From mboxrd@z Thu Jan 1 00:00:00 1970 From: Eric Dumazet Subject: Re: txqueuelen has wrong units; should be time Date: Tue, 01 Mar 2011 06:36:58 +0100 Message-ID: <1298957818.2676.15.camel@edumazet-laptop> References: <1298793252.8726.45.camel@edumazet-laptop> <20110227125540.40754c5y78j9u2m8@hayate.sektori.org> <1298837273.8726.128.camel@edumazet-laptop> <20110228161115.GB2515@tuxdriver.com> <1298911694.2941.639.camel@edumazet-laptop> <20110228165501.GC2515@tuxdriver.com> <1298955683.2676.2.camel@edumazet-laptop> Mime-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: QUOTED-PRINTABLE Cc: John Heffner , "John W. Linville" , Jussi Kivilinna , Mikael Abrahamsson , linux-kernel , netdev@vger.kernel.org To: Albert Cahalan Return-path: In-Reply-To: <1298955683.2676.2.camel@edumazet-laptop> Sender: linux-kernel-owner@vger.kernel.org List-Id: netdev.vger.kernel.org Le mardi 01 mars 2011 =C3=A0 06:01 +0100, Eric Dumazet a =C3=A9crit : > Le lundi 28 f=C3=A9vrier 2011 =C3=A0 23:11 -0500, Albert Cahalan a =C3= =A9crit : >=20 > > It sounds like you need a callback or similar, so that TCP can be > > informed later that the drop has occurred. >=20 > There is the thing called skb destructor / skb_orphan() mess, that is > not stackable... Might extend this to something more clever, and be a= ble > to call functions (into TCP stack for example) giving a status of skb= : > Sent, or dropped somewhere in the stack... >=20 One problem of such schem is the huge extra cost involved, extra locking, extra memory allocations, extra atomic operations...