From mboxrd@z Thu Jan 1 00:00:00 1970 Return-path: Received: from mail-wg0-f48.google.com ([74.125.82.48]:32926 "EHLO mail-wg0-f48.google.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1751562AbbCORSN (ORCPT ); Sun, 15 Mar 2015 13:18:13 -0400 From: Christian Lamparter To: Florian Westphal Cc: linux-wireless@vger.kernel.org, netdev@vger.kernel.org Subject: Re: [PATCH next] drivers: wireless: carl9170: shrink carl9170_tx_info Date: Sun, 15 Mar 2015 18:17:56 +0100 Message-ID: <2002983.Zu6fLi5xz6@debian64> (sfid-20150315_181818_584510_ECF0ACD0) In-Reply-To: <20150314175532.GH14761@breakpoint.cc> References: <1426261045-22373-1-git-send-email-fw@strlen.de> <1689955.Ikz24aORLX@debian64> <20150314175532.GH14761@breakpoint.cc> MIME-Version: 1.0 Content-Type: text/plain; charset="us-ascii" Sender: linux-wireless-owner@vger.kernel.org List-ID: On Saturday, March 14, 2015 06:55:32 PM Florian Westphal wrote: > Christian Lamparter wrote: > > On Friday, March 13, 2015 04:37:25 PM Florian Westphal wrote: > > > its embededded inside rate_driver_data of the ieee80211_tx_info struct, > > > which in turn is stored in skb->cb[]. > > > > > > In order to shrink cb, we need to shrink ieee80211_tx_info which means > > > to downsize all users first. > > > > > > Alternatively, one might be able to remove kref but > > > its less intrusive/simpler to use u32 for timeout handling. > > > > > > Signed-off-by: Florian Westphal > > > --- > > u32 jiffies... that's a lot of pointing (well not so much) and really > > ugly casting (a lot). I guess it would be easier to just use a "per-queue" > > timeout watchdog like almost everybody else. This way, the driver will > > be ready for the next skb->cb shrink as well. > > > > carl9170_tx_ampdu_timeout can be completely removed. > > Maybe, but there is only so much I am willing to do with "compile > tested only" patches... well, reinventing jiffies_32 is just a ugly hack and not "intrusive/simpler" than removing kref here. Please try again. Don't worry about testing, I'm sure this is not a problem and someone will test the patches with the hw on 32-bit and 64-bit platforms. Regards, Christian