Groups | Search | Server Info | Login | Register


Groups > comp.security.misc > #1549

Re: Finding backdoors

Organization Timetravellers Anonymous
References <vd3uf2$7ng3$1@dont-email.me> <wwvbk08uoui.fsf@LkoBDZeT.terraraq.uk> <vdfs6p$2jf3c$1@dont-email.me> <A3AQO.382679$FzW1.372916@fx14.iad> <veurfk$3gv4u$1@dont-email.me>
From rlhamil@smart.net (Richard L. Hamilton)
Subject Re: Finding backdoors
Newsgroups comp.security.misc
Message-ID <oHEQO.291923$v8v2.282917@fx18.iad> (permalink)
Date 2024-10-19 02:20 +0000

Show all headers | View raw


In article <veurfk$3gv4u$1@dont-email.me>,
	<bp@www.zefox.net> writes:
> Richard L. Hamilton <rlhamil@smart.net> wrote:
>> 
>> It's worse than that even. See
>> 
>> Reflections on Trusting trust
>> 
>> https://dl.acm.org/doi/pdf/10.1145/358198.358210
>> 
>> by Ken Thompson, the co-creator of Unix along with DMR.
>> 
>> It describes adding some code to the C compiler that will insert into
>> login.c a magic password for root power, and into a re-compile of the
>> compiler, the code needed to propagate itself and that login.c backdoor.
[...]
> 
> A fascinating article, entirely new to me. 
> 
> Have any examples of such a trojan been found, or even suspected?
> After forty years, one might expect to see it "in the wild" or at
> least see plausible consequences if it's viable in practice.
> 
> Thank you!

I gather he ran it briefly on some system(s) he administered, so it
definitely worked. It wouldn't propagate except if someone copied the
gimmicked compiler binaries.

The source code for the changes is out there. Of course, compilers
have changed a lot since then, and the big open source compilers (gcc,
clang) have a lot of eyes on their source, and the binaries to start
rebuilding the toolchain with are probably built by "trusted" people
and accompanied with gpg signatures. Which is not to say that it
couldn't be done. Think about how gradually the xz based backdoor was
introduced; although that did have a lot less eyes on it.

While I've had interesting enough connections to have had a reasonably
high level of general awareness expected of me, that did not extend to
the point of either using such things or front line defense against
them.  So I couldn't say that something like it hasn't appeared. If in
doubt, I would assume that all known possible threats and plenty of
unknown ones existed.

Consider if someone makes changes to a web based source control system
(git, hg, etc) or publishing software (wordpress, etc); that could do
something interesting to a lot of people's code, or to a lot of
generated web pages. Doubtless a few such things have happened,
although I don't recall specifics.

Back to comp.security.misc | Previous | NextPrevious in thread | Find similar


Thread

Finding backdoors <bp@www.zefox.net> - 2024-09-26 15:26 +0000
  Re: Finding backdoors Marco Moock <mm+usenet-es@dorfdsl.de> - 2024-09-26 20:15 +0200
    Re: Finding backdoors William Unruh <unruh@invalid.ca> - 2024-09-26 19:57 +0000
      Re: Finding backdoors Marco Moock <mm+usenet-es@dorfdsl.de> - 2024-09-27 17:31 +0200
        Re: Finding backdoors William Unruh <unruh@invalid.ca> - 2024-09-27 16:26 +0000
          Re: Finding backdoors Marco Moock <mm+usenet-es@dorfdsl.de> - 2024-09-27 20:11 +0200
  Re: Finding backdoors Richard Kettlewell <invalid@invalid.invalid> - 2024-09-28 10:09 +0100
    Re: Finding backdoors <bp@www.zefox.net> - 2024-10-01 04:01 +0000
      Re: Finding backdoors Richard Kettlewell <invalid@invalid.invalid> - 2024-10-01 16:44 +0100
      Re: Finding backdoors rlhamil@smart.net (Richard L. Hamilton) - 2024-10-18 21:04 +0000
        Re: Finding backdoors <bp@www.zefox.net> - 2024-10-18 23:37 +0000
          Re: Finding backdoors rlhamil@smart.net (Richard L. Hamilton) - 2024-10-19 02:20 +0000

csiph-web