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
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
|
/**********************************************************
* Copyright 2008-2009 VMware, Inc. All rights reserved.
*
* 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 OR COPYRIGHT HOLDERS
* 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 "util/u_inlines.h"
#include "pipe/p_defines.h"
#include "svga_screen.h"
#include "svga_context.h"
#include "svga_state.h"
#include "svga_cmd.h"
#include "svga_tgsi.h"
#include "svga_debug.h"
#include "svga_hw_reg.h"
/*
* Don't try to send more than 4kb of successive constants.
*/
#define MAX_CONST_REG_COUNT 256 /**< number of float[4] constants */
/**
* Convert from PIPE_SHADER_* to SVGA3D_SHADERTYPE_*
*/
static unsigned
svga_shader_type(unsigned shader)
{
switch (shader) {
case PIPE_SHADER_VERTEX:
return SVGA3D_SHADERTYPE_VS;
case PIPE_SHADER_FRAGMENT:
return SVGA3D_SHADERTYPE_PS;
default:
assert(!"Unexpected shader type");
return SVGA3D_SHADERTYPE_VS;
}
}
/**
* Check and emit one shader constant register.
* \param shader PIPE_SHADER_FRAGMENT or PIPE_SHADER_VERTEX
* \param i which float[4] constant to change
* \param value the new float[4] value
*/
static enum pipe_error
emit_const(struct svga_context *svga, unsigned shader, unsigned i,
const float *value)
{
enum pipe_error ret = PIPE_OK;
assert(shader < PIPE_SHADER_TYPES);
assert(i < SVGA3D_CONSTREG_MAX);
if (memcmp(svga->state.hw_draw.cb[shader][i], value,
4 * sizeof(float)) != 0) {
if (SVGA_DEBUG & DEBUG_CONSTS)
debug_printf("%s %s %u: %f %f %f %f\n",
__FUNCTION__,
shader == PIPE_SHADER_VERTEX ? "VERT" : "FRAG",
i,
value[0],
value[1],
value[2],
value[3]);
ret = SVGA3D_SetShaderConst( svga->swc,
i,
svga_shader_type(shader),
SVGA3D_CONST_TYPE_FLOAT,
value );
if (ret != PIPE_OK)
return ret;
memcpy(svga->state.hw_draw.cb[shader][i], value, 4 * sizeof(float));
}
return ret;
}
/*
* Check and emit a range of shader constant registers, trying to coalesce
* successive shader constant updates in a single command in order to save
* space on the command buffer. This is a HWv8 feature.
*/
static enum pipe_error
emit_const_range(struct svga_context *svga,
unsigned shader,
unsigned offset,
unsigned count,
const float (*values)[4])
{
unsigned i, j;
enum pipe_error ret;
#ifdef DEBUG
if (offset + count > SVGA3D_CONSTREG_MAX) {
debug_printf("svga: too many constants (offset + count = %u)\n",
offset + count);
}
#endif
if (offset > SVGA3D_CONSTREG_MAX) {
/* This isn't OK, but if we propagate an error all the way up we'll
* just get into more trouble.
* XXX note that offset is always zero at this time so this is moot.
*/
return PIPE_OK;
}
if (offset + count > SVGA3D_CONSTREG_MAX) {
/* Just drop the extra constants for now.
* Ideally we should not have allowed the app to create a shader
* that exceeds our constant buffer size but there's no way to
* express that in gallium at this time.
*/
count = SVGA3D_CONSTREG_MAX - offset;
}
i = 0;
while (i < count) {
if (memcmp(svga->state.hw_draw.cb[shader][offset + i],
values[i],
4 * sizeof(float)) != 0) {
/* Found one dirty constant
*/
if (SVGA_DEBUG & DEBUG_CONSTS)
debug_printf("%s %s %d: %f %f %f %f\n",
__FUNCTION__,
shader == PIPE_SHADER_VERTEX ? "VERT" : "FRAG",
offset + i,
values[i][0],
values[i][1],
values[i][2],
values[i][3]);
/* Look for more consecutive dirty constants.
*/
j = i + 1;
while (j < count &&
j < i + MAX_CONST_REG_COUNT &&
memcmp(svga->state.hw_draw.cb[shader][offset + j],
values[j],
4 * sizeof(float)) != 0) {
if (SVGA_DEBUG & DEBUG_CONSTS)
debug_printf("%s %s %d: %f %f %f %f\n",
__FUNCTION__,
shader == PIPE_SHADER_VERTEX ? "VERT" : "FRAG",
offset + j,
values[j][0],
values[j][1],
values[j][2],
values[j][3]);
++j;
}
assert(j >= i + 1);
/* Send them all together.
*/
ret = SVGA3D_SetShaderConsts(svga->swc,
offset + i, j - i,
svga_shader_type(shader),
SVGA3D_CONST_TYPE_FLOAT,
values + i);
if (ret != PIPE_OK) {
return ret;
}
/*
* Local copy of the hardware state.
*/
memcpy(svga->state.hw_draw.cb[shader][offset + i],
values[i],
(j - i) * 4 * sizeof(float));
i = j + 1;
} else {
++i;
}
}
return PIPE_OK;
}
/**
* Emit all the constants in a constant buffer for a shader stage.
*/
static enum pipe_error
emit_consts(struct svga_context *svga, unsigned shader)
{
struct svga_screen *ss = svga_screen(svga->pipe.screen);
struct pipe_transfer *transfer = NULL;
unsigned count;
const float (*data)[4] = NULL;
unsigned i;
enum pipe_error ret = PIPE_OK;
const unsigned offset = 0;
assert(shader < PIPE_SHADER_TYPES);
if (svga->curr.cb[shader] == NULL)
goto done;
count = svga->curr.cb[shader]->width0 / (4 * sizeof(float));
data = (const float (*)[4])pipe_buffer_map(&svga->pipe,
svga->curr.cb[shader],
PIPE_TRANSFER_READ,
&transfer);
if (data == NULL) {
ret = PIPE_ERROR_OUT_OF_MEMORY;
goto done;
}
if (ss->hw_version >= SVGA3D_HWVERSION_WS8_B1) {
ret = emit_const_range( svga, shader, offset, count, data );
if (ret != PIPE_OK) {
goto done;
}
} else {
for (i = 0; i < count; i++) {
ret = emit_const( svga, shader, offset + i, data[i] );
if (ret != PIPE_OK) {
goto done;
}
}
}
done:
if (data)
pipe_buffer_unmap(&svga->pipe, transfer);
return ret;
}
static enum pipe_error
emit_fs_consts(struct svga_context *svga, unsigned dirty)
{
const struct svga_shader_variant *variant = svga->state.hw_draw.fs;
enum pipe_error ret = PIPE_OK;
ret = emit_consts( svga, PIPE_SHADER_FRAGMENT );
if (ret != PIPE_OK)
return ret;
/* The internally generated fragment shader for xor blending
* doesn't have a 'variant' struct. It should be fixed to avoid
* this special case, but work around it with a NULL check:
*/
if (variant) {
const struct svga_fs_compile_key *key = &variant->key.fkey;
if (key->num_unnormalized_coords) {
const unsigned offset =
variant->shader->info.file_max[TGSI_FILE_CONSTANT] + 1;
unsigned i;
for (i = 0; i < key->num_textures; i++) {
if (key->tex[i].unnormalized) {
struct pipe_resource *tex = svga->curr.sampler_views[i]->texture;
float data[4];
data[0] = 1.0f / (float) tex->width0;
data[1] = 1.0f / (float) tex->height0;
data[2] = 1.0f;
data[3] = 1.0f;
ret = emit_const(svga,
PIPE_SHADER_FRAGMENT,
key->tex[i].width_height_idx + offset,
data);
if (ret != PIPE_OK) {
return ret;
}
}
}
}
}
return PIPE_OK;
}
struct svga_tracked_state svga_hw_fs_constants =
{
"hw fs params",
(SVGA_NEW_FS_CONST_BUFFER |
SVGA_NEW_FS_VARIANT |
SVGA_NEW_TEXTURE_BINDING),
emit_fs_consts
};
static enum pipe_error
emit_vs_consts(struct svga_context *svga, unsigned dirty)
{
const struct svga_shader_variant *variant = svga->state.hw_draw.vs;
const struct svga_vs_compile_key *key;
enum pipe_error ret = PIPE_OK;
unsigned offset;
/* SVGA_NEW_VS_VARIANT
*/
if (variant == NULL)
return PIPE_OK;
key = &variant->key.vkey;
/* SVGA_NEW_VS_CONST_BUFFER
*/
ret = emit_consts( svga, PIPE_SHADER_VERTEX );
if (ret != PIPE_OK)
return ret;
/* offset = number of constants in the VS const buffer */
offset = variant->shader->info.file_max[TGSI_FILE_CONSTANT] + 1;
/* SVGA_NEW_VS_PRESCALE
* Put the viewport pre-scale/translate values into the const buffer.
*/
if (key->need_prescale) {
ret = emit_const( svga, PIPE_SHADER_VERTEX, offset++,
svga->state.hw_clear.prescale.scale );
if (ret != PIPE_OK)
return ret;
ret = emit_const( svga, PIPE_SHADER_VERTEX, offset++,
svga->state.hw_clear.prescale.translate );
if (ret != PIPE_OK)
return ret;
}
return PIPE_OK;
}
struct svga_tracked_state svga_hw_vs_constants =
{
"hw vs params",
(SVGA_NEW_PRESCALE |
SVGA_NEW_VS_CONST_BUFFER |
SVGA_NEW_VS_VARIANT),
emit_vs_consts
};
|