Received-SPF: pass (google.com: best guess record for domain of linux-kernel-owner@vger.kernel.org designates 209.132.180.67 as permitted sender) client-ip=209.132.180.67;
Date:   Mon, 3 Sep 2018 15:50:20 +0200
From:   Jiri Benc <jbenc@redhat.com>
To:     Kirill Tkhai <ktkhai@virtuozzo.com>
Cc:     Christian Brauner <christian.brauner@canonical.com>,
        netdev@vger.kernel.org, linux-kernel@vger.kernel.org,
        davem@davemloft.net, kuznet@ms2.inr.ac.ru, yoshfuji@linux-ipv6.org,
        pombredanne@nexb.com, kstewart@linuxfoundation.org,
        gregkh@linuxfoundation.org, dsahern@gmail.com, fw@strlen.de,
        lucien.xin@gmail.com, jakub.kicinski@netronome.com,
        nicolas.dichtel@6wind.com
Subject: Re: [PATCH net-next 0/5] rtnetlink: add IFA_IF_NETNSID for
 RTM_GETADDR
Message-ID: <20180903155020.6112f773@redhat.com>
In-Reply-To: <2319a029-7aca-b7aa-2e8f-4dfdeedcb6df@virtuozzo.com>
References: <20180828231859.29758-1-christian@brauner.io>
        <adc2fae5-e22b-3a74-d531-01570e7970ee@virtuozzo.com>
        <20180829181303.4sacopk7y3p5xyou@gmail.com>
        <81379a4f-7149-10ff-2453-886314d0b0c4@virtuozzo.com>
        <20180830144544.tpross4jd6awou4u@gmail.com>
        <20180901013427.tj3t2mlik4t7hlt5@gmail.com>
        <2319a029-7aca-b7aa-2e8f-4dfdeedcb6df@virtuozzo.com>
MIME-Version: 1.0
Content-Type: text/plain; charset=US-ASCII
Content-Transfer-Encoding: 7bit
Sender: linux-kernel-owner@vger.kernel.org
Precedence: bulk

On Mon, 3 Sep 2018 16:41:45 +0300, Kirill Tkhai wrote:
> But this is a synthetic test, while I asked about real workflow.
> Is this real problem for lxd, and there is observed performance
> decrease?

It's actually not as much a performance problem but rather the only way
to get the data in some situations. Namely, when you have only netnsid.
This happens e.g. when you want to query a veth peer in another netns.

setns() requires a file descriptor which you don't have. Nor there is
a way to convert netnsid to a fd.

While developing the IFLA_IF_NETNSID patch, I was first thinking about
implementing an API doing the conversion. The problem is there's no
good place to put this into. It can't be done over netlink: netlink is
unreliable and you can't have the kernel open a fd for you and lose it.
There's no ioctl to use. So we'd be left with a procfs/sysfs or a
syscall.

Using netnsid to refer to the target netns seems to be a nice solution -
after all, netnsid is the identifier to use in netlink.

 Jiri