Return-path: Received: from mga01.intel.com ([192.55.52.88]:28610 "EHLO mga01.intel.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1752761AbZK3BLQ (ORCPT ); Sun, 29 Nov 2009 20:11:16 -0500 Subject: Re: [PATCH v2 2/2] iwmc3200wifi: rx aggregation support From: Zhu Yi To: Johannes Berg Cc: "linville@tuxdriver.com" , "linux-wireless@vger.kernel.org" In-Reply-To: <1259314800.5428.13.camel@johannes.local> References: <1259291917-1399-1-git-send-email-yi.zhu@intel.com> <1259291917-1399-2-git-send-email-yi.zhu@intel.com> <1259314800.5428.13.camel@johannes.local> Content-Type: text/plain; charset="UTF-8" Date: Mon, 30 Nov 2009 09:11:21 +0800 Message-ID: <1259543481.12157.18.camel@debian> Mime-Version: 1.0 Sender: linux-wireless-owner@vger.kernel.org List-ID: On Fri, 2009-11-27 at 17:40 +0800, Johannes Berg wrote: > You did the right thing in mac80211, but not here -- if it returns an > error it hasn't freed 'skb', only modified it. Indeed. > I think that we should have ieee80211_amsdu_to_8023s() consume it in > all cases, just not fill the list (and return an error I guess) when > it can't be decoded properly. Yeah, since I didn't even use it correctly myself, I couldn't expect others do. Besides, the nontransparent reuse of skb for the last item in the list (skb == rx->skb) looks dirty. I'll use RX_QUEUED in v3. Thanks, -yi