[SSTIC 2009] Day 2
it goes on for the second day. Already, some have trouble getting up for the first conf. the day. Well, the goal is to make a post shorter than for day 1!
Thursday, June 4, 2009
Mustang Stage 2 Specs
Morning
A morning dedicated to
features offered by the fuzzer; robustness;
performance. The idea here is to consider the model of fuzzing (mutation, pre-generated/scripted, block based, model based). So, here we want to stick to current fuzz and automate best. For example, we seek to address the specific protocol being audited. Finally, we Ari offers a checklist for choosing or design its fuzzer. The second
Improvements are already underway but the tool has already found vulnerabilities in programs like
readelf
You still need to know that our malicious program is not signed. popup could warn user installation. But the change (via rapid
) a security policy can address this problem;)
In summary, the law is not always adapted to the digital world and here is proof. This results in gaps that perpetrators attempt to use.
Finally, I echo the words of the late tois announcer: "
accountability
communication and control
We start with Nicolas RUFF with "Why security is failing." The title says it all with the meadows. was almost tempted to depressing because safety is our reason for living (especially if you have no social life side:) but it was learned during the rump session that there are "girl geek" PHEW!) Good The major problems date back 30 years ago as our speaker (Internet, C language, ...): Phew, I was not born, it's not my fault! I stop and depress I say I'm still going to make the safety! During his presentation, Nicolas shows us the big Windows flaws that have occurred and in which he participated and not really new at the end. We can only confirm that Windows environments are be beaten in internal testing. It was explained in the conf. lack of prevention in securing IF by entraprises and unfortunately, they end up paying the consequences ... It remains to convince these companies to pay "if" ... Otherwise, I suggest we take a comm 'of 1% every time they had warned the ongoing problem of box and no correction was made ... Before the rumps, we had presentations from projects to secure an OS for a user who must be able to do all the standard office (mail, word processing, e-commerce .. .): the project
SEC & SI
. The three projects have different approaches but interesting. We find the principles of proxy servers, sandbox, of hardening, etc.. No prognosis yet but we are forced to mark in our annals of SSTIC as "sh, that we did well shourrir":)) As for rumps, it was everything. As a listener, I especially noted the tools
IMA , reconet
,
evalSMSI
netif and I will test as soon as possible and why not to make a post.
PS failed, I failed to make a post shorter than yesterday and I will retry tomorrow!
- conf is given by Gabriel Campana. She made the transition with the previous tool "Fuzzgrind. The latter is based on two existing tools: Valgrind and STP. The idea is that for every new protocol is a long time to rebuild his fuzzer. This is where Fuzzgrind happens. This tool will work within the following 3 steps:
- symbolic execution of the program audited;
- taking into account the conditions (all cases);
Improvements are already underway but the tool has already found vulnerabilities in programs like
readelf
- and
- swfextract ... in less than an hour! Bravo! The last
- conf on this theme is presented by Lawrence fuzzing BUTTI. It is now tackling - in the context of an "architecture of fixed-mobile convergence" - in so-called security protocols such as IKEv2 and EAP. And where it starts to fear is found flaws there too ... In reality, these protocols appear to be safe but the idea to remember is that we should not overlook the point supposed to be secure during testing of any kind because the flaws will come network engineering and implementation of its protocols. In summary, "do not trust any (one physical phone, time to inject the program. Then we may act from a distance. To avoid this necessity, access to telephone the victim, several solutions are reviewing and the winner is to use the workstation with the smartphone that syncs. The pattern is so simple:
You still need to know that our malicious program is not signed. popup could warn user installation. But the change (via rapid
) a security policy can address this problem;)
- For the next conf, it really changes the subject to talk with Teddy FURON watermarking in a presentation titled "tracing traitor in multimedia. " As the title suggests, it is to find the person or persons guilty of making an illegal copy of a media owner (uh ... not good!). The challenge is that results can be strong enough to serve as legal evidence. We realize that this is not simple but that tattoo is successful based on weaknesses of human perception (visual and sound). Finally, we note that if the file has not been too Degard by hackers, we can find the attackers by estimating the probability of their guilt. The last
- conf this morning is presented by Marie BAREL, lawyer of record of SSTIC;) It presents a concrete case and we get to be audited regularly: the theft of information by an intern (which we simulate for a pentest internal in this case). The definition of theft provided by the French laws do not account for digital data:
- per flight there should be "subtraction" or theft of information does not imply that the company hit its provisions no longer the stolen data;
In summary, the law is not always adapted to the digital world and here is proof. This results in gaps that perpetrators attempt to use.
Finally, I echo the words of the late tois announcer: "
accountability
communication and control
- .
- Afternoon
We start with Nicolas RUFF with "Why security is failing." The title says it all with the meadows. was almost tempted to depressing because safety is our reason for living (especially if you have no social life side:) but it was learned during the rump session that there are "girl geek" PHEW!) Good The major problems date back 30 years ago as our speaker (Internet, C language, ...): Phew, I was not born, it's not my fault! I stop and depress I say I'm still going to make the safety! During his presentation, Nicolas shows us the big Windows flaws that have occurred and in which he participated and not really new at the end. We can only confirm that Windows environments are be beaten in internal testing. It was explained in the conf. lack of prevention in securing IF by entraprises and unfortunately, they end up paying the consequences ... It remains to convince these companies to pay "if" ... Otherwise, I suggest we take a comm 'of 1% every time they had warned the ongoing problem of box and no correction was made ... Before the rumps, we had presentations from projects to secure an OS for a user who must be able to do all the standard office (mail, word processing, e-commerce .. .): the project
SEC & SI
. The three projects have different approaches but interesting. We find the principles of proxy servers, sandbox, of hardening, etc.. No prognosis yet but we are forced to mark in our annals of SSTIC as "sh, that we did well shourrir":)) As for rumps, it was everything. As a listener, I especially noted the tools
IMA , reconet
,
evalSMSI
netif and I will test as soon as possible and why not to make a post.
PS failed, I failed to make a post shorter than yesterday and I will retry tomorrow!
Subscribe to:
Post Comments (Atom)
0 comments:
Post a Comment