From mboxrd@z Thu Jan 1 00:00:00 1970 Return-path: Received: from fgwmail7.fujitsu.co.jp ([192.51.44.37]) by bombadil.infradead.org with esmtp (Exim 4.69 #1 (Red Hat Linux)) id 1MHBUl-000851-0o for kexec@lists.infradead.org; Thu, 18 Jun 2009 06:53:23 +0000 Received: from m2.gw.fujitsu.co.jp ([10.0.50.72]) by fgwmail7.fujitsu.co.jp (Fujitsu Gateway) with ESMTP id n5I6r91u026656 for (envelope-from seto.hidetoshi@jp.fujitsu.com); Thu, 18 Jun 2009 15:53:09 +0900 Received: from smail (m2 [127.0.0.1]) by outgoing.m2.gw.fujitsu.co.jp (Postfix) with ESMTP id CDD3045DE63 for ; Thu, 18 Jun 2009 15:53:08 +0900 (JST) Received: from s2.gw.fujitsu.co.jp (s2.gw.fujitsu.co.jp [10.0.50.92]) by m2.gw.fujitsu.co.jp (Postfix) with ESMTP id 785EF45DE5D for ; Thu, 18 Jun 2009 15:53:08 +0900 (JST) Received: from s2.gw.fujitsu.co.jp (localhost.localdomain [127.0.0.1]) by s2.gw.fujitsu.co.jp (Postfix) with ESMTP id 40BA61DB804B for ; Thu, 18 Jun 2009 15:53:08 +0900 (JST) Received: from m107.s.css.fujitsu.com (m107.s.css.fujitsu.com [10.249.87.107]) by s2.gw.fujitsu.co.jp (Postfix) with ESMTP id C2D091DB803E for ; Thu, 18 Jun 2009 15:53:07 +0900 (JST) Message-ID: <4A39E451.2020100@jp.fujitsu.com> Date: Thu, 18 Jun 2009 15:53:05 +0900 From: Hidetoshi Seto MIME-Version: 1.0 Subject: [PATCH 7/7] ia64, kdump: Short path to freeze CPUs References: <4A39E247.4030908@jp.fujitsu.com> In-Reply-To: <4A39E247.4030908@jp.fujitsu.com> List-Id: List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Content-Type: text/plain; charset="us-ascii" Content-Transfer-Encoding: 7bit Sender: kexec-bounces@lists.infradead.org Errors-To: kexec-bounces+dwmw2=infradead.org@lists.infradead.org To: linux-ia64@vger.kernel.org, linux-kernel@vger.kernel.org Cc: Haren Myneni , kexec@lists.infradead.org, Vivek Goyal Setting monarch_cpu = -1 to let slaves frozen might not work, because there might be slaves being late, not entered the rendezvous yet. Such slaves might be caught in while (monarch_cpu == -1) loop. Use kdump_in_progress instead of monarch_cpus to break INIT rendezvous and let all slaves enter DIE_INIT_SLAVE_LEAVE smoothly. And monarch no longer need to manage rendezvous if once kdump_in_progress is set, catch the monarch in DIE_INIT_MONARCH_ENTER then. Signed-off-by: Hidetoshi Seto Cc: Vivek Goyal Cc: Haren Myneni Cc: kexec@lists.infradead.org --- arch/ia64/kernel/crash.c | 15 ++++++--------- arch/ia64/kernel/mca.c | 15 +++++++++++++-- 2 files changed, 19 insertions(+), 11 deletions(-) diff --git a/arch/ia64/kernel/crash.c b/arch/ia64/kernel/crash.c index f43eccf..cf833af 100644 --- a/arch/ia64/kernel/crash.c +++ b/arch/ia64/kernel/crash.c @@ -129,14 +129,13 @@ machine_crash_shutdown(struct pt_regs *pt) * If an INIT is asserted here: * - All receivers might be slaves, since some of cpus could already * be frozen and INIT might be masked on monarch. In this case, - * all slaves will park in while (monarch_cpu == -1) loop before - * DIE_INIT_SLAVE_ENTER that for waiting monarch enters. - * => TBD: freeze all slaves + * all slaves will be frozen soon since kdump_in_progress will let + * them into DIE_INIT_SLAVE_LEAVE. * - One might be a monarch, but INIT rendezvous will fail since * at least this cpu already have INIT masked so it never join * to the rendezvous. In this case, all slaves and monarch will - * be frozen after timeout of the INIT rendezvous. - * => TBD: freeze them without waiting timeout + * be frozen soon with no wait since the INIT rendezvous is skipped + * by kdump_in_progress. */ kdump_smp_send_stop(); /* not all cpu response to IPI, send INIT to freeze them */ @@ -189,6 +188,7 @@ kdump_init_notifier(struct notifier_block *self, unsigned long val, void *data) break; /* fall through */ case DIE_INIT_SLAVE_LEAVE: + case DIE_INIT_MONARCH_ENTER: case DIE_MCA_RENDZVOUS_LEAVE: unw_init_running(kdump_cpu_freeze, NULL); break; @@ -219,7 +219,6 @@ kdump_init_notifier(struct notifier_block *self, unsigned long val, void *data) if (kdump_on_init && (nd->sos->rv_rc != 1)) { if (atomic_inc_return(&kdump_in_progress) != 1) kdump_freeze_monarch = 1; - *(nd->monarch_cpu) = -1; } break; case DIE_INIT_MONARCH_LEAVE: @@ -230,10 +229,8 @@ kdump_init_notifier(struct notifier_block *self, unsigned long val, void *data) case DIE_MCA_MONARCH_LEAVE: /* *(nd->data) indicate if MCA is recoverable */ if (kdump_on_fatal_mca && !(*(nd->data))) { - if (atomic_inc_return(&kdump_in_progress) == 1) { - *(nd->monarch_cpu) = -1; + if (atomic_inc_return(&kdump_in_progress) == 1) machine_kdump_on_init(); - } /* We got fatal MCA while kdump!? No way!! */ } break; diff --git a/arch/ia64/kernel/mca.c b/arch/ia64/kernel/mca.c index 8f33a88..7a34c6c 100644 --- a/arch/ia64/kernel/mca.c +++ b/arch/ia64/kernel/mca.c @@ -1682,14 +1682,25 @@ ia64_init_handler(struct pt_regs *regs, struct switch_stack *sw, if (!sos->monarch) { ia64_mc_info.imi_rendez_checkin[cpu] = IA64_MCA_RENDEZ_CHECKIN_INIT; + +#ifdef CONFIG_KEXEC + while (monarch_cpu == -1 && !atomic_read(&kdump_in_progress)) + udelay(1000); +#else while (monarch_cpu == -1) - cpu_relax(); /* spin until monarch enters */ + cpu_relax(); /* spin until monarch enters */ +#endif NOTIFY_INIT(DIE_INIT_SLAVE_ENTER, regs, (long)&nd, 1); NOTIFY_INIT(DIE_INIT_SLAVE_PROCESS, regs, (long)&nd, 1); +#ifdef CONFIG_KEXEC + while (monarch_cpu != -1 && !atomic_read(&kdump_in_progress)) + udelay(1000); +#else while (monarch_cpu != -1) - cpu_relax(); /* spin until monarch leaves */ + cpu_relax(); /* spin until monarch leaves */ +#endif NOTIFY_INIT(DIE_INIT_SLAVE_LEAVE, regs, (long)&nd, 1); -- 1.6.0 _______________________________________________ kexec mailing list kexec@lists.infradead.org http://lists.infradead.org/mailman/listinfo/kexec