT
>
(
mask
)
*
factor_
);
}
private:
MT
factor_
;
};
template
<
typename
T
>
template
<
typename
T
>
void
DropoutGradGPUKernelDriver
(
const
phi
::
GPUContext
&
dev_ctx
,
void
DropoutGradGPUKernelDriver
(
const
phi
::
GPUContext
&
dev_ctx
,
const
std
::
string
dropout_implementation
,
const
std
::
string
dropout_implementation
,
...
...
paddle/phi/kernels/funcs/distribution_helper.h
浏览文件 @
95d3ebc8
...
@@ -114,13 +114,19 @@ struct normal_transform {
...
@@ -114,13 +114,19 @@ struct normal_transform {
namespace
kps
=
phi
::
kps
;
namespace
kps
=
phi
::
kps
;
/*********************** Distribution Function *************************/
/*********************** Distribution Function *************************/
template
<
typename
T
>
struct
uniform_distribution
;
template
<
typename
T
>
template
<
typename
T
>
struct
normal_distribution
;
struct
normal_distribution
;
#if defined(__NVCC__)
#if defined(__NVCC__)
template
<
typename
T
>
struct
uniform_distribution
{
__device__
inline
T
operator
()(
curandStatePhilox4_32_10_t
*
state
)
const
{
return
static_cast
<
T
>
(
curand_uniform
(
state
));
}
static
constexpr
int
kReturnsCount
=
1
;
};
template
<
>
template
<
>
struct
uniform_distribution
<
float
>
{
struct
uniform_distribution
<
float
>
{
__device__
inline
float4
operator
()(
curandStatePhilox4_32_10_t
*
state
)
const
{
__device__
inline
float4
operator
()(
curandStatePhilox4_32_10_t
*
state
)
const
{
...
@@ -177,6 +183,14 @@ struct normal_distribution<double> {
...
@@ -177,6 +183,14 @@ struct normal_distribution<double> {
};
};
#else
#else
template
<
typename
T
>
struct
uniform_distribution
{
__device__
inline
T
operator
()(
hiprandStatePhilox4_32_10_t
*
state
)
const
{
return
hiprand_uniform
(
state
);
}
static
constexpr
int
kReturnsCount
=
1
;
};
template
<
>
template
<
>
struct
uniform_distribution
<
float
>
{
struct
uniform_distribution
<
float
>
{
__device__
inline
float4
operator
()(
__device__
inline
float4
operator
()(
...
...
paddle/phi/kernels/gpu/masked_select_grad_kernel.cu
浏览文件 @
95d3ebc8
...
@@ -17,11 +17,10 @@
...
@@ -17,11 +17,10 @@
#include <thrust/reverse.h>
#include <thrust/reverse.h>
#include <thrust/scan.h>
#include <thrust/scan.h>
#include "paddle/phi/core/kernel_registry.h"
#include "paddle/phi/kernels/funcs/select_impl.cu.h"
#include "paddle/phi/kernels/funcs/select_impl.cu.h"
#include "paddle/phi/kernels/masked_select_grad_kernel.h"
#include "paddle/phi/kernels/masked_select_grad_kernel.h"
#include "paddle/phi/backends/cpu/cpu_context.h"
#include "paddle/phi/core/kernel_registry.h"
namespace
phi
{
namespace
phi
{
template
<
typename
MT
,
typename
InT
,
typename
OutT
>
template
<
typename
MT
,
typename
InT
,
typename
OutT
>
...
@@ -50,7 +49,7 @@ void MaskedSelectGradKernel(const Context& dev_ctx,
...
@@ -50,7 +49,7 @@ void MaskedSelectGradKernel(const Context& dev_ctx,
const
DenseTensor
&
mask
,
const
DenseTensor
&
mask
,
DenseTensor
*
x_grad
)
{
DenseTensor
*
x_grad
)
{
auto
mask_size
=
mask
.
numel
();
auto
mask_size
=
mask
.
numel
();
auto
*
out_data
=
x_grad
->
mutable_data
<
T
>
(
dev_ctx
.
GetPlace
()
);
dev_ctx
.
template
Alloc
<
T
>(
x_grad
);
if
(
mask_size
<=
0
)
return
;
if
(
mask_size
<=
0
)
return
;
using
Functor
=
MaskedSelectGradFunctor
<
bool
,
T
,
T
>
;
using
Functor
=
MaskedSelectGradFunctor
<
bool
,
T
,
T
>
;
phi
::
funcs
::
SelectKernel
<
bool
,
T
,
T
,
2
,
Functor
>
(
phi
::
funcs
::
SelectKernel
<
bool
,
T
,
T
,
2
,
Functor
>
(
...
...
编辑
预览
Markdown
is supported
0%
请重试
或
添加新附件
.
添加附件
取消
You are about to add
0
people
to the discussion. Proceed with caution.
先完成此消息的编辑!
取消
想要评论请
注册
或
登录