[comp.virus] obscure procedure in Yankee Doodle

8326442@AWIWUW11.BITNET (Martin Zejma) (01/10/91)

hello virus-proofed community |

Last week i found the ( or a ) oh-so-old-but-never-found Yankee Doodle
Virus at a friend , savely jailed on a floppy disk.

I worked through the code quite heavy ,found nothing unbelieveable
clever but : after copying the virus-code to the top of memory ( i
worked hard to figure out the meaning of TOM in recent issues) , it
gets the size of the absolute system memory for DOS from a word in the
BIOS-segment ( 280h) multiplies this by various things to get the end
of memory ( A000:0000 )

AND THEN ::: checksums 61 words starting from A000:014E ( or 012E ,
i'm not sure without the source next to me ) , simply adding all these
61 words together , and if the result is something like 0b52 , it
writes a jump instruct ion into high memory , pointing to a small
procedure which changes Int 13h (disk interupt).

On my system ( a 286 Neat with 2 MB Ram running at 20 MHz 1 WS ) there
is nothing accessible after a000:0000 , everything just HIGH-VALUE
(FFh), not possible to change a byte .

I tried using Shadow RAM enabled at A000  , but that also failed .

SO THE ONE AND ONLY QUESTION :

Are there systems where this part of memory is accessible or would the
virus just overwrite a resident other virus when the value in the
BIOS-segment is below 280h due to a previous (already running)
infection ?

                                 Please many answers and soon , i'm puzzled

                                 Sincerly yours , Martin

+-----------------------------------------------------------------------+
| Martin Zejma                                8326442 @ AWIWUW11.BITNET |
|                                                                       |
| Wirtschaftsuniversitaet Wien  ---   Univ. of Economics Vienna/Austria |
+-----------------------------------------------------------------------+

cjimenez@anyware.es (Carlos Jimenez) (01/10/91)

>Send by Martin Zejma <8326442@AWIWUW11.BITNET>:
>
>hello virus-proofed community |
>Last week i found the ( or a ) oh-so-old-but-never-found Yankee ...
>...
>SO THE ONE AND ONLY QUESTION :
>Are there systems where this part of memory is accessible or would the
>virus just overwrite a resident other virus when the value in the
>BIOS-segment is below 280h due to a previous (already running)
>infection ?

The segment A000h of computer is used by graphics cards like EGA, MCGA
& VGA to implement graphics modes 0Dh to 13h and new modes of higher
resolution.  This segment of memory isn't used in text modes. Thus,
when you use text modes (the normal situation if you don't work in
Windows) the virus can use the segment A000h. Probably you have a CGA
or Hercules Graphic Card and then you can't use this segment (There
isn't RAM for the virus in this segment).  I hope this comment can
help you.

Carlos Jimenez    R+D Manager                    Phone: +34 1 556 92 15
                  ANYWARE Information Security          +34 1 556 92 16
                  General Peron, 32                Fax: +34 1 556 91 58
                  28020 Madrid (SPAIN)           EUnet: cjimenez@anyware.es

CHESS@YKTVMV.BITNET (David.M.Chess) (01/11/91)

Martin Zejma <8326442@AWIWUW11.BITNET>:
> Are there systems where this part of memory is accessible or would the
> virus just overwrite a resident other virus when the value in the
> BIOS-segment is below 280h due to a previous (already running)
> infection ?

I haven't verified it myself, but a reasonably authoritative rumor
says that the checksum the virus does will detect a Bouncing Ball
(a.k.a. "Ping Pong") infection active in memory, and patch it so that
it (eventually?) stops infecting.

There are a few other cases of viruses that look for other viruses;
the Den Zuk / Ohio family look for and remove the Brain (before
installing themselves), the TPxxVIR look for and remove earlier
members of the family, and so on.

DC