From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1750840AbaGBXJV (ORCPT ); Wed, 2 Jul 2014 19:09:21 -0400 Received: from mail-ie0-f176.google.com ([209.85.223.176]:40703 "EHLO mail-ie0-f176.google.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1750796AbaGBXJU (ORCPT ); Wed, 2 Jul 2014 19:09:20 -0400 Date: Wed, 2 Jul 2014 16:09:17 -0700 (PDT) From: David Rientjes X-X-Sender: rientjes@chino.kir.corp.google.com To: Alexey Dobriyan cc: Andrew Morton , linux-kernel@vger.kernel.org Subject: Re: [PATCH] headers: remove fs.h from poll.h In-Reply-To: <20140701220642.GA3150@p183.telecom.by> Message-ID: References: <20140701200150.GA31103@p183.telecom.by> <20140701134338.1d221dd127ffef4c25e86800@linux-foundation.org> <20140701220642.GA3150@p183.telecom.by> User-Agent: Alpine 2.02 (DEB 1266 2009-07-14) MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org On Wed, 2 Jul 2014, Alexey Dobriyan wrote: > fs.h inclusion is unnecessary because it is only needed > for "struct file" forward declaration. > > Those fs.h inclusions are fixing inevitable compilation failures. > There maybe more compilation failures. > > What I did to get 2000 number is: > * make allmodconfig > * make > * touch include/linux/fs.h > * make > * [count how many files were rebuilt] > > After more header tweaking the number of files that were rebuilt dropped > by ~2000 but code still compiles, which means fs.h inclusions were useless. > Lots of #ifdefs in fs.h, I hope your series includes building for all possible combinations. > This is first patch in series. Same thing we did with sched.h earlier. > Is it possible to post the whole series together so it makes sense and it's possible to put it through some thorough build testing? If you can point me to a git tree that I can pull, then I can do that on infrastructure that I have probably quicker than most.