summaryrefslogtreecommitdiff
path: root/src/gallium/drivers/nvc0/nvc0_tex.c
blob: f6c4ab39bd99c132fd707877471296997f39dfdb (plain)
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
/*
 * Copyright 2008 Ben Skeggs
 *
 * Permission is hereby granted, free of charge, to any person obtaining a
 * copy of this software and associated documentation files (the "Software"),
 * to deal in the Software without restriction, including without limitation
 * the rights to use, copy, modify, merge, publish, distribute, sublicense,
 * and/or sell copies of the Software, and to permit persons to whom the
 * Software is furnished to do so, subject to the following conditions:
 *
 * The above copyright notice and this permission notice shall be included in
 * all copies or substantial portions of the Software.
 *
 * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
 * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
 * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
 * THE AUTHORS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY,
 * WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF
 * OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
 * SOFTWARE.
 */

#include "nvc0_context.h"
#include "nvc0_resource.h"
#include "nv50/nv50_texture.xml.h"

#include "util/u_format.h"

#define NV50_TIC_0_SWIZZLE__MASK                      \
   (NV50_TIC_0_MAPA__MASK | NV50_TIC_0_MAPB__MASK |   \
    NV50_TIC_0_MAPG__MASK | NV50_TIC_0_MAPR__MASK)

static INLINE uint32_t
nv50_tic_swizzle(uint32_t tc, unsigned swz, boolean tex_int)
{
   switch (swz) {
   case PIPE_SWIZZLE_RED:
      return (tc & NV50_TIC_0_MAPR__MASK) >> NV50_TIC_0_MAPR__SHIFT;
   case PIPE_SWIZZLE_GREEN:
      return (tc & NV50_TIC_0_MAPG__MASK) >> NV50_TIC_0_MAPG__SHIFT;
   case PIPE_SWIZZLE_BLUE:
      return (tc & NV50_TIC_0_MAPB__MASK) >> NV50_TIC_0_MAPB__SHIFT;
   case PIPE_SWIZZLE_ALPHA:
      return (tc & NV50_TIC_0_MAPA__MASK) >> NV50_TIC_0_MAPA__SHIFT;
   case PIPE_SWIZZLE_ONE:
      return tex_int ? NV50_TIC_MAP_ONE_INT : NV50_TIC_MAP_ONE_FLOAT;
   case PIPE_SWIZZLE_ZERO:
   default:
      return NV50_TIC_MAP_ZERO;
   }
}

struct pipe_sampler_view *
nvc0_create_sampler_view(struct pipe_context *pipe,
                         struct pipe_resource *texture,
                         const struct pipe_sampler_view *templ)
{
   const struct util_format_description *desc;
   uint64_t address;
   uint32_t *tic;
   uint32_t swz[4];
   uint32_t depth;
   struct nv50_tic_entry *view;
   struct nv50_miptree *mt;
   boolean tex_int;

   view = MALLOC_STRUCT(nv50_tic_entry);
   if (!view)
      return NULL;
   mt = nv50_miptree(texture);

   view->pipe = *templ;
   view->pipe.reference.count = 1;
   view->pipe.texture = NULL;
   view->pipe.context = pipe;

   view->id = -1;

   pipe_resource_reference(&view->pipe.texture, texture);

   tic = &view->tic[0];

   desc = util_format_description(view->pipe.format);

   tic[0] = nvc0_format_table[view->pipe.format].tic;

   tex_int = util_format_is_pure_integer(view->pipe.format);

   swz[0] = nv50_tic_swizzle(tic[0], view->pipe.swizzle_r, tex_int);
   swz[1] = nv50_tic_swizzle(tic[0], view->pipe.swizzle_g, tex_int);
   swz[2] = nv50_tic_swizzle(tic[0], view->pipe.swizzle_b, tex_int);
   swz[3] = nv50_tic_swizzle(tic[0], view->pipe.swizzle_a, tex_int);
   tic[0] = (tic[0] & ~NV50_TIC_0_SWIZZLE__MASK) |
      (swz[0] << NV50_TIC_0_MAPR__SHIFT) |
      (swz[1] << NV50_TIC_0_MAPG__SHIFT) |
      (swz[2] << NV50_TIC_0_MAPB__SHIFT) |
      (swz[3] << NV50_TIC_0_MAPA__SHIFT);

   address = mt->base.address;

   tic[2] = 0x10001000 | NV50_TIC_2_NO_BORDER;

   if (desc->colorspace == UTIL_FORMAT_COLORSPACE_SRGB)
      tic[2] |= NV50_TIC_2_COLORSPACE_SRGB;

   /* check for linear storage type */
   if (unlikely(!nouveau_bo_memtype(nv04_resource(texture)->bo))) {
      if (texture->target == PIPE_BUFFER) {
         address +=
            view->pipe.u.buf.first_element * desc->block.bits / 8;
         tic[2] |= NV50_TIC_2_LINEAR | NV50_TIC_2_TARGET_BUFFER;
         tic[3] = 0;
         tic[4] = /* width */
            view->pipe.u.buf.last_element - view->pipe.u.buf.first_element + 1;
         tic[5] = 0;
      } else {
         /* must be 2D texture without mip maps */
         tic[2] |= NV50_TIC_2_LINEAR | NV50_TIC_2_TARGET_RECT;
         if (texture->target != PIPE_TEXTURE_RECT)
            tic[2] |= NV50_TIC_2_NORMALIZED_COORDS;
         tic[3] = mt->level[0].pitch;
         tic[4] = mt->base.base.width0;
         tic[5] = (1 << 16) | mt->base.base.height0;
      }
      tic[6] =
      tic[7] = 0;
      tic[1] = address;
      tic[2] |= address >> 32;
      return &view->pipe;
   }

   if (mt->base.base.target != PIPE_TEXTURE_RECT)
      tic[2] |= NV50_TIC_2_NORMALIZED_COORDS;

   tic[2] |=
      ((mt->level[0].tile_mode & 0x0f0) << (22 - 4)) |
      ((mt->level[0].tile_mode & 0xf00) << (25 - 8));

   depth = MAX2(mt->base.base.array_size, mt->base.base.depth0);

   if (mt->base.base.array_size > 1) {
      /* there doesn't seem to be a base layer field in TIC */
      address += view->pipe.u.tex.first_layer * mt->layer_stride;
      depth = view->pipe.u.tex.last_layer - view->pipe.u.tex.first_layer + 1;
   }
   tic[1] = address;
   tic[2] |= address >> 32;

   switch (mt->base.base.target) {
   case PIPE_TEXTURE_1D:
      tic[2] |= NV50_TIC_2_TARGET_1D;
      break;
/* case PIPE_TEXTURE_2D_MS: */
   case PIPE_TEXTURE_2D:
      tic[2] |= NV50_TIC_2_TARGET_2D;
      break;
   case PIPE_TEXTURE_RECT:
      tic[2] |= NV50_TIC_2_TARGET_RECT;
      break;
   case PIPE_TEXTURE_3D:
      tic[2] |= NV50_TIC_2_TARGET_3D;
      break;
   case PIPE_TEXTURE_CUBE:
      depth /= 6;
      if (depth > 1)
         tic[2] |= NV50_TIC_2_TARGET_CUBE_ARRAY;
      else
         tic[2] |= NV50_TIC_2_TARGET_CUBE;
      break;
   case PIPE_TEXTURE_1D_ARRAY:
      tic[2] |= NV50_TIC_2_TARGET_1D_ARRAY;
      break;
/* case PIPE_TEXTURE_2D_ARRAY_MS: */
   case PIPE_TEXTURE_2D_ARRAY:
      tic[2] |= NV50_TIC_2_TARGET_2D_ARRAY;
      break;
   default:
      NOUVEAU_ERR("invalid texture target: %d\n", mt->base.base.target);
      return FALSE;
   }

   if (mt->base.base.target == PIPE_BUFFER)
      tic[3] = mt->base.base.width0;
   else
      tic[3] = 0x00300000;

   tic[4] = (1 << 31) | (mt->base.base.width0 << mt->ms_x);

   tic[5] = (mt->base.base.height0 << mt->ms_y) & 0xffff;
   tic[5] |= depth << 16;
   tic[5] |= mt->base.base.last_level << 28;

   tic[6] = (mt->ms_x > 1) ? 0x88000000 : 0x03000000; /* sampling points */

   tic[7] = (view->pipe.u.tex.last_level << 4) | view->pipe.u.tex.first_level;

   /*
   if (mt->base.base.target == PIPE_TEXTURE_2D_MS ||
       mt->base.base.target == PIPE_TEXTURE_2D_ARRAY_MS)
      tic[7] |= mt->ms_mode << 12;
   */

   return &view->pipe;
}

static boolean
nvc0_validate_tic(struct nvc0_context *nvc0, int s)
{
   uint32_t commands[32];
   struct nouveau_pushbuf *push = nvc0->base.pushbuf;
   struct nouveau_bo *txc = nvc0->screen->txc;
   unsigned i;
   unsigned n = 0;
   boolean need_flush = FALSE;

   for (i = 0; i < nvc0->num_textures[s]; ++i) {
      struct nv50_tic_entry *tic = nv50_tic_entry(nvc0->textures[s][i]);
      struct nv04_resource *res;
      const boolean dirty = !!(nvc0->textures_dirty[s] & (1 << i));

      if (!tic) {
         if (dirty)
            commands[n++] = (i << 1) | 0;
         continue;
      }
      res = nv04_resource(tic->pipe.texture);

      if (tic->id < 0) {
         tic->id = nvc0_screen_tic_alloc(nvc0->screen, tic);

         PUSH_SPACE(push, 17);
         BEGIN_NVC0(push, NVC0_M2MF(OFFSET_OUT_HIGH), 2);
         PUSH_DATAh(push, txc->offset + (tic->id * 32));
         PUSH_DATA (push, txc->offset + (tic->id * 32));
         BEGIN_NVC0(push, NVC0_M2MF(LINE_LENGTH_IN), 2);
         PUSH_DATA (push, 32);
         PUSH_DATA (push, 1);
         BEGIN_NVC0(push, NVC0_M2MF(EXEC), 1);
         PUSH_DATA (push, 0x100111);
         BEGIN_NIC0(push, NVC0_M2MF(DATA), 8);
         PUSH_DATAp(push, &tic->tic[0], 8);

         need_flush = TRUE;
      } else
      if (res->status & NOUVEAU_BUFFER_STATUS_GPU_WRITING) {
         BEGIN_NVC0(push, NVC0_3D(TEX_CACHE_CTL), 1);
         PUSH_DATA (push, (tic->id << 4) | 1);
      }
      nvc0->screen->tic.lock[tic->id / 32] |= 1 << (tic->id % 32);

      res->status &= ~NOUVEAU_BUFFER_STATUS_GPU_WRITING;
      res->status |=  NOUVEAU_BUFFER_STATUS_GPU_READING;

      if (!dirty)
         continue;
      commands[n++] = (tic->id << 9) | (i << 1) | 1;

      BCTX_REFN(nvc0->bufctx_3d, TEX(s, i), res, RD);
   }
   for (; i < nvc0->state.num_textures[s]; ++i)
      commands[n++] = (i << 1) | 0;

   nvc0->state.num_textures[s] = nvc0->num_textures[s];

   if (n) {
      BEGIN_NIC0(push, NVC0_3D(BIND_TIC(s)), n);
      PUSH_DATAp(push, commands, n);
   }
   nvc0->textures_dirty[s] = 0;

   return need_flush;
}

void nvc0_validate_textures(struct nvc0_context *nvc0)
{
   boolean need_flush;

   need_flush  = nvc0_validate_tic(nvc0, 0);
   need_flush |= nvc0_validate_tic(nvc0, 3);
   need_flush |= nvc0_validate_tic(nvc0, 4);

   if (need_flush) {
      BEGIN_NVC0(nvc0->base.pushbuf, NVC0_3D(TIC_FLUSH), 1);
      PUSH_DATA (nvc0->base.pushbuf, 0);
   }
}

static boolean
nvc0_validate_tsc(struct nvc0_context *nvc0, int s)
{
   uint32_t commands[16];
   struct nouveau_pushbuf *push = nvc0->base.pushbuf;
   unsigned i;
   unsigned n = 0;
   boolean need_flush = FALSE;

   for (i = 0; i < nvc0->num_samplers[s]; ++i) {
      struct nv50_tsc_entry *tsc = nv50_tsc_entry(nvc0->samplers[s][i]);

      if (!(nvc0->samplers_dirty[s] & (1 << i)))
         continue;
      if (!tsc) {
         commands[n++] = (i << 4) | 0;
         continue;
      }
      if (tsc->id < 0) {
         tsc->id = nvc0_screen_tsc_alloc(nvc0->screen, tsc);

         nvc0_m2mf_push_linear(&nvc0->base, nvc0->screen->txc,
                               65536 + tsc->id * 32, NOUVEAU_BO_VRAM,
                               32, tsc->tsc);
         need_flush = TRUE;
      }
      nvc0->screen->tsc.lock[tsc->id / 32] |= 1 << (tsc->id % 32);

      commands[n++] = (tsc->id << 12) | (i << 4) | 1;
   }
   for (; i < nvc0->state.num_samplers[s]; ++i)
      commands[n++] = (i << 4) | 0;

   nvc0->state.num_samplers[s] = nvc0->num_samplers[s];

   if (n) {
      BEGIN_NIC0(push, NVC0_3D(BIND_TSC(s)), n);
      PUSH_DATAp(push, commands, n);
   }
   nvc0->samplers_dirty[s] = 0;

   return need_flush;
}

void nvc0_validate_samplers(struct nvc0_context *nvc0)
{
   boolean need_flush;

   need_flush  = nvc0_validate_tsc(nvc0, 0);
   need_flush |= nvc0_validate_tsc(nvc0, 3);
   need_flush |= nvc0_validate_tsc(nvc0, 4);

   if (need_flush) {
      BEGIN_NVC0(nvc0->base.pushbuf, NVC0_3D(TSC_FLUSH), 1);
      PUSH_DATA (nvc0->base.pushbuf, 0);
   }
}