1 /****************************************************************************
2 * Copyright (C) 2014-2015 Intel Corporation. All Rights Reserved.
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
25 * @brief Definitions for API state.
27 ******************************************************************************/
30 #include "common/formats.h"
31 #include "common/simdintrin.h"
34 #define SWR_CLEAR_NONE 0
35 #define SWR_CLEAR_COLOR (1 << 0)
36 #define SWR_CLEAR_DEPTH (1 << 1)
37 #define SWR_CLEAR_STENCIL (1 << 2)
45 //////////////////////////////////////////////////////////////////////////
46 /// PRIMITIVE_TOPOLOGY.
47 //////////////////////////////////////////////////////////////////////////
48 enum PRIMITIVE_TOPOLOGY
54 TOP_TRIANGLE_LIST
= 0x4,
55 TOP_TRIANGLE_STRIP
= 0x5,
56 TOP_TRIANGLE_FAN
= 0x6,
59 TOP_LINE_LIST_ADJ
= 0x9,
60 TOP_LISTSTRIP_ADJ
= 0xA,
61 TOP_TRI_LIST_ADJ
= 0xB,
62 TOP_TRI_STRIP_ADJ
= 0xC,
63 TOP_TRI_STRIP_REVERSE
= 0xD,
67 TOP_POINT_LIST_BF
= 0x11,
68 TOP_LINE_STRIP_CONT
= 0x12,
69 TOP_LINE_STRIP_BF
= 0x13,
70 TOP_LINE_STRIP_CONT_BF
= 0x14,
71 TOP_TRIANGLE_FAN_NOSTIPPLE
= 0x16,
72 TOP_TRIANGLE_DISC
= 0x17, /// @todo What is this??
74 TOP_PATCHLIST_BASE
= 0x1F, // Invalid topology, used to calculate num verts for a patchlist.
75 TOP_PATCHLIST_1
= 0x20, // List of 1-vertex patches
76 TOP_PATCHLIST_2
= 0x21,
77 TOP_PATCHLIST_3
= 0x22,
78 TOP_PATCHLIST_4
= 0x23,
79 TOP_PATCHLIST_5
= 0x24,
80 TOP_PATCHLIST_6
= 0x25,
81 TOP_PATCHLIST_7
= 0x26,
82 TOP_PATCHLIST_8
= 0x27,
83 TOP_PATCHLIST_9
= 0x28,
84 TOP_PATCHLIST_10
= 0x29,
85 TOP_PATCHLIST_11
= 0x2A,
86 TOP_PATCHLIST_12
= 0x2B,
87 TOP_PATCHLIST_13
= 0x2C,
88 TOP_PATCHLIST_14
= 0x2D,
89 TOP_PATCHLIST_15
= 0x2E,
90 TOP_PATCHLIST_16
= 0x2F,
91 TOP_PATCHLIST_17
= 0x30,
92 TOP_PATCHLIST_18
= 0x31,
93 TOP_PATCHLIST_19
= 0x32,
94 TOP_PATCHLIST_20
= 0x33,
95 TOP_PATCHLIST_21
= 0x34,
96 TOP_PATCHLIST_22
= 0x35,
97 TOP_PATCHLIST_23
= 0x36,
98 TOP_PATCHLIST_24
= 0x37,
99 TOP_PATCHLIST_25
= 0x38,
100 TOP_PATCHLIST_26
= 0x39,
101 TOP_PATCHLIST_27
= 0x3A,
102 TOP_PATCHLIST_28
= 0x3B,
103 TOP_PATCHLIST_29
= 0x3C,
104 TOP_PATCHLIST_30
= 0x3D,
105 TOP_PATCHLIST_31
= 0x3E,
106 TOP_PATCHLIST_32
= 0x3F, // List of 32-vertex patches
109 //////////////////////////////////////////////////////////////////////////
111 //////////////////////////////////////////////////////////////////////////
124 //////////////////////////////////////////////////////////////////////////
125 /// SWR_RENDERTARGET_ATTACHMENT
126 /// @todo Its not clear what an "attachment" means. Its not common term.
127 //////////////////////////////////////////////////////////////////////////
128 enum SWR_RENDERTARGET_ATTACHMENT
130 SWR_ATTACHMENT_COLOR0
,
131 SWR_ATTACHMENT_COLOR1
,
132 SWR_ATTACHMENT_COLOR2
,
133 SWR_ATTACHMENT_COLOR3
,
134 SWR_ATTACHMENT_COLOR4
,
135 SWR_ATTACHMENT_COLOR5
,
136 SWR_ATTACHMENT_COLOR6
,
137 SWR_ATTACHMENT_COLOR7
,
138 SWR_ATTACHMENT_DEPTH
,
139 SWR_ATTACHMENT_STENCIL
,
144 #define SWR_NUM_RENDERTARGETS 8
146 #define SWR_ATTACHMENT_COLOR0_BIT 0x001
147 #define SWR_ATTACHMENT_COLOR1_BIT 0x002
148 #define SWR_ATTACHMENT_COLOR2_BIT 0x004
149 #define SWR_ATTACHMENT_COLOR3_BIT 0x008
150 #define SWR_ATTACHMENT_COLOR4_BIT 0x010
151 #define SWR_ATTACHMENT_COLOR5_BIT 0x020
152 #define SWR_ATTACHMENT_COLOR6_BIT 0x040
153 #define SWR_ATTACHMENT_COLOR7_BIT 0x080
154 #define SWR_ATTACHMENT_DEPTH_BIT 0x100
155 #define SWR_ATTACHMENT_STENCIL_BIT 0x200
156 #define SWR_ATTACHMENT_MASK_ALL 0x3ff
157 #define SWR_ATTACHMENT_MASK_COLOR 0x0ff
160 //////////////////////////////////////////////////////////////////////////
161 /// @brief SWR Inner Tessellation factor ID
162 /// See above GetTessFactorOutputPosition code for documentation
163 enum SWR_INNER_TESSFACTOR_ID
165 SWR_QUAD_U_TRI_INSIDE
,
168 SWR_NUM_INNER_TESS_FACTORS
,
171 //////////////////////////////////////////////////////////////////////////
172 /// @brief SWR Outer Tessellation factor ID
173 /// See above GetTessFactorOutputPosition code for documentation
174 enum SWR_OUTER_TESSFACTOR_ID
176 SWR_QUAD_U_EQ0_TRI_U_LINE_DETAIL
,
177 SWR_QUAD_V_EQ0_TRI_V_LINE_DENSITY
,
178 SWR_QUAD_U_EQ1_TRI_W
,
181 SWR_NUM_OUTER_TESS_FACTORS
,
185 /////////////////////////////////////////////////////////////////////////
187 /// @brief Defines a vertex element that holds all the data for SIMD vertices.
188 /// Contains position in clip space, hardcoded to attribute 0,
189 /// space for up to 32 attributes, as well as any SGV values generated
191 /////////////////////////////////////////////////////////////////////////
192 #define VERTEX_POSITION_SLOT 0
193 #define VERTEX_ATTRIB_START_SLOT 1
194 #define VERTEX_ATTRIB_END_SLOT 32
195 #define VERTEX_RTAI_SLOT 33 // GS writes RenderTargetArrayIndex here
196 #define VERTEX_PRIMID_SLOT 34 // GS writes PrimId here
197 #define VERTEX_CLIPCULL_DIST_LO_SLOT 35 // VS writes lower 4 clip/cull dist
198 #define VERTEX_CLIPCULL_DIST_HI_SLOT 36 // VS writes upper 4 clip/cull dist
199 #define VERTEX_POINT_SIZE_SLOT 37 // VS writes point size here
203 simdvector attrib
[KNOB_NUM_ATTRIBUTES
];
206 //////////////////////////////////////////////////////////////////////////
208 /// @brief Input to vertex shader
209 /////////////////////////////////////////////////////////////////////////
210 struct SWR_VS_CONTEXT
212 simdvertex
* pVin
; // IN: SIMD input vertex data store
213 simdvertex
* pVout
; // OUT: SIMD output vertex data store
215 uint32_t InstanceID
; // IN: Instance ID, constant across all verts of the SIMD
216 simdscalari VertexID
; // IN: Vertex ID
217 simdscalari mask
; // IN: Active mask for shader
220 /////////////////////////////////////////////////////////////////////////
222 /// @brief defines a control point element as passed from the output
223 /// of the hull shader to the input of the domain shader
224 /////////////////////////////////////////////////////////////////////////
235 ScalarAttrib attrib
[KNOB_NUM_ATTRIBUTES
];
238 //////////////////////////////////////////////////////////////////////////
239 /// SWR_TESSELLATION_FACTORS
240 /// @brief Tessellation factors structure (non-vector)
241 /////////////////////////////////////////////////////////////////////////
242 struct SWR_TESSELLATION_FACTORS
244 float OuterTessFactors
[SWR_NUM_OUTER_TESS_FACTORS
];
245 float InnerTessFactors
[SWR_NUM_INNER_TESS_FACTORS
];
248 #define MAX_NUM_VERTS_PER_PRIM 32 // support up to 32 control point patches
251 SWR_TESSELLATION_FACTORS tessFactors
;
252 ScalarCPoint cp
[MAX_NUM_VERTS_PER_PRIM
];
253 ScalarCPoint patchData
;
256 //////////////////////////////////////////////////////////////////////////
258 /// @brief Input to hull shader
259 /////////////////////////////////////////////////////////////////////////
260 struct SWR_HS_CONTEXT
262 simdvertex vert
[MAX_NUM_VERTS_PER_PRIM
]; // IN: (SIMD) input primitive data
263 simdscalari PrimitiveID
; // IN: (SIMD) primitive ID generated from the draw call
264 simdscalari mask
; // IN: Active mask for shader
265 ScalarPatch
* pCPout
; // OUT: Output control point patch
266 // SIMD-sized-array of SCALAR patches
269 //////////////////////////////////////////////////////////////////////////
271 /// @brief Input to domain shader
272 /////////////////////////////////////////////////////////////////////////
273 struct SWR_DS_CONTEXT
275 uint32_t PrimitiveID
; // IN: (SCALAR) PrimitiveID for the patch associated with the DS invocation
276 uint32_t vectorOffset
; // IN: (SCALAR) vector index offset into SIMD data.
277 uint32_t vectorStride
; // IN: (SCALAR) stride (in vectors) of output data per attribute-component
278 ScalarPatch
* pCpIn
; // IN: (SCALAR) Control patch
279 simdscalar
* pDomainU
; // IN: (SIMD) Domain Point U coords
280 simdscalar
* pDomainV
; // IN: (SIMD) Domain Point V coords
281 simdscalari mask
; // IN: Active mask for shader
282 simdscalar
* pOutputData
; // OUT: (SIMD) Vertex Attributes (2D array of vectors, one row per attribute-component)
285 //////////////////////////////////////////////////////////////////////////
287 /// @brief Input to geometry shader.
288 /////////////////////////////////////////////////////////////////////////
289 struct SWR_GS_CONTEXT
291 simdvertex vert
[MAX_NUM_VERTS_PER_PRIM
]; // IN: input primitive data for SIMD prims
292 simdscalari PrimitiveID
; // IN: input primitive ID generated from the draw call
293 uint32_t InstanceID
; // IN: input instance ID
294 simdscalari mask
; // IN: Active mask for shader
295 uint8_t* pStream
; // OUT: output stream (contains vertices for all output streams)
296 uint8_t* pCutOrStreamIdBuffer
; // OUT: cut or stream id buffer
297 simdscalari vertexCount
; // OUT: num vertices emitted per SIMD lane
300 struct PixelPositions
308 #define SWR_MAX_NUM_MULTISAMPLES 16
310 //////////////////////////////////////////////////////////////////////////
312 /// @brief Input to pixel shader.
313 /////////////////////////////////////////////////////////////////////////
314 struct SWR_PS_CONTEXT
316 PixelPositions vX
; // IN: x location(s) of pixels
317 PixelPositions vY
; // IN: x location(s) of pixels
318 simdscalar vZ
; // INOUT: z location of pixels
319 simdscalari activeMask
; // OUT: mask for kill
320 simdscalar inputMask
; // IN: input coverage mask for all samples
321 simdscalari oMask
; // OUT: mask for output coverage
323 PixelPositions vI
; // barycentric coords evaluated at pixel center, sample position, centroid
325 PixelPositions vOneOverW
; // IN: 1/w
327 const float* pAttribs
; // IN: pointer to attribute barycentric coefficients
328 const float* pPerspAttribs
; // IN: pointer to attribute/w barycentric coefficients
329 const float* pRecipW
; // IN: pointer to 1/w coord for each vertex
330 const float *I
; // IN: Barycentric A, B, and C coefs used to compute I
331 const float *J
; // IN: Barycentric A, B, and C coefs used to compute J
332 float recipDet
; // IN: 1/Det, used when barycentric interpolating attributes
333 const float* pSamplePosX
; // IN: array of sample positions
334 const float* pSamplePosY
; // IN: array of sample positions
335 simdvector shaded
[SWR_NUM_RENDERTARGETS
];
336 // OUT: result color per rendertarget
338 uint32_t frontFace
; // IN: front- 1, back- 0
339 uint32_t primID
; // IN: primitive ID
340 uint32_t sampleIndex
; // IN: sampleIndex
344 //////////////////////////////////////////////////////////////////////////
346 /// @brief Input to compute shader.
347 /////////////////////////////////////////////////////////////////////////
348 struct SWR_CS_CONTEXT
350 // The ThreadGroupId is the current thread group index relative
351 // to all thread groups in the Dispatch call. The ThreadId, ThreadIdInGroup,
352 // and ThreadIdInGroupFlattened can be derived from ThreadGroupId in the shader.
354 // Compute shader accepts the following system values.
355 // o ThreadId - Current thread id relative to all other threads in dispatch.
356 // o ThreadGroupId - Current thread group id relative to all other groups in dispatch.
357 // o ThreadIdInGroup - Current thread relative to all threads in the current thread group.
358 // o ThreadIdInGroupFlattened - Flattened linear id derived from ThreadIdInGroup.
360 // All of these system values can be computed in the shader. They will be
361 // derived from the current tile counter. The tile counter is an atomic counter that
362 // resides in the draw context and is initialized to the product of the dispatch dims.
364 // tileCounter = dispatchDims.x * dispatchDims.y * dispatchDims.z
366 // Each CPU worker thread will atomically decrement this counter and passes the current
367 // count into the shader. When the count reaches 0 then all thread groups in the
368 // dispatch call have been completed.
370 uint32_t tileCounter
; // The tile counter value for this thread group.
372 // Dispatch dimensions used by shader to compute system values from the tile counter.
373 uint32_t dispatchDims
[3];
375 uint8_t* pTGSM
; // Thread Group Shared Memory pointer.
377 uint8_t* pSpillFillBuffer
; // Spill/fill buffer for barrier support
383 SWR_TILE_NONE
= 0x0, // Linear mode (no tiling)
384 SWR_TILE_MODE_WMAJOR
, // W major tiling
385 SWR_TILE_MODE_XMAJOR
, // X major tiling
386 SWR_TILE_MODE_YMAJOR
, // Y major tiling
387 SWR_TILE_SWRZ
, // SWR-Z tiling
392 enum SWR_SURFACE_TYPE
399 SURFACE_STRUCTURED_BUFFER
= 5,
428 enum SWR_BLEND_FACTOR
431 BLENDFACTOR_SRC_COLOR
,
432 BLENDFACTOR_SRC_ALPHA
,
433 BLENDFACTOR_DST_ALPHA
,
434 BLENDFACTOR_DST_COLOR
,
435 BLENDFACTOR_SRC_ALPHA_SATURATE
,
436 BLENDFACTOR_CONST_COLOR
,
437 BLENDFACTOR_CONST_ALPHA
,
438 BLENDFACTOR_SRC1_COLOR
,
439 BLENDFACTOR_SRC1_ALPHA
,
441 BLENDFACTOR_INV_SRC_COLOR
,
442 BLENDFACTOR_INV_SRC_ALPHA
,
443 BLENDFACTOR_INV_DST_ALPHA
,
444 BLENDFACTOR_INV_DST_COLOR
,
445 BLENDFACTOR_INV_CONST_COLOR
,
446 BLENDFACTOR_INV_CONST_ALPHA
,
447 BLENDFACTOR_INV_SRC1_COLOR
,
448 BLENDFACTOR_INV_SRC1_ALPHA
464 LOGICOP_AND_INVERTED
,
465 LOGICOP_COPY_INVERTED
,
480 struct SWR_SURFACE_STATE
482 uint8_t *pBaseAddress
;
483 SWR_SURFACE_TYPE type
; // @llvm_enum
484 SWR_FORMAT format
; // @llvm_enum
489 uint32_t samplePattern
;
492 uint32_t minLod
; // for sampled surfaces, the most detailed LOD that can be accessed by sampler
493 uint32_t maxLod
; // for sampled surfaces, the max LOD that can be accessed
494 float resourceMinLod
; // for sampled surfaces, the most detailed fractional mip that can be accessed by sampler
495 uint32_t lod
; // for render targets, the lod being rendered to
496 uint32_t arrayIndex
; // for render targets, the array index being rendered to for arrayed surfaces
497 SWR_TILE_MODE tileMode
; // @llvm_enum
503 uint32_t lodOffsets
[2][15]; // lod offsets for sampled surfaces
505 uint8_t *pAuxBaseAddress
; // Used for compression, append/consume counter, etc.
507 bool bInterleavedSamples
; // are MSAA samples stored interleaved or planar
510 // vertex fetch state
511 // WARNING- any changes to this struct need to be reflected
512 // in the fetch shader jit
513 struct SWR_VERTEX_BUFFER_STATE
517 const uint8_t *pData
;
520 uint32_t maxVertex
; // size / pitch. precalculated value used by fetch shader for OOB checks
521 uint32_t partialInboundsSize
; // size % pitch. precalculated value used by fetch shader for partially OOB vertices
524 struct SWR_INDEX_BUFFER_STATE
526 // Format type for indices (e.g. UINT16, UINT32, etc.)
527 SWR_FORMAT format
; // @llvm_enum
528 const void *pIndices
;
533 //////////////////////////////////////////////////////////////////////////
534 /// SWR_FETCH_CONTEXT
535 /// @brief Input to fetch shader.
536 /// @note WARNING - Changes to this struct need to be reflected in the
537 /// fetch shader jit.
538 /////////////////////////////////////////////////////////////////////////
539 struct SWR_FETCH_CONTEXT
541 const SWR_VERTEX_BUFFER_STATE
* pStreams
; // IN: array of bound vertex buffers
542 const int32_t* pIndices
; // IN: pointer to index buffer for indexed draws
543 const int32_t* pLastIndex
; // IN: pointer to end of index buffer, used for bounds checking
544 uint32_t CurInstance
; // IN: current instance
545 uint32_t BaseVertex
; // IN: base vertex
546 uint32_t StartVertex
; // IN: start vertex
547 uint32_t StartInstance
; // IN: start instance
548 simdscalari VertexID
; // OUT: vector of vertex IDs
549 simdscalari CutMask
; // OUT: vector mask of indices which have the cut index value
552 //////////////////////////////////////////////////////////////////////////
555 /// @brief All statistics generated by SWR go here. These are public
557 /////////////////////////////////////////////////////////////////////////
561 uint64_t DepthPassCount
; // Number of passing depth tests. Not exact.
564 uint64_t IaVertices
; // Number of Fetch Shader vertices
565 uint64_t IaPrimitives
; // Number of PA primitives.
566 uint64_t VsInvocations
; // Number of Vertex Shader invocations
567 uint64_t HsInvocations
; // Number of Hull Shader invocations
568 uint64_t DsInvocations
; // Number of Domain Shader invocations
569 uint64_t GsInvocations
; // Number of Geometry Shader invocations
570 uint64_t PsInvocations
; // Number of Pixel Shader invocations
571 uint64_t CsInvocations
; // Number of Compute Shader invocations
572 uint64_t CInvocations
; // Number of clipper invocations
573 uint64_t CPrimitives
; // Number of clipper primitives.
574 uint64_t GsPrimitives
; // Number of prims GS outputs.
577 uint32_t SoWriteOffset
[4];
578 uint64_t SoPrimStorageNeeded
[4];
579 uint64_t SoNumPrimsWritten
[4];
582 //////////////////////////////////////////////////////////////////////////
583 /// STREAMOUT_BUFFERS
584 /////////////////////////////////////////////////////////////////////////
586 #define MAX_SO_STREAMS 4
587 #define MAX_ATTRIBUTES 32
589 struct SWR_STREAMOUT_BUFFER
593 // Pointers to streamout buffers.
596 // Size of buffer in dwords.
599 // Vertex pitch of buffer in dwords.
602 // Offset into buffer in dwords. SOS will increment this offset.
603 uint32_t streamOffset
;
605 // Offset to the SO write offset. If not null then we update offset here.
606 uint32_t* pWriteOffset
;
610 //////////////////////////////////////////////////////////////////////////
612 /////////////////////////////////////////////////////////////////////////
613 struct SWR_STREAMOUT_STATE
615 // This disables stream output.
618 // which streams are enabled for streamout
619 bool streamEnable
[MAX_SO_STREAMS
];
621 // If set then do not send any streams to the rasterizer.
622 bool rasterizerDisable
;
624 // Specifies which stream to send to the rasterizer.
625 uint32_t streamToRasterizer
;
627 // The stream masks specify which attributes are sent to which streams.
628 // These masks help the FE to setup the pPrimData buffer that is passed
629 // the the Stream Output Shader (SOS) function.
630 uint32_t streamMasks
[MAX_SO_STREAMS
];
632 // Number of attributes, including position, per vertex that are streamed out.
633 // This should match number of bits in stream mask.
634 uint32_t streamNumEntries
[MAX_SO_STREAMS
];
637 //////////////////////////////////////////////////////////////////////////
638 /// STREAMOUT_CONTEXT - Passed to SOS
639 /////////////////////////////////////////////////////////////////////////
640 struct SWR_STREAMOUT_CONTEXT
643 SWR_STREAMOUT_BUFFER
* pBuffer
[MAX_SO_STREAMS
];
645 // Num prims written for this stream
646 uint32_t numPrimsWritten
;
648 // Num prims that should have been written if there were no overflow.
649 uint32_t numPrimStorageNeeded
;
652 //////////////////////////////////////////////////////////////////////////
653 /// SWR_GS_STATE - Geometry shader state
654 /////////////////////////////////////////////////////////////////////////
659 // number of input attributes per vertex. used by the frontend to
660 // optimize assembling primitives for GS
661 uint32_t numInputAttribs
;
663 // output topology - can be point, tristrip, or linestrip
664 PRIMITIVE_TOPOLOGY outputTopology
; // @llvm_enum
666 // maximum number of verts that can be emitted by a single instance of the GS
667 uint32_t maxNumVerts
;
670 uint32_t instanceCount
;
672 // geometry shader emits renderTargetArrayIndex
673 bool emitsRenderTargetArrayIndex
;
675 // geometry shader emits PrimitiveID
676 bool emitsPrimitiveID
;
678 // if true, geometry shader emits a single stream, with separate cut buffer.
679 // if false, geometry shader emits vertices for multiple streams to the stream buffer, with a separate StreamID buffer
680 // to map vertices to streams
683 // when single stream is enabled, singleStreamID dictates which stream is being output.
684 // field ignored if isSingleStream is false
685 uint32_t singleStreamID
;
689 //////////////////////////////////////////////////////////////////////////
690 /// SWR_TS_OUTPUT_TOPOLOGY - Defines data output by the tessellator / DS
691 /////////////////////////////////////////////////////////////////////////
692 enum SWR_TS_OUTPUT_TOPOLOGY
696 SWR_TS_OUTPUT_TRI_CW
,
697 SWR_TS_OUTPUT_TRI_CCW
,
699 SWR_TS_OUTPUT_TOPOLOGY_COUNT
702 //////////////////////////////////////////////////////////////////////////
703 /// SWR_TS_PARTITIONING - Defines tessellation algorithm
704 /////////////////////////////////////////////////////////////////////////
705 enum SWR_TS_PARTITIONING
708 SWR_TS_ODD_FRACTIONAL
,
709 SWR_TS_EVEN_FRACTIONAL
,
711 SWR_TS_PARTITIONING_COUNT
714 //////////////////////////////////////////////////////////////////////////
715 /// SWR_TS_DOMAIN - Defines Tessellation Domain
716 /////////////////////////////////////////////////////////////////////////
726 //////////////////////////////////////////////////////////////////////////
727 /// SWR_TS_STATE - Tessellation state
728 /////////////////////////////////////////////////////////////////////////
732 SWR_TS_OUTPUT_TOPOLOGY tsOutputTopology
; // @llvm_enum
733 SWR_TS_PARTITIONING partitioning
; // @llvm_enum
734 SWR_TS_DOMAIN domain
; // @llvm_enum
736 PRIMITIVE_TOPOLOGY postDSTopology
; // @llvm_enum
738 uint32_t numHsInputAttribs
;
739 uint32_t numHsOutputAttribs
;
740 uint32_t numDsOutputAttribs
;
743 // output merger state
744 struct SWR_RENDER_TARGET_BLEND_STATE
746 uint8_t writeDisableRed
: 1;
747 uint8_t writeDisableGreen
: 1;
748 uint8_t writeDisableBlue
: 1;
749 uint8_t writeDisableAlpha
: 1;
751 static_assert(sizeof(SWR_RENDER_TARGET_BLEND_STATE
) == 1, "Invalid SWR_RENDER_TARGET_BLEND_STATE size");
753 enum SWR_MULTISAMPLE_COUNT
755 SWR_MULTISAMPLE_1X
= 0,
760 SWR_MULTISAMPLE_TYPE_MAX
763 struct SWR_BLEND_STATE
765 // constant blend factor color in RGBA float
766 float constantColor
[4];
768 // alpha test reference value in unorm8 or float32
769 uint32_t alphaTestReference
;
771 // all RT's have the same sample count
772 ///@todo move this to Output Merger state when we refactor
773 SWR_MULTISAMPLE_COUNT sampleCount
; // @llvm_enum
775 SWR_RENDER_TARGET_BLEND_STATE renderTarget
[SWR_NUM_RENDERTARGETS
];
777 static_assert(sizeof(SWR_BLEND_STATE
) == 36, "Invalid SWR_BLEND_STATE size");
779 //////////////////////////////////////////////////////////////////////////
780 /// FUNCTION POINTERS FOR SHADERS
782 typedef void(__cdecl
*PFN_FETCH_FUNC
)(SWR_FETCH_CONTEXT
& fetchInfo
, simdvertex
& out
);
783 typedef void(__cdecl
*PFN_VERTEX_FUNC
)(HANDLE hPrivateData
, SWR_VS_CONTEXT
* pVsContext
);
784 typedef void(__cdecl
*PFN_HS_FUNC
)(HANDLE hPrivateData
, SWR_HS_CONTEXT
* pHsContext
);
785 typedef void(__cdecl
*PFN_DS_FUNC
)(HANDLE hPrivateData
, SWR_DS_CONTEXT
* pDsContext
);
786 typedef void(__cdecl
*PFN_GS_FUNC
)(HANDLE hPrivateData
, SWR_GS_CONTEXT
* pGsContext
);
787 typedef void(__cdecl
*PFN_CS_FUNC
)(HANDLE hPrivateData
, SWR_CS_CONTEXT
* pCsContext
);
788 typedef void(__cdecl
*PFN_SO_FUNC
)(SWR_STREAMOUT_CONTEXT
& soContext
);
789 typedef void(__cdecl
*PFN_PIXEL_KERNEL
)(HANDLE hPrivateData
, SWR_PS_CONTEXT
*pContext
);
790 typedef void(__cdecl
*PFN_CPIXEL_KERNEL
)(HANDLE hPrivateData
, SWR_PS_CONTEXT
*pContext
);
791 typedef void(__cdecl
*PFN_BLEND_JIT_FUNC
)(const SWR_BLEND_STATE
*, simdvector
&, simdvector
&, uint32_t, uint8_t*, simdvector
&, simdscalari
*, simdscalari
*);
792 typedef simdscalar(*PFN_QUANTIZE_DEPTH
)(simdscalar
);
794 //////////////////////////////////////////////////////////////////////////
796 /////////////////////////////////////////////////////////////////////////
797 struct SWR_FRONTEND_STATE
799 // skip clip test, perspective divide, and viewport transform
800 // intended for verts in screen space
801 bool vpTransformDisable
;
802 bool bEnableCutIndex
;
808 uint32_t lineStripList
: 1;
809 uint32_t triStripList
: 2;
813 uint32_t topologyProvokingVertex
; // provoking vertex for the draw topology
816 //////////////////////////////////////////////////////////////////////////
818 /////////////////////////////////////////////////////////////////////////
819 struct SWR_VIEWPORT_MATRIX
829 //////////////////////////////////////////////////////////////////////////
831 /////////////////////////////////////////////////////////////////////////
842 //////////////////////////////////////////////////////////////////////////
844 //////////////////////////////////////////////////////////////////////////
856 SWR_FILLMODE_WIREFRAME
,
860 enum SWR_FRONTWINDING
867 enum SWR_MSAA_SAMPLE_PATTERN
869 SWR_MSAA_CENTER_PATTERN
,
870 SWR_MSAA_STANDARD_PATTERN
,
871 SWR_MSAA_SAMPLE_PATTERN_MAX
874 enum SWR_PIXEL_LOCATION
876 SWR_PIXEL_LOCATION_CENTER
,
877 SWR_PIXEL_LOCATION_UL
,
880 // fixed point screen space sample locations within a pixel
881 struct SWR_MULTISAMPLE_POS
887 enum SWR_MSAA_RASTMODE
889 SWR_MSAA_RASTMODE_OFF_PIXEL
,
890 SWR_MSAA_RASTMODE_OFF_PATTERN
,
891 SWR_MSAA_RASTMODE_ON_PIXEL
,
892 SWR_MSAA_RASTMODE_ON_PATTERN
895 //////////////////////////////////////////////////////////////////////////
897 //////////////////////////////////////////////////////////////////////////
900 uint32_t cullMode
: 2;
901 uint32_t fillMode
: 2;
902 uint32_t frontWinding
: 1;
903 uint32_t scissorEnable
: 1;
904 uint32_t depthClipEnable
: 1;
905 uint32_t pointParam
: 1;
906 uint32_t pointSpriteEnable
: 1;
907 uint32_t pointSpriteTopOrigin
: 1;
908 uint32_t msaaRastEnable
: 1;
909 uint32_t forcedSampleCount
: 1;
910 uint32_t pixelOffset
: 1;
911 uint32_t depthBiasPreAdjusted
: 1; ///< depth bias constant is in float units, not per-format Z units
912 uint32_t conservativeRast
: 1;
918 float slopeScaledDepthBias
;
919 float depthBiasClamp
;
920 SWR_FORMAT depthFormat
; // @llvm_enum
923 // multisample state for MSAA lines
924 SWR_MSAA_RASTMODE rastMode
; // @llvm_enum
926 // sample count the rasterizer is running at
927 SWR_MULTISAMPLE_COUNT sampleCount
; // @llvm_enum
928 uint32_t pixelLocation
; // UL or Center
929 SWR_MULTISAMPLE_POS iSamplePos
[SWR_MAX_NUM_MULTISAMPLES
];
930 SWR_MSAA_SAMPLE_PATTERN samplePattern
; // @llvm_enum
932 // user clip/cull distance enables
933 uint8_t cullDistanceMask
;
934 uint8_t clipDistanceMask
;
938 struct SWR_BACKEND_STATE
940 uint32_t constantInterpolationMask
;
941 uint32_t pointSpriteTexCoordMask
;
942 uint8_t numAttributes
;
943 uint8_t numComponents
[KNOB_NUM_ATTRIBUTES
];
947 union SWR_DEPTH_STENCIL_STATE
952 uint32_t depthWriteEnable
: 1;
953 uint32_t depthTestEnable
: 1;
954 uint32_t stencilWriteEnable
: 1;
955 uint32_t stencilTestEnable
: 1;
956 uint32_t doubleSidedStencilTestEnable
: 1;
958 uint32_t depthTestFunc
: 3;
959 uint32_t stencilTestFunc
: 3;
961 uint32_t backfaceStencilPassDepthPassOp
: 3;
962 uint32_t backfaceStencilPassDepthFailOp
: 3;
963 uint32_t backfaceStencilFailOp
: 3;
964 uint32_t backfaceStencilTestFunc
: 3;
965 uint32_t stencilPassDepthPassOp
: 3;
966 uint32_t stencilPassDepthFailOp
: 3;
967 uint32_t stencilFailOp
: 3;
970 uint8_t backfaceStencilWriteMask
;
971 uint8_t backfaceStencilTestMask
;
972 uint8_t stencilWriteMask
;
973 uint8_t stencilTestMask
;
976 uint8_t backfaceStencilRefValue
;
977 uint8_t stencilRefValue
;
982 enum SWR_SHADING_RATE
984 SWR_SHADING_RATE_PIXEL
,
985 SWR_SHADING_RATE_SAMPLE
,
986 SWR_SHADING_RATE_MAX
,
989 enum SWR_INPUT_COVERAGE
991 SWR_INPUT_COVERAGE_NONE
,
992 SWR_INPUT_COVERAGE_NORMAL
,
993 SWR_INPUT_COVERAGE_INNER_CONSERVATIVE
,
994 SWR_INPUT_COVERAGE_MAX
,
997 enum SWR_PS_POSITION_OFFSET
999 SWR_PS_POSITION_SAMPLE_NONE
,
1000 SWR_PS_POSITION_SAMPLE_OFFSET
,
1001 SWR_PS_POSITION_CENTROID_OFFSET
,
1002 SWR_PS_POSITION_OFFSET_MAX
,
1005 enum SWR_BARYCENTRICS_MASK
1007 SWR_BARYCENTRIC_PER_PIXEL_MASK
= 0x1,
1008 SWR_BARYCENTRIC_CENTROID_MASK
= 0x2,
1009 SWR_BARYCENTRIC_PER_SAMPLE_MASK
= 0x4,
1010 SWR_BARYCENTRICS_MASK_MAX
= 0x8
1013 // pixel shader state
1017 PFN_PIXEL_KERNEL pfnPixelShader
; // @llvm_pfn
1020 uint32_t killsPixel
: 1; // pixel shader can kill pixels
1021 uint32_t inputCoverage
: 1; // ps uses input coverage
1022 uint32_t writesODepth
: 1; // pixel shader writes to depth
1023 uint32_t usesSourceDepth
: 1; // pixel shader reads depth
1024 uint32_t shadingRate
: 2; // shading per pixel / sample / coarse pixel
1025 uint32_t numRenderTargets
: 4; // number of render target outputs in use (0-8)
1026 uint32_t posOffset
: 2; // type of offset (none, sample, centroid) to add to pixel position
1027 uint32_t barycentricsMask
: 3; // which type(s) of barycentric coords does the PS interpolate attributes with
1028 uint32_t usesUAV
: 1; // pixel shader accesses UAV
1029 uint32_t forceEarlyZ
: 1; // force execution of early depth/stencil test