News for hackers. Stuff that matters.

A comment on Branch Privilege Injection: Exploiting branch predictor race conditions. Parent comment

Re:Branch Privilege Injection: Exploiting branch predictor r... by titzer on Tuesday May 13, 2025 @04:52PM (#43977596)
That's a vast underestimate. Putting in lfence before every branch is on the order of 10X slowdown.
[ Parent | on HN ]
Re:Branch Privilege Injection: Exploiting branch predictor r... by grumbelbart2 on Wednesday May 14, 2025 @03:56AM (#43982052)
There is of course a slight chicken-egg-thing here: If there was no (dynamic) branch prediction, we (as in compilers) would emit different code that is faster for non-predicting CPUs (and presumably slower for predicting CPUs). That would mitigate a bit of that 10x.
[ Parent | on HN ]
Re:Branch Privilege Injection: Exploiting branch predictor r... by anyfoo on Wednesday May 14, 2025 @06:12PM (#43989729)
A bit. I think we've shown time and time again that letting the compiler do what the CPU is doing doesn't work out, most recently with Itanium.
[ Parent | on HN ]
Re:Branch Privilege Injection: Exploiting branch predictor r... by thesz on Wednesday May 14, 2025 @12:29PM (#43986446)
The issue is with indirect branches. Most branches are direct ones.
[ Parent | on HN ]