From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1751673AbdFGUga (ORCPT ); Wed, 7 Jun 2017 16:36:30 -0400 Received: from mail-io0-f193.google.com ([209.85.223.193]:36480 "EHLO mail-io0-f193.google.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1751299AbdFGUg3 (ORCPT ); Wed, 7 Jun 2017 16:36:29 -0400 MIME-Version: 1.0 In-Reply-To: References: <20170530181306.GV141096@google.com> <20170531235519.GX141096@google.com> <20170606212354.GZ141096@google.com> From: Linus Torvalds Date: Wed, 7 Jun 2017 13:36:27 -0700 X-Google-Sender-Auth: iOW0V9JPEhm93J1uojL9Ljsi4Dw Message-ID: Subject: Re: [RFC] clang: 'unused-function' warning on static inline functions To: Arnd Bergmann Cc: Matthias Kaehlcke , Linux Kernel Mailing List , Andrew Morton , Greg Kroah-Hartman , Ingo Molnar , Thomas Gleixner , Christoph Hellwig , Jens Axboe , Steven Rostedt , David Rientjes , Douglas Anderson , Guenter Roeck , Mark Brown , David Miller , Tom Herbert Content-Type: text/plain; charset="UTF-8" Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org On Wed, Jun 7, 2017 at 12:43 PM, Arnd Bergmann wrote: > > The main reason I see for it is that a lot of the unused inline functions > in C files are mistakes, Bah. Blah blah blah. The clang warnign doesn't actually really buy us anything, and it's a completely pointless difference to gcc. I'm not in the least interested in supporting these kinds of pointless differences. The people who are interested in making the kernel compile well with clang should care about the things that matter, not annoying people with idiotic patches. So stop the idiotic patches. When clang actually adds _value_, that's one thing. Right now it's just stupid noise. For some reason compiler people think that "more warnings are good". No. They are not. More noise without any value is absolutely not good, and an unused inline function si by definition not something we care about. Really. Fit the clang noise. Get clang to generate good code. Once clang has actually proven itself, and we haev years of clang under our belt, and clang isn't just a toy and a source of bugs and pointless warnings as far as kernel builds are concerned, THEN we can start talking about actually making use of clang features. Right now it should be about "don't be a f*cking pain in the arse!" Linus