[Top] [All Lists]

Re: System crash in tcp_fragment()

To: "David S. Miller" <davem@xxxxxxxxxx>
Subject: Re: System crash in tcp_fragment()
From: Andi Kleen <ak@xxxxxxx>
Date: Tue, 21 May 2002 11:49:22 +0200
Cc: george@xxxxxxxxxx, niv@xxxxxxxxxx, kuznet@xxxxxxxxxxxxx, ak@xxxxxxx, netdev@xxxxxxxxxxx, linux-net@xxxxxxxxxxxxxxx, ak@xxxxxx, pekkas@xxxxxxxxxx
In-reply-to: <20020520.230021.29510217.davem@xxxxxxxxxx>
References: <Pine.LNX.4.33.0205201836160.9301-100000@xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx> <3CE9E466.AC2358EE@xxxxxxxxxx> <20020520.230021.29510217.davem@xxxxxxxxxx>
Sender: owner-netdev@xxxxxxxxxxx
User-agent: Mutt/
> That's not the problem.  We use per-cpu values for each counter (and
> when the user asks for the value, we add together the values from
> each processor).

At least on x86 gcc usually seems to just generate an incl, which should
be ok because it is atomic enough (even when a reschedule happens it will 
act as a full memory barrier)

So it'll likely just be a problem for load-store architectures.


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