Messages in this thread | | | Date | Sat, 8 Nov 2008 10:41:16 -0800 | From | Arjan van de Ven <> | Subject | Re: [git pull] scheduler updates |
| |
On Sat, 8 Nov 2008 10:28:21 -0800 (PST) Linus Torvalds <torvalds@linux-foundation.org> wrote:
> > > On Sat, 8 Nov 2008, Ingo Molnar wrote: > > > > Ingo Molnar (2): > > sched: improve sched_clock() performance > > sched: optimize sched_clock() a bit > > Btw, why do we do that _idiotic_ rdtsc_barrier() AT ALL? > > No sane user can possibly want it. If you do 'rdtsc', there's nothing > you can do about a few cycles difference due to OoO _anyway_. Adding > barriers is entirely meaningless - it's not going to make the return > value mean anything else. > > Can we please just remove that idiocy? Or can somebody give a _sane_ > argument for it?
historically it was for early AMD cpus (K7, not sure if early K8 did this) where 2 consecutive rdtsc's in the same codestream would get reordered compared to eachother, so you could observe the tsc go backwards...
-- Arjan van de Ven Intel Open Source Technology Centre For development, discussion and tips for power savings, visit http://www.lesswatts.org
| |