* jamal <[EMAIL PROTECTED]> 2005-12-07 07:56
> How is this different, conceptually, from any other flag setting being
> lost - for example a promisc or admin up/down?
> In other words if you want to reliably transmit state, shouldnt the
> "responsible system" have to worry about the reliability?
>
> I will let Stefan address the other questions. This specific one was
> brought up by Krzysztof as well and the discussion went on a tangent.
I'm not talking about the same races. It's a conceptual question
which can be solved in two ways:
a) The system noticing and initiating the state change actually
executes it, i.e. netif_carrier_(on|off) also makes sure to
complete the down -> dormant transition etc. This is the way
my patch solved things.
b) The system noticing and initiating the state change relies on
an event listener running async. to perform the transition of
the operational state. This leaves a _wide_ gap open where the
link state and the operational state do not match.
The problem is not the delayed transition change, that is perfectly
fine. The problem is the time where the link state does not match
the operational state for quite a while.
-
To unsubscribe from this list: send the line "unsubscribe netdev" in
the body of a message to [EMAIL PROTECTED]
More majordomo info at http://vger.kernel.org/majordomo-info.html
