BONUS: cevags

Paul Rusty Russell
Thu, 30 Sep 1999 17:45:43 +0930

[BTW: Just a bit of fun which you can skip 8-)]

I split the two points between James Morris (who got the `why cevags?'
question right) and Albert Hui (who got the purpose right).

Albert said:

> They are no-ops, just return 0.  The "unhelp" operation modifies the
> binary to make all forms of printf() do nothing.  "cevags" is called
> that because it is exactly 6 character long, so you can do a global
> replace with sed without messing up the binary file.  Since the
> substitution will match also vfprintf(), fprintf() and friends,
> corresponding stub functions of *evags() are prepared. :)

James said:

> I'm taking a wild guess that 'cevags' is rot13'd 'printf'.
> It seems to be part of a general mechanism for wreaking havoc, and
> drawing attention to no error checking being done on library calls
> like dlopen() in iptables.

[`general mechanism for wreaking havoc!']

I'll come up with a new easter egg sometime before 1.0 (ie. not for a
while).  Another +3 points to the person who finds and explains
it. 8-)

At least it got people reading the code...
Hacking time.