radeonsi: enable DCC fast clear for 128-bit formats

Reviewed-by: Bas Nieuwenhuizen <bas@basnieuwenhuizen.nl>
Reviewed-by: Nicolai Hähnle <nicolai.haehnle@amd.com>

diff --git a/src/gallium/drivers/radeon/r600_texture.c b/src/gallium/drivers/radeon/r600_texture.c
index aee768f..074fed8 100644 (file)
--- a/src/gallium/drivers/radeon/r600_texture.c
+++ b/src/gallium/drivers/radeon/r600_texture.c
@@ -2200,7 +2200,16 @@ static void evergreen_set_clear_color(struct r600_texture *rtex,
 
        memset(&uc, 0, sizeof(uc));
 
-       if (util_format_is_pure_uint(surface_format)) {
+       if (util_format_get_blocksizebits(surface_format) == 128) {
+               /* DCC fast clear only:
+                *   CLEAR_WORD0 = R = G = B
+                *   CLEAR_WORD1 = A
+                */
+               assert(color->ui[0] == color->ui[1] &&
+                      color->ui[0] == color->ui[2]);
+               uc.ui[0] = color->ui[0];
+               uc.ui[1] = color->ui[3];
+       } else if (util_format_is_pure_uint(surface_format)) {
                util_format_write_4ui(surface_format, color->ui, 0, &uc, 0, 0, 0, 1, 1);
        } else if (util_format_is_pure_sint(surface_format)) {
                util_format_write_4i(surface_format, color->i, 0, &uc, 0, 0, 0, 1, 1);
@@ -2211,7 +2220,7 @@ static void evergreen_set_clear_color(struct r600_texture *rtex,
        memcpy(rtex->color_clear_value, &uc, 2 * sizeof(uint32_t));
 }
 
-static void vi_get_fast_clear_parameters(enum pipe_format surface_format,
+static bool vi_get_fast_clear_parameters(enum pipe_format surface_format,
                                         const union pipe_color_union *color,
                                         uint32_t* reset_value,
                                         bool* clear_words_needed)
@@ -2223,6 +2232,11 @@ static void vi_get_fast_clear_parameters(enum pipe_format surface_format,
        int extra_channel;
        const struct util_format_description *desc = util_format_description(surface_format);
 
+       if (desc->block.bits == 128 &&
+           (color->ui[0] != color->ui[1] ||
+            color->ui[0] != color->ui[2]))
+               return false;
+
        *clear_words_needed = true;
        *reset_value = 0x20202020U;
 
@@ -2242,7 +2256,7 @@ static void vi_get_fast_clear_parameters(enum pipe_format surface_format,
                else
                        extra_channel = 0;
        } else
-               return;
+               return true;
 
        for (i = 0; i < 4; ++i) {
                int index = desc->swizzle[i] - PIPE_SWIZZLE_X;
@@ -2258,7 +2272,7 @@ static void vi_get_fast_clear_parameters(enum pipe_format surface_format,
 
                        values[i] = color->i[i] != 0;
                        if (color->i[i] != 0 && MIN2(color->i[i], max) != max)
-                               return;
+                               return true;
                } else if (desc->channel[i].pure_integer &&
                           desc->channel[i].type == UTIL_FORMAT_TYPE_UNSIGNED) {
                        /* Use the maximum value for clamping the clear color. */
@@ -2266,11 +2280,11 @@ static void vi_get_fast_clear_parameters(enum pipe_format surface_format,
 
                        values[i] = color->ui[i] != 0U;
                        if (color->ui[i] != 0U && MIN2(color->ui[i], max) != max)
-                               return;
+                               return true;
                } else {
                        values[i] = color->f[i] != 0.0F;
                        if (color->f[i] != 0.0F && color->f[i] != 1.0F)
-                               return;
+                               return true;
                }
 
                if (index == extra_channel)
@@ -2284,7 +2298,7 @@ static void vi_get_fast_clear_parameters(enum pipe_format surface_format,
                    desc->swizzle[i] - PIPE_SWIZZLE_X != extra_channel &&
                    desc->swizzle[i] >= PIPE_SWIZZLE_X &&
                    desc->swizzle[i] <= PIPE_SWIZZLE_W)
-                       return;
+                       return true;
 
        *clear_words_needed = false;
        if (main_value)
@@ -2292,6 +2306,7 @@ static void vi_get_fast_clear_parameters(enum pipe_format surface_format,
 
        if (extra_value)
                *reset_value |= 0x40404040U;
+       return true;
 }
 
 void vi_dcc_clear_level(struct r600_common_context *rctx,
@@ -2424,11 +2439,6 @@ void evergreen_do_fast_color_clear(struct r600_common_context *rctx,
 
                tex = (struct r600_texture *)fb->cbufs[i]->texture;
 
-               /* 128-bit formats are unusupported */
-               if (util_format_get_blocksizebits(fb->cbufs[i]->format) > 64) {
-                       continue;
-               }
-
                /* the clear is allowed if all layers are bound */
                if (fb->cbufs[i]->u.tex.first_layer != 0 ||
                    fb->cbufs[i]->u.tex.last_layer != util_max_layer(&tex->resource.b.b, 0)) {
@@ -2489,13 +2499,22 @@ void evergreen_do_fast_color_clear(struct r600_common_context *rctx,
                        if (rctx->screen->chip_class >= SI)
                                si_set_optimal_micro_tile_mode(rctx->screen, tex);
 
-                       vi_get_fast_clear_parameters(fb->cbufs[i]->format, color, &reset_value, &clear_words_needed);
+                       if (!vi_get_fast_clear_parameters(fb->cbufs[i]->format,
+                                                         color, &reset_value,
+                                                         &clear_words_needed))
+                               continue;
+
                        vi_dcc_clear_level(rctx, tex, 0, reset_value);
 
                        if (clear_words_needed)
                                tex->dirty_level_mask |= 1 << fb->cbufs[i]->u.tex.level;
                        tex->separate_dcc_dirty = true;
                } else {
+                       /* 128-bit formats are unusupported */
+                       if (util_format_get_blocksizebits(fb->cbufs[i]->format) > 64) {
+                               continue;
+                       }
+
                        /* Stoney/RB+ doesn't work with CMASK fast clear. */
                        if (rctx->family == CHIP_STONEY)
                                continue;