Return-path: Received: from nbd.name ([46.4.11.11]:58317 "EHLO nbd.name" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1754944Ab1HBUUS (ORCPT ); Tue, 2 Aug 2011 16:20:18 -0400 Message-ID: <4E385BFA.90406@openwrt.org> (sfid-20110802_222040_277348_759A2386) Date: Tue, 02 Aug 2011 22:20:10 +0200 From: Felix Fietkau MIME-Version: 1.0 To: Larry Finger CC: John W Linville , geert@linux-m68k.org, chaoming_li@realsil.com.cn, linux-wireless@vger.kernel.org, linux-kernel@vger.kernel.org Subject: Re: [PATCH] rtlwifi: Fix build errors on sh4 architecture References: <4e3855e4.8qPfeuR18XtAq7Vs%Larry.Finger@lwfinger.net> In-Reply-To: <4e3855e4.8qPfeuR18XtAq7Vs%Larry.Finger@lwfinger.net> Content-Type: text/plain; charset=ISO-8859-1; format=flowed Sender: linux-wireless-owner@vger.kernel.org List-ID: On 2011-08-02 9:54 PM, Larry Finger wrote: > When rtlwifi is built for the sh4 architecture, build errors of the following > type occur. An explicit reference to should fix these. I have not > setup the cross-build environment, thus this patch is untested on that platform. > It does not cause any problems on i386 or x86_64. > > v3.0/sh4/sh-allmodconfig v3.0/sh4/sh-allyesconfig > src/drivers/net/wireless/rtlwifi/rtl8192c/../pci.h:290: error: implicit declaration of function 'outl': 4 errors in 2 logs > v3.0/sh4/sh-allmodconfig v3.0/sh4/sh-allyesconfig > src/drivers/net/wireless/rtlwifi/rtl8192c/../pci.h:295: error: implicit declaration of function 'outb': 4 errors in 2 logs > v3.0/sh4/sh-allmodconfig v3.0/sh4/sh-allyesconfig > src/drivers/net/wireless/rtlwifi/rtl8192c/../pci.h:300: error: implicit declaration of function 'inb': 4 errors in 2 logs > v3.0/sh4/sh-allmodconfig v3.0/sh4/sh-allyesconfig > src/drivers/net/wireless/rtlwifi/rtl8192c/../pci.h:305: error: implicit declaration of function 'inw': 4 errors in 2 logs > > Signed-off-by: Larry Finger > Cc: Geert Uytterhoeven > --- > > John, > > I'm not sure where this goes. I guess that 3.2 would be OK, unless someone > actually tries to build it on sh4. The only time I have seen the errors is > in Geert's listing of errors and warnings. I looked at the code that uses raw port access, and it looks to me like this is just a bunch of x86 specific hacks to mess with the ASPM settings of the PCI bridge. This code should probably be either #ifdef'd, removed or rewritten. I doubt it'll work properly on other architectures. - Felix