From f5bb7fddecc0066a6cfb580538f297cec1def5d8 Mon Sep 17 00:00:00 2001 From: Franklin Wei Date: Fri, 9 Aug 2019 20:31:24 -0400 Subject: quake: merge the remaining Pocket Quake PQ_OPT changes I didn't do _OPT3 because it's disabled in the PQ source. This gives as +0.2FPS boost over the last commit. Change-Id: I9c3c8fb7bd23262beb810da6e9469d6b6c4b2a81 --- apps/plugins/sdl/progs/quake/r_draw.c | 491 +++++++++++++++++++++++++++++++++- 1 file changed, 489 insertions(+), 2 deletions(-) (limited to 'apps/plugins/sdl/progs/quake/r_draw.c') diff --git a/apps/plugins/sdl/progs/quake/r_draw.c b/apps/plugins/sdl/progs/quake/r_draw.c index c4bf95f112..c7da5a8a7e 100644 --- a/apps/plugins/sdl/progs/quake/r_draw.c +++ b/apps/plugins/sdl/progs/quake/r_draw.c @@ -60,6 +60,11 @@ int intsintable[SIN_BUFFER_SIZE]; mvertex_t r_leftenter, r_leftexit; mvertex_t r_rightenter, r_rightexit; +#ifdef USE_PQ_OPT2 +mvertex_fxp_t r_leftenter_fxp, r_leftexit_fxp; +mvertex_fxp_t r_rightenter_fxp, r_rightexit_fxp; +#endif + typedef struct { float u,v; @@ -259,6 +264,311 @@ void R_EmitEdge (mvertex_t *pv0, mvertex_t *pv1) removeedges[v2] = edge; } +#ifdef USE_PQ_OPT2 +void R_EmitEdgeFXP_fxp (mvertex_fxp_t *pv0, mvertex_fxp_t *pv1) +{ + //This is just like R_EmitEdge_fxp, except that the vertex parameters are + //already fixed point, and don't need to be converted. + + edge_t *edge, *pcheck; + int u_check; + //float u, u_step; + int u_fxp, u_step_fxp; + //vec3_t local, transformed; + int local_fxp[3], transformed_fxp[3]; + int *world; + int v, v2, ceilv0; + //float scale, lzi0, u0, v0; + int scale_fxp, scale2_fxp, lzi0_fxp, u0_fxp, v0_fxp; + int side; + + if (r_lastvertvalid) + { + u0_fxp = r_u1_fxp; + v0_fxp = r_v1_fxp; + lzi0_fxp = r_lzi1_fxp; + //lzi0 = r_lzi1; + ceilv0 = r_ceilv1; + } + else + { + //world_fxp=(int)(pv0->position[0]*(float)(2^16)); + world = &pv0->position[0]; + + // transform and project + //VectorSubtract (world, modelorg, local); + //Vector Subtract (and convert) + local_fxp[0]=world[0]-modelorg_fxp[0]; + local_fxp[1]=world[1]-modelorg_fxp[1]; + local_fxp[2]=world[2]-modelorg_fxp[2]; + + //TransformVector (local, transformed); + //transformed_fxp[0] = (int)(local_fxp[0]*vright[0])+(int)(local_fxp[1]*vright[1])+(int)(local_fxp[2]*vright[2]); + //transformed_fxp[1] = (int)(local_fxp[0]*vup[0])+(int)(local_fxp[1]*vup[1])+(int)(local_fxp[2]*vup[2]); + //transformed_fxp[2] = (int)(local_fxp[0]*vpn[0])+(int)(local_fxp[1]*vpn[1])+(int)(local_fxp[2]*vpn[2]); + + //13.19 / 24.8 = 21.11 + transformed_fxp[0] = local_fxp[0]/vright_fxp[0]+local_fxp[1]/vright_fxp[1]+local_fxp[2]/vright_fxp[2]; + transformed_fxp[1] = local_fxp[0]/vup_fxp[0]+local_fxp[1]/vup_fxp[1]+local_fxp[2]/vup_fxp[2]; + transformed_fxp[2] = local_fxp[0]/vpn_fxp[0]+local_fxp[1]/vpn_fxp[1]+local_fxp[2]/vpn_fxp[2]; + + if (transformed_fxp[2] < (int)(NEAR_CLIP*2048.0)) + transformed_fxp[2] = (int)(NEAR_CLIP*2048.0); + + transformed_fxp[0]/=16; //21.11->25.7 + transformed_fxp[1]/=16; //21.11->25.7 + transformed_fxp[2]/=8; //21.11->24.8 + + lzi0_fxp=transformed_fxp[2]; + //lzi0 = (float)(1.0 / transformed[2]); + + // FIXME: build x/yscale into transform? + //scale = xscale * lzi0; + //u0 = (xcenter + scale*transformed[0]); + + //Dan: Is this needed? + if (!transformed_fxp[2]) scale_fxp=0; + else scale_fxp=xscale_fxp/transformed_fxp[2]; //9.23 / 24.8 = 17.15 + scale2_fxp=transformed_fxp[0]*(scale_fxp); // 25.7 * 17.15 = 10.22 + + if (transformed_fxp[0]<0) { + if (scale2_fxp>0) scale2_fxp=-511*4194304; + } else { + if (scale2_fxp<0) scale2_fxp=511*4194304; + } + + u0_fxp=scale2_fxp+xcenter_fxp; + + if (u0_fxp < r_refdef_fvrectx_adj_fxp) + u0_fxp = r_refdef_fvrectx_adj_fxp; + if (u0_fxp > r_refdef_fvrectright_adj_fxp) + u0_fxp = r_refdef_fvrectright_adj_fxp; + + //scale = yscale * lzi0; + //v0 = (ycenter - scale*transformed[1]); + + //Dan: Is this needed? + if (!transformed_fxp[2]) scale_fxp=0; + else scale_fxp=yscale_fxp/transformed_fxp[2]; //9.23 / 24.8 = 17.15 + scale2_fxp=transformed_fxp[1]*(scale_fxp); // 25.7 * 17.15 = 10.22 + + if (transformed_fxp[1]<0) { + if (scale2_fxp>0) scale2_fxp=-511*4194304; + } else { + if (scale2_fxp<0) scale2_fxp=511*4194304; //255*8388608; + } + + v0_fxp = ycenter_fxp-scale2_fxp; + + if (v0_fxp < r_refdef_fvrecty_adj_fxp) + v0_fxp = r_refdef_fvrecty_adj_fxp; + if (v0_fxp > r_refdef_fvrectbottom_adj_fxp) + v0_fxp = r_refdef_fvrectbottom_adj_fxp; + + ceilv0 = v0_fxp/4194304; + if (v0_fxp&0x3FFFFF) ceilv0++; + } + + //world(pv1->position[0]*(float)(2^16)); + world = &pv1->position[0]; + +// transform and project + //VectorSubtract (world, modelorg, local); + //Vector Subtract (and convert) + local_fxp[0]=world[0]-modelorg_fxp[0]; + local_fxp[1]=world[1]-modelorg_fxp[1]; + local_fxp[2]=world[2]-modelorg_fxp[2]; + + //TransformVector (local, transformed); + //transformed_fxp[0] = ((int)(local_fxp[0]*vright[0]))+((int)(local_fxp[1]*vright[1]))+((int)(local_fxp[2]*vright[2])); + //transformed_fxp[1] = ((int)(local_fxp[0]*vup[0]))+((int)(local_fxp[1]*vup[1]))+((int)(local_fxp[2]*vup[2])); + //transformed_fxp[2] = ((int)(local_fxp[0]*vpn[0]))+((int)(local_fxp[1]*vpn[1]))+((int)(local_fxp[2]*vpn[2])); + + transformed_fxp[0] = local_fxp[0]/vright_fxp[0]+local_fxp[1]/vright_fxp[1]+local_fxp[2]/vright_fxp[2]; + transformed_fxp[1] = local_fxp[0]/vup_fxp[0]+local_fxp[1]/vup_fxp[1]+local_fxp[2]/vup_fxp[2]; + transformed_fxp[2] = local_fxp[0]/vpn_fxp[0]+local_fxp[1]/vpn_fxp[1]+local_fxp[2]/vpn_fxp[2]; + + if (transformed_fxp[2] < (int)(NEAR_CLIP*2048.0)) + transformed_fxp[2] = (int)(NEAR_CLIP*2048.0); + + transformed_fxp[0]/=16; + transformed_fxp[1]/=16; + transformed_fxp[2]/=8; + + r_lzi1_fxp=transformed_fxp[2]; + //r_lzi1 = (float)(1.0 / transformed[2]); + //scale = xscale * r_lzi1; + + //Dan: Is this needed? + if (!transformed_fxp[2]) scale_fxp=0; + else scale_fxp=xscale_fxp/transformed_fxp[2]; //9.23 / 24.8 = 17.15 + scale2_fxp=transformed_fxp[0]*(scale_fxp); // 24.8 * 17.15 = 9.23 //21.11 + + if (transformed_fxp[0]<0) { + if (scale2_fxp>0) scale2_fxp=-511*4194304; + } else { + if (scale2_fxp<0) scale2_fxp=511*4194304; + } + + //r_u1 = (xcenter + scale*transformed[0]); + r_u1_fxp = xcenter_fxp + scale2_fxp; + if (r_u1_fxp < r_refdef_fvrectx_adj_fxp) + r_u1_fxp = r_refdef_fvrectx_adj_fxp; + if (r_u1_fxp > r_refdef_fvrectright_adj_fxp) + r_u1_fxp = r_refdef_fvrectright_adj_fxp; + + //scale = yscale * r_lzi1; + //r_v1 = (ycenter - scale*transformed[1]); + + //Dan: Is this needed? + if (!transformed_fxp[2]) scale_fxp=0; + else scale_fxp=yscale_fxp/transformed_fxp[2]; //9.23 / 24.8 = 17.15 + scale2_fxp=transformed_fxp[1]*(scale_fxp); // 23.9 * 17.15 = 9.23 //21.11 + + if (transformed_fxp[1]<0) { + if (scale2_fxp>0) scale2_fxp=-511*4194304; + } else { + if (scale2_fxp<0) scale2_fxp=511*4194304; + } + + r_v1_fxp = ycenter_fxp - scale2_fxp; + if (r_v1_fxp < r_refdef_fvrecty_adj_fxp) + r_v1_fxp = r_refdef_fvrecty_adj_fxp; + if (r_v1_fxp > r_refdef_fvrectbottom_adj_fxp) + r_v1_fxp = r_refdef_fvrectbottom_adj_fxp; + + //if (r_lzi1 > lzi0) + // lzi0 = r_lzi1; + if (r_lzi1_fxp < lzi0_fxp) + lzi0_fxp = r_lzi1_fxp; + + //if (lzi0 > r_nearzi) // for mipmap finding + // r_nearzi = lzi0; + if (/*128*/128.0/lzi0_fxp > r_nearzi) { // for mipmap finding + //if (!lzi0_fxp) r_nearzi=0; + //else + r_nearzi = (float)(128.0/lzi0_fxp); + } + +// for right edges, all we want is the effect on 1/z + if (r_nearzionly) + return; + + r_emitted = 1; + + //r_ceilv1 = (int) ceil(r_v1); + r_ceilv1 = r_v1_fxp/4194304; + if (r_v1_fxp&0x3FFFFF) r_ceilv1++; + + +// create the edge + if (ceilv0 == r_ceilv1) + { + // we cache unclipped horizontal edges as fully clipped + if (cacheoffset != 0x7FFFFFFF) + { + cacheoffset = FULLY_CLIPPED_CACHED | + (r_framecount & FRAMECOUNT_MASK); + } + + return; // horizontal edge + } + + side = ceilv0 > r_ceilv1; + + edge = edge_p++; + + edge->owner = NULL; + + edge->owner = r_pedge; + + //Dan: Is this needed? + if (!lzi0_fxp) edge->nearzi=0.0; + else edge->nearzi = (float)(128.0f/*256.0*//lzi0_fxp); + + if (side == 0) + { + int tmp; + // trailing edge (go from p1 to p2) + v = ceilv0; + v2 = r_ceilv1 - 1; + + edge->surfs[0] = surface_p - surfaces; + edge->surfs[1] = 0; + + //u_step = ((r_u1 - u0) / (r_v1 - v0)); + //u = u0 + ((float)v - v0) * u_step; + + tmp=((r_v1_fxp - v0_fxp)>>10); + if (tmp) + u_step_fxp=(r_u1_fxp - u0_fxp) / tmp; //10.22 / 15.12 = 22.10 + else + u_step_fxp=0; + u_fxp = u0_fxp + ((v*4194304 - v0_fxp)>>12 * u_step_fxp>>12); + } + else + { + int tmp; + // leading edge (go from p2 to p1) + v2 = ceilv0 - 1; + v = r_ceilv1; + + edge->surfs[0] = 0; + edge->surfs[1] = surface_p - surfaces; + + //u_step = ((u0 - r_u1) / (v0 - r_v1)); + //u = r_u1 + ((float)v - r_v1) * u_step; + + tmp=((v0_fxp - r_v1_fxp)>>10); + if (tmp) + u_step_fxp = (u0_fxp - r_u1_fxp) / tmp; + else + u_step_fxp=0; + u_fxp = r_u1_fxp + ((v*4194304 - r_v1_fxp)>>12 * u_step_fxp>>12); + + } + //edge->u_step = tmp*0x100000; + //edge->u = (int)(/*(u_fxp/65536)*0x100000*/u_fxp/4 + 0xFFFFF); + + edge->u_step = u_step_fxp*1024;///16; //tmp*0x100000; + edge->u = (int)(/*(u_fxp/65536)*0x100000*/u_fxp/4 + 0xFFFFF); + +// we need to do this to avoid stepping off the edges if a very nearly +// horizontal edge is less than epsilon above a scan, and numeric error causes +// it to incorrectly extend to the scan, and the extension of the line goes off +// the edge of the screen +// FIXME: is this actually needed? + if (edge->u < r_refdef.vrect_x_adj_shift20) + edge->u = r_refdef.vrect_x_adj_shift20; + if (edge->u > r_refdef.vrectright_adj_shift20) + edge->u = r_refdef.vrectright_adj_shift20; + +// +// sort the edge in normally +// + u_check = edge->u; + if (edge->surfs[0]) + u_check++; // sort trailers after leaders + + if (!newedges[v] || newedges[v]->u >= u_check) + { + edge->next = newedges[v]; + newedges[v] = edge; + } + else + { + pcheck = newedges[v]; + while (pcheck->next && pcheck->next->u < u_check) + pcheck = pcheck->next; + edge->next = pcheck->next; + pcheck->next = edge; + } + + edge->nextremove = removeedges[v2]; + removeedges[v2] = edge; +} +#endif + #ifdef USE_PQ_OPT1 void R_EmitEdge_fxp (mvertex_t *pv0, mvertex_t *pv1) { @@ -560,6 +870,138 @@ void R_EmitEdge_fxp (mvertex_t *pv0, mvertex_t *pv1) } #endif +#ifdef USE_PQ_OPT2 +void R_ClipEdge_fxp (mvertex_fxp_t pv0[3], mvertex_fxp_t pv1[3], clipplane_fxp_t *clip) +{ + int d0_fxp, d1_fxp, f_fxp; + mvertex_fxp_t clipvert_fxp; + //12.20 + + if (clip) + { + do + { + //13.19 / 20.12 = 25.7 + d0_fxp=(pv0->position[0])/clip->normal[0]+(pv0->position[1])/clip->normal[1]+(pv0->position[2])/clip->normal[2]; + d1_fxp=(pv1->position[0])/clip->normal[0]+(pv1->position[1])/clip->normal[1]+(pv1->position[2])/clip->normal[2]; + d0_fxp-=(clip->dist); + d1_fxp-=(clip->dist); + //d0 = DotProduct (pv0->position, clip->normal) - clip->dist; + //d1 = DotProduct (pv1->position, clip->normal) - clip->dist; + + if (d0_fxp >= 0) + { + // point 0 is unclipped + if (d1_fxp >= 0) + { + // both points are unclipped + continue; + } + + // only point 1 is clipped + + // we don't cache clipped edges + cacheoffset = 0x7FFFFFFF; + + if (!(d0_fxp)) + f_fxp=2<<29; + else { + f_fxp = (((d0_fxp - d1_fxp)<<7)/(d0_fxp)); //(25.7->18.14) / 25.7 = 25.7 + if (!f_fxp) f_fxp=2<<29; + } + + + clipvert_fxp.position[0] = pv0->position[0] + + (((pv1->position[0] - pv0->position[0])/f_fxp)<<7); //13.19 / 25.7 = 20.12 + + clipvert_fxp.position[1] = pv0->position[1] + + (((pv1->position[1] - pv0->position[1])/f_fxp)<<7); + + clipvert_fxp.position[2] = pv0->position[2] + + (((pv1->position[2] - pv0->position[2])/f_fxp)<<7); + + if (clip->leftedge) + { + r_leftclipped = true; + r_leftexit_fxp = clipvert_fxp; + } + else if (clip->rightedge) + { + r_rightclipped = true; + r_rightexit_fxp = clipvert_fxp; + } + + R_ClipEdge_fxp (pv0, &clipvert_fxp, clip->next); + return; + } + else + { + // point 0 is clipped + if (d1_fxp < 0) + { + // both points are clipped + // we do cache fully clipped edges + if (!r_leftclipped) + cacheoffset = FULLY_CLIPPED_CACHED | + (r_framecount & FRAMECOUNT_MASK); + return; + } + + // only point 0 is clipped + r_lastvertvalid = false; + + // we don't cache partially clipped edges + cacheoffset = 0x7FFFFFFF; + + if (!d0_fxp) + f_fxp=2<<29; + else { + f_fxp = (((d0_fxp - d1_fxp)<<7)/d0_fxp); //12.20 / 20.12 = 24.8 + if (!f_fxp) f_fxp=2<<29; + } + + clipvert_fxp.position[0] = pv0->position[0] + + (((pv1->position[0] - pv0->position[0])/f_fxp)<<7); //12.20 / 24.8 = 20.12 + + clipvert_fxp.position[1] = pv0->position[1] + + (((pv1->position[1] - pv0->position[1])/f_fxp)<<7); + + clipvert_fxp.position[2] = pv0->position[2] + + (((pv1->position[2] - pv0->position[2])/f_fxp)<<7); + + if (clip->leftedge) + { + r_leftclipped = true; + r_leftenter_fxp = clipvert_fxp; + } + else if (clip->rightedge) + { + r_rightclipped = true; + r_rightenter_fxp = clipvert_fxp; + } + + R_ClipEdge_fxp (&clipvert_fxp, pv1, clip->next); + return; + } + } while ((clip = clip->next) != NULL); + } +/* +{ + mvertex_t p0, p1; + p0.position[0]=pv0->position[0]/524288.0f; + p0.position[1]=pv0->position[1]/524288.0f; + p0.position[2]=pv0->position[2]/524288.0f; + p1.position[0]=pv1->position[0]/524288.0f; + p1.position[1]=pv1->position[1]/524288.0f; + p1.position[2]=pv1->position[2]/524288.0f; + R_EmitEdge (&p0, &p1); +} +*/ +// add the edge + R_EmitEdgeFXP_fxp (pv0, pv1); +} +#endif + /* ================ R_ClipEdge @@ -705,6 +1147,9 @@ void R_RenderFace (msurface_t *fa, int clipflags) vec3_t p_normal; medge_t *pedges, tedge; clipplane_t *pclip; +#ifdef USE_PQ_OPT2 + clipplane_fxp_t *pclip_fxp; +#endif // skip out if no more surfs if ((surface_p) >= surf_max) @@ -734,6 +1179,19 @@ void R_RenderFace (msurface_t *fa, int clipflags) } } +#ifdef USE_PQ_OPT2 + pclip_fxp = NULL; + + for (i=3, mask = 0x08 ; i>=0 ; i--, mask >>= 1) + { + if (clipflags & mask) + { + view_clipplanes_fxp[i].next = pclip_fxp; + pclip_fxp = &view_clipplanes_fxp[i]; + } + } +#endif + // push the edges through r_emitted = 0; r_nearzi = 0; @@ -779,9 +1237,16 @@ void R_RenderFace (msurface_t *fa, int clipflags) // assume it's cacheable cacheoffset = (byte *)edge_p - (byte *)r_edges; r_leftclipped = r_rightclipped = false; +#ifndef USE_PQ_OPT2 R_ClipEdge (&r_pcurrentvertbase[r_pedge->v[0]], &r_pcurrentvertbase[r_pedge->v[1]], pclip); +#else + R_ClipEdge_fxp (&r_pcurrentvertbase_fxp[r_pedge->v[0]], + &r_pcurrentvertbase_fxp[r_pedge->v[1]], + pclip_fxp); +#endif + r_pedge->cachededgeoffset = cacheoffset; if (r_leftclipped) @@ -825,9 +1290,15 @@ void R_RenderFace (msurface_t *fa, int clipflags) // assume it's cacheable cacheoffset = (byte *)edge_p - (byte *)r_edges; r_leftclipped = r_rightclipped = false; +#ifndef USE_PQ_OPT2 R_ClipEdge (&r_pcurrentvertbase[r_pedge->v[1]], &r_pcurrentvertbase[r_pedge->v[0]], pclip); +#else + R_ClipEdge_fxp (&r_pcurrentvertbase_fxp[r_pedge->v[1]], + &r_pcurrentvertbase_fxp[r_pedge->v[0]], + pclip_fxp); +#endif r_pedge->cachededgeoffset = cacheoffset; if (r_leftclipped) @@ -845,7 +1316,11 @@ void R_RenderFace (msurface_t *fa, int clipflags) { r_pedge = &tedge; r_lastvertvalid = false; +#ifndef USE_PQ_OPT2 R_ClipEdge (&r_leftexit, &r_leftenter, pclip->next); +#else + R_ClipEdge_fxp (&r_leftexit_fxp, &r_leftenter_fxp, pclip_fxp->next); +#endif } // if there was a clip off the right edge, get the right r_nearzi @@ -854,7 +1329,11 @@ void R_RenderFace (msurface_t *fa, int clipflags) r_pedge = &tedge; r_lastvertvalid = false; r_nearzionly = true; +#ifndef USE_PQ_OPT2 R_ClipEdge (&r_rightexit, &r_rightenter, view_clipplanes[1].next); +#else + R_ClipEdge_fxp (&r_rightexit_fxp, &r_rightenter_fxp, view_clipplanes_fxp[1].next); +#endif } // if no edges made it out, return without posting the surface @@ -876,14 +1355,22 @@ void R_RenderFace (msurface_t *fa, int clipflags) // FIXME: cache this? TransformVector (pplane->normal, p_normal); // FIXME: cache this? - distinv = 1.0 / (pplane->dist - DotProduct (modelorg, pplane->normal)); + distinv = (float) 1.0 / (pplane->dist - DotProduct (modelorg, pplane->normal)); +#ifndef USE_PQ_OPT3 surface_p->d_zistepu = p_normal[0] * xscaleinv * distinv; surface_p->d_zistepv = -p_normal[1] * yscaleinv * distinv; surface_p->d_ziorigin = p_normal[2] * distinv - xcenter * surface_p->d_zistepu - ycenter * surface_p->d_zistepv; - +#else + surface_p->d_zistepu_fxp=(int)(p_normal[0] * xscaleinv * distinv*4194304.0f); + surface_p->d_zistepv_fxp=(int)(-p_normal[1] * yscaleinv * distinv*4194304.0f); + surface_p->d_ziorigin_fxp=((int)(p_normal[2] * distinv * 4194304.0f)) - + ((int)(xcenter * surface_p->d_zistepu_fxp)) - + ((int)(ycenter * surface_p->d_zistepv_fxp)); +#endif + //JDC VectorCopy (r_worldmodelorg, surface_p->modelorg); surface_p++; } -- cgit v1.2.3