From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1754282AbcEPOXf (ORCPT ); Mon, 16 May 2016 10:23:35 -0400 Received: from devils.ext.ti.com ([198.47.26.153]:35966 "EHLO devils.ext.ti.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1753936AbcEPOXc (ORCPT ); Mon, 16 May 2016 10:23:32 -0400 Subject: Re: Can not play audio with screen off To: Pavel Machek , , , kernel list , linux-arm-kernel , , , , , , , References: <20160515193106.GA18685@amd> From: Nishanth Menon Message-ID: <5739D7B6.7060102@ti.com> Date: Mon, 16 May 2016 09:22:46 -0500 User-Agent: Mozilla/5.0 (X11; Linux x86_64; rv:38.0) Gecko/20100101 Thunderbird/38.6.0 MIME-Version: 1.0 In-Reply-To: <20160515193106.GA18685@amd> Content-Type: text/plain; charset="windows-1252" Content-Transfer-Encoding: 7bit Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org On 05/15/2016 02:31 PM, Pavel Machek wrote: > Hi! > > I'm playing some more with N900, this time I'd like to use it as a > alarm clock... > > And now I realized I can not, because system performance is so > horrible with screen off that mpg123 + pulseaudio skips horribly. This > is without tweaking serial parameters, so CPU is definitely not going > into the idle/off modes. > > Loading CPU fixes the issue (I used cat /dev/zero > /dev/null). > > Hmm. And it seems to work better if I ssh in and launch mpg123 > remotely. Weird. > > Kernel 4.5-rc0. Will try on different ones. > > Any ideas? Is there interface to tell kernel that better performance > is needed for a while? > pm_qos cpu_dma_latency? Drivers will have to be properly instrumented with appropriate latency values -> this will control the depth in C states that the system will attempt. NOTE: it is a dual edged sword, bad configurations can result in none of the C-states being entered. -- Regards, Nishanth Menon From mboxrd@z Thu Jan 1 00:00:00 1970 From: Nishanth Menon Subject: Re: Can not play audio with screen off Date: Mon, 16 May 2016 09:22:46 -0500 Message-ID: <5739D7B6.7060102@ti.com> References: <20160515193106.GA18685@amd> Mime-Version: 1.0 Content-Type: text/plain; charset="windows-1252" Content-Transfer-Encoding: 7bit Return-path: In-Reply-To: <20160515193106.GA18685@amd> Sender: linux-kernel-owner@vger.kernel.org To: Pavel Machek , pali.rohar@gmail.com, sre@kernel.org, kernel list , linux-arm-kernel , linux-omap@vger.kernel.org, tony@atomide.com, khilman@kernel.org, aaro.koskinen@iki.fi, ivo.g.dimitrov.75@gmail.com, patrikbachan@gmail.com, serge@hallyn.com List-Id: linux-omap@vger.kernel.org On 05/15/2016 02:31 PM, Pavel Machek wrote: > Hi! > > I'm playing some more with N900, this time I'd like to use it as a > alarm clock... > > And now I realized I can not, because system performance is so > horrible with screen off that mpg123 + pulseaudio skips horribly. This > is without tweaking serial parameters, so CPU is definitely not going > into the idle/off modes. > > Loading CPU fixes the issue (I used cat /dev/zero > /dev/null). > > Hmm. And it seems to work better if I ssh in and launch mpg123 > remotely. Weird. > > Kernel 4.5-rc0. Will try on different ones. > > Any ideas? Is there interface to tell kernel that better performance > is needed for a while? > pm_qos cpu_dma_latency? Drivers will have to be properly instrumented with appropriate latency values -> this will control the depth in C states that the system will attempt. NOTE: it is a dual edged sword, bad configurations can result in none of the C-states being entered. -- Regards, Nishanth Menon From mboxrd@z Thu Jan 1 00:00:00 1970 From: nm@ti.com (Nishanth Menon) Date: Mon, 16 May 2016 09:22:46 -0500 Subject: Can not play audio with screen off In-Reply-To: <20160515193106.GA18685@amd> References: <20160515193106.GA18685@amd> Message-ID: <5739D7B6.7060102@ti.com> To: linux-arm-kernel@lists.infradead.org List-Id: linux-arm-kernel.lists.infradead.org On 05/15/2016 02:31 PM, Pavel Machek wrote: > Hi! > > I'm playing some more with N900, this time I'd like to use it as a > alarm clock... > > And now I realized I can not, because system performance is so > horrible with screen off that mpg123 + pulseaudio skips horribly. This > is without tweaking serial parameters, so CPU is definitely not going > into the idle/off modes. > > Loading CPU fixes the issue (I used cat /dev/zero > /dev/null). > > Hmm. And it seems to work better if I ssh in and launch mpg123 > remotely. Weird. > > Kernel 4.5-rc0. Will try on different ones. > > Any ideas? Is there interface to tell kernel that better performance > is needed for a while? > pm_qos cpu_dma_latency? Drivers will have to be properly instrumented with appropriate latency values -> this will control the depth in C states that the system will attempt. NOTE: it is a dual edged sword, bad configurations can result in none of the C-states being entered. -- Regards, Nishanth Menon