热门标签
热门文章
- 1Jinjia2-语法入门
- 2react+Antd表单用Form.Item互相校验_antd + vue循环form-item的校验并在其中一项给值后清除另一项的校验
- 3远程显示协议对比：RemoteFX vs. HDX vs. PCoIP
- 4【机器学习】特征选择之最小冗余最大相关性(mRMR)与随机森林(RF)_最小冗余最大相关matlab代码
- 5GPT SOVITS项目 一分钟克隆 （文字输出）
- 6python的异步编程async_python comm.spider_communal @async
- 7不归类
- 8明翰游戏学笔记V0.2（持续更新）
- 9Kubectl 常用命令, 开发人员常用k8s命令_kubectl delete job
- 10二十.python中的输入函数（两种方法）_python怎么输入函数
当前位置: article > 正文
在Linux上运行procmem和procrank_proc_mem
作者：繁依Fanyi0 | 2024-02-28 18:30:19
赞
踩
proc_mem
Android系统中提供了两个命令行工具procrank、procmem用于查看系统中的内存使用情况。procrank可以查看系统中所有进程的整体内存占用情况，并按照规则排序。而procmem可以针对某个特定的进程分析其堆、栈、共享库等内存占用情况。这两个工具对于我们分析内存相关问题非常有效。由于Android系统使用的是Linux内核，理论上这样的工具可以在Linux上运行。
编译
参考Android系统中procrank、procmem的Android.mk发现他们依赖于libpagemap库，需要将其一并移植到Linux系统中。
Makefile可以通过CMake自动生成，当然前提是需要安装cmake工具。
sudo apt-get install cmake- 1
参考Android.mk，可以写出如下的CMakeLists.txt.
cmake_minimum_required(VERSION 2.8)
PROJECT (mem_proc)
file(GLOB SOURCES "${PROJECT_SOURCE_DIR}/libpagemap/*.c")
add_definitions (
-D_LARGEFILE64_SOURCE
)
include_directories (${PROJECT_SOURCE_DIR}/libpagemap/include)
add_library(pagemap ${SOURCES})
add_executable(procmem ${PROJECT_SOURCE_DIR}/procmem/procmem.c)
target_link_libraries(procmem pagemap)
add_executable(procrank ${PROJECT_SOURCE_DIR}/procrank/procrank.c)
target_link_libraries(procrank pagemap)- 1
- 2
- 3
- 4
- 5
- 6
- 7
- 8
- 9
- 10
- 11
- 12
- 13
编译过程如下：
-从GitHub下载源码
david-virtual-machine% git clone https://github.com/sunao2002002/mem_proc.git
正克隆到 'mem_proc'...
remote: Counting objects: 22, done.
remote: Compressing objects: 100% (15/15), done.
remote: Total 22 (delta 4), reused 22 (delta 4), pack-reused 0
展开对象中: 100% (22/22), 完成.
检查连接... 完成。- 1
- 2
- 3
- 4
- 5
- 6
- 7
-运行cmake生成Makefile
david-virtual-machine% cd mem_proc
david-virtual-machine% ls
build.sh CMakeLists.txt libpagemap procmem procrank
david-virtual-machine% mkdir out
david-virtual-machine% cd out
david-virtual-machine% cmake ..
-- The C compiler identification is GNU 5.4.0
-- The CXX compiler identification is GNU 5.4.0
-- Check for working C compiler: /usr/bin/cc
-- Check for working C compiler: /usr/bin/cc -- works
-- Detecting C compiler ABI info
-- Detecting C compiler ABI info - done
-- Detecting C compile features
-- Detecting C compile features - done
-- Check for working CXX compiler: /usr/bin/c++
-- Check for working CXX compiler: /usr/bin/c++ -- works
-- Detecting CXX compiler ABI info
-- Detecting CXX compiler ABI info - done
-- Detecting CXX compile features
-- Detecting CXX compile features - done
-- Configuring done
-- Generating done
-- Build files have been written to: /home/david/github/mem_proc/out
david-virtual-machine% ls
CMakeCache.txt CMakeFiles cmake_install.cmake Makefile
david-virtual-machine%
- 1
- 2
- 3
- 4
- 5
- 6
- 7
- 8
- 9
- 10
- 11
- 12
- 13
- 14
- 15
- 16
- 17
- 18
- 19
- 20
- 21
- 22
- 23
- 24
- 25
- 26
-make生成可执行文件
david-virtual-machine% make
Scanning dependencies of target pagemap
[ 11%] Building C object CMakeFiles/pagemap.dir/libpagemap/pm_memusage.c.o
[ 22%] Building C object CMakeFiles/pagemap.dir/libpagemap/pm_process.c.o
[ 33%] Building C object CMakeFiles/pagemap.dir/libpagemap/pm_map.c.o
[ 44%] Building C object CMakeFiles/pagemap.dir/libpagemap/pm_kernel.c.o
[ 55%] Linking C static library libpagemap.a
[ 55%] Built target pagemap
Scanning dependencies of target procmem
[ 66%] Building C object CMakeFiles/procmem.dir/procmem/procmem.c.o
[ 77%] Linking C executable procmem
[ 77%] Built target procmem
Scanning dependencies of target procrank
[ 88%] Building C object CMakeFiles/procrank.dir/procrank/procrank.c.o
/home/david/github/mem_proc/procrank/procrank.c: In function ‘getprocname’:
/home/david/github/mem_proc/procrank/procrank.c:381:9: warning: implicit declaration of function ‘asprintf’ [-Wimplicit-function-declaration]
if (asprintf(&filename, "/proc/%d/cmdline", pid) < 0) {
^
/home/david/github/mem_proc/procrank/procrank.c:408:56: warning: comparison between pointer and integer
if (strncpy(buf, unknown_cmdline, (size_t)len) >= (size_t)len) {
^
[100%] Linking C executable procrank
[100%] Built target procrank
david-virtual-machine% file procrank procmem
procrank: ELF 64-bit LSB executable, x86-64, version 1 (SYSV), dynamically linked, interpreter /lib64/ld-linux-x86-64.so.2, for GNU/Linux 2.6.32, BuildID[sha1]=19cdcb1d4674c66a3ea62e3bc026f01773a43ce0, not stripped
procmem: ELF 64-bit LSB executable, x86-64, version 1 (SYSV), dynamically linked, interpreter /lib64/ld-linux-x86-64.so.2, for GNU/Linux 2.6.32, BuildID[sha1]=20e20cce4fd40bc859eca4d91d298f4d5165c8ff, not stripped
david-virtual-machine%- 1
- 2
- 3
- 4
- 5
- 6
- 7
- 8
- 9
- 10
- 11
- 12
- 13
- 14
- 15
- 16
- 17
- 18
- 19
- 20
- 21
- 22
- 23
- 24
- 25
- 26
- 27
-也可以直接运行build.sh完成整个流程。
运行
procrank
procrank是按照内存占用情况对进程进行排序。因为它需要遍历／proc下的所有进程获取内存占用情况，所以在运行时候需要有root权限。可用排序的有VSS、RSS、PSS、USS。
VSS：Virtual Set Size，虚拟内存耗用内存，包括共享库的内存
RSS：Resident Set Size，实际使用物理内存，包括共享库
PSS：Proportional Set Size，实际使用的物理内存，共享库按比例分配
USS：Unique Set Size，进程独占的物理内存，不计算共享库，也可以理解为将进程杀死能释放出的内存
一般VSS >= RSS >= PSS >= USS
procrank默认是按照PSS降序排列
david-virtual-machine% sudo ./procrank -h
Usage: ./procrank [ -W ] [ -v | -r | -p | -u | -s | -h ]
-v Sort by VSS.
-r Sort by RSS.
-p Sort by PSS.
-u Sort by USS.
-s Sort by swap.
(Default sort order is PSS.)
-R Reverse sort order (default is descending).
-c Only show cached (storage backed) pages
-C Only show non-cached (ram/swap backed) pages
-k Only show pages collapsed by KSM
-w Display statistics for working set only.
-W Reset working set of all processes.
-h Display this help screen.
david-virtual-machine%- 1
- 2
- 3
- 4
- 5
- 6
- 7
- 8
- 9
- 10
- 11
- 12
- 13
- 14
- 15
- 16
运行如下截图
david-virtual-machine% sudo ./procrank
PID Vss Rss Pss Uss cmdline
1078 283472K 40276K 33492K 32176K /usr/lib/xorg/Xorg
1746 637444K 50136K 21732K 16172K /usr/bin/python3
1724 237872K 28724K 17579K 14408K /usr/bin/python3
......
1793 4504K 1688K 261K 244K /bin/sh
749 4396K 1760K 240K 220K /usr/sbin/acpid
3158 14868K 1788K 188K 164K gnome-pty-helper
------ ------ ------
340421K 270184K TOTAL
RAM: 8157592K total, 6676160K free, 116188K buffers, 834428K cached, 12048K shmem, 148268K slab- 1
- 2
- 3
- 4
- 5
- 6
- 7
- 8
- 9
- 10
- 11
- 12
- 13
procmem
procrank 给出了系统整体的内存占用情况，针对每一个进程具体分析就要使用procmem工具了。同样运行procmem需要root权限。
以上图1793号进程 ／bin/sh为例，分析内存占用情况：
david-virtual-machine% sudo ./procmem 1793
Vss Rss Pss Uss ShCl ShDi PrCl PrDi Name
------- ------- ------- ------- ------- ------- ------- -------
144K 144K 144K 144K 0K 0K 144K 0K /bin/dash
8K 8K 8K 8K 0K 0K 8K 0K /bin/dash
4K 4K 4K 4K 0K 0K 4K 0K /bin/dash
8K 4K 4K 4K 0K 0K 4K 0K
132K 8K 8K 8K 0K 0K 8K 0K [heap]
1788K 1288K 16K 0K 1288K 0K 0K 0K /lib/x86_64-linux-gnu/libc-2.23.so
2048K 0K 0K 0K 0K 0K 0K 0K /lib/x86_64-linux-gnu/libc-2.23.so
16K 16K 16K 16K 0K 0K 16K 0K /lib/x86_64-linux-gnu/libc-2.23.so
8K 8K 8K 8K 0K 0K 8K 0K /lib/x86_64-linux-gnu/libc-2.23.so
16K 8K 8K 8K 0K 0K 8K 0K
152K 152K 1K 0K 152K 0K 0K 0K /lib/x86_64-linux-gnu/ld-2.23.so
12K 12K 12K 12K 0K 0K 12K 0K
8K 8K 8K 8K 0K 0K 8K 0K
4K 4K 4K 4K 0K 0K 4K 0K /lib/x86_64-linux-gnu/ld-2.23.so
4K 4K 4K 4K 0K 0K 4K 0K /lib/x86_64-linux-gnu/ld-2.23.so
4K 4K 4K 4K 0K 0K 4K 0K
132K 12K 12K 12K 0K 0K 12K 0K [stack]
8K 0K 0K 0K 0K 0K 0K 0K [vvar]
8K 4K 0K 0K 4K 0K 0K 0K [vdso]
0K 0K 0K 0K 0K 0K 0K 0K [vsyscall]
------- ------- ------- ------- ------- ------- ------- -------
4504K 1688K 261K 244K 1444K 0K 244K 0K TOTAL- 1
- 2
- 3
- 4
- 5
- 6
- 7
- 8
- 9
- 10
- 11
- 12
- 13
- 14
- 15
- 16
- 17
- 18
- 19
- 20
- 21
- 22
- 23
- 24
- 25
procmem 给出了procrank中VSS、RSS、PSS、USS的每一部分组成，包括进程可执行程序本身、共享库、堆、栈的内存占用。这里的信息应该是从/proc/pid/maps文件中获取的.
david-virtual-machine% sudo cat /proc/1793/maps
559358022000-559358046000 r-xp 00000000 08:01 1835041 /bin/dash
559358245000-559358247000 r--p 00023000 08:01 1835041 /bin/dash
559358247000-559358248000 rw-p 00025000 08:01 1835041 /bin/dash
559358248000-55935824a000 rw-p 00000000 00:00 0
559358920000-559358941000 rw-p 00000000 00:00 0 [heap]
7fab1342c000-7fab135eb000 r-xp 00000000 08:01 7606546 /lib/x86_64-linux-gnu/libc-2.23.so
7fab135eb000-7fab137eb000 ---p 001bf000 08:01 7606546 /lib/x86_64-linux-gnu/libc-2.23.so
7fab137eb000-7fab137ef000 r--p 001bf000 08:01 7606546 /lib/x86_64-linux-gnu/libc-2.23.so
7fab137ef000-7fab137f1000 rw-p 001c3000 08:01 7606546 /lib/x86_64-linux-gnu/libc-2.23.so
7fab137f1000-7fab137f5000 rw-p 00000000 00:00 0
7fab137f5000-7fab1381b000 r-xp 00000000 08:01 7602273 /lib/x86_64-linux-gnu/ld-2.23.so
7fab139fa000-7fab139fd000 rw-p 00000000 00:00 0
7fab13a18000-7fab13a1a000 rw-p 00000000 00:00 0
7fab13a1a000-7fab13a1b000 r--p 00025000 08:01 7602273 /lib/x86_64-linux-gnu/ld-2.23.so
7fab13a1b000-7fab13a1c000 rw-p 00026000 08:01 7602273 /lib/x86_64-linux-gnu/ld-2.23.so
7fab13a1c000-7fab13a1d000 rw-p 00000000 00:00 0
7ffe0a1f9000-7ffe0a21a000 rw-p 00000000 00:00 0 [stack]
7ffe0a273000-7ffe0a275000 r--p 00000000 00:00 0 [vvar]
7ffe0a275000-7ffe0a277000 r-xp 00000000 00:00 0 [vdso]
ffffffffff600000-ffffffffff601000 r-xp 00000000 00:00 0 [vsyscall]- 1
- 2
- 3
- 4
- 5
- 6
- 7
- 8
- 9
- 10
- 11
- 12
- 13
- 14
- 15
- 16
- 17
- 18
- 19
- 20
- 21
声明：本文内容由网友自发贡献，转载请注明出处：【wpsshop博客】
推荐阅读
相关标签
