Accéder au contenu.
Menu Sympa

starpu-devel - Re: [Starpu-devel] [PATCH 5/7] src/sched_policies/deque_modeling_policy_data_aware.c: add a compute_all_performance_predictions() function.

Objet : Developers list for StarPU

Archives de la liste

Re: [Starpu-devel] [PATCH 5/7] src/sched_policies/deque_modeling_policy_data_aware.c: add a compute_all_performance_predictions() function.


Chronologique Discussions 
  • From: Samuel Thibault <samuel.thibault@ens-lyon.org>
  • To: Cyril Roelandt <cyril.roelandt@inria.fr>
  • Cc: starpu-devel@lists.gforge.inria.fr
  • Subject: Re: [Starpu-devel] [PATCH 5/7] src/sched_policies/deque_modeling_policy_data_aware.c: add a compute_all_performance_predictions() function.
  • Date: Fri, 19 Oct 2012 17:42:21 +0200
  • List-archive: <http://lists.gforge.inria.fr/pipermail/starpu-devel>
  • List-id: "Developers list. For discussion of new features, code changes, etc." <starpu-devel.lists.gforge.inria.fr>

Cyril Roelandt, le Tue 16 Oct 2012 23:45:22 +0200, a écrit :
>
> /* Sometimes workers didn't take the tasks as early
> as we expected */
> fifo->exp_start = STARPU_MAX(fifo->exp_start,
> starpu_timing_now());
> + /* XXX: next 3 lines: should we use
> exp_end[worker][nimpl] instead of
> + fifo->exp_end, as it's done in heft ? */
> fifo->exp_end = fifo->exp_start + fifo->exp_len;

Yes. It's not useful to update the fifo exp_end before we actually push
the task.

Samuel





Archives gérées par MHonArc 2.6.19+.

Haut de le page