quakeforge/libs/models/alias/gl_mesh.c
Bill Currie 12c84046f3 [cvar] Make cvars properly typed
This is an extremely extensive patch as it hits every cvar, and every
usage of the cvars. Cvars no longer store the value they control,
instead, they use a cexpr value object to reference the value and
specify the value's type (currently, a null type is used for strings).
Non-string cvars are passed through cexpr, allowing expressions in the
cvars' settings. Also, cvars have returned to an enhanced version of the
original (id quake) registration scheme.

As a minor benefit, relevant code having direct access to the
cvar-controlled variables is probably a slight optimization as it
removed a pointer dereference, and the variables can be located for data
locality.

The static cvar descriptors are made private as an additional safety
layer, though there's nothing stopping external modification via
Cvar_FindVar (which is needed for adding listeners).

While not used yet (partly due to working out the design), cvars can
have a validation function.

Registering a cvar allows a primary listener (and its data) to be
specified: it will always be called first when the cvar is modified. The
combination of proper listeners and direct access to the controlled
variable greatly simplifies the more complex cvar interactions as much
less null checking is required, and there's no need for one cvar's
callback to call another's.

nq-x11 is known to work at least well enough for the demos. More testing
will come.
2022-04-24 19:15:22 +09:00

544 lines
14 KiB
C

/*
gl_mesh.c
gl_mesh.c: triangle model functions
Copyright (C) 1996-1997 Id Software, Inc.
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
#ifdef HAVE_STRING_H
# include <string.h>
#endif
#ifdef HAVE_STRINGS_H
# include <strings.h>
#endif
#include <stdio.h>
#include "QF/cvar.h"
#include "QF/dstring.h"
#include "QF/mdfour.h"
#include "QF/quakefs.h"
#include "QF/sys.h"
#include "mod_internal.h"
#include "compat.h"
// ALIAS MODEL DISPLAY LIST GENERATION ========================================
static qboolean *used;
static int used_size;
// the command list holds counts and s/t values that are valid for every frame
static int *commands;
static int numcommands;
static int commands_size;
// all frames will have their vertexes rearranged and expanded
// so they are in the order expected by the command list
static int *vertexorder;
static int numorder;
static int vertexorder_size;
static int allverts, alltris;
static int *stripverts;
static int *striptris;
static int stripcount;
static int strip_size;
static inline void
alloc_used (int size)
{
if (size <= used_size)
return;
size = (size + 1023) & ~1023;
used = realloc (used, size * sizeof (used[0]));
if (!used)
Sys_Error ("gl_mesh: out of memory");
used_size = size;
}
static inline void
add_command (int cmd)
{
if (numcommands + 1 > commands_size) {
commands_size += 1024;
commands = realloc (commands, commands_size * sizeof (commands[0]));
if (!commands)
Sys_Error ("gl_mesh: out of memory");
}
commands[numcommands++] = cmd;
}
static inline void
add_vertex (int vert)
{
if (numorder + 1 > vertexorder_size) {
vertexorder_size += 1024;
vertexorder = realloc (vertexorder, vertexorder_size * sizeof (vertexorder[0]));
if (!vertexorder)
Sys_Error ("gl_mesh: out of memory");
}
vertexorder[numorder++] = vert;
}
static inline void
add_strip (int vert, int tri)
{
if (stripcount + 1 > strip_size) {
strip_size += 1024;
stripverts = realloc (stripverts, strip_size * sizeof (stripverts[0]));
striptris = realloc (striptris, strip_size * sizeof (striptris[0]));
if (!stripverts || !striptris)
Sys_Error ("gl_mesh: out of memory");
}
stripverts[stripcount] = vert;
striptris[stripcount] = tri;
stripcount++;
}
static int
StripLength (mod_alias_ctx_t *alias_ctx, int starttri, int startv)
{
aliashdr_t *header = alias_ctx->header;
int m1, m2, j, k;
mtriangle_t *last, *check;
used[starttri] = 2;
last = &alias_ctx->triangles.a[starttri];
stripcount = 0;
add_strip (last->vertindex[(startv) % 3], starttri);
add_strip (last->vertindex[(startv + 1) % 3], starttri);
add_strip (last->vertindex[(startv + 2) % 3], starttri);
m1 = last->vertindex[(startv + 2) % 3];
m2 = last->vertindex[(startv + 1) % 3];
// look for a matching triangle
nexttri:
for (j = starttri + 1, check = &alias_ctx->triangles.a[starttri + 1];
j < header->mdl.numtris; j++, check++) {
if (check->facesfront != last->facesfront)
continue;
for (k = 0; k < 3; k++) {
if (check->vertindex[k] != m1)
continue;
if (check->vertindex[(k + 1) % 3] != m2)
continue;
// this is the next part of the fan
// if we can't use this triangle, this tristrip is done
if (used[j])
goto done;
// the new edge
if (stripcount & 1)
m2 = check->vertindex[(k + 2) % 3];
else
m1 = check->vertindex[(k + 2) % 3];
add_strip (check->vertindex[(k + 2) % 3], j);
used[j] = 2;
goto nexttri;
}
}
done:
// clear the temp used flags
for (j = starttri + 1; j < header->mdl.numtris; j++)
if (used[j] == 2)
used[j] = 0;
return stripcount - 2;
}
static int
FanLength (mod_alias_ctx_t *alias_ctx, int starttri, int startv)
{
aliashdr_t *header = alias_ctx->header;
int m1, m2, j, k;
mtriangle_t *last, *check;
used[starttri] = 2;
last = &alias_ctx->triangles.a[starttri];
stripcount = 0;
add_strip (last->vertindex[(startv) % 3], starttri);
add_strip (last->vertindex[(startv + 1) % 3], starttri);
add_strip (last->vertindex[(startv + 2) % 3], starttri);
m1 = last->vertindex[(startv + 0) % 3];
m2 = last->vertindex[(startv + 2) % 3];
// look for a matching triangle
nexttri:
for (j = starttri + 1, check = &alias_ctx->triangles.a[starttri + 1];
j < header->mdl.numtris; j++, check++) {
if (check->facesfront != last->facesfront)
continue;
for (k = 0; k < 3; k++) {
if (check->vertindex[k] != m1)
continue;
if (check->vertindex[(k + 1) % 3] != m2)
continue;
// this is the next part of the fan
// if we can't use this triangle, this tristrip is done
if (used[j])
goto done;
// the new edge
m2 = check->vertindex[(k + 2) % 3];
add_strip (m2, j);
used[j] = 2;
goto nexttri;
}
}
done:
// clear the temp used flags
for (j = starttri + 1; j < header->mdl.numtris; j++)
if (used[j] == 2)
used[j] = 0;
return stripcount - 2;
}
/*
BuildTris
Generate a list of trifans or strips
for the model, which holds for all frames
*/
static void
BuildTris (mod_alias_ctx_t *alias_ctx)
{
aliashdr_t *header = alias_ctx->header;
float s, t;
int bestlen, len, startv, type, i, j, k;
int besttype = 0;
int *bestverts = 0, *besttris = 0;
// build tristrips
numorder = 0;
numcommands = 0;
stripcount = 0;
alloc_used (header->mdl.numtris);
memset (used, 0, used_size * sizeof (used[0]));
for (i = 0; i < header->mdl.numtris; i++) {
// pick an unused triangle and start the trifan
if (used[i])
continue;
bestlen = 0;
for (type = 0; type < 2; type++) {
// type = 1;
for (startv = 0; startv < 3; startv++) {
if (type == 1)
len = StripLength (alias_ctx, i, startv);
else
len = FanLength (alias_ctx, i, startv);
if (len > bestlen) {
besttype = type;
bestlen = len;
if (bestverts)
free (bestverts);
if (besttris)
free (besttris);
bestverts = stripverts;
besttris = striptris;
stripverts = striptris = 0;
strip_size = 0;
}
}
}
// mark the tris on the best strip as used
for (j = 0; j < bestlen; j++)
used[besttris[j + 2]] = 1;
if (besttype == 1)
add_command (bestlen + 2);
else
add_command (-(bestlen + 2));
for (j = 0; j < bestlen + 2; j++) {
int tmp;
// emit a vertex into the reorder buffer
k = bestverts[j];
add_vertex (k);
// emit s/t coords into the commands stream
s = alias_ctx->stverts.a[k].s;
t = alias_ctx->stverts.a[k].t;
if (!alias_ctx->triangles.a[besttris[0]].facesfront
&& alias_ctx->stverts.a[k].onseam)
s += header->mdl.skinwidth / 2; // on back side
s = (s + 0.5) / header->mdl.skinwidth;
t = (t + 0.5) / header->mdl.skinheight;
memcpy (&tmp, &s, 4);
add_command (tmp);
memcpy (&tmp, &t, 4);
add_command (tmp);
}
}
add_command (0); // end of list marker
Sys_MaskPrintf (SYS_dev, "%3i tri %3i vert %3i cmd\n",
header->mdl.numtris, numorder, numcommands);
allverts += numorder;
alltris += header->mdl.numtris;
if (bestverts)
free (bestverts);
if (besttris)
free (besttris);
}
void
gl_Mod_MakeAliasModelDisplayLists (mod_alias_ctx_t *alias_ctx, void *_m,
int _s, int extra)
{
aliashdr_t *header = alias_ctx->header;
dstring_t *cache, *fullpath;
unsigned char model_digest[MDFOUR_DIGEST_BYTES];
unsigned char mesh_digest[MDFOUR_DIGEST_BYTES];
int i, j;
int *cmds;
QFile *f;
qboolean remesh = true;
qboolean do_cache = false;
cache = dstring_new ();
fullpath = dstring_new ();
if (!gl_alias_render_tri) {
if (gl_mesh_cache && gl_mesh_cache <= header->mdl.numtris) {
do_cache = true;
mdfour (model_digest, (unsigned char *) _m, _s);
// look for a cached version
dstring_copystr (cache, "glquake/");
dstring_appendstr (cache, alias_ctx->mod->path);
QFS_StripExtension (alias_ctx->mod->path + strlen ("progs/"),
cache->str + strlen ("glquake/"));
dstring_appendstr (cache, ".qfms");
f = QFS_FOpenFile (cache->str);
if (f) {
unsigned char d1[MDFOUR_DIGEST_BYTES];
unsigned char d2[MDFOUR_DIGEST_BYTES];
struct mdfour md;
int len, vers;
int nc = 0, no = 0;
int *c = 0, *vo = 0;
memset (d1, 0, sizeof (d1));
memset (d2, 0, sizeof (d2));
Qread (f, &vers, sizeof (int));
Qread (f, &len, sizeof (int));
Qread (f, &nc, sizeof (int));
Qread (f, &no, sizeof (int));
if (vers == 1 && (nc + no) == len) {
c = malloc (((nc + 1023) & ~1023) * sizeof (c[0]));
vo = malloc (((no + 1023) & ~1023) * sizeof (vo[0]));
if (!c || !vo)
Sys_Error ("gl_mesh.c: out of memory");
Qread (f, c, nc * sizeof (c[0]));
Qread (f, vo, no * sizeof (vo[0]));
Qread (f, d1, MDFOUR_DIGEST_BYTES);
Qread (f, d2, MDFOUR_DIGEST_BYTES);
Qclose (f);
mdfour_begin (&md);
mdfour_update (&md, (unsigned char *) &vers, sizeof(int));
mdfour_update (&md, (unsigned char *) &len, sizeof(int));
mdfour_update (&md, (unsigned char *) &nc, sizeof(int));
mdfour_update (&md, (unsigned char *) &no, sizeof(int));
mdfour_update (&md, (unsigned char *) c, nc * sizeof (c[0]));
mdfour_update (&md, (unsigned char *) vo, no * sizeof (vo[0]));
mdfour_update (&md, d1, MDFOUR_DIGEST_BYTES);
mdfour_result (&md, mesh_digest);
if (memcmp (d2, mesh_digest, MDFOUR_DIGEST_BYTES) == 0
&& memcmp (d1, model_digest, MDFOUR_DIGEST_BYTES) == 0) {
remesh = false;
numcommands = nc;
numorder = no;
if (numcommands > commands_size) {
if (commands)
free (commands);
commands_size = (numcommands + 1023) & ~1023;
commands = c;
} else {
memcpy (commands, c, numcommands * sizeof (c[0]));
free(c);
}
if (numorder > vertexorder_size) {
if (vertexorder)
free (vertexorder);
vertexorder_size = (numorder + 1023) & ~1023;
vertexorder = vo;
} else {
memcpy (vertexorder, vo, numorder * sizeof (vo[0]));
free (vo);
}
}
}
}
}
if (remesh) {
// build it from scratch
Sys_MaskPrintf (SYS_dev, "meshing %s...\n", alias_ctx->mod->path);
BuildTris (alias_ctx); // trifans or lists
if (do_cache) {
// save out the cached version
dsprintf (fullpath, "%s/%s", qfs_gamedir->dir.def, cache->str);
f = QFS_WOpen (fullpath->str, 9);
if (f) {
struct mdfour md;
int vers = 1;
int len = numcommands + numorder;
mdfour_begin (&md);
mdfour_update (&md, (unsigned char *) &vers, sizeof (int));
mdfour_update (&md, (unsigned char *) &len, sizeof (int));
mdfour_update (&md, (unsigned char *) &numcommands,
sizeof (int));
mdfour_update (&md, (unsigned char *) &numorder, sizeof (int));
mdfour_update (&md, (unsigned char *) commands,
numcommands * sizeof (commands[0]));
mdfour_update (&md, (unsigned char *) vertexorder,
numorder * sizeof (vertexorder[0]));
mdfour_update (&md, model_digest, MDFOUR_DIGEST_BYTES);
mdfour_result (&md, mesh_digest);
Qwrite (f, &vers, sizeof (int));
Qwrite (f, &len, sizeof (int));
Qwrite (f, &numcommands, sizeof (int));
Qwrite (f, &numorder, sizeof (int));
Qwrite (f, commands, numcommands * sizeof (commands[0]));
Qwrite (f, vertexorder, numorder * sizeof (vertexorder[0]));
Qwrite (f, model_digest, MDFOUR_DIGEST_BYTES);
Qwrite (f, mesh_digest, MDFOUR_DIGEST_BYTES);
Qclose (f);
}
}
}
// save the data out
header->poseverts = numorder;
cmds = Hunk_Alloc (0, numcommands * sizeof (int));
header->commands = (byte *) cmds - (byte *) header;
memcpy (cmds, commands, numcommands * sizeof (int));
} else {
tex_coord_t *tex_coord;
numorder = 0;
for (i=0; i < header->mdl.numtris; i++) {
add_vertex(alias_ctx->triangles.a[i].vertindex[0]);
add_vertex(alias_ctx->triangles.a[i].vertindex[1]);
add_vertex(alias_ctx->triangles.a[i].vertindex[2]);
}
header->poseverts = numorder;
tex_coord = Hunk_Alloc (0, numorder * sizeof(tex_coord_t));
header->tex_coord = (byte *) tex_coord - (byte *) header;
for (i=0; i < numorder; i++) {
float s, t;
int k;
k = vertexorder[i];
s = alias_ctx->stverts.a[k].s;
t = alias_ctx->stverts.a[k].t;
if (!alias_ctx->triangles.a[i/3].facesfront
&& alias_ctx->stverts.a[k].onseam)
s += header->mdl.skinwidth / 2; // on back side
s = (s + 0.5) / header->mdl.skinwidth;
t = (t + 0.5) / header->mdl.skinheight;
tex_coord[i].st[0] = s;
tex_coord[i].st[1] = t;
}
}
if (extra) {
trivertx16_t *verts;
verts = Hunk_Alloc (0, header->numposes * header->poseverts
* sizeof (trivertx16_t));
header->posedata = (byte *) verts - (byte *) header;
for (i = 0; i < header->numposes; i++) {
trivertx_t *pv = alias_ctx->poseverts.a[i];
for (j = 0; j < numorder; j++) {
trivertx16_t v;
// convert MD16's split coordinates into something a little
// saner. The first chunk of vertices is fully compatible with
// IDPO alias models (even the scale). The second chunk is the
// fractional bits of the vertex, giving 8.8. However, it's
// easier for us to multiply everything by 256 and adjust the
// model scale appropriately
VectorMultAdd (pv[vertexorder[j] + header->mdl.numverts].v,
256, pv[vertexorder[j]].v, v.v);
v.lightnormalindex =
alias_ctx->poseverts.a[i][vertexorder[j]].lightnormalindex;
*verts++ = v;
}
}
} else {
trivertx_t *verts;
verts = Hunk_Alloc (0, header->numposes * header->poseverts
* sizeof (trivertx_t));
header->posedata = (byte *) verts - (byte *) header;
for (i = 0; i < header->numposes; i++) {
for (j = 0; j < numorder; j++)
*verts++ = alias_ctx->poseverts.a[i][vertexorder[j]];
}
}
dstring_delete (cache);
dstring_delete (fullpath);
}