From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: from e28smtp03.in.ibm.com ([122.248.162.3]:42909 "EHLO e28smtp03.in.ibm.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1752694AbaBNIED (ORCPT ); Fri, 14 Feb 2014 03:04:03 -0500 Received: from /spool/local by e28smtp03.in.ibm.com with IBM ESMTP SMTP Gateway: Authorized Use Only! Violators will be prosecuted for from ; Fri, 14 Feb 2014 13:34:00 +0530 From: "Srivatsa S. Bhat" Subject: [PATCH v2 41/52] oprofile, nmi-timer: Fix CPU hotplug callback registration Date: Fri, 14 Feb 2014 13:28:32 +0530 Message-ID: <20140214075831.22701.70305.stgit@srivatsabhat.in.ibm.com> In-Reply-To: <20140214074750.22701.47330.stgit@srivatsabhat.in.ibm.com> References: <20140214074750.22701.47330.stgit@srivatsabhat.in.ibm.com> MIME-Version: 1.0 Content-Type: text/plain; charset="utf-8" Content-Transfer-Encoding: 7bit Sender: linux-arch-owner@vger.kernel.org List-ID: To: paulus@samba.org, oleg@redhat.com, mingo@kernel.org, rusty@rustcorp.com.au, peterz@infradead.org, tglx@linutronix.de, akpm@linux-foundation.org Cc: paulmck@linux.vnet.ibm.com, tj@kernel.org, walken@google.com, ego@linux.vnet.ibm.com, linux@arm.linux.org.uk, rjw@rjwysocki.net, linux-kernel@vger.kernel.org, linux-arch@vger.kernel.org, srivatsa.bhat@linux.vnet.ibm.com, Robert Richter "Srivatsa S. Bhat" Message-ID: <20140214075832.T8qbiKzZvNDVg-vtGpEJKRMhzoJr3JTEmweALjl8oxU@z> Subsystems that want to register CPU hotplug callbacks, as well as perform initialization for the CPUs that are already online, often do it as shown below: get_online_cpus(); for_each_online_cpu(cpu) init_cpu(cpu); register_cpu_notifier(&foobar_cpu_notifier); put_online_cpus(); This is wrong, since it is prone to ABBA deadlocks involving the cpu_add_remove_lock and the cpu_hotplug.lock (when running concurrently with CPU hotplug operations). Instead, the correct and race-free way of performing the callback registration is: cpu_notifier_register_begin(); for_each_online_cpu(cpu) init_cpu(cpu); /* Note the use of the double underscored version of the API */ __register_cpu_notifier(&foobar_cpu_notifier); cpu_notifier_register_done(); Fix the nmi-timer code in oprofile by using this latter form of callback registration. Cc: Robert Richter Cc: Ingo Molnar Signed-off-by: Srivatsa S. Bhat --- drivers/oprofile/nmi_timer_int.c | 23 +++++++++++++---------- 1 file changed, 13 insertions(+), 10 deletions(-) diff --git a/drivers/oprofile/nmi_timer_int.c b/drivers/oprofile/nmi_timer_int.c index 76f1c93..9559829 100644 --- a/drivers/oprofile/nmi_timer_int.c +++ b/drivers/oprofile/nmi_timer_int.c @@ -108,8 +108,8 @@ static void nmi_timer_shutdown(void) struct perf_event *event; int cpu; - get_online_cpus(); - unregister_cpu_notifier(&nmi_timer_cpu_nb); + cpu_notifier_register_begin(); + __unregister_cpu_notifier(&nmi_timer_cpu_nb); for_each_possible_cpu(cpu) { event = per_cpu(nmi_timer_events, cpu); if (!event) @@ -119,7 +119,7 @@ static void nmi_timer_shutdown(void) perf_event_release_kernel(event); } - put_online_cpus(); + cpu_notifier_register_done(); } static int nmi_timer_setup(void) @@ -132,20 +132,23 @@ static int nmi_timer_setup(void) do_div(period, HZ); nmi_timer_attr.sample_period = period; - get_online_cpus(); - err = register_cpu_notifier(&nmi_timer_cpu_nb); + cpu_notifier_register_begin(); + err = __register_cpu_notifier(&nmi_timer_cpu_nb); if (err) goto out; + /* can't attach events to offline cpus: */ for_each_online_cpu(cpu) { err = nmi_timer_start_cpu(cpu); - if (err) - break; + if (err) { + cpu_notifier_register_done(); + nmi_timer_shutdown(); + return err; + } } - if (err) - nmi_timer_shutdown(); + out: - put_online_cpus(); + cpu_notifier_register_done(); return err; }