netdev - Re: [net-next] ppp: rtnetlink device handling

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: Windows password security audit tool. GUI, reports in PDF.

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

Message-ID: <CA+icZUWSytUFWPJ43Da-Kkct=B75JfoQLx8608gJ=mxaCSyeww@mail.gmail.com>
Date:	Thu, 31 Dec 2015 13:03:50 +0100
From:	Sedat Dilek <sedat.dilek@...il.com>
To:	Guillaume Nault <g.nault@...halink.fr>
Cc:	David Miller <davem@...emloft.net>,
	"netdev@...r.kernel.org" <netdev@...r.kernel.org>
Subject: Re: [net-next] ppp: rtnetlink device handling

On Thu, Dec 31, 2015 at 12:01 PM, Sedat Dilek <sedat.dilek@...il.com> wrote:
> On Thu, Dec 31, 2015 at 11:41 AM, Guillaume Nault <g.nault@...halink.fr> wrote:
>> On Thu, Dec 31, 2015 at 08:46:59AM +0100, Sedat Dilek wrote:
>>> Hi Guillaume,
>>>
>>> can you explain why you moved ppp to rtnetlink device handling?
>>> Benefits, etc.?
>>>
>> The objective is to bring all the flexibility of rtnetlink device
>> creation to ppp interfaces. This is particularly useful for Network
>> Access Servers which need to define device properties at creation
>> time. My use case includes setting device name and netns, which is
>> impossible with the ioctl-based interface without generating transient
>> effects.
>>
>
> What do you mean by "my use-case"?
> Example?
>
>>> Does anything change when using NetworkManager/ModemManager/pppd for
>>> my network setup/handling (here: Ubuntu/precise AMD64)?
>>>
>> No. The ioctl-based ABI remains unchanged, so programs using it won't
>> see any difference. The rtnl ABI only brings a new way to create PPP
>> devices.
>>
>> Whether a ppp device is created with the ioctl or rtnl ABI doesn't
>> change its behaviour. In particular ppp devices created with rtnl can
>> still be configured with the old ioctls. The only difference I can
>> think of, is that managing ppp devices with rtnl allows a user to
>> remove it with "ip link del", while this is not possible with ppp
>> devices created with ioctl.
>>
>
> That all sounds like an improvement.
>
>>> Thanks in advance.
>>>
>> You're welcome.
>>
>
> Building net-next-20151231 with your 3 patches on top...
>
> Guillaume Nault (8):
>       ppp: don't set sk_state to PPPOX_ZOMBIE in pppoe_disc_rcv()
>       ppp: remove PPPOX_ZOMBIE socket state
>       pppox: use standard module auto-loading feature
>       ppp: define "ppp" device type
>       ppp: declare ppp devices as enumerated interfaces
>       ppp: define reusable device creation functions <--- (1)
>       ppp: implement rtnetlink device handling <--- (2)
>       l2tp: rely on ppp layer for skb scrubbing <--- (3)
>
> ...will report.
>
> - Sedat -

# LC_ALL=C ip link show
1: lo: <LOOPBACK,UP,LOWER_UP> mtu 65536 qdisc noqueue state UNKNOWN mode DEFAULT
    link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
2: eth0: <NO-CARRIER,BROADCAST,MULTICAST,UP> mtu 1500 qdisc pfifo_fast
state DOWN mode DEFAULT qlen 1000
    link/ether e8:03:9a:36:17:a9 brd ff:ff:ff:ff:ff:ff
3: wlan0: <BROADCAST,MULTICAST> mtu 1500 qdisc noop state DOWN mode
DEFAULT qlen 1000
    link/ether 88:53:2e:ac:c3:12 brd ff:ff:ff:ff:ff:ff
4: wwan0: <BROADCAST,MULTICAST> mtu 1500 qdisc noop state DOWN mode
DEFAULT qlen 1000
    link/ether 02:50:f3:00:00:00 brd ff:ff:ff:ff:ff:ff
5: ppp0: <POINTOPOINT,MULTICAST,NOARP,UP,LOWER_UP> mtu 1500 qdisc
pfifo_fast state UNKNOWN mode DEFAULT qlen 3
    link/ppp

- Sedat -

Download attachment "run_htb-with-fq_codel_ppp0.sh" of type "application/x-sh" (1481 bytes)