Linux/Unix用valgrind检测内存泄漏

时间:2022-12-17 15:18:01

c\c++编程中,内存管理是比较头疼的问题,一不小心就会造成内存泄漏,除了养成良好的编程习惯外(使用智能指针等),使用工具也能帮助检测内存泄漏,valgrind是Unix\Linux下一款不错的工具。(mac下暂时有bug)


Valgrind 安装
  1. 到www.valgrind.org下载最新版valgrind-3.2.3.tar.bz2
  2. 解压安装包:tar –jxvf valgrind-3.2.3.tar.bz2
  3. 解压后生成目录valgrind-3.2.3
  4. cd valgrind-3.2.3
  5. 运行./autogen.sh设置环境(需要标准的autoconf工具)(可选)
  6. ./configure;配置Valgrind,生成MakeFile文件,具体参数信息详见INSTALL文件。一般只需要设置—prefix=/where/you/want/it/installed
  7. Make;编译Valgrind
  8. make install;安装Valgrind

ubuntu 直接用命令 sudo apt-get install valgrind安装
Valgrind包含的工具
Valgrind支持很多工具:memcheck,addrcheck,cachegrind,Massif,helgrind和Callgrind等。在运行Valgrind时,你必须指明想用的工具,如果省略工具名,默认运行memcheck。

1、memcheck

memcheck探测程序中内存管理存在的问题。它检查所有对内存的读/写操作,并截取所有的malloc/new/free/delete调用。因此memcheck工具能够探测到以下问题:
1)使用未初始化的内存
2)读/写已经被释放的内存
3)读/写内存越界
4)读/写不恰当的内存栈空间
5)内存泄漏
6)使用malloc/new/new[]和free/delete/delete[]不匹配。
7)src和dst的重叠

2、cachegrind

cachegrind是一个cache剖析器。它模拟执行CPU中的L1, D1和L2 cache,因此它能很精确的指出代码中的cache未命中。如果你需要,它可以打印出cache未命中的次数,内存引用和发生cache未命中的每一行代码,每一个函数,每一个模块和整个程序的摘要。如果你要求更细致的信息,它可以打印出每一行机器码的未命中次数。在x86和amd64上, cachegrind通过CPUID自动探测机器的cache配置,所以在多数情况下它不再需要更多的配置信息了。

3、helgrind

helgrind查找多线程程序中的竞争数据。helgrind查找内存地址,那些被多于一条线程访问的内存地址,但是没有使用一致的锁就会被查出。这表示这些地址在多线程间访问的时候没有进行同步,很可能会引起很难查找的时序问题。
它主要用来检查多线程程序中出现的竞争问题。Helgrind 寻找内存中被多个线程访问,而又没有一贯加锁的区域,这些区域往往是线程之间失去同步的地方,而且会导致难以发掘的错误。Helgrind实现了名为”Eraser” 的竞争检测算法,并做了进一步改进,减少了报告错误的次数。

4、Callgrind

Callgrind收集程序运行时的一些数据,函数调用关系等信息,还可以有选择地进行cache 模拟。在运行结束时,它会把分析数据写入一个文件。callgrind_annotate可以把这个文件的内容转化成可读的形式。
一般用法:
$valgrind —tool=callgrind ./sec_infod
会在当前目录下生成callgrind.out.[pid], 如果我们想结束程序, 可以
$killall callgrind
然后我们可以用
$callgrind_annotate —auto=yes callgrind.out.[pid] > log
$vi log

5、Massif

堆栈分析器,它能测量程序在堆栈中使用了多少内存,告诉我们堆块,堆管理块和栈的大小。Massif能帮助我们减少内存的使用,在带有虚拟内存的现代系统中,它还能够加速我们程序的运行,减少程序停留在交换区中的几率。

6、lackey

lackey是一个示例程序,以其为模版可以创建你自己的工具。在程序结束后,它打印出一些基本的关于程序执行统计数据。

Valgrind的参数

用法: valgrind [options] prog-and-args [options]: 常用选项,适用于所有####Valgrind工具
—tool=
最常用的选项。运行 valgrind中名为toolname的工具。默认memcheck。
-h —help
显示所有选项的帮助,包括内核和选定的工具两者。
—version
显示valgrind内核的版本,每个工具都有各自的版本。
-q —quiet
安静地运行,只打印错误信息。
—verbose
更详细的信息。
—trace-children=
跟踪子线程? [default: no]
—track-fds=
跟踪打开的文件描述?[default: no]
—time-stamp=
增加时间戳到LOG信息? [default: no]
—log-fd=
输出LOG到描述符文件 [2=stderr]
—log-file=
将输出的信息写入到filename.PID的文件里,PID是运行程序的进行ID
—log-file-exactly=
输出LOG信息到 file
LOG信息输出
—xml=yes
将信息以xml格式输出,只有memcheck可用
—num-callers=
show callers in stack traces [12]
—error-exitcode=
如果发现错误则返回错误代码 [0=disable]
—db-attach=
当出现错误,valgrind会自动启动调试器gdb。[default: no]
—db-command=
启动调试器的命令行选项[gdb -nw %f %p]
适用于Memcheck工具的相关选项:
—leak-check=
要求对leak给出详细信息? Leak是指,存在一块没有被引用的内存空间,或没有被释放的内存空间,如summary,只反馈一些总结信息,告诉你有多少个malloc,多少个free 等;如果是full将输出所有的leaks,也就是定位到某一个malloc/free。 [default: summary]
—show-reachable=
如果为no,只输出没有引用的内存leaks,或指向malloc返回的内存块中部某处的leaks [default: no]
更详细的参数指令见附录A。
Valgrind的使用
首先,在编译程序的时候打开调试模式(gcc编译器的-g选项)。如果没有调试信息,即使最好的valgrind工具也将中能够猜测特定的代码是属于哪一个函数。打开调试选项进行编译后再用valgrind检查,valgrind将会给你的个详细的报告,比如哪一行代码出现了内存泄漏。
当检查的是C++程序的时候,还应该考虑另一个选项 -fno-inline。它使得函数调用链很清晰,这样可以减少你在浏览大型C++程序时的混乱。比如在使用这个选项的时候,用memcheck检查openoffice就很容易。当然,你可能不会做这项工作,但是使用这一选项使得valgrind生成更精确的错误报告和减少混乱。
一些编译优化选项(比如-O2或者更高的优化选项),可能会使得memcheck提交错误的未初始化报告,因此,为了使得valgrind的报告更精确,在编译的时候最好不要使用优化选项。
如果程序是通过脚本启动的,可以修改脚本里启动程序的代码,或者使用—trace-children=yes选项来运行脚本。
下面是用memcheck检查sample.c的例子
这里用到的示例程序文件名为:sample.c(如下所示),选用的编译器为gcc。

123456789101112
/* sample.c */#include <stdlib.h>void fun(){    int *p = (int*)malloc(10*sizeof(int));    p[10] = 0;}int main(void){    fun();    return 0;}

生成可执行程序
gcc –g sample.c –o sample
运行Valgrind
valgrind --tool=memcheck ./sample
以下是运行上述命令后的输出
==23913== Memcheck, a memory error detector
==23913== Copyright (C) 2002-2013, and GNU GPL’d, by Julian Seward et al.
==23913== Using Valgrind-3.10.0.SVN and LibVEX; rerun with -h for copyright info
==23913== Command: ./sample
==23913==
==23913== Invalid write of size 4
==23913== at 0x8048438: fun (sample.c:6)
==23913== by 0x804844A: main (sample.c:11)
==23913== Address 0x4204050 is 0 bytes after a block of size 40 alloc’d
==23913== at 0x402A17C: malloc (in /usr/lib/valgrind/vgpreload_memcheck-x86-linux.so)
==23913== by 0x804842E: fun (sample.c:5)
==23913== by 0x804844A: main (sample.c:11)
==23913==
==23913==
==23913== HEAP SUMMARY:
==23913== in use at exit: 40 bytes in 1 blocks
==23913== total heap usage: 1 allocs, 0 frees, 40 bytes allocated
==23913==
==23913== LEAK SUMMARY:
==23913== definitely lost: 40 bytes in 1 blocks
==23913== indirectly lost: 0 bytes in 0 blocks
==23913== possibly lost: 0 bytes in 0 blocks
==23913== still reachable: 0 bytes in 0 blocks
==23913== suppressed: 0 bytes in 0 blocks
==23913== Rerun with —leak-check=full to see details of leaked memory
==23913==
==23913== For counts of detected and suppressed errors, rerun with: -v
==23913== ERROR SUMMARY: 1 errors from 1 contexts (suppressed: 0 from 0)
左边显示类似行号的数字(23913)表示的是 Process ID。
最上面的红色方框表示的是 valgrind 的版本信息。
中间的红色方框表示 valgrind 通过运行被测试程序,发现的内存问题。通过阅读这些信息,可以发现:
l 这是一个对内存的非法写操作,非法写操作的内存是4 bytes。
l 发生错误时的函数堆栈,以及具体的源代码行号。
l 非法写操作的具体地址空间。
最下面的红色方框是对发现的内存问题和内存泄漏问题的总结。内存泄漏的大小(40 bytes)也能够被检测出来。
Valgrind的示例
例1.使用未初始化的内存
代码:

12345678910
#include <stdio.h>int main(){    int x;    if(x == 0)    {    printf("X is zero");    }    return 0;}

Valgrind提示如下
==14222== Conditional jump or move depends on uninitialised value(s)
==14222== at 0x400484: main (sample2.c:6)
X is zero==14222==
==14222== ERROR SUMMARY: 1 errors from 1 contexts (suppressed: 5 from 1)
==14222== malloc/free: in use at exit: 0 bytes in 0 blocks.
==14222== malloc/free: 0 allocs, 0 frees, 0 bytes allocated.
==14222== For counts of detected errors, rerun with: -v
==14222== All heap blocks were freed — no leaks are possible.
例2.内存读写越界
代码如下:

123456789101112131415
#include <stdlib.h>#include <stdio.h>int main(int argc,char *argv[]){    int len=5;    int i;    int *pt=(int*)malloc(len*sizeof(int));    int *p=pt;    for(i=0;i<len;i++)    {p++;}    *p=5;    printf(“%d”,*p);    return;}

Valgrind提示如下
==23045== Invalid write of size 4
==23045== at 0x40050A: main (sample2.c:11)
==23045== Address 0x4C2E044 is 0 bytes after a block of size 20 alloc’d
==23045== at 0x4A05809: malloc (vg_replace_malloc.c:149)
==23045== by 0x4004DF: main (sample2.c:7)
==23045==
==23045== Invalid read of size 4
==23045== at 0x400514: main (sample2.c:12)
==23045== Address 0x4C2E044 is 0 bytes after a block of size 20 alloc’d
==23045== at 0x4A05809: malloc (vg_replace_malloc.c:149)
==23045== by 0x4004DF: main (sample2.c:7)
5==23045==
==23045== ERROR SUMMARY: 2 errors from 2 contexts (suppressed: 5 from 1)
==23045== malloc/free: in use at exit: 20 bytes in 1 blocks.
==23045== malloc/free: 1 allocs, 0 frees, 20 bytes allocated.
==23045== For counts of detected errors, rerun with: -v
==23045== searching for pointers to 1 not-freed blocks.
==23045== checked 66,584 bytes.
==23045==
==23045== LEAK SUMMARY:
==23045== definitely lost: 20 bytes in 1 blocks.
==23045== possibly lost: 0 bytes in 0 blocks.
==23045== still reachable: 0 bytes in 0 blocks.
==23045== suppressed: 0 bytes in 0 blocks.
==23045== Use —leak-check=full to see details of leaked memory.
例3.src和dst内存覆盖
代码如下:

123456789101112131415161718
#include <stdlib.h>#include <stdio.h>#include <string.h>int main(int argc,char *argv[]){     char x[50];    int i;    for(i=0;i<50;i++)    {x[i]=i;}    strncpy(x+20,x,20); //Good    strncpy(x+20,x,21); //Overlap    x[39]=’\0’;    strcpy(x,x+20); //Good    x[39]=40;    x[40]=’\0’;    strcpy(x,x+20); //Overlap    return 0;}

Valgrind提示如下
==24139== Source and destination overlap in strncpy(0x7FEFFFC09, 0x7FEFFFBF5, 21)
==24139== at 0x4A0724F: strncpy (mc_replace_strmem.c:116)
==24139== by 0x400527: main (sample3.c:10)
==24139==
==24139== Source and destination overlap in strcpy(0x7FEFFFBE0, 0x7FEFFFBF4)
==24139== at 0x4A06E47: strcpy (mc_replace_strmem.c:106)
==24139== by 0x400555: main (sample3.c:15)
==24139==
==24139== ERROR SUMMARY: 2 errors from 2 contexts (suppressed: 5 from 1)
==24139== malloc/free: in use at exit: 0 bytes in 0 blocks.
==24139== malloc/free: 0 allocs, 0 frees, 0 bytes allocated.
==24139== For counts of detected errors, rerun with: -v
==24139== All heap blocks were freed — no leaks are possible.
例4.动态内存管理错误
常见的内存分配方式分三种:静态存储,栈上分配,堆上分配。全局变量属于静态存储,它们是在编译时就被分配了存储空间,函数内的局部变量属于栈上分配,而最灵活的内存使用方式当属堆上分配,也叫做内存动态分配了。常用的内存动态分配函数包括:malloc, alloc, realloc, new等,动态释放函数包括free, delete。
一旦成功申请了动态内存,我们就需要自己对其进行内存管理,而这又是最容易犯错误的。常见的内存动态管理错误包括:
l 申请和释放不一致
由于 C++ 兼容 C,而 C 与 C++ 的内存申请和释放函数是不同的,因此在 C++ 程序中,就有两套动态内存管理函数。一条不变的规则就是采用 C 方式申请的内存就用 C 方式释放;用 C++ 方式申请的内存,用 C++ 方式释放。也就是用 malloc/alloc/realloc 方式申请的内存,用 free 释放;用 new 方式申请的内存用 delete 释放。在上述程序中,用 malloc 方式申请了内存却用 delete 来释放,虽然这在很多情况下不会有问题,但这绝对是潜在的问题。
l 申请和释放不匹配
申请了多少内存,在使用完成后就要释放多少。如果没有释放,或者少释放了就是内存泄露;多释放了也会产生问题。上述程序中,指针p和pt指向的是同一块内存,却被先后释放两次。
l 释放后仍然读写
本质上说,系统会在堆上维护一个动态内存链表,如果被释放,就意味着该块内存可以继续被分配给其他部分,如果内存被释放后再访问,就可能覆盖其他部分的信息,这是一种严重的错误,上述程序第16行中就在释放后仍然写这块内存。
下面的一段程序,就包括了内存动态管理中常见的错误.

1234567891011121314
#include <stdlib.h>#include <stdio.h>int main(int argc,char *argv[]){     char *p=(char*)malloc(10);    char *pt=p;    int i;    for(i=0;i<10;i++)    {p[i]=’z’;}    delete p;    p[1]=’a’;    free(pt);    return 0;}

Valgrind提示如下
==25811== Mismatched free() / delete / delete []
==25811== at 0x4A05130: operator delete(void) (vg_replace_malloc.c:244)
==25811== by 0x400654: main (sample4.c:9)
==25811== Address 0x4C2F030 is 0 bytes inside a block of size 10 alloc’d
==25811== at 0x4A05809: malloc (vg_replace_malloc.c:149)
==25811== by 0x400620: main (sample4.c:4)
==25811==
==25811== Invalid write of size 1
==25811== at 0x40065D: main (sample4.c:10)
==25811== Address 0x4C2F031 is 1 bytes inside a block of size 10 free’d
==25811== at 0x4A05130: operator delete(void
) (vg_replace_malloc.c:244)
==25811== by 0x400654: main (sample4.c:9)
==25811==
==25811== Invalid free() / delete / delete[]
==25811== at 0x4A0541E: free (vg_replace_malloc.c:233)
==25811== by 0x400668: main (sample4.c:11)
==25811== Address 0x4C2F030 is 0 bytes inside a block of size 10 free’d
==25811== at 0x4A05130: operator delete(void*) (vg_replace_malloc.c:244)
==25811== by 0x400654: main (sample4.c:9)
==25811==
==25811== ERROR SUMMARY: 3 errors from 3 contexts (suppressed: 5 from 1)
==25811== malloc/free: in use at exit: 0 bytes in 0 blocks.
==25811== malloc/free: 1 allocs, 2 frees, 10 bytes allocated.
==25811== For counts of detected errors, rerun with: -v
==25811== All heap blocks were freed — no leaks are possible.
例5.内存泄漏
代码如下:

12345678910
#include <stdlib.h>int main(){    char *x = (char*)malloc(20);    char *y = (char*)malloc(20);    x=y;    free(x);    free(y);    return 0;}

Valgrind提示如下
==19013== Invalid free() / delete / delete[]
==19013== at 0x4A0541E: free (vg_replace_malloc.c:233)
==19013== by 0x4004F5: main (sample5.c:8)
==19013== Address 0x4C2E078 is 0 bytes inside a block of size 20 free’d
==19013== at 0x4A0541E: free (vg_replace_malloc.c:233)
==19013== by 0x4004EC: main (sample5.c:7)
==19013==
==19013== ERROR SUMMARY: 1 errors from 1 contexts (suppressed: 5 from 1)
==19013== malloc/free: in use at exit: 20 bytes in 1 blocks.
==19013== malloc/free: 2 allocs, 2 frees, 40 bytes allocated.
==19013== For counts of detected errors, rerun with: -v
==19013== searching for pointers to 1 not-freed blocks.
==19013== checked 66,584 bytes.
==19013==
==19013== LEAK SUMMARY:
==19013== definitely lost: 20 bytes in 1 blocks.
==19013== possibly lost: 0 bytes in 0 blocks.
==19013== still reachable: 0 bytes in 0 blocks.
==19013== suppressed: 0 bytes in 0 blocks.
==19013== Use —leak-check=full to see details of leaked memory.
例6.非法写/读
代码如下:

123456789
int main(){    int i, *x;    x = (int *)malloc(10*sizeof(int));    for (i=0; i<11; i++)        x[i] = i;    free(x);}

Valgrind提示如下
==21483== Invalid write of size 4
==21483== at 0x4004EA: main (sample6.c:6)
==21483== Address 0x4C2E058 is 0 bytes after a block of size 40 alloc’d
==21483== at 0x4A05809: malloc (vg_replace_malloc.c:149)
==21483== by 0x4004C9: main (sample6.c:4)
==21483==
==21483== ERROR SUMMARY: 1 errors from 1 contexts (suppressed: 5 from 1)
==21483== malloc/free: in use at exit: 0 bytes in 0 blocks.
==21483== malloc/free: 1 allocs, 1 frees, 40 bytes allocated.
==21483== For counts of detected errors, rerun with: -v
==21483== All heap blocks were freed — no leaks are possible.
例7.无效指针
代码如下:

123456789
#include <stdlib.h>int main(){    char *x = malloc(10);    x[10] = 'a';    free(x);    return 0;}

Valgrind提示如下
==15262== Invalid write of size 1
==15262== at 0x4004D6: main (sample7.c:5)
==15262== Address 0x4C2E03A is 0 bytes after a block of size 10 alloc’d
==15262== at 0x4A05809: malloc (vg_replace_malloc.c:149)
==15262== by 0x4004C9: main (sample7.c:4)
==15262==
==15262== ERROR SUMMARY: 1 errors from 1 contexts (suppressed: 5 from 1)
==15262== malloc/free: in use at exit: 0 bytes in 0 blocks.
==15262== malloc/free: 1 allocs, 1 frees, 10 bytes allocated.
==15262== For counts of detected errors, rerun with: -v
==15262== All heap blocks were freed — no leaks are possible.
例8.重复释放
代码:

123456789
#include <stdlib.h>int main(){    char *x = malloc(10);    free(x);    free(x);    return 0;}

Valgrind提示如下
==15005== Invalid free() / delete / delete[]
==15005== at 0x4A0541E: free (vg_replace_malloc.c:233)
==15005== by 0x4004DF: main (sample8.c:6)
==15005== Address 0x4C2E030 is 0 bytes inside a block of size 10 free’d
==15005== at 0x4A0541E: free (vg_replace_malloc.c:233)
==15005== by 0x4004D6: main (sample8.c:5)
==15005==
==15005== ERROR SUMMARY: 1 errors from 1 contexts (suppressed: 5 from 1)
==15005== malloc/free: in use at exit: 0 bytes in 0 blocks.
==15005== malloc/free: 1 allocs, 2 frees, 10 bytes allocated.
==15005== For counts of detected errors, rerun with: -v
==15005== All heap blocks were freed — no leaks are possible.

Valgrind的局限

l Valgrind不对静态数组(分配在栈上)进行边界检查。如果在程序中声明了一个数组:

12345
int main(){    char x[10];    x[11] = 'a';}

Valgrind则不会警告你,你可以把数组改为动态在堆上分配的数组,这样就可能进行边界检查了。这个方法好像有点得不偿失的感觉。
l Valgrind占用了更多的内存—可达两倍于你程序的正常使用量。如果你用Valgrind来检测使用大量内存的程序就会遇到问题,它可能会用很长的时间来运行测试。大多数情况下,这都不是问题,即使速度慢也仅是检测时速度慢,如果你用Valgrind来检测一个正常运行时速度就很慢的程序,这下问题就大了。 Valgrind不可能检测出你在程序中犯下的所有错误—如果你不检查缓冲区溢出,Valgrind也不会告诉你代码写了它不应该写的内存。


参考:

Valgrind User Manual
应用Valgrind发现Linux程序的内存问题