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
兽拳
Paddle
提交
30c350b9
P
Paddle
项目概览
兽拳
/
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 >>
未验证
提交
30c350b9
编写于
5月 15, 2018
作者:
X
Xin Pan
提交者:
GitHub
5月 15, 2018
浏览文件
操作
浏览文件
下载
差异文件
Merge pull request #10659 from panyx0718/fix_rpc
Delete prefetch_ctx_ after use.
上级
13836b04
32e05562
变更
4
隐藏空白更改
内联
并排
Showing
4 changed file
with
7 addition
and
7 deletion
+7
-7
paddle/fluid/operators/detail/grpc_server.cc
paddle/fluid/operators/detail/grpc_server.cc
+1
-1
paddle/fluid/operators/detail/grpc_server.h
paddle/fluid/operators/detail/grpc_server.h
+4
-3
paddle/fluid/operators/detail/grpc_server_test.cc
paddle/fluid/operators/detail/grpc_server_test.cc
+1
-1
paddle/fluid/operators/listen_and_serv_op.cc
paddle/fluid/operators/listen_and_serv_op.cc
+1
-2
未找到文件。
paddle/fluid/operators/detail/grpc_server.cc
浏览文件 @
30c350b9
...
...
@@ -306,7 +306,7 @@ void AsyncGRPCServer::TryToRegisterNewPrefetchOne() {
}
RequestPrefetch
*
prefetch
=
new
RequestPrefetch
(
&
service_
,
cq_prefetch_
.
get
(),
sync_mode_
,
scope_
,
dev_ctx_
,
executor_
,
program_
,
prefetch_ctx_
);
dev_ctx_
,
executor_
,
program_
,
prefetch_ctx_
.
get
()
);
VLOG
(
4
)
<<
"Create RequestPrefetch status:"
<<
prefetch
->
Status
();
}
...
...
paddle/fluid/operators/detail/grpc_server.h
浏览文件 @
30c350b9
...
...
@@ -64,8 +64,9 @@ class AsyncGRPCServer final {
void
SetExecutor
(
framework
::
Executor
*
executor
)
{
executor_
=
executor
;
}
void
SetPrefetchPreparedCtx
(
framework
::
ExecutorPrepareContext
*
prepared
)
{
prefetch_ctx_
=
prepared
;
void
SetPrefetchPreparedCtx
(
std
::
unique_ptr
<
framework
::
ExecutorPrepareContext
>
prepared
)
{
prefetch_ctx_
.
reset
(
prepared
.
release
());
}
int
GetSelectedPort
()
const
{
return
selected_port_
;
}
...
...
@@ -116,7 +117,7 @@ class AsyncGRPCServer final {
std
::
unique_ptr
<
std
::
thread
>
t_get_
;
std
::
unique_ptr
<
std
::
thread
>
t_prefetch_
;
framework
::
ExecutorPrepareContext
*
prefetch_ctx_
;
std
::
unique_ptr
<
framework
::
ExecutorPrepareContext
>
prefetch_ctx_
;
framework
::
ProgramDesc
*
program_
;
framework
::
Executor
*
executor_
;
int
selected_port_
;
...
...
paddle/fluid/operators/detail/grpc_server_test.cc
浏览文件 @
30c350b9
...
...
@@ -100,7 +100,7 @@ void StartServer(const std::string& endpoint) {
InitTensorsOnServer
(
&
scope
,
&
place
,
10
);
rpc_service_
->
SetProgram
(
&
program
);
rpc_service_
->
SetPrefetchPreparedCtx
(
prepared
.
get
(
));
rpc_service_
->
SetPrefetchPreparedCtx
(
std
::
move
(
prepared
));
rpc_service_
->
SetDevCtx
(
&
ctx
);
rpc_service_
->
SetScope
(
&
scope
);
rpc_service_
->
SetExecutor
(
&
exe
);
...
...
paddle/fluid/operators/listen_and_serv_op.cc
浏览文件 @
30c350b9
...
...
@@ -322,8 +322,7 @@ void ListenAndServOp::RunImpl(const framework::Scope &scope,
// prepare for prefetch
VLOG
(
3
)
<<
"prefetch block id is "
<<
prefetch_block
->
ID
();
auto
prefetch_prepared
=
executor
.
Prepare
(
*
program
,
prefetch_block
->
ID
());
rpc_service_
->
SetPrefetchPreparedCtx
(
prefetch_prepared
.
get
());
prefetch_prepared
.
release
();
rpc_service_
->
SetPrefetchPreparedCtx
(
std
::
move
(
prefetch_prepared
));
// start the server listening after all member initialized.
server_thread_
.
reset
(
new
std
::
thread
(
RunServer
,
rpc_service_
));
...
...
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