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
|
/*
* Copyright 2014 Advanced Micro Devices, 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 (including the next
* paragraph) 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 "si_build_pm4.h"
/* For MSAA sample positions. */
#define FILL_SREG(s0x, s0y, s1x, s1y, s2x, s2y, s3x, s3y) \
((((unsigned)(s0x) & 0xf) << 0) | (((unsigned)(s0y) & 0xf) << 4) | \
(((unsigned)(s1x) & 0xf) << 8) | (((unsigned)(s1y) & 0xf) << 12) | \
(((unsigned)(s2x) & 0xf) << 16) | (((unsigned)(s2y) & 0xf) << 20) | \
(((unsigned)(s3x) & 0xf) << 24) | (((unsigned)(s3y) & 0xf) << 28))
/* For obtaining location coordinates from registers */
#define SEXT4(x) ((int)((x) | ((x) & 0x8 ? 0xfffffff0 : 0)))
#define GET_SFIELD(reg, index) SEXT4(((reg) >> ((index) * 4)) & 0xf)
#define GET_SX(reg, index) GET_SFIELD((reg)[(index) / 4], ((index) % 4) * 2)
#define GET_SY(reg, index) GET_SFIELD((reg)[(index) / 4], ((index) % 4) * 2 + 1)
/* The following sample ordering is required by EQAA.
*
* Sample 0 is approx. in the top-left quadrant.
* Sample 1 is approx. in the bottom-right quadrant.
*
* Sample 2 is approx. in the bottom-left quadrant.
* Sample 3 is approx. in the top-right quadrant.
* (sample I={2,3} adds more detail to the vicinity of sample I-2)
*
* Sample 4 is approx. in the same quadrant as sample 0. (top-left)
* Sample 5 is approx. in the same quadrant as sample 1. (bottom-right)
* Sample 6 is approx. in the same quadrant as sample 2. (bottom-left)
* Sample 7 is approx. in the same quadrant as sample 3. (top-right)
* (sample I={4,5,6,7} adds more detail to the vicinity of sample I-4)
*
* The next 8 samples add more detail to the vicinity of the previous samples.
* (sample I (I >= 8) adds more detail to the vicinity of sample I-8)
*
* The ordering is specified such that:
* If we take the first 2 samples, we should get good 2x MSAA.
* If we add 2 more samples, we should get good 4x MSAA with the same sample locations.
* If we add 4 more samples, we should get good 8x MSAA with the same sample locations.
* If we add 8 more samples, we should get perfect 16x MSAA with the same sample locations.
*
* The ordering also allows finding samples in the same vicinity.
*
* Group N of 2 samples in the same vicinity in 16x MSAA: {N,N+8}
* Group N of 2 samples in the same vicinity in 8x MSAA: {N,N+4}
* Group N of 2 samples in the same vicinity in 4x MSAA: {N,N+2}
*
* Groups of 4 samples in the same vicinity in 16x MSAA:
* Top left: {0,4,8,12}
* Bottom right: {1,5,9,13}
* Bottom left: {2,6,10,14}
* Top right: {3,7,11,15}
*
* Groups of 4 samples in the same vicinity in 8x MSAA:
* Left half: {0,2,4,6}
* Right half: {1,3,5,7}
*
* Groups of 8 samples in the same vicinity in 16x MSAA:
* Left half: {0,2,4,6,8,10,12,14}
* Right half: {1,3,5,7,9,11,13,15}
*/
/* 1x MSAA */
static const uint32_t sample_locs_1x =
FILL_SREG( 0, 0, 0, 0, 0, 0, 0, 0); /* S1, S2, S3 fields are not used by 1x */
static const uint64_t centroid_priority_1x = 0x0000000000000000ull;
/* 2x MSAA */
static const uint32_t sample_locs_2x =
FILL_SREG(-4,-4, 4, 4, 0, 0, 0, 0); /* S2 & S3 fields are not used by 2x MSAA */
static const uint64_t centroid_priority_2x = 0x1010101010101010ull;
/* 4x, 8x, and 16x MSAA
* - The first 4 locations happen to be optimal for 4x MSAA, better than
* the standard DX 4x locations.
* - The first 8 locations happen to be almost as good as 8x DX locations,
* but the DX locations are horrible for worst-case EQAA 8s4f and 8s2f.
*/
static const uint32_t sample_locs_4x_8x_16x[] = {
FILL_SREG(-5,-2, 5, 3, -2, 6, 3,-5),
FILL_SREG(-6,-7, 1, 1, -6, 4, 7,-3),
FILL_SREG(-1,-3, 6, 7, -3, 2, 0,-7),
FILL_SREG(-4,-6, 2, 5, -8, 0, 4,-1),
};
static const uint64_t centroid_priority_4x = 0x2310231023102310ull;
static const uint64_t centroid_priority_8x = 0x4762310547623105ull;
static const uint64_t centroid_priority_16x = 0x49e7c6b231d0fa85ull;
static void si_get_sample_position(struct pipe_context *ctx, unsigned sample_count,
unsigned sample_index, float *out_value)
{
const uint32_t *sample_locs;
switch (sample_count) {
case 1:
default:
sample_locs = &sample_locs_1x;
break;
case 2:
sample_locs = &sample_locs_2x;
break;
case 4:
case 8:
case 16:
sample_locs = sample_locs_4x_8x_16x;
break;
}
out_value[0] = (GET_SX(sample_locs, sample_index) + 8) / 16.0f;
out_value[1] = (GET_SY(sample_locs, sample_index) + 8) / 16.0f;
}
static void si_emit_max_4_sample_locs(struct radeon_winsys_cs *cs,
uint64_t centroid_priority,
uint32_t sample_locs)
{
radeon_set_context_reg_seq(cs, R_028BD4_PA_SC_CENTROID_PRIORITY_0, 2);
radeon_emit(cs, centroid_priority);
radeon_emit(cs, centroid_priority >> 32);
radeon_set_context_reg(cs, R_028BF8_PA_SC_AA_SAMPLE_LOCS_PIXEL_X0Y0_0, sample_locs);
radeon_set_context_reg(cs, R_028C08_PA_SC_AA_SAMPLE_LOCS_PIXEL_X1Y0_0, sample_locs);
radeon_set_context_reg(cs, R_028C18_PA_SC_AA_SAMPLE_LOCS_PIXEL_X0Y1_0, sample_locs);
radeon_set_context_reg(cs, R_028C28_PA_SC_AA_SAMPLE_LOCS_PIXEL_X1Y1_0, sample_locs);
}
static void si_emit_max_16_sample_locs(struct radeon_winsys_cs *cs,
uint64_t centroid_priority,
const uint32_t *sample_locs,
unsigned num_samples)
{
radeon_set_context_reg_seq(cs, R_028BD4_PA_SC_CENTROID_PRIORITY_0, 2);
radeon_emit(cs, centroid_priority);
radeon_emit(cs, centroid_priority >> 32);
radeon_set_context_reg_seq(cs, R_028BF8_PA_SC_AA_SAMPLE_LOCS_PIXEL_X0Y0_0,
num_samples == 8 ? 14 : 16);
radeon_emit_array(cs, sample_locs, 4);
radeon_emit_array(cs, sample_locs, 4);
radeon_emit_array(cs, sample_locs, 4);
radeon_emit_array(cs, sample_locs, num_samples == 8 ? 2 : 4);
}
void si_emit_sample_locations(struct radeon_winsys_cs *cs, int nr_samples)
{
switch (nr_samples) {
default:
case 1:
si_emit_max_4_sample_locs(cs, centroid_priority_1x, sample_locs_1x);
break;
case 2:
si_emit_max_4_sample_locs(cs, centroid_priority_2x, sample_locs_2x);
break;
case 4:
si_emit_max_4_sample_locs(cs, centroid_priority_4x, sample_locs_4x_8x_16x[0]);
break;
case 8:
si_emit_max_16_sample_locs(cs, centroid_priority_8x, sample_locs_4x_8x_16x, 8);
break;
case 16:
si_emit_max_16_sample_locs(cs, centroid_priority_16x, sample_locs_4x_8x_16x, 16);
break;
}
}
void si_init_msaa_functions(struct si_context *sctx)
{
int i;
sctx->b.get_sample_position = si_get_sample_position;
si_get_sample_position(&sctx->b, 1, 0, sctx->sample_locations_1x[0]);
for (i = 0; i < 2; i++)
si_get_sample_position(&sctx->b, 2, i, sctx->sample_locations_2x[i]);
for (i = 0; i < 4; i++)
si_get_sample_position(&sctx->b, 4, i, sctx->sample_locations_4x[i]);
for (i = 0; i < 8; i++)
si_get_sample_position(&sctx->b, 8, i, sctx->sample_locations_8x[i]);
for (i = 0; i < 16; i++)
si_get_sample_position(&sctx->b, 16, i, sctx->sample_locations_16x[i]);
}
|
