> And even if you get that right, it may not work well on hardware. We > saw cases with linux migration, while migrating from one x86 to > another, where valid FP values would cause the target to get an FP > trap. Made no sense, but it's what happened, because the two x86's > were different *implementations* of the same architecture. So, > migration works well for all the really easy cases -- CPU you migrate > to was fabbed by the same vendor in the same place with the same mask > and microcode as the one you migrate from -- and can fail on anything > tricky. That's why I don't like it either :-)
the existance of broken hardware doesn't say anything about the idea. we don't give up on page tables because some mmus don't work. - erik
