[<prev] [next>] [day] [month] [year] [list]
Message-ID: <448e9a321003122228i48a6f9edg8c965bb6b893768@mail.gmail.com>
Date: Fri, 12 Mar 2010 22:28:01 -0800
From: Michal Zalewski <lcamtuf@...edump.cx>
To: bugtraq <bugtraq@...urityfocus.com>,
full-disclosure <full-disclosure@...ts.grok.org.uk>
Subject: ...because you can't get enough of clickjacking
[ I promise to post something more interesting shortly - but in the
meantime, I wanted to drop a quick note about something kinda amusing.
]
There was a considerable amount of buzz around clickjacking [1] in the
past year or so. It is commonly believed that this simple attack can
only be realistically employed to exploit one-click UI actions. Alas,
a related vector is generally overlooked: JS focus semantics - a
source of considerable amount of grief in the past [2] - can be abused
to execute multi-step attacks by altering focus between a hidden frame
and the edited document while the user is simply typing something in.
No need for pixel-accurate positioning of the target, too.
Consider this whimsical proof-of-concept exploit (works on Windows,
WebKit-based browsers only):
http://lcamtuf.coredump.cx/focus-webkit/
It's not very serious, but more cuter than clickjacking proper. WebKit
focus behavior on Windows makes this particular PoC easier there, but
I believe that no browser is designed to counter this general attack
pattern in any particular way. The usual opt-in mitigations
(X-Frame-Options, frame busting) should offer a reasonable degree of
protection already.
[1] http://code.google.com/p/browsersec/wiki/Part2#Arbitrary_page_mashups_%28UI_redressing%29
[2] http://lcamtuf.coredump.cx/focusbug/ and so forth
Powered by blists - more mailing lists