[Top] [All Lists]


To: jgarzik@xxxxxxxxxxxxxxxx (Jeff Garzik)
Subject: Re: IFF_PROMISC bug?
From: kuznet@xxxxxxxxxxxxx
Date: Wed, 13 Feb 2002 21:36:17 +0300 (MSK)
Cc: davem@xxxxxxxxxx, ak@xxxxxxx, linux-net@xxxxxxxxxxxxxxx, netdev@xxxxxxxxxxx
In-reply-to: <3C6A2789.5DCC1C8F@xxxxxxxxxxxxxxxx> from "Jeff Garzik" at Feb 13, 2 03:44:57 am
Sender: owner-netdev@xxxxxxxxxxx

> For SIOCGIFFLAGS, my preference would be to manage the IFF_PROMISC bit

Before all SIOCGIFFLAGS must be managed consistently with SIOCSIFFLAGS.
All the rest of requirements are details comparing to this.

SIOCSIFFLAGS is not deprecated as api element. It is deprecated
for uses sort of made by libpcap, for uses to enable bridging etc.
I.e. any which are supposed to be interleaved.

Its global function remains for those who want this.

> actually enabled or disabled.  Thus IFF_PROMISC will reflect the true
> state of the hardware,

Are you about IFF_PROMISC or about ifconfig yet? It is not one thing. :-)

Anyway, ifconfig should show that state which it changes.
If you want to look at "true" state you may use "ip link".
What is the difference? The difference is that "ip link" _does_ _not_
_allow_ to change this flag. It is just an external state from
its viewpoint.

No problems with "improving" ifconfig to show true state too,
provided untrue one remains on its place too.


<Prev in Thread] Current Thread [Next in Thread>