Subject: Re: workqueue(9), per-CPU queues (part 2)
To: None <tech-kern@netbsd.org>
From: Mindaugas R. <rmind@NetBSD.org>
List: tech-kern
Date: 07/29/2007 06:12:37
yamt@mwd.biglobe.ne.jp (YAMAMOTO Takashi) wrote:
> i'm curious what actually needs these functionalities.
> 
> when i wrote it, i tried to make the worker code minimum and intentionally
> omitted functionalities which can be implemented in the caller.
> it doesn't mean i will automatically object against them, tho. :)

Something like workqueue_empty/count() could not be implemented in the caller
without violating of the API.

workqueue_wait() could be useful for those who use workqueue_destroy().
Currently, only sysmon_envsys uses and implements it "outside".
Also, I think it could be useful for handling the finish of work, let's say
in asynchronous I/O.

Of course, it is not a necessity, at this moment. Perhaps one could implement
it only if there will be a concrete need.

-- 
Best regards,
Mindaugas
www.NetBSD.org