Dustin Puryear wrote: > Scott Harney wrote: > >> Dustin Puryear wrote: >> >>> >>> Well, one issue with the x86 is that you can execute code on the >>> stack and [possibly] in the data segment (it's been a while since I >>> did assembly, so I don't really remember this one). But then again, >>> I think this occurs on a lot of other CPU architectures as well. >>> >>> If you couldn't execute code on the stack then a lot of symptoms >>> would go away, but the root cause of bad code would remain. >> >> Compilers can check for this at compile time too. > > Yeah, but exploits that use code on the stack are runtime issues.
More grist for the Microsoft vs Linux mill I think. You'd better hold this meeting when I'm in town. Edmund Cramp -- Never send an e-mail when a bathroom wall will do.
