Messages in this thread | | | Date | Thu, 18 Jun 2009 10:52:24 -0700 (PDT) | From | Davide Libenzi <> | Subject | Re: [KVM-RFC PATCH 1/2] eventfd: add an explicit srcu based notifier interface |
| |
On Thu, 18 Jun 2009, Michael S. Tsirkin wrote:
> On Wed, Jun 17, 2009 at 04:21:19PM -0700, Davide Libenzi wrote: > > The interface is just ugly IMO. You have eventfd_signal() that can sleep, > > or not, depending on the registered ->signal() function implementations. > > This is pretty bad, a lot worse than the theoretical us spent in the > > schedule_work() processing. > > I agree. How about the idea of introducing eventfd_signal_from_task > which can sleep? Does this sound same?
You're basically asking to double the size of eventfd, make the signal path heavier, make the eventf size bigger, w/out having provided any *real life* measurement whatsoever to build a case for it. WAY too much stuff went in by just branding the latest coolest names as reasons for them. And all this to remove the wakeup of a *kernel* thread going to run in the same CPU where the work has been scheduled. Come back with *replicable* real life benchmarks, and then we'll see what the best approach to address it will be.
- Davide
| |