linux-kernel - RE: microsecond event scheduling in an application

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [day] [month] [year] [list]

Message-ID: <005601ca3b4e$d5361660$7fa24320$@kaist.ac.kr>
Date:	Tue, 22 Sep 2009 15:35:09 +0900
From:	"Junhee Lee" <junhee@...sys.kaist.ac.kr>
To:	"'Arjan van de Ven'" <arjan@...radead.org>,
	"'Robert Hancock'" <hancockrwd@...il.com>
Cc:	<linux-kernel@...r.kernel.org>
Subject: RE: microsecond event scheduling in an application

Thanks Arjan van de Ven. 
Your comment is very useful to me.
Now I am trying to solve microsecond events scheduling using hrtimer.

-----Original Message-----
From: Arjan van de Ven [mailto:arjan@...radead.org] 
Sent: Monday, September 21, 2009 3:31 AM
To: Robert Hancock
Cc: Junhee Lee; linux-kernel@...r.kernel.org
Subject: Re: microsecond event scheduling in an application

On Sun, 20 Sep 2009 12:26:09 -0600
Robert Hancock <hancockrwd@...il.com> wrote:

> On 09/08/2009 08:27 AM, Junhee Lee wrote:
> > I am working on event scheduler which handles events in microsecond
> > level. Actual this program is a network emulator using simulation
> > codes. I'd like to expect that network emulator is working as
> > simulation behaviors. Thus high resolution timer interrupt is
> > required. But high resolution timer interrupt derived by high tick
> > frequency (jiffies clock) must effect the system performance.
> > Are there any comments or ways to support microsecond event
> > scheduling without performance degradation?
> 
> Just increasing HZ will degrade performance, yes, but we have
> hrtimers now which should be able to use granularities smaller than
> one jiffy, so it shouldn't be needed..


select/poll use hrtimers, which are jiffies independent....

you'll be hard pressed to notice jiffies granularity in userspace
nowadays..



-- 
Arjan van de Ven 	Intel Open Source Technology Centre
For development, discussion and tips for power savings, 
visit http://www.lesswatts.org

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