1 files changed, 66 insertions, 65 deletions

diff --git a/drivers/gpu/drm/i915/intel_color.c b/drivers/gpu/drm/i915/intel_color.c
index aa66e952a95d..c6a7beabd58d 100644
--- a/drivers/gpu/drm/i915/intel_color.c
+++ b/drivers/gpu/drm/i915/intel_color.c
@@ -39,7 +39,7 @@
 #define CTM_COEFF_NEGATIVE(coeff)	(((coeff) & CTM_COEFF_SIGN) != 0)
 #define CTM_COEFF_ABS(coeff)		((coeff) & (CTM_COEFF_SIGN - 1))
 
-#define LEGACY_LUT_LENGTH		(sizeof(struct drm_color_lut) * 256)
+#define LEGACY_LUT_LENGTH		256
 
 /* Post offset values for RGB->YCBCR conversion */
 #define POSTOFF_RGB_TO_YUV_HI 0x800
@@ -66,48 +66,49 @@
  * of the CTM coefficient and we write the value from bit 3. We also round the
  * value.
  */
-#define I9XX_CSC_COEFF_FP(coeff, fbits)	\
+#define ILK_CSC_COEFF_FP(coeff, fbits)	\
 	(clamp_val(((coeff) >> (32 - (fbits) - 3)) + 4, 0, 0xfff) & 0xff8)
 
-#define I9XX_CSC_COEFF_LIMITED_RANGE	\
-	I9XX_CSC_COEFF_FP(CTM_COEFF_LIMITED_RANGE, 9)
-#define I9XX_CSC_COEFF_1_0		\
-	((7 << 12) | I9XX_CSC_COEFF_FP(CTM_COEFF_1_0, 8))
+#define ILK_CSC_COEFF_LIMITED_RANGE	\
+	ILK_CSC_COEFF_FP(CTM_COEFF_LIMITED_RANGE, 9)
+#define ILK_CSC_COEFF_1_0		\
+	((7 << 12) | ILK_CSC_COEFF_FP(CTM_COEFF_1_0, 8))
 
 static bool crtc_state_is_legacy_gamma(struct drm_crtc_state *state)
 {
 	return !state->degamma_lut &&
 		!state->ctm &&
 		state->gamma_lut &&
-		state->gamma_lut->length == LEGACY_LUT_LENGTH;
+		drm_color_lut_size(state->gamma_lut) == LEGACY_LUT_LENGTH;
 }
 
 /*
  * When using limited range, multiply the matrix given by userspace by
- * the matrix that we would use for the limited range. We do the
- * multiplication in U2.30 format.
+ * the matrix that we would use for the limited range.
  */
-static void ctm_mult_by_limited(uint64_t *result, int64_t *input)
+static u64 *ctm_mult_by_limited(u64 *result, const u64 *input)
 {
 	int i;
 
-	for (i = 0; i < 9; i++)
-		result[i] = 0;
+	for (i = 0; i < 9; i++) {
+		u64 user_coeff = input[i];
+		u32 limited_coeff = CTM_COEFF_LIMITED_RANGE;
+		u32 abs_coeff = clamp_val(CTM_COEFF_ABS(user_coeff), 0,
+					  CTM_COEFF_4_0 - 1) >> 2;
 
-	for (i = 0; i < 3; i++) {
-		int64_t user_coeff = input[i * 3 + i];
-		uint64_t limited_coeff = CTM_COEFF_LIMITED_RANGE >> 2;
-		uint64_t abs_coeff = clamp_val(CTM_COEFF_ABS(user_coeff),
-					       0,
-					       CTM_COEFF_4_0 - 1) >> 2;
-
-		result[i * 3 + i] = (limited_coeff * abs_coeff) >> 27;
-		if (CTM_COEFF_NEGATIVE(user_coeff))
-			result[i * 3 + i] |= CTM_COEFF_SIGN;
+		/*
+		 * By scaling every co-efficient with limited range (16-235)
+		 * vs full range (0-255) the final o/p will be scaled down to
+		 * fit in the limited range supported by the panel.
+		 */
+		result[i] = mul_u32_u32(limited_coeff, abs_coeff) >> 30;
+		result[i] |= user_coeff & CTM_COEFF_SIGN;
 	}
+
+	return result;
 }
 
-static void i9xx_load_ycbcr_conversion_matrix(struct intel_crtc *intel_crtc)
+static void ilk_load_ycbcr_conversion_matrix(struct intel_crtc *intel_crtc)
 {
 	int pipe = intel_crtc->pipe;
 	struct drm_i915_private *dev_priv = to_i915(intel_crtc->base.dev);
@@ -131,8 +132,7 @@ static void i9xx_load_ycbcr_conversion_matrix(struct intel_crtc *intel_crtc)
 	I915_WRITE(PIPE_CSC_MODE(pipe), 0);
 }
 
-/* Set up the pipe CSC unit. */
-static void i9xx_load_csc_matrix(struct drm_crtc_state *crtc_state)
+static void ilk_load_csc_matrix(struct drm_crtc_state *crtc_state)
 {
 	struct drm_crtc *crtc = crtc_state->crtc;
 	struct drm_i915_private *dev_priv = to_i915(crtc->dev);
@@ -140,21 +140,27 @@ static void i9xx_load_csc_matrix(struct drm_crtc_state *crtc_state)
 	int i, pipe = intel_crtc->pipe;
 	uint16_t coeffs[9] = { 0, };
 	struct intel_crtc_state *intel_crtc_state = to_intel_crtc_state(crtc_state);
+	bool limited_color_range = false;
+
+	/*
+	 * FIXME if there's a gamma LUT after the CSC, we should
+	 * do the range compression using the gamma LUT instead.
+	 */
+	if (INTEL_GEN(dev_priv) >= 8 || IS_HASWELL(dev_priv))
+		limited_color_range = intel_crtc_state->limited_color_range;
 
 	if (intel_crtc_state->ycbcr420) {
-		i9xx_load_ycbcr_conversion_matrix(intel_crtc);
+		ilk_load_ycbcr_conversion_matrix(intel_crtc);
 		return;
 	} else if (crtc_state->ctm) {
-		struct drm_color_ctm *ctm =
-			(struct drm_color_ctm *)crtc_state->ctm->data;
-		uint64_t input[9] = { 0, };
-
-		if (intel_crtc_state->limited_color_range) {
-			ctm_mult_by_limited(input, ctm->matrix);
-		} else {
-			for (i = 0; i < ARRAY_SIZE(input); i++)
-				input[i] = ctm->matrix[i];
-		}
+		struct drm_color_ctm *ctm = crtc_state->ctm->data;
+		const u64 *input;
+		u64 temp[9];
+
+		if (limited_color_range)
+			input = ctm_mult_by_limited(temp, ctm->matrix);
+		else
+			input = ctm->matrix;
 
 		/*
 		 * Convert fixed point S31.32 input to format supported by the
@@ -175,21 +181,21 @@ static void i9xx_load_csc_matrix(struct drm_crtc_state *crtc_state)
 
 			if (abs_coeff < CTM_COEFF_0_125)
 				coeffs[i] |= (3 << 12) |
-					I9XX_CSC_COEFF_FP(abs_coeff, 12);
+					ILK_CSC_COEFF_FP(abs_coeff, 12);
 			else if (abs_coeff < CTM_COEFF_0_25)
 				coeffs[i] |= (2 << 12) |
-					I9XX_CSC_COEFF_FP(abs_coeff, 11);
+					ILK_CSC_COEFF_FP(abs_coeff, 11);
 			else if (abs_coeff < CTM_COEFF_0_5)
 				coeffs[i] |= (1 << 12) |
-					I9XX_CSC_COEFF_FP(abs_coeff, 10);
+					ILK_CSC_COEFF_FP(abs_coeff, 10);
 			else if (abs_coeff < CTM_COEFF_1_0)
-				coeffs[i] |= I9XX_CSC_COEFF_FP(abs_coeff, 9);
+				coeffs[i] |= ILK_CSC_COEFF_FP(abs_coeff, 9);
 			else if (abs_coeff < CTM_COEFF_2_0)
 				coeffs[i] |= (7 << 12) |
-					I9XX_CSC_COEFF_FP(abs_coeff, 8);
+					ILK_CSC_COEFF_FP(abs_coeff, 8);
 			else
 				coeffs[i] |= (6 << 12) |
-					I9XX_CSC_COEFF_FP(abs_coeff, 7);
+					ILK_CSC_COEFF_FP(abs_coeff, 7);
 		}
 	} else {
 		/*
@@ -201,11 +207,11 @@ static void i9xx_load_csc_matrix(struct drm_crtc_state *crtc_state)
 		 * into consideration.
 		 */
 		for (i = 0; i < 3; i++) {
-			if (intel_crtc_state->limited_color_range)
+			if (limited_color_range)
 				coeffs[i * 3 + i] =
-					I9XX_CSC_COEFF_LIMITED_RANGE;
+					ILK_CSC_COEFF_LIMITED_RANGE;
 			else
-				coeffs[i * 3 + i] = I9XX_CSC_COEFF_1_0;
+				coeffs[i * 3 + i] = ILK_CSC_COEFF_1_0;
 		}
 	}
 
@@ -225,7 +231,7 @@ static void i9xx_load_csc_matrix(struct drm_crtc_state *crtc_state)
 	if (INTEL_GEN(dev_priv) > 6) {
 		uint16_t postoff = 0;
 
-		if (intel_crtc_state->limited_color_range)
+		if (limited_color_range)
 			postoff = (16 * (1 << 12) / 255) & 0x1fff;
 
 		I915_WRITE(PIPE_CSC_POSTOFF_HI(pipe), postoff);
@@ -236,7 +242,7 @@ static void i9xx_load_csc_matrix(struct drm_crtc_state *crtc_state)
 	} else {
 		uint32_t mode = CSC_MODE_YUV_TO_RGB;
 
-		if (intel_crtc_state->limited_color_range)
+		if (limited_color_range)
 			mode |= CSC_BLACK_SCREEN_OFFSET;
 
 		I915_WRITE(PIPE_CSC_MODE(pipe), mode);
@@ -255,8 +261,7 @@ static void cherryview_load_csc_matrix(struct drm_crtc_state *state)
 	uint32_t mode;
 
 	if (state->ctm) {
-		struct drm_color_ctm *ctm =
-			(struct drm_color_ctm *) state->ctm->data;
+		struct drm_color_ctm *ctm = state->ctm->data;
 		uint16_t coeffs[9] = { 0, };
 		int i;
 
@@ -323,7 +328,7 @@ static void i9xx_load_luts_internal(struct drm_crtc *crtc,
 	}
 
 	if (blob) {
-		struct drm_color_lut *lut = (struct drm_color_lut *) blob->data;
+		struct drm_color_lut *lut = blob->data;
 		for (i = 0; i < 256; i++) {
 			uint32_t word =
 				(drm_color_lut_extract(lut[i].red, 8) << 16) |
@@ -393,8 +398,7 @@ static void bdw_load_degamma_lut(struct drm_crtc_state *state)
 		   PAL_PREC_SPLIT_MODE | PAL_PREC_AUTO_INCREMENT);
 
 	if (state->degamma_lut) {
-		struct drm_color_lut *lut =
-			(struct drm_color_lut *) state->degamma_lut->data;
+		struct drm_color_lut *lut = state->degamma_lut->data;
 
 		for (i = 0; i < lut_size; i++) {
 			uint32_t word =
@@ -428,8 +432,7 @@ static void bdw_load_gamma_lut(struct drm_crtc_state *state, u32 offset)
 		   offset);
 
 	if (state->gamma_lut) {
-		struct drm_color_lut *lut =
-			(struct drm_color_lut *) state->gamma_lut->data;
+		struct drm_color_lut *lut = state->gamma_lut->data;
 
 		for (i = 0; i < lut_size; i++) {
 			uint32_t word =
@@ -561,7 +564,7 @@ static void cherryview_load_luts(struct drm_crtc_state *state)
 	}
 
 	if (state->degamma_lut) {
-		lut = (struct drm_color_lut *) state->degamma_lut->data;
+		lut = state->degamma_lut->data;
 		lut_size = INTEL_INFO(dev_priv)->color.degamma_lut_size;
 		for (i = 0; i < lut_size; i++) {
 			/* Write LUT in U0.14 format. */
@@ -576,7 +579,7 @@ static void cherryview_load_luts(struct drm_crtc_state *state)
 	}
 
 	if (state->gamma_lut) {
-		lut = (struct drm_color_lut *) state->gamma_lut->data;
+		lut = state->gamma_lut->data;
 		lut_size = INTEL_INFO(dev_priv)->color.gamma_lut_size;
 		for (i = 0; i < lut_size; i++) {
 			/* Write LUT in U0.10 format. */
@@ -616,19 +619,17 @@ int intel_color_check(struct drm_crtc *crtc,
 	struct drm_i915_private *dev_priv = to_i915(crtc->dev);
 	size_t gamma_length, degamma_length;
 
-	degamma_length = INTEL_INFO(dev_priv)->color.degamma_lut_size *
-		sizeof(struct drm_color_lut);
-	gamma_length = INTEL_INFO(dev_priv)->color.gamma_lut_size *
-		sizeof(struct drm_color_lut);
+	degamma_length = INTEL_INFO(dev_priv)->color.degamma_lut_size;
+	gamma_length = INTEL_INFO(dev_priv)->color.gamma_lut_size;
 
 	/*
 	 * We allow both degamma & gamma luts at the right size or
 	 * NULL.
 	 */
 	if ((!crtc_state->degamma_lut ||
-	     crtc_state->degamma_lut->length == degamma_length) &&
+	     drm_color_lut_size(crtc_state->degamma_lut) == degamma_length) &&
 	    (!crtc_state->gamma_lut ||
-	     crtc_state->gamma_lut->length == gamma_length))
+	     drm_color_lut_size(crtc_state->gamma_lut) == gamma_length))
 		return 0;
 
 	/*
@@ -651,14 +652,14 @@ void intel_color_init(struct drm_crtc *crtc)
 		dev_priv->display.load_csc_matrix = cherryview_load_csc_matrix;
 		dev_priv->display.load_luts = cherryview_load_luts;
 	} else if (IS_HASWELL(dev_priv)) {
-		dev_priv->display.load_csc_matrix = i9xx_load_csc_matrix;
+		dev_priv->display.load_csc_matrix = ilk_load_csc_matrix;
 		dev_priv->display.load_luts = haswell_load_luts;
 	} else if (IS_BROADWELL(dev_priv) || IS_GEN9_BC(dev_priv) ||
 		   IS_BROXTON(dev_priv)) {
-		dev_priv->display.load_csc_matrix = i9xx_load_csc_matrix;
+		dev_priv->display.load_csc_matrix = ilk_load_csc_matrix;
 		dev_priv->display.load_luts = broadwell_load_luts;
 	} else if (IS_GEMINILAKE(dev_priv) || IS_CANNONLAKE(dev_priv)) {
-		dev_priv->display.load_csc_matrix = i9xx_load_csc_matrix;
+		dev_priv->display.load_csc_matrix = ilk_load_csc_matrix;
 		dev_priv->display.load_luts = glk_load_luts;
 	} else {
 		dev_priv->display.load_luts = i9xx_load_luts;