From mboxrd@z Thu Jan 1 00:00:00 1970 From: Sylvain BERTRAND Subject: atombios powerplay cac table Date: Thu, 7 Nov 2013 04:57:21 +0100 Message-ID: <20131107035721.GA4330@freedom> Mime-Version: 1.0 Content-Type: text/plain; charset="us-ascii" Content-Transfer-Encoding: 7bit Return-path: Received: from smtpauth-vit.online.net (smtpauth-vit.online.net [88.190.253.76]) by gabe.freedesktop.org (Postfix) with ESMTP id 06E7C383C9 for ; Wed, 6 Nov 2013 19:57:23 -0800 (PST) Received: from [88.188.175.165] (helo=localhost) by smtpauth-vit.online.net with esmtpsa (TLS1.2:RSA_AES_128_CBC_SHA1:128) (Exim 4.80) (envelope-from ) id 1VeGig-0007Qz-57 for dri-devel@lists.freedesktop.org; Thu, 07 Nov 2013 04:57:22 +0100 Content-Disposition: inline List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Sender: dri-devel-bounces@lists.freedesktop.org Errors-To: dri-devel-bounces@lists.freedesktop.org To: dri-devel@lists.freedesktop.org List-Id: dri-devel@lists.freedesktop.org Hi, Which GPU generations/families do use the "new" powerplay CAC table, defined when the atombios platform capabilities have the ATOM_PP_PLATFORM_CAP_NEW_CAC_VOLTAGE flag? ---- On drm-next-3.13 branch, commit 70471860ff9f335c60c004d42ebd48945bfa5403, I noticed the following: For the "new" power CAC table, there is a relationship with the table defining the link between vddc and the engine clock. My tahiti based board does use the "old" cac table, but I noticed the use of a virtual value (0xff??) for vddc in the vddc dependency on engine clock table which is not translated to a real voltage value. This is found in si_dpm.c, si_get_std_voltage_value function. Would that would explain why there are 2 lookup loops in si_get_std_voltage_value function? (one looking for the exact value, the other one seems to be a trick to deal with untranslated virtual voltage). regards, -- Sylvain