[prev in list] [next in list] [prev in thread] [next in thread]
List: kde-devel
Subject: Re: KLineEdit Security
From: Thomas =?iso-8859-1?q?L=FCbking?= <thomas.luebking () web ! de>
Date: 2009-05-21 16:46:02
Message-ID: 200905211846.02597.thomas.luebking () web ! de
[Download RAW message or body]
[Attachment #2 (multipart/alternative)]
You don't need XTest - it's just more conveniant.
In general you can XSendEvent to any WId but the XKeyEvent has a flag for
send_event that iff set by XSendEvent (i DON'T know this) would kill your
efforts in this direction :-(
The kernel device should be read protected except for root (at least it is
here, but you can of course add some udev rule to free it like often necessary
for scanners)
If anything has however already root access you're doomed anyway.
The (only really hard and) proper solution to this is btw. an encrypted
keyboard -> app chain, but there're few keyboards that can encrypt strokes
(and i don't even know whether X11 supports such at all, so you'd have to
bring your own keyboard driver)
TPM was supposed to bring a protocol for this, but then: TPM had some other
issues... :-(
Am Thursday 21 May 2009 schrieb Andreas Pakulat:
> Well, first of all you'd have to make sure that your somehow creating real
> X11 events - at least - because a keylogger would sit at the X11 level
> looking for Keyboard Events. This could be possible with the XTest library,
> but I'm not sure wether those events really don't provide a way to find out
> they're generated and not "original" events from the keyboard. Not sure if
> that would really be enough, as it might be possible to directly read
> events from the interfaces that the kernel provides and you're not going to
> be able to fake those I think (unless the kernel provides an interface to
> do that)..
>
> Andreas
>
> >> Visit http://mail.kde.org/mailman/listinfo/kde-devel#unsub to
> >> unsubscribe <<
[Attachment #5 (text/html)]
<!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 4.0//EN" \
"http://www.w3.org/TR/REC-html40/strict.dtd"><html><head><meta name="qrichtext" \
content="1" /><style type="text/css">p, li { white-space: pre-wrap; \
}</style></head><body style=" font-family:'Segoe'; font-size:10pt; font-weight:400; \
font-style:normal;">You don't need XTest - it's just more conveniant.<br> In general \
you can XSendEvent to any WId but the XKeyEvent has a flag for send_event that iff \
set by XSendEvent (i DON'T know this) would kill your efforts in this direction \
:-(<br> <p style="-qt-paragraph-type:empty; margin-top:0px; margin-bottom:0px; \
margin-left:0px; margin-right:0px; -qt-block-indent:0; text-indent:0px; \
-qt-user-state:0;"><br></p>The kernel device should be read protected except for root \
(at least it is here, but you can of course add some udev rule to free it like often \
necessary for scanners)<br> <p style="-qt-paragraph-type:empty; margin-top:0px; \
margin-bottom:0px; margin-left:0px; margin-right:0px; -qt-block-indent:0; \
text-indent:0px; -qt-user-state:0;"><br></p>If anything has however already root \
access you're doomed anyway.<br> <p style="-qt-paragraph-type:empty; margin-top:0px; \
margin-bottom:0px; margin-left:0px; margin-right:0px; -qt-block-indent:0; \
text-indent:0px; -qt-user-state:0;"><br></p>The (only really hard and) proper \
solution to this is btw. an encrypted keyboard -> app chain, but there're few \
keyboards that can encrypt strokes (and i don't even know whether X11 supports such \
at all, so you'd have to bring your own keyboard driver)<br> <p \
style="-qt-paragraph-type:empty; margin-top:0px; margin-bottom:0px; margin-left:0px; \
margin-right:0px; -qt-block-indent:0; text-indent:0px; -qt-user-state:0;"><br></p>TPM \
was supposed to bring a protocol for this, but then: TPM had some other issues... \
:-(<br> <p style="-qt-paragraph-type:empty; margin-top:0px; margin-bottom:0px; \
margin-left:0px; margin-right:0px; -qt-block-indent:0; text-indent:0px; \
-qt-user-state:0;"><br></p>Am Thursday 21 May 2009 schrieb Andreas Pakulat:<br> > \
Well, first of all you'd have to make sure that your somehow creating real<br> > \
X11 events - at least - because a keylogger would sit at the X11 level<br> > \
looking for Keyboard Events. This could be possible with the XTest library,<br> > \
but I'm not sure wether those events really don't provide a way to find out<br> > \
they're generated and not "original" events from the keyboard. Not sure if<br> > \
that would really be enough, as it might be possible to directly read<br> > events \
from the interfaces that the kernel provides and you're not going to<br> > be able \
to fake those I think (unless the kernel provides an interface to<br> > do \
that)..<br> ><br>
> Andreas<br>
><br>
> >> Visit http://mail.kde.org/mailman/listinfo/kde-devel#unsub to<br>
> >> unsubscribe <<<br>
<p style="-qt-paragraph-type:empty; margin-top:0px; margin-bottom:0px; \
margin-left:0px; margin-right:0px; -qt-block-indent:0; text-indent:0px; \
-qt-user-state:0;"><br></p><p style="-qt-paragraph-type:empty; margin-top:0px; \
margin-bottom:0px; margin-left:0px; margin-right:0px; -qt-block-indent:0; \
text-indent:0px; -qt-user-state:0;"><br></p></body></html>
>> Visit http://mail.kde.org/mailman/listinfo/kde-devel#unsub to unsubscribe <<
[prev in list] [next in list] [prev in thread] [next in thread]
Configure |
About |
News |
Add a list |
Sponsored by KoreLogic