News for hackers. Stuff that matters.

A comment on A 40-line fix eliminated a 400x performance gap. Parent comment

Re:A 40-line fix eliminated a 400x performance gap by jerrinot on Tuesday January 13, 2026 @07:47PM (#46610840)
Only for some clocks (CLOCK_MONOTONIC, etc) and some clock sources. For VIRT/SCHED, the vDSO shim still has to invoke the actual syscall. You can't avoid the kernel transition when you need per-thread accounting.
[ Parent | on HN ]
Re:A 40-line fix eliminated a 400x performance gap by higherhalf on Tuesday January 13, 2026 @08:32PM (#46611200)
Thanks, I really should've looked deeper than that.
[ Parent | on HN ]
Re:A 40-line fix eliminated a 400x performance gap by jerrinot on Tuesday January 13, 2026 @08:35PM (#46611228)
no problem at all, I was confused too when I saw the profile for the first time.
[ Parent | on HN ]
Re:A 40-line fix eliminated a 400x performance gap by touisteur on Wednesday January 14, 2026 @02:48AM (#46613421)
Oh for some time after its introduction, CLOCK_MONOTONIC_RAW wasn't vDSO'd and it took some time and syscall profiling ('huh, why do I see these as syscalls in perf record -e syscalls' ...) to understand what was going on.
[ Parent | on HN ]