Liste des Groupes | Revenir à c arch |
Thomas Koenig <tkoenig@netcologne.de> wrote:The easiest way to state this succinctly is::Seems like Intel branch predictors have been pretty completely>
reverse-engineered. The following paper promises to for very
interesting reading:
>
https://www.usenix.org/conference/usenixsecurity24/presentation/li-luyi
>
I wonder what you think of this...
There are more papers on this topic. There were several papers
on variations of Spectre. I think that there is simple condition
which guarantees that nothing Spectre-related affects given
processor: the sequence of microarchitecutral operations (incuding
speculative operations) should depend only on architecturaly
executed instructions.
So, processor may do widely speculativeThe processor can maintain look-ahead predictor state and use it
things, but only base speculation on architecturaly executed
instructions.
Some people try to just close single hole atSo far its been almost t10 years of closing one hole after another
a time, IMO it is hopeless, there are too many possible
variations.
And weaker conditions, like "cancelling" effectsThese also lead to replay-storms.
of speculative instructions are likely to fail.
Les messages affichés proviennent d'usenet.