From patchwork Wed Oct 10 07:29:23 2012 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit X-Patchwork-Submitter: Deepthi Dharwar X-Patchwork-Id: 190571 Return-Path: X-Original-To: patchwork-incoming@ozlabs.org Delivered-To: patchwork-incoming@ozlabs.org Received: from ozlabs.org (localhost [IPv6:::1]) by ozlabs.org (Postfix) with ESMTP id A173D2C058C for ; Wed, 10 Oct 2012 18:30:56 +1100 (EST) Received: from e23smtp08.au.ibm.com (e23smtp08.au.ibm.com [202.81.31.141]) (using TLSv1 with cipher DHE-RSA-AES256-SHA (256/256 bits)) (Client CN "e23smtp08.au.ibm.com", Issuer "GeoTrust SSL CA" (not verified)) by ozlabs.org (Postfix) with ESMTPS id 3F8D22C022F for ; Wed, 10 Oct 2012 18:29:27 +1100 (EST) Received: from /spool/local by e23smtp08.au.ibm.com with IBM ESMTP SMTP Gateway: Authorized Use Only! Violators will be prosecuted for from ; Wed, 10 Oct 2012 17:28:42 +1000 Received: from d23relay05.au.ibm.com (202.81.31.247) by e23smtp08.au.ibm.com (202.81.31.205) with IBM ESMTP SMTP Gateway: Authorized Use Only! Violators will be prosecuted; Wed, 10 Oct 2012 17:28:40 +1000 Received: from d23av02.au.ibm.com (d23av02.au.ibm.com [9.190.235.138]) by d23relay05.au.ibm.com (8.13.8/8.13.8/NCO v10.0) with ESMTP id q9A7JX4p57147564 for ; Wed, 10 Oct 2012 18:19:33 +1100 Received: from d23av02.au.ibm.com (loopback [127.0.0.1]) by d23av02.au.ibm.com (8.14.4/8.13.1/NCO v10.0 AVout) with ESMTP id q9A7TMO0026681 for ; Wed, 10 Oct 2012 18:29:22 +1100 Received: from deepthi.in.ibm.com (deepthi.in.ibm.com [9.124.35.83]) by d23av02.au.ibm.com (8.14.4/8.13.1/NCO v10.0 AVin) with ESMTP id q9A7TLN2026654; Wed, 10 Oct 2012 18:29:22 +1100 From: Deepthi Dharwar Subject: [RESEND PATCH v1 3/3] cpuidle: (POWER) Fix snooze state problem persistant in the current cpuidle design on pseries. To: linux-kernel@vger.kernel.org, linuxppc-dev@lists.ozlabs.org, benh@kernel.crashing.org Date: Wed, 10 Oct 2012 12:59:23 +0530 Message-ID: <20121010072921.29317.67508.stgit@deepthi.in.ibm.com> In-Reply-To: <20121010072859.29317.71472.stgit@deepthi.in.ibm.com> References: <20121010072859.29317.71472.stgit@deepthi.in.ibm.com> User-Agent: StGIT/0.14.3 MIME-Version: 1.0 x-cbid: 12101007-5140-0000-0000-0000022F152B X-BeenThere: linuxppc-dev@lists.ozlabs.org X-Mailman-Version: 2.1.15 Precedence: list List-Id: Linux on PowerPC Developers Mail List List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Errors-To: linuxppc-dev-bounces+patchwork-incoming=ozlabs.org@lists.ozlabs.org Sender: "Linuxppc-dev" Earlier without cpuidle framework on pseries, the native arch idle routine comprised of both snooze and nap states. smt_snooze_delay variable was used to delay the idle process entry to deeper idle state like nap. With the coming of cpuidle, this arch specific idle was replaced by two different idle routines, one for supporting snooze and other for nap. This enabled addition of more low level idle states on pseries in the future. On adopting the generic cpuidle framework for POWER systems, the decision of which idle state to choose from, given a predicted idle time is taken by the menu governor based on target_residency and exit_latency of the idle states. target_residency is the minimum time to be resident in that idle state. Exit_latency is time taken to exit out of idle state. Deeper the idle state, both the target residency and exit latency would be higher. In the current design, smt_snooze_delay is used as target_residency for the snooze state which is incorrect, as it is not the minimum but the maximum duration to be in snooze state. This would result in the governor in taking bad decision, as presently target_residency of nap < target_residency of snooze inspite of nap being deeper idle state. This patch aims to fix this problem by replacing the smt_snooze_delay loop in snooze state, with the need_resched() as the governor is aware of entry and exit of various idle transitions based on which next idle time prediction. The governor is intelligent enough to determine the idle state the needs to be transitioned to and maintains a whole of heuristics including io load, previous idle states predictions etc for the same, based on which idle state entry decision is taken. With this fix, of setting target_residency of snooze to 0 nap to smt_snooze_delay if the predicted idle time is less than smt_snooze_delay (target_residency of nap) value governor would pick snooze state, else nap. This adhers to the previous native idle design. Signed-off-by: Deepthi Dharwar --- arch/powerpc/platforms/pseries/processor_idle.c | 36 ++++++++--------------- 1 file changed, 13 insertions(+), 23 deletions(-) diff --git a/arch/powerpc/platforms/pseries/processor_idle.c b/arch/powerpc/platforms/pseries/processor_idle.c index a32d56d..45d00e5 100644 --- a/arch/powerpc/platforms/pseries/processor_idle.c +++ b/arch/powerpc/platforms/pseries/processor_idle.c @@ -59,32 +59,22 @@ static int snooze_loop(struct cpuidle_device *dev, { unsigned long in_purr; ktime_t kt_before; - unsigned long start_snooze; - long snooze = drv->states[0].target_residency; + int cpu = dev->cpu; idle_loop_prolog(&in_purr, &kt_before); + local_irq_enable(); + set_thread_flag(TIF_POLLING_NRFLAG); - if (snooze) { - start_snooze = get_tb() + snooze * tb_ticks_per_usec; - local_irq_enable(); - set_thread_flag(TIF_POLLING_NRFLAG); - - while ((snooze < 0) || (get_tb() < start_snooze)) { - if (need_resched() || cpu_is_offline(dev->cpu)) - goto out; - ppc64_runlatch_off(); - HMT_low(); - HMT_very_low(); - } - - HMT_medium(); - clear_thread_flag(TIF_POLLING_NRFLAG); - smp_mb(); - local_irq_disable(); + while ((!need_resched()) && cpu_online(cpu)) { + ppc64_runlatch_off(); + HMT_low(); + HMT_very_low(); } -out: HMT_medium(); + clear_thread_flag(TIF_POLLING_NRFLAG); + smp_mb(); + dev->last_residency = (int)idle_loop_epilog(in_purr, kt_before); return index; @@ -165,8 +155,8 @@ static struct cpuidle_state dedicated_states[MAX_IDLE_STATE_COUNT] = { .name = "CEDE", .desc = "CEDE", .flags = CPUIDLE_FLAG_TIME_VALID, - .exit_latency = 1, - .target_residency = 10, + .exit_latency = 10, + .target_residency = 100, .enter = &dedicated_cede_loop }, }; @@ -197,7 +187,7 @@ void update_smt_snooze_delay(int cpu, int residency) dev->states_usage[1].disable = 1; } else if (drv) - drv->states[0].target_residency = residency; + drv->states[1].target_residency = residency; } static int pseries_cpuidle_add_cpu_notifier(struct notifier_block *n,