From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1753115AbaKJVUl (ORCPT ); Mon, 10 Nov 2014 16:20:41 -0500 Received: from mail-qg0-f43.google.com ([209.85.192.43]:38467 "EHLO mail-qg0-f43.google.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1752662AbaKJVUf (ORCPT ); Mon, 10 Nov 2014 16:20:35 -0500 MIME-Version: 1.0 In-Reply-To: References: <1415471865-1893-1-git-send-email-oded.gabbay@amd.com> <54606AF7.8070506@amd.com> Date: Mon, 10 Nov 2014 16:20:34 -0500 Message-ID: Subject: Re: [PATCH v5 00/24] AMDKFD Kernel Driver From: Alex Deucher To: Dave Airlie Cc: Oded Gabbay , Alex Deucher , dri-devel , LKML Content-Type: text/plain; charset=UTF-8 Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org On Mon, Nov 10, 2014 at 4:16 PM, Dave Airlie wrote: >>> How does amdkfd interfact with runtime pm on the radeon driver? I'd expect >>> some calls to the runtime get/put functions in some places. >>> >>> Dave. >>> >> Hi Dave, >> Per Jerome's request from the first time he saw the driver, we removed all >> the "register bashing" code from amdkfd and moved them to the interface file >> between amdkfd and radeon, which is part of the radeon driver. >> See "[PATCH v5 05/24] drm/radeon: Add radeon <--> amdkfd interface" for the >> implementation of that interface file. > > So in GPU offload mode, how does amdkfd wake the GPU up to execute > stuff? > > GPU offload is where the radeon GPU is a secondary GPU and we power it > down automatically. > > Granted I think our powerdown should probably check if the GPU is executing > anything a bit more, but I don't think amdkfd can even wake it up to submit > work at the moment. Currently we never power down APUs (only dGPUs) and kfd only supports APUs so they should always be available. It will definitely need to be handled properly if we eventually support dGPUs with the kfd. Alex