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
|
/*
* Copyright © 2012 Intel Corporation
*
* 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.
*/
/** @file gen8_instruction.h
*
* A representation of a Gen8+ EU instruction, with helper methods to get
* and set various fields. This is the actual hardware format.
*/
#ifndef GEN8_INSTRUCTION_H
#define GEN8_INSTRUCTION_H
#include <stdio.h>
#include <stdint.h>
#include "brw_compat.h"
#include "brw_reg.h"
struct gen8_instruction {
uint32_t data[4];
};
static inline unsigned gen8_bits(struct gen8_instruction *insn,
unsigned high,
unsigned low);
static inline void gen8_set_bits(struct gen8_instruction *insn,
unsigned high,
unsigned low,
unsigned value);
#define F(name, high, low) \
static inline void gen8_set_##name(struct gen8_instruction *insn, unsigned v) \
{ \
gen8_set_bits(insn, high, low, v); \
} \
static inline unsigned gen8_##name(struct gen8_instruction *insn) \
{ \
return gen8_bits(insn, high, low); \
}
/**
* Direct addressing only:
* @{
*/
F(src1_da_reg_nr, 108, 101);
F(src0_da_reg_nr, 76, 69);
F(dst_da1_hstride, 62, 61);
F(dst_da_reg_nr, 60, 53);
F(dst_da16_subreg_nr, 52, 52);
F(dst_da1_subreg_nr, 52, 48);
F(da16_writemask, 51, 48); /* Dst.ChanEn */
/** @} */
F(src1_vert_stride, 120, 117)
F(src1_da1_width, 116, 114)
F(src1_da16_swiz_w, 115, 114)
F(src1_da16_swiz_z, 113, 112)
F(src1_da1_hstride, 113, 112)
F(src1_address_mode, 111, 111)
/** Src1.SrcMod @{ */
F(src1_negate, 110, 110)
F(src1_abs, 109, 109)
/** @} */
F(src1_da16_subreg_nr, 100, 100)
F(src1_da1_subreg_nr, 100, 96)
F(src1_da16_swiz_y, 99, 98)
F(src1_da16_swiz_x, 97, 96)
F(src1_reg_type, 94, 91)
F(src1_reg_file, 90, 89)
F(src0_vert_stride, 88, 85)
F(src0_da1_width, 84, 82)
F(src0_da16_swiz_w, 83, 82)
F(src0_da16_swiz_z, 81, 80)
F(src0_da1_hstride, 81, 80)
F(src0_address_mode, 79, 79)
/** Src0.SrcMod @{ */
F(src0_negate, 78, 78)
F(src0_abs, 77, 77)
/** @} */
F(src0_da16_subreg_nr, 68, 68)
F(src0_da1_subreg_nr, 68, 64)
F(src0_da16_swiz_y, 67, 66)
F(src0_da16_swiz_x, 65, 64)
F(dst_address_mode, 63, 63)
F(src0_reg_type, 46, 43)
F(src0_reg_file, 42, 41)
F(dst_reg_type, 40, 37)
F(dst_reg_file, 36, 35)
F(mask_control, 34, 34)
F(flag_reg_nr, 33, 33)
F(flag_subreg_nr, 32, 32)
F(saturate, 31, 31)
F(branch_control, 30, 30)
F(debug_control, 30, 30)
F(cmpt_control, 29, 29)
F(acc_wr_control, 28, 28)
F(cond_modifier, 27, 24)
F(exec_size, 23, 21)
F(pred_inv, 20, 20)
F(pred_control, 19, 16)
F(thread_control, 15, 14)
F(qtr_control, 13, 12)
F(nib_control, 11, 11)
F(dep_control, 10, 9)
F(access_mode, 8, 8)
/* Bit 7 is Reserve d (for future Opcode expansion) */
F(opcode, 6, 0)
/**
* Three-source instructions:
* @{
*/
F(src2_3src_reg_nr, 125, 118)
F(src2_3src_subreg_nr, 117, 115)
F(src2_3src_swizzle, 114, 107)
F(src2_3src_rep_ctrl, 106, 106)
F(src1_3src_reg_nr, 104, 97)
F(src1_3src_subreg_hi, 96, 96)
F(src1_3src_subreg_lo, 95, 94)
F(src1_3src_swizzle, 93, 86)
F(src1_3src_rep_ctrl, 85, 85)
F(src0_3src_reg_nr, 83, 76)
F(src0_3src_subreg_nr, 75, 73)
F(src0_3src_swizzle, 72, 65)
F(src0_3src_rep_ctrl, 64, 64)
F(dst_3src_reg_nr, 63, 56)
F(dst_3src_subreg_nr, 55, 53)
F(dst_3src_writemask, 52, 49)
F(dst_3src_type, 48, 46)
F(src_3src_type, 45, 43)
F(src2_3src_negate, 42, 42)
F(src2_3src_abs, 41, 41)
F(src1_3src_negate, 40, 40)
F(src1_3src_abs, 39, 39)
F(src0_3src_negate, 38, 38)
F(src0_3src_abs, 37, 37)
/** @} */
/**
* Fields for SEND messages:
* @{
*/
F(eot, 127, 127)
F(mlen, 124, 121)
F(rlen, 120, 116)
F(header_present, 115, 115)
F(function_control, 114, 96)
F(sfid, 27, 24)
F(math_function, 27, 24)
/** @} */
/**
* URB message function control bits:
* @{
*/
F(urb_per_slot_offset, 113, 113)
F(urb_interleave, 111, 111)
F(urb_global_offset, 110, 100)
F(urb_opcode, 99, 96)
/** @} */
/**
* Sampler message function control bits:
* @{
*/
F(sampler_simd_mode, 114, 113)
F(sampler_msg_type, 112, 108)
F(sampler, 107, 104)
F(binding_table_index, 103, 96)
/** @} */
/**
* Data port message function control bits:
* @ {
*/
F(dp_category, 114, 114)
F(dp_message_type, 113, 110)
F(dp_message_control, 109, 104)
F(dp_binding_table_index, 103, 96)
/** @} */
/**
* Thread Spawn message function control bits:
* @ {
*/
F(ts_resource_select, 100, 100)
F(ts_request_type, 97, 97)
F(ts_opcode, 96, 96)
/** @} */
/**
* Video Motion Estimation message function control bits:
* @ {
*/
F(vme_message_type, 110, 109)
F(vme_binding_table_index, 103, 96)
/** @} */
/**
* Check & Refinement Engine message function control bits:
* @ {
*/
F(cre_message_type, 110, 109)
F(cre_binding_table_index, 103, 96)
/** @} */
/* Addr Mode */
F(dst_addr_mode, 63, 63)
F(src0_addr_mode, 79, 79)
F(src1_addr_mode, 111, 111)
/* Indirect access mode for Align1. */
F(dst_ida1_sub_nr, 60, 57)
F(src0_ida1_sub_nr, 76, 73)
F(src1_ida1_sub_nr, 108, 105)
/* Imm[8:0] of Immediate addr offset under Indirect mode */
F(dst_ida1_imm8, 56, 48)
F(src0_ida1_imm8, 72, 64)
F(src1_ida1_imm8, 104, 96)
/* Imm Bit9 of Immediate addr offset under Indirect mode */
F(dst_ida1_imm9, 47, 47)
F(src0_ida1_imm9, 95, 95)
F(src1_ida1_imm9, 121, 121)
#undef F
#define IMM8_MASK 0x1FF
#define IMM9_MASK 0x200
/**
* Flow control instruction bits:
* @{
*/
static inline unsigned gen8_uip(struct gen8_instruction *insn)
{
return insn->data[2];
}
static inline void gen8_set_uip(struct gen8_instruction *insn, unsigned uip)
{
insn->data[2] = uip;
}
static inline unsigned gen8_jip(struct gen8_instruction *insn)
{
return insn->data[3];
}
static inline void gen8_set_jip(struct gen8_instruction *insn, unsigned jip)
{
insn->data[3] = jip;
}
/** @} */
static inline int gen8_src1_imm_d(struct gen8_instruction *insn)
{
return insn->data[3];
}
static inline unsigned gen8_src1_imm_ud(struct gen8_instruction *insn)
{
return insn->data[3];
}
static inline float gen8_src1_imm_f(struct gen8_instruction *insn)
{
fi_type ft;
ft.u = insn->data[3];
return ft.f;
}
void gen8_set_dst(struct gen8_instruction *insn, struct brw_reg reg);
void gen8_set_src0(struct gen8_instruction *insn, struct brw_reg reg);
void gen8_set_src1(struct gen8_instruction *insn, struct brw_reg reg);
void gen8_set_urb_message(struct gen8_instruction *insn,
unsigned opcode, unsigned mlen, unsigned rlen,
bool eot, unsigned offset, bool interleave);
void gen8_set_sampler_message(struct gen8_instruction *insn,
unsigned binding_table_index, unsigned sampler,
unsigned msg_type, unsigned rlen, unsigned mlen,
bool header_present, unsigned simd_mode);
void gen8_set_dp_message(struct gen8_instruction *insn,
enum brw_message_target sfid,
unsigned binding_table_index,
unsigned msg_type,
unsigned msg_control,
unsigned msg_length,
unsigned response_length,
bool header_present,
bool end_of_thread);
/** Disassemble the instruction. */
int gen8_disassemble(FILE *file, struct gen8_instruction *insn, int gen);
/**
* Fetch a set of contiguous bits from the instruction.
*
* Bits indexes range from 0..127; fields may not cross 32-bit boundaries.
*/
static inline unsigned
gen8_bits(struct gen8_instruction *insn, unsigned high, unsigned low)
{
/* We assume the field doesn't cross 32-bit boundaries. */
const unsigned word = high / 32;
assert(word == low / 32);
high %= 32;
low %= 32;
const unsigned mask = (((1 << (high - low + 1)) - 1) << low);
return (insn->data[word] & mask) >> low;
}
/**
* Set bits in the instruction, with proper shifting and masking.
*
* Bits indexes range from 0..127; fields may not cross 32-bit boundaries.
*/
static inline void
gen8_set_bits(struct gen8_instruction *insn,
unsigned high,
unsigned low,
unsigned value)
{
const unsigned word = high / 32;
assert(word == low / 32);
high %= 32;
low %= 32;
const unsigned mask = (((1 << (high - low + 1)) - 1) << low);
insn->data[word] = (insn->data[word] & ~mask) | ((value << low) & mask);
}
#endif
|