summaryrefslogtreecommitdiff
path: root/src/mesa/drivers/dri/i965/brw_context.c
blob: 46505537837c644392d6e3af5c1f0cb3879e54a3 (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
/*
 Copyright (C) Intel Corp.  2006.  All Rights Reserved.
 Intel funded Tungsten Graphics (http://www.tungstengraphics.com) to
 develop this 3D driver.
 
 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 COPYRIGHT OWNER(S) AND/OR ITS SUPPLIERS 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.
 
 **********************************************************************/
 /*
  * Authors:
  *   Keith Whitwell <keith@tungstengraphics.com>
  */


#include "main/api_exec.h"
#include "main/imports.h"
#include "main/macros.h"
#include "main/simple_list.h"
#include "main/version.h"
#include "main/vtxfmt.h"

#include "vbo/vbo_context.h"

#include "brw_context.h"
#include "brw_defines.h"
#include "brw_draw.h"
#include "brw_state.h"

#include "intel_fbo.h"
#include "intel_mipmap_tree.h"
#include "intel_regions.h"
#include "intel_tex.h"
#include "intel_tex_obj.h"

#include "tnl/t_pipeline.h"
#include "glsl/ralloc.h"

/***************************************
 * Mesa's Driver Functions
 ***************************************/

static size_t
brw_query_samples_for_format(struct gl_context *ctx, GLenum target,
                             GLenum internalFormat, int samples[16])
{
   struct intel_context *intel = intel_context(ctx);

   (void) target;

   switch (intel->gen) {
   case 7:
      samples[0] = 8;
      samples[1] = 4;
      return 2;

   case 6:
      samples[0] = 4;
      return 1;

   default:
      samples[0] = 1;
      return 1;
   }
}

static void brwInitDriverFunctions(struct intel_screen *screen,
				   struct dd_function_table *functions)
{
   intelInitDriverFunctions( functions );

   brwInitFragProgFuncs( functions );
   brw_init_queryobj_functions(functions);

   functions->QuerySamplesForFormat = brw_query_samples_for_format;
   functions->BeginTransformFeedback = brw_begin_transform_feedback;

   if (screen->gen >= 7)
      functions->EndTransformFeedback = gen7_end_transform_feedback;
   else
      functions->EndTransformFeedback = brw_end_transform_feedback;

   if (screen->gen >= 6)
      functions->GetSamplePosition = gen6_get_sample_position;
}

bool
brwCreateContext(int api,
	         const struct gl_config *mesaVis,
		 __DRIcontext *driContextPriv,
                 unsigned major_version,
                 unsigned minor_version,
                 uint32_t flags,
                 unsigned *error,
	         void *sharedContextPrivate)
{
   __DRIscreen *sPriv = driContextPriv->driScreenPriv;
   struct intel_screen *screen = sPriv->driverPrivate;
   struct dd_function_table functions;
   unsigned i;

   struct brw_context *brw = rzalloc(NULL, struct brw_context);
   if (!brw) {
      printf("%s: failed to alloc context\n", __FUNCTION__);
      *error = __DRI_CTX_ERROR_NO_MEMORY;
      return false;
   }

   /* brwInitVtbl needs to know the chipset generation so that it can set the
    * right pointers.
    */
   brw->intel.gen = screen->gen;

   brwInitVtbl( brw );

   brwInitDriverFunctions(screen, &functions);

   struct intel_context *intel = &brw->intel;
   struct gl_context *ctx = &intel->ctx;

   if (!intelInitContext( intel, api, major_version, minor_version,
                          mesaVis, driContextPriv,
			  sharedContextPrivate, &functions,
			  error)) {
      ralloc_free(brw);
      return false;
   }

   brw_init_surface_formats(brw);

   /* Initialize swrast, tnl driver tables: */
   TNLcontext *tnl = TNL_CONTEXT(ctx);
   if (tnl)
      tnl->Driver.RunPipeline = _tnl_run_pipeline;

   ctx->DriverFlags.NewTransformFeedback = BRW_NEW_TRANSFORM_FEEDBACK;
   ctx->DriverFlags.NewRasterizerDiscard = BRW_NEW_RASTERIZER_DISCARD;

   ctx->Const.MaxDualSourceDrawBuffers = 1;
   ctx->Const.MaxDrawBuffers = BRW_MAX_DRAW_BUFFERS;
   ctx->Const.MaxTextureImageUnits = BRW_MAX_TEX_UNIT;
   ctx->Const.MaxTextureCoordUnits = 8; /* Mesa limit */
   ctx->Const.MaxTextureUnits = MIN2(ctx->Const.MaxTextureCoordUnits,
                                     ctx->Const.MaxTextureImageUnits);
   ctx->Const.MaxVertexTextureImageUnits = BRW_MAX_TEX_UNIT;
   ctx->Const.MaxCombinedTextureImageUnits =
      ctx->Const.MaxVertexTextureImageUnits +
      ctx->Const.MaxTextureImageUnits;

   ctx->Const.MaxTextureLevels = 14; /* 8192 */
   if (ctx->Const.MaxTextureLevels > MAX_TEXTURE_LEVELS)
	   ctx->Const.MaxTextureLevels = MAX_TEXTURE_LEVELS;
   ctx->Const.Max3DTextureLevels = 9;
   ctx->Const.MaxCubeTextureLevels = 12;

   if (intel->gen >= 7)
      ctx->Const.MaxArrayTextureLayers = 2048;
   else
      ctx->Const.MaxArrayTextureLayers = 512;

   ctx->Const.MaxTextureRectSize = (1<<12);
   
   ctx->Const.MaxTextureMaxAnisotropy = 16.0;

   /* Hardware only supports a limited number of transform feedback buffers.
    * So we need to override the Mesa default (which is based only on software
    * limits).
    */
   ctx->Const.MaxTransformFeedbackBuffers = BRW_MAX_SOL_BUFFERS;

   /* On Gen6, in the worst case, we use up one binding table entry per
    * transform feedback component (see comments above the definition of
    * BRW_MAX_SOL_BINDINGS, in brw_context.h), so we need to advertise a value
    * for MAX_TRANSFORM_FEEDBACK_INTERLEAVED_COMPONENTS equal to
    * BRW_MAX_SOL_BINDINGS.
    *
    * In "separate components" mode, we need to divide this value by
    * BRW_MAX_SOL_BUFFERS, so that the total number of binding table entries
    * used up by all buffers will not exceed BRW_MAX_SOL_BINDINGS.
    */
   ctx->Const.MaxTransformFeedbackInterleavedComponents = BRW_MAX_SOL_BINDINGS;
   ctx->Const.MaxTransformFeedbackSeparateComponents =
      BRW_MAX_SOL_BINDINGS / BRW_MAX_SOL_BUFFERS;

   if (intel->gen == 6) {
      ctx->Const.MaxSamples = 4;
      ctx->Const.MaxColorTextureSamples = 4;
      ctx->Const.MaxDepthTextureSamples = 4;
      ctx->Const.MaxIntegerSamples = 4;
   }
   else if (intel->gen >= 7) {
      ctx->Const.MaxSamples = 8;
      ctx->Const.MaxColorTextureSamples = 8;
      ctx->Const.MaxDepthTextureSamples = 8;
      ctx->Const.MaxIntegerSamples = 8;
   }

   /* if conformance mode is set, swrast can handle any size AA point */
   ctx->Const.MaxPointSizeAA = 255.0;

   /* We want the GLSL compiler to emit code that uses condition codes */
   for (i = 0; i <= MESA_SHADER_FRAGMENT; i++) {
      ctx->ShaderCompilerOptions[i].MaxIfDepth = intel->gen < 6 ? 16 : UINT_MAX;
      ctx->ShaderCompilerOptions[i].EmitCondCodes = true;
      ctx->ShaderCompilerOptions[i].EmitNoNoise = true;
      ctx->ShaderCompilerOptions[i].EmitNoMainReturn = true;
      ctx->ShaderCompilerOptions[i].EmitNoIndirectInput = true;
      ctx->ShaderCompilerOptions[i].EmitNoIndirectOutput = true;

      ctx->ShaderCompilerOptions[i].EmitNoIndirectUniform =
	 (i == MESA_SHADER_FRAGMENT);
      ctx->ShaderCompilerOptions[i].EmitNoIndirectTemp =
	 (i == MESA_SHADER_FRAGMENT);
      ctx->ShaderCompilerOptions[i].LowerClipDistance = true;
   }

   ctx->Const.VertexProgram.MaxNativeInstructions = (16 * 1024);
   ctx->Const.VertexProgram.MaxAluInstructions = 0;
   ctx->Const.VertexProgram.MaxTexInstructions = 0;
   ctx->Const.VertexProgram.MaxTexIndirections = 0;
   ctx->Const.VertexProgram.MaxNativeAluInstructions = 0;
   ctx->Const.VertexProgram.MaxNativeTexInstructions = 0;
   ctx->Const.VertexProgram.MaxNativeTexIndirections = 0;
   ctx->Const.VertexProgram.MaxNativeAttribs = 16;
   ctx->Const.VertexProgram.MaxNativeTemps = 256;
   ctx->Const.VertexProgram.MaxNativeAddressRegs = 1;
   ctx->Const.VertexProgram.MaxNativeParameters = 1024;
   ctx->Const.VertexProgram.MaxEnvParams =
      MIN2(ctx->Const.VertexProgram.MaxNativeParameters,
	   ctx->Const.VertexProgram.MaxEnvParams);

   ctx->Const.FragmentProgram.MaxNativeInstructions = (1 * 1024);
   ctx->Const.FragmentProgram.MaxNativeAluInstructions = (1 * 1024);
   ctx->Const.FragmentProgram.MaxNativeTexInstructions = (1 * 1024);
   ctx->Const.FragmentProgram.MaxNativeTexIndirections = (1 * 1024);
   ctx->Const.FragmentProgram.MaxNativeAttribs = 12;
   ctx->Const.FragmentProgram.MaxNativeTemps = 256;
   ctx->Const.FragmentProgram.MaxNativeAddressRegs = 0;
   ctx->Const.FragmentProgram.MaxNativeParameters = 1024;
   ctx->Const.FragmentProgram.MaxEnvParams =
      MIN2(ctx->Const.FragmentProgram.MaxNativeParameters,
	   ctx->Const.FragmentProgram.MaxEnvParams);

   /* Fragment shaders use real, 32-bit twos-complement integers for all
    * integer types.
    */
   ctx->Const.FragmentProgram.LowInt.RangeMin = 31;
   ctx->Const.FragmentProgram.LowInt.RangeMax = 30;
   ctx->Const.FragmentProgram.LowInt.Precision = 0;
   ctx->Const.FragmentProgram.HighInt = ctx->Const.FragmentProgram.MediumInt
      = ctx->Const.FragmentProgram.LowInt;

   /* Gen6 converts quads to polygon in beginning of 3D pipeline,
      but we're not sure how it's actually done for vertex order,
      that affect provoking vertex decision. Always use last vertex
      convention for quad primitive which works as expected for now. */
   if (intel->gen >= 6)
       ctx->Const.QuadsFollowProvokingVertexConvention = false;

   ctx->Const.QueryCounterBits.Timestamp = 36;

   if (intel->is_g4x || intel->gen >= 5) {
      brw->CMD_VF_STATISTICS = GM45_3DSTATE_VF_STATISTICS;
      brw->CMD_PIPELINE_SELECT = CMD_PIPELINE_SELECT_GM45;
      brw->has_surface_tile_offset = true;
      if (intel->gen < 6)
	  brw->has_compr4 = true;
      brw->has_aa_line_parameters = true;
      brw->has_pln = true;
  } else {
      brw->CMD_VF_STATISTICS = GEN4_3DSTATE_VF_STATISTICS;
      brw->CMD_PIPELINE_SELECT = CMD_PIPELINE_SELECT_965;
   }

   /* WM maximum threads is number of EUs times number of threads per EU. */
   assert(intel->gen <= 7);

   if (intel->is_haswell) {
      if (intel->gt == 1) {
	 brw->max_wm_threads = 102;
	 brw->max_vs_threads = 70;
	 brw->urb.size = 128;
	 brw->urb.max_vs_entries = 640;
	 brw->urb.max_gs_entries = 256;
      } else if (intel->gt == 2) {
	 brw->max_wm_threads = 204;
	 brw->max_vs_threads = 280;
	 brw->urb.size = 256;
	 brw->urb.max_vs_entries = 1664;
	 brw->urb.max_gs_entries = 640;
      }
   } else if (intel->gen == 7) {
      if (intel->gt == 1) {
	 brw->max_wm_threads = 48;
	 brw->max_vs_threads = 36;
	 brw->max_gs_threads = 36;
	 brw->urb.size = 128;
	 brw->urb.max_vs_entries = 512;
	 brw->urb.max_gs_entries = 192;
      } else if (intel->gt == 2) {
	 brw->max_wm_threads = 172;
	 brw->max_vs_threads = 128;
	 brw->max_gs_threads = 128;
	 brw->urb.size = 256;
	 brw->urb.max_vs_entries = 704;
	 brw->urb.max_gs_entries = 320;
      } else {
	 assert(!"Unknown gen7 device.");
      }
   } else if (intel->gen == 6) {
      if (intel->gt == 2) {
	 brw->max_wm_threads = 80;
	 brw->max_vs_threads = 60;
	 brw->max_gs_threads = 60;
	 brw->urb.size = 64;            /* volume 5c.5 section 5.1 */
	 brw->urb.max_vs_entries = 256; /* volume 2a (see 3DSTATE_URB) */
	 brw->urb.max_gs_entries = 256;
      } else {
	 brw->max_wm_threads = 40;
	 brw->max_vs_threads = 24;
	 brw->max_gs_threads = 21; /* conservative; 24 if rendering disabled */
	 brw->urb.size = 32;            /* volume 5c.5 section 5.1 */
	 brw->urb.max_vs_entries = 256; /* volume 2a (see 3DSTATE_URB) */
	 brw->urb.max_gs_entries = 256;
      }
      brw->urb.gen6_gs_previously_active = false;
   } else if (intel->gen == 5) {
      brw->urb.size = 1024;
      brw->max_vs_threads = 72;
      brw->max_gs_threads = 32;
      brw->max_wm_threads = 12 * 6;
   } else if (intel->is_g4x) {
      brw->urb.size = 384;
      brw->max_vs_threads = 32;
      brw->max_gs_threads = 2;
      brw->max_wm_threads = 10 * 5;
   } else if (intel->gen < 6) {
      brw->urb.size = 256;
      brw->max_vs_threads = 16;
      brw->max_gs_threads = 2;
      brw->max_wm_threads = 8 * 4;
      brw->has_negative_rhw_bug = true;
   }

   if (intel->gen <= 7) {
      brw->needs_unlit_centroid_workaround = true;
   }

   brw->prim_restart.in_progress = false;
   brw->prim_restart.enable_cut_index = false;
   intel->hw_ctx = drm_intel_gem_context_create(intel->bufmgr);

   brw_init_state( brw );

   brw->curbe.last_buf = calloc(1, 4096);
   brw->curbe.next_buf = calloc(1, 4096);

   brw->state.dirty.mesa = ~0;
   brw->state.dirty.brw = ~0;

   brw->emit_state_always = 0;

   intel->batch.need_workaround_flush = true;

   ctx->VertexProgram._MaintainTnlProgram = true;
   ctx->FragmentProgram._MaintainTexEnvProgram = true;

   brw_draw_init( brw );

   brw->precompile = driQueryOptionb(&intel->optionCache, "shader_precompile");

   ctx->Const.NativeIntegers = true;
   ctx->Const.UniformBooleanTrue = 1;
   ctx->Const.UniformBufferOffsetAlignment = 16;

   ctx->Const.ForceGLSLExtensionsWarn = driQueryOptionb(&intel->optionCache, "force_glsl_extensions_warn");

   ctx->Const.DisableGLSLLineContinuations = driQueryOptionb(&intel->optionCache, "disable_glsl_line_continuations");

   ctx->Const.ContextFlags = 0;
   if ((flags & __DRI_CTX_FLAG_FORWARD_COMPATIBLE) != 0)
      ctx->Const.ContextFlags |= GL_CONTEXT_FLAG_FORWARD_COMPATIBLE_BIT;

   if ((flags & __DRI_CTX_FLAG_DEBUG) != 0) {
      ctx->Const.ContextFlags |= GL_CONTEXT_FLAG_DEBUG_BIT;

      /* Turn on some extra GL_ARB_debug_output generation. */
      intel->perf_debug = true;
   }

   brw_fs_alloc_reg_sets(brw);

   if (INTEL_DEBUG & DEBUG_SHADER_TIME)
      brw_init_shader_time(brw);

   _mesa_compute_version(ctx);

   _mesa_initialize_dispatch_tables(ctx);
   _mesa_initialize_vbo_vtxfmt(ctx);

   return true;
}