OrdinaryMagician/vkdoom_m
OrdinaryMagician/vkdoom_m
1
Fork 0
Code Activity

- Fix voxel speed regression in the palette software renderer

Browse source
This commit is contained in:
Magnus Norddahl 2017-06-18 19:45:15 +02:00
commit 8beafbb7da
10 changed files with 271 additions and 44 deletions
Download patch file Download diff file Expand all files Collapse all files

103
src/swrenderer/things/r_voxel.cpp
View file

@ -220,11 +220,12 @@ namespace swrenderer
{
CheckOffscreenBuffer(viewport->RenderTarget->GetWidth(), viewport->RenderTarget->GetHeight(), !!(flags & DVF_SPANSONLY));
}
*/
if (spr->bInMirror)
{
flags |= DVF_MIRRORED;
}
*/
// Render the voxel, either directly to the screen or offscreen.
DrawVoxel(thread, drawerargs, spr->pa.vpos, spr->pa.vang, spr->gpos, spr->Angle,
@ -276,18 +277,18 @@ namespace swrenderer
fixed_t cosang, sinang, sprcosang, sprsinang;
int backx, backy, gxinc, gyinc;
int daxscalerecip, dayscalerecip, cnt, gxstart, gystart, dazscale;
int lx, rx, nx, ny, x1=0, y1=0, x2=0, y2=0, yinc=0;
int yoff, xs=0, ys=0, xe, ye, xi=0, yi=0, cbackx, cbacky, dagxinc, dagyinc;
int lx, rx, nx, ny, x1 = 0, y1 = 0, x2 = 0, y2 = 0, yinc = 0;
int yoff, xs = 0, ys = 0, xe, ye, xi = 0, yi = 0, cbackx, cbacky, dagxinc, dagyinc;
kvxslab_t *voxptr, *voxend;
FVoxelMipLevel *mip;
int z1a[64], z2a[64], yplc[64];
auto viewport = thread->Viewport.get();
const int nytooclose = viewport->viewwindow.centerxwide * 2100, nytoofar = 32768*32768 - 1048576;
const int nytooclose = viewport->viewwindow.centerxwide * 2100, nytoofar = 32768 * 32768 - 1048576;
const int xdimenscale = FLOAT2FIXED(viewport->viewwindow.centerxwide * viewport->YaspectMul / 160);
const double centerxwide_f = viewport->viewwindow.centerxwide;
const double centerxwidebig_f = centerxwide_f * 65536*65536*8;
const double centerxwidebig_f = centerxwide_f * 65536 * 65536 * 8;
// Convert to Build's coordinate system.
fixed_t globalposx = xs_Fix<4>::ToFix(globalpos.X);
@ -329,14 +330,14 @@ namespace swrenderer
minslabz >>= k;
maxslabz >>= k;
daxscale <<= (k+8); dayscale <<= (k+8);
daxscale <<= (k + 8); dayscale <<= (k + 8);
dazscale = FixedDiv(dayscale, FLOAT2FIXED(viewport->BaseYaspectMul));
daxscale = fixed_t(daxscale / viewport->YaspectMul);
daxscale = Scale(daxscale, xdimenscale, viewport->viewwindow.centerxwide << 9);
dayscale = Scale(dayscale, FixedMul(xdimenscale, viewport->viewingrangerecip), viewport->viewwindow.centerxwide << 9);
daxscalerecip = (1<<30) / daxscale;
dayscalerecip = (1<<30) / dayscale;
daxscalerecip = (1 << 30) / daxscale;
dayscalerecip = (1 << 30) / dayscale;
fixed_t piv_x = fixed_t(mip->Pivot.X*256.);
fixed_t piv_y = fixed_t(mip->Pivot.Y*256.);
@ -344,7 +345,7 @@ namespace swrenderer
x = FixedMul(globalposx - dasprx, daxscalerecip);
y = FixedMul(globalposy - daspry, daxscalerecip);
backx = (DMulScale10(x, sprcosang, y, sprsinang) + piv_x) >> 8;
backx = (DMulScale10(x, sprcosang, y, sprsinang) + piv_x) >> 8;
backy = (DMulScale10(y, sprcosang, x, -sprsinang) + piv_y) >> 8;
cbackx = clamp(backx, 0, mip->SizeX - 1);
cbacky = clamp(backy, 0, mip->SizeY - 1);
@ -353,7 +354,7 @@ namespace swrenderer
sprsinang = MulScale14(daxscale, sprsinang);
x = (dasprx - globalposx) - DMulScale18(piv_x, sprcosang, piv_y, -sprsinang);
y = (daspry - globalposy) - DMulScale18(piv_y, sprcosang, piv_x, sprsinang);
y = (daspry - globalposy) - DMulScale18(piv_y, sprcosang, piv_x, sprsinang);
cosang = FixedMul(cosang, dayscalerecip);
sinang = FixedMul(sinang, dayscalerecip);
@ -361,7 +362,7 @@ namespace swrenderer
gxstart = y*cosang - x*sinang;
gystart = x*cosang + y*sinang;
gxinc = DMulScale10(sprsinang, cosang, sprcosang, -sinang);
gyinc = DMulScale10(sprcosang, cosang, sprsinang, sinang);
gyinc = DMulScale10(sprcosang, cosang, sprsinang, sinang);
if ((abs(globalposz - dasprz) >> 10) >= abs(dazscale)) return;
x = 0; y = 0; j = MAX(mip->SizeX, mip->SizeY);
@ -376,18 +377,24 @@ namespace swrenderer
syoff = DivScale21(globalposz - dasprz, FixedMul(dazscale, 0xE800)) + (piv_z << 7);
yoff = (abs(gxinc) + abs(gyinc)) >> 1;
bool useSlabDataBgra = !drawerargs.DrawerNeedsPalInput() && viewport->RenderTarget->IsBgra();
const int maxoutblocks = 100;
VoxelBlock *outblocks = thread->FrameMemory->AllocMemory<VoxelBlock>(maxoutblocks);
int nextoutblock = 0;
for (cnt = 0; cnt < 8; cnt++)
{
switch (cnt)
{
case 0: xs = 0; ys = 0; xi = 1; yi = 1; break;
case 1: xs = mip->SizeX-1; ys = 0; xi = -1; yi = 1; break;
case 2: xs = 0; ys = mip->SizeY-1; xi = 1; yi = -1; break;
case 3: xs = mip->SizeX-1; ys = mip->SizeY-1; xi = -1; yi = -1; break;
case 4: xs = 0; ys = cbacky; xi = 1; yi = 2; break;
case 5: xs = mip->SizeX-1; ys = cbacky; xi = -1; yi = 2; break;
case 6: xs = cbackx; ys = 0; xi = 2; yi = 1; break;
case 7: xs = cbackx; ys = mip->SizeY-1; xi = 2; yi = -1; break;
case 0: xs = 0; ys = 0; xi = 1; yi = 1; break;
case 1: xs = mip->SizeX - 1; ys = 0; xi = -1; yi = 1; break;
case 2: xs = 0; ys = mip->SizeY - 1; xi = 1; yi = -1; break;
case 3: xs = mip->SizeX - 1; ys = mip->SizeY - 1; xi = -1; yi = -1; break;
case 4: xs = 0; ys = cbacky; xi = 1; yi = 2; break;
case 5: xs = mip->SizeX - 1; ys = cbacky; xi = -1; yi = 2; break;
case 6: xs = cbackx; ys = 0; xi = 2; yi = 1; break;
case 7: xs = cbackx; ys = mip->SizeY - 1; xi = 2; yi = -1; break;
}
xe = cbackx; ye = cbacky;
if (cnt < 4)
@ -407,28 +414,28 @@ namespace swrenderer
}
i = sgn(ys - backy) + sgn(xs - backx) * 3 + 4;
switch(i)
switch (i)
{
case 6: case 7: x1 = 0; y1 = 0; break;
case 8: case 5: x1 = gxinc; y1 = gyinc; break;
case 0: case 3: x1 = gyinc; y1 = -gxinc; break;
case 2: case 1: x1 = gxinc+gyinc; y1 = gyinc-gxinc; break;
case 6: case 7: x1 = 0; y1 = 0; break;
case 8: case 5: x1 = gxinc; y1 = gyinc; break;
case 0: case 3: x1 = gyinc; y1 = -gxinc; break;
case 2: case 1: x1 = gxinc + gyinc; y1 = gyinc - gxinc; break;
}
switch(i)
switch (i)
{
case 2: case 5: x2 = 0; y2 = 0; break;
case 0: case 1: x2 = gxinc; y2 = gyinc; break;
case 8: case 7: x2 = gyinc; y2 = -gxinc; break;
case 6: case 3: x2 = gxinc+gyinc; y2 = gyinc-gxinc; break;
case 2: case 5: x2 = 0; y2 = 0; break;
case 0: case 1: x2 = gxinc; y2 = gyinc; break;
case 8: case 7: x2 = gyinc; y2 = -gxinc; break;
case 6: case 3: x2 = gxinc + gyinc; y2 = gyinc - gxinc; break;
}
uint8_t oand = (1 << int(xs<backx)) + (1 << (int(ys<backy)+2));
uint8_t oand = (1 << int(xs < backx)) + (1 << (int(ys < backy) + 2));
uint8_t oand16 = oand + 16;
uint8_t oand32 = oand + 32;
if (yi > 0) { dagxinc = gxinc; dagyinc = FixedMul(gyinc, viewport->viewingrangerecip); }
else { dagxinc = -gxinc; dagyinc = -FixedMul(gyinc, viewport->viewingrangerecip); }
if (yi > 0) { dagxinc = gxinc; dagyinc = FixedMul(gyinc, viewport->viewingrangerecip); }
else { dagxinc = -gxinc; dagyinc = -FixedMul(gyinc, viewport->viewingrangerecip); }
/* Fix for non 90 degree viewing ranges */
/* Fix for non 90 degree viewing ranges */
nxoff = FixedMul(x2 - x1, viewport->viewingrangerecip);
x1 = FixedMul(x1, viewport->viewingrangerecip);
@ -446,7 +453,7 @@ namespace swrenderer
{
if ((ny <= nytooclose) || (ny >= nytoofar)) continue;
voxptr = (kvxslab_t *)(slabxoffs + xyoffs[y]);
voxend = (kvxslab_t *)(slabxoffs + xyoffs[y+1]);
voxend = (kvxslab_t *)(slabxoffs + xyoffs[y + 1]);
if (voxptr >= voxend) continue;
lx = xs_RoundToInt(nx * centerxwide_f / (ny + y1)) + viewport->viewwindow.centerx;
@ -516,7 +523,7 @@ namespace swrenderer
}
else
{
if (z2-z1 >= 1024) yinc = FixedDiv(zleng, z2 - z1);
if (z2 - z1 >= 1024) yinc = FixedDiv(zleng, z2 - z1);
else yinc = (((1 << 24) - 1) / (z2 - z1)) * zleng >> 8;
}
// [RH] Clip each column separately, not just by the first one.
@ -549,7 +556,7 @@ namespace swrenderer
}
const uint8_t *columnColors = col;
if (!drawerargs.DrawerNeedsPalInput() && viewport->RenderTarget->IsBgra())
if (useSlabDataBgra)
{
// The true color slab data array is identical, except its using uint32_t instead of uint8.
//
@ -575,12 +582,31 @@ namespace swrenderer
break;
}
outblocks[nextoutblock].x = lxt + xxl;
outblocks[nextoutblock].y = z1;
outblocks[nextoutblock].width = xxr - xxl;
outblocks[nextoutblock].height = z2 - z1;
outblocks[nextoutblock].vPos = yplc[xxl];
outblocks[nextoutblock].vStep = yinc;
outblocks[nextoutblock].voxels = columnColors;
outblocks[nextoutblock].voxelsCount = zleng;
nextoutblock++;
if (nextoutblock == maxoutblocks)
{
drawerargs.DrawVoxelBlocks(thread, outblocks, nextoutblock);
outblocks = thread->FrameMemory->AllocMemory<VoxelBlock>(maxoutblocks);
nextoutblock = 0;
}
/*
for (int x = xxl; x < xxr; ++x)
{
drawerargs.SetDest(viewport, lxt + x, z1);
drawerargs.SetCount(z2 - z1);
drawerargs.DrawVoxelColumn(thread, yplc[xxl], yinc, columnColors, zleng);
}
*/
/*
if (!(flags & DVF_OFFSCREEN))
@ -618,6 +644,11 @@ namespace swrenderer
}
}
}
if (nextoutblock != 0)
{
drawerargs.DrawVoxelBlocks(thread, outblocks, nextoutblock);
}
}
kvxslab_t *RenderVoxel::GetSlabStart(const FVoxelMipLevel &mip, int x, int y)