quakeforge/tools/qfvis/source/flow.c
Bill Currie d88bd96390 [qfvis] Use cmem for portal flow stack node allocation
The portal flow stack nodes contain a simd vector, which requires
16-byte alignment. However, on 32-bit Windows, malloc returns 8-byte
aligned memory, leading to eventual segfaults. Since pstack_t is 48
bytes on 32-bit systems, it fits nicely into a 64-byte aligned cache
line (or two on 64-bit systems due to being 80 bytes).
2021-08-13 11:33:03 +09:00

526 lines
13 KiB
C

/*
flow.c
PVS PHS generator tool
Copyright (C) 1996-1997 Id Software, Inc.
Copyright (C) 2002 Colin Thompson
This program is free software; you can redistribute it and/or
modify it under the terms of the GNU General Public License
as published by the Free Software Foundation; either version 2
of the License, or (at your option) any later version.
This program is distributed in the hope that it will be useful,
but WITHOUT ANY WARRANTY; without even the implied warranty of
MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.
See the GNU General Public License for more details.
You should have received a copy of the GNU General Public License
along with this program; if not, write to:
Free Software Foundation, Inc.
59 Temple Place - Suite 330
Boston, MA 02111-1307, USA
*/
#ifdef HAVE_CONFIG_H
# include "config.h"
#endif
#include <getopt.h>
#include <stdlib.h>
#ifdef HAVE_UNISTD_H
# include <unistd.h>
#endif
#ifdef HAVE_IO_H
# include <io.h>
#endif
#ifdef HAVE_STRING_H
# include <string.h>
#endif
#ifdef HAVE_STRINGS_H
# include <strings.h>
#endif
#include <getopt.h>
#include <errno.h>
#include <string.h>
#include <stdlib.h>
#include "QF/bspfile.h"
#include "QF/cmem.h"
#include "QF/cmd.h"
#include "QF/mathlib.h"
#include "QF/quakefs.h"
#include "QF/sys.h"
#include "tools/qfvis/include/vis.h"
#include "tools/qfvis/include/options.h"
static int
CheckStack (cluster_t *cluster, threaddata_t *thread)
{
pstack_t *portal;
for (portal = thread->pstack_head.next; portal; portal = portal->next) {
if (portal->cluster == cluster) {
printf ("CheckStack: cluster recursion\n");
return 1;
}
if (!portal->cluster)
break;
}
return 0;
}
static pstack_t *
new_stack (threaddata_t *td)
{
pstack_t *stack;
stack = cmemalloc (td->memsuper, sizeof (pstack_t));
stack->next = 0;
stack->mightsee = set_new_size_r (&td->set_pool, portalclusters);
td->stats.stack_alloc++;
return stack;
}
static sep_t *
new_separator (threaddata_t *thread)
{
sep_t *sep;
sep = CMEMALLOC (32, sep_t, thread->sep, thread->memsuper);
thread->stats.sep_alloc++;
return sep;
}
static void
delete_separator (threaddata_t *thread, sep_t *sep)
{
thread->stats.sep_free++;
CMEMFREE (thread->sep, sep);
}
static void
free_separators (threaddata_t *thread, sep_t *sep_list)
{
unsigned count = thread->stats.sep_alloc - thread->stats.sep_free;
if (count > thread->stats.sep_highwater) {
thread->stats.sep_highwater = count;
}
count = 0;
while (sep_list) {
sep_t *sep = sep_list;
sep_list = sep->next;
delete_separator (thread, sep);
count++;
}
if (count > thread->stats.sep_maxbulk) {
thread->stats.sep_maxbulk = count;
}
}
static inline int
test_zero (float d)
{
if (d < -ON_EPSILON)
return -1;
else if (d > ON_EPSILON)
return 1;
return 0;
}
static int
calc_plane (vec4f_t v1, vec4f_t v2, int flip, vec4f_t p, vec4f_t *plane)
{
vec4f_t length;
if (flip < 0) {
//CrossProduct (v2, v1, plane.normal);
(*plane)[0] = v2[1] * v1[2] - v2[2] * v1[1];
(*plane)[1] = v2[2] * v1[0] - v2[0] * v1[2];
(*plane)[2] = v2[0] * v1[1] - v2[1] * v1[0];
} else {
//CrossProduct (v1, v2, plane.normal);
(*plane)[0] = v1[1] * v2[2] - v1[2] * v2[1];
(*plane)[1] = v1[2] * v2[0] - v1[0] * v2[2];
(*plane)[2] = v1[0] * v2[1] - v1[1] * v2[0];
}
(*plane)[3] = 0;
length = dotf (*plane, *plane);
// if points don't make a valid plane, skip it
if (length[0] < ON_EPSILON)
return 0;
*plane /= vsqrtf (length);
(*plane)[3] = -dotf (p, *plane)[0];
return 1;
}
static inline int
test_plane (vec4f_t plane, const winding_t *pass, int index)
{
int s1, s2;
int k;
vec4f_t d;
k = (index + 1) % pass->numpoints;
d = dotf (pass->points[k], plane);
s1 = test_zero (d[0]);
k = (index + pass->numpoints - 1) % pass->numpoints;
d = dotf (pass->points[k], plane);
s2 = test_zero (d[0]);
if (s1 == 0 && s2 == 0)
return 0;
if (s1 < 0 || s2 < 0)
return 0;
return 1;
}
static inline sep_t *
create_separator (threaddata_t *thread, vec4f_t src_pl, vec4f_t p1, vec4f_t v1,
const winding_t *pass, int index, int flip)
{
int fliptest;
vec4f_t d;
vec4f_t v2;
vec4f_t plane;
sep_t *sep;
d = dotf (pass->points[index], src_pl);
if ((fliptest = test_zero (d[0])) == 0)
return 0; // The point lies in the source plane
v2 = pass->points[index] - p1;
if (!calc_plane (v1, v2, fliptest, pass->points[index], &plane))
return 0; // point does not form a valid plane
if (!test_plane (plane, pass, index))
return 0; // not the right point
sep = new_separator (thread);
// flip the normal if we want the back side
if (flip) {
sep->plane = -plane;
} else {
sep->plane = plane;
}
return sep;
}
/*
Find the planes separating source from pass. The planes form a double
pyramid with source as the base (ie, source's edges will all be in one
plane each) and the vertex of the pyramid is between source and pass.
Edges from pass may or may not be in a plane, but each vertex will be in
at least one plane.
If flip is false, the planes will be such that the space enclosed by the
planes and on the pass side of the vertex are on the front sides of the
planes. If flip is true, then the space on the source side of the vertex
and enclosed by the planes is on the front side of the planes.
// find a vertex of pass that makes a plane that puts all of the
// vertexes of pass on the front side and all of the vertexes of
// source on the back side
*/
static sep_t *
FindSeparators (threaddata_t *thread,
const winding_t *source, vec4f_t src_pl,
const winding_t *pass, int flip)
{
unsigned i, j, l;
vec4f_t v1;
sep_t *separators = 0, *sep;
for (i = 0; i < source->numpoints; i++) {
l = (i + 1) % source->numpoints;
v1 = source->points[l] - source->points[i];
for (j = 0; j < pass->numpoints; j++) {
sep = create_separator (thread, src_pl, source->points[i], v1,
pass, j, flip);
if (sep) {
sep->next = separators;
separators = sep;
break;
}
}
}
return separators;
}
static winding_t *
ClipToSeparators (threaddata_t *thread, const sep_t *separators,
winding_t *target)
{
const sep_t *sep;
for (sep = separators; target && sep; sep = sep->next) {
target = ClipWinding (thread, target, sep->plane, false);
}
return target;
}
static inline set_t *
select_test_set (portal_t *portal, threaddata_t *thread)
{
set_t *test;
if (portal->status == stat_done) {
thread->stats.vistest++;
test = portal->visbits;
} else {
thread->stats.mighttest++;
test = portal->mightsee;
}
return test;
}
static inline int
mightsee_more (set_t *might, const set_t *prev_might, const set_t *test,
const set_t *vis)
{
unsigned i;
set_bits_t more = 0;
// might = intersection (prev_might, test)
// more = (might is not a subset of vis)
for (i = 0; i < SET_WORDS (might); i++) {
might->map[i] = prev_might->map[i] & test->map[i];
more |= might->map[i] & ~vis->map[i];
}
return more != 0;
}
static void
free_winding_memory (threaddata_t *thread, size_t winding_mark)
{
thread->stats.winding_mark = max (thread->stats.winding_mark,
Hunk_LowMark (thread->hunk));
Hunk_RawFreeToLowMark (thread->hunk, winding_mark);
}
/*
RecursiveClusterFlow
Flood fill through the clusters
*/
static void
RecursiveClusterFlow (int clusternum, threaddata_t *thread, pstack_t *prevstack)
{
int i;
set_t *might;
const set_t *test, *vis;
cluster_t *cluster;
pstack_t *stack;
portal_t *target_portal;
vec4f_t backplane;
vec4f_t source_plane, pass_plane;
const winding_t *pass_winding;
winding_t *source_winding, *target_winding;
thread->stats.chains++;
if (!prevstack->next)
prevstack->next = new_stack (thread);
stack = prevstack->next;
stack->cluster = 0;
cluster = &clusters[clusternum];
if (CheckStack(cluster, thread))
return;
// mark the cluster as visible
if (!set_is_member (thread->clustervis, clusternum)) {
set_add (thread->clustervis, clusternum);
thread->base->numcansee++;
}
stack->cluster = cluster;
stack->pass_portal = NULL;
stack->separators[0] = 0;
stack->separators[1] = 0;
might = stack->mightsee;
vis = thread->clustervis;
source_plane = thread->pstack_head.pass_plane;
pass_winding = prevstack->pass_winding;
pass_plane = prevstack->pass_plane;
size_t winding_mark = Hunk_LowMark (thread->hunk);
// check all portals for flowing into other clusters
for (i = 0; i < cluster->numportals; i++) {
target_portal = &cluster->portals[i];
if (!set_is_member (prevstack->mightsee, target_portal->cluster))
continue; // can't possibly see it
// if target_portal can't see anything we haven't already seen, skip it
test = select_test_set (target_portal, thread);
if (!mightsee_more (might, prevstack->mightsee, test, vis)) {
// can't see anything new
continue;
}
// get plane of target_portal, point normal into the neighbor cluster
backplane = -target_portal->plane;
vec4f_t diff = vabsf (pass_plane - backplane);
vec4i_t cmp = diff > (vec4f_t) {0.001, 0.001, 0.001, 0.001};
if (!(cmp[0] || cmp[1] || cmp[2])) { // dist isn't interesting
continue; // can't go out a coplanar face
}
free_winding_memory (thread, winding_mark);
thread->stats.portalcheck++;
target_winding = target_portal->winding;
target_winding = ClipWinding (thread, target_winding, source_plane, false);
if (!target_winding)
continue;
if (!pass_winding) {
// the second cluster can be blocked only if coplanar
stack->source_winding = prevstack->source_winding;
stack->pass_winding = target_winding;
stack->pass_plane = target_portal->plane;
stack->pass_portal = target_portal;
RecursiveClusterFlow (target_portal->cluster, thread, stack);
continue;
}
target_winding = ClipWinding (thread, target_winding, pass_plane, false);
if (!target_winding)
continue;
// copy source_winding because it likely is already a copy and thus
// if it gets clipped away, earlier stack levels will get corrupted
source_winding = CopyWinding (thread, prevstack->source_winding);
source_winding = ClipWinding (thread, source_winding, backplane, false);
if (!source_winding) {
continue;
}
thread->stats.portaltest++;
thread->stats.targettested++;
if (options.level > 0) {
winding_t *old = target_winding;
if (!stack->separators[0])
stack->separators[0] = FindSeparators (thread,
source_winding,
source_plane,
pass_winding, 0);
target_winding = ClipToSeparators (thread,
stack->separators[0],
target_winding);
if (!target_winding) {
thread->stats.targetclipped++;
continue;
}
if (target_winding != old)
thread->stats.targettrimmed++;
}
if (options.level > 1) {
winding_t *old = target_winding;
if (!stack->separators[1])
stack->separators[1] = FindSeparators (thread,
pass_winding,
pass_plane,
source_winding, 1);
target_winding = ClipToSeparators (thread,
stack->separators[1],
target_winding);
if (!target_winding) {
thread->stats.targetclipped++;
continue;
}
if (target_winding != old)
thread->stats.targettrimmed++;
}
thread->stats.sourcetested++;
if (options.level > 2) {
winding_t *old = source_winding;
sep_t *sep;
sep = FindSeparators (thread,
target_winding, target_portal->plane,
pass_winding, 0);
source_winding = ClipToSeparators (thread, sep, source_winding);
free_separators (thread, sep);
if (!source_winding) {
thread->stats.sourceclipped++;
continue;
}
if (source_winding != old)
thread->stats.sourcetrimmed++;
}
if (options.level > 3) {
winding_t *old = source_winding;
sep_t *sep;
sep = FindSeparators (thread, pass_winding, pass_plane,
target_winding, 1);
source_winding = ClipToSeparators (thread, sep, source_winding);
free_separators (thread, sep);
if (!source_winding) {
thread->stats.sourceclipped++;
continue;
}
if (source_winding != old)
thread->stats.sourcetrimmed++;
}
stack->source_winding = source_winding;
stack->pass_winding = target_winding;
stack->pass_plane = target_portal->plane;
stack->pass_portal = target_portal;
thread->stats.portalpass++;
// flow through it for real
RecursiveClusterFlow (target_portal->cluster, thread, stack);
}
free_winding_memory (thread, winding_mark);
free_separators (thread, stack->separators[1]);
free_separators (thread, stack->separators[0]);
}
void
PortalFlow (threaddata_t *data, portal_t *portal)
{
WRLOCK_PORTAL (portal);
if (portal->status != stat_selected)
Sys_Error ("PortalFlow: reflowed");
portal->status = stat_working;
UNLOCK_PORTAL (portal);
portal->visbits = set_new_size_r (&data->set_pool, portalclusters);
data->clustervis = portal->visbits;
data->base = portal;
data->pstack_head.cluster = 0;
data->pstack_head.pass_portal = portal;
data->pstack_head.source_winding = portal->winding;
data->pstack_head.pass_winding = 0;
data->pstack_head.pass_plane = portal->plane;
data->pstack_head.mightsee = portal->mightsee;
data->pstack_head.separators[0] = 0;
data->pstack_head.separators[1] = 0;
RecursiveClusterFlow (portal->cluster, data, &data->pstack_head);
}