From mboxrd@z Thu Jan 1 00:00:00 1970 From: Chris Wilson Subject: Re: [PATCH 3/3] drm/i915: don't rewrite the GTT on resume Date: Mon, 15 Oct 2012 10:42:03 +0100 Message-ID: <453bf0$631iug@azsmga001.ch.intel.com> References: <1350267038-3599-1-git-send-email-jbarnes@virtuousgeek.org> <1350267038-3599-3-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 mga14.intel.com (mga14.intel.com [143.182.124.37]) by gabe.freedesktop.org (Postfix) with ESMTP id CB6139E916 for ; Mon, 15 Oct 2012 02:42:20 -0700 (PDT) In-Reply-To: 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 , intel-gfx@lists.freedesktop.org List-Id: intel-gfx@lists.freedesktop.org On Mon, 15 Oct 2012 09:37:00 +0100, Chris Wilson wrote: > On Sun, 14 Oct 2012 19:10:38 -0700, Jesse Barnes wrote: > > The BIOS shouldn't be touching this memory across suspend/resume, so > > just leave it alone. This saves us ~50ms on resume on my T420. > > > > Signed-off-by: Jesse Barnes > > --- > > drivers/gpu/drm/i915/i915_drv.c | 7 ++++++- > > 1 file changed, 6 insertions(+), 1 deletion(-) > > > > diff --git a/drivers/gpu/drm/i915/i915_drv.c b/drivers/gpu/drm/i915/i915_drv.c > > index 824e3c8..95e2b8b 100644 > > --- a/drivers/gpu/drm/i915/i915_drv.c > > +++ b/drivers/gpu/drm/i915/i915_drv.c > > @@ -119,6 +119,10 @@ module_param_named(i915_enable_ppgtt, i915_enable_ppgtt, int, 0600); > > MODULE_PARM_DESC(i915_enable_ppgtt, > > "Enable PPGTT (default: true)"); > > > > +int i915_needs_gtt_restore __read_mostly = 0; > > +module_param_named(gtt_restore, i915_needs_gtt_restore, int, 0600); > > +MODULE_PARM_DESC(gtt_restore, "Rewrite GTT on resume (default: false)"); > > Wrong way around. This code exists purely because KMS resume was broken > on a number of machines without reinitialising the GTT, ergo this is a > regression. Looking into the history, at the time we wrote the GTT reinit, I believe we did not have a GFX_FLUSH_CNTL in the resume path. Now we do, so I'd be more inclined to believe that we have fixed a bug in the meantime. If so, we only need to be extra careful for gen2. -Chris -- Chris Wilson, Intel Open Source Technology Centre