许多有用的工具被作为标准而提供。
- Memcheck是一个内存错误检测器。它有助于使你的程序,尤其是那些用C和C++写的程序,更加准确。
- Cachegrind是一个缓存和分支预测分析器。它有助于使你的程序运行更快。
- Callgrind是一个调用图缓存生成分析器。它与Cachegrind的功能有重叠,但也收集Cachegrind不收集的一些信息。
- Helgrind是一个线程错误检测器。它有助于使你的多线程程序更加准确。
- DRD也是一个线程错误检测器。它和Helgrind相似,但使用不同的分析技术,所以可能找到不同的问题。
- Massif是一个堆分析器。它有助于使你的程序使用更少的内存。
- DHAT是另一种不同的堆分析器。它有助于理解块的生命期、块的使用和布局的低效等问题。
- SGcheck是一个实验工具,用来检测堆和全局数组的溢出。它的功能和Memcheck互补:SGcheck找到Memcheck无法找到的问题,反之亦然。
- BBV是个实验性质的SimPoint基本块矢量生成器。它对于进行计算机架构的研究和开发很有用处。
也有一些对大多数用户没有用的小工具:Lackey是演示仪器基础的示例工具;Nulgrind是一个最小化的Valgrind工具,不做分析或者操作,仅用于测试目的。
在这篇文章我们将关注“memcheck”工具。
使用 Valgrind Memcheck
memcheck工具的使用方式如下:
1 | valgrind --tool=memcheck ./a.out |
从上面的命令可以清楚的看到, 主要的命令是valgrind,而我们想使用的工具是通过'-tool'选项来指定的. 上面的‘a.out’指的是我们想使用memcheck运行的可执行文件.
该工具可以检测下列与内存相关的问题 :
- 未释放内存的使用
- 对释放后内存的读/写
- 对已分配内存块尾部的读/写
- 内存泄露
- 不匹配的使用malloc/new/new[] 和 free/delete/delete[]
- 重复释放内存
注意: 上面列出的并不很全面,但却包含了能被该工具检测到的很多普遍的问题.
让我们一个一个地对上面的场景进行讨论:
注意: 下面讨论的所有测试代码都应该使用gcc并且加上-g选项(用来在memcheck的输出中生成行号)进行编译. 就想我们之前讨论过的 C程序被编译成可执行文件, 它需要经历四个不同的阶段.
1. 使用未初始化的内存
Code :
10 | printf ( "\n [%c]\n" ,c); |
在上面的代码中,我们尝试使用未初始化的指针 ‘p’.
让我们运行Memcheck来看下结果.
01 | $ valgrind --tool=memcheck ./val |
02 | ==2862== Memcheck, a memory error detector |
03 | ==2862== Copyright (C) 2002-2009, and GNU GPL'd, by Julian Seward et al. |
04 | ==2862== Using Valgrind-3.6.0.SVN-Debian and LibVEX; rerun with -h for copyright info |
05 | ==2862== Command: ./val |
07 | ==2862== Use of uninitialised value of size 8 |
08 | ==2862== at 0x400530: main (valgrind.c:8) |
14 | ==2862== in use at exit : 0 bytes in 0 blocks |
15 | ==2862== total heap usage: 0 allocs, 0 frees, 0 bytes allocated |
17 | ==2862== All heap blocks were freed -- no leaks are possible |
19 | ==2862== For counts of detected and suppressed errors, rerun with: -v |
20 | ==2862== Use --track-origins=yes to see where uninitialized values come from |
21 | ==2862== ERROR SUMMARY: 1 errors from 1 contexts (suppressed: 4 from 4) |
从上面的输出可以看到,Valgrind检测到了未初始化的变量,然后给出了警告(上面加粗的几行(译者注:貌似上面没有加粗的)).
2. 在内存被释放后进行读/写
Code :
11 | printf ( "\n [%c]\n" ,c); |
上面的代码中,我们有一个释放了内存的指针 ‘p’ 然后我们又尝试利用指针获取值.
让我们运行memcheck来看一下Valgrind对这种情况是如何反应的.
01 | $ valgrind --tool=memcheck ./val |
02 | ==2849== Memcheck, a memory error detector |
03 | ==2849== Copyright (C) 2002-2009, and GNU GPL'd, by Julian Seward et al. |
04 | ==2849== Using Valgrind-3.6.0.SVN-Debian and LibVEX; rerun with -h for copyright info |
05 | ==2849== Command: ./val |
09 | ==2849== Invalid read of size 1 |
10 | ==2849== at 0x400603: main (valgrind.c:30) |
11 | ==2849== Address 0x51b0040 is 0 bytes inside a block of size 1 free 'd |
12 | ==2849== at 0x4C270BD: free (vg_replace_malloc.c:366) |
13 | ==2849== by 0x4005FE: main (valgrind.c:29) |
17 | ==2849== in use at exit : 0 bytes in 0 blocks |
18 | ==2849== total heap usage: 1 allocs, 1 frees, 1 bytes allocated |
20 | ==2849== All heap blocks were freed -- no leaks are possible |
22 | ==2849== For counts of detected and suppressed errors, rerun with: -v |
23 | ==2849== ERROR SUMMARY: 1 errors from 1 contexts (suppressed: 4 from 4) |
从上面的输出内容可以看到,Valgrind检测到了无效的读取操作然后输出了警告 ‘Invalid read of size 1′.
另注,使用gdb来调试c程序.
3. 从已分配内存块的尾部进行读/写
Code :
11 | printf ( "\n [%c]\n" ,c); |
在上面的代码中,我们已经为‘p’分配了一个字节的内存,但我们在将值读取到 ‘c’中的时候使用的是地址p+1.
现在我们使用Valgrind运行上面的代码 :
01 | $ valgrind --tool=memcheck ./val |
02 | ==2835== Memcheck, a memory error detector |
03 | ==2835== Copyright (C) 2002-2009, and GNU GPL'd, by Julian Seward et al. |
04 | ==2835== Using Valgrind-3.6.0.SVN-Debian and LibVEX; rerun with -h for copyright info |
05 | ==2835== Command: ./val |
07 | ==2835== Invalid read of size 1 |
08 | ==2835== at 0x4005D9: main (valgrind.c:25) |
09 | ==2835== Address 0x51b0041 is 0 bytes after a block of size 1 alloc'd |
10 | ==2835== at 0x4C274A8: malloc (vg_replace_malloc.c:236) |
11 | ==2835== by 0x4005C5: main (valgrind.c:22) |
17 | ==2835== in use at exit : 0 bytes in 0 blocks |
18 | ==2835== total heap usage: 1 allocs, 1 frees, 1 bytes allocated |
20 | ==2835== All heap blocks were freed -- no leaks are possible |
22 | ==2835== For counts of detected and suppressed errors, rerun with: -v |
23 | ==2835== ERROR SUMMARY: 1 errors from 1 contexts (suppressed: 4 from 4) |
同样,该工具在这种情况下也检测到了无效的读取操作.
4. 内存泄露
Code:
11 | printf ( "\n [%c]\n" ,c); |
在这次的代码中, 我们申请了一个字节但是没有将它释放.现在让我们运行Valgrind看看会发生什么:
01 | $ valgrind --tool=memcheck --leak-check=full ./val |
02 | ==2888== Memcheck, a memory error detector |
03 | ==2888== Copyright (C) 2002-2009, and GNU GPL'd, by Julian Seward et al. |
04 | ==2888== Using Valgrind-3.6.0.SVN-Debian and LibVEX; rerun with -h for copyright info |
05 | ==2888== Command: ./val |
11 | ==2888== in use at exit : 1 bytes in 1 blocks |
12 | ==2888== total heap usage: 1 allocs, 0 frees, 1 bytes allocated |
14 | ==2888== 1 bytes in 1 blocks are definitely lost in loss record 1 of 1 |
15 | ==2888== at 0x4C274A8: malloc (vg_replace_malloc.c:236) |
16 | ==2888== by 0x400575: main (valgrind.c:6) |
19 | ==2888== definitely lost: 1 bytes in 1 blocks |
20 | ==2888== indirectly lost: 0 bytes in 0 blocks |
21 | ==2888== possibly lost: 0 bytes in 0 blocks |
22 | ==2888== still reachable: 0 bytes in 0 blocks |
23 | ==2888== suppressed: 0 bytes in 0 blocks |
25 | ==2888== For counts of detected and suppressed errors, rerun with: -v |
26 | ==2888== ERROR SUMMARY: 1 errors from 1 contexts (suppressed: 4 from 4) |
输出行(上面加粗的部分)显示,该工具能够检测到内存的泄露.
注意: 在这里我们增加了一个选项‘–leak-check=full’来得到内存泄露的详细细节.
5. 不匹配地使用malloc/new/new[] 和 free/delete/delete[]
Code:
07 | char *p = ( char *) malloc (1); |
12 | printf ( "\n [%c]\n" ,c); |
上面的代码中,我们使用了malloc()来分配内存,但是使用了delete操作符来删除内存.
注意 : 使用g++来编译上面的代码,因为delete操作符是在C++中引进的,而要编译C++需要使用g++.
让我们运行来看一下 :
01 | $ valgrind --tool=memcheck --leak-check=full ./val |
02 | ==2972== Memcheck, a memory error detector |
03 | ==2972== Copyright (C) 2002-2009, and GNU GPL'd, by Julian Seward et al. |
04 | ==2972== Using Valgrind-3.6.0.SVN-Debian and LibVEX; rerun with -h for copyright info |
05 | ==2972== Command: ./val |
09 | ==2972== Mismatched free () / delete / delete [] |
10 | ==2972== at 0x4C26DCF: operator delete ( void *) (vg_replace_malloc.c:387) |
11 | ==2972== by 0x40080B: main (valgrind.c:13) |
12 | ==2972== Address 0x595e040 is 0 bytes inside a block of size 1 alloc'd |
13 | ==2972== at 0x4C274A8: malloc (vg_replace_malloc.c:236) |
14 | ==2972== by 0x4007D5: main (valgrind.c:7) |
18 | ==2972== in use at exit : 0 bytes in 0 blocks |
19 | ==2972== total heap usage: 1 allocs, 1 frees, 1 bytes allocated |
21 | ==2972== All heap blocks were freed -- no leaks are possible |
23 | ==2972== For counts of detected and suppressed errors, rerun with: -v |
24 | ==2972== ERROR SUMMARY: 1 errors from 1 contexts (suppressed: 4 from 4) |
从上面的输出可以看到 (加粗的行), Valgrind清楚的说明了‘不匹配的使用了free() / delete / delete []‘
你可以尝试在测试代码中使用'new'和'free'进行组合来看看Valgrind给出的结果是什么.
6. 两次释放内存
Code :
06 | char *p = ( char *) malloc (1); |
10 | printf ( "\n [%c]\n" ,c); |
在上面的代码中, 我们两次释放了'p'指向的内存. 现在让我们运行memcheck :
01 | $ valgrind --tool=memcheck --leak-check=full ./val |
02 | ==3167== Memcheck, a memory error detector |
03 | ==3167== Copyright (C) 2002-2009, and GNU GPL'd, by Julian Seward et al. |
04 | ==3167== Using Valgrind-3.6.0.SVN-Debian and LibVEX; rerun with -h for copyright info |
05 | ==3167== Command: ./val |
09 | ==3167== Invalid free () / delete / delete [] |
10 | ==3167== at 0x4C270BD: free (vg_replace_malloc.c:366) |
11 | ==3167== by 0x40060A: main (valgrind.c:12) |
12 | ==3167== Address 0x51b0040 is 0 bytes inside a block of size 1 free 'd |
13 | ==3167== at 0x4C270BD: free (vg_replace_malloc.c:366) |
14 | ==3167== by 0x4005FE: main (valgrind.c:11) |
18 | ==3167== in use at exit : 0 bytes in 0 blocks |
19 | ==3167== total heap usage: 1 allocs, 2 frees, 1 bytes allocated |
21 | ==3167== All heap blocks were freed -- no leaks are possible |
23 | ==3167== For counts of detected and suppressed errors, rerun with: -v |
24 | ==3167== ERROR SUMMARY: 1 errors from 1 contexts (suppressed: 4 from 4) |
从上面的输出可以看到(加粗的行), 该功能检测到我们对同一个指针调用了两次释放内存操作.
在本文中,我们把注意力放在了内存管理框架Valgrind,然后使用memcheck(Valgrind框架提供的)工具来了
解它是如何降低需要经常操作内存的程序员的负担的. 该工具能够检测到很多手动检测不到的与内存相关的问题