summaryrefslogtreecommitdiffstats
path: root/src/intel/vulkan/gen7_cmd_buffer.c
blob: 23327ec0724ca23dac5ed8fed95613d3a38638fc (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
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
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
446
447
448
449
450
451
452
453
454
455
456
457
458
459
460
461
462
463
464
465
466
467
468
469
470
471
472
473
474
475
476
477
478
479
480
481
482
483
484
485
486
487
488
489
490
491
492
493
494
495
496
497
498
499
500
501
502
503
504
505
506
507
508
509
510
511
512
513
514
515
516
517
518
519
520
521
522
523
524
525
526
527
528
529
530
531
532
533
534
535
536
537
538
539
540
541
542
543
544
545
546
547
548
549
550
551
552
553
554
555
556
557
558
559
560
561
562
563
564
565
566
567
568
569
570
571
572
573
574
575
576
577
578
579
580
581
582
583
584
585
586
587
588
589
/*
 * Copyright © 2015 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.
 */

#include <assert.h>
#include <stdbool.h>
#include <string.h>
#include <unistd.h>
#include <fcntl.h>

#include "anv_private.h"

#include "genxml/gen7_pack.h"
#include "genxml/gen75_pack.h"

static uint32_t
cmd_buffer_flush_push_constants(struct anv_cmd_buffer *cmd_buffer)
{
   static const uint32_t push_constant_opcodes[] = {
      [MESA_SHADER_VERTEX]                      = 21,
      [MESA_SHADER_TESS_CTRL]                   = 25, /* HS */
      [MESA_SHADER_TESS_EVAL]                   = 26, /* DS */
      [MESA_SHADER_GEOMETRY]                    = 22,
      [MESA_SHADER_FRAGMENT]                    = 23,
      [MESA_SHADER_COMPUTE]                     = 0,
   };

   VkShaderStageFlags flushed = 0;

   anv_foreach_stage(stage, cmd_buffer->state.push_constants_dirty) {
      if (stage == MESA_SHADER_COMPUTE)
         continue;

      struct anv_state state = anv_cmd_buffer_push_constants(cmd_buffer, stage);

      if (state.offset == 0)
         continue;

      anv_batch_emit(&cmd_buffer->batch, GEN7_3DSTATE_CONSTANT_VS,
                     ._3DCommandSubOpcode = push_constant_opcodes[stage],
                     .ConstantBody = {
                        .PointerToConstantBuffer0 = { .offset = state.offset },
                        .ConstantBuffer0ReadLength = DIV_ROUND_UP(state.alloc_size, 32),
                     });

      flushed |= mesa_to_vk_shader_stage(stage);
   }

   cmd_buffer->state.push_constants_dirty &= ~flushed;

   return flushed;
}

GENX_FUNC(GEN7, GEN7) void
genX(cmd_buffer_emit_descriptor_pointers)(struct anv_cmd_buffer *cmd_buffer,
                                          uint32_t stages)
{
   static const uint32_t sampler_state_opcodes[] = {
      [MESA_SHADER_VERTEX]                      = 43,
      [MESA_SHADER_TESS_CTRL]                   = 44, /* HS */
      [MESA_SHADER_TESS_EVAL]                   = 45, /* DS */
      [MESA_SHADER_GEOMETRY]                    = 46,
      [MESA_SHADER_FRAGMENT]                    = 47,
      [MESA_SHADER_COMPUTE]                     = 0,
   };

   static const uint32_t binding_table_opcodes[] = {
      [MESA_SHADER_VERTEX]                      = 38,
      [MESA_SHADER_TESS_CTRL]                   = 39,
      [MESA_SHADER_TESS_EVAL]                   = 40,
      [MESA_SHADER_GEOMETRY]                    = 41,
      [MESA_SHADER_FRAGMENT]                    = 42,
      [MESA_SHADER_COMPUTE]                     = 0,
   };

   anv_foreach_stage(s, stages) {
      if (cmd_buffer->state.samplers[s].alloc_size > 0) {
         anv_batch_emit(&cmd_buffer->batch,
                        GEN7_3DSTATE_SAMPLER_STATE_POINTERS_VS,
                        ._3DCommandSubOpcode  = sampler_state_opcodes[s],
                        .PointertoVSSamplerState = cmd_buffer->state.samplers[s].offset);
      }

      /* Always emit binding table pointers if we're asked to, since on SKL
       * this is what flushes push constants. */
      anv_batch_emit(&cmd_buffer->batch,
                     GEN7_3DSTATE_BINDING_TABLE_POINTERS_VS,
                     ._3DCommandSubOpcode  = binding_table_opcodes[s],
                     .PointertoVSBindingTable = cmd_buffer->state.binding_tables[s].offset);
   }
}

GENX_FUNC(GEN7, GEN7) uint32_t
genX(cmd_buffer_flush_descriptor_sets)(struct anv_cmd_buffer *cmd_buffer)
{
   VkShaderStageFlags dirty = cmd_buffer->state.descriptors_dirty &
                              cmd_buffer->state.pipeline->active_stages;

   VkResult result = VK_SUCCESS;
   anv_foreach_stage(s, dirty) {
      result = anv_cmd_buffer_emit_samplers(cmd_buffer, s,
                                            &cmd_buffer->state.samplers[s]);
      if (result != VK_SUCCESS)
         break;
      result = anv_cmd_buffer_emit_binding_table(cmd_buffer, s,
                                                 &cmd_buffer->state.binding_tables[s]);
      if (result != VK_SUCCESS)
         break;
   }

   if (result != VK_SUCCESS) {
      assert(result == VK_ERROR_OUT_OF_DEVICE_MEMORY);

      result = anv_cmd_buffer_new_binding_table_block(cmd_buffer);
      assert(result == VK_SUCCESS);

      /* Re-emit state base addresses so we get the new surface state base
       * address before we start emitting binding tables etc.
       */
      anv_cmd_buffer_emit_state_base_address(cmd_buffer);

      /* Re-emit all active binding tables */
      dirty |= cmd_buffer->state.pipeline->active_stages;
      anv_foreach_stage(s, dirty) {
         result = anv_cmd_buffer_emit_samplers(cmd_buffer, s,
                                               &cmd_buffer->state.samplers[s]);
         if (result != VK_SUCCESS)
            return result;
         result = anv_cmd_buffer_emit_binding_table(cmd_buffer, s,
                                                    &cmd_buffer->state.binding_tables[s]);
         if (result != VK_SUCCESS)
            return result;
      }
   }

   cmd_buffer->state.descriptors_dirty &= ~dirty;

   return dirty;
}

static inline int64_t
clamp_int64(int64_t x, int64_t min, int64_t max)
{
   if (x < min)
      return min;
   else if (x < max)
      return x;
   else
      return max;
}

static void
emit_scissor_state(struct anv_cmd_buffer *cmd_buffer,
                   uint32_t count, const VkRect2D *scissors)
{
   struct anv_state scissor_state =
      anv_cmd_buffer_alloc_dynamic_state(cmd_buffer, count * 8, 32);

   for (uint32_t i = 0; i < count; i++) {
      const VkRect2D *s = &scissors[i];

      /* Since xmax and ymax are inclusive, we have to have xmax < xmin or
       * ymax < ymin for empty clips.  In case clip x, y, width height are all
       * 0, the clamps below produce 0 for xmin, ymin, xmax, ymax, which isn't
       * what we want. Just special case empty clips and produce a canonical
       * empty clip. */
      static const struct GEN7_SCISSOR_RECT empty_scissor = {
         .ScissorRectangleYMin = 1,
         .ScissorRectangleXMin = 1,
         .ScissorRectangleYMax = 0,
         .ScissorRectangleXMax = 0
      };

      const int max = 0xffff;
      struct GEN7_SCISSOR_RECT scissor = {
         /* Do this math using int64_t so overflow gets clamped correctly. */
         .ScissorRectangleYMin = clamp_int64(s->offset.y, 0, max),
         .ScissorRectangleXMin = clamp_int64(s->offset.x, 0, max),
         .ScissorRectangleYMax = clamp_int64((uint64_t) s->offset.y + s->extent.height - 1, 0, max),
         .ScissorRectangleXMax = clamp_int64((uint64_t) s->offset.x + s->extent.width - 1, 0, max)
      };

      if (s->extent.width <= 0 || s->extent.height <= 0) {
         GEN7_SCISSOR_RECT_pack(NULL, scissor_state.map + i * 8,
                                &empty_scissor);
      } else {
         GEN7_SCISSOR_RECT_pack(NULL, scissor_state.map + i * 8, &scissor);
      }
   }

   anv_batch_emit(&cmd_buffer->batch, GEN7_3DSTATE_SCISSOR_STATE_POINTERS,
                  .ScissorRectPointer = scissor_state.offset);

   if (!cmd_buffer->device->info.has_llc)
      anv_state_clflush(scissor_state);
}

GENX_FUNC(GEN7, GEN7) void
genX(cmd_buffer_emit_scissor)(struct anv_cmd_buffer *cmd_buffer)
{
   if (cmd_buffer->state.dynamic.scissor.count > 0) {
      emit_scissor_state(cmd_buffer, cmd_buffer->state.dynamic.scissor.count,
                         cmd_buffer->state.dynamic.scissor.scissors);
   } else {
      /* Emit a default scissor based on the currently bound framebuffer */
      emit_scissor_state(cmd_buffer, 1,
                         &(VkRect2D) {
                            .offset = { .x = 0, .y = 0, },
                            .extent = {
                               .width = cmd_buffer->state.framebuffer->width,
                               .height = cmd_buffer->state.framebuffer->height,
                            },
                         });
   }
}

static const uint32_t vk_to_gen_index_type[] = {
   [VK_INDEX_TYPE_UINT16]                       = INDEX_WORD,
   [VK_INDEX_TYPE_UINT32]                       = INDEX_DWORD,
};

static const uint32_t restart_index_for_type[] = {
   [VK_INDEX_TYPE_UINT16]                    = UINT16_MAX,
   [VK_INDEX_TYPE_UINT32]                    = UINT32_MAX,
};

void genX(CmdBindIndexBuffer)(
    VkCommandBuffer                             commandBuffer,
    VkBuffer                                    _buffer,
    VkDeviceSize                                offset,
    VkIndexType                                 indexType)
{
   ANV_FROM_HANDLE(anv_cmd_buffer, cmd_buffer, commandBuffer);
   ANV_FROM_HANDLE(anv_buffer, buffer, _buffer);

   cmd_buffer->state.dirty |= ANV_CMD_DIRTY_INDEX_BUFFER;
   if (ANV_IS_HASWELL)
      cmd_buffer->state.restart_index = restart_index_for_type[indexType];
   cmd_buffer->state.gen7.index_buffer = buffer;
   cmd_buffer->state.gen7.index_type = vk_to_gen_index_type[indexType];
   cmd_buffer->state.gen7.index_offset = offset;
}

static VkResult
flush_compute_descriptor_set(struct anv_cmd_buffer *cmd_buffer)
{
   struct anv_device *device = cmd_buffer->device;
   struct anv_pipeline *pipeline = cmd_buffer->state.compute_pipeline;
   struct anv_state surfaces = { 0, }, samplers = { 0, };
   VkResult result;

   result = anv_cmd_buffer_emit_samplers(cmd_buffer,
                                         MESA_SHADER_COMPUTE, &samplers);
   if (result != VK_SUCCESS)
      return result;
   result = anv_cmd_buffer_emit_binding_table(cmd_buffer,
                                              MESA_SHADER_COMPUTE, &surfaces);
   if (result != VK_SUCCESS)
      return result;

   struct anv_state push_state = anv_cmd_buffer_cs_push_constants(cmd_buffer);

   const struct brw_cs_prog_data *cs_prog_data = &pipeline->cs_prog_data;
   const struct brw_stage_prog_data *prog_data = &cs_prog_data->base;

   unsigned local_id_dwords = cs_prog_data->local_invocation_id_regs * 8;
   unsigned push_constant_data_size =
      (prog_data->nr_params + local_id_dwords) * 4;
   unsigned reg_aligned_constant_size = ALIGN(push_constant_data_size, 32);
   unsigned push_constant_regs = reg_aligned_constant_size / 32;

   if (push_state.alloc_size) {
      anv_batch_emit(&cmd_buffer->batch, GENX(MEDIA_CURBE_LOAD),
                     .CURBETotalDataLength = push_state.alloc_size,
                     .CURBEDataStartAddress = push_state.offset);
   }

   assert(prog_data->total_shared <= 64 * 1024);
   uint32_t slm_size = 0;
   if (prog_data->total_shared > 0) {
      /* slm_size is in 4k increments, but must be a power of 2. */
      slm_size = 4 * 1024;
      while (slm_size < prog_data->total_shared)
         slm_size <<= 1;
      slm_size /= 4 * 1024;
   }

   struct anv_state state =
      anv_state_pool_emit(&device->dynamic_state_pool,
                          GEN7_INTERFACE_DESCRIPTOR_DATA, 64,
                          .KernelStartPointer = pipeline->cs_simd,
                          .BindingTablePointer = surfaces.offset,
                          .SamplerStatePointer = samplers.offset,
                          .ConstantURBEntryReadLength =
                             push_constant_regs,
                          .ConstantURBEntryReadOffset = 0,
                          .BarrierEnable = cs_prog_data->uses_barrier,
                          .SharedLocalMemorySize = slm_size,
                          .NumberofThreadsinGPGPUThreadGroup =
                             pipeline->cs_thread_width_max);

   const uint32_t size = GEN7_INTERFACE_DESCRIPTOR_DATA_length * sizeof(uint32_t);
   anv_batch_emit(&cmd_buffer->batch, GEN7_MEDIA_INTERFACE_DESCRIPTOR_LOAD,
                  .InterfaceDescriptorTotalLength = size,
                  .InterfaceDescriptorDataStartAddress = state.offset);

   return VK_SUCCESS;
}

void
genX(cmd_buffer_flush_compute_state)(struct anv_cmd_buffer *cmd_buffer)
{
   struct anv_pipeline *pipeline = cmd_buffer->state.compute_pipeline;
   VkResult result;

   assert(pipeline->active_stages == VK_SHADER_STAGE_COMPUTE_BIT);

   if (cmd_buffer->state.current_pipeline != GPGPU) {
      anv_batch_emit(&cmd_buffer->batch, GEN7_PIPELINE_SELECT,
                     .PipelineSelection = GPGPU);
      cmd_buffer->state.current_pipeline = GPGPU;
   }

   if (cmd_buffer->state.compute_dirty & ANV_CMD_DIRTY_PIPELINE)
      anv_batch_emit_batch(&cmd_buffer->batch, &pipeline->batch);

   if ((cmd_buffer->state.descriptors_dirty & VK_SHADER_STAGE_COMPUTE_BIT) ||
       (cmd_buffer->state.compute_dirty & ANV_CMD_DIRTY_PIPELINE)) {
      /* FIXME: figure out descriptors for gen7 */
      result = flush_compute_descriptor_set(cmd_buffer);
      assert(result == VK_SUCCESS);
      cmd_buffer->state.descriptors_dirty &= ~VK_SHADER_STAGE_COMPUTE_BIT;
   }

   cmd_buffer->state.compute_dirty = 0;
}

void
genX(cmd_buffer_flush_state)(struct anv_cmd_buffer *cmd_buffer)
{
   struct anv_pipeline *pipeline = cmd_buffer->state.pipeline;
   uint32_t *p;

   uint32_t vb_emit = cmd_buffer->state.vb_dirty & pipeline->vb_used;

   assert((pipeline->active_stages & VK_SHADER_STAGE_COMPUTE_BIT) == 0);

   genX(flush_pipeline_select_3d)(cmd_buffer);

   if (vb_emit) {
      const uint32_t num_buffers = __builtin_popcount(vb_emit);
      const uint32_t num_dwords = 1 + num_buffers * 4;

      p = anv_batch_emitn(&cmd_buffer->batch, num_dwords,
                          GEN7_3DSTATE_VERTEX_BUFFERS);
      uint32_t vb, i = 0;
      for_each_bit(vb, vb_emit) {
         struct anv_buffer *buffer = cmd_buffer->state.vertex_bindings[vb].buffer;
         uint32_t offset = cmd_buffer->state.vertex_bindings[vb].offset;

         struct GEN7_VERTEX_BUFFER_STATE state = {
            .VertexBufferIndex = vb,
            .BufferAccessType = pipeline->instancing_enable[vb] ? INSTANCEDATA : VERTEXDATA,
            .VertexBufferMemoryObjectControlState = GEN7_MOCS,
            .AddressModifyEnable = true,
            .BufferPitch = pipeline->binding_stride[vb],
            .BufferStartingAddress = { buffer->bo, buffer->offset + offset },
            .EndAddress = { buffer->bo, buffer->offset + buffer->size - 1},
            .InstanceDataStepRate = 1
         };

         GEN7_VERTEX_BUFFER_STATE_pack(&cmd_buffer->batch, &p[1 + i * 4], &state);
         i++;
      }
   }

   if (cmd_buffer->state.dirty & ANV_CMD_DIRTY_PIPELINE) {
      /* If somebody compiled a pipeline after starting a command buffer the
       * scratch bo may have grown since we started this cmd buffer (and
       * emitted STATE_BASE_ADDRESS).  If we're binding that pipeline now,
       * reemit STATE_BASE_ADDRESS so that we use the bigger scratch bo. */
      if (cmd_buffer->state.scratch_size < pipeline->total_scratch)
         gen7_cmd_buffer_emit_state_base_address(cmd_buffer);

      anv_batch_emit_batch(&cmd_buffer->batch, &pipeline->batch);
   }

   if (cmd_buffer->state.descriptors_dirty & VK_SHADER_STAGE_VERTEX_BIT ||
       cmd_buffer->state.push_constants_dirty & VK_SHADER_STAGE_VERTEX_BIT) {
      /* From the IVB PRM Vol. 2, Part 1, Section 3.2.1:
       *
       *    "A PIPE_CONTROL with Post-Sync Operation set to 1h and a depth
       *    stall needs to be sent just prior to any 3DSTATE_VS,
       *    3DSTATE_URB_VS, 3DSTATE_CONSTANT_VS,
       *    3DSTATE_BINDING_TABLE_POINTER_VS,
       *    3DSTATE_SAMPLER_STATE_POINTER_VS command.  Only one
       *    PIPE_CONTROL needs to be sent before any combination of VS
       *    associated 3DSTATE."
       */
      anv_batch_emit(&cmd_buffer->batch, GEN7_PIPE_CONTROL,
                     .DepthStallEnable = true,
                     .PostSyncOperation = WriteImmediateData,
                     .Address = { &cmd_buffer->device->workaround_bo, 0 });
   }

   uint32_t dirty = 0;
   if (cmd_buffer->state.descriptors_dirty) {
      dirty = gen7_cmd_buffer_flush_descriptor_sets(cmd_buffer);
      gen7_cmd_buffer_emit_descriptor_pointers(cmd_buffer, dirty);
   }

   if (cmd_buffer->state.push_constants_dirty)
      cmd_buffer_flush_push_constants(cmd_buffer);

   /* We use the gen8 state here because it only contains the additional
    * min/max fields and, since they occur at the end of the packet and
    * don't change the stride, they work on gen7 too.
    */
   if (cmd_buffer->state.dirty & ANV_CMD_DIRTY_DYNAMIC_VIEWPORT)
      gen8_cmd_buffer_emit_viewport(cmd_buffer);

   if (cmd_buffer->state.dirty & ANV_CMD_DIRTY_DYNAMIC_SCISSOR)
      gen7_cmd_buffer_emit_scissor(cmd_buffer);

   if (cmd_buffer->state.dirty & (ANV_CMD_DIRTY_PIPELINE |
                                  ANV_CMD_DIRTY_RENDER_TARGETS |
                                  ANV_CMD_DIRTY_DYNAMIC_LINE_WIDTH |
                                  ANV_CMD_DIRTY_DYNAMIC_DEPTH_BIAS)) {

      bool enable_bias = cmd_buffer->state.dynamic.depth_bias.bias != 0.0f ||
         cmd_buffer->state.dynamic.depth_bias.slope != 0.0f;

      const struct anv_image_view *iview =
         anv_cmd_buffer_get_depth_stencil_view(cmd_buffer);
      const struct anv_image *image = iview ? iview->image : NULL;
      const uint32_t depth_format = image ?
         isl_surf_get_depth_format(&cmd_buffer->device->isl_dev,
                                   &image->depth_surface.isl) : D16_UNORM;

      uint32_t sf_dw[GEN7_3DSTATE_SF_length];
      struct GEN7_3DSTATE_SF sf = {
         GEN7_3DSTATE_SF_header,
         .DepthBufferSurfaceFormat = depth_format,
         .LineWidth = cmd_buffer->state.dynamic.line_width,
         .GlobalDepthOffsetEnableSolid = enable_bias,
         .GlobalDepthOffsetEnableWireframe = enable_bias,
         .GlobalDepthOffsetEnablePoint = enable_bias,
         .GlobalDepthOffsetConstant = cmd_buffer->state.dynamic.depth_bias.bias,
         .GlobalDepthOffsetScale = cmd_buffer->state.dynamic.depth_bias.slope,
         .GlobalDepthOffsetClamp = cmd_buffer->state.dynamic.depth_bias.clamp
      };
      GEN7_3DSTATE_SF_pack(NULL, sf_dw, &sf);

      anv_batch_emit_merge(&cmd_buffer->batch, sf_dw, pipeline->gen7.sf);
   }

   if (cmd_buffer->state.dirty & (ANV_CMD_DIRTY_DYNAMIC_BLEND_CONSTANTS |
                                  ANV_CMD_DIRTY_DYNAMIC_STENCIL_REFERENCE)) {
      struct anv_state cc_state =
         anv_cmd_buffer_alloc_dynamic_state(cmd_buffer,
                                            GEN7_COLOR_CALC_STATE_length * 4,
                                            64);
      struct GEN7_COLOR_CALC_STATE cc = {
         .BlendConstantColorRed = cmd_buffer->state.dynamic.blend_constants[0],
         .BlendConstantColorGreen = cmd_buffer->state.dynamic.blend_constants[1],
         .BlendConstantColorBlue = cmd_buffer->state.dynamic.blend_constants[2],
         .BlendConstantColorAlpha = cmd_buffer->state.dynamic.blend_constants[3],
         .StencilReferenceValue =
            cmd_buffer->state.dynamic.stencil_reference.front,
         .BackFaceStencilReferenceValue =
            cmd_buffer->state.dynamic.stencil_reference.back,
      };
      GEN7_COLOR_CALC_STATE_pack(NULL, cc_state.map, &cc);
      if (!cmd_buffer->device->info.has_llc)
         anv_state_clflush(cc_state);

      anv_batch_emit(&cmd_buffer->batch,
                     GEN7_3DSTATE_CC_STATE_POINTERS,
                     .ColorCalcStatePointer = cc_state.offset);
   }

   if (cmd_buffer->state.dirty & (ANV_CMD_DIRTY_PIPELINE |
                                  ANV_CMD_DIRTY_RENDER_TARGETS |
                                  ANV_CMD_DIRTY_DYNAMIC_STENCIL_COMPARE_MASK |
                                  ANV_CMD_DIRTY_DYNAMIC_STENCIL_WRITE_MASK)) {
      uint32_t depth_stencil_dw[GEN7_DEPTH_STENCIL_STATE_length];

      const struct anv_image_view *iview =
         anv_cmd_buffer_get_depth_stencil_view(cmd_buffer);

      struct GEN7_DEPTH_STENCIL_STATE depth_stencil = {
         .StencilBufferWriteEnable = iview && (iview->aspect_mask & VK_IMAGE_ASPECT_STENCIL_BIT),

         .StencilTestMask =
            cmd_buffer->state.dynamic.stencil_compare_mask.front & 0xff,
         .StencilWriteMask =
            cmd_buffer->state.dynamic.stencil_write_mask.front & 0xff,

         .BackfaceStencilTestMask =
            cmd_buffer->state.dynamic.stencil_compare_mask.back & 0xff,
         .BackfaceStencilWriteMask =
            cmd_buffer->state.dynamic.stencil_write_mask.back & 0xff,
      };
      GEN7_DEPTH_STENCIL_STATE_pack(NULL, depth_stencil_dw, &depth_stencil);

      struct anv_state ds_state =
         anv_cmd_buffer_merge_dynamic(cmd_buffer, depth_stencil_dw,
                                      pipeline->gen7.depth_stencil_state,
                                      GEN7_DEPTH_STENCIL_STATE_length, 64);

      anv_batch_emit(&cmd_buffer->batch,
                     GEN7_3DSTATE_DEPTH_STENCIL_STATE_POINTERS,
                     .PointertoDEPTH_STENCIL_STATE = ds_state.offset);
   }

   if (cmd_buffer->state.gen7.index_buffer &&
       cmd_buffer->state.dirty & (ANV_CMD_DIRTY_PIPELINE |
                                  ANV_CMD_DIRTY_INDEX_BUFFER)) {
      struct anv_buffer *buffer = cmd_buffer->state.gen7.index_buffer;
      uint32_t offset = cmd_buffer->state.gen7.index_offset;

      if (ANV_IS_HASWELL) {
         anv_batch_emit(&cmd_buffer->batch, GEN75_3DSTATE_VF,
                        .IndexedDrawCutIndexEnable = pipeline->primitive_restart,
                        .CutIndex = cmd_buffer->state.restart_index);
      }

      anv_batch_emit(&cmd_buffer->batch, GEN7_3DSTATE_INDEX_BUFFER,
                     .CutIndexEnable = pipeline->primitive_restart,
                     .IndexFormat = cmd_buffer->state.gen7.index_type,
                     .MemoryObjectControlState = GEN7_MOCS,
                     .BufferStartingAddress = { buffer->bo, buffer->offset + offset },
                     .BufferEndingAddress = { buffer->bo, buffer->offset + buffer->size });
   }

   cmd_buffer->state.vb_dirty &= ~vb_emit;
   cmd_buffer->state.dirty = 0;
}

void genX(CmdSetEvent)(
    VkCommandBuffer                             commandBuffer,
    VkEvent                                     event,
    VkPipelineStageFlags                        stageMask)
{
   stub();
}

void genX(CmdResetEvent)(
    VkCommandBuffer                             commandBuffer,
    VkEvent                                     event,
    VkPipelineStageFlags                        stageMask)
{
   stub();
}

void genX(CmdWaitEvents)(
    VkCommandBuffer                             commandBuffer,
    uint32_t                                    eventCount,
    const VkEvent*                              pEvents,
    VkPipelineStageFlags                        srcStageMask,
    VkPipelineStageFlags                        destStageMask,
    uint32_t                                    memoryBarrierCount,
    const VkMemoryBarrier*                      pMemoryBarriers,
    uint32_t                                    bufferMemoryBarrierCount,
    const VkBufferMemoryBarrier*                pBufferMemoryBarriers,
    uint32_t                                    imageMemoryBarrierCount,
    const VkImageMemoryBarrier*                 pImageMemoryBarriers)
{
   stub();
}