Messages in this thread | | | Date | Wed, 15 Jun 2011 15:15:32 -0700 | From | Andi Kleen <> | Subject | Re: REGRESSION: Performance regressions from switching anon_vma->lock to mutex |
| |
(needing to know the > number of cpus to optimize locking sounds like a trainwreck all of its > own, needing it _that_ often instead of just once at startup is even > worse).
libdb does it only once per startup, it's just that it gets restarted for every child (it's a library, not a server)
Really the kernel just needs to provide a faster way to get that. Requiring /proc/stat for that is just insane.
I'll resend the sysconf patchkit I guess :-)
-Andi
| |