Messages in this thread | | | Date | Fri, 28 Nov 2008 13:44:05 +0900 | From | Tejun Heo <> | Subject | Re: [PATCH] poll: allow f_op->poll to sleep, take#6 |
| |
Davide Libenzi wrote: > Looks OK to me, although it'd be better if some more folks eye it, in > order to avoid painful mistakes. > Did you test it al all in a live system?
Depends on what you call a live system. I tested the functionalities and they worked okay but considering the failure scenarios, it's highly unlikely to happen especially on x86 cpus with all its nice memory ordering rules. Now that it has smp_wmb() on the wake up side and full mb on waking up side, it's as conservative as it gets, so I think we're in the safe.
Thanks.
-- tejun
| |