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
fe10e86d
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
提交
fe10e86d
编写于
10月 04, 2017
作者:
Q
qijun
浏览文件
操作
浏览文件
下载
电子邮件补丁
差异文件
fix gpu build error
上级
cb198fa7
变更
3
隐藏空白更改
内联
并排
Showing
3 changed file
with
17 addition
and
8 deletion
+17
-8
paddle/framework/executor.cc
paddle/framework/executor.cc
+14
-5
paddle/framework/executor.h
paddle/framework/executor.h
+2
-2
paddle/platform/gpu_info.cc
paddle/platform/gpu_info.cc
+1
-1
未找到文件。
paddle/framework/executor.cc
浏览文件 @
fe10e86d
...
@@ -27,12 +27,12 @@ Executor::Executor(const std::vector<platform::Place>& places) {
...
@@ -27,12 +27,12 @@ Executor::Executor(const std::vector<platform::Place>& places) {
device_contexts_
.
resize
(
places
.
size
());
device_contexts_
.
resize
(
places
.
size
());
for
(
size_t
i
=
0
;
i
<
places
.
size
();
i
++
)
{
for
(
size_t
i
=
0
;
i
<
places
.
size
();
i
++
)
{
if
(
platform
::
is_cpu_place
(
places
[
i
]))
{
if
(
platform
::
is_cpu_place
(
places
[
i
]))
{
device_contexts_
[
i
]
.
reset
(
new
platform
::
CPUDeviceContext
(
device_contexts_
[
i
]
=
new
platform
::
CPUDeviceContext
(
boost
::
get
<
platform
::
CPUPlace
>
(
places
[
i
]))
)
;
boost
::
get
<
platform
::
CPUPlace
>
(
places
[
i
]));
}
else
{
}
else
if
(
platform
::
is_gpu_place
(
places
[
i
]))
{
#ifndef PADDLE_ONLY_CPU
#ifndef PADDLE_ONLY_CPU
device_contexts_
[
i
]
.
reset
(
new
platform
::
CUDADeviceContext
(
device_contexts_
[
i
]
=
new
platform
::
CUDADeviceContext
(
boost
::
get
<
platform
::
CPUPlace
>
(
places
[
i
])
));
boost
::
get
<
platform
::
GPUPlace
>
(
places
[
i
]
));
#else
#else
PADDLE_THROW
(
"'GPUPlace' is not supported in CPU only device."
);
PADDLE_THROW
(
"'GPUPlace' is not supported in CPU only device."
);
#endif
#endif
...
@@ -40,6 +40,14 @@ Executor::Executor(const std::vector<platform::Place>& places) {
...
@@ -40,6 +40,14 @@ Executor::Executor(const std::vector<platform::Place>& places) {
}
}
}
}
Executor
::~
Executor
()
{
for
(
auto
&
device_context
:
device_contexts_
)
{
if
(
device_context
)
{
delete
device_context
;
}
}
}
void
Executor
::
Run
(
const
ProgramDesc
&
pdesc
,
Scope
*
scope
,
void
Executor
::
Run
(
const
ProgramDesc
&
pdesc
,
Scope
*
scope
,
std
::
vector
<
Tensor
>*
outputs
)
{
std
::
vector
<
Tensor
>*
outputs
)
{
// TODO(tonyyang-svail):
// TODO(tonyyang-svail):
...
@@ -59,6 +67,7 @@ void Executor::Run(const ProgramDesc& pdesc, Scope* scope,
...
@@ -59,6 +67,7 @@ void Executor::Run(const ProgramDesc& pdesc, Scope* scope,
for
(
auto
&
op_desc
:
block
.
ops
())
{
for
(
auto
&
op_desc
:
block
.
ops
())
{
auto
op
=
paddle
::
framework
::
OpRegistry
::
CreateOp
(
op_desc
);
auto
op
=
paddle
::
framework
::
OpRegistry
::
CreateOp
(
op_desc
);
std
::
cout
<<
op
->
DebugString
()
<<
std
::
endl
;
op
->
Run
(
*
scope
,
*
device
);
op
->
Run
(
*
scope
,
*
device
);
}
}
...
...
paddle/framework/executor.h
浏览文件 @
fe10e86d
...
@@ -25,11 +25,11 @@ namespace framework {
...
@@ -25,11 +25,11 @@ namespace framework {
class
Executor
{
class
Executor
{
public:
public:
explicit
Executor
(
const
std
::
vector
<
platform
::
Place
>&
places
);
explicit
Executor
(
const
std
::
vector
<
platform
::
Place
>&
places
);
~
Executor
()
{}
~
Executor
()
;
void
Run
(
const
ProgramDesc
&
,
Scope
*
,
std
::
vector
<
Tensor
>*
);
void
Run
(
const
ProgramDesc
&
,
Scope
*
,
std
::
vector
<
Tensor
>*
);
private:
private:
std
::
vector
<
std
::
unique_ptr
<
platform
::
DeviceContext
>
>
device_contexts_
;
std
::
vector
<
platform
::
DeviceContext
*
>
device_contexts_
;
};
};
}
// namespace framework
}
// namespace framework
...
...
paddle/platform/gpu_info.cc
浏览文件 @
fe10e86d
...
@@ -43,7 +43,7 @@ int GetCurrentDeviceId() {
...
@@ -43,7 +43,7 @@ int GetCurrentDeviceId() {
}
}
void
SetDeviceId
(
int
id
)
{
void
SetDeviceId
(
int
id
)
{
PADDLE_ENFORCE
(
id
<
GetDeviceCount
(),
"id must less than GPU count"
)
PADDLE_ENFORCE
(
id
<
GetDeviceCount
(),
"id must less than GPU count"
)
;
PADDLE_ENFORCE
(
cudaSetDevice
(
id
),
PADDLE_ENFORCE
(
cudaSetDevice
(
id
),
"cudaSetDevice failed in paddle::platform::SetDeviceId"
);
"cudaSetDevice failed in paddle::platform::SetDeviceId"
);
}
}
...
...
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