From: Daniel Vetter Subject: Re: [RFC 00/44] GPU scheduler for i915 driver Date: Mon, 7 Jul 2014 17:57:41 +0200 Message-ID: <20140707155741.GK5821@phenom.ffwll.local> References: <1403803475-16337-1-git-send-email-John.C.Harrison@Intel.com> Mime-Version: 1.0 Content-Type: text/plain; charset="us-ascii" Content-Transfer-Encoding: 7bit Return-path: Received: from mail-wg0-f49.google.com (mail-wg0-f49.google.com [74.125.82.49]) by gabe.freedesktop.org (Postfix) with ESMTP id 7F95489F24 for ; Mon, 7 Jul 2014 08:57:32 -0700 (PDT) Received: by mail-wg0-f49.google.com with SMTP id a1so1242111wgh.8 for ; Mon, 07 Jul 2014 08:57:31 -0700 (PDT) Content-Disposition: inline In-Reply-To: List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Errors-To: intel-gfx-bounces@lists.freedesktop.org Sender: "Intel-gfx" To: Dave Airlie Cc: "intel-gfx@lists.freedesktop.org" List-Id: intel-gfx@lists.freedesktop.org On Fri, Jun 27, 2014 at 06:44:04AM +1000, Dave Airlie wrote: > > > > Implemented a batch buffer submission scheduler for the i915 DRM driver. > > > > While this seems very interesting, you might want to address in the commit msg > or the cover email > > a) why this is needed, > b) any improvements in speed, power consumption or throughput it generates, > i.e. benchmarks. > > also some notes on what hw supports preemption. Also tests to both exercise the rescheduling (i.e. let a high-prio batch compete agains a large pile of low-prio batches) and augmenting our gpu sync tests (from Damien) with that, too. -Daniel -- Daniel Vetter Software Engineer, Intel Corporation +41 (0) 79 365 57 48 - http://blog.ffwll.ch