[Intel-gfx] [PATCH v2] drm/i915/skl: Make sure to allocate mininum sizes in the DDB

[Damien Lespiau]

I overlooked the fact that we need to allocate a minimum 8 blocks and
that just allocating the planes depending on how much they need to fetch
from the DDB in proportion of how much memory bw is necessary for the
whole display can lead to cases where we don't respect those minima (and
thus overrun).

So, instead, start by allocating 8 blocks to each active display plane
and then allocate the remaining blocks like before.

v2: Rebase on top of -nightly

Cc: Mahesh Kumar <mahesh1.kumar at intel.com>
Signed-off-by: Damien Lespiau <damien.lespiau at intel.com>
---
 drivers/gpu/drm/i915/intel_pm.c | 22 ++++++++++++++++++----
 1 file changed, 18 insertions(+), 4 deletions(-)

diff --git a/drivers/gpu/drm/i915/intel_pm.c b/drivers/gpu/drm/i915/intel_pm.c
index bebefe7..f6c7e53 100644
--- a/drivers/gpu/drm/i915/intel_pm.c
+++ b/drivers/gpu/drm/i915/intel_pm.c
@@ -2502,6 +2502,7 @@ skl_allocate_pipe_ddb(struct drm_crtc *crtc,
 	enum pipe pipe = intel_crtc->pipe;
 	struct skl_ddb_entry *alloc = &ddb->pipe[pipe];
 	uint16_t alloc_size, start, cursor_blocks;
+	uint16_t minimum[I915_MAX_PLANES];
 	unsigned int total_data_rate;
 	int plane;
 
@@ -2520,9 +2521,21 @@ skl_allocate_pipe_ddb(struct drm_crtc *crtc,
 	alloc_size -= cursor_blocks;
 	alloc->end -= cursor_blocks;
 
+	/* 1. Allocate the mininum required blocks for each active plane */
+	for_each_plane(pipe, plane) {
+		const struct intel_plane_wm_parameters *p;
+
+		p = &params->plane[plane];
+		if (!p->enabled)
+			continue;
+
+		minimum[plane] = 8;
+		alloc_size -= minimum[plane];
+	}
+
 	/*
-	 * Each active plane get a portion of the remaining space, in
-	 * proportion to the amount of data they need to fetch from memory.
+	 * 2. Distribute the remaining space in proportion to the amount of
+	 * data each plane needs to fetch from memory.
 	 *
 	 * FIXME: we may not allocate every single block here.
 	 */
@@ -2544,8 +2557,9 @@ skl_allocate_pipe_ddb(struct drm_crtc *crtc,
 		 * promote the expression to 64 bits to avoid overflowing, the
 		 * result is < available as data_rate / total_data_rate < 1
 		 */
-		plane_blocks = div_u64((uint64_t)alloc_size * data_rate,
-				       total_data_rate);
+		plane_blocks = minimum[plane];
+		plane_blocks += div_u64((uint64_t)alloc_size * data_rate,
+					total_data_rate);
 
 		ddb->plane[pipe][plane].start = start;
 		ddb->plane[pipe][plane].end = start + plane_blocks;
-- 
1.8.3.1