It's not a 32-bit limitation. There are buffer overruns in 64-bit code too, and they are just as likely to occur using 64-bit processors.Originally posted by: juspatrick
I read in one of my many weekly geek-mails that over half of XP's venerabilities regarding the
running of malicious code are due to buffer overruns, an inherent limitation of hardware the
32 bit code runs on
It is important to emphasize that these NX features do NOT eliminate buffer overrun attacks! It is still quite possible to control and/or influence application behavior by overflowing buffers, and it is even possible given the right circumstances to execute arbitrary code despite the NX protection. It just makes such attacks much more difficult to achieve.
Shouldn't they be developing a 64bit version of Longhorn, as well?
Yes, which confused me at first because it actually requires an entire new level for page tables, and I couldn't see Microsoft making such a drastic change to the kernel for 32-bit X86. (They have to do it already for 64-bit addressing modes, so it's a no-brainer there.)Originally posted by: biostud666
aren't the NX flag activated in SP2?
Encoding and encryption is where 64bit may well shine. I have seen some 64bit encoding betas that show massive performance boosts in 64bit.
I think another big boost will come when the extra 8 GPR's are utilized in 64bit Longmode that are currently sitting useless (they are not utilized in 32bit even in the 64bit OS) (along with all the other 64bit goodness [at least on the MS side of OS land])
Originally posted by: gflores
So, if Windows XP64 isn't coming out until late 2005 (IIRC), then Longhorn will be out less than half a year later. Shouldn't they be developing a 64bit version of Longhorn, as well?