From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1754765Ab0HIGk2 (ORCPT ); Mon, 9 Aug 2010 02:40:28 -0400 Received: from mail-pz0-f46.google.com ([209.85.210.46]:34407 "EHLO mail-pz0-f46.google.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1751775Ab0HIGk1 (ORCPT ); Mon, 9 Aug 2010 02:40:27 -0400 DomainKey-Signature: a=rsa-sha1; c=nofws; d=gmail.com; s=gamma; h=subject:from:to:cc:in-reply-to:references:content-type:date :message-id:mime-version:x-mailer:content-transfer-encoding; b=VkDokw7EWFOjp0PLyICBKutIDEaT6m2haGeNEldvmKpHdamOnNYf0D7PqEoyuBhuT3 X20nq5QhTqeYNOZbpzjf114hCrKptnns8I5QTzB8wd43T4rlxgOm2+1eLWhTW3++Nc4L 752UZ8mANAmVzLOFcb9BZQXgrrcUAl27wuEA8= Subject: Re: [PATCH] x86: remove __phys_reloc_hide From: Namhyung Kim To: Andi Kleen Cc: Thomas Gleixner , Ingo Molnar , "H. Peter Anvin" , x86@kernel.org, linux-kernel@vger.kernel.org In-Reply-To: <87k4o08fv9.fsf@basil.nowhere.org> References: <1281303490-390-1-git-send-email-namhyung@gmail.com> <87k4o08fv9.fsf@basil.nowhere.org> Content-Type: text/plain; charset="UTF-8" Date: Mon, 09 Aug 2010 15:40:31 +0900 Message-ID: <1281336031.976.11.camel@leonhard> Mime-Version: 1.0 X-Mailer: Evolution 2.28.3 Content-Transfer-Encoding: 8bit Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org 2010-08-09 (월), 08:22 +0200, Andi Kleen: > It's for the benefit of the compiler, we've had miscompilations > due to undefined overflow for addresses in the past. The optimizer > assumes this won't happen. > > Given the x86-64 version normally doesn't overflow, but it's > still safer to have it. > > -Andi > Hi, I'm not talking about the RELOC_HIDE itself. I do know we need it for some specific case, ie. percpu. But in this case __pa_symbol(x) is expanded to RELOC_HIDE((unsigned long)(x), 0) which does nothing meaningful. I believe the overflow is not a concern here. -- Regards, Namhyung Kim