compile and install the static library of fluid inference #7827

luotao1 · 2018-01-24T09:57:48Z

related #7231

The structure of the installed shared library is as follows. It uses default cmake option + -DWITH_PYTHON=OFF

PaddleRoot/
├── bin
│   └── paddle
├── include
│   └── paddle
│       ├── framework
│       ├── inference
│       ├── memory
│       ├── platform
│       └── string
├── lib
│   ├── libpaddle_fluid.a
│   └── libpaddle_fluid.so
├── opt
│   └── paddle
│       └── bin
└── third_party
    ├── eigen3
    │   ├── Eigen
    │   └── unsupported
    ├── gflags
    │   ├── include
    │   └── lib
    ├── glog
    │   ├── include
    │   └── lib
    ├── openblas
    │   ├── include
    │   └── lib
    └── protobuf
        ├── include
        └── lib

du -sh *
12K     bin
700K    include
188M    lib
200M    opt
92M     third_party

128M    libpaddle_fluid.a
61M     libpaddle_fluid.so

The DEPS of static library is manually added one by one now, it always missing some new DEPS. So adding DEPS automatically will be done later.
Examples: https://github.com/luotao1/fluid_inference_example.

wangkuiyi

Please refer to my comment #7231 (comment) 8 days ago.

It seems that this PR and the issue aforementioned are related to the binary release of PaddlePaddle.

However, to what I understand is that our binary release would be in the form of pip packages, but not as binary libraries (/bin, /include, /lib directories).

Also, even if we are going to add binary libraries as an additional form of binary release, we are not supposed to put third-party binaries in our directory hierarchy. The right way is to build only ours and pack a .deb file, which describes on what other .deb files it depends.

My suggestion is that we discard this PR and possibly revert another one (I vaguely remember, had been merged), and rethink about (1) if we are going to do this, and (2) what's the right way to doing it.

luotao1 · 2018-01-25T07:54:49Z

Please refer to my comment #7231 (comment) 8 days ago

Discussed with @Xreki, we have already answered in #7572 (comment), but sorry to forget @ you.

It seems that this PR and the issue aforementioned are related to the binary release of PaddlePaddle.

The binary directory /bin and /opt is left for C-API since WITH_FLUID option (#6346) has not been finished now. After finishing #6346, the final fluid inference library directory likes as follows, and it is not a binary library:

PaddleRoot/
├── include
│   └── paddle
│       ├── framework
│       ├── inference
│       ├── memory
│       ├── platform
│       └── string
├── lib
│   ├── libpaddle_fluid.a
│   └── libpaddle_fluid.so
└── third_party
    ├── eigen3
    │   ├── Eigen
    │   └── unsupported
    ├── gflags
    │   ├── include
    │   └── lib
    ├── glog
    │   ├── include
    │   └── lib
    ├── openblass
    │   ├── include
    │   └── lib
    └── protobuf
        ├── include
        └── lib

However, to what I understand is that our binary release would be in the form of pip packages

As both IOS and Android don't support pip packages since they don't support python, the fluid inference library should not be in the form of this way.

Also, even if we are going to add binary libraries as an additional form of binary release, we are not supposed to put third-party binaries in our directory hierarchy.

Our third-party libraries only have include/ and lib/, and don't contain any binaries.

There are two reasons why we put third-party libraries in our directory:

For some platforms (such as IOS) and usages, users must link static library. Thus, they must link the corresponding third-party libraries. If we don't provide them, users must download and compile them manually, which is not convenient and sometimes hard for users. For example, like openblas.cmake, the compile option is different from IOS, Android, RPI, Apple and Linux.

Paddle/cmake/external/openblas.cmake

Lines 35 to 67 in b468567

    
           IF(CMAKE_CROSSCOMPILING) 
        
               SET(OPTIONAL_ARGS HOSTCC=${HOST_C_COMPILER}) 
        
               GET_FILENAME_COMPONENT(CROSS_SUFFIX ${CMAKE_C_COMPILER} DIRECTORY) 
        
               SET(CROSS_SUFFIX ${CROSS_SUFFIX}/) 
        
               IF(ANDROID) 
        
                   IF(ANDROID_ABI MATCHES "^armeabi(-v7a)?$") 
        
                       # use softfp 
        
                       SET(OPTIONAL_ARGS ${OPTIONAL_ARGS} TARGET=ARMV7 ARM_SOFTFP_ABI=1 USE_THREAD=0) 
        
                   ELSEIF(ANDROID_ABI STREQUAL "arm64-v8a") 
        
                       SET(OPTIONAL_ARGS ${OPTIONAL_ARGS} TARGET=ARMV8 BINARY=64 USE_THREAD=0) 
        
                   ENDIF() 
        
               ELSEIF(IOS) 
        
                   IF(CMAKE_OSX_ARCHITECTURES MATCHES "arm64") 
        
                       SET(OPENBLAS_CC "${OPENBLAS_CC} ${CMAKE_C_FLAGS} -isysroot ${CMAKE_OSX_SYSROOT}") 
        
                       SET(OPENBLAS_CC "${OPENBLAS_CC} -arch arm64") 
        
                       SET(OPTIONAL_ARGS ${OPTIONAL_ARGS} TARGET=ARMV8 BINARY=64 USE_THREAD=0 CROSS_SUFFIX=${CROSS_SUFFIX}) 
        
                   ELSE() 
        
                       MESSAGE(FATAL_ERROR "OpenBLAS only support arm64 architectures on iOS. " 
        
                              "You can set IOS_USE_VECLIB_FOR_BLAS=ON or USE_EIGEN_FOR_BLAS=ON to use other blas library instead.") 
        
                   ENDIF() 
        
               ELSEIF(RPI) 
        
                   # use hardfp 
        
                   SET(OPTIONAL_ARGS ${OPTIONAL_ARGS} TARGET=ARMV7 USE_THREAD=0) 
        
               ENDIF() 
        
           ELSE() 
        
               IF(APPLE) 
        
                   SET(OPENBLAS_CC "${CMAKE_C_COMPILER} -isysroot ${CMAKE_OSX_SYSROOT}") 
        
               ENDIF() 
        
               SET(OPTIONAL_ARGS "") 
        
               IF(CMAKE_SYSTEM_PROCESSOR MATCHES "^x86(_64)?$") 
        
                   SET(OPTIONAL_ARGS DYNAMIC_ARCH=1 NUM_THREADS=64) 
        
               ENDIF() 
        
           ENDIF()

It is easy for users to know how many third-party libraries depended by paddle, and it is easy for users to link these libraries themselves.

The right way is to build only ours and pack a .deb file, which describes on what other .deb files it depends.

.deb can only be installed on Ubuntu, but it can't be installed on Mac, IOS and Andriod. Thus, .deb may not be a good choice.

wangkuiyi · 2018-01-29T07:29:16Z

I agree that we publish our inference library, but with two points different from the current status of this PR:

make inference_lib_dist instead of make install
the directory hierarchy in the generated tgz file should be the same as that in ./build.

luotao1 · 2018-01-29T11:59:37Z

@wangkuiyi

the directory hierarchy in the generated tgz file should be the same as that in ./build

If the same as that in ./build, the final directory hierarchy would like as follows:

PaddleRoot
├── paddle
│   ├── framework
│   │   ├── attribute.h
│   │   ├── backward.h ...
│   │   └── var_type_inference.h
│   ├── inference
│   │   ├── inference.h
│   │   └── libpaddle_fluid.so
│   ├── memory
│   │   ├── detail
│   │   ├── memcpy.h
│   │   └── memory.h
│   ├── platform
│   │   ├── assert.h
│   │   ├── call_once.h ...
│   │   └── variant.h
│   └── string
│       ├── piece.h
│       ├── printf.h
│       ├── tinyformat
│       └── to_string.h
└── third_party
    ├── eigen3
    │   ├── Eigen
    │   │   ├── Core
    │   │   └── src
    │   └── unsupported
    │       └── Eigen
    └── install
        ├── gflags
        │   ├── include
        │   └── lib
        ├── glog
        │   ├── include
        │   └── lib
        └── protobuf
            ├── include
            └── lib

Xreki · 2018-02-09T06:27:52Z

cmake/external/openblas.cmake

@@ -82,6 +82,7 @@ IF(NOT ${CBLAS_FOUND})
        CONFIGURE_COMMAND   ""
    )
    SET(CBLAS_PROVIDER openblas)
+    FILE(REMOVE_RECURSE ${CBLAS_INSTALL_DIR}/lib/cmake ${CBLAS_INSTALL_DIR}/lib/pkgconfig)


FILE命令应该是在执行cmake命令的时候执行的。extern_openblas是在make阶段执行的，也就是说，必须要执行完extern_openblas这个任务之后，才会生成${CBLAS_INSTALL_DIR}/lib/cmake和${CBLAS_INSTALL_DIR}/lib/pkgconfig这两个目录。所以这条语句应该达不到删除的目的。应该使用execute_process语句删除，并且该语句需要依赖extern_openblas这个任务。

Xreki · 2018-02-09T06:40:25Z

paddle/inference/CMakeLists.txt

 # Merge all modules into a single static library
-cc_library(paddle_fluid DEPS paddle_fluid_api ${FLUID_CORE_MODULES} ${GLOB_OP_LIB})
+cc_library(paddle_fluid DEPS paddle_fluid_api 


像memory模块一样，我们在每个模块最后加一个cc_library，将该模块里面所有的target都合并成一个静态库？

Paddle/paddle/memory/CMakeLists.txt

Lines 6 to 14 in 1185a1b

cc_library(paddle_memory

DEPS

memory

memcpy

meta_data

meta_cache

memory_block

buddy_allocator

system_allocator)

这样，在inference里面，我们只需要依赖paddle_memroy, paddle_framework, paddle_operators, paddle_string, paddle_platform。不过这样还是不能自动添加。

CLAassistant · 2018-02-10T05:40:19Z

All committers have signed the CLA.

* use libprotobuf.a instead of libprotobuf-lite.a for profiler

Xreki · 2018-03-07T09:49:43Z

cmake/inference_lib.cmake

      DSTS ${dst_dir} ${dst_dir}/lib
    )
 ENDIF(NOT PROTOBUF_FOUND)

+IF(NOT WITH_MKL)


这里应该是if(NOT CBLAS_FOUND)，cblas.cmake里面会去找系统里的一些blas库，只有CBLAS_FOUND为OFF时才会自动去下载、编译和安装openblas。原则上，不是Paddle自动下载、编译的库，我们不要拷贝吧。mklml在加入的时候，也在cblas.cmake里面设置了CBLAS_FOUND，这种情况另外再考虑吧。

另外，大小写统一一下？:smile:

Xreki

LGTM and have tested. It is glad to see all fluid C++ modules can be collected automatically and merged into a single static library. Great work!

luotao1 added 2 commits January 24, 2018 16:36

compile and install the static library of fluid inference

284d137

Merge branch 'develop' into fluid_infer

9c908fa

Xreki added the 预测原名Inference，包含Capi预测问题等 label Jan 24, 2018

fix dynload_cuda not in CPU mode

ff06fac

wangkuiyi requested changes Jan 24, 2018

View reviewed changes

This was referenced Jan 30, 2018

make inference_lib_dist for fluid inference shared library #7977

Merged

Add make clean in docker/build.sh to avoid link problem among different environment #8075

Closed

Merge branch 'develop' into fluid_infer

eaa48d7

luotao1 closed this in #7977 Feb 7, 2018

luotao1 reopened this Feb 8, 2018

luotao1 added 2 commits February 8, 2018 16:45

Merge branch 'develop' into fluid_infer

932a46e

update shared library and adjust the deploy of openblas

750d2e6

Xreki reviewed Feb 9, 2018

View reviewed changes

Merge branch 'develop' into fluid_infer

bb334f2

luotao1 closed this Feb 10, 2018

luotao1 reopened this Feb 10, 2018

luotao1 added 3 commits February 11, 2018 17:15

Merge branch 'develop' into fluid_infer

30a3969

adjust the deploy of openblas

52d3bff

Merge branch 'develop' into fluid_infer

0f84de8

luotao1 mentioned this pull request Feb 27, 2018

combine batch_size_like.cc into batch_size_like.h #8602

Closed

luotao1 added 5 commits March 2, 2018 14:59

Merge branch 'develop' into fluid_infer

a1b6799

* auto add all fluid modules for static library

47f1d47

* use libprotobuf.a instead of libprotobuf-lite.a for profiler

Merge branch 'develop' into fluid_infer

48b701a

Merge branch 'develop' into fluid_infer

91d19b6

Merge branch 'develop' into fluid_infer

780867e

luotao1 added 2 commits March 7, 2018 11:25

use set_property to set the global varible instead of ENV

9bb9962

add gpu depends of fluid modules, auto add inference_lib_dist depends

9b6b845

Xreki reviewed Mar 7, 2018

View reviewed changes

change the condition of openblas_lib, and fix a typo

565278b

Xreki approved these changes Mar 7, 2018

View reviewed changes

wangkuiyi approved these changes Mar 7, 2018

View reviewed changes

wangkuiyi merged commit 6f50dee into PaddlePaddle:develop Mar 7, 2018

luotao1 deleted the fluid_infer branch May 22, 2018 05:46

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

compile and install the static library of fluid inference #7827

compile and install the static library of fluid inference #7827

luotao1 commented Jan 24, 2018 •

edited

Loading

wangkuiyi left a comment

luotao1 commented Jan 25, 2018

wangkuiyi commented Jan 29, 2018

luotao1 commented Jan 29, 2018 •

edited

Loading

Xreki Feb 9, 2018

Xreki Feb 9, 2018

CLAassistant commented Feb 10, 2018 •

edited

Loading

Xreki Mar 7, 2018

luotao1 Mar 7, 2018

Xreki left a comment •

edited

Loading

	cc_library(paddle_memory
	DEPS
	memory
	memcpy
	meta_data
	meta_cache
	memory_block
	buddy_allocator
	system_allocator)

compile and install the static library of fluid inference #7827

compile and install the static library of fluid inference #7827

Conversation

luotao1 commented Jan 24, 2018 • edited Loading

wangkuiyi left a comment

Choose a reason for hiding this comment

luotao1 commented Jan 25, 2018

wangkuiyi commented Jan 29, 2018

luotao1 commented Jan 29, 2018 • edited Loading

Xreki Feb 9, 2018

Choose a reason for hiding this comment

Xreki Feb 9, 2018

Choose a reason for hiding this comment

CLAassistant commented Feb 10, 2018 • edited Loading

Xreki Mar 7, 2018

Choose a reason for hiding this comment

luotao1 Mar 7, 2018

Choose a reason for hiding this comment

Xreki left a comment • edited Loading

Choose a reason for hiding this comment

luotao1 commented Jan 24, 2018 •

edited

Loading

luotao1 commented Jan 29, 2018 •

edited

Loading

CLAassistant commented Feb 10, 2018 •

edited

Loading

Xreki left a comment •

edited

Loading