From mboxrd@z Thu Jan 1 00:00:00 1970 From: Krzysztof Kozlowski Subject: Re: [next] Odroid XU3 easily over-heats Date: Mon, 15 Feb 2016 15:36:18 +0900 Message-ID: <56C171E2.90306@samsung.com> References: <56C01963.3080106@gmail.com> <20160215052843.GA6334@vireshk-i7> Mime-Version: 1.0 Content-Type: text/plain; charset=windows-1252 Content-Transfer-Encoding: 7bit Return-path: Received: from mailout2.w1.samsung.com ([210.118.77.12]:59816 "EHLO mailout2.w1.samsung.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1751868AbcBOGg0 (ORCPT ); Mon, 15 Feb 2016 01:36:26 -0500 In-reply-to: <20160215052843.GA6334@vireshk-i7> Sender: linux-samsung-soc-owner@vger.kernel.org List-Id: linux-samsung-soc@vger.kernel.org To: Viresh Kumar , Krzysztof Kozlowski Cc: Lukasz Majewski , Bartlomiej Zolnierkiewicz , "Rafael J. Wysocki" , Linux Kernel Mailing List , Zhang Rui , Eduardo Valentin , linux-arm-kernel@lists.infradead.org, linux-samsung-soc@vger.kernel.org, Javier Martinez Canillas , Kukjin Kim On 15.02.2016 14:28, Viresh Kumar wrote: > On 14-02-16, 15:06, Krzysztof Kozlowski wrote: >> Hi all, >> >> With recent addition of cpufreq-dt support to Exynos5422 >> it is very easy to over-heat the Odroid XU3 leading to critical shutdown: > > How were OPPs passed earlier? I failed to see them in kernel somehow. > You must have been using opp-v1 earlier, right? Till now there was no support for cpufreq for Exynos542x. For some time Bartlomiej was developing these patches - using opp-v1 at beginning. I don't know whether he encountered this before. > > You used boost-freqs then? You used cpufreq as a cooling device then? > >> What about cpufreq as cooling device? Probably that part >> is missing now, right? > > Maybe, how was that earlier though ? Bootloader initialized frequency and regulators to some safe defaults. I think the voltage set by uboot is 1 V (for big and little) which corresponds to ~800 MHz of frequency (I don't know the exact frequency because it is hardcoded as divider value...). However there was no cooling device except fan. This did not change. So the only difference seems to be that now at critical shutdown moment it operates at 1300 MHz / 1.275 V (little) and 1800 MHz / 1.250 V (big). Best regards, Krzysztof From mboxrd@z Thu Jan 1 00:00:00 1970 From: k.kozlowski@samsung.com (Krzysztof Kozlowski) Date: Mon, 15 Feb 2016 15:36:18 +0900 Subject: [next] Odroid XU3 easily over-heats In-Reply-To: <20160215052843.GA6334@vireshk-i7> References: <56C01963.3080106@gmail.com> <20160215052843.GA6334@vireshk-i7> Message-ID: <56C171E2.90306@samsung.com> To: linux-arm-kernel@lists.infradead.org List-Id: linux-arm-kernel.lists.infradead.org On 15.02.2016 14:28, Viresh Kumar wrote: > On 14-02-16, 15:06, Krzysztof Kozlowski wrote: >> Hi all, >> >> With recent addition of cpufreq-dt support to Exynos5422 >> it is very easy to over-heat the Odroid XU3 leading to critical shutdown: > > How were OPPs passed earlier? I failed to see them in kernel somehow. > You must have been using opp-v1 earlier, right? Till now there was no support for cpufreq for Exynos542x. For some time Bartlomiej was developing these patches - using opp-v1 at beginning. I don't know whether he encountered this before. > > You used boost-freqs then? You used cpufreq as a cooling device then? > >> What about cpufreq as cooling device? Probably that part >> is missing now, right? > > Maybe, how was that earlier though ? Bootloader initialized frequency and regulators to some safe defaults. I think the voltage set by uboot is 1 V (for big and little) which corresponds to ~800 MHz of frequency (I don't know the exact frequency because it is hardcoded as divider value...). However there was no cooling device except fan. This did not change. So the only difference seems to be that now at critical shutdown moment it operates at 1300 MHz / 1.275 V (little) and 1800 MHz / 1.250 V (big). Best regards, Krzysztof