Skip site navigation (1)Skip section navigation (2)
Date:      Tue, 14 Nov 2017 11:07:18 -0700
From:      Scott Long <scottl@netflix.com>
To:        Larry McVoy <lm@mcvoy.com>
Cc:        Mateusz Guzik <mjguzik@gmail.com>, freebsd-arch@freebsd.org, Kevin Bowling <kbowling@llnw.com>, gallatin@netflix.com
Subject:   Re: small patch for numactl.  Comments?
Message-ID:  <D6D19BFB-CD33-4666-8D11-962FDC893AD5@netflix.com>
In-Reply-To: <20171114172134.GD6265@mcvoy.com>
References:  <20171114020138.GA18863@mcvoy.com> <20171114171032.ez6pxk3yrlczplvi@mguzik> <20171114172134.GD6265@mcvoy.com>

next in thread | previous in thread | raw e-mail | index | archive | help

> On Nov 14, 2017, at 10:21 AM, Larry McVoy <lm@mcvoy.com> wrote:
>=20
> On Tue, Nov 14, 2017 at 06:10:34PM +0100, Mateusz Guzik wrote:
>>> I'm wacking LMbench to be numa aware and this patch would help me =
make
>>> sure that when you are a numa machine you could insist that people=20=

>>> run the benchmark via numactl (imma gonna blog about numa, it sucks
>>> unless you are numa aware).
>>>=20
>>=20
>> Well, I think the right thing to do is to query the existing policy =
and
>> complain when it turns out nothing is set. Perhaps exit by default =
and
>> add a switch to proceed anyway.
>=20
> As already stated, that means #ifdef-ing portable code.  Not a fan of =
that.
>=20
> I believe someone already approved env var approach anyway.

FWIW, communicating state via environment variables is not a common =
pattern in
the core FreeBSD userland tools.  It=E2=80=99s a bit more common in the =
contributed tools.
That=E2=80=99s probably why people are wishy-washy about your proposal.  =
That said,
there=E2=80=99s no architectural rule against what you=E2=80=99re =
proposing, it=E2=80=99s useful, and it=E2=80=99s not
invasive and needing long discussion.

Scott




Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?D6D19BFB-CD33-4666-8D11-962FDC893AD5>