From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1755318Ab3KGPzv (ORCPT ); Thu, 7 Nov 2013 10:55:51 -0500 Received: from mail-ee0-f49.google.com ([74.125.83.49]:46111 "EHLO mail-ee0-f49.google.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1752435Ab3KGPzs (ORCPT ); Thu, 7 Nov 2013 10:55:48 -0500 Date: Thu, 7 Nov 2013 16:55:44 +0100 From: Ingo Molnar To: Vince Weaver Cc: Peter Zijlstra , hpa@zytor.com, anton@samba.org, mathieu.desnoyers@polymtl.ca, linux-kernel@vger.kernel.org, michael@ellerman.id.au, paulmck@linux.vnet.ibm.com, benh@kernel.crashing.org, fweisbec@gmail.com, VICTORK@il.ibm.com, tglx@linutronix.de, oleg@redhat.com, mikey@neuling.org, linux-tip-commits@vger.kernel.org, Arnaldo Carvalho de Melo , Jiri Olsa , Namhyung Kim , David Ahern Subject: Re: [tip:perf/core] tools/perf: Add required memory barriers Message-ID: <20131107155544.GA9619@gmail.com> References: <20131030104246.GH16117@laptop.programming.kicks-ass.net> <20131106140011.GL10651@twins.programming.kicks-ass.net> <20131106144456.GI26785@twins.programming.kicks-ass.net> <20131106160720.GK26785@twins.programming.kicks-ass.net> <20131106182437.GJ16117@laptop.programming.kicks-ass.net> <20131107082122.GC32438@gmail.com> MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline In-Reply-To: User-Agent: Mutt/1.5.21 (2010-09-15) Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org * Vince Weaver wrote: > On Thu, 7 Nov 2013, Ingo Molnar wrote: > > > I don't want a library that is external and under-tested: for example > > quite a few of the PAPI breakages were found very late, after a new > > kernel has been released - that's the big disadvantage of > > librarization and decentralization. The decentralized library model > > might work if all you want to create is a second-class also-ran GUI, > > but it just doesn't work very well for actively developed kernel code. > > I would argue that PAPI's problem was because it was trying to use > perf_event_open() which is a complex, poorly documented kernel interface > with a lot of code churn. Usually it's the job of the kernel not to > break user-space, not the other way around. As Linus said it on the Kernel Summit: breakages that don't get reported or don't get noticed essentially don't exist. We can only fix what gets reported in time. > It's too late on the decentralized library issue. PAPI has to support > kernels going back to 2.6.32 so it's going to have its own copy of the > mmap parsing code even if a new syscall gets introduced. > > There are a lot of tools out there now that open-code a perf_event > interface. I don't think it's really possible to say "anyone using the > syscall without using our kernel library is unsupported". I'm not saying that at all - but you appear to expect perfect kernel code and fixes done before you report them: that's an impossible expectation. It's your choice to live outside the space that we readily test and it's your choice to not test your bits with a new kernel in time. Others do not test your code with -rc kernels, they don't report the bugs, so some bugs that affect the PAPI library go unnoticed. Yet you try to blame it on us, which is really backwards ... Thanks, Ingo