dolphin/Source/Core/VideoCommon/VertexShaderGen.cpp
Scott Mansell 1b771deb56 Move worldpos into it's own varying.
Previously it was packed into spare slots in clippos.xy and normal.w,
but it's ugly and more importantly it's causing bugs.

This was discovered during the debugging of a zfreeze branch, which
expected clippos.xy to be xy position coordinates in clipspace (as
the name suggested).

Turns out the stereoscopy shader had also run into this trap, modifying
clippos.x (introducing errors with per-pixel lighting).

This commit has been moved outside of the zfreeze PR for fast merging.
2015-01-03 09:23:09 +13:00

434 lines
15 KiB
C++

// Copyright 2013 Dolphin Emulator Project
// Licensed under GPLv2
// Refer to the license.txt file included.
#include <cmath>
#include "VideoCommon/BPMemory.h"
#include "VideoCommon/CPMemory.h"
#include "VideoCommon/DriverDetails.h"
#include "VideoCommon/LightingShaderGen.h"
#include "VideoCommon/NativeVertexFormat.h"
#include "VideoCommon/VertexShaderGen.h"
#include "VideoCommon/VideoConfig.h"
static char text[16768];
template<class T>
static inline void GenerateVertexShader(T& out, u32 components, API_TYPE api_type)
{
// Non-uid template parameters will write to the dummy data (=> gets optimized out)
vertex_shader_uid_data dummy_data;
vertex_shader_uid_data* uid_data = out.template GetUidData<vertex_shader_uid_data>();
if (uid_data == nullptr)
uid_data = &dummy_data;
out.SetBuffer(text);
const bool is_writing_shadercode = (out.GetBuffer() != nullptr);
if (is_writing_shadercode)
text[sizeof(text) - 1] = 0x7C; // canary
_assert_(bpmem.genMode.numtexgens == xfmem.numTexGen.numTexGens);
_assert_(bpmem.genMode.numcolchans == xfmem.numChan.numColorChans);
out.Write("%s", s_lighting_struct);
// uniforms
if (api_type == API_OPENGL)
out.Write("layout(std140%s) uniform VSBlock {\n", g_ActiveConfig.backend_info.bSupportsBindingLayout ? ", binding = 2" : "");
else
out.Write("cbuffer VSBlock {\n");
out.Write(s_shader_uniforms);
out.Write("};\n");
out.Write("struct VS_OUTPUT {\n");
GenerateVSOutputMembers<T>(out, api_type);
out.Write("};\n");
uid_data->numTexGens = xfmem.numTexGen.numTexGens;
uid_data->components = components;
uid_data->pixel_lighting = g_ActiveConfig.bEnablePixelLighting;
if (api_type == API_OPENGL)
{
out.Write("in float4 rawpos; // ATTR%d,\n", SHADER_POSITION_ATTRIB);
if (components & VB_HAS_POSMTXIDX)
out.Write("in int posmtx; // ATTR%d,\n", SHADER_POSMTX_ATTRIB);
if (components & VB_HAS_NRM0)
out.Write("in float3 rawnorm0; // ATTR%d,\n", SHADER_NORM0_ATTRIB);
if (components & VB_HAS_NRM1)
out.Write("in float3 rawnorm1; // ATTR%d,\n", SHADER_NORM1_ATTRIB);
if (components & VB_HAS_NRM2)
out.Write("in float3 rawnorm2; // ATTR%d,\n", SHADER_NORM2_ATTRIB);
if (components & VB_HAS_COL0)
out.Write("in float4 color0; // ATTR%d,\n", SHADER_COLOR0_ATTRIB);
if (components & VB_HAS_COL1)
out.Write("in float4 color1; // ATTR%d,\n", SHADER_COLOR1_ATTRIB);
for (int i = 0; i < 8; ++i)
{
u32 hastexmtx = (components & (VB_HAS_TEXMTXIDX0<<i));
if ((components & (VB_HAS_UV0<<i)) || hastexmtx)
out.Write("in float%d tex%d; // ATTR%d,\n", hastexmtx ? 3 : 2, i, SHADER_TEXTURE0_ATTRIB + i);
}
if (g_ActiveConfig.backend_info.bSupportsGeometryShaders)
{
out.Write("out VertexData {\n");
GenerateVSOutputMembers<T>(out, api_type, g_ActiveConfig.backend_info.bSupportsBindingLayout ? "centroid" : "centroid out");
out.Write("} vs;\n");
}
else
{
// Let's set up attributes
for (size_t i = 0; i < 8; ++i)
{
if (i < xfmem.numTexGen.numTexGens)
{
out.Write("centroid out float3 uv%d;\n", i);
}
}
out.Write("centroid out float4 clipPos;\n");
if (g_ActiveConfig.bEnablePixelLighting)
{
out.Write("centroid out float3 Normal;\n");
out.Write("centroid out float3 WorldPos;\n");
}
out.Write("centroid out float4 colors_0;\n");
out.Write("centroid out float4 colors_1;\n");
}
out.Write("void main()\n{\n");
}
else // D3D
{
out.Write("VS_OUTPUT main(\n");
// inputs
if (components & VB_HAS_NRM0)
out.Write(" float3 rawnorm0 : NORMAL0,\n");
if (components & VB_HAS_NRM1)
out.Write(" float3 rawnorm1 : NORMAL1,\n");
if (components & VB_HAS_NRM2)
out.Write(" float3 rawnorm2 : NORMAL2,\n");
if (components & VB_HAS_COL0)
out.Write(" float4 color0 : COLOR0,\n");
if (components & VB_HAS_COL1)
out.Write(" float4 color1 : COLOR1,\n");
for (int i = 0; i < 8; ++i)
{
u32 hastexmtx = (components & (VB_HAS_TEXMTXIDX0<<i));
if ((components & (VB_HAS_UV0<<i)) || hastexmtx)
out.Write(" float%d tex%d : TEXCOORD%d,\n", hastexmtx ? 3 : 2, i, i);
}
if (components & VB_HAS_POSMTXIDX)
out.Write(" int posmtx : BLENDINDICES,\n");
out.Write(" float4 rawpos : POSITION) {\n");
}
out.Write("VS_OUTPUT o;\n");
// transforms
if (components & VB_HAS_POSMTXIDX)
{
if (is_writing_shadercode && (DriverDetails::HasBug(DriverDetails::BUG_NODYNUBOACCESS) && !DriverDetails::HasBug(DriverDetails::BUG_ANNIHILATEDUBOS)) )
{
// This'll cause issues, but it can't be helped
out.Write("float4 pos = float4(dot(" I_TRANSFORMMATRICES"[0], rawpos), dot(" I_TRANSFORMMATRICES"[1], rawpos), dot(" I_TRANSFORMMATRICES"[2], rawpos), 1);\n");
if (components & VB_HAS_NRMALL)
out.Write("float3 N0 = " I_NORMALMATRICES"[0].xyz, N1 = " I_NORMALMATRICES"[1].xyz, N2 = " I_NORMALMATRICES"[2].xyz;\n");
}
else
{
out.Write("float4 pos = float4(dot(" I_TRANSFORMMATRICES"[posmtx], rawpos), dot(" I_TRANSFORMMATRICES"[posmtx+1], rawpos), dot(" I_TRANSFORMMATRICES"[posmtx+2], rawpos), 1);\n");
if (components & VB_HAS_NRMALL)
{
out.Write("int normidx = posmtx >= 32 ? (posmtx-32) : posmtx;\n");
out.Write("float3 N0 = " I_NORMALMATRICES"[normidx].xyz, N1 = " I_NORMALMATRICES"[normidx+1].xyz, N2 = " I_NORMALMATRICES"[normidx+2].xyz;\n");
}
}
if (components & VB_HAS_NRM0)
out.Write("float3 _norm0 = normalize(float3(dot(N0, rawnorm0), dot(N1, rawnorm0), dot(N2, rawnorm0)));\n");
if (components & VB_HAS_NRM1)
out.Write("float3 _norm1 = float3(dot(N0, rawnorm1), dot(N1, rawnorm1), dot(N2, rawnorm1));\n");
if (components & VB_HAS_NRM2)
out.Write("float3 _norm2 = float3(dot(N0, rawnorm2), dot(N1, rawnorm2), dot(N2, rawnorm2));\n");
}
else
{
out.Write("float4 pos = float4(dot(" I_POSNORMALMATRIX"[0], rawpos), dot(" I_POSNORMALMATRIX"[1], rawpos), dot(" I_POSNORMALMATRIX"[2], rawpos), 1.0);\n");
if (components & VB_HAS_NRM0)
out.Write("float3 _norm0 = normalize(float3(dot(" I_POSNORMALMATRIX"[3].xyz, rawnorm0), dot(" I_POSNORMALMATRIX"[4].xyz, rawnorm0), dot(" I_POSNORMALMATRIX"[5].xyz, rawnorm0)));\n");
if (components & VB_HAS_NRM1)
out.Write("float3 _norm1 = float3(dot(" I_POSNORMALMATRIX"[3].xyz, rawnorm1), dot(" I_POSNORMALMATRIX"[4].xyz, rawnorm1), dot(" I_POSNORMALMATRIX"[5].xyz, rawnorm1));\n");
if (components & VB_HAS_NRM2)
out.Write("float3 _norm2 = float3(dot(" I_POSNORMALMATRIX"[3].xyz, rawnorm2), dot(" I_POSNORMALMATRIX"[4].xyz, rawnorm2), dot(" I_POSNORMALMATRIX"[5].xyz, rawnorm2));\n");
}
if (!(components & VB_HAS_NRM0))
out.Write("float3 _norm0 = float3(0.0, 0.0, 0.0);\n");
out.Write("o.pos = float4(dot(" I_PROJECTION"[0], pos), dot(" I_PROJECTION"[1], pos), dot(" I_PROJECTION"[2], pos), dot(" I_PROJECTION"[3], pos));\n");
out.Write("int4 lacc;\n"
"float3 ldir, h;\n"
"float dist, dist2, attn;\n");
uid_data->numColorChans = xfmem.numChan.numColorChans;
if (xfmem.numChan.numColorChans == 0)
{
if (components & VB_HAS_COL0)
out.Write("o.colors_0 = color0;\n");
else
out.Write("o.colors_0 = float4(1.0, 1.0, 1.0, 1.0);\n");
}
GenerateLightingShader<T>(out, uid_data->lighting, components, "color", "o.colors_");
if (xfmem.numChan.numColorChans < 2)
{
if (components & VB_HAS_COL1)
out.Write("o.colors_1 = color1;\n");
else
out.Write("o.colors_1 = o.colors_0;\n");
}
// special case if only pos and tex coord 0 and tex coord input is AB11
// donko - this has caused problems in some games. removed for now.
bool texGenSpecialCase = false;
/*bool texGenSpecialCase =
((g_main_cp_state.vtx_desc.Hex & 0x60600L) == g_main_cp_state.vtx_desc.Hex) && // only pos and tex coord 0
(g_main_cp_state.vtx_desc.Tex0Coord != NOT_PRESENT) &&
(xfmem.texcoords[0].texmtxinfo.inputform == XF_TEXINPUT_AB11);
*/
// transform texcoords
out.Write("float4 coord = float4(0.0, 0.0, 1.0, 1.0);\n");
for (unsigned int i = 0; i < xfmem.numTexGen.numTexGens; ++i)
{
TexMtxInfo& texinfo = xfmem.texMtxInfo[i];
out.Write("{\n");
out.Write("coord = float4(0.0, 0.0, 1.0, 1.0);\n");
uid_data->texMtxInfo[i].sourcerow = xfmem.texMtxInfo[i].sourcerow;
switch (texinfo.sourcerow)
{
case XF_SRCGEOM_INROW:
_assert_( texinfo.inputform == XF_TEXINPUT_ABC1 );
out.Write("coord = rawpos;\n"); // pos.w is 1
break;
case XF_SRCNORMAL_INROW:
if (components & VB_HAS_NRM0)
{
_assert_( texinfo.inputform == XF_TEXINPUT_ABC1 );
out.Write("coord = float4(rawnorm0.xyz, 1.0);\n");
}
break;
case XF_SRCCOLORS_INROW:
_assert_( texinfo.texgentype == XF_TEXGEN_COLOR_STRGBC0 || texinfo.texgentype == XF_TEXGEN_COLOR_STRGBC1 );
break;
case XF_SRCBINORMAL_T_INROW:
if (components & VB_HAS_NRM1)
{
_assert_( texinfo.inputform == XF_TEXINPUT_ABC1 );
out.Write("coord = float4(rawnorm1.xyz, 1.0);\n");
}
break;
case XF_SRCBINORMAL_B_INROW:
if (components & VB_HAS_NRM2)
{
_assert_( texinfo.inputform == XF_TEXINPUT_ABC1 );
out.Write("coord = float4(rawnorm2.xyz, 1.0);\n");
}
break;
default:
_assert_(texinfo.sourcerow <= XF_SRCTEX7_INROW);
if (components & (VB_HAS_UV0<<(texinfo.sourcerow - XF_SRCTEX0_INROW)) )
out.Write("coord = float4(tex%d.x, tex%d.y, 1.0, 1.0);\n", texinfo.sourcerow - XF_SRCTEX0_INROW, texinfo.sourcerow - XF_SRCTEX0_INROW);
break;
}
// first transformation
uid_data->texMtxInfo[i].texgentype = xfmem.texMtxInfo[i].texgentype;
switch (texinfo.texgentype)
{
case XF_TEXGEN_EMBOSS_MAP: // calculate tex coords into bump map
if (components & (VB_HAS_NRM1|VB_HAS_NRM2))
{
// transform the light dir into tangent space
uid_data->texMtxInfo[i].embosslightshift = xfmem.texMtxInfo[i].embosslightshift;
uid_data->texMtxInfo[i].embosssourceshift = xfmem.texMtxInfo[i].embosssourceshift;
out.Write("ldir = normalize(" LIGHT_POS".xyz - pos.xyz);\n", LIGHT_POS_PARAMS(texinfo.embosslightshift));
out.Write("o.tex%d.xyz = o.tex%d.xyz + float3(dot(ldir, _norm1), dot(ldir, _norm2), 0.0);\n", i, texinfo.embosssourceshift);
}
else
{
_assert_(0); // should have normals
uid_data->texMtxInfo[i].embosssourceshift = xfmem.texMtxInfo[i].embosssourceshift;
out.Write("o.tex%d.xyz = o.tex%d.xyz;\n", i, texinfo.embosssourceshift);
}
break;
case XF_TEXGEN_COLOR_STRGBC0:
_assert_(texinfo.sourcerow == XF_SRCCOLORS_INROW);
out.Write("o.tex%d.xyz = float3(o.colors_0.x, o.colors_0.y, 1);\n", i);
break;
case XF_TEXGEN_COLOR_STRGBC1:
_assert_(texinfo.sourcerow == XF_SRCCOLORS_INROW);
out.Write("o.tex%d.xyz = float3(o.colors_1.x, o.colors_1.y, 1);\n", i);
break;
case XF_TEXGEN_REGULAR:
default:
uid_data->texMtxInfo_n_projection |= xfmem.texMtxInfo[i].projection << i;
if (components & (VB_HAS_TEXMTXIDX0<<i))
{
out.Write("int tmp = int(tex%d.z);\n", i);
if (texinfo.projection == XF_TEXPROJ_STQ)
out.Write("o.tex%d.xyz = float3(dot(coord, " I_TRANSFORMMATRICES"[tmp]), dot(coord, " I_TRANSFORMMATRICES"[tmp+1]), dot(coord, " I_TRANSFORMMATRICES"[tmp+2]));\n", i);
else
out.Write("o.tex%d.xyz = float3(dot(coord, " I_TRANSFORMMATRICES"[tmp]), dot(coord, " I_TRANSFORMMATRICES"[tmp+1]), 1);\n", i);
}
else
{
if (texinfo.projection == XF_TEXPROJ_STQ)
out.Write("o.tex%d.xyz = float3(dot(coord, " I_TEXMATRICES"[%d]), dot(coord, " I_TEXMATRICES"[%d]), dot(coord, " I_TEXMATRICES"[%d]));\n", i, 3*i, 3*i+1, 3*i+2);
else
out.Write("o.tex%d.xyz = float3(dot(coord, " I_TEXMATRICES"[%d]), dot(coord, " I_TEXMATRICES"[%d]), 1);\n", i, 3*i, 3*i+1);
}
break;
}
uid_data->dualTexTrans_enabled = xfmem.dualTexTrans.enabled;
// CHECKME: does this only work for regular tex gen types?
if (xfmem.dualTexTrans.enabled && texinfo.texgentype == XF_TEXGEN_REGULAR)
{
const PostMtxInfo& postInfo = xfmem.postMtxInfo[i];
uid_data->postMtxInfo[i].index = xfmem.postMtxInfo[i].index;
int postidx = postInfo.index;
out.Write("float4 P0 = " I_POSTTRANSFORMMATRICES"[%d];\n"
"float4 P1 = " I_POSTTRANSFORMMATRICES"[%d];\n"
"float4 P2 = " I_POSTTRANSFORMMATRICES"[%d];\n",
postidx&0x3f, (postidx+1)&0x3f, (postidx+2)&0x3f);
if (texGenSpecialCase)
{
// no normalization
// q of input is 1
// q of output is unknown
// multiply by postmatrix
out.Write("o.tex%d.xyz = float3(dot(P0.xy, o.tex%d.xy) + P0.z + P0.w, dot(P1.xy, o.tex%d.xy) + P1.z + P1.w, 0.0);\n", i, i, i);
}
else
{
uid_data->postMtxInfo[i].normalize = xfmem.postMtxInfo[i].normalize;
if (postInfo.normalize)
out.Write("o.tex%d.xyz = normalize(o.tex%d.xyz);\n", i, i);
// multiply by postmatrix
out.Write("o.tex%d.xyz = float3(dot(P0.xyz, o.tex%d.xyz) + P0.w, dot(P1.xyz, o.tex%d.xyz) + P1.w, dot(P2.xyz, o.tex%d.xyz) + P2.w);\n", i, i, i, i);
}
}
out.Write("}\n");
}
// clipPos/w needs to be done in pixel shader, not here
out.Write("o.clipPos = o.pos;\n");
if (g_ActiveConfig.bEnablePixelLighting)
{
out.Write("o.Normal = _norm0;\n");
out.Write("o.WorldPos = pos.xyz;\n");
if (components & VB_HAS_COL0)
out.Write("o.colors_0 = color0;\n");
if (components & VB_HAS_COL1)
out.Write("o.colors_1 = color1;\n");
}
//write the true depth value, if the game uses depth textures pixel shaders will override with the correct values
//if not early z culling will improve speed
if (api_type == API_D3D)
{
out.Write("o.pos.z = o.pos.w + o.pos.z;\n");
}
else // OGL
{
// this results in a scale from -1..0 to -1..1 after perspective
// divide
out.Write("o.pos.z = o.pos.w + o.pos.z * 2.0;\n");
// the next steps of the OGL pipeline are:
// (x_c,y_c,z_c,w_c) = o.pos //switch to OGL spec terminology
// clipping to -w_c <= (x_c,y_c,z_c) <= w_c
// (x_d,y_d,z_d) = (x_c,y_c,z_c)/w_c//perspective divide
// z_w = (f-n)/2*z_d + (n+f)/2
// z_w now contains the value to go to the 0..1 depth buffer
//trying to get the correct semantic while not using glDepthRange
//seems to get rather complicated
}
// The console GPU places the pixel center at 7/12 in screen space unless
// antialiasing is enabled, while D3D and OpenGL place it at 0.5. This results
// in some primitives being placed one pixel too far to the bottom-right,
// which in turn can be critical if it happens for clear quads.
// Hence, we compensate for this pixel center difference so that primitives
// get rasterized correctly.
out.Write("o.pos.xy = o.pos.xy - " I_PIXELCENTERCORRECTION".xy;\n");
if (api_type == API_OPENGL)
{
if (g_ActiveConfig.backend_info.bSupportsGeometryShaders)
{
AssignVSOutputMembers(out, "vs", "o");
}
else
{
// TODO: Pass interface blocks between shader stages even if geometry shaders
// are not supported, however that will require at least OpenGL 3.2 support.
for (unsigned int i = 0; i < xfmem.numTexGen.numTexGens; ++i)
out.Write("uv%d.xyz = o.tex%d;\n", i, i);
out.Write("clipPos = o.clipPos;\n");
if (g_ActiveConfig.bEnablePixelLighting)
{
out.Write("Normal = o.Normal;\n");
out.Write("WorldPos = o.WorldPos;\n");
}
out.Write("colors_0 = o.colors_0;\n");
out.Write("colors_1 = o.colors_1;\n");
}
out.Write("gl_Position = o.pos;\n");
}
else // D3D
{
out.Write("return o;\n");
}
out.Write("}\n");
if (is_writing_shadercode)
{
if (text[sizeof(text) - 1] != 0x7C)
PanicAlert("VertexShader generator - buffer too small, canary has been eaten!");
}
}
void GetVertexShaderUid(VertexShaderUid& object, u32 components, API_TYPE api_type)
{
GenerateVertexShader<VertexShaderUid>(object, components, api_type);
}
void GenerateVertexShaderCode(VertexShaderCode& object, u32 components, API_TYPE api_type)
{
GenerateVertexShader<VertexShaderCode>(object, components, api_type);
}