drm/amd/display: Fix programming of gamma end points
authorAnthony Koo <Anthony.Koo@amd.com>
Fri, 16 Dec 2016 17:12:03 +0000 (12:12 -0500)
committerAlex Deucher <alexander.deucher@amd.com>
Tue, 26 Sep 2017 21:05:54 +0000 (17:05 -0400)
Signed-off-by: Anthony Koo <anthony.koo@amd.com>
Reviewed-by: Anthony Koo <Anthony.Koo@amd.com>
Acked-by: Harry Wentland <Harry.Wentland@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
drivers/gpu/drm/amd/display/dc/calcs/gamma_calcs.c

index 729652a3763d2e7da173f7a7ff8d4e2056e54eec..5cd408dadb201b5be16ec661673a07e4ff247a4f 100644 (file)
@@ -245,6 +245,8 @@ static bool build_hw_curve_configuration(
        uint32_t segments = 0;
        uint32_t max_number;
 
+       int8_t num_regions = 0;
+
        bool result = false;
 
        if (!number_of_points) {
@@ -273,6 +275,7 @@ static bool build_hw_curve_configuration(
                ASSERT(curve_config->segments[i] >= 0);
 
                segments += (1 << curve_config->segments[i]);
+               ++num_regions;
 
                ++i;
        }
@@ -284,12 +287,14 @@ static bool build_hw_curve_configuration(
                uint32_t offset = 0;
                int8_t begin = curve_config->begin;
                int32_t region_number = 0;
+               struct fixed31_32 magic_number =
+                               dal_fixed31_32_from_fraction(249, 1000);
 
                i = begin;
 
                while ((index < max_number) &&
                        (region_number < max_regions_number) &&
-                       (i <= 1)) {
+                       (i < (begin + num_regions))) {
                        int32_t j = 0;
 
                        segments = curve_config->segments[region_number];
@@ -345,8 +350,7 @@ static bool build_hw_curve_configuration(
                                divisor);
 
                        points[index].x = region1;
-
-                       round_custom_float_6_12(points + index);
+                       points[index].adjusted_x = region1;
 
                        ++index;
                        ++region_number;
@@ -366,9 +370,10 @@ static bool build_hw_curve_configuration(
                        ++i;
                }
 
-               points[index].x = region1;
-
-               round_custom_float_6_12(points + index);
+               points[index].x =
+                               dal_fixed31_32_add(region1, magic_number);
+               points[index].adjusted_x =
+                               dal_fixed31_32_add(region1, magic_number);
 
                *number_of_points = index;
 
@@ -1215,15 +1220,11 @@ static void rebuild_curve_configuration_magic(
                const struct hw_x_point *coordinates_x,
                uint32_t hw_points_num)
 {
-       const struct fixed31_32 magic_number =
-               dal_fixed31_32_from_fraction(249, 1000);
-
        struct fixed31_32 y_r;
        struct fixed31_32 y_g;
        struct fixed31_32 y_b;
 
        struct fixed31_32 y1_min;
-       struct fixed31_32 y2_max;
        struct fixed31_32 y3_max;
 
        y_r = rgb_resulted[0].red;
@@ -1238,29 +1239,31 @@ static void rebuild_curve_configuration_magic(
                                        arr_points[0].y,
                                        arr_points[0].x);
 
-       arr_points[1].x = dal_fixed31_32_add(
-                       coordinates_x[hw_points_num - 1].adjusted_x,
-                       magic_number);
-
-       arr_points[2].x = arr_points[1].x;
-
-       y_r = rgb_resulted[hw_points_num - 1].red;
-       y_g = rgb_resulted[hw_points_num - 1].green;
-       y_b = rgb_resulted[hw_points_num - 1].blue;
-
-       y2_max = dal_fixed31_32_max(y_r, dal_fixed31_32_max(y_g, y_b));
-
-       arr_points[1].y = y2_max;
+       /* this should be cleaned up as it's confusing my understanding (KK) is
+        * that REGAMMA_CNTLA_EXP_REGION_END is the X value for the region end
+        * REGAMMA_CNTLA_EXP_REGION_END_BASE is Y value for the above X
+        * REGAMMA_CNTLA_EXP_REGION_END_SLOPE is the slope beyond (X,Y) above
+        * currently when programming REGION_END = m_arrPoints[1].x,
+        * REGION_END_BASE = m_arrPoints[1].y, REGION_END_SLOPE=1
+        * we don't use m_arrPoints[2] at all after this function,
+        * and its purpose isn't clear to me
+        */
+       arr_points[1].x = coordinates_x[hw_points_num].adjusted_x;
+       arr_points[2].x = coordinates_x[hw_points_num].adjusted_x;
 
        y_r = rgb_resulted[hw_points_num].red;
        y_g = rgb_resulted[hw_points_num].green;
        y_b = rgb_resulted[hw_points_num].blue;
 
+       /* see comment above, m_arrPoints[1].y should be the Y value for the
+        * region end (m_numOfHwPoints), not last HW point(m_numOfHwPoints - 1)
+        */
        y3_max = dal_fixed31_32_max(y_r, dal_fixed31_32_max(y_g, y_b));
 
+       arr_points[1].y = y3_max;
        arr_points[2].y = y3_max;
 
-       arr_points[2].slope = dal_fixed31_32_one;
+       arr_points[2].slope = dal_fixed31_32_zero;
 }
 
 static bool convert_to_custom_float_format(