In a message dated , David Noon wrote to Robert King:
DN> Not under OS/2, it isn't. The OS/2 loader opens the .EXE with sharing
DN> denied to write applications; try running LINK386.EXE on a file you
DN> are currently running. You _can_ modify the extended attributes, but
DN> these are not loaded as part of the executable code.

LINK386.EXE isn't designed to perform writes on an open file. 

DN> I suppose "page descriptor table" means nothing also. Such concepts 
DN> do
DN> tend to limit the memory access of ring 3 (normal program)
DN> applications. Getting into ring 0 (operating system) privileges is
DN> impossible without being booted as a device driver or the kernel and
DN> having ring 0 privileges established at boot time. All ring 3 access
DN> to ring 0 functions is done by API calls that have ring transitions
DN> established when the system is booted, or inherited from such.

 You sure don't know much about programming device drivers..

DN> Actually, the hardware itself prevents this. There is a flag bit
DN> called I/O Privilege Level that can only be set at ring 0 (when the
DN> descriptor table entry is built) and VDM's do not have this 
DN> privilege.
DN> When IOPL is off, you cannot touch the I/O ports. All VDM's have 
DN> their
DN> legitimate I/O requests (INT 21H, various functions) escalated back 
DN> to
DN> OS/2; illegitimate attempts result in the program being killed by the
DN> operating system. A virus dies with its host.
DN> 

 Toggling IOPL isn't exactly dificult...

DN> Central Point Anti-Virus (for DOS and Windows) works fine in a VDM, 
DN> at
DN> least as far as the signatures it has currently installed are
DN> concerned. You have to keep updating the signature file (from CP,
DN> nowadays Symantec) if you want to keep up-to-date with the morons. 
DN> Any
DN> virus scanner is no better than its signature file. Incidentally, the
DN> plural of "virus" is "viruses".

Clearly you don't speak latin, virus, virusi....
 Robert

--- Simplex/2 BBS (v1.07.00Beta [OS/2])