netdev - Re: [PATCH] net: don't unnecessarily load kernel modules in dev

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 for Android: free password hash cracker in your pocket

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

Message-ID: <c2d378b6-80b3-d9e1-7087-95f143875067@gmail.com>
Date:   Thu, 8 Mar 2018 10:05:12 -0800
From:   Eric Dumazet <eric.dumazet@...il.com>
To:     David Miller <davem@...emloft.net>, stephen@...workplumber.org
Cc:     pmoore@...hat.com, netdev@...r.kernel.org, viro@...iv.linux.org.uk,
        dac.override@...il.com, selinux@...ho.nsa.gov,
        linux-security-module@...r.kernel.org
Subject: Re: [PATCH] net: don't unnecessarily load kernel modules in
 dev_ioctl()



On 03/08/2018 09:34 AM, David Miller wrote:
> From: Stephen Hemminger <stephen@...workplumber.org>
> Date: Tue, 6 Mar 2018 15:59:20 -0800
> 
>> Actually dev_load by ethernet name is really a legacy thing that
>> should just die,
>>
>> It was kept around so that some very tunnel configuration using special names.
>>
>> 	# ifconfig sit0
>>
>> which probably several web pages still tell users to do...
>> We have much better control now with ip commands so that this is just
>> baggage.
> 
> As you say, some people use this stuff, so we really can't break
> it at this point.

Another problematic legacy behavior is the automatic creation of 
fallback tunnels, which hurts netns creation/deletion.

Some environments want to create a netns for every job/task, and they do 
not care if the init netns has these tunnels or not.

We have a local patch adding yet another knob to control this, since it 
saves a lot of cpu cycles (about 10ms per netns create/delete pair here)

lpk43:~# echo 0 >/proc/sys/net/core/fb_tunnels_only_for_init_net
lpk43:~# time for i in {1..1000}; do unshare -n /bin/false;done

real	0m14.939s
user	0m0.152s
sys	0m1.496s

lpk43:~# unshare -n
lpk43:~# ip link
1: lo: <LOOPBACK> mtu 65536 qdisc noop state DOWN mode DEFAULT group 
default qlen 1000
     link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
2: tunl0@...E: <NOARP> mtu 1480 qdisc noop state DOWN mode DEFAULT group 
default qlen 1000
     link/ipip 0.0.0.0 brd 0.0.0.0
3: gre0@...E: <NOARP> mtu 1476 qdisc noop state DOWN mode DEFAULT group 
default qlen 1000
     link/gre 0.0.0.0 brd 0.0.0.0
4: gretap0@...E: <BROADCAST,MULTICAST> mtu 1462 qdisc noop state DOWN 
mode DEFAULT group default qlen 1000
     link/ether 00:00:00:00:00:00 brd ff:ff:ff:ff:ff:ff
5: erspan0@...E: <BROADCAST,MULTICAST> mtu 1450 qdisc noop state DOWN 
mode DEFAULT group default qlen 1000
     link/ether 00:00:00:00:00:00 brd ff:ff:ff:ff:ff:ff
6: sit0@...E: <NOARP> mtu 1480 qdisc noop state DOWN mode DEFAULT group 
default qlen 1000
     link/sit 0.0.0.0 brd 0.0.0.0
7: ip6tnl0@...E: <NOARP> mtu 1452 qdisc noop state DOWN mode DEFAULT 
group default qlen 1000
     link/tunnel6 :: brd ::
8: ip6gre0@...E: <NOARP> mtu 1448 qdisc noop state DOWN mode DEFAULT 
group default qlen 1000
     link/gre6 00:00:00:00:00:00:00:00:00:00:00:00:00:00:00:00 brd 
00:00:00:00:00:00:00:00:00:00:00:00:00:00:00:00
lpk43:~# exit
logout

lpk43:~# echo 1 >/proc/sys/net/core/fb_tunnels_only_for_init_net

lpk43:~# time for i in {1..1000}; do unshare -n /bin/false;done

real	0m4.169s
user	0m0.202s
sys	0m0.875s

lpk43:~# unshare -n

lpk43:~# ip link
1: lo: <LOOPBACK> mtu 65536 qdisc noop state DOWN mode DEFAULT group 
default qlen 1000
     link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
lpk43:~#