intel/blorp: Use the surface format for computing offsets
[mesa.git] / src / intel / blorp / blorp_priv.h
1 /*
2 * Copyright © 2012 Intel Corporation
3 *
4 * Permission is hereby granted, free of charge, to any person obtaining a
5 * copy of this software and associated documentation files (the "Software"),
6 * to deal in the Software without restriction, including without limitation
7 * the rights to use, copy, modify, merge, publish, distribute, sublicense,
8 * and/or sell copies of the Software, and to permit persons to whom the
9 * Software is furnished to do so, subject to the following conditions:
10 *
11 * The above copyright notice and this permission notice (including the next
12 * paragraph) shall be included in all copies or substantial portions of the
13 * Software.
14 *
15 * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
16 * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
17 * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL
18 * THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
19 * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
20 * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS
21 * IN THE SOFTWARE.
22 */
23
24 #pragma once
25
26 #include <stdint.h>
27
28 #include "compiler/nir/nir.h"
29 #include "brw_compiler.h"
30
31 #include "blorp.h"
32
33 #ifdef __cplusplus
34 extern "C" {
35 #endif
36
37 /**
38 * Binding table indices used by BLORP.
39 */
40 enum {
41 BLORP_RENDERBUFFER_BT_INDEX,
42 BLORP_TEXTURE_BT_INDEX,
43 BLORP_NUM_BT_ENTRIES
44 };
45
46 enum blorp_fast_clear_op {
47 BLORP_FAST_CLEAR_OP_NONE = 0,
48 BLORP_FAST_CLEAR_OP_CLEAR,
49 BLORP_FAST_CLEAR_OP_RESOLVE_PARTIAL,
50 BLORP_FAST_CLEAR_OP_RESOLVE_FULL,
51 };
52
53 struct brw_blorp_surface_info
54 {
55 struct isl_surf surf;
56 struct blorp_address addr;
57
58 struct isl_surf aux_surf;
59 struct blorp_address aux_addr;
60 enum isl_aux_usage aux_usage;
61
62 union isl_color_value clear_color;
63
64 struct isl_view view;
65
66 /* Z offset into a 3-D texture or slice of a 2-D array texture. */
67 uint32_t z_offset;
68
69 uint32_t tile_x_sa, tile_y_sa;
70 };
71
72 void
73 brw_blorp_surface_info_init(struct blorp_context *blorp,
74 struct brw_blorp_surface_info *info,
75 const struct blorp_surf *surf,
76 unsigned int level, unsigned int layer,
77 enum isl_format format, bool is_render_target);
78
79
80 struct brw_blorp_coord_transform
81 {
82 float multiplier;
83 float offset;
84 };
85
86 /**
87 * Bounding rectangle telling pixel discard which pixels are not to be
88 * touched. This is needed in when surfaces are configured as something else
89 * what they really are:
90 *
91 * - writing W-tiled stencil as Y-tiled
92 * - writing interleaved multisampled as single sampled.
93 *
94 * See blorp_nir_discard_if_outside_rect().
95 */
96 struct brw_blorp_discard_rect
97 {
98 uint32_t x0;
99 uint32_t x1;
100 uint32_t y0;
101 uint32_t y1;
102 };
103
104 /**
105 * Grid needed for blended and scaled blits of integer formats, see
106 * blorp_nir_manual_blend_bilinear().
107 */
108 struct brw_blorp_rect_grid
109 {
110 float x1;
111 float y1;
112 float pad[2];
113 };
114
115 struct brw_blorp_wm_inputs
116 {
117 struct brw_blorp_discard_rect discard_rect;
118 struct brw_blorp_rect_grid rect_grid;
119 struct brw_blorp_coord_transform coord_transform[2];
120
121 /* Minimum layer setting works for all the textures types but texture_3d
122 * for which the setting has no effect. Use the z-coordinate instead.
123 */
124 uint32_t src_z;
125
126 /* Pad out to an integral number of registers */
127 uint32_t pad[3];
128 };
129
130 struct brw_blorp_prog_data
131 {
132 bool dispatch_8;
133 bool dispatch_16;
134
135 uint8_t first_curbe_grf_0;
136 uint8_t first_curbe_grf_2;
137
138 uint32_t ksp_offset_2;
139
140 /**
141 * True if the WM program should be run in MSDISPMODE_PERSAMPLE with more
142 * than one sample per pixel.
143 */
144 bool persample_msaa_dispatch;
145
146 /**
147 * Mask of which FS inputs are marked flat by the shader source. This is
148 * needed for setting up 3DSTATE_SF/SBE.
149 */
150 uint32_t flat_inputs;
151 unsigned num_varying_inputs;
152 uint64_t inputs_read;
153 };
154
155 static inline unsigned
156 brw_blorp_get_urb_length(const struct brw_blorp_prog_data *prog_data)
157 {
158 if (prog_data == NULL)
159 return 1;
160
161 /* From the BSpec: 3D Pipeline - Strips and Fans - 3DSTATE_SBE
162 *
163 * read_length = ceiling((max_source_attr+1)/2)
164 */
165 return MAX2((prog_data->num_varying_inputs + 1) / 2, 1);
166 }
167
168 struct blorp_params
169 {
170 uint32_t x0;
171 uint32_t y0;
172 uint32_t x1;
173 uint32_t y1;
174 struct brw_blorp_surface_info depth;
175 uint32_t depth_format;
176 struct brw_blorp_surface_info src;
177 struct brw_blorp_surface_info dst;
178 enum blorp_hiz_op hiz_op;
179 enum blorp_fast_clear_op fast_clear_op;
180 bool color_write_disable[4];
181 struct brw_blorp_wm_inputs wm_inputs;
182 unsigned num_draw_buffers;
183 unsigned num_layers;
184 uint32_t wm_prog_kernel;
185 struct brw_blorp_prog_data *wm_prog_data;
186 };
187
188 void blorp_params_init(struct blorp_params *params);
189
190 struct brw_blorp_blit_prog_key
191 {
192 /* Number of samples per pixel that have been configured in the surface
193 * state for texturing from.
194 */
195 unsigned tex_samples;
196
197 /* MSAA layout that has been configured in the surface state for texturing
198 * from.
199 */
200 enum isl_msaa_layout tex_layout;
201
202 enum isl_aux_usage tex_aux_usage;
203
204 /* Actual number of samples per pixel in the source image. */
205 unsigned src_samples;
206
207 /* Actual MSAA layout used by the source image. */
208 enum isl_msaa_layout src_layout;
209
210 /* Number of samples per pixel that have been configured in the render
211 * target.
212 */
213 unsigned rt_samples;
214
215 /* MSAA layout that has been configured in the render target. */
216 enum isl_msaa_layout rt_layout;
217
218 /* Actual number of samples per pixel in the destination image. */
219 unsigned dst_samples;
220
221 /* Actual MSAA layout used by the destination image. */
222 enum isl_msaa_layout dst_layout;
223
224 /* Type of the data to be read from the texture (one of
225 * nir_type_(int|uint|float)).
226 */
227 nir_alu_type texture_data_type;
228
229 /* True if the source image is W tiled. If true, the surface state for the
230 * source image must be configured as Y tiled, and tex_samples must be 0.
231 */
232 bool src_tiled_w;
233
234 /* True if the destination image is W tiled. If true, the surface state
235 * for the render target must be configured as Y tiled, and rt_samples must
236 * be 0.
237 */
238 bool dst_tiled_w;
239
240 /* True if all source samples should be blended together to produce each
241 * destination pixel. If true, src_tiled_w must be false, tex_samples must
242 * equal src_samples, and tex_samples must be nonzero.
243 */
244 bool blend;
245
246 /* True if the rectangle being sent through the rendering pipeline might be
247 * larger than the destination rectangle, so the WM program should kill any
248 * pixels that are outside the destination rectangle.
249 */
250 bool use_kill;
251
252 /**
253 * True if the WM program should be run in MSDISPMODE_PERSAMPLE with more
254 * than one sample per pixel.
255 */
256 bool persample_msaa_dispatch;
257
258 /* True for scaled blitting. */
259 bool blit_scaled;
260
261 /* Scale factors between the pixel grid and the grid of samples. We're
262 * using grid of samples for bilinear filetring in multisample scaled blits.
263 */
264 float x_scale;
265 float y_scale;
266
267 /* True for blits with filter = GL_LINEAR. */
268 bool bilinear_filter;
269 };
270
271 /**
272 * \name BLORP internals
273 * \{
274 *
275 * Used internally by gen6_blorp_exec() and gen7_blorp_exec().
276 */
277
278 void brw_blorp_init_wm_prog_key(struct brw_wm_prog_key *wm_key);
279
280 const unsigned *
281 brw_blorp_compile_nir_shader(struct blorp_context *blorp, struct nir_shader *nir,
282 const struct brw_wm_prog_key *wm_key,
283 bool use_repclear,
284 struct brw_blorp_prog_data *prog_data,
285 unsigned *program_size);
286
287 /** \} */
288
289 #ifdef __cplusplus
290 } /* end extern "C" */
291 #endif /* __cplusplus */