From mboxrd@z Thu Jan 1 00:00:00 1970 From: Ben Widawsky Subject: Re: [PATCH] drm/i915: optionally check GTT checksum across suspend/resume Date: Wed, 12 Sep 2012 23:02:43 -0700 Message-ID: <20120912230243.1b8e6679@bwidawsk.net> References: <1347481630-2986-1-git-send-email-jbarnes@virtuousgeek.org> Mime-Version: 1.0 Content-Type: text/plain; charset="us-ascii" Content-Transfer-Encoding: 7bit Return-path: Received: from shiva.chad-versace.us (209-20-75-48.static.cloud-ips.com [209.20.75.48]) by gabe.freedesktop.org (Postfix) with ESMTP id 214D79E7D5 for ; Wed, 12 Sep 2012 23:02:54 -0700 (PDT) In-Reply-To: <1347481630-2986-1-git-send-email-jbarnes@virtuousgeek.org> List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Sender: intel-gfx-bounces+gcfxdi-intel-gfx=m.gmane.org@lists.freedesktop.org Errors-To: intel-gfx-bounces+gcfxdi-intel-gfx=m.gmane.org@lists.freedesktop.org To: Jesse Barnes Cc: intel-gfx@lists.freedesktop.org List-Id: intel-gfx@lists.freedesktop.org On Wed, 12 Sep 2012 13:27:10 -0700 Jesse Barnes wrote: > This adds a module parameter, gtt_suspend_verify, that will enable a > simple GTT checksum across suspend and resume. > > We currently spend a good amount of time (>20ms) clearing all the GTT > PTEs at resume time, even though this may not be necessary on some > machines. This debug feature is intended to help determine whether > newer machines don't need the unconditional GTT clear on resume. > > Signed-off-by: Jesse Barnes It's a cool idea, but I'd vote against adding a module parameter to verify it. Daniel's trees should be well enough tested that you could just disable it on IVB+ or whatever and let it go through QA and developer testing. Aside from that, a debugfs flag to toggle it would probably be a little better. Also, as a bikeshed you could probably get a much better detection with a CRC or something similar. I dunno, just a thought, -- Ben Widawsky, Intel Open Source Technology Center