I don't expect anything special. If there is no contention, there will
no syscalls and no slowdown. If there is contention, you will have many
syscalls and associated context switches, and meltdown mitigation
slowdown will apply.
It's not any different from any other syscall intensive workload.
On 01/16/2018 04:22 PM, [email protected] wrote:
Hi,
I saw a considerable slowdown with a microbenchmark on Linux systems
with KPTI enabled vs. it being disabled.
The benchmark performs no I/O but it is very futex-syscall-heavy. I
unfortunately did not have the time to
investigate further, but I also have the suspicion that the pagetable
switching overhead when performing these
syscalls is to blame. I would also be interested in more input on this.
Regards,
Marvin
On Monday, January 15, 2018 at 6:40:07 PM UTC+1, Francesco Nigro wrote:
Hi guys!
Any of you have already measured (or simply know) if OS mutexes
are somehow affected by Spectre/Meltdown?
Cheers,
Franz
--
You received this message because you are subscribed to the Google
Groups "mechanical-sympathy" group.
To unsubscribe from this group and stop receiving emails from it, send
an email to [email protected]
<mailto:[email protected]>.
For more options, visit https://groups.google.com/d/optout.
--
You received this message because you are subscribed to the Google Groups
"mechanical-sympathy" group.
To unsubscribe from this group and stop receiving emails from it, send an email
to [email protected].
For more options, visit https://groups.google.com/d/optout.
