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  PHC 
Open Source and information security mailing list archives
 
Hash Suite for Android: free password hash cracker in your pocket
[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Date:	Mon, 17 Dec 2007 13:37:01 -0500
From:	"Parag Warudkar" <parag.warudkar@...il.com>
To:	"Stephen Hemminger" <shemminger@...ux-foundation.org>
Cc:	"Eric Dumazet" <dada1@...mosbay.com>,
	"David Miller" <davem@...emloft.net>,
	"netdev@...r.kernel.org" <netdev@...r.kernel.org>,
	"Arjan van de Ven" <arjan@...radead.org>
Subject: Re: init_timer_deferrable conversion

On Dec 17, 2007 1:13 PM, Stephen Hemminger
<shemminger@...ux-foundation.org> wrote:
> On Mon, 17 Dec 2007 12:47:59 -0500
> "Parag Warudkar" <parag.warudkar@...il.com> wrote:
>
>
> > On Dec 17, 2007 12:00 PM, Stephen Hemminger
> > <shemminger@...ux-foundation.org> wrote:
> > > > > >
> > > > > > a) drivers/net/sky2.c - watchdog_timer. This was showing up high on
> > > > > > Powertop's list of things that cause routine wakeups from idle. After
> > > > > > converting to init_timer_deferrable() the wakeups went down and this one
> > > > > > no longer shows up in powertop's list. 25% reduction.
> > >
> > > This surprises me because it is a 1 hz timer and uses round_jiffies() in
> > > the current kernel.
> >
> > I am using the current git and I already have low wakeups per second
> > to begin with - 5-7  and out of that 25% are attributed to sky2. Not
> > sure if that matches up with the 1 hz + round_jiffies() logic.
> >
> > But is it conceptually ok to make this deferrable? I suppose yes as
> > it's just a watchdog that checks if the link is up and delaying that
> > would not make a difference?
>
> I think you are going to wake up once a second anyway, so all it
> ends up changing is the accounting. Please check with the powertop
> developers.

As I understand it the advantage of deferrable is that sky2 won't have
to wakeup the CPU just for itself.
It can be coupled with other things that need to wake up the CPU. So
hopefully this isn't just a powertop accounting fixup :)

>
> I'm fine with changing sky2, but it would be good if you could
> go through all the network drivers and fix them as well.
>

Arjan - if there is value in converting netdev watchdogs to deferrable
from a PM perpective I will fix up other drivers as well.

Thanks

Parag
--
To unsubscribe from this list: send the line "unsubscribe netdev" in
the body of a message to majordomo@...r.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

Powered by blists - more mailing lists