From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S933417Ab0I0TBc (ORCPT ); Mon, 27 Sep 2010 15:01:32 -0400 Received: from mail-iw0-f174.google.com ([209.85.214.174]:33542 "EHLO mail-iw0-f174.google.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S933403Ab0I0TB3 convert rfc822-to-8bit (ORCPT ); Mon, 27 Sep 2010 15:01:29 -0400 MIME-Version: 1.0 X-Originating-IP: [46.116.86.31] In-Reply-To: <20100927184445.GF11086@tuxdriver.com> References: <20100927154408.55a28127.sfr@canb.auug.org.au> <20100927180412.GD11086@tuxdriver.com> <20100927.111818.149846920.davem@davemloft.net> <20100927182546.GE11086@tuxdriver.com> <20100927184445.GF11086@tuxdriver.com> From: Ohad Ben-Cohen Date: Mon, 27 Sep 2010 21:01:08 +0200 Message-ID: Subject: Re: linux-next: build failure after merge of the final tree (net tree related) To: "John W. Linville" Cc: David Miller , sfr@canb.auug.org.au, netdev@vger.kernel.org, linux-next@vger.kernel.org, linux-kernel@vger.kernel.org Content-Type: text/plain; charset=ISO-8859-1 Content-Transfer-Encoding: 8BIT Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org On Mon, Sep 27, 2010 at 8:44 PM, John W. Linville wrote: > On Mon, Sep 27, 2010 at 08:38:50PM +0200, Ohad Ben-Cohen wrote: >> On Mon, Sep 27, 2010 at 8:25 PM, John W. Linville >> wrote: >> >> BTW, if this fuglet going to work if the driver is built modular? >> >> >> >> The idea of this things seems to be to allow arch platform code to call >> >> into it to set things.  Such calls from arch platform code are going >> >> to be "obj-y" >> >> >> >> So if this is built into the modular driver, I can't see how it >> >> can work. >> > >> > Dave is right -- these Kconfig dependencies (or lack thereof) seem >> > to be wrong. >> >> This is why I had to put it in drivers/net/wireless/Makefile (in both >> the original patch and in the fix I sent earlier) and not in >> drivers/net/wireless/wl12xx/Makefile (as it was in the fix that Dave >> was referring to). >> >> This way it works with a modular driver. > > That doesn't seem very helpful if someone selects > > CONFIG_MACH_OMAP_ZOOM2=y > CONFIG_WL12XX_PLATFORM_DATA=n > CONFIG_WL12XX_PLATFORM_DATA is selected automatically if the relevant driver is selected (this is independent of the underlying arch/board): config WL12XX_PLATFORM_DATA bool depends on WL1271_SDIO != n default y All we have to make sure is that it will be compiled built-in. If we use drivers/net/wireless/wl12xx/Makefile, and we build a modular wl1271_sdio, it won't (despite it being selected as y). That's why I put it in drivers/net/wireless/Makefile. In the original patch I put the code itself in drivers/net/wireless/wl12xx to make it look nicer, but as Stephen reported that seem to break building with a separate object folder... So the fix I sent just moves that builtin code to drivers/net/wireless. This maintains the builtin property even if the driver is modular, and fixes the issue reported by Stephen. Please tell me if you have any thoughts/suggestions. Thanks, Ohad. > or am I missing something? > > John > -- > John W. Linville                Someday the world will need a hero, and you > linville@tuxdriver.com                  might be all we have.  Be ready. >