With that said, it is amazingly clever and certainly some ppl want it as a short-term mitigation. https://support.google.com/faqs/answer/7625886 …
-
-
Show this threadThanks. Twitter will use this to make your timeline better. UndoUndo
-
-
-
I wonder if this will simply start an arms race: Intel et al. make a branch predictor that sees through this, and then the whole things starts all over again.
-
that /is/ my experience with how well hardware teams communicate with compiler teams :/
-
IMO, they should be the same team. Chip mfr’s that figure that out will do much better in the long run.
-
Not sure if they should be the same team, but they should be close and talk to each other often. At the very least this will lead to chip designs with features that can actually be exploited sanely in compilers. But that's going at bit off-topic here.
- 1 more reply
New conversation -
-
-
"If ability to run existing binaries safely IS one of your requirements, throw out x86 already"
-
throw out the C language too when you're at it (throw the entire baby with the water and the bathtub too, they are all rotten)
End of conversation
New conversation -
-
-
It isn’t just x86 that’s vulnerable here. The ideas behind the attack work on other CPU architectures with speculative execution.
-
the idea is that CPU leaves behind side effects from failed speculative execution unbeknownst to the program
-
More specifically, the cache and BTB don’t roll back if speculation fails. Normal code doesn’t care about that — and it wasn’t obvious that it was exploitable.
End of conversation
New conversation -
Loading seems to be taking a while.
Twitter may be over capacity or experiencing a momentary hiccup. Try again or visit Twitter Status for more information.