From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1754892AbaHGR7N (ORCPT ); Thu, 7 Aug 2014 13:59:13 -0400 Received: from mga09.intel.com ([134.134.136.24]:33122 "EHLO mga09.intel.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1750763AbaHGR7M (ORCPT ); Thu, 7 Aug 2014 13:59:12 -0400 X-ExtLoop1: 1 X-IronPort-AV: E=Sophos;i="5.01,819,1400050800"; d="scan'208";a="555293079" Subject: [PATCH] [BUGFIX] x86: mm: fix rcu splat from new TLB tracepoints To: linux-kernel@vger.kernel.org Cc: Dave Hansen , dave.hansen@linux.intel.com, davej@redhat.com, paulmck@linux.vnet.ibm.com, hpa@zytor.com, x86@kernel.org From: Dave Hansen Date: Thu, 07 Aug 2014 10:58:41 -0700 Message-Id: <20140807175841.5C92D878@viggo.jf.intel.com> Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org From: Dave Hansen This is a bug fix that needs to be pushed to mainline before the 3.17 release. Dave Jones reported seeing a bug from one of my TLB tracepoints: http://lkml.kernel.org/r/20140806181801.GA4605@redhat.com According to Paul McKenney, the right way to fix this is adding an _rcuidle suffix to the tracepoint. http://lkml.kernel.org/r/20140807065055.GA5821@linux.vnet.ibm.com This patch does just that. Signed-off-by: Dave Hansen Reported-by: Dave Jones , Cc: paulmck@linux.vnet.ibm.com --- b/arch/x86/mm/tlb.c | 8 +++++++- 1 file changed, 7 insertions(+), 1 deletion(-) diff -puN arch/x86/mm/tlb.c~mm-fix-rcu-splat-from-tlb-tracepoints arch/x86/mm/tlb.c --- a/arch/x86/mm/tlb.c~mm-fix-rcu-splat-from-tlb-tracepoints 2014-08-07 10:55:43.312283542 -0700 +++ b/arch/x86/mm/tlb.c 2014-08-07 10:56:49.763352575 -0700 @@ -49,7 +49,13 @@ void leave_mm(int cpu) if (cpumask_test_cpu(cpu, mm_cpumask(active_mm))) { cpumask_clear_cpu(cpu, mm_cpumask(active_mm)); load_cr3(swapper_pg_dir); - trace_tlb_flush(TLB_FLUSH_ON_TASK_SWITCH, TLB_FLUSH_ALL); + /* + * This gets called in the idle path where RCU + * functions differently. Tracing normally + * uses RCU, so we have to call the tracepoint + * specially here. + */ + trace_tlb_flush_rcuidle(TLB_FLUSH_ON_TASK_SWITCH, TLB_FLUSH_ALL); } } EXPORT_SYMBOL_GPL(leave_mm); _