Return-path: Received: from static-ip-62-75-166-246.inaddr.intergenia.de ([62.75.166.246]:33153 "EHLO vs166246.vserver.de" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1751734AbXBEQyQ (ORCPT ); Mon, 5 Feb 2007 11:54:16 -0500 From: Michael Buesch To: "Jouni Malinen" Subject: Re: [PATCH] d80211: Add API to generate RTS and CTS-to-self frames Date: Mon, 5 Feb 2007 17:53:44 +0100 Cc: John Linville , linux-wireless@vger.kernel.org References: <200702051412.50703.mb@bu3sch.de> <200702051628.46518.mb@bu3sch.de> <20070205164738.GA31058@instant802.com> In-Reply-To: <20070205164738.GA31058@instant802.com> MIME-Version: 1.0 Content-Type: text/plain; charset="iso-8859-1" Message-Id: <200702051753.45093.mb@bu3sch.de> Sender: linux-wireless-owner@vger.kernel.org List-ID: On Monday 05 February 2007 17:47, Jouni Malinen wrote: > On Mon, Feb 05, 2007 at 04:28:46PM +0100, Michael Buesch wrote: > > > As I said in response to Johannes' comment, if a driver needs just > > the duration, we should export the functions to calculate it > > and remove the calculation from the tx_control and the tx_control > > setup path. > > In that case, I would like to see this change being included in the same > patch/patchset.. Without this, I can only NAK this kind of change since > it removes functionality needed for some designs and that is not > something I would like to see happening with net/d80211 since the goal > is to allow all reasonable hardware designs to work with it. Ok, I already resent the patchset. ;) > > Currently there is no driver using it, so I simply removed it. > > There are, but maybe not in wireless-dev.git. > > > BTW: Which cards need the duration, but not the whole frame? > > Any card that supports different transmit rates for hw retries involving > RTS/CTS/fragmentation (e.g., Atheros). Ok, nice. I just want to make sure that we don't have unused features (aka bloat) in the stack. -- Greetings Michael.