Perf的安装与简单使用

Linux Perf 的安装与使用

Ubuntu下安装Perf（Ubuntu18.04）

1. 使用apt包工具安装和下载

# 下载linux-tools-common
zouren@ubuntu:~$ sudo apt-get install linux-tools-common
# 查看是否存在perf
zouren@ubuntu:~$ perf --version
# 如果不存在，可以下载特定的内核版本下的tools，根据命令行的提示，我的命令是， 这个命令需要下载特定内核版本的工具， 命令 uname -r 查看内核版本
zouren@ubuntu:~$ sudo apt-get install linux-tools-5.4.0-122-generic
# 再检查一下
zouren@ubuntu:~$ perf --version
perf version 5.4.192

详细过程可查：基于Ubuntu 18.04 安装perf工具 - 简书 (jianshu.com)

2. 源码安装Perf, 拿到Perf的源码

# 查看自己内核的版本，到官网上去下载特定的内核源码
>$ uname -r
5.4.0-122-generic
# 去官网上下载内核源码，可以手动下载，也可以使用wget
>$ wget http://ftp.sjtu.edu.cn/sites/ftp.kernel.org/pub/linux/kernel/v5.x/linux-5.4.122.tar.gz
# 下载完毕之后，接下内核源代码
>$ tar -zxvf linux-5.4.122.tar.gz
# 进入如下目录
>$ cd linux-5.4.122/tools/perf/
# 源码级安装, 如有些依赖包没有安装，得安装一下，依赖包在下面第二个链接
>$ make -j10 && make install
# 查看perf的安装情况
>$ perf --version
perf version 5.4.192

Linux内核源码下载网站：Index of /sites/ftp.kernel.org/pub/linux/kernel/ (sjtu.edu.cn)

Perf源码安装的依赖：ubuntu源码安装性能分析工具perf - 知乎 (zhihu.com)

Perf的使用

>$ perf --help

 usage: perf [--version] [--help] [OPTIONS] COMMAND [ARGS]

 The most commonly used perf commands are:
   annotate        Read perf.data (created by perf record) and display annotated code
   archive         Create archive with object files with build-ids found in perf.data file
   bench           General framework for benchmark suites
   buildid-cache   Manage build-id cache.
   buildid-list    List the buildids in a perf.data file
   c2c             Shared Data C2C/HITM Analyzer.
   config          Get and set variables in a configuration file.
   data            Data file related processing
   diff            Read perf.data files and display the differential profile
   evlist          List the event names in a perf.data file
   ftrace          simple wrapper for kernel's ftrace functionality
   inject          Filter to augment the events stream with additional information
   kallsyms        Searches running kernel for symbols
   kmem            Tool to trace/measure kernel memory properties
   kvm             Tool to trace/measure kvm guest os
   list            List all symbolic event types
   lock            Analyze lock events
   mem             Profile memory accesses
   record          Run a command and record its profile into perf.data
   report          Read perf.data (created by perf record) and display the profile
   sched           Tool to trace/measure scheduler properties (latencies)
   script          Read perf.data (created by perf record) and display trace output
   stat            Run a command and gather performance counter statistics
   test            Runs sanity tests.
   timechart       Tool to visualize total system behavior during a workload
   top             System profiling tool.
   version         display the version of perf binary
   probe           Define new dynamic tracepoints
   trace           strace inspired tool

 See 'perf help COMMAND' for more information on a specific command.

示例程序(矩阵的乘法)

// example.c
#include <stdio.h>
#include <stdlib.h>
#include <time.h>


#define N 1024

double A[N][N];
double B[N][N];
double C[N][N];


int main(){
    /* init srand */
    srand((unsigned)time(NULL));

    /* init matrix */
    for(int i = 0 ; i < N ; ++i){
        for(int j = 0 ; j < N ; ++j){
            A[i][j] = rand()/32767.0;
            B[i][j] = rand()/32767.0;
            C[i][j] = 0;
        }
    }

    /* mult matrix */
    for(int i = 0 ; i < N ; ++i){
        for(int j = 0 ; j < N ; ++j){
            for(int k = 0 ; k < N ; ++k){
                C[i][j] += A[i][k] * B[k][j];
            }
        }
    }

    return 0;
}

简单使用perf

# 1. 编译目标位可执行文件
>$ gcc -g example.c -o example
# 2. 使用perf分析性能
>$ perf stat exapmle

 Performance counter stats for './example':

          3,617.38 msec task-clock                #    0.998 CPUs utilized          
                10      context-switches          #    0.003 K/sec                  
                 0      cpu-migrations            #    0.000 K/sec                  
             6,189      page-faults               #    0.002 M/sec                  
   <not supported>      cycles                                                      
   <not supported>      instructions                                                
   <not supported>      branches                                                    
   <not supported>      branch-misses                                               

       3.624806187 seconds time elapsed

       3.614249000 seconds user
       0.003993000 seconds sys
# 如果没有上面的报告，报错了，可能要配置一下/etc/sysctl.conf，将里面的 kernel.perf_event_paranoid设置为-1， 具体操作如下
>$ sudo gedit /etc/sysctl.conf
# 然后将kernel.perf_event_paranoid = -1 写入，更新一下
>$ sudo sysctl -p