All of lore.kernel.org
 help / color / mirror / Atom feed
From: Matt Roper <matthew.d.roper@intel.com>
To: ville.syrjala@linux.intel.com
Cc: intel-gfx@lists.freedesktop.org, Paulo Zanoni <paulo.r.zanoni@intel.com>
Subject: Re: [PATCH v2 07/10] drm/i915: Try to make sure cxsr is disabled around plane enable/disable
Date: Wed, 1 Jul 2015 13:38:35 -0700	[thread overview]
Message-ID: <20150701203835.GP5645@intel.com> (raw)
In-Reply-To: <1435778018-9412-1-git-send-email-ville.syrjala@linux.intel.com>

On Wed, Jul 01, 2015 at 10:13:38PM +0300, ville.syrjala@linux.intel.com wrote:
> From: Ville Syrjälä <ville.syrjala@linux.intel.com>
> 
> CxSR (or maxfifo on VLV/CHV) blocks somne changes to the plane control
> register (enable bit at least, not quite sure about the rest). So in
> order to have the plane enable/disable when we want we need to first
> kick the hardware out of cxsr.
> 
> Unfortunateloy this requires some extra vblank waits. For the CxSR
> enable after the plane update we should eventually use an async
> vblank worker, but since we don't have that just do sync vblank
> waits. For the disable case we have no choice but to do it
> synchronously.
> 
> v2: Don't add a spurious intel_pre_plane_update() to crtc disable
> 
> Cc: Paulo Zanoni <paulo.r.zanoni@intel.com>
> Reviewed-by: Clint Taylor <Clinton.A.Taylor@intel.com>
> Tested-by: Clint Taylor <Clinton.A.Taylor@intel.com>
> Signed-off-by: Ville Syrjälä <ville.syrjala@linux.intel.com>
> ---
> Paulo noticed some frontbuffer_bits WARNs from this patch, and sure enough
> I accidentally added another intel_pre_plane_update() to the crtc disable loop.
> I failed to notice because I had commented out the frontbuffer_bits WARNs earlier
> from my tree since they were too noisy.

I can confirm that your v2 fixes the warning spam caused by v1.

Regarding the v2 fix:
Tested-by: Matt Roper <matthew.d.roper@intel.com>

> 
>  drivers/gpu/drm/i915/intel_display.c | 34 +++++++++++++++++++++++++++++-----
>  drivers/gpu/drm/i915/intel_drv.h     |  3 +++
>  drivers/gpu/drm/i915/intel_pm.c      | 11 ++++-------
>  3 files changed, 36 insertions(+), 12 deletions(-)
> 
> diff --git a/drivers/gpu/drm/i915/intel_display.c b/drivers/gpu/drm/i915/intel_display.c
> index d67b5f1..defc4ce 100644
> --- a/drivers/gpu/drm/i915/intel_display.c
> +++ b/drivers/gpu/drm/i915/intel_display.c
> @@ -4716,6 +4716,9 @@ static void intel_post_plane_update(struct intel_crtc *crtc)
>  
>  	intel_frontbuffer_flip(dev, atomic->fb_bits);
>  
> +	if (atomic->disable_cxsr)
> +		crtc->wm.cxsr_allowed = true;
> +
>  	if (crtc->atomic.update_wm_post)
>  		intel_update_watermarks(&crtc->base);
>  
> @@ -4765,6 +4768,11 @@ static void intel_pre_plane_update(struct intel_crtc *crtc)
>  
>  	if (atomic->pre_disable_primary)
>  		intel_pre_disable_primary(&crtc->base);
> +
> +	if (atomic->disable_cxsr) {
> +		crtc->wm.cxsr_allowed = false;
> +		intel_set_memory_cxsr(dev_priv, false);
> +	}
>  }
>  
>  static void intel_crtc_disable_planes(struct drm_crtc *crtc, unsigned plane_mask)
> @@ -11646,12 +11654,26 @@ int intel_plane_atomic_calc_changes(struct drm_crtc_state *crtc_state,
>  			 plane->base.id, was_visible, visible,
>  			 turn_off, turn_on, mode_changed);
>  
> -	if (turn_on)
> +	if (turn_on) {
>  		intel_crtc->atomic.update_wm_pre = true;
> -	else if (turn_off)
> +		/* must disable cxsr around plane enable/disable */
> +		if (plane->type != DRM_PLANE_TYPE_CURSOR) {
> +			intel_crtc->atomic.disable_cxsr = true;
> +			/* to potentially re-enable cxsr */
> +			intel_crtc->atomic.wait_vblank = true;
> +			intel_crtc->atomic.update_wm_post = true;
> +		}
> +	} else if (turn_off) {
>  		intel_crtc->atomic.update_wm_post = true;
> -	else if (intel_wm_need_update(plane, plane_state))
> +		/* must disable cxsr around plane enable/disable */
> +		if (plane->type != DRM_PLANE_TYPE_CURSOR) {
> +			if (is_crtc_enabled)
> +				intel_crtc->atomic.wait_vblank = true;
> +			intel_crtc->atomic.disable_cxsr = true;
> +		}
> +	} else if (intel_wm_need_update(plane, plane_state)) {
>  		intel_crtc->atomic.update_wm_pre = true;
> +	}
>  
>  	if (visible)
>  		intel_crtc->atomic.fb_bits |=
> @@ -11808,8 +11830,8 @@ static int intel_crtc_atomic_check(struct drm_crtc *crtc,
>  	if (pipe_config->quirks & PIPE_CONFIG_QUIRK_INITIAL_PLANES)
>  		intel_crtc_check_initial_planes(crtc, crtc_state);
>  
> -	if (mode_changed)
> -		intel_crtc->atomic.update_wm_post = !crtc_state->active;
> +	if (mode_changed && !crtc_state->active)
> +		intel_crtc->atomic.update_wm_post = true;
>  
>  	if (mode_changed && crtc_state->enable &&
>  	    dev_priv->display.crtc_compute_clock &&
> @@ -14089,6 +14111,8 @@ static void intel_crtc_init(struct drm_device *dev, int pipe)
>  	intel_crtc->cursor_cntl = ~0;
>  	intel_crtc->cursor_size = ~0;
>  
> +	intel_crtc->wm.cxsr_allowed = true;
> +
>  	BUG_ON(pipe >= ARRAY_SIZE(dev_priv->plane_to_crtc_mapping) ||
>  	       dev_priv->plane_to_crtc_mapping[intel_crtc->plane] != NULL);
>  	dev_priv->plane_to_crtc_mapping[intel_crtc->plane] = &intel_crtc->base;
> diff --git a/drivers/gpu/drm/i915/intel_drv.h b/drivers/gpu/drm/i915/intel_drv.h
> index f26a680..4e8d13e 100644
> --- a/drivers/gpu/drm/i915/intel_drv.h
> +++ b/drivers/gpu/drm/i915/intel_drv.h
> @@ -507,6 +507,7 @@ struct intel_crtc_atomic_commit {
>  	/* Sleepable operations to perform before commit */
>  	bool wait_for_flips;
>  	bool disable_fbc;
> +	bool disable_cxsr;
>  	bool pre_disable_primary;
>  	bool update_wm_pre, update_wm_post;
>  	unsigned disabled_planes;
> @@ -565,6 +566,8 @@ struct intel_crtc {
>  		struct intel_pipe_wm active;
>  		/* SKL wm values currently in use */
>  		struct skl_pipe_wm skl_active;
> +		/* allow CxSR on this pipe */
> +		bool cxsr_allowed;
>  	} wm;
>  
>  	int scanline_offset;
> diff --git a/drivers/gpu/drm/i915/intel_pm.c b/drivers/gpu/drm/i915/intel_pm.c
> index c7c90ce..b65817d 100644
> --- a/drivers/gpu/drm/i915/intel_pm.c
> +++ b/drivers/gpu/drm/i915/intel_pm.c
> @@ -335,6 +335,7 @@ void intel_set_memory_cxsr(struct drm_i915_private *dev_priv, bool enable)
>  	if (IS_VALLEYVIEW(dev)) {
>  		I915_WRITE(FW_BLC_SELF_VLV, enable ? FW_CSPWRDWNEN : 0);
>  		POSTING_READ(FW_BLC_SELF_VLV);
> +		dev_priv->wm.vlv.cxsr = enable;
>  	} else if (IS_G4X(dev) || IS_CRESTLINE(dev)) {
>  		I915_WRITE(FW_BLC_SELF, enable ? FW_BLC_SELF_EN : 0);
>  		POSTING_READ(FW_BLC_SELF);
> @@ -1116,7 +1117,7 @@ static void vlv_compute_wm(struct intel_crtc *crtc)
>  
>  	memset(wm_state, 0, sizeof(*wm_state));
>  
> -	wm_state->cxsr = crtc->pipe != PIPE_C;
> +	wm_state->cxsr = crtc->pipe != PIPE_C && crtc->wm.cxsr_allowed;
>  	if (IS_CHERRYVIEW(dev))
>  		wm_state->num_levels = CHV_WM_NUM_LEVELS;
>  	else
> @@ -1369,10 +1370,8 @@ static void vlv_update_wm(struct drm_crtc *crtc)
>  	    dev_priv->wm.vlv.level >= VLV_WM_LEVEL_PM5)
>  		chv_set_memory_pm5(dev_priv, false);
>  
> -	if (!wm.cxsr && dev_priv->wm.vlv.cxsr) {
> +	if (!wm.cxsr && dev_priv->wm.vlv.cxsr)
>  		intel_set_memory_cxsr(dev_priv, false);
> -		intel_wait_for_vblank(dev, pipe);
> -	}
>  
>  	/* FIXME should be part of crtc atomic commit */
>  	vlv_pipe_set_fifo_size(intel_crtc);
> @@ -1385,10 +1384,8 @@ static void vlv_update_wm(struct drm_crtc *crtc)
>  		      wm.pipe[pipe].sprite[0], wm.pipe[pipe].sprite[1],
>  		      wm.sr.plane, wm.sr.cursor, wm.level, wm.cxsr);
>  
> -	if (wm.cxsr && !dev_priv->wm.vlv.cxsr) {
> -		intel_wait_for_vblank(dev, pipe);
> +	if (wm.cxsr && !dev_priv->wm.vlv.cxsr)
>  		intel_set_memory_cxsr(dev_priv, true);
> -	}
>  
>  	if (wm.level >= VLV_WM_LEVEL_PM5 &&
>  	    dev_priv->wm.vlv.level < VLV_WM_LEVEL_PM5)
> -- 
> 2.3.6
> 
> _______________________________________________
> Intel-gfx mailing list
> Intel-gfx@lists.freedesktop.org
> http://lists.freedesktop.org/mailman/listinfo/intel-gfx

-- 
Matt Roper
Graphics Software Engineer
IoTG Platform Enabling & Development
Intel Corporation
(916) 356-2795
_______________________________________________
Intel-gfx mailing list
Intel-gfx@lists.freedesktop.org
http://lists.freedesktop.org/mailman/listinfo/intel-gfx

  parent reply	other threads:[~2015-07-01 20:38 UTC|newest]

Thread overview: 29+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2015-06-24 19:00 [PATCH 00/10] drm/i915: Another WM rewrite to enable DDR DVFS on CHV ville.syrjala
2015-06-24 19:00 ` [PATCH 01/10] drm/i915: POSTING_READ() in intel_set_memory_cxsr() ville.syrjala
2015-06-26 20:22   ` Clint Taylor
2015-06-24 19:00 ` [PATCH 02/10] drm/i915: Split atomic wm update to pre and post variants ville.syrjala
2015-06-26 20:22   ` Clint Taylor
2015-06-24 19:00 ` [PATCH 03/10] drm/i915: Read wm values from hardware at init on CHV ville.syrjala
2015-06-26 20:23   ` Clint Taylor
2015-06-24 19:00 ` [PATCH 04/10] drm/i915: CHV DDR DVFS support and another watermark rewrite ville.syrjala
2015-06-26 17:56   ` Clint Taylor
2015-06-26 19:48     ` Ville Syrjälä
2015-06-26 20:21       ` Clint Taylor
2015-06-29  8:03       ` Jani Nikula
2015-06-29  8:54         ` Daniel Vetter
2015-06-24 19:00 ` [PATCH 05/10] drm/i915: Compute display FIFO split dynamically for CHV ville.syrjala
2015-06-26 20:23   ` Clint Taylor
2015-06-24 19:00 ` [PATCH 06/10] drm/i915: Use the memory latency based WM computation on VLV too ville.syrjala
2015-06-26 20:23   ` Clint Taylor
2015-06-24 19:00 ` [PATCH 07/10] drm/i915: Try to make sure cxsr is disabled around plane enable/disable ville.syrjala
2015-06-26 20:23   ` Clint Taylor
2015-07-01 19:13   ` [PATCH v2 " ville.syrjala
2015-07-01 19:36     ` Paulo Zanoni
2015-07-01 20:38     ` Matt Roper [this message]
2015-06-24 19:00 ` [PATCH 08/10] drm/i915: Don't do PM5/DDR DVFS with multiple pipes ville.syrjala
2015-06-26 20:23   ` Clint Taylor
2015-06-24 19:00 ` [PATCH 09/10] drm/i915: Add debugfs knobs for VLVCHV memory latency values ville.syrjala
2015-06-26 20:24   ` Clint Taylor
2015-06-24 19:00 ` [PATCH 10/10] drm/i915: Zero unused WM1 watermarks on VLV/CHV ville.syrjala
2015-06-26 20:24   ` Clint Taylor
2015-06-29  9:00     ` 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=20150701203835.GP5645@intel.com \
    --to=matthew.d.roper@intel.com \
    --cc=intel-gfx@lists.freedesktop.org \
    --cc=paulo.r.zanoni@intel.com \
    --cc=ville.syrjala@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.