diff mbox series

[v5,6/7] drm/i915/scaler: Check if vblank is sufficient for scaler

Message ID 20250115151714.3665211-7-mitulkumar.ajitkumar.golani@intel.com (mailing list archive)
State New
Headers show
Series Check Scaler and DSC Prefill Latency Against Vblank | expand

Commit Message

Golani, Mitulkumar Ajitkumar Jan. 15, 2025, 3:17 p.m. UTC
High refresh rate panels which may have small line times
and vblank sizes, Check if vblank size is sufficient for
enabled scaler users.

Bspec: 70151

--v2:
- Use hweight* family of functions for counting bits. [Jani]
- Update precision handling for hscale and vscale. [Ankit]
- Consider chroma downscaling factor during latency
calculation. [Ankit]
- Replace function name from scaler_prefill_time to
scaler_prefill_latency.

--v3:
- hscale_k and vscale_k values are already left shifted
by 16, after multiplying by 1000, those need to be right
shifted to 16. [Ankit]
- Replace YCBCR444 to YCBCR420. [Ankit]
- Divide by 1000 * 1000 in end to get correct precision. [Ankit]
- Initialise latency to 0 to avoid any garbage.

--v4:
- Elaborate commit message and add Bspec number. [Ankit]
- Improvise latency calculation. [Ankit]
- Use ceiling value for down scaling factor when less than 1
as per bspec. [Ankit]
- Correct linetime calculation. [Ankit]
- Consider cdclk prefill adjustment while prefill
computation.[Ankit]

Signed-off-by: Mitul Golani <mitulkumar.ajitkumar.golani@intel.com>
---
 drivers/gpu/drm/i915/display/skl_watermark.c | 49 ++++++++++++++++++++
 1 file changed, 49 insertions(+)
diff mbox series

Patch

diff --git a/drivers/gpu/drm/i915/display/skl_watermark.c b/drivers/gpu/drm/i915/display/skl_watermark.c
index f4458d1185b3..ca9d5677c356 100644
--- a/drivers/gpu/drm/i915/display/skl_watermark.c
+++ b/drivers/gpu/drm/i915/display/skl_watermark.c
@@ -2292,6 +2292,54 @@  static int icl_build_plane_wm(struct intel_crtc_state *crtc_state,
 	return 0;
 }
 
+static int
+cdclk_prefill_adjustment(const struct intel_crtc_state *crtc_state)
+{
+	struct intel_atomic_state *state =
+		to_intel_atomic_state(crtc_state->uapi.state);
+	const struct intel_cdclk_state *cdclk_state;
+
+	cdclk_state = intel_atomic_get_cdclk_state(state);
+	if (IS_ERR(cdclk_state))
+		return PTR_ERR(cdclk_state);
+
+	return min(1, DIV_ROUND_UP(crtc_state->pixel_rate,
+				   2 * cdclk_state->logical.cdclk));
+}
+
+static int
+scaler_prefill_latency(const struct intel_crtc_state *crtc_state)
+{
+	const struct intel_crtc_scaler_state *scaler_state =
+					&crtc_state->scaler_state;
+	int num_scaler_users = hweight32(scaler_state->scaler_users);
+	int latency = 0;
+	int linetime =
+		intel_usecs_to_scanlines(&crtc_state->hw.adjusted_mode,
+					 DIV_ROUND_UP(1000 * crtc_state->hw.adjusted_mode.htotal,
+						      crtc_state->pixel_rate));
+	long long  hscale_k, vscale_k;
+
+	if (!num_scaler_users)
+		return latency;
+
+	latency = intel_usecs_to_scanlines(&crtc_state->hw.adjusted_mode, 4 * linetime);
+
+	if (num_scaler_users > 1) {
+		int chroma_downscaling_factor =
+			crtc_state->output_format == INTEL_OUTPUT_FORMAT_YCBCR420 ? 2 : 1;
+		hscale_k = max(1000, mul_u32_u32(scaler_state->scalers[0].hscale, 1000) >> 16);
+		vscale_k = max(1000, mul_u32_u32(scaler_state->scalers[0].vscale, 1000) >> 16);
+
+		latency += chroma_downscaling_factor *
+			   DIV_ROUND_UP_ULL((4 * linetime * hscale_k * vscale_k),
+					    1000000);
+		latency *= cdclk_prefill_adjustment(crtc_state);
+	}
+
+	return latency;
+}
+
 static bool
 skl_is_vblank_too_short(const struct intel_crtc_state *crtc_state,
 			int wm0_lines, int latency)
@@ -2302,6 +2350,7 @@  skl_is_vblank_too_short(const struct intel_crtc_state *crtc_state,
 	/* FIXME missing scaler and DSC pre-fill time */
 	return crtc_state->framestart_delay +
 		intel_usecs_to_scanlines(adjusted_mode, latency) +
+		scaler_prefill_latency(crtc_state) +
 		wm0_lines >
 		adjusted_mode->crtc_vtotal - adjusted_mode->crtc_vblank_start;
 }