Re: [PATCH 13/13] workqueue: record global worker ID instead of poolID in work->data when off-queue

[Lai Jiangshan]

On Tue, Feb 5, 2013 at 6:12 AM, Tejun Heo <tj@xxxxxxxxxx> wrote:
> Hello,
>
> One more thing.
>
> On Fri, Feb 01, 2013 at 02:41:36AM +0800, Lai Jiangshan wrote:
>> @@ -1216,7 +1198,11 @@ static int try_to_grab_pending(struct work_struct *work, bool is_dwork,
>>               list_del_init(&work->entry);
>>               cwq_dec_nr_in_flight(get_work_cwq(work), get_work_color(work));
>>
>> -             clear_work_cwq(work, pool->id);
>> +             /* Does the work is still running? */
>> +             worker = find_worker_executing_work(pool, work);
>> +             worker_gwid = worker ? worker->gwid: WORK_OFFQ_WORKER_NONE;
>> +             clear_work_cwq(work, worker_gwid);
>
> Any chance we can remove the busy_hash?  Having to keep it around
> isn't a big deal but it would be nice if can get rid of it.

It is very possible, but it needs a different trade-off.
And I didn't plan to implement it in the series.
I will discus it when this series is settled down.

Thanks,
lai

>
> Thanks.
>
> --
> tejun
> --
> To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
> the body of a message to majordomo@xxxxxxxxxxxxxxx
> More majordomo info at  http://vger.kernel.org/majordomo-info.html
> Please read the FAQ at  http://www.tux.org/lkml/
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/