Groups | Search | Server Info | Keyboard shortcuts | Login | Register [http] [https] [nntp] [nntps]

Groups > comp.programming > #6004 > unrolled thread

finding backdoors and security holes

Back to article view | Back to comp.programming

Contents

  finding backdoors and security holes G G <gdotone@gmail.com> - 2015-12-02 17:59 -0800
    Re: finding backdoors and security holes Robert Wessel <robertwessel2@yahoo.com> - 2015-12-03 01:58 -0600
    Re: finding backdoors and security holes Richard Heathfield <rjh@cpax.org.uk> - 2015-12-03 11:04 +0000
      Re: finding backdoors and security holes "Osmium" <r124c4u102@comcast.net> - 2015-12-09 09:22 -0600
        Re: finding backdoors and security holes Richard Heathfield <rjh@cpax.org.uk> - 2015-12-09 15:58 +0000
        Re: finding backdoors and security holes "J. Clarke" <j.clarke.873638@gmail.com> - 2015-12-13 22:59 -0500
    Re: finding backdoors and security holes Öö Tiib <ootiib@hot.ee> - 2015-12-03 05:42 -0800
      Re: finding backdoors and security holes "Osmium" <r124c4u102@comcast.net> - 2015-12-03 08:12 -0600
        Re: finding backdoors and security holes G G <gdotone@gmail.com> - 2015-12-03 06:55 -0800
          Re: finding backdoors and security holes "Osmium" <r124c4u102@comcast.net> - 2015-12-03 09:48 -0600
            Re: finding backdoors and security holes Öö Tiib <ootiib@hot.ee> - 2015-12-03 10:01 -0800
    Re: finding backdoors and security holes Victor Porton <porton@narod.ru> - 2015-12-03 20:49 +0200
      Re: finding backdoors and security holes G G <gdotone@gmail.com> - 2015-12-04 15:56 -0800

#6004 — finding backdoors and security holes

How are these guys (generic) find these backdoors and security holes?

Some of the code they are looking at has to be 100s of thousands
line of code.

In the case of Arris's backdoor, what, how?

Any really good books on the subject matter?

Thanks,

g.

[toc] | [next] | [standalone]

#6005

On Wed, 2 Dec 2015 17:59:33 -0800 (PST), G G <gdotone@gmail.com>
wrote:

>How are these guys (generic) find these backdoors and security holes?
>
>Some of the code they are looking at has to be 100s of thousands
>line of code.
>
>In the case of Arris's backdoor, what, how?
>
>Any really good books on the subject matter?


Half the time you can find security holes just by running the source
through Lint.

[toc] | [prev] | [next] | [standalone]

#6006

On 03/12/15 01:59, G G wrote:
> How are these guys (generic) find these backdoors and security holes?

Do they have the source? If so, imagine trying to find a shoe shop on 
Oxford Street (not a particular shop - any shoe shop will do), looking 
for a hill in Wales, tracking down a cow in Gloucestershire, or 
identifying a crank in sci.crypt. These are tasks of comparable length 
and difficulty.

-- 
Richard Heathfield
Email: rjh at cpax dot org dot uk
"Usenet is a strange place" - dmr 29 July 1999
Sig line 4 vacant - apply within

[toc] | [prev] | [next] | [standalone]

#6034

"Richard Heathfield" wrote:

> On 03/12/15 01:59, G G wrote:
>> How are these guys (generic) find these backdoors and security holes?
>
> Do they have the source? If so, imagine trying to find a shoe shop on 
> Oxford Street (not a particular shop - any shoe shop will do), looking for 
> a hill in Wales, tracking down a cow in Gloucestershire, or identifying a 
> crank in sci.crypt. These are tasks of comparable length and difficulty.

I will make a wild guess that they sell shoes on Oxford Street and Wales is 
hilly, therefore finding backdoors is a trivial exercise..

I would describe a program that does something other than what was expected 
by the user, as having a backdoor. Using that base, one link on Google says 
there are 482,000 VW diesels in the US that have backdoors in their 
software.

Here's one link describing how that particular backdoor after several years, 
was finally found.  Sounds like a good base plot line for a movie on the 
Airdbus 380 or whatver they call that big thing.

http://www.bbc.com/news/business-34519184

[toc] | [prev] | [next] | [standalone]

#6035

On 09/12/15 15:22, Osmium wrote:
> "Richard Heathfield" wrote:
>
>> On 03/12/15 01:59, G G wrote:
>>> How are these guys (generic) find these backdoors and security holes?
>>
>> Do they have the source? If so, imagine trying to find a shoe shop on
>> Oxford Street (not a particular shop - any shoe shop will do), looking
>> for a hill in Wales, tracking down a cow in Gloucestershire, or
>> identifying a crank in sci.crypt. These are tasks of comparable length
>> and difficulty.
>
> I will make a wild guess that they sell shoes on Oxford Street and Wales
> is hilly, therefore finding backdoors is a trivial exercise..

Finding security holes is a trivial exercise. Finding backdoors can be 
rather harder. Cf "Reflections on trusting trust" (kt).

> I would describe a program that does something other than what was
> expected by the user, as having a backdoor.

I would call that a bug.

> Using that base, one link on
> Google says there are 482,000 VW diesels in the US that have backdoors
> in their software.

Or perhaps Google was being Google, and actually there are 482,000 
hatchback VW diesels in the USA, and the word "software" also appears on 
the page. (Although on this occasion it is of course more likely that 
you are right. I think it's reasonable to call the VW hack a "back door".)

> Here's one link describing how that particular backdoor after several
> years, was finally found.  Sounds like a good base plot line for a movie
> on the Airdbus 380 or whatver they call that big thing.
>
> http://www.bbc.com/news/business-34519184

The only reason it took several years to find it was that it took 
several years for someone to look. All they had to do was measure the 
on-the-road emissions. (It is, however, quite true that back doors into, 
say, cryptographic algorithms can be rather more difficult to find.)

But security holes? You can't chuck a brick without hitting half a dozen.

-- 
Richard Heathfield
Email: rjh at cpax dot org dot uk
"Usenet is a strange place" - dmr 29 July 1999
Sig line 4 vacant - apply within

[toc] | [prev] | [next] | [standalone]

#6043

In article <dcqv96FpihiU1@mid.individual.net>, r124c4u102@comcast.net 
says...
> 
> "Richard Heathfield" wrote:
> 
> > On 03/12/15 01:59, G G wrote:
> >> How are these guys (generic) find these backdoors and security holes?
> >
> > Do they have the source? If so, imagine trying to find a shoe shop on 
> > Oxford Street (not a particular shop - any shoe shop will do), looking for 
> > a hill in Wales, tracking down a cow in Gloucestershire, or identifying a 
> > crank in sci.crypt. These are tasks of comparable length and difficulty.
> 
> I will make a wild guess that they sell shoes on Oxford Street and Wales is 
> hilly, therefore finding backdoors is a trivial exercise..
> 
> I would describe a program that does something other than what was expected 
> by the user, as having a backdoor.

You are redfining "backdoor" to mean "bug".

> Using that base, one link on Google says 
> there are 482,000 VW diesels in the US that have backdoors in their 
> software.
> 
> Here's one link describing how that particular backdoor after several years, 
> was finally found.  Sounds like a good base plot line for a movie on the 
> Airdbus 380 or whatver they call that big thing.
> 
> http://www.bbc.com/news/business-34519184

[toc] | [prev] | [next] | [standalone]

#6007

On Thursday, 3 December 2015 03:59:38 UTC+2, G G  wrote:
> How are these guys (generic) find these backdoors and security holes?

They search for those.
 
> 
> Some of the code they are looking at has to be 100s of thousands
> line of code.
>
> In the case of Arris's backdoor, what, how?

It was likely deliberately made by author of modem software
for testing and debugging purposes and then forgotten to remove
from production software.

> 
> Any really good books on the subject matter?

There are plenty of books about how to prevent attacks and to reduce
risks of security vulnerabilities. These books teach how to look for
those.

[toc] | [prev] | [next] | [standalone]

#6008

"Öö Tiib" wrote:

> On Thursday, 3 December 2015 03:59:38 UTC+2, G G  wrote:
>> How are these guys (generic) find these backdoors and security holes?
>
> They search for those.
>
>>
>> Some of the code they are looking at has to be 100s of thousands
>> line of code.
>>
>> In the case of Arris's backdoor, what, how?
>
> It was likely deliberately made by author of modem software
> for testing and debugging purposes and then forgotten to remove
> from production software.
>
>>
>> Any really good books on the subject matter?
>
> There are plenty of books about how to prevent attacks and to reduce
> risks of security vulnerabilities. These books teach how to look for
> those.

To the OP:
Perhaps if you would limit your question to backdoors you might start 
getting useful answers. You gave a sneak path by mentioning security 
issues - it's kind of like electricity, it takes the easiest path.

I only have guesses, my guess is that an awful lot of stars have to align to 
find a backdoor. 

[toc] | [prev] | [next] | [standalone]

#6009

On Thursday, December 3, 2015 at 9:13:05 AM UTC-5, Osmium wrote:
 
> To the OP:
> Perhaps if you would limit your question to backdoors you might start 
> getting useful answers. You gave a sneak path by mentioning security 
> issues - it's kind of like electricity, it takes the easiest path.
> 
> I only have guesses, my guess is that an awful lot of stars have to align to 
> find a backdoor.

What do you mean? You gave a sneak path by mentioning security 
> issues - it's kind of like electricity, it takes the easiest path."

If a backdoor in a device or software, isn't that a security issue too.

I would love to hear from anyone with insight or just reasonable guess
work (thinking). 

[toc] | [prev] | [next] | [standalone]

#6011

"G G" wrote:

> On Thursday, December 3, 2015 at 9:13:05 AM UTC-5, Osmium wrote:
>
>> To the OP:
>> Perhaps if you would limit your question to backdoors you might start
>> getting useful answers. You gave a sneak path by mentioning security
>> issues - it's kind of like electricity, it takes the easiest path.
>>
>> I only have guesses, my guess is that an awful lot of stars have to align 
>> to
>> find a backdoor.
>
> What do you mean? You gave a sneak path by mentioning security
>> issues - it's kind of like electricity, it takes the easiest path."
>
> If a backdoor in a device or software, isn't that a security issue too.
>
> I would love to hear from anyone with insight or just reasonable guess
> work (thinking).

I mean you asked two questions. One about security issues and one about 
backdoors.  If you ask two questions people will naturally answer the easier 
one. A huge number of security issues are well known and I am quite sure I 
could find a whole book on the subject by spending a few minutes with 
Google.  Back doors are a *specific* security issue and the answers are 
going to become much more sparse. For example, I have no reason in the world 
to believe that my OS (Windows XP) does not contain one or more backdoors. 
It is such a behemoth that it takes several CDs to contain a copy of it.

[toc] | [prev] | [next] | [standalone]

#6013

On Thursday, 3 December 2015 17:48:40 UTC+2, Osmium  wrote:
> "G G" wrote:
> 
> > On Thursday, December 3, 2015 at 9:13:05 AM UTC-5, Osmium wrote:
> >
> >> To the OP:
> >> Perhaps if you would limit your question to backdoors you might start
> >> getting useful answers. You gave a sneak path by mentioning security
> >> issues - it's kind of like electricity, it takes the easiest path.
> >>
> >> I only have guesses, my guess is that an awful lot of stars have to align 
> >> to
> >> find a backdoor.
> >
> > What do you mean? You gave a sneak path by mentioning security
> >> issues - it's kind of like electricity, it takes the easiest path."
> >
> > If a backdoor in a device or software, isn't that a security issue too.
> >
> > I would love to hear from anyone with insight or just reasonable guess
> > work (thinking).
> 
> I mean you asked two questions. One about security issues and one about 
> backdoors.  If you ask two questions people will naturally answer the easier 
> one. A huge number of security issues are well known and I am quite sure I 
> could find a whole book on the subject by spending a few minutes with 
> Google.

More like whole bookshelf of decent books and a library of crappy ones.

> Back doors are a *specific* security issue and the answers are 
> going to become much more sparse.

It is because writing testing shortcuts and back-doors into product code
is considered anti-pattern by most professional programmers. Hiding
maliciously made back-doors is also possibility but may bring charges
to authors. Works of that contest may give inspiration how to: 
https://en.wikipedia.org/wiki/Underhanded_C_Contest

> For example, I have no reason in the world 
> to believe that my OS (Windows XP) does not contain one or more backdoors. 
> It is such a behemoth that it takes several CDs to contain a copy of it.

Actually so large companies like Microsoft have their quality standards 
and enough budget to check, review, test and validate their code-bases
over and over. Deliberate backdoor is possible perhaps only when very
cleverly hidden.

It is far more likely to find some non-deliberate vulnerability in 
Windows code. Also it may be more fruitful not to scan the operating
system itself but various third party software that you may run on it.
Some software upgrader, multiplayer game, document viewer or editor, 
web browser plugin ... things like that. ;)

[toc] | [prev] | [next] | [standalone]

#6014

G G wrote:

> How are these guys (generic) find these backdoors and security holes?
> 
> Some of the code they are looking at has to be 100s of thousands
> line of code.

One thing which helps, is using a language with increased reliability such 
as Ada (including latest 2012 version of Ada).

-- 
Victor Porton - http://portonvictor.org

[toc] | [prev] | [next] | [standalone]

#6019

On Thursday, December 3, 2015 at 1:49:33 PM UTC-5, Victor Porton wrote:
> One thing which helps, is using a language with increased reliability such 
> as Ada (including latest 2012 version of Ada).
> 

I wish Ada was more popular. 

[toc] | [prev] | [standalone]

Back to top | Article view | comp.programming

csiph-web