public inbox for intel-gfx@lists.freedesktop.org
 help / color / mirror / Atom feed
From: "Ville Syrjälä" <ville.syrjala@linux.intel.com>
To: intel-gfx@lists.freedesktop.org
Subject: Re: [PATCH 4/9] drm/i915: Allocate enough DDB for the cursor
Date: Tue, 19 Mar 2019 12:51:23 +0200	[thread overview]
Message-ID: <20190319105123.GU3888@intel.com> (raw)
In-Reply-To: <20190312205844.6339-5-ville.syrjala@linux.intel.com>

On Tue, Mar 12, 2019 at 10:58:39PM +0200, Ville Syrjala wrote:
> From: Ville Syrjälä <ville.syrjala@linux.intel.com>
> 
> Currently we just assume that 32 or 8 blocks of ddb is sufficient
> for the cursor. The 32 might be, but the 8 is certainly not. The
> minimum we need is at least what level 0 watermarks need, but that
> is a bit restrictive, so instead let's calculate what level 7
> would need for a 256x256 cursor. We'll use that to determine the
> fixed ddb allocation for the cursor. This way the cursor will never
> be responsible for missing out on deeper power saving states.
> 
> Cc: Neel Desai <neel.desai@intel.com>
> Signed-off-by: Ville Syrjälä <ville.syrjala@linux.intel.com>
> ---
>  drivers/gpu/drm/i915/intel_pm.c | 33 ++++++++++++++++++++++++++++-----
>  1 file changed, 28 insertions(+), 5 deletions(-)
> 
> diff --git a/drivers/gpu/drm/i915/intel_pm.c b/drivers/gpu/drm/i915/intel_pm.c
> index 9e7b4412f7a8..ae9a0ce4dcb9 100644
> --- a/drivers/gpu/drm/i915/intel_pm.c
> +++ b/drivers/gpu/drm/i915/intel_pm.c
> @@ -3924,12 +3924,35 @@ skl_ddb_get_pipe_allocation_limits(struct drm_i915_private *dev_priv,
>  	alloc->end = ddb_size * (width_before_pipe + pipe_width) / total_width;
>  }
>  
> -static unsigned int skl_cursor_allocation(int num_active)
> +static int skl_compute_wm_params(const struct intel_crtc_state *crtc_state,
> +				 int width, const struct drm_format_info *format,
> +				 u64 modifier, unsigned int rotation,
> +				 u32 plane_pixel_rate, struct skl_wm_params *wp,
> +				 int color_plane);
> +static void skl_compute_plane_wm(const struct intel_crtc_state *cstate,
> +				 int level,
> +				 const struct skl_wm_params *wp,
> +				 const struct skl_wm_level *result_prev,
> +				 struct skl_wm_level *result /* out */);
> +
> +static unsigned int
> +skl_cursor_allocation(const struct intel_crtc_state *crtc_state,
> +		      int num_active)
>  {
> -	if (num_active == 1)
> -		return 32;
> +	struct skl_wm_level wm = {};
> +	struct skl_wm_params wp;
> +	int ret;
> +
> +	ret = skl_compute_wm_params(crtc_state, 256,
> +				    drm_format_info(DRM_FORMAT_ARGB8888),
> +				    DRM_FORMAT_MOD_LINEAR,
> +				    DRM_MODE_ROTATE_0,
> +				    crtc_state->pixel_rate, &wp, 0);
> +	WARN_ON(ret);
> +
> +	skl_compute_plane_wm(crtc_state, 7, &wp, &wm, &wm);

Hmm. This borked if latency[7]==0. I guess I actually need to loop
here...

>  
> -	return 8;
> +	return max_t(int, num_active == 1 ? 32 : 8, wm.min_ddb_alloc);
>  }
>  
>  static void skl_ddb_entry_init_from_hw(struct drm_i915_private *dev_priv,
> @@ -4354,7 +4377,7 @@ skl_allocate_pipe_ddb(struct intel_crtc_state *cstate,
>  		return 0;
>  
>  	/* Allocate fixed number of blocks for cursor. */
> -	total[PLANE_CURSOR] = skl_cursor_allocation(num_active);
> +	total[PLANE_CURSOR] = skl_cursor_allocation(cstate, num_active);
>  	alloc_size -= total[PLANE_CURSOR];
>  	cstate->wm.skl.plane_ddb_y[PLANE_CURSOR].start =
>  		alloc->end - total[PLANE_CURSOR];
> -- 
> 2.19.2

-- 
Ville Syrjälä
Intel
_______________________________________________
Intel-gfx mailing list
Intel-gfx@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/intel-gfx

  reply	other threads:[~2019-03-19 10:51 UTC|newest]

Thread overview: 25+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2019-03-12 20:58 [PATCH 0/9] skl+ cursor DDB allocation fixes Ville Syrjala
2019-03-12 20:58 ` [PATCH 1/9] drm/i915: Accept alloc_size == blocks Ville Syrjala
2019-03-12 20:58 ` [PATCH 2/9] drm/i915: Don't pass plane state to skl_compute_plane_wm() Ville Syrjala
2019-03-12 20:58 ` [PATCH 3/9] drm/i915: Extract skl_compute_wm_params() Ville Syrjala
2019-03-12 20:58 ` [PATCH 4/9] drm/i915: Allocate enough DDB for the cursor Ville Syrjala
2019-03-19 10:51   ` Ville Syrjälä [this message]
2019-03-19 16:03   ` [PATCH v2 " Ville Syrjala
2019-03-12 20:58 ` [PATCH 5/9] drm/i915: Make sure cursor has enough ddb for the selected wm level Ville Syrjala
2019-03-12 20:58 ` [PATCH 6/9] drm/i915: Keep plane watermarks enabled more aggressively Ville Syrjala
2019-03-19  0:09   ` Matt Roper
2019-03-12 20:58 ` [PATCH 7/9] drm/i915: Move some variables to tighter scope Ville Syrjala
2019-03-19  0:10   ` Matt Roper
2019-03-20 15:27     ` Ville Syrjälä
2019-03-12 20:58 ` [PATCH 8/9] drm/i915: Don't pass pipe_wm around so much Ville Syrjala
2019-03-19  0:10   ` Matt Roper
2019-03-12 20:58 ` [PATCH 9/9] drm/i915: Inline skl_build_pipe_wm() into its only caller Ville Syrjala
2019-03-19  0:11   ` Matt Roper
2019-03-19 10:49     ` Ville Syrjälä
2019-03-13 15:34 ` ✗ Fi.CI.SPARSE: warning for skl+ cursor DDB allocation fixes Patchwork
2019-03-13 16:15 ` ✓ Fi.CI.BAT: success " Patchwork
2019-03-13 20:41 ` ✗ Fi.CI.IGT: failure " Patchwork
2019-03-14 17:11   ` Ville Syrjälä
2019-03-19 16:41 ` ✗ Fi.CI.SPARSE: warning for skl+ cursor DDB allocation fixes (rev2) Patchwork
2019-03-19 17:14 ` ✓ Fi.CI.BAT: success " Patchwork
2019-03-20  2:41 ` ✓ Fi.CI.IGT: " Patchwork

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=20190319105123.GU3888@intel.com \
    --to=ville.syrjala@linux.intel.com \
    --cc=intel-gfx@lists.freedesktop.org \
    /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 a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox