All of lore.kernel.org
 help / color / mirror / Atom feed
From: Daniel Vetter <daniel@ffwll.ch>
To: Jani Nikula <jani.nikula@linux.intel.com>
Cc: Daniel Vetter <daniel.vetter@ffwll.ch>, intel-gfx@lists.freedesktop.org
Subject: Re: [PATCH 14/15] drm/i915: Sanitize watermarks after hardware state readout
Date: Thu, 1 Oct 2015 18:12:52 +0200	[thread overview]
Message-ID: <20151001161252.GA3383@phenom.ffwll.local> (raw)
In-Reply-To: <874miabshd.fsf@intel.com>

On Thu, Oct 01, 2015 at 04:58:38PM +0300, Jani Nikula wrote:
> On Fri, 25 Sep 2015, Matt Roper <matthew.d.roper@intel.com> wrote:
> > Although we can do a good job of reading out hardware state, the
> > graphics firmware may have programmed the watermarks in a creative way
> > that doesn't match how i915 would have chosen to program them.  We
> > shouldn't trust the firmware's watermark programming, but should rather
> > re-calculate how we think WM's should be programmed and then shove those
> > values into the hardware.
> >
> > We can do this pretty easily by creating a dummy top-level state,
> > running it through the check process to calculate all the values, and
> > then just programming the watermarks for each CRTC.
> >
> > Cc: Maarten Lankhorst <maarten.lankhorst@linux.intel.com>
> > Signed-off-by: Matt Roper <matthew.d.roper@intel.com>
> 
> Bisect tells me this patch, i.e.
> 
> commit c103f2b83dfa3b6fb6b5819ae0362fee6cf4b242
> Author: Matt Roper <matthew.d.roper@intel.com>
> Date:   Thu Sep 24 15:53:19 2015 -0700
> 
>     drm/i915: Sanitize watermarks after hardware state readout
> 
> gives me black screen on a BDW NUC pciid=0x1616.
> 
> Daniel, please drop it.

Done, plus the next patch due to conflicts. Matt, I'll reapply them as
soon as we have a fix for this boot issue.
-Daniel

> 
> BR,
> Jani.
> 
> 
> > ---
> > Maarten, does this solve the problem you were seeing on Ironlake?  You
> > indicated that your firmware had sprite watermarks programmed even though
> > sprites themselves were off and I don't have any kind of system that can
> > reproduce that setup.  I'm hoping this will patch will do the trick.
> >
> >  drivers/gpu/drm/i915/i915_drv.h      |  1 +
> >  drivers/gpu/drm/i915/intel_display.c | 51 ++++++++++++++++++++++++++++++++++++
> >  drivers/gpu/drm/i915/intel_pm.c      | 14 +++++-----
> >  3 files changed, 60 insertions(+), 6 deletions(-)
> >
> > diff --git a/drivers/gpu/drm/i915/i915_drv.h b/drivers/gpu/drm/i915/i915_drv.h
> > index 8b7c8f9..a9bac1e 100644
> > --- a/drivers/gpu/drm/i915/i915_drv.h
> > +++ b/drivers/gpu/drm/i915/i915_drv.h
> > @@ -629,6 +629,7 @@ struct drm_i915_display_funcs {
> >  			  struct dpll *best_clock);
> >  	int (*compute_pipe_wm)(struct intel_crtc *crtc,
> >  			       struct drm_atomic_state *state);
> > +	void (*program_watermarks)(struct intel_crtc_state *cstate);
> >  	void (*update_wm)(struct drm_crtc *crtc);
> >  	int (*modeset_calc_cdclk)(struct drm_atomic_state *state);
> >  	void (*modeset_commit_cdclk)(struct drm_atomic_state *state);
> > diff --git a/drivers/gpu/drm/i915/intel_display.c b/drivers/gpu/drm/i915/intel_display.c
> > index e2a0777..0c3783c 100644
> > --- a/drivers/gpu/drm/i915/intel_display.c
> > +++ b/drivers/gpu/drm/i915/intel_display.c
> > @@ -15310,6 +15310,54 @@ static void intel_modeset_readout_hw_state(struct drm_device *dev)
> >  	}
> >  }
> >  
> > +/*
> > + * Calculate what we think the watermarks should be for the state we've read
> > + * out of the hardware and then immediately program those watermarks so that
> > + * we ensure the hardware settings match our internal state.
> > + */
> > +static void sanitize_watermarks(struct drm_device *dev)
> > +{
> > +	struct drm_i915_private *dev_priv = to_i915(dev);
> > +	struct drm_atomic_state *state;
> > +	struct drm_crtc *crtc;
> > +	struct drm_crtc_state *cstate;
> > +	int ret;
> > +	int i;
> > +
> > +	/* Only supported on platforms that use atomic watermark design */
> > +	if (!dev_priv->display.program_watermarks)
> > +		return;
> > +
> > +	/*
> > +	 * Calculate what we think WM's should be by creating a dummy state and
> > +	 * running it through the atomic check code.
> > +	 */
> > +	state = drm_atomic_helper_duplicate_state(dev,
> > +						  dev->mode_config.acquire_ctx);
> > +	if (WARN_ON(IS_ERR(state)))
> > +		return;
> > +
> > +	ret = intel_atomic_check(dev, state);
> > +	if (ret) {
> > +		/*
> > +		 * Just give up and leave watermarks untouched if we get an
> > +		 * error back from 'check'
> > +		 */
> > +		DRM_DEBUG_KMS("Could not determine valid watermarks for inherited state\n");
> > +		return;
> > +	}
> > +
> > +	/* Write calculated watermark values back */
> > +	to_i915(dev)->wm.config = to_intel_atomic_state(state)->wm_config;
> > +	for_each_crtc_in_state(state, crtc, cstate, i) {
> > +		struct intel_crtc_state *cs = to_intel_crtc_state(cstate);
> > +
> > +		dev_priv->display.program_watermarks(cs);
> > +	}
> > +
> > +	drm_atomic_state_free(state);
> > +}
> > +
> >  /* Scan out the current hw modeset state,
> >   * and sanitizes it to the current state
> >   */
> > @@ -15365,6 +15413,9 @@ intel_modeset_setup_hw_state(struct drm_device *dev)
> >  			modeset_put_power_domains(dev_priv, put_domains);
> >  	}
> >  	intel_display_set_init_power(dev_priv, false);
> > +
> > +	/* Make sure hardware watermarks really match the state we read out */
> > +	sanitize_watermarks(dev);
> >  }
> >  
> >  void intel_display_resume(struct drm_device *dev)
> > diff --git a/drivers/gpu/drm/i915/intel_pm.c b/drivers/gpu/drm/i915/intel_pm.c
> > index f3652bb..988893e 100644
> > --- a/drivers/gpu/drm/i915/intel_pm.c
> > +++ b/drivers/gpu/drm/i915/intel_pm.c
> > @@ -3664,15 +3664,19 @@ static void skl_update_wm(struct drm_crtc *crtc)
> >  	dev_priv->wm.skl_hw = *results;
> >  }
> >  
> > -static void ilk_program_watermarks(struct drm_i915_private *dev_priv)
> > +static void ilk_program_watermarks(struct intel_crtc_state *cstate)
> >  {
> > -	struct drm_device *dev = dev_priv->dev;
> > +	struct drm_crtc *crtc = cstate->base.crtc;
> > +	struct drm_device *dev = crtc->dev;
> > +	struct drm_i915_private *dev_priv = to_i915(dev);
> >  	struct intel_pipe_wm lp_wm_1_2 = {}, lp_wm_5_6 = {}, *best_lp_wm;
> >  	struct ilk_wm_maximums max;
> >  	struct intel_wm_config *config = &dev_priv->wm.config;
> >  	struct ilk_wm_values results = {};
> >  	enum intel_ddb_partitioning partitioning;
> >  
> > +	to_intel_crtc(crtc)->wm.active.ilk = cstate->wm.optimal.ilk;
> > +
> >  	ilk_compute_wm_maximums(dev, 1, config, INTEL_DDB_PART_1_2, &max);
> >  	ilk_wm_merge(dev, config, &max, &lp_wm_1_2);
> >  
> > @@ -3697,7 +3701,6 @@ static void ilk_program_watermarks(struct drm_i915_private *dev_priv)
> >  
> >  static void ilk_update_wm(struct drm_crtc *crtc)
> >  {
> > -	struct drm_i915_private *dev_priv = to_i915(crtc->dev);
> >  	struct intel_crtc *intel_crtc = to_intel_crtc(crtc);
> >  	struct intel_crtc_state *cstate = to_intel_crtc_state(crtc->state);
> >  
> > @@ -3715,9 +3718,7 @@ static void ilk_update_wm(struct drm_crtc *crtc)
> >  		intel_wait_for_vblank(crtc->dev, intel_crtc->pipe);
> >  	}
> >  
> > -	intel_crtc->wm.active.ilk = cstate->wm.optimal.ilk;
> > -
> > -	ilk_program_watermarks(dev_priv);
> > +	ilk_program_watermarks(cstate);
> >  }
> >  
> >  static void skl_pipe_wm_active_state(uint32_t val,
> > @@ -7051,6 +7052,7 @@ void intel_init_pm(struct drm_device *dev)
> >  		     dev_priv->wm.spr_latency[0] && dev_priv->wm.cur_latency[0])) {
> >  			dev_priv->display.update_wm = ilk_update_wm;
> >  			dev_priv->display.compute_pipe_wm = ilk_compute_pipe_wm;
> > +			dev_priv->display.program_watermarks = ilk_program_watermarks;
> >  		} else {
> >  			DRM_DEBUG_KMS("Failed to read display plane latency. "
> >  				      "Disable CxSR\n");
> > -- 
> > 2.1.4
> >
> > _______________________________________________
> > Intel-gfx mailing list
> > Intel-gfx@lists.freedesktop.org
> > http://lists.freedesktop.org/mailman/listinfo/intel-gfx
> 
> -- 
> Jani Nikula, Intel Open Source Technology Center

-- 
Daniel Vetter
Software Engineer, Intel Corporation
http://blog.ffwll.ch
_______________________________________________
Intel-gfx mailing list
Intel-gfx@lists.freedesktop.org
http://lists.freedesktop.org/mailman/listinfo/intel-gfx

  reply	other threads:[~2015-10-01 16:09 UTC|newest]

Thread overview: 29+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2015-09-24 22:53 [PATCH 00/15] Atomic watermark updates (v5) Matt Roper
2015-09-24 22:53 ` [PATCH 01/15] drm/i915: Drop redundant watermark programming Matt Roper
2015-09-24 22:53 ` [PATCH 02/15] drm/i915: Eliminate usage of plane_wm_parameters from ILK-style WM code (v2) Matt Roper
2016-01-05 12:49   ` Flicker caused by "drm/i915: Eliminate usage of plane_wm_parameters from ILK-style WM code (v2)" Jan Niehusmann
2016-01-05 21:58     ` Matt Roper
2016-01-06  0:46       ` Jan Niehusmann
2015-09-24 22:53 ` [PATCH 03/15] drm/i915: Eliminate usage of pipe_wm_parameters from ILK-style WM (v2) Matt Roper
2015-09-24 22:53 ` [PATCH 04/15] drm/i915: Determine I915_MAX_PLANES from plane enum Matt Roper
2015-09-24 22:53 ` [PATCH 05/15] drm/i915/skl: Simplify wm structures slightly (v2) Matt Roper
2015-09-30 15:13   ` Daniel Vetter
2015-09-24 22:53 ` [PATCH 06/15] drm/i915/skl: Eliminate usage of pipe_wm_parameters from SKL-style WM (v3) Matt Roper
2015-09-24 22:53 ` [PATCH 07/15] drm/i915/ivb: Move WaCxSRDisabledForSpriteScaling w/a to atomic check Matt Roper
2015-09-24 22:53 ` [PATCH 08/15] drm/i915: Drop intel_update_sprite_watermarks Matt Roper
2015-09-24 22:53 ` [PATCH 09/15] drm/i915: Refactor ilk_update_wm (v3) Matt Roper
2015-09-24 22:53 ` [PATCH 10/15] drm/i915: Calculate pipe watermarks into CRTC state (v3) Matt Roper
2015-09-24 22:53 ` [PATCH 11/15] drm/i915: Calculate ILK-style watermarks during atomic check (v3) Matt Roper
2015-09-24 22:53 ` [PATCH 12/15] drm/i915: Don't set plane visible during HW readout if CRTC is off Matt Roper
2015-09-24 22:53 ` [PATCH 13/15] drm/i915: Calculate watermark configuration during atomic check (v2) Matt Roper
2015-09-24 22:53 ` [PATCH 14/15] drm/i915: Sanitize watermarks after hardware state readout Matt Roper
2015-10-01 13:58   ` Jani Nikula
2015-10-01 16:12     ` Daniel Vetter [this message]
2015-10-01 16:53     ` [PATCH] drm/i915: Convert hsw_compute_linetime_wm to use in-flight state Matt Roper
2015-10-06 14:34       ` Jani Nikula
2015-09-24 22:53 ` [PATCH 15/15] drm/i915: Add two-stage ILK-style watermark programming (v5) Matt Roper
2015-09-30 15:20 ` [PATCH 00/15] Atomic watermark updates (v5) Daniel Vetter
2015-09-30 22:21   ` Zanoni, Paulo R
2015-10-01 23:03     ` [PATCH] fixup! drm/i915/skl: Eliminate usage of pipe_wm_parameters from SKL-style WM (v3) Matt Roper
2015-10-02 18:43       ` Zanoni, Paulo R
2015-10-06 10:13         ` [PATCH] fixup! drm/i915/skl: Eliminate usage of pipe_wm_parameters from SKL-style WM (v3) [regression] Daniel Vetter

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=20151001161252.GA3383@phenom.ffwll.local \
    --to=daniel@ffwll.ch \
    --cc=daniel.vetter@ffwll.ch \
    --cc=intel-gfx@lists.freedesktop.org \
    --cc=jani.nikula@linux.intel.com \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.