From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1756478Ab1EWQTS (ORCPT ); Mon, 23 May 2011 12:19:18 -0400 Received: from pfepb.post.tele.dk ([195.41.46.236]:55849 "EHLO pfepb.post.tele.dk" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1756290Ab1EWQTR (ORCPT ); Mon, 23 May 2011 12:19:17 -0400 Date: Mon, 23 May 2011 18:19:14 +0200 From: Sam Ravnborg To: Ingo Molnar , Christopher Li Cc: Linus Torvalds , Stephen Rothwell , Thomas Gleixner , "H. Peter Anvin" , David Miller , linux-next@vger.kernel.org, linux-kernel@vger.kernel.org, x86@kernel.org, Peter Zijlstra , Andrew Morton Subject: Re: [bloat] Measuring header file bloat effects on kernel build performance: a more than 2x slowdown ... Message-ID: <20110523161914.GA15464@merkur.ravnborg.org> References: <20110520161210.81bbef3a.sfr@canb.auug.org.au> <20110523090918.GA5474@elte.hu> MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline In-Reply-To: <20110523090918.GA5474@elte.hu> User-Agent: Mutt/1.5.18 (2008-05-17) Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org > > Such thing happen due to: > > - header files only get added, almost never removed > > The key thing was that the build did not break when prefetch.h was kept > dangling. Not sure what to do about that - for humans a dangling header is > absolutely non-obvious to find - we'd need tooling help. Long time ago I discussed this briefly with Christopher Li, and he hacked support for this in sparse. The branch is still around: http://git.kernel.org/?p=devel/sparse/chrisl/sparse.git;a=shortlog;h=unused-include-files I never got around to play with it for various reasons. IIRC it will list include files not used in _current_ configuration, so use if CONFIG_ etc. may give different results. Sam