From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1751725AbZL0KLy (ORCPT ); Sun, 27 Dec 2009 05:11:54 -0500 Received: (majordomo@vger.kernel.org) by vger.kernel.org id S1751551AbZL0KLx (ORCPT ); Sun, 27 Dec 2009 05:11:53 -0500 Received: from bombadil.infradead.org ([18.85.46.34]:39476 "EHLO bombadil.infradead.org" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1751471AbZL0KLw (ORCPT ); Sun, 27 Dec 2009 05:11:52 -0500 Subject: Re: tbench regression with 2.6.33-rc1 From: Peter Zijlstra To: Mike Galbraith Cc: Lin Ming , mingo@elte.hu, tglx@linutronix.de, linux-kernel , "Zhang, Yanmin" , =?ISO-8859-1?Q?Fr=E9d=E9ric?= Weisbecker In-Reply-To: <1261906883.5451.1.camel@marge.simson.net> References: <1261739467.10685.18.camel@minggr.sh.intel.com> <1261902760.6201.33.camel@marge.simson.net> <1261904349.15854.2.camel@laptop> <1261906883.5451.1.camel@marge.simson.net> Content-Type: text/plain; charset="UTF-8" Date: Sun, 27 Dec 2009 11:11:02 +0100 Message-ID: <1261908662.15854.4.camel@laptop> Mime-Version: 1.0 X-Mailer: Evolution 2.28.1 Content-Transfer-Encoding: 7bit Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org On Sun, 2009-12-27 at 10:41 +0100, Mike Galbraith wrote: > On Sun, 2009-12-27 at 09:59 +0100, Peter Zijlstra wrote: > > On Sun, 2009-12-27 at 09:32 +0100, Mike Galbraith wrote: > > > Perf also isn't free, and in 33, we lost the ability to config it out, > > > > Uhm, that's a plain bug, we should be able to build a kernel without > > perf support, does the hw breakpoint support have an unfortunate select > > some place or something? > > Yeah, it's selected in arch/x86/Kconfig. The kernel won't build without > it either. Came from this :/ Any way we can fix this? Frederic/Ingo? --- commit 99e8c5a3b875a34d894a711c9a3669858d6adf45 Author: Frederic Weisbecker Date: Thu Dec 17 01:33:54 2009 +0100 hw-breakpoints: Fix hardware breakpoints -> perf events dependency The kbuild's select command doesn't propagate through the config dependencies. Hence the current rules of hardware breakpoint's config can't ensure perf can never be disabled under us. We have: config X86 selects HAVE_HW_BREAKPOINTS config HAVE_HW_BREAKPOINTS select PERF_EVENTS config PERF_EVENTS [...] x86 will select the breakpoints but that won't propagate to perf events. The user can still disable the latter, but it is necessary for the breakpoints. What we need is: - x86 selects HAVE_HW_BREAKPOINTS and PERF_EVENTS - HAVE_HW_BREAKPOINTS depends on PERF_EVENTS so that we ensure PERF_EVENTS is enabled and frozen for x86. This fixes the following kind of build errors: In file included from arch/x86/kernel/hw_breakpoint.c:31: include/linux/hw_breakpoint.h: In function 'hw_breakpoint_addr': include/linux/hw_breakpoint.h:39: error: 'struct perf_event' has no member named 'attr' v2: Select also ANON_INODES from x86, required for perf Reported-by: Cyrill Gorcunov Reported-by: Michal Marek Reported-by: Andrew Randrianasulu Signed-off-by: Frederic Weisbecker Cc: Peter Zijlstra Cc: Arnaldo Carvalho de Melo Cc: Paul Mackerras Cc: Randy Dunlap Cc: K.Prasad LKML-Reference: <1261010034-7786-1-git-send-regression-fweisbec@gmail.com> Signed-off-by: Ingo Molnar diff --git a/arch/Kconfig b/arch/Kconfig index d828758..9d055b4 100644 --- a/arch/Kconfig +++ b/arch/Kconfig @@ -135,9 +135,7 @@ config HAVE_DEFAULT_NO_SPIN_MUTEXES config HAVE_HW_BREAKPOINT bool - depends on HAVE_PERF_EVENTS - select ANON_INODES - select PERF_EVENTS + depends on PERF_EVENTS config HAVE_USER_RETURN_NOTIFIER bool diff --git a/arch/x86/Kconfig b/arch/x86/Kconfig index 3b2a5ac..55298e8 100644 --- a/arch/x86/Kconfig +++ b/arch/x86/Kconfig @@ -50,6 +50,8 @@ config X86 select HAVE_KERNEL_BZIP2 select HAVE_KERNEL_LZMA select HAVE_HW_BREAKPOINT + select PERF_EVENTS + select ANON_INODES select HAVE_ARCH_KMEMCHECK select HAVE_USER_RETURN_NOTIFIER