Return-path: Received: from wolverine02.qualcomm.com ([199.106.114.251]:65234 "EHLO wolverine02.qualcomm.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S932198Ab1JYMqK (ORCPT ); Tue, 25 Oct 2011 08:46:10 -0400 Message-ID: <4EA6AF8C.20206@qca.qualcomm.com> (sfid-20111025_144615_906531_FFE513FA) Date: Tue, 25 Oct 2011 15:46:04 +0300 From: Kalle Valo MIME-Version: 1.0 To: Vasanthakumar Thiagarajan CC: , , Subject: Re: [PATCH 34/34] ath6kl: Enable not more than one interface to workaround target assert References: <1319537430-12459-1-git-send-email-vthiagar@qca.qualcomm.com> <1319537430-12459-35-git-send-email-vthiagar@qca.qualcomm.com> <4EA6A0CF.70806@qca.qualcomm.com> <20111025120539.GA22698@vasanth-laptop> In-Reply-To: <20111025120539.GA22698@vasanth-laptop> Content-Type: text/plain; charset="ISO-8859-1" Sender: linux-wireless-owner@vger.kernel.org List-ID: On 10/25/2011 03:05 PM, Vasanthakumar Thiagarajan wrote: > On Tue, Oct 25, 2011 at 02:43:11PM +0300, Kalle Valo wrote: >> On 10/25/2011 01:10 PM, Vasanthakumar Thiagarajan wrote: >>> Configuring the firmware for three interfaces causes random >>> target assert. So restrict the number of vifs to just 1 to avoid >>> this issue for now. Once it is fixed in firmware, this patch >>> would be reverted. >> >> Patches 32-34 are problematic. They are changing functionality of >> patches earlier in the series. (I guess you created those patches >> later.) Instead you should edit the original patches, not just add >> extra patches on top. > > Patch 32 and 33 are clean ups, i don't think integrating these > changes into respective patches would be trivial with the number of > patches that need to be rebased after the change. Yeah, that's always a problem when you are sending a huge patchset like this. A solution for this is to work split work to smaller patches, that way changes in patches would cause less work. > So I can drop these two patches and send them sometime later. Dropping doesn't make any difference here actually, still the original code needing for cleanup will stay in the history. > Regarding patch 34, the target configuration changes needed for multi > vif are done in more than one patches and integrating this change > into those changes would not make sense to those configurations. Let > me see how I can integrate this. Thanks. Kalle