Messages in this thread |  | | | Subject | Re: REGRESSION: Performance regressions from switching anon_vma->lock to mutex | | From | Linus Torvalds <> | | Date | Wed, 15 Jun 2011 14:51:34 -0700 |
| |
Peter Zijlstra <a.p.zijlstra@chello.nl> wrote: >removing >that sysconf() wreckage closed the gap considerably (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).
Yeah, I think it's ridiculous to say that glibc is not doing something stupid and that it's a problem with kernel interfaces.
Do the proc file parsing once and cache the result in a static variable. Doing it over and over again is just crazy.
Adding new system calls because glibc is crazy is insane.
Linus
|  |