On Tue, 31 Aug 1999 17:13:21 -0000, Crispin Cowan <[EMAIL PROTECTED]> said:
> Tobias Haustein wrote:
> > I was about to write a
> > binary translator, that reads an executeable, locates every function
> > prolog and epilog, adds the nescessary code to detect buffer
> > overflows, and writes a new version of the executeable.
>
> How do you make room for the extra code in prolog & epilog without re-linking
> the entire program?
The old-time IBM mainframe people would consider it a "piece of cake".
Basically, you replace one instruction with a branch to your code
stored someplace else, which would do your added code, then execute
the replaced instruction, then branch back to the next original opcode.
For example:
old code:
STM R2,R12,12(R13)
L R5,24(,R1)
ST R6,20(,R1)
new code:
STM R2,R12,12(R13)
B 2400(,R15) -----> SR R2,R5
ST R6,20(,R1) <---+ MVI 36(R3),C'+'
| L R5,24(,R1)
+-- B 8(,R15)
You get the idea. Of course, quite often you had to first perform
*one* 'zap' to free up space inside the 4K-per-base-register
restriction, and then a second 'zap' to put code into there. I've
personally hacked up HMASPZAP jobs that would end up replacing 5 or 6
different patches of code, 20 to 30 bytes at a stretch. You would
quite often find a string of 16 to 32 bytes of zeros at the end of a
module, specifically intended for patch space.
Valdis Kletnieks
Computer Systems Senior Engineer
Virginia Tech