我正在解决hackerbank问题https://www.hackerrank.com/challenges/crush/problem.这是代码：
long arrayManipulation(int n, vector<vector<int>> queries) {
long ans[n] = {0};
cout << sizeof(ans)/sizeof(long) << endl;
cout << queries.size() << " "<< n << endl;
for (long i = 0; i < queries.size(); i++) {
long s = queries[i][0];
long e = queries[i][1];
long k = queries[i][2];
//cout << s << " start " << e << " end " << endl;
ans[s-1] += k;
if (e != n-1) {
ans[e] -= k;
}
//cout << ans[s-1] << " " << ans[e] << endl;
}
//cout << sizeof(ans)/sizeof(long) << endl;
/*for (auto i = 0; i < n; i++) {
cout << ans[i] << " ";
}
cout << endl;
for (long i = 1; i < n; i++) {
ans[i] += ans[i-1];
}
for (auto i = 0; i < n; i++) {
cout << ans[i] << " ";
}
cout << endl;
*/
sort(ans, ans+n+1);
/* for (auto i = 0; i < n; i++) {
cout << ans[i] << " ";
}*/
return ans[n-1];
}
问题是，它在少量输入时工作良好，但当输入约为1000000时，它会出现分段错误，当我全局声明一个固定大小为10000000的数组时，它在大量输入时工作。以下是当我用上面的代码运行程序时valgrind的输出：
❯ less ip.txt | valgrind ./am==7397== Memcheck, a memory error detector
==7397== Copyright (C) 2002-2017, and GNU GPL'd, by Julian Seward et al.
==7397== Using Valgrind-3.16.1 and LibVEX; rerun with -h for copyright info
==7397== Command: ./am
==7397==
==7397== Warning: client switching stacks? SP change: 0x1ffefffe30 --> 0x1ffa3b4a30
==7397== to suppress, use: --max-stackframe=80000000 or greater
==7397== Invalid write of size 8
==7397== at 0x1092E1: arrayManipulation(int, std::vector<std::vector<int, std::allocator<int> >, std::allocator<std::vector<int, std::allocator<int> > > >) (in /home/krillxox/Documents/CPP/am)==7397== Address 0x1ffa3b4a30 is on thread 1's stack
==7397==
==7397==
==7397== Process terminating with default action of signal 11 (SIGSEGV): dumping core
==7397== Access not within mapped region at address 0x1FFA3B4A30
==7397== at 0x1092E1: arrayManipulation(int, std::vector<std::vector<int, std::allocator<int> >, std::allocator<std::vector<int, std::allocator<int> > > >) (in /home/krillxox/Documents/CPP/am)
==7397== If you believe this happened as a result of a stack
==7397== overflow in your program's main thread (unlikely but
==7397== possible), you can try to increase the size of the
==7397== main thread stack using the --main-stacksize= flag.
==7397== The main thread stack size used in this run was 8388608.
==7397== Invalid write of size 8
==7397== at 0x482F120: _vgnU_freeres (vg_preloaded.c:57)
==7397== Address 0x1ffa3b4a28 is on thread 1's stack
==7397==
==7397==
==7397== Process terminating with default action of signal 11 (SIGSEGV)
==7397== Access not within mapped region at address 0x1FFA3B4A28
==7397== at 0x482F120: _vgnU_freeres (vg_preloaded.c:57)
==7397== If you believe this happened as a result of a stack
==7397== overflow in your program's main thread (unlikely but
==7397== possible), you can try to increase the size of the
==7397== main thread stack using the --main-stacksize= flag.
==7397== The main thread stack size used in this run was 8388608.
==7397==
==7397== HEAP SUMMARY:
==7397== in use at exit: 7,276,800 bytes in 200,004 blocks
==7397== total heap usage: 200,004 allocs, 0 frees, 7,276,800 bytes allocated
==7397==
==7397== LEAK SUMMARY:
==7397== definitely lost: 0 bytes in 0 blocks
==7397== indirectly lost: 0 bytes in 0 blocks
==7397== possibly lost: 0 bytes in 0 blocks
==7397== still reachable: 7,276,800 bytes in 200,004 blocks
==7397== suppressed: 0 bytes in 0 blocks
==7397== Rerun with --leak-check=full to see details of leaked memory
==7397==
==7397== For lists of detected and suppressed errors, rerun with: -s
==7397== ERROR SUMMARY: 2 errors from 2 contexts (suppressed: 0 from 0)
[1] 7396 done less ip.txt |
7397 segmentation fault (core dumped) valgrind ./am
当我在函数中声明数组时，分段错误的原因是什么？
您正在堆栈上放置一个巨大的数组。这是一个有限的资源。您最好使用std:：vector