From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1752443Ab1FUGLu (ORCPT ); Tue, 21 Jun 2011 02:11:50 -0400 Received: from mail-ew0-f46.google.com ([209.85.215.46]:37834 "EHLO mail-ew0-f46.google.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1751589Ab1FUGLs convert rfc822-to-8bit (ORCPT ); Tue, 21 Jun 2011 02:11:48 -0400 DomainKey-Signature: a=rsa-sha1; c=nofws; d=gmail.com; s=gamma; h=mime-version:in-reply-to:references:date:message-id:subject:from:to :content-type:content-transfer-encoding; b=oyFWZA1ItQlBVwn4ndjg550ngFCkynVZ1iYjvvHZsq+dUdpGrWk/+KzEsrap9I4HLr hsITIXNTYRnYwUghMLsP+sH1PFEwlVzXb/cARKyNJic9Q99dIEgHPcLT94FohQtlTzN3 55/vZWaw07xRMyriGeD9PKetG6QA0spe/73QQ= MIME-Version: 1.0 In-Reply-To: <201106201500.58617.arnd@arndb.de> References: <201106201500.58617.arnd@arndb.de> Date: Tue, 21 Jun 2011 15:11:47 +0900 Message-ID: Subject: Re: 'hweight_long' include/linux/bitops.h compile warning problem. From: NamJae Jeon To: Arnd Bergmann , linux-kernel@vger.kernel.org, linux-arch@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 Hello Arnd. I attached patch for this warning issue. If there is a problem of patch, plz reply .... Thanks. -------------------------------------------------------------------------------------------------------------------------------------- >>From 28589ea9aabb080020ab3c79adcecad5d7315f39 Mon Sep 17 00:00:00 2001 From: Namjae Jeon Date: Tue, 21 Jun 2011 14:39:03 +0900 Subject: [PATCH] compilation warning is caused when using hweight_long with -Wsign-compare option. Warning message is ' include/linux/bitops.h: In function 'hweight_long': include/linux/bitops.h:49: warning: signed and unsigned type in conditional expression The reason is that the default return value of this macro is signed. So, need type casting to remove warning. Signed-off-by: Namjae Jeon --- include/asm-generic/bitops/const_hweight.h | 2 +- 1 files changed, 1 insertions(+), 1 deletions(-) diff --git a/include/asm-generic/bitops/const_hweight.h b/include/asm-generic/bitops/const_hweight.h index fa2a50b..d91b1aa 100644 --- a/include/asm-generic/bitops/const_hweight.h +++ b/include/asm-generic/bitops/const_hweight.h @@ -4,7 +4,7 @@ /* * Compile time versions of __arch_hweightN() */ -#define __const_hweight8(w) \ +#define __const_hweight8(w) (unsigned long) \ ( (!!((w) & (1ULL << 0))) + \ (!!((w) & (1ULL << 1))) + \ (!!((w) & (1ULL << 2))) + \ -- 1.7.4-rc2 -------------------------------------------------------------------------------------------------------------------------------------- 2011/6/20 Arnd Bergmann : > On Monday 20 June 2011, NamJae Jeon wrote: >> I found compile warning while compiling module when using -Wsign-compare option. >> >> include/linux/bitops.h: In function 'hweight_long': >> include/linux/bitops.h:49: warning: signed and unsigned type in >> conditional expression >> >> I found the reason of this problem that the default return value of >> the below macro is signed. >> >> #define __const_hweight8(w)        \ >>       ( (!!((w) & (1ULL << 0))) +       \ >>         (!!((w) & (1ULL << 1))) +       \ >>         (!!((w) & (1ULL << 2))) +       \ >>         (!!((w) & (1ULL << 3))) +       \ >>         (!!((w) & (1ULL << 4))) +       \ >>         (!!((w) & (1ULL << 5))) +       \ >>         (!!((w) & (1ULL << 6))) +       \ >>         (!!((w) & (1ULL << 7))) ) >> >> So, I try to add (unsigned long) in __const_hweight8 like this. >> >> #define __const_hweight8(w)     (unsigned long )        \ >>       ( (!!((w) & (1ULL << 0))) +       \ >>         (!!((w) & (1ULL << 1))) +       \ >>         (!!((w) & (1ULL << 2))) +       \ >>         (!!((w) & (1ULL << 3))) +       \ >>         (!!((w) & (1ULL << 4))) +       \ >>         (!!((w) & (1ULL << 5))) +       \ >>         (!!((w) & (1ULL << 6))) +       \ >>         (!!((w) & (1ULL << 7))) ) >> >> so, I can't see compile warning after fixing it, Would you plz check this ? > > Yes, this looks correct to me. Could you send the change as a proper patch > with your Signed-off-by:? > >        Arnd >