full-disclosure - [anti-XSS]about CERT/CC:malicious_code

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]

Message-ID: <411836B3.6050504@ghettohackers.net>
From: dd at ghettohackers.net (dd)
Subject: [anti-XSS]about CERT/CC:malicious_code_mitigation

Woops, O'Brian is pissed again ;)

Real solution is to have "per input" input validation which will always 
let some "potentially bad" things through, but help mitigate exposures 
and then do things "right".  Right of course means always binding data 
to sql statements and properly performing context sensitive output encoding.

dd

Valdis.Kletnieks@...edu wrote:
> On Sat, 07 Aug 2004 06:25:00 -0000, bitlance winter said:
> 
> 
>>#! The first function takes the negative approach.
>>#! Use a list of bad characters to filter the data
>>sub FilterNeg {
>>    local( $fd ) = @_;
>>    $fd =~ s/[\<\>\"\'\%\;\)\(\&\+]//g;
>>    return( $fd ) ;
>>}
> 
> 
> *BZZT!!* Wrong.  Don't do this in production code, because...
> 
> 
>>I have understood that bad characters are
>>< > " ' % ; ) ( & +
> 
> 
> If it turns out that * (asterisk) is a "bad character", you're screwed.
> 
> If it turns out that *any other* character is "bad", you're screwed.
> 
> The *proper* way to do the filtering is to *remove* *all* characters
> not known to be good.  Something like:
> 
> $fd =~ s/[^-_ a-zA-Z0-9]//g;
> 
> Only pass alphabetic, numeric, space, hyphen, and underscore.  Add other
> characters *only* if you can show they are *not* a problem.