Messages in this thread | | | Date | Mon, 17 Oct 2016 12:52:50 +0200 | From | Peter Zijlstra <> | Subject | Re: About group scheduling for SCHED_DEADLINE |
| |
On Sun, Oct 16, 2016 at 09:34:20PM +0200, Luca Abeni wrote: > Hi Peter, > > first of all, sorry for the delay in my answer
Not a problem, we're all busy. As it happens I had this thing in Berlin which delayed me reading email in any case ;-)
> > However, I think there's a third alternative. I have memories of a > > paper from UNC (I'd have to dig through the site to see if I can > > still find it) where they argue that for a hierarchical (G-)FIFO you > > should use minimal concurrency, that is run the minimal number of > > (v)cpu servers.
> Ok, I need to find and read that paper
I had a real quick look and it could be this one, but memories are vague...
https://cs.unc.edu/~anderson/papers/ecrts08c.pdf
In particular 3.P.
| |