piwalker/dolphin
1
0
Fork 0
mirror of https://github.com/dolphin-emu/dolphin.git synced 2025-07-26 07:39:45 -06:00
Code Issues Packages Projects Releases Wiki Activity

First steps of making the pixel shader uid struct more compact.

Browse Source
This commit is contained in:
NeoBrainX
2013-05-01 11:39:30 +02:00
parent 9cb263ad48
commit 32b78a8572
5 changed files with 181 additions and 205 deletions
Download Patch File Download Diff File Expand all files Collapse all files

91
Source/Core/VideoCommon/Src/PixelShaderGen.cpp
View File

@ -241,6 +241,8 @@ static void GeneratePixelShader(T& out, DSTALPHA_MODE dstAlphaMode, API_TYPE Api
pixel_shader_uid_data& uid_data = (&out.template GetUidData<pixel_shader_uid_data>() != NULL)
? out.template GetUidData<pixel_shader_uid_data>() : dummy_data;
ERROR_LOG(VIDEO, "%lu", sizeof(pixel_shader_uid_data));
out.SetBuffer(text);
if (out.GetBuffer() != NULL)
setlocale(LC_NUMERIC, "C"); // Reset locale for compilation
@ -257,9 +259,9 @@ static void GeneratePixelShader(T& out, DSTALPHA_MODE dstAlphaMode, API_TYPE Api
numStages, numTexgen, bpmem.genMode.numindstages);
uid_data.dstAlphaMode = dstAlphaMode;
uid_data.genMode.numindstages = bpmem.genMode.numindstages;
uid_data.genMode.numtevstages = bpmem.genMode.numtevstages;
uid_data.genMode.numtexgens = bpmem.genMode.numtexgens;
uid_data.genMode_numindstages = bpmem.genMode.numindstages;
uid_data.genMode_numtevstages = bpmem.genMode.numtevstages;
uid_data.genMode_numtexgens = bpmem.genMode.numtexgens;
if (ApiType == API_OPENGL)
{
@ -466,7 +468,7 @@ static void GeneratePixelShader(T& out, DSTALPHA_MODE dstAlphaMode, API_TYPE Api
for (unsigned int i = 0; i < numTexgen; ++i)
{
// optional perspective divides
uid_data.texMtxInfo[i].projection = xfregs.texMtxInfo[i].projection;
uid_data.texMtxInfo_n_projection |= xfregs.texMtxInfo[i].projection << i;
if (xfregs.texMtxInfo[i].projection == XF_TEXPROJ_STQ)
{
out.Write("\tif (uv%d.z != 0.0f)", i);
@ -496,7 +498,7 @@ static void GeneratePixelShader(T& out, DSTALPHA_MODE dstAlphaMode, API_TYPE Api
unsigned int texcoord = bpmem.tevindref.getTexCoord(i);
unsigned int texmap = bpmem.tevindref.getTexMap(i);
uid_data.tevindref.SetValues(i, texcoord, texmap);
uid_data.SetTevindrefValues(i, texcoord, texmap);
if (texcoord < numTexgen)
{
out.SetConstantsUsed(C_INDTEXSCALE+i/2,C_INDTEXSCALE+i/2);
@ -566,7 +568,7 @@ static void GeneratePixelShader(T& out, DSTALPHA_MODE dstAlphaMode, API_TYPE Api
// Note: depth textures are disabled if early depth test is enabled
uid_data.ztex_op = bpmem.ztex2.op;
uid_data.per_pixel_depth = per_pixel_depth;
uid_data.fog.fsel = bpmem.fog.c_proj_fsel.fsel;
uid_data.fog_fsel = bpmem.fog.c_proj_fsel.fsel;
// depth texture can safely be ignored if the result won't be written to the depth buffer (early_ztest) and isn't used for fog either
bool skip_ztexture = !per_pixel_depth && !bpmem.fog.c_proj_fsel.fsel;
@ -687,13 +689,12 @@ static void WriteStage(T& out, pixel_shader_uid_data& uid_data, int n, API_TYPE
out.Write("// TEV stage %d\n", n);
uid_data.bHasIndStage |= bHasIndStage << n;
if (n < 8) { uid_data.tevorders_n_texcoord1 |= texcoord << (3 * n); }
else uid_data.tevorders_n_texcoord2 |= texcoord << (3 * n - 24);
uid_data.tevorders_n_texcoord |= texcoord << (3 * n);
if (bHasIndStage)
{
uid_data.tevind_n.bs |= bpmem.tevind[n].bs << (2*n);
uid_data.tevind_n.bt |= bpmem.tevind[n].bt << (2*n);
uid_data.tevind_n.fmt |= bpmem.tevind[n].fmt << (2*n);
uid_data.tevind_n_bs |= bpmem.tevind[n].bs << (2*n);
uid_data.tevind_n_bt |= bpmem.tevind[n].bt << (2*n);
uid_data.tevind_n_fmt |= bpmem.tevind[n].fmt << (2*n);
out.Write("// indirect op\n");
// perform the indirect op on the incoming regular coordinates using indtex%d as the offset coords
@ -708,12 +709,12 @@ static void WriteStage(T& out, pixel_shader_uid_data& uid_data, int n, API_TYPE
out.Write("float3 indtevcrd%d = indtex%d * %s;\n", n, bpmem.tevind[n].bt, tevIndFmtScale[bpmem.tevind[n].fmt]);
// bias
uid_data.tevind_n.Set_bias(n, bpmem.tevind[n].bias);
uid_data.Set_tevind_bias(n, bpmem.tevind[n].bias);
if (bpmem.tevind[n].bias != ITB_NONE )
out.Write("indtevcrd%d.%s += %s;\n", n, tevIndBiasField[bpmem.tevind[n].bias], tevIndBiasAdd[bpmem.tevind[n].fmt]);
// multiply by offset matrix and scale
uid_data.tevind_n.Set_mid(n, bpmem.tevind[n].mid);
uid_data.Set_tevind_mid(n, bpmem.tevind[n].mid);
if (bpmem.tevind[n].mid != 0)
{
if (bpmem.tevind[n].mid <= 3)
@ -750,9 +751,9 @@ static void WriteStage(T& out, pixel_shader_uid_data& uid_data, int n, API_TYPE
// ---------
// Wrapping
// ---------
uid_data.tevind_n.Set_sw(n, bpmem.tevind[n].sw);
uid_data.tevind_n.Set_tw(n, bpmem.tevind[n].tw);
uid_data.tevind_n.fb_addprev |= bpmem.tevind[n].fb_addprev << n;
uid_data.Set_tevind_sw(n, bpmem.tevind[n].sw);
uid_data.Set_tevind_tw(n, bpmem.tevind[n].tw);
uid_data.tevind_n_fb_addprev |= bpmem.tevind[n].fb_addprev << n;
// wrap S
if (bpmem.tevind[n].sw == ITW_OFF)
@ -779,8 +780,26 @@ static void WriteStage(T& out, pixel_shader_uid_data& uid_data, int n, API_TYPE
TevStageCombiner::ColorCombiner &cc = bpmem.combiners[n].colorC;
TevStageCombiner::AlphaCombiner &ac = bpmem.combiners[n].alphaC;
uid_data.combiners[n].colorC.hex = cc.hex & 0xFFFFFF;
uid_data.combiners[n].alphaC.hex = ac.hex & 0xFFFFFF;
uid_data.cc_n_d = cc.d;
uid_data.cc_n_c = cc.c;
uid_data.cc_n_b = cc.b;
uid_data.cc_n_a = cc.a;
uid_data.cc_n_bias = cc.bias;
uid_data.cc_n_op = cc.op;
uid_data.cc_n_clamp = cc.clamp;
uid_data.cc_n_shift = cc.shift;
uid_data.cc_n_dest = cc.dest;
uid_data.ac_n_rswap = ac.rswap;
uid_data.ac_n_tswap = ac.tswap;
uid_data.ac_n_d = ac.d;
uid_data.ac_n_c = ac.c;
uid_data.ac_n_b = ac.b;
uid_data.ac_n_a = ac.a;
uid_data.ac_n_bias = ac.bias;
uid_data.ac_n_op = ac.op;
uid_data.ac_n_clamp = ac.clamp;
uid_data.ac_n_shift = ac.shift;
uid_data.ac_n_dest = ac.dest;
if(cc.a == TEVCOLORARG_RASA || cc.a == TEVCOLORARG_RASC
|| cc.b == TEVCOLORARG_RASA || cc.b == TEVCOLORARG_RASC
@ -790,10 +809,10 @@ static void WriteStage(T& out, pixel_shader_uid_data& uid_data, int n, API_TYPE
|| ac.c == TEVALPHAARG_RASA || ac.d == TEVALPHAARG_RASA)
{
const int i = bpmem.combiners[n].alphaC.rswap;
uid_data.tevksel[i*2 ].swap1 = bpmem.tevksel[i*2 ].swap1;
uid_data.tevksel[i*2+1].swap1 = bpmem.tevksel[i*2+1].swap1;
uid_data.tevksel[i*2 ].swap2 = bpmem.tevksel[i*2 ].swap2;
uid_data.tevksel[i*2+1].swap2 = bpmem.tevksel[i*2+1].swap2;
uid_data.tevksel_n_swap1 = bpmem.tevksel[i*2 ].swap1 << (2 * (i*2 ));
uid_data.tevksel_n_swap1 = bpmem.tevksel[i*2+1].swap1 << (2 * (i*2+1));
uid_data.tevksel_n_swap2 = bpmem.tevksel[i*2 ].swap2 << (2 * (i*2 ));
uid_data.tevksel_n_swap2 = bpmem.tevksel[i*2+1].swap2 << (2 * (i*2+1));
char *rasswap = swapModeTable[bpmem.combiners[n].alphaC.rswap];
out.Write("rastemp = %s.%s;\n", tevRasTable[bpmem.tevorders[n / 2].getColorChan(n & 1)], rasswap);
@ -813,14 +832,14 @@ static void WriteStage(T& out, pixel_shader_uid_data& uid_data, int n, API_TYPE
}
const int i = bpmem.combiners[n].alphaC.tswap;
uid_data.tevksel[i*2 ].swap1 = bpmem.tevksel[i*2 ].swap1;
uid_data.tevksel[i*2+1].swap1 = bpmem.tevksel[i*2+1].swap1;
uid_data.tevksel[i*2 ].swap2 = bpmem.tevksel[i*2 ].swap2;
uid_data.tevksel[i*2+1].swap2 = bpmem.tevksel[i*2+1].swap2;
uid_data.tevksel_n_swap1 = bpmem.tevksel[i*2 ].swap1 << (2 * (i*2 ));
uid_data.tevksel_n_swap1 = bpmem.tevksel[i*2+1].swap1 << (2 * (i*2+1));
uid_data.tevksel_n_swap2 = bpmem.tevksel[i*2 ].swap2 << (2 * (i*2 ));
uid_data.tevksel_n_swap2 = bpmem.tevksel[i*2+1].swap2 << (2 * (i*2+1));
char *texswap = swapModeTable[bpmem.combiners[n].alphaC.tswap];
int texmap = bpmem.tevorders[n/2].getTexMap(n&1);
uid_data.tevindref.SetTexmap(i, texmap);
uid_data.SetTevindrefTexmap(i, texmap);
SampleTexture<T>(out, "textemp", "tevcoord", texswap, texmap, ApiType);
}
else
@ -834,8 +853,8 @@ static void WriteStage(T& out, pixel_shader_uid_data& uid_data, int n, API_TYPE
{
int kc = bpmem.tevksel[n / 2].getKC(n & 1);
int ka = bpmem.tevksel[n / 2].getKA(n & 1);
uid_data.tevksel[n/2].set_kcsel(n & 1, kc);
uid_data.tevksel[n/2].set_kasel(n & 1, ka);
uid_data.set_tevksel_kcsel(n/2, n & 1, kc);
uid_data.set_tevksel_kasel(n/2, n & 1, ka);
out.Write("konsttemp = float4(%s, %s);\n", tevKSelTableC[kc], tevKSelTableA[ka]);
if(kc > 7 || ka > 7)
{
@ -1086,9 +1105,9 @@ static void WriteAlphaTest(T& out, pixel_shader_uid_data& uid_data, API_TYPE Api
// using discard then return works the same in cg and dx9 but not in dx11
out.Write("\tif(!( ");
uid_data.alpha_test.comp0 = bpmem.alpha_test.comp0;
uid_data.alpha_test.logic = bpmem.alpha_test.comp1;
uid_data.alpha_test.logic = bpmem.alpha_test.logic;
uid_data.alpha_test_comp0 = bpmem.alpha_test.comp0;
uid_data.alpha_test_logic = bpmem.alpha_test.comp1;
uid_data.alpha_test_logic = bpmem.alpha_test.logic;
// Lookup the first component from the alpha function table
int compindex = bpmem.alpha_test.comp0;
@ -1117,7 +1136,7 @@ static void WriteAlphaTest(T& out, pixel_shader_uid_data& uid_data, API_TYPE Api
// when the alpha test fail. This is not a correct implementation because
// even if the depth test fails the fragment could be alpha blended, but
// we don't have a choice.
uid_data.alpha_test.use_zcomploc_hack = bpmem.zcontrol.early_ztest && bpmem.zmode.updateenable;
uid_data.alpha_test_use_zcomploc_hack = bpmem.zcontrol.early_ztest && bpmem.zmode.updateenable;
if (!(bpmem.zcontrol.early_ztest && bpmem.zmode.updateenable))
{
out.Write("\t\tdiscard;\n");
@ -1143,11 +1162,11 @@ static const char *tevFogFuncsTable[] =
template<class T>
static void WriteFog(T& out, pixel_shader_uid_data& uid_data)
{
uid_data.fog.fsel = bpmem.fog.c_proj_fsel.fsel;
uid_data.fog_fsel = bpmem.fog.c_proj_fsel.fsel;
if(bpmem.fog.c_proj_fsel.fsel == 0)
return; // no Fog
uid_data.fog.proj = bpmem.fog.c_proj_fsel.proj;
uid_data.fog_proj = bpmem.fog.c_proj_fsel.proj;
out.SetConstantsUsed(C_FOG, C_FOG+1);
if (bpmem.fog.c_proj_fsel.proj == 0)
@ -1166,7 +1185,7 @@ static void WriteFog(T& out, pixel_shader_uid_data& uid_data)
// x_adjust = sqrt((x-center)^2 + k^2)/k
// ze *= x_adjust
// this is completely theoretical as the real hardware seems to use a table intead of calculating the values.
uid_data.fog.RangeBaseEnabled = bpmem.fogRange.Base.Enabled;
uid_data.fog_RangeBaseEnabled = bpmem.fogRange.Base.Enabled;
if (bpmem.fogRange.Base.Enabled)
{
out.SetConstantsUsed(C_FOG+2, C_FOG+2);