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
wzpei9
Paddle
提交
8e3fdc6e
P
Paddle
项目概览
wzpei9
/
Paddle
与 Fork 源项目一致
Fork自
PaddlePaddle / Paddle
通知
1
Star
1
Fork
0
代码
文件
提交
分支
Tags
贡献者
分支图
Diff
Issue
1
列表
看板
标记
里程碑
合并请求
0
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
Issue
1
列表
看板
标记
里程碑
合并请求
0
合并请求
0
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
体验新版 GitCode,发现更多精彩内容 >>
提交
8e3fdc6e
编写于
9月 29, 2018
作者:
Y
Yu Yang
浏览文件
操作
浏览文件
下载
电子邮件补丁
差异文件
Fix SetDevice on init
上级
524f6e9b
变更
7
显示空白变更内容
内联
并排
Showing
7 changed file
with
52 addition
and
6 deletion
+52
-6
paddle/fluid/memory/allocation/CMakeLists.txt
paddle/fluid/memory/allocation/CMakeLists.txt
+2
-0
paddle/fluid/memory/allocation/allocation_and_eigen_test.cu
paddle/fluid/memory/allocation/allocation_and_eigen_test.cu
+45
-0
paddle/fluid/memory/allocation/allocator_facade.cc
paddle/fluid/memory/allocation/allocator_facade.cc
+0
-1
paddle/fluid/memory/allocation/cuda_allocator.cc
paddle/fluid/memory/allocation/cuda_allocator.cc
+0
-1
paddle/fluid/operators/math/CMakeLists.txt
paddle/fluid/operators/math/CMakeLists.txt
+1
-1
paddle/fluid/platform/device_context.cc
paddle/fluid/platform/device_context.cc
+2
-2
paddle/fluid/platform/init.cc
paddle/fluid/platform/init.cc
+2
-1
未找到文件。
paddle/fluid/memory/allocation/CMakeLists.txt
浏览文件 @
8e3fdc6e
...
...
@@ -42,3 +42,5 @@ cc_library(allocator_facade SRCS allocator_facade.cc DEPS
naive_managed_allocator
aligned_allocator
cuda_device_guard
)
nv_test
(
allocation_and_eigen_test SRCS allocation_and_eigen_test.cu DEPS allocator_facade
)
paddle/fluid/memory/allocation/allocation_and_eigen_test.cu
0 → 100644
浏览文件 @
8e3fdc6e
// Copyright (c) 2018 PaddlePaddle Authors. All Rights Reserved.
//
// 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.
#include "gtest/gtest.h"
#include "paddle/fluid/framework/eigen.h"
#include "paddle/fluid/framework/tensor.h"
#include "paddle/fluid/platform/device_context.h"
#include "paddle/fluid/platform/for_range.h"
#include "unsupported/Eigen/CXX11/Tensor"
struct
FillZero
{
public:
float
*
ptr_
;
__device__
void
operator
()(
size_t
i
)
{
ptr_
[
i
]
=
0.0
f
;
}
};
namespace
paddle
{
TEST
(
Eigen
,
main
)
{
framework
::
Tensor
tensor
;
platform
::
CUDAPlace
gpu
(
0
);
float
*
ptr
=
tensor
.
mutable_data
<
float
>
({
10
,
10
},
gpu
);
auto
&
dev_ctx
=
*
reinterpret_cast
<
platform
::
CUDADeviceContext
*>
(
platform
::
DeviceContextPool
::
Instance
().
Get
(
gpu
));
PADDLE_ENFORCE
(
cudaMemset
(
ptr
,
0
,
sizeof
(
float
)
*
100
));
platform
::
ForRange
<
platform
::
CUDADeviceContext
>
for_range
(
dev_ctx
,
100
);
for_range
(
FillZero
{
ptr
});
dev_ctx
.
Wait
();
auto
eigen_vec
=
framework
::
EigenVector
<
float
>::
Flatten
(
tensor
);
auto
&
eigen_dev
=
*
dev_ctx
.
eigen_device
();
eigen_vec
.
device
(
eigen_dev
)
=
eigen_vec
.
constant
(
0.0
f
);
}
}
// namespace paddle
paddle/fluid/memory/allocation/allocator_facade.cc
浏览文件 @
8e3fdc6e
...
...
@@ -46,7 +46,6 @@ class AllocatorFacadePrivate {
}
AllocatorFacadePrivate
()
{
std
::
cout
<<
"Init Allocator Facade"
<<
std
::
endl
;
InitCPUAllocator
();
InitCUDAAllocator
();
}
...
...
paddle/fluid/memory/allocation/cuda_allocator.cc
浏览文件 @
8e3fdc6e
...
...
@@ -31,7 +31,6 @@ std::unique_ptr<Allocation> CUDAAllocator::Allocate(size_t size, Attr attr) {
"Cannot allocate %d on GPU %d, cuda status %d, %s"
,
size
,
place_
.
device
,
status
,
cudaGetErrorString
(
status
)));
}
return
std
::
unique_ptr
<
Allocation
>
(
new
CUDAAllocation
(
ptr
,
size
,
platform
::
Place
(
place_
)));
}
...
...
paddle/fluid/operators/math/CMakeLists.txt
浏览文件 @
8e3fdc6e
...
...
@@ -72,7 +72,7 @@ cc_test(vol2col_test SRCS vol2col_test.cc DEPS vol2col)
cc_test
(
sequence_padding_test SRCS sequence_padding_test.cc DEPS sequence_padding
)
if
(
WITH_GPU
)
nv_test
(
math_function_gpu_test SRCS math_function_test.cu DEPS math_function
)
nv_test
(
selected_rows_functor_gpu_test SRCS selected_rows_functor_test.cu DEPS selected_rows_functor math_function
)
nv_test
(
selected_rows_functor_gpu_test SRCS selected_rows_functor_test.cu
.cc
DEPS selected_rows_functor math_function
)
endif
()
cc_test
(
concat_test SRCS concat_test.cc DEPS concat
)
cc_test
(
cpu_vec_test SRCS cpu_vec_test.cc DEPS blas cpu_info
)
paddle/fluid/platform/device_context.cc
浏览文件 @
8e3fdc6e
...
...
@@ -9,11 +9,11 @@ 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. */
#include "paddle/fluid/platform/device_context.h"
#include <set>
#include <string>
#include <unordered_set>
#include <vector>
#include "paddle/fluid/platform/cuda_device_guard.h"
#include "paddle/fluid/memory/memory.h"
#ifdef PADDLE_WITH_CUDA
...
...
@@ -205,7 +205,7 @@ class CudnnHolder {
CUDADeviceContext
::
CUDADeviceContext
(
CUDAPlace
place
)
:
place_
(
place
),
cudnn_holder_
(
nullptr
)
{
SetDeviceI
d
(
place_
.
device
);
CUDADeviceGuard
guar
d
(
place_
.
device
);
compute_capability
=
GetCUDAComputeCapability
(
place_
.
device
);
multi_process
=
GetCUDAMultiProcessors
(
place_
.
device
);
max_threads_per_mp
=
GetCUDAMaxThreadsPerMultiProcessor
(
place_
.
device
);
...
...
paddle/fluid/platform/init.cc
浏览文件 @
8e3fdc6e
...
...
@@ -19,6 +19,7 @@ limitations under the License. */
#include "paddle/fluid/framework/operator.h"
#include "paddle/fluid/platform/cpu_helper.h"
#include "paddle/fluid/platform/cpu_info.h"
#include "paddle/fluid/platform/cuda_device_guard.h"
#include "paddle/fluid/platform/device_context.h"
#include "paddle/fluid/platform/init.h"
#include "paddle/fluid/platform/place.h"
...
...
@@ -64,7 +65,7 @@ void InitP2P(std::vector<int> devices) {
LOG
(
WARNING
)
<<
"Cannot enable P2P access from "
<<
devices
[
i
]
<<
" to "
<<
devices
[
j
];
}
else
{
cudaSetDevice
(
devices
[
i
]);
platform
::
CUDADeviceGuard
guard
(
devices
[
i
]);
cudaDeviceEnablePeerAccess
(
devices
[
j
],
0
);
}
}
...
...
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