Extent of a memory/hard drive wipe?

[YT2095]

SmokingSkillz said in post # :

it would probably work on a spectrum. but i think you'd get a load of GPF's on windows. (each process has it's own virtual memory space and can't address outside of it). Having said that WinHex has a RAM editor that manages to do it so god knows.

 

it does indeed work with Spectrums, a CBM 64`s and BBC micros etc... (of course you must adapt the syntax a little for each computer).

but the algorithm for the program is valid right across the board really and is easily converted (due to its simplicity) into any lang including ASM or MC.

I understand your point about allocated mem areas though, most modern puters and clients do tend to "Stick you in a box".

it would have be run at a low level and as it`s own program I should imagine, a bit like any program on a PC would, and from there it would be free to interogate any section of ram, although you`de have to add more code to compensate for memory violation errors or "mem in use" errors.

certainly not impossible IF you know what your doing, the X86 series I`m unfamiliar with and so have NO idea

[SmokingSkillz]

I think there is a concept of memory rings, with the kernel running in say ring 0. It then allocates ring 1 memory for use by client processes. So your program would have to be run in ring 0, maybe if you could inject or buffer overun the code into the kernels call stack you could get a system wide mem dump!

[YT2095]

that sounds good, but you`de have to factor in IRQs too, looking at the simple desktop clock mem area, would need an interupt, and then there would be the prob of the all the cards they use, gfx and sound cards, they all have onboard mem, it sure wouldn`t be a task I`de like to undertake