您的位置:首页 > 移动开发 > Android开发

Android ps命令执行后的各项参数含义

2016-03-11 16:04 871 查看
直接输入ps后可以看到如下信息:

# ps

ps

USER PID PPID VSIZE RSS WCHAN PC NAME

root 1 0 276 188 c0099f1c 000086e8 S /init

root 2 0 0 0 c004df64 00000000 S kthreadd

root 3 2 0 0 c003fa28 00000000 S ksoftirqd/0

root 4 2 0 0 c004abc0 00000000 S events/0

root 5 2 0 0 c004abc0 00000000 S khelper

root 6 2 0 0 c004abc0 00000000 S suspend

root 7 2 0 0 c004abc0 00000000 S kblockd/0

root 8 2 0 0 c004abc0 00000000 S cqueue

root 9 2 0 0 c01780d0 00000000 S kseriod

root 10 2 0 0 c004abc0 00000000 S kmmcd

root 11 2 0 0 c006efa8 00000000 S pdflush

root 12 2 0 0 c006efa8 00000000 S pdflush

root 13 2 0 0 c0073480 00000000 S kswapd0

root 14 2 0 0 c004abc0 00000000 S aio/0

root 22 2 0 0 c0175900 00000000 S mtdblockdc

那么这些项各代表着什么意思呢?

USER: 进程的当前用户;

PID : 毫无疑问, process ID的缩写,也就进程号;

PPID :process parent ID,父进程ID

VSIZE : virtual size,进程虚拟地址空间大小;

RSS : 进程正在使用的物理内存的大小;

WCHAN :进程如果处于休眠状态的话,在内核中的地址;

PC : program counter,

NAME: process name,进程的名称

对比linux下的参数信息:

root@long-desktop:~# ps aux

USER PID %CPU %MEM VSZ RSS TTY STAT START TIME COMMAND

root 1 0.0 0.2 2804 1684 ? Ss 21:11 0:01 /sbin/init

root 2 0.0 0.0 0 0 ? S 21:11 0:00 [kthreadd]

root 2286 0.0 0.4 8888 3172 ? Ss 21:52 0:00 sshd: root@pts/0

root 2352 0.0 0.2 4684 2048 pts/0 Ss+ 21:52 0:00 -bash

root 2454 0.4 0.4 8888 3164 ? Rs 22:06 0:00 sshd: root@pts/1

root 2518 0.1 0.2 4684 2004 pts/1 Ss 22:06 0:00 -bash

root 2551 0.0 0.1 2736 1088 pts/1 R+ 22:06 0:00 ps aux

在Linux下ps命令各项含义如下所示:

%CPU 进程的cpu占用率

%MEM 进程的内存占用率

VSZ 进程所使用的虚存的大小

RSS 进程使用的驻留集大小或者是实际内存的大小

TTY 与进程关联的终端(tty)

STAT 检查的状态:进程状态使用字符表示的,如R(running正在运行或准备运行)、S(sleeping睡眠)、I(idle空闲)、Z (僵死)、D(不可中断的睡眠,通常是I/O)、P(等待交换页)、W(换出,表示当前页面不在内存)、N(低优先级任务)T(terminate终 止)、W has no resident pages

START (进程启动时间和日期)

TIME ;(进程使用的总cpu时间)

COMMAND (正在执行的命令行命令)

NI (nice)优先级

PRI 进程优先级编号

PPID 父进程的进程ID(parent process id)

SID 会话ID(session id)

WCHAN 进程正在睡眠的内核函数名称;该函数的名称是从/root/system.map文件中获得的。

FLAGS 与进程相关的数字标识

源码如下:

android下ps命令的源码的位置:android/system/core/toolbox/ps.c,其实现如下:

int ps_main(int argc, char **argv)

{

DIR *d;

struct dirent *de;

char *namefilter = 0;

int pidfilter = 0;

int threads = 0;

d = opendir("/proc");

if(d == 0) return -1;

while(argc > 1){

if(!strcmp(argv[1],"-t")) {

threads = 1;

} else if(!strcmp(argv[1],"-x")) {

display_flags |= SHOW_TIME;

} else if(!strcmp(argv[1],"-p")) {

display_flags |= SHOW_PRIO;

} else if(isdigit(argv[1][0])){

pidfilter = atoi(argv[1]);

} else {

namefilter = argv[1];

}

argc--;

argv++;

}

printf("USER PID PPID VSIZE RSS %sWCHAN PC NAME\n",

(display_flags&SHOW_PRIO)?"PRIO NICE RTPRI SCHED ":"");

while((de = readdir(d)) != 0){

if(isdigit(de->d_name[0])){

int pid = atoi(de->d_name);

if(!pidfilter || (pidfilter == pid)) {

ps_line(pid, 0, namefilter);

if(threads) ps_threads(pid, namefilter);

}

}

}

closedir(d);

return 0;

}

每一行数据是如何获得的:

static int ps_line(int pid, int tid, char *namefilter)

{

char statline[1024];

char cmdline[1024];

char user[32];

struct stat stats;

int fd, r;

char *ptr, *name, *state;

int ppid, tty;

unsigned wchan, rss, vss, eip;

unsigned utime, stime;

int prio, nice, rtprio, sched;

struct passwd *pw;

sprintf(statline, "/proc/%d", pid);

stat(statline, &stats);

if(tid) {

sprintf(statline, "/proc/%d/task/%d/stat", pid, tid);

cmdline[0] = 0;

} else {

sprintf(statline, "/proc/%d/stat", pid);

sprintf(cmdline, "/proc/%d/cmdline", pid);

fd = open(cmdline, O_RDONLY);

if(fd == 0) {

r = 0;

} else {

r = read(fd, cmdline, 1023);

close(fd);

if(r < 0) r = 0;

}

cmdline[r] = 0;

}

fd = open(statline, O_RDONLY);

if(fd == 0) return -1;

r = read(fd, statline, 1023);

close(fd);

if(r < 0) return -1;

statline[r] = 0;

ptr = statline;

nexttok(&ptr); // skip pid

ptr++; // skip "("

name = ptr;

ptr = strrchr(ptr, ')'); // Skip to *last* occurence of ')',

*ptr++ = '\0'; // and null-terminate name.

ptr++; // skip " "

state = nexttok(&ptr);

ppid = atoi(nexttok(&ptr));

nexttok(&ptr); // pgrp

nexttok(&ptr); // sid

tty = atoi(nexttok(&ptr));

nexttok(&ptr); // tpgid

nexttok(&ptr); // flags

nexttok(&ptr); // minflt

nexttok(&ptr); // cminflt

nexttok(&ptr); // majflt

nexttok(&ptr); // cmajflt

#if 1

utime = atoi(nexttok(&ptr));

stime = atoi(nexttok(&ptr));

#else

nexttok(&ptr); // utime

nexttok(&ptr); // stime

#endif

nexttok(&ptr); // cutime

nexttok(&ptr); // cstime

prio = atoi(nexttok(&ptr));

nice = atoi(nexttok(&ptr));

nexttok(&ptr); // threads

nexttok(&ptr); // itrealvalue

nexttok(&ptr); // starttime

vss = strtoul(nexttok(&ptr), 0, 10); // vsize

rss = strtoul(nexttok(&ptr), 0, 10); // rss

nexttok(&ptr); // rlim

nexttok(&ptr); // startcode

nexttok(&ptr); // endcode

nexttok(&ptr); // startstack

nexttok(&ptr); // kstkesp

eip = strtoul(nexttok(&ptr), 0, 10); // kstkeip

nexttok(&ptr); // signal

nexttok(&ptr); // blocked

nexttok(&ptr); // sigignore

nexttok(&ptr); // sigcatch

wchan = strtoul(nexttok(&ptr), 0, 10); // wchan

nexttok(&ptr); // nswap

nexttok(&ptr); // cnswap

nexttok(&ptr); // exit signal

nexttok(&ptr); // processor

rtprio = atoi(nexttok(&ptr)); // rt_priority

sched = atoi(nexttok(&ptr)); // scheduling policy

tty = atoi(nexttok(&ptr));

if(tid != 0) {

ppid = pid;

pid = tid;

}

pw = getpwuid(stats.st_uid);

if(pw == 0) {

sprintf(user,"%d",(int)stats.st_uid);

} else {

strcpy(user,pw->pw_name);

}

if(!namefilter || !strncmp(name, namefilter, strlen(namefilter))) {

printf("%-8s %-5d %-5d %-5d %-5d", user, pid, ppid, vss / 1024, rss * 4);

if(display_flags&SHOW_PRIO)

printf(" %-5d %-5d %-5d %-5d", prio, nice, rtprio, sched);

printf(" %08x %08x %s %s", wchan, eip, state, cmdline[0] ? cmdline : name);

if(display_flags&SHOW_TIME)

printf(" (u:%d, s:%d)", utime, stime);

printf("\n");

}

return 0;

}
内容来自用户分享和网络整理,不保证内容的准确性,如有侵权内容,可联系管理员处理 点击这里给我发消息
标签: