On Sat, 6 Oct 2018 00:03:58 +0800 (CST) <[email protected]> wrote:
> >Hi Peng, > > >On Sat, Oct 06, 2018 at 06:22:11AM +0800, Peng Hao wrote: > >> find_lock_lowest_rq may or not releease rq lock when return > >> lowest_rq=NULL, but it is fuzzy. > >> If not releasing rq lock, it is unnecessary to re-call > >> pick_next_pushable_task. > > >IIRC, deadline.c uses a similar pattern (c.f., find_lock_later_rq() and > >pick_next_pushable_dl_task()): should it be considered for this change? > peterz asked the same question. > at first I thought dl's retry action is simple. But now the change is > simpler, I will > add it. > I would still do that as a separate patch though. -- Steve
