From minchan@kernel.org Thu Oct 25 08:47:45 2012
From: Minchan Kim <minchan@kernel.org>
To: linaro-kernel@lists.linaro.org
Subject:
 Re: [RFC v2 0/2] vmevent: A bit reworked pressure attribute + docs +	man page
Date: Thu, 25 Oct 2012 17:53:09 +0900
Message-ID: <20121025085309.GC15767@bbox>
In-Reply-To:
 <CAOJsxLGsjTe13WjY_Q=BLBELwQXOjuwo7PiEKwONHUfR4mQmig@mail.gmail.com>
MIME-Version: 1.0
Content-Type: multipart/mixed; boundary="===============4978251934007229311=="

--===============4978251934007229311==
Content-Type: text/plain; charset="utf-8"
Content-Transfer-Encoding: quoted-printable

Hi Pekka,

On Thu, Oct 25, 2012 at 09:44:52AM +0300, Pekka Enberg wrote:
> On Thu, Oct 25, 2012 at 9:40 AM, Minchan Kim <minchan(a)kernel.org> wrote:
> > Your description doesn't include why we need new vmevent_fd(2).
> > Of course, it's very flexible and potential to add new VM knob easily but
> > the thing we is about to use now is only VMEVENT_ATTR_PRESSURE.
> > Is there any other use cases for swap or free? or potential user?
> > Adding vmevent_fd without them is rather overkill.
>=20
> What ABI would you use instead?

I thought /dev/some_knob like mem_notify and epoll is enough but please keep =
in mind
that I'm not against vmevent_fd strongly. My point is that description should=
 include
explain about why other candidate is not good or why vmevent_fd is better.
(But at least, I don't like vmevent timer polling still and I hope we use it
as last resort once we can find another)

>=20
> On Thu, Oct 25, 2012 at 9:40 AM, Minchan Kim <minchan(a)kernel.org> wrote:
> > I don't object but we need rationale for adding new system call which sho=
uld
> > be maintained forever once we add it.
>=20
> Agreed.
>=20
> --
> To unsubscribe, send a message with 'unsubscribe linux-mm' in
> the body to majordomo(a)kvack.org.  For more info on Linux MM,
> see: http://www.linux-mm.org/ .
> Don't email: <a href=3Dmailto:"dont(a)kvack.org"> email(a)kvack.org </a>

--=20
Kind regards,
Minchan Kim


--===============4978251934007229311==--