其實Linus Torvalds有表態
https://lkml.org/lkml/2018/11/19/37
STIBP by default.. Revert?
This was marked for stable, and honestly, nowhere in the discussion did I see an
y mention of just *how* bad the performance impact of this was.
When performance goes down by 50% on some loads, people need to start asking the
mselves whether it was worth it.
It's apparently better to just disable SMT entirely, which is what security-cons
cious people do anyway.
So why do that STIBP slow-down by default when the people who *really* care alre
ady disabled SMT?
I think we should use the same logic as for L1TF: we default to something that d
oesn't kill performance.
Warn once about it, and let the crazy people say "I'd rather take a 50% performa
nce hit than worry about a theoretical issue".
Linus