Messages in this thread | | | Subject | Re: 'global' rq->clock | From | Peter Zijlstra <> | Date | Sat, 03 May 2008 00:07:34 +0200 |
| |
On Fri, 2008-05-02 at 14:48 -0700, David Miller wrote: > From: Peter Zijlstra <a.p.zijlstra@chello.nl> > Date: Fri, 02 May 2008 21:56:26 +0200 > > > Ok, the the below would need something that relates tick_timestamp's to > > one another.. probably sucks without it.. > > > > OTOH, Andi said he was working on a fastish global sched_clock() thingy, > > Andi got a link to that code? > > While I'm fine with this kind of stuff being added to constantly cope > with x86's joke of a TSC register implementation, it's starting to > become an enormous burdon for platforms where the TICK source actually > works properly. > > Heck, on my Niagara2 chips, all 64 cpus use the same physical register > for the TICK source, it physically can't get desynchronized :-) > > So, a way to turn all of this muck off would be much appreciated. > I'm happy to test anything on sparc64, and I'm sure the powerpc > folks are as well. > > And I also heard a rumor that Peter has access to a machine with a > stable tick source for testing :-)
Yeah, I was thinking of a way for architectures to signify how much help they need building the various clocks. Sparc, power and s390 would not need any help.
| |