2 * Copyright © 2011 Intel Corporation
4 * Permission is hereby granted, free of charge, to any person obtaining a
5 * copy of this software and associated documentation files (the "Software"),
6 * to deal in the Software without restriction, including without limitation
7 * the rights to use, copy, modify, merge, publish, distribute, sublicense,
8 * and/or sell copies of the Software, and to permit persons to whom the
9 * Software is furnished to do so, subject to the following conditions:
11 * The above copyright notice and this permission notice (including the next
12 * paragraph) shall be included in all copies or substantial portions of the
15 * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
16 * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
17 * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL
18 * THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
19 * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
20 * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS
24 #include "brw_context.h"
25 #include "brw_state.h"
26 #include "brw_defines.h"
28 #include "main/macros.h"
29 #include "intel_batchbuffer.h"
32 upload_sbe_state(struct brw_context
*brw
)
34 struct intel_context
*intel
= &brw
->intel
;
35 struct gl_context
*ctx
= &intel
->ctx
;
36 struct brw_vue_map vue_map
;
37 uint32_t urb_entry_read_length
;
38 /* CACHE_NEW_VS_PROG */
39 GLbitfield64 vs_outputs_written
= brw
->vs
.prog_data
->outputs_written
;
40 /* BRW_NEW_FRAGMENT_PROGRAM */
41 uint32_t num_outputs
= _mesa_bitcount_64(brw
->fragment_program
->Base
.InputsRead
);
43 bool shade_model_flat
= ctx
->Light
.ShadeModel
== GL_FLAT
;
44 uint32_t dw1
, dw10
, dw11
;
46 int attr
= 0, input_index
= 0;
48 int urb_entry_read_offset
= 1;
49 bool userclip_active
= (ctx
->Transform
.ClipPlanesEnabled
!= 0);
50 uint16_t attr_overrides
[FRAG_ATTRIB_MAX
];
52 brw_compute_vue_map(&vue_map
, intel
, userclip_active
, vs_outputs_written
);
53 urb_entry_read_length
= (vue_map
.num_slots
+ 1)/2 - urb_entry_read_offset
;
54 if (urb_entry_read_length
== 0) {
55 /* Setting the URB entry read length to 0 causes undefined behavior, so
56 * if we have no URB data to read, set it to 1.
58 urb_entry_read_length
= 1;
61 /* FINISHME: Attribute Swizzle Control Mode? */
63 GEN7_SBE_SWIZZLE_ENABLE
|
64 num_outputs
<< GEN7_SBE_NUM_OUTPUTS_SHIFT
|
65 urb_entry_read_length
<< GEN7_SBE_URB_ENTRY_READ_LENGTH_SHIFT
|
66 urb_entry_read_offset
<< GEN7_SBE_URB_ENTRY_READ_OFFSET_SHIFT
;
69 if (ctx
->Point
.SpriteOrigin
== GL_LOWER_LEFT
)
70 dw1
|= GEN6_SF_POINT_SPRITE_LOWERLEFT
;
75 /* Create the mapping from the FS inputs we produce to the VS outputs
78 for (; attr
< FRAG_ATTRIB_MAX
; attr
++) {
79 enum glsl_interp_qualifier interp_qualifier
=
80 brw
->fragment_program
->InterpQualifier
[attr
];
81 bool is_gl_Color
= attr
== FRAG_ATTRIB_COL0
|| attr
== FRAG_ATTRIB_COL1
;
83 if (!(brw
->fragment_program
->Base
.InputsRead
& BITFIELD64_BIT(attr
)))
86 if (ctx
->Point
.PointSprite
&&
87 attr
>= FRAG_ATTRIB_TEX0
&& attr
<= FRAG_ATTRIB_TEX7
&&
88 ctx
->Point
.CoordReplace
[attr
- FRAG_ATTRIB_TEX0
]) {
89 dw10
|= (1 << input_index
);
92 if (attr
== FRAG_ATTRIB_PNTC
)
93 dw10
|= (1 << input_index
);
96 if (interp_qualifier
== INTERP_QUALIFIER_FLAT
||
97 (shade_model_flat
&& is_gl_Color
&&
98 interp_qualifier
== INTERP_QUALIFIER_NONE
))
99 dw11
|= (1 << input_index
);
101 /* The hardware can only do the overrides on 16 overrides at a
102 * time, and the other up to 16 have to be lined up so that the
103 * input index = the output index. We'll need to do some
104 * tweaking to make sure that's the case.
106 assert(input_index
< 16 || attr
== input_index
);
108 attr_overrides
[input_index
++] =
109 get_attr_override(&vue_map
, urb_entry_read_offset
, attr
,
110 ctx
->VertexProgram
._TwoSideEnabled
);
113 for (; attr
< FRAG_ATTRIB_MAX
; attr
++)
114 attr_overrides
[input_index
++] = 0;
117 OUT_BATCH(_3DSTATE_SBE
<< 16 | (14 - 2));
120 /* Output dwords 2 through 9 */
121 for (i
= 0; i
< 8; i
++) {
122 OUT_BATCH(attr_overrides
[i
* 2] | attr_overrides
[i
* 2 + 1] << 16);
125 OUT_BATCH(dw10
); /* point sprite texcoord bitmask */
126 OUT_BATCH(dw11
); /* constant interp bitmask */
127 OUT_BATCH(0); /* wrapshortest enables 0-7 */
128 OUT_BATCH(0); /* wrapshortest enables 8-15 */
132 const struct brw_tracked_state gen7_sbe_state
= {
134 .mesa
= (_NEW_LIGHT
|
137 .brw
= (BRW_NEW_CONTEXT
|
138 BRW_NEW_FRAGMENT_PROGRAM
),
139 .cache
= CACHE_NEW_VS_PROG
141 .emit
= upload_sbe_state
,
145 upload_sf_state(struct brw_context
*brw
)
147 struct intel_context
*intel
= &brw
->intel
;
148 struct gl_context
*ctx
= &intel
->ctx
;
149 uint32_t dw1
, dw2
, dw3
;
152 bool render_to_fbo
= brw
->intel
.ctx
.DrawBuffer
->Name
!= 0;
154 dw1
= GEN6_SF_STATISTICS_ENABLE
| GEN6_SF_VIEWPORT_TRANSFORM_ENABLE
;
157 dw1
|= (brw_depthbuffer_format(brw
) << GEN7_SF_DEPTH_BUFFER_SURFACE_FORMAT_SHIFT
);
160 if ((ctx
->Polygon
.FrontFace
== GL_CCW
) ^ render_to_fbo
)
161 dw1
|= GEN6_SF_WINDING_CCW
;
163 if (ctx
->Polygon
.OffsetFill
)
164 dw1
|= GEN6_SF_GLOBAL_DEPTH_OFFSET_SOLID
;
166 if (ctx
->Polygon
.OffsetLine
)
167 dw1
|= GEN6_SF_GLOBAL_DEPTH_OFFSET_WIREFRAME
;
169 if (ctx
->Polygon
.OffsetPoint
)
170 dw1
|= GEN6_SF_GLOBAL_DEPTH_OFFSET_POINT
;
172 switch (ctx
->Polygon
.FrontMode
) {
174 dw1
|= GEN6_SF_FRONT_SOLID
;
178 dw1
|= GEN6_SF_FRONT_WIREFRAME
;
182 dw1
|= GEN6_SF_FRONT_POINT
;
190 switch (ctx
->Polygon
.BackMode
) {
192 dw1
|= GEN6_SF_BACK_SOLID
;
196 dw1
|= GEN6_SF_BACK_WIREFRAME
;
200 dw1
|= GEN6_SF_BACK_POINT
;
210 if (ctx
->Polygon
.CullFlag
) {
211 switch (ctx
->Polygon
.CullFaceMode
) {
213 dw2
|= GEN6_SF_CULL_FRONT
;
216 dw2
|= GEN6_SF_CULL_BACK
;
218 case GL_FRONT_AND_BACK
:
219 dw2
|= GEN6_SF_CULL_BOTH
;
226 dw2
|= GEN6_SF_CULL_NONE
;
230 if (ctx
->Scissor
.Enabled
)
231 dw2
|= GEN6_SF_SCISSOR_ENABLE
;
234 dw2
|= U_FIXED(CLAMP(ctx
->Line
.Width
, 0.0, 7.99), 7) <<
235 GEN6_SF_LINE_WIDTH_SHIFT
;
236 if (ctx
->Line
.SmoothFlag
) {
237 dw2
|= GEN6_SF_LINE_AA_ENABLE
;
238 dw2
|= GEN6_SF_LINE_AA_MODE_TRUE
;
239 dw2
|= GEN6_SF_LINE_END_CAP_WIDTH_1_0
;
242 /* FINISHME: Last Pixel Enable? Vertex Sub Pixel Precision Select?
243 * FINISHME: AA Line Distance Mode?
249 if (!(ctx
->VertexProgram
.PointSizeEnabled
|| ctx
->Point
._Attenuated
))
250 dw3
|= GEN6_SF_USE_STATE_POINT_WIDTH
;
252 /* Clamp to ARB_point_parameters user limits */
253 point_size
= CLAMP(ctx
->Point
.Size
, ctx
->Point
.MinSize
, ctx
->Point
.MaxSize
);
255 /* Clamp to the hardware limits and convert to fixed point */
256 dw3
|= U_FIXED(CLAMP(point_size
, 0.125, 255.875), 3);
259 if (ctx
->Light
.ProvokingVertex
!= GL_FIRST_VERTEX_CONVENTION
) {
261 (2 << GEN6_SF_TRI_PROVOKE_SHIFT
) |
262 (2 << GEN6_SF_TRIFAN_PROVOKE_SHIFT
) |
263 (1 << GEN6_SF_LINE_PROVOKE_SHIFT
);
265 dw3
|= (1 << GEN6_SF_TRIFAN_PROVOKE_SHIFT
);
269 OUT_BATCH(_3DSTATE_SF
<< 16 | (7 - 2));
273 OUT_BATCH_F(ctx
->Polygon
.OffsetUnits
* 2); /* constant. copied from gen4 */
274 OUT_BATCH_F(ctx
->Polygon
.OffsetFactor
); /* scale */
275 OUT_BATCH_F(0.0); /* XXX: global depth offset clamp */
279 const struct brw_tracked_state gen7_sf_state
= {
281 .mesa
= (_NEW_LIGHT
|
288 .brw
= (BRW_NEW_CONTEXT
),
289 .cache
= CACHE_NEW_VS_PROG
291 .emit
= upload_sf_state
,