qf/quakeforge
0
0
Fork 0
mirror of https://git.code.sf.net/p/quake/quakeforge synced 2024-11-27 06:34:11 +00:00
Code Issues Projects Releases Packages Wiki Activity
quakeforge/libs/video/renderer/gl/gl_rmain.c

296 lines
6.9 KiB
C
Raw Normal View History

initial checkin of most recent newtree and nuq(?) source
2001-02-19 21:15:25 +00:00
/*
gl_rmain.c
*beats CVS with the cluebat*
2003-05-09 01:41:56 +00:00
(no description)
initial checkin of most recent newtree and nuq(?) source
2001-02-19 21:15:25 +00:00
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
re-arrange things so __attribute__ and __builtin_expect are properly autoconfiscated so rcsid will continue to work with gcc 3.3
2003-01-15 15:31:36 +00:00
First pass at updating nq particles. Compiles. Crashes. Taniwha wanted a look. Flee in terror. Also has bunches of cleanups, including string.h/strings.h ifdefs, whitespace, and removal of pointless comments.
2001-05-09 05:41:34 +00:00
#ifdef HAVE_STRING_H
# include <string.h>
#endif
#ifdef HAVE_STRINGS_H
# include <strings.h>
#endif
initial checkin of most recent newtree and nuq(?) source
2001-02-19 21:15:25 +00:00
#include <math.h>
#include <stdio.h>
#include <stdlib.h>
no header includes cvar.h now
2001-05-31 03:41:35 +00:00
#include "QF/cvar.h"
patches to allow for multiple view ports
2004-05-07 03:54:35 +00:00
#include "QF/draw.h"
diff noise reduction
2001-05-14 20:35:32 +00:00
#include "QF/mathlib.h"
First pass at updating nq particles. Compiles. Crashes. Taniwha wanted a look. Flee in terror. Also has bunches of cleanups, including string.h/strings.h ifdefs, whitespace, and removal of pointless comments.
2001-05-09 05:41:34 +00:00
#include "QF/qargs.h"
reduce gl_rmain's dependency on cl*.h
2001-05-21 00:22:35 +00:00
#include "QF/render.h"
diff noise reduction
2001-05-14 20:35:32 +00:00
#include "QF/skin.h"
#include "QF/sound.h"
move the api headers into include/QF and clean up (most of) the resulting mess. target specific files that I don't build won't compile yet. just put QF/ infront of the offending headers. Also move ver_check into libqfutils
2001-03-27 20:33:07 +00:00
#include "QF/sys.h"
diff noise reduction
2001-05-14 20:35:32 +00:00
#include "QF/vid.h"
[scene] Rename libQFentity to libQFscene And start working on scene management.
2021-07-24 05:19:52 +00:00
#include "QF/scene/entity.h"
This is changing more then I wanted to at once, however. The major change is that we no longer require libGL to even exist on the system at compile time for the GL targets, we dynamicly link to the libGL of choice at run time. (This probably breaks most non-linux systems, and all GL targets except -glx, some fixup will be needed.) (This also kills glquake, dead dead DEAD! GONE FOREVER! WHEE!) Some gl_draw cleanup. Commented out equake alias model occlusion test stuff, very experimental. Added the .lo and .la patterns to the .gitignore files. Some minor sbar cleanup. (We don't use the disc in use symbol for anything.)
2001-06-24 09:25:55 +00:00
#include "QF/GL/defines.h"
#include "QF/GL/funcs.h"
[renderer] Remove namehack.h Finally. I never liked it (felt bad adding it in the first place), and it has caused confusion with function and global variable names, but it did let me get the render plugins working.
2022-03-14 05:31:23 +00:00
#include "QF/GL/qf_alias.h"
Make a start on putting the pieces together. Still many undefined and multiply defined symbols, but there seems to be a light in this darkness. I just hope it's not an oncoming train :)
2012-02-22 12:53:17 +00:00
#include "QF/GL/qf_draw.h"
Render iqm models in GL. There are still many issues to sort out, but the basics are working. Problems: rendered fullbright (no lighting done) normals are ignored extra textures (glow etc) not used/loaded 4 models on the screen don't seem to be a problem.
2012-05-17 06:58:29 +00:00
#include "QF/GL/qf_iqm.h"
[renderer] Remove namehack.h Finally. I never liked it (felt bad adding it in the first place), and it has caused confusion with function and global variable names, but it did let me get the render plugins working.
2022-03-14 05:31:23 +00:00
#include "QF/GL/qf_particles.h"
Whitespace.
2001-08-26 17:13:33 +00:00
#include "QF/GL/qf_rlight.h"
make gcc more anal about prototypes, string constants and function visibility (ie, global functions must have a prototype)
2003-01-06 18:28:13 +00:00
#include "QF/GL/qf_rmain.h"
Whitespace.
2001-08-26 17:13:33 +00:00
#include "QF/GL/qf_rsurf.h"
[renderer] Remove namehack.h Finally. I never liked it (felt bad adding it in the first place), and it has caused confusion with function and global variable names, but it did let me get the render plugins working.
2022-03-14 05:31:23 +00:00
#include "QF/GL/qf_sprite.h"
This is changing more then I wanted to at once, however. The major change is that we no longer require libGL to even exist on the system at compile time for the GL targets, we dynamicly link to the libGL of choice at run time. (This probably breaks most non-linux systems, and all GL targets except -glx, some fixup will be needed.) (This also kills glquake, dead dead DEAD! GONE FOREVER! WHEE!) Some gl_draw cleanup. Commented out equake alias model occlusion test stuff, very experimental. Added the .lo and .la patterns to the .gitignore files. Some minor sbar cleanup. (We don't use the disc in use symbol for anything.)
2001-06-24 09:25:55 +00:00
#include "QF/GL/qf_vid.h"
(Optionally) Use glTexSubImage2D for lightmap updates. Doesn't give as large a boost as I'd hoped, but will matter far more as other optimizations happen. New cvars gl_lightmap_align does nothing yet, and gl_lightmap_subimage only works in mode 0 and 1. Mode 2 (should be fastest mode on most cards) will eventully get implemented...
2001-08-02 02:18:04 +00:00
#include "compat.h"
Take the first step towards render plugins. No clients link. Even if they did, nothing would work.
2012-02-14 08:28:09 +00:00
#include "r_internal.h"
Make sprite rendering use vertex arrays, as a test for detection of vertex array support (tests only for OpenGL 1.1 or better, at the moment, no driver blacklist yet).
2003-03-20 19:58:18 +00:00
#include "varrays.h"
[renderer] Merge the fog support code GL still has its own functions for enabling and disabling fog while rendering, but GLSL doesn't need such (thanks to the shaders), nor will vulkan (and the software renderers don't support fog).
2022-03-07 17:10:47 +00:00
#include "vid_gl.h"
initial checkin of most recent newtree and nuq(?) source
2001-02-19 21:15:25 +00:00
Rename the public symbols for gl, glsl and sw32. Evil hack, but it does the job.
2012-02-22 07:32:34 +00:00
float gl_r_world_matrix[16];
[renderer] Remove more old fields from entity_t The only transform related field remaining is old_origin. This also brings the renderer closer to using simd (lots of stuff to fix still, though).
2021-03-19 11:18:45 +00:00
//FIXME static float r_base_world_matrix[16];
run indent over all the .c files using qw's .indent.pro. The real reason for this is I mistakenly did so while making some other changes (which I made sure were NOT in the checkin:)
2001-02-26 06:48:02 +00:00
Clean up global symbols for the gl renderer. Where possible, symbols have been made static, prefixed with gl_/GL_ or moved into the code shared by all renderers. This will make doing plugins easier but done now for link testing.
2012-02-17 09:33:07 +00:00
//vec3_t gl_shadecolor; // Ender (Extend) Colormod
float gl_modelalpha; // Ender (Extend) Alpha
run indent over all the .c files using qw's .indent.pro. The real reason for this is I mistakenly did so while making some other changes (which I made sure were NOT in the checkin:)
2001-02-26 06:48:02 +00:00
void
95% merge gl_rmain.c nq is actually generally working again, except alias models are black
2001-05-14 06:47:38 +00:00
glrmain_init (void)
initial checkin of most recent newtree and nuq(?) source
2001-02-19 21:15:25 +00:00
{
Clean up glDepthFunc and glDepthRange so it's properly initted, and not set needlessly every frame.
2001-10-02 00:01:05 +00:00
gldepthmin = 0;
gldepthmax = 1;
qfglDepthFunc (GL_LEQUAL);
qfglDepthRange (gldepthmin, gldepthmax);
[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.
2022-04-23 03:22:45 +00:00
gl_overbright_f (0, 0);
gl_multitexture_f (0, 0);
95% merge gl_rmain.c nq is actually generally working again, except alias models are black
2001-05-14 06:47:38 +00:00
}
initial checkin of most recent newtree and nuq(?) source
2001-02-19 21:15:25 +00:00
patch from phrosty for vc2005 support. includes fixes for a bunch of gcc-isms that crept in over the years.
2007-03-22 23:20:57 +00:00
void
Rename the public symbols for gl, glsl and sw32. Evil hack, but it does the job.
2012-02-22 07:32:34 +00:00
gl_R_RotateForEntity (entity_t *e)
initial checkin of most recent newtree and nuq(?) source
2001-02-19 21:15:25 +00:00
{
[renderer] Clean up entity_t to a certain extent This is the first step towards component-based entities. There's still some transform-related stuff in the struct that needs to be moved, but it's all entirely client related (rather than renderer) and will probably go into a "client" component. Also, the current components are directly included structs rather than references as I didn't want to deal with the object management at this stage. As part of the process (because transforms use simd) this also starts the process of moving QF to using simd for vectors and matrices. There's now a mess of simd and sisd code mixed together, but it works surprisingly well together.
2021-03-09 14:52:40 +00:00
mat4f_t mat;
Transform_GetWorldMatrix (e->transform, mat);
Update vec3_t/vec4f_t hacks to work with clang Still work with gcc, of course, and I still need to fix them properly, but now they're actually slightly easier to find as they all have vec_t and FIXME on the same line.
2022-03-30 15:07:20 +00:00
qfglMultMatrixf ((vec_t*)&mat[0]);//FIXME
initial checkin of most recent newtree and nuq(?) source
2001-02-19 21:15:25 +00:00
}
Clean up glError handling.
2003-05-08 22:16:30 +00:00
[renderer] Move most of the scene rendering into r_screen r_screen isn't really the right place, but it gets the scene rendering out of the low-level renderers and will make it easier to sort out later, and hopefully easier to figure out a good design for vulkan.
2022-03-17 08:50:38 +00:00
void
gl_R_RenderEntities (entqueue_t *queue)
initial checkin of most recent newtree and nuq(?) source
2001-02-19 21:15:25 +00:00
{
[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.
2022-04-23 03:22:45 +00:00
if (!r_drawentities)
initial checkin of most recent newtree and nuq(?) source
2001-02-19 21:15:25 +00:00
return;
// LordHavoc: split into 3 loops to simplify state changes
Lots of optimizations, particularly for GL state changes, and tweak lighting a bit more. Also a cleanup from Grievre: Use VectorIsZero in two places in gl_mod_alias.c.
2004-03-15 21:32:19 +00:00
if (gl_mtex_active_tmus >= 2) {
qglActiveTexture (gl_mtex_enum + 1);
qfglEnable (GL_TEXTURE_2D);
qfglTexEnvf (GL_TEXTURE_ENV, GL_TEXTURE_ENV_MODE, GL_DECAL);
qfglDisable (GL_TEXTURE_2D);
qglActiveTexture (gl_mtex_enum + 0);
}
[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.
2022-04-23 03:22:45 +00:00
if (gl_affinemodels)
Fix failure to free cache for non-shadowed stuff (lightning, grenades, etc). Also a minor speedup (only mess with glHints once per frame, instead of dozens of times).
2001-12-19 18:32:26 +00:00
qfglHint (GL_PERSPECTIVE_CORRECTION_HINT, GL_NICEST);
Do a final cleanup before doing the big rename.
2012-02-22 02:09:09 +00:00
if (gl_tess)
TruForm support. Controlled by gl_tessellate
2004-01-26 00:13:47 +00:00
qfglEnable (GL_PN_TRIANGLES_ATI);
Grievre's lighting patch, plus some of my own changes. My changes: Emit normals if truform is enabled. Attempt to avoid 0,0,0 normals (makes lighting & tesselation unhappy). Fix some (ancient) apparent bugs in GetAliasFrameVerts16(). Clamp minlight, instead of adding it. Apply colormod as glColor, rather than adding it to emission, to prevent QSG2 issues with fullbrights. Rearrange init code, and don't go quite as wild with responsiveness to lights. Material & lightmode settings will need tweaking & testing to work well on all cards. Feedback needed there.
2004-03-10 08:14:38 +00:00
qfglEnable (GL_CULL_FACE);
Massive whitespace cleanup. Lots of trailing whitespace and otherwise blank lines.
2012-05-21 23:23:22 +00:00
[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.
2022-04-23 03:22:45 +00:00
if (gl_vector_light) {
the rest of Grievre's lighting tweaks
2004-05-02 21:21:00 +00:00
qfglEnable (GL_LIGHTING);
qfglEnable (GL_NORMALIZE);
Do a final cleanup before doing the big rename.
2012-02-22 02:09:09 +00:00
} else if (gl_tess) {
the rest of Grievre's lighting tweaks
2004-05-02 21:21:00 +00:00
qfglEnable (GL_NORMALIZE);
}
Render iqm models in GL. There are still many issues to sort out, but the basics are working. Problems: rendered fullbright (no lighting done) normals are ignored extra textures (glow etc) not used/loaded 4 models on the screen don't seem to be a problem.
2012-05-17 06:58:29 +00:00
[renderer] Move most of the scene rendering into r_screen r_screen isn't really the right place, but it gets the scene rendering out of the low-level renderers and will make it easier to sort out later, and hopefully easier to figure out a good design for vulkan.
2022-03-17 08:50:38 +00:00
for (size_t i = 0; i < queue->ent_queues[mod_alias].size; i++) { \
entity_t *ent = queue->ent_queues[mod_alias].a[i]; \
[renderer] Remove currententity from non-sw renderers Really needed only for vulkan, but removing currententity from gl and glsl made testing easier.
2021-07-22 06:39:28 +00:00
gl_R_DrawAliasModel (ent);
initial checkin of most recent newtree and nuq(?) source
2001-02-19 21:15:25 +00:00
}
Fix a bunch of state leaks that made multitextured fullbrights smear the world. Now just have one (known) issue with bmodels with fullbright textures not rendering their base texture. Floating fullbrights are nifty! Also a number of optimizations, which need wider performance testing.
2004-02-21 05:36:19 +00:00
qfglColor3ubv (color_white);
Massive whitespace cleanup. Lots of trailing whitespace and otherwise blank lines.
2012-05-21 23:23:22 +00:00
Lots of optimizations, particularly for GL state changes, and tweak lighting a bit more. Also a cleanup from Grievre: Use VectorIsZero in two places in gl_mod_alias.c.
2004-03-15 21:32:19 +00:00
qfglDisable (GL_NORMALIZE);
qfglDisable (GL_LIGHTING);
the rest of Grievre's lighting tweaks
2004-05-02 21:21:00 +00:00
Do a final cleanup before doing the big rename.
2012-02-22 02:09:09 +00:00
if (gl_tess)
TruForm support. Controlled by gl_tessellate
2004-01-26 00:13:47 +00:00
qfglDisable (GL_PN_TRIANGLES_ATI);
[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.
2022-04-23 03:22:45 +00:00
if (gl_affinemodels)
Fix failure to free cache for non-shadowed stuff (lightning, grenades, etc). Also a minor speedup (only mess with glHints once per frame, instead of dozens of times).
2001-12-19 18:32:26 +00:00
qfglHint (GL_PERSPECTIVE_CORRECTION_HINT, GL_DONT_CARE);
Optimize fullbright testing for brushes, assorted other cleanups, test for functions built into current GL drivers, as well as the old extensions, and add incorrectly rendering BGRA support. Whee!
2004-02-24 20:50:55 +00:00
if (gl_mtex_active_tmus >= 2) { // FIXME: Ugly, but faster than cleaning
// up in every R_DrawAliasModel()!
Fix a bunch of state leaks that made multitextured fullbrights smear the world. Now just have one (known) issue with bmodels with fullbright textures not rendering their base texture. Floating fullbrights are nifty! Also a number of optimizations, which need wider performance testing.
2004-02-21 05:36:19 +00:00
qglActiveTexture (gl_mtex_enum + 1);
qfglEnable (GL_TEXTURE_2D);
[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.
2022-04-23 03:22:45 +00:00
if (gl_combine_capable && gl_overbright) {
Remove gl_doublebright, remove fractional overbrights, clean up init a bit, uglify init a bit. Stuff for taniwha to look at. Nothing to see here, move along.
2004-03-07 23:51:29 +00:00
qfglTexEnvf (GL_TEXTURE_ENV, GL_TEXTURE_ENV_MODE, GL_COMBINE);
qfglTexEnvf (GL_TEXTURE_ENV, GL_COMBINE_RGB, GL_MODULATE);
Do a final cleanup before doing the big rename.
2012-02-22 02:09:09 +00:00
qfglTexEnvf (GL_TEXTURE_ENV, GL_RGB_SCALE, gl_rgb_scale);
Fix a bunch of state leaks that made multitextured fullbrights smear the world. Now just have one (known) issue with bmodels with fullbright textures not rendering their base texture. Floating fullbrights are nifty! Also a number of optimizations, which need wider performance testing.
2004-02-21 05:36:19 +00:00
} else {
qfglTexEnvf (GL_TEXTURE_ENV, GL_TEXTURE_ENV_MODE, GL_MODULATE);
}
qfglDisable (GL_TEXTURE_2D);
qglActiveTexture (gl_mtex_enum + 0);
}
initial checkin of most recent newtree and nuq(?) source
2001-02-19 21:15:25 +00:00
[renderer] Move most of the scene rendering into r_screen r_screen isn't really the right place, but it gets the scene rendering out of the low-level renderers and will make it easier to sort out later, and hopefully easier to figure out a good design for vulkan.
2022-03-17 08:50:38 +00:00
for (size_t i = 0; i < queue->ent_queues[mod_iqm].size; i++) { \
entity_t *ent = queue->ent_queues[mod_iqm].a[i]; \
[renderer] Remove currententity from non-sw renderers Really needed only for vulkan, but removing currententity from gl and glsl made testing easier.
2021-07-22 06:39:28 +00:00
gl_R_DrawIQMModel (ent);
Render iqm models in GL. There are still many issues to sort out, but the basics are working. Problems: rendered fullbright (no lighting done) normals are ignored extra textures (glow etc) not used/loaded 4 models on the screen don't seem to be a problem.
2012-05-17 06:58:29 +00:00
}
qfglColor3ubv (color_white);
qfglDisable (GL_CULL_FACE);
Instead of glEnable/glDisable (GL_ALPHA_TEST) pairs for every sprite rendered, let's do it once for the entire sprite loop. Minor hit when there are no sprites, but 2+ it's a gain. Better worst case, rather than better peak performance.
2001-10-02 01:48:51 +00:00
qfglEnable (GL_ALPHA_TEST);
Make sprite rendering use vertex arrays, as a test for detection of vertex array support (tests only for OpenGL 1.1 or better, at the moment, no driver blacklist yet).
2003-03-20 19:58:18 +00:00
if (gl_va_capable)
Clean up global symbols for the gl renderer. Where possible, symbols have been made static, prefixed with gl_/GL_ or moved into the code shared by all renderers. This will make doing plugins easier but done now for link testing.
2012-02-17 09:33:07 +00:00
qfglInterleavedArrays (GL_T2F_C4UB_V3F, 0, gl_spriteVertexArray);
[renderer] Move most of the scene rendering into r_screen r_screen isn't really the right place, but it gets the scene rendering out of the low-level renderers and will make it easier to sort out later, and hopefully easier to figure out a good design for vulkan.
2022-03-17 08:50:38 +00:00
for (size_t i = 0; i < queue->ent_queues[mod_sprite].size; i++) { \
entity_t *ent = queue->ent_queues[mod_sprite].a[i]; \
[renderer] Remove namehack.h Finally. I never liked it (felt bad adding it in the first place), and it has caused confusion with function and global variable names, but it did let me get the render plugins working.
2022-03-14 05:31:23 +00:00
gl_R_DrawSpriteModel (ent);
initial checkin of most recent newtree and nuq(?) source
2001-02-19 21:15:25 +00:00
}
Instead of glEnable/glDisable (GL_ALPHA_TEST) pairs for every sprite rendered, let's do it once for the entire sprite loop. Minor hit when there are no sprites, but 2+ it's a gain. Better worst case, rather than better peak performance.
2001-10-02 01:48:51 +00:00
qfglDisable (GL_ALPHA_TEST);
initial checkin of most recent newtree and nuq(?) source
2001-02-19 21:15:25 +00:00
}
run indent over all the .c files using qw's .indent.pro. The real reason for this is I mistakenly did so while making some other changes (which I made sure were NOT in the checkin:)
2001-02-26 06:48:02 +00:00
static void
R_DrawViewModel (void)
initial checkin of most recent newtree and nuq(?) source
2001-02-19 21:15:25 +00:00
{
[renderer] Remove currententity from non-sw renderers Really needed only for vulkan, but removing currententity from gl and glsl made testing easier.
2021-07-22 06:39:28 +00:00
entity_t *ent = vr_data.view_model;
Take the first step towards render plugins. No clients link. Even if they did, nothing would work.
2012-02-14 08:28:09 +00:00
if (vr_data.inhibit_viewmodel
[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.
2022-04-23 03:22:45 +00:00
|| !r_drawviewmodel
|| !r_drawentities
[renderer] Remove currententity from non-sw renderers Really needed only for vulkan, but removing currententity from gl and glsl made testing easier.
2021-07-22 06:39:28 +00:00
|| !ent->renderer.model)
initial checkin of most recent newtree and nuq(?) source
2001-02-19 21:15:25 +00:00
return;
* taniwha mutters about end-runs (sorry, Grievre)
2004-04-28 04:07:50 +00:00
// hack the depth range to prevent view model from poking into walls
qfglDepthRange (gldepthmin, gldepthmin + 0.3 * (gldepthmax - gldepthmin));
qfglEnable (GL_CULL_FACE);
the rest of Grievre's lighting tweaks
2004-05-02 21:21:00 +00:00
[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.
2022-04-23 03:22:45 +00:00
if (gl_vector_light) {
the rest of Grievre's lighting tweaks
2004-05-02 21:21:00 +00:00
qfglEnable (GL_LIGHTING);
qfglEnable (GL_NORMALIZE);
Do a final cleanup before doing the big rename.
2012-02-22 02:09:09 +00:00
} else if (gl_tess) {
the rest of Grievre's lighting tweaks
2004-05-02 21:21:00 +00:00
qfglEnable (GL_NORMALIZE);
}
[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.
2022-04-23 03:22:45 +00:00
if (gl_affinemodels)
Might as well allow gl_affinemodels to affect the viewmodel as well...
2001-12-19 20:15:38 +00:00
qfglHint (GL_PERSPECTIVE_CORRECTION_HINT, GL_NICEST);
Lots of optimizations, particularly for GL state changes, and tweak lighting a bit more. Also a cleanup from Grievre: Use VectorIsZero in two places in gl_mod_alias.c.
2004-03-15 21:32:19 +00:00
if (gl_mtex_active_tmus >= 2) {
qglActiveTexture (gl_mtex_enum + 1);
qfglEnable (GL_TEXTURE_2D);
qfglTexEnvf (GL_TEXTURE_ENV, GL_TEXTURE_ENV_MODE, GL_DECAL);
qfglDisable (GL_TEXTURE_2D);
qglActiveTexture (gl_mtex_enum + 0);
}
Fix a bunch of state leaks that made multitextured fullbrights smear the world. Now just have one (known) issue with bmodels with fullbright textures not rendering their base texture. Floating fullbrights are nifty! Also a number of optimizations, which need wider performance testing.
2004-02-21 05:36:19 +00:00
[renderer] Remove currententity from non-sw renderers Really needed only for vulkan, but removing currententity from gl and glsl made testing easier.
2021-07-22 06:39:28 +00:00
gl_R_DrawAliasModel (ent);
Lots of optimizations, particularly for GL state changes, and tweak lighting a bit more. Also a cleanup from Grievre: Use VectorIsZero in two places in gl_mod_alias.c.
2004-03-15 21:32:19 +00:00
Fix a bunch of state leaks that made multitextured fullbrights smear the world. Now just have one (known) issue with bmodels with fullbright textures not rendering their base texture. Floating fullbrights are nifty! Also a number of optimizations, which need wider performance testing.
2004-02-21 05:36:19 +00:00
qfglColor3ubv (color_white);
Multitexture state fix. Harmless so far, but testing carefully helps us avoid state leak horrors.
2004-02-27 21:42:52 +00:00
if (gl_mtex_active_tmus >= 2) { // FIXME: Ugly, but faster than cleaning
// up in every R_DrawAliasModel()!
Fix a bunch of state leaks that made multitextured fullbrights smear the world. Now just have one (known) issue with bmodels with fullbright textures not rendering their base texture. Floating fullbrights are nifty! Also a number of optimizations, which need wider performance testing.
2004-02-21 05:36:19 +00:00
qglActiveTexture (gl_mtex_enum + 1);
qfglEnable (GL_TEXTURE_2D);
[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.
2022-04-23 03:22:45 +00:00
if (gl_combine_capable && gl_overbright) {
Remove gl_doublebright, remove fractional overbrights, clean up init a bit, uglify init a bit. Stuff for taniwha to look at. Nothing to see here, move along.
2004-03-07 23:51:29 +00:00
qfglTexEnvf (GL_TEXTURE_ENV, GL_TEXTURE_ENV_MODE, GL_COMBINE);
qfglTexEnvf (GL_TEXTURE_ENV, GL_COMBINE_RGB, GL_MODULATE);
Do a final cleanup before doing the big rename.
2012-02-22 02:09:09 +00:00
qfglTexEnvf (GL_TEXTURE_ENV, GL_RGB_SCALE, gl_rgb_scale);
Fix a bunch of state leaks that made multitextured fullbrights smear the world. Now just have one (known) issue with bmodels with fullbright textures not rendering their base texture. Floating fullbrights are nifty! Also a number of optimizations, which need wider performance testing.
2004-02-21 05:36:19 +00:00
} else {
qfglTexEnvf (GL_TEXTURE_ENV, GL_TEXTURE_ENV_MODE, GL_MODULATE);
}
qfglDisable (GL_TEXTURE_2D);
qglActiveTexture (gl_mtex_enum + 0);
}
[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.
2022-04-23 03:22:45 +00:00
if (gl_affinemodels)
Might as well allow gl_affinemodels to affect the viewmodel as well...
2001-12-19 20:15:38 +00:00
qfglHint (GL_PERSPECTIVE_CORRECTION_HINT, GL_DONT_CARE);
Massive whitespace cleanup. Lots of trailing whitespace and otherwise blank lines.
2012-05-21 23:23:22 +00:00
Lots of optimizations, particularly for GL state changes, and tweak lighting a bit more. Also a cleanup from Grievre: Use VectorIsZero in two places in gl_mod_alias.c.
2004-03-15 21:32:19 +00:00
qfglDisable (GL_NORMALIZE);
qfglDisable (GL_LIGHTING);
the rest of Grievre's lighting tweaks
2004-05-02 21:21:00 +00:00
Massive whitespace cleanup. Lots of trailing whitespace and otherwise blank lines.
2012-05-21 23:23:22 +00:00
Grievre's lighting patch, plus some of my own changes. My changes: Emit normals if truform is enabled. Attempt to avoid 0,0,0 normals (makes lighting & tesselation unhappy). Fix some (ancient) apparent bugs in GetAliasFrameVerts16(). Clamp minlight, instead of adding it. Apply colormod as glColor, rather than adding it to emission, to prevent QSG2 issues with fullbrights. Rearrange init code, and don't go quite as wild with responsiveness to lights. Material & lightmode settings will need tweaking & testing to work well on all cards. Feedback needed there.
2004-03-10 08:14:38 +00:00
qfglDisable (GL_CULL_FACE);
Clean up and repair of the run time linking, does NOT work until deek commits his gl_funcs.c.
2001-06-26 02:26:46 +00:00
qfglDepthRange (gldepthmin, gldepthmax);
initial checkin of most recent newtree and nuq(?) source
2001-02-19 21:15:25 +00:00
}
sw and gl fisheye patches from Arkadi Shishlov (arkadi@it.lv). I don't think /I/ can play that way (dizzy:)
2003-01-15 16:53:34 +00:00
static void
R_SetupGL (void)
{
[renderer] Clean up FOV and viewport handling Viewport and FOV updates are now separate so updating one doesn't cause recalculations of the other. Also, perspective setup is now done directly from the tangents of the half angles for fov_x and fov_y making the renderers independent of fov/aspect mode. I imagine things are a bit of a mess with view size changes, and especially screen size changes (not supported yet anyway), and vulkan winds up updating its projection matrices every frame, but everything that's expected to work does (vulkan errors out for fisheye or warp due to frame buffer creation not being supported yet).
2022-03-30 05:55:32 +00:00
qfglMatrixMode (GL_PROJECTION);
Update vec3_t/vec4f_t hacks to work with clang Still work with gcc, of course, and I still need to fix them properly, but now they're actually slightly easier to find as they all have vec_t and FIXME on the same line.
2022-03-30 15:07:20 +00:00
qfglLoadMatrixf ((vec_t*)&gl_ctx->projection[0]);//FIXME
run indent over all the .c files using qw's .indent.pro. The real reason for this is I mistakenly did so while making some other changes (which I made sure were NOT in the checkin:)
2001-02-26 06:48:02 +00:00
[gl] Clean out some missed mirror code
2022-03-14 15:09:37 +00:00
qfglFrontFace (GL_CW);
initial checkin of most recent newtree and nuq(?) source
2001-02-19 21:15:25 +00:00
Clean up and repair of the run time linking, does NOT work until deek commits his gl_funcs.c.
2001-06-26 02:26:46 +00:00
qfglMatrixMode (GL_MODELVIEW);
qfglLoadIdentity ();
initial checkin of most recent newtree and nuq(?) source
2001-02-19 21:15:25 +00:00
[renderer] Remove more old fields from entity_t The only transform related field remaining is old_origin. This also brings the renderer closer to using simd (lots of stuff to fix still, though).
2021-03-19 11:18:45 +00:00
static mat4f_t z_up = {
[gl] Make perspective matrix setup consistent Now GL perspective matrix setup matches that of GLSL and Vulkan, and GL's z_up matrix matches GLSL's (as it should, since they're really going through the same API). GL also needs the depth adjustmet matrix now. Other than having to google the docs for glFrustum, there's nothing wrong with the function itself, but it's nice to have direct control over the matrices. In the process, I discovered how horribly confused I've been at times with respect to the handedness of GL and Quake: GL is right-handed (y-up, z-out, x-right), as is Quake itself (but z-up, y-left, x-in), but as the perspective matrix used in the three renderers expects z-in, having x-right and y-up makes the matrix effectively left-handed (not for Vulkan though, because there it's y-down, x-right, z-up, so right-handed again).
2022-03-27 04:15:15 +00:00
{ 0, 0, 1, 0},
[renderer] Remove more old fields from entity_t The only transform related field remaining is old_origin. This also brings the renderer closer to using simd (lots of stuff to fix still, though).
2021-03-19 11:18:45 +00:00
{-1, 0, 0, 0},
{ 0, 1, 0, 0},
{ 0, 0, 0, 1},
};
mat4f_t view;
[renderer] Clean up use of vup/vright/vpn This moves the common camera setup code out of the individual drivers, and completely removes vup/vright/vpn from the non-software renderers. This has highlighted the craziness around AngleVectors with it putting +X forward, -Y right and +Z up. The main issue with this is it requires a 90 degree pre-rotation about the Z axis to get the camera pointing in the right direction, and that's for the native sw renderer (vulkan needs a 90 degree pre-rotation about X, and gl and glsl need to invert an axis, too), though at least it's just a matrix swizzle and vector negation. However, it does mean the camera matrices can't be used directly. Also rename vpn to vfwd (still abbreviated, but fwd is much clearer in meaning (to me, at least) than pn (plane normal, I guess, but which way?)).
2022-03-14 00:34:24 +00:00
mmulf (view, z_up, r_refdef.camera_inverse);
Update vec3_t/vec4f_t hacks to work with clang Still work with gcc, of course, and I still need to fix them properly, but now they're actually slightly easier to find as they all have vec_t and FIXME on the same line.
2022-03-30 15:07:20 +00:00
qfglLoadMatrixf ((vec_t*)&view[0]);//FIXME
initial checkin of most recent newtree and nuq(?) source
2001-02-19 21:15:25 +00:00
Rename the public symbols for gl, glsl and sw32. Evil hack, but it does the job.
2012-02-22 07:32:34 +00:00
qfglGetFloatv (GL_MODELVIEW_MATRIX, gl_r_world_matrix);
initial checkin of most recent newtree and nuq(?) source
2001-02-19 21:15:25 +00:00
// set drawing parms
* taniwha mutters about end-runs (sorry, Grievre)
2004-04-28 04:07:50 +00:00
// qfglEnable (GL_CULL_FACE);
Clean up and repair of the run time linking, does NOT work until deek commits his gl_funcs.c.
2001-06-26 02:26:46 +00:00
qfglDisable (GL_ALPHA_TEST);
qfglAlphaFunc (GL_GREATER, 0.5);
qfglEnable (GL_DEPTH_TEST);
[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.
2022-04-23 03:22:45 +00:00
if (gl_dlight_smooth)
Clean up and repair of the run time linking, does NOT work until deek commits his gl_funcs.c.
2001-06-26 02:26:46 +00:00
qfglShadeModel (GL_SMOOTH);
95% merge gl_rmain.c nq is actually generally working again, except alias models are black
2001-05-14 06:47:38 +00:00
else
Clean up and repair of the run time linking, does NOT work until deek commits his gl_funcs.c.
2001-06-26 02:26:46 +00:00
qfglShadeModel (GL_FLAT);
initial checkin of most recent newtree and nuq(?) source
2001-02-19 21:15:25 +00:00
}
[gl] Get fisheye working with frame buffers Of course, it's not as correct as glsl or sw due to using polygons and uvs rather than a fragment shader (not that such is out of the question since GL 3.0 is requested, but I don't feel like getting shaders going just for a couple of post-processing effects in an obsolete renderer).
2022-03-26 03:08:06 +00:00
void
gl_R_RenderView (void)
initial checkin of most recent newtree and nuq(?) source
2001-02-19 21:15:25 +00:00
{
[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.
2022-04-23 03:22:45 +00:00
if (r_norefresh) {
initial checkin of most recent newtree and nuq(?) source
2001-02-19 21:15:25 +00:00
return;
[renderer] Clean up entity_t to a certain extent This is the first step towards component-based entities. There's still some transform-related stuff in the struct that needs to be moved, but it's all entirely client related (rather than renderer) and will probably go into a "client" component. Also, the current components are directly included structs rather than references as I didn't want to deal with the object management at this stage. As part of the process (because transforms use simd) this also starts the process of moving QF to using simd for vectors and matrices. There's now a mess of simd and sisd code mixed together, but it works surprisingly well together.
2021-03-09 14:52:40 +00:00
}
[renderer] move r_worldentity and r_viewleaf into refdef More cleanup of globals that seem to be quake specific.
2022-03-14 06:27:43 +00:00
if (!r_refdef.worldmodel) {
[renderer] Make R_RenderView private This is actually a better solution to the renderer directly accessing client code than provided by 7e078c7f9c8ab27af92ff268f49bac070c9ff617. Essentially, V_RenderView should not have been calling R_RenderView, and CL_UpdateScreen should have been calling V_RenderView directly. The issue was that the renderers expected the world entity model to be valid at all times. Now, R_RenderView checks the world entity model's validity and immediately bails if it is not, and R_ClearState (which is called whenever the client disconnects and thus no longer has a world to render) clears the world entity model. Thus R_RenderView can (and is) now called unconditionally from within the renderer, simplifying renderer-specific variants.
2021-12-02 09:30:57 +00:00
return;
[renderer] Clean up entity_t to a certain extent This is the first step towards component-based entities. There's still some transform-related stuff in the struct that needs to be moved, but it's all entirely client related (rather than renderer) and will probably go into a "client" component. Also, the current components are directly included structs rather than references as I didn't want to deal with the object management at this stage. As part of the process (because transforms use simd) this also starts the process of moving QF to using simd for vectors and matrices. There's now a mess of simd and sisd code mixed together, but it works surprisingly well together.
2021-03-09 14:52:40 +00:00
}
initial checkin of most recent newtree and nuq(?) source
2001-02-19 21:15:25 +00:00
[gl] Get fisheye working with frame buffers Of course, it's not as correct as glsl or sw due to using polygons and uvs rather than a fragment shader (not that such is out of the question since GL 3.0 is requested, but I don't feel like getting shaders going just for a couple of post-processing effects in an obsolete renderer).
2022-03-26 03:08:06 +00:00
R_SetupGL ();
gl_Fog_EnableGFog ();
sw and gl fisheye patches from Arkadi Shishlov (arkadi@it.lv). I don't think /I/ can play that way (dizzy:)
2003-01-15 16:53:34 +00:00
[gl] Get fisheye working with frame buffers Of course, it's not as correct as glsl or sw due to using polygons and uvs rather than a fragment shader (not that such is out of the question since GL 3.0 is requested, but I don't feel like getting shaders going just for a couple of post-processing effects in an obsolete renderer).
2022-03-26 03:08:06 +00:00
gl_R_DrawWorld ();
S_ExtraUpdate (); // don't let sound get messed up if going slow
gl_R_RenderDlights ();
[renderer] Don't try to draw a nonexistent view model Doing so doesn't end well (segfault). This is a bit of a hack until I come up with a design for configurable scene rendering.
2022-04-25 22:22:45 +00:00
if (vr_data.view_model) {
R_DrawViewModel ();
}
sw and gl fisheye patches from Arkadi Shishlov (arkadi@it.lv). I don't think /I/ can play that way (dizzy:)
2003-01-15 16:53:34 +00:00
[gl] Get fisheye working with frame buffers Of course, it's not as correct as glsl or sw due to using polygons and uvs rather than a fragment shader (not that such is out of the question since GL 3.0 is requested, but I don't feel like getting shaders going just for a couple of post-processing effects in an obsolete renderer).
2022-03-26 03:08:06 +00:00
gl_Fog_DisableGFog ();
[renderer] Get fisheye working again Only for gl and sw at the moment (want to merge things further before I do anything for glsl or vulkan). However, with with I've learned getting gl and sw to work, glsl and vulkan will be trivial.
2022-03-14 14:51:30 +00:00
}
Clean up the mess around D_FlushCaches. Create an R_ClearState and call that from the clients. It calls any necessary functions itself.
2011-12-23 03:51:47 +00:00
Remove VISIBLE from the files that will go into the plugins.
2012-02-18 13:28:42 +00:00
void
Rename the public symbols for gl, glsl and sw32. Evil hack, but it does the job.
2012-02-22 07:32:34 +00:00
gl_R_ClearState (void)
Clean up the mess around D_FlushCaches. Create an R_ClearState and call that from the clients. It calls any necessary functions itself.
2011-12-23 03:51:47 +00:00
{
[renderer] move r_worldentity and r_viewleaf into refdef More cleanup of globals that seem to be quake specific.
2022-03-14 06:27:43 +00:00
r_refdef.worldmodel = 0;
Clean up the mess around D_FlushCaches. Create an R_ClearState and call that from the clients. It calls any necessary functions itself.
2011-12-23 03:51:47 +00:00
R_ClearEfrags ();
R_ClearDlights ();
[particles] Merge particle spawning into client After yesterday's crazy marathon editing all the particles files, and starting to do another big change to them today, I realized that I really do need to merge them down. All the actual spawning is now in the client library (though particle insertion will need to be moved). GLSL particle rendering is semi-broken in that it now does only points (until I come up with a way to select between points and quads (probably a context object, which I need anyway for Vulkan)).
2021-12-19 04:08:39 +00:00
R_ClearParticles ();
Clean up the mess around D_FlushCaches. Create an R_ClearState and call that from the clients. It calls any necessary functions itself.
2011-12-23 03:51:47 +00:00
}
Reference in a new issue Copy permalink