blob: c1f2208f5133e3ae3cba1330d1f3415d63b50c81 (
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
|
#ifndef __NV40_SHADER_H__
#define __NV40_SHADER_H__
/* Vertex programs instruction set
*
* The NV40 instruction set is very similar to NV30. Most fields are in
* a slightly different position in the instruction however.
*
* Merged instructions
* In some cases it is possible to put two instructions into one opcode
* slot. The rules for when this is OK is not entirely clear to me yet.
*
* There are separate writemasks and dest temp register fields for each
* grouping of instructions. There is however only one field with the
* ID of a result register. Writing to temp/result regs is selected by
* setting VEC_RESULT/SCA_RESULT.
*
* Temporary registers
* The source/dest temp register fields have been extended by 1 bit, to
* give a total of 32 temporary registers.
*
* Relative Addressing
* NV40 can use an address register to index into vertex attribute regs.
* This is done by putting the offset value into INPUT_SRC and setting
* the INDEX_INPUT flag.
*
* Conditional execution (see NV_vertex_program{2,3} for details)
* There is a second condition code register on NV40, it's use is enabled
* by setting the COND_REG_SELECT_1 flag.
*
* Texture lookup
* TODO
*/
/* ---- OPCODE BITS 127:96 / data DWORD 0 --- */
#define NV40_VP_INST_VEC_RESULT (1 << 30)
/* uncertain.. */
#define NV40_VP_INST_COND_UPDATE_ENABLE ((1 << 14)|1<<29)
/* use address reg as index into attribs */
#define NV40_VP_INST_INDEX_INPUT (1 << 27)
#define NV40_VP_INST_SATURATE (1 << 26)
#define NV40_VP_INST_COND_REG_SELECT_1 (1 << 25)
#define NV40_VP_INST_ADDR_REG_SELECT_1 (1 << 24)
#define NV40_VP_INST_SRC2_ABS (1 << 23)
#define NV40_VP_INST_SRC1_ABS (1 << 22)
#define NV40_VP_INST_SRC0_ABS (1 << 21)
#define NV40_VP_INST_VEC_DEST_TEMP_SHIFT 15
#define NV40_VP_INST_VEC_DEST_TEMP_MASK (0x3F << 15)
#define NV40_VP_INST_COND_TEST_ENABLE (1 << 13)
#define NV40_VP_INST_COND_SHIFT 10
#define NV40_VP_INST_COND_MASK (0x7 << 10)
#define NV40_VP_INST_COND_SWZ_X_SHIFT 8
#define NV40_VP_INST_COND_SWZ_X_MASK (3 << 8)
#define NV40_VP_INST_COND_SWZ_Y_SHIFT 6
#define NV40_VP_INST_COND_SWZ_Y_MASK (3 << 6)
#define NV40_VP_INST_COND_SWZ_Z_SHIFT 4
#define NV40_VP_INST_COND_SWZ_Z_MASK (3 << 4)
#define NV40_VP_INST_COND_SWZ_W_SHIFT 2
#define NV40_VP_INST_COND_SWZ_W_MASK (3 << 2)
#define NV40_VP_INST_COND_SWZ_ALL_SHIFT 2
#define NV40_VP_INST_COND_SWZ_ALL_MASK (0xFF << 2)
#define NV40_VP_INST_ADDR_SWZ_SHIFT 0
#define NV40_VP_INST_ADDR_SWZ_MASK (0x03 << 0)
#define NV40_VP_INST0_KNOWN ( \
NV40_VP_INST_INDEX_INPUT | \
NV40_VP_INST_COND_REG_SELECT_1 | \
NV40_VP_INST_ADDR_REG_SELECT_1 | \
NV40_VP_INST_SRC2_ABS | \
NV40_VP_INST_SRC1_ABS | \
NV40_VP_INST_SRC0_ABS | \
NV40_VP_INST_VEC_DEST_TEMP_MASK | \
NV40_VP_INST_COND_TEST_ENABLE | \
NV40_VP_INST_COND_MASK | \
NV40_VP_INST_COND_SWZ_ALL_MASK | \
NV40_VP_INST_ADDR_SWZ_MASK)
/* ---- OPCODE BITS 95:64 / data DWORD 1 --- */
#define NV40_VP_INST_VEC_OPCODE_SHIFT 22
#define NV40_VP_INST_VEC_OPCODE_MASK (0x1F << 22)
#define NV40_VP_INST_SCA_OPCODE_SHIFT 27
#define NV40_VP_INST_SCA_OPCODE_MASK (0x1F << 27)
#define NV40_VP_INST_CONST_SRC_SHIFT 12
#define NV40_VP_INST_CONST_SRC_MASK (0xFF << 12)
#define NV40_VP_INST_INPUT_SRC_SHIFT 8
#define NV40_VP_INST_INPUT_SRC_MASK (0x0F << 8)
#define NV40_VP_INST_SRC0H_SHIFT 0
#define NV40_VP_INST_SRC0H_MASK (0xFF << 0)
#define NV40_VP_INST1_KNOWN ( \
NV40_VP_INST_VEC_OPCODE_MASK | \
NV40_VP_INST_SCA_OPCODE_MASK | \
NV40_VP_INST_CONST_SRC_MASK | \
NV40_VP_INST_INPUT_SRC_MASK | \
NV40_VP_INST_SRC0H_MASK \
)
/* ---- OPCODE BITS 63:32 / data DWORD 2 --- */
#define NV40_VP_INST_SRC0L_SHIFT 23
#define NV40_VP_INST_SRC0L_MASK (0x1FF << 23)
#define NV40_VP_INST_SRC1_SHIFT 6
#define NV40_VP_INST_SRC1_MASK (0x1FFFF << 6)
#define NV40_VP_INST_SRC2H_SHIFT 0
#define NV40_VP_INST_SRC2H_MASK (0x3F << 0)
#define NV40_VP_INST_IADDRH_SHIFT 0
#define NV40_VP_INST_IADDRH_MASK (0x3F << 0)
/* ---- OPCODE BITS 31:0 / data DWORD 3 --- */
#define NV40_VP_INST_IADDRL_SHIFT 29
#define NV40_VP_INST_IADDRL_MASK (7 << 29)
#define NV40_VP_INST_SRC2L_SHIFT 21
#define NV40_VP_INST_SRC2L_MASK (0x7FF << 21)
#define NV40_VP_INST_SCA_WRITEMASK_SHIFT 17
#define NV40_VP_INST_SCA_WRITEMASK_MASK (0xF << 17)
# define NV40_VP_INST_SCA_WRITEMASK_X (1 << 20)
# define NV40_VP_INST_SCA_WRITEMASK_Y (1 << 19)
# define NV40_VP_INST_SCA_WRITEMASK_Z (1 << 18)
# define NV40_VP_INST_SCA_WRITEMASK_W (1 << 17)
#define NV40_VP_INST_VEC_WRITEMASK_SHIFT 13
#define NV40_VP_INST_VEC_WRITEMASK_MASK (0xF << 13)
# define NV40_VP_INST_VEC_WRITEMASK_X (1 << 16)
# define NV40_VP_INST_VEC_WRITEMASK_Y (1 << 15)
# define NV40_VP_INST_VEC_WRITEMASK_Z (1 << 14)
# define NV40_VP_INST_VEC_WRITEMASK_W (1 << 13)
#define NV40_VP_INST_SCA_RESULT (1 << 12)
#define NV40_VP_INST_SCA_DEST_TEMP_SHIFT 7
#define NV40_VP_INST_SCA_DEST_TEMP_MASK (0x1F << 7)
#define NV40_VP_INST_DEST_SHIFT 2
#define NV40_VP_INST_DEST_MASK (31 << 2)
# define NV40_VP_INST_DEST_POS 0
# define NV40_VP_INST_DEST_COL0 1
# define NV40_VP_INST_DEST_COL1 2
# define NV40_VP_INST_DEST_BFC0 3
# define NV40_VP_INST_DEST_BFC1 4
# define NV40_VP_INST_DEST_FOGC 5
# define NV40_VP_INST_DEST_PSZ 6
# define NV40_VP_INST_DEST_TC0 7
# define NV40_VP_INST_DEST_TC(n) (7+n)
# define NV40_VP_INST_DEST_TEMP 0x1F
#define NV40_VP_INST_INDEX_CONST (1 << 1)
#define NV40_VP_INST3_KNOWN ( \
NV40_VP_INST_SRC2L_MASK |\
NV40_VP_INST_SCA_WRITEMASK_MASK |\
NV40_VP_INST_VEC_WRITEMASK_MASK |\
NV40_VP_INST_SCA_DEST_TEMP_MASK |\
NV40_VP_INST_DEST_MASK |\
NV40_VP_INST_INDEX_CONST)
/* Useful to split the source selection regs into their pieces */
#define NV40_VP_SRC0_HIGH_SHIFT 9
#define NV40_VP_SRC0_HIGH_MASK 0x0001FE00
#define NV40_VP_SRC0_LOW_MASK 0x000001FF
#define NV40_VP_SRC2_HIGH_SHIFT 11
#define NV40_VP_SRC2_HIGH_MASK 0x0001F800
#define NV40_VP_SRC2_LOW_MASK 0x000007FF
/* Source selection - these are the bits you fill NV40_VP_INST_SRCn with */
#define NV40_VP_SRC_NEGATE (1 << 16)
#define NV40_VP_SRC_SWZ_X_SHIFT 14
#define NV40_VP_SRC_SWZ_X_MASK (3 << 14)
#define NV40_VP_SRC_SWZ_Y_SHIFT 12
#define NV40_VP_SRC_SWZ_Y_MASK (3 << 12)
#define NV40_VP_SRC_SWZ_Z_SHIFT 10
#define NV40_VP_SRC_SWZ_Z_MASK (3 << 10)
#define NV40_VP_SRC_SWZ_W_SHIFT 8
#define NV40_VP_SRC_SWZ_W_MASK (3 << 8)
#define NV40_VP_SRC_SWZ_ALL_SHIFT 8
#define NV40_VP_SRC_SWZ_ALL_MASK (0xFF << 8)
#define NV40_VP_SRC_TEMP_SRC_SHIFT 2
#define NV40_VP_SRC_TEMP_SRC_MASK (0x1F << 2)
#define NV40_VP_SRC_REG_TYPE_SHIFT 0
#define NV40_VP_SRC_REG_TYPE_MASK (3 << 0)
# define NV40_VP_SRC_REG_TYPE_UNK0 0
# define NV40_VP_SRC_REG_TYPE_TEMP 1
# define NV40_VP_SRC_REG_TYPE_INPUT 2
# define NV40_VP_SRC_REG_TYPE_CONST 3
#include "nvfx_shader.h"
#endif
|
