qf/quakeforge
0
0
Fork 0
mirror of https://git.code.sf.net/p/quake/quakeforge synced 2024-12-11 05:11:16 +00:00
Code Issues Projects Releases Packages Wiki Activity
quakeforge/libs/video/renderer/vulkan/vulkan_main.c

254 lines
6 KiB
C
Raw Normal View History

[vulkan] Make a lot of progress for brush models Light maps are maybe updating, but as nothing is actually rendered yet, it's hard to tell.
2021-01-19 16:25:54 +00:00
/*
vulkan_main.c
Vulkan rendering
Copyright (C) 2021 Bill Currie <bill@taniwha.org>
Author: Bill Currie <bill@taniwha.org>
Date: 2021/1/19
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 "QF/cmd.h"
#include "QF/cvar.h"
#include "QF/image.h"
#include "QF/render.h"
#include "QF/screen.h"
#include "QF/sys.h"
#include "QF/Vulkan/qf_vid.h"
[vulkan] Start work on alias model rendering
2021-01-26 11:58:24 +00:00
#include "QF/Vulkan/qf_alias.h"
[vulkan] Make a lot of progress for brush models Light maps are maybe updating, but as nothing is actually rendered yet, it's hard to tell.
2021-01-19 16:25:54 +00:00
#include "QF/Vulkan/qf_bsp.h"
//#include "QF/Vulkan/qf_iqm.h"
[vulkan] Load the map's lights into an array It's not used yet as work needs to be done to better support generic entities, but this is the next step to real-time lighting (though, to be honest, I expect it will be too slow to be usable).
2021-03-03 09:14:16 +00:00
#include "QF/Vulkan/qf_lighting.h"
[vulkan] Make a lot of progress for brush models Light maps are maybe updating, but as nothing is actually rendered yet, it's hard to tell.
2021-01-19 16:25:54 +00:00
#include "QF/Vulkan/qf_lightmap.h"
#include "QF/Vulkan/qf_main.h"
#include "QF/Vulkan/qf_particles.h"
//#include "QF/Vulkan/qf_textures.h"
[vulkan] Support multiple render passes Multiple render passes are needed for supporting shadow mapping, and this is a huge step towards breaking the Vulkan render free of Quake, and hopefully will lead the way for breaking the GL renderers free as well.
2021-12-02 12:58:29 +00:00
#include "QF/Vulkan/renderpass.h"
[vulkan] Make a lot of progress for brush models Light maps are maybe updating, but as nothing is actually rendered yet, it's hard to tell.
2021-01-19 16:25:54 +00:00
#include "mod_internal.h"
#include "r_internal.h"
#include "vid_vulkan.h"
static void
setup_frame (vulkan_ctx_t *ctx)
{
R_AnimateLight ();
R_ClearEnts ();
r_framecount++;
[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
VectorCopy (r_refdef.viewposition, r_origin);
VectorCopy (qvmulf (r_refdef.viewrotation, (vec4f_t) { 1, 0, 0, 0 }), vpn);
VectorCopy (qvmulf (r_refdef.viewrotation, (vec4f_t) { 0, -1, 0, 0 }), vright);
VectorCopy (qvmulf (r_refdef.viewrotation, (vec4f_t) { 0, 0, 1, 0 }), vup);
[vulkan] Make a lot of progress for brush models Light maps are maybe updating, but as nothing is actually rendered yet, it's hard to tell.
2021-01-19 16:25:54 +00:00
R_SetFrustum ();
[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
r_viewleaf = Mod_PointInLeaf (r_origin, r_worldentity.renderer.model);
[vulkan] Make a lot of progress for brush models Light maps are maybe updating, but as nothing is actually rendered yet, it's hard to tell.
2021-01-19 16:25:54 +00:00
}
static void
setup_view (vulkan_ctx_t *ctx)
{
[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
mat4f_t view;
static mat4f_t z_up = {
{ 0, 0, -1, 0},
{-1, 0, 0, 0},
{ 0, 1, 0, 0},
{ 0, 0, 0, 1},
[vulkan] Make a lot of progress for brush models Light maps are maybe updating, but as nothing is actually rendered yet, it's hard to tell.
2021-01-19 16:25:54 +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
vec4f_t offset = { 0, 0, 0, 1 };
[vulkan] Make a lot of progress for brush models Light maps are maybe updating, but as nothing is actually rendered yet, it's hard to tell.
2021-01-19 16:25:54 +00:00
/*x = r_refdef.vrect.x;
y = (vid.height - (r_refdef.vrect.y + r_refdef.vrect.height));
w = r_refdef.vrect.width;
h = r_refdef.vrect.height;
qfeglViewport (x, y, w, h);*/
[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
mat4fquat (view, qconjf (r_refdef.viewrotation));
mmulf (view, z_up, view);
offset = -r_refdef.viewposition;
offset[3] = 1;
view[3] = mvmulf (view, offset);
memcpy (ctx->matrices.view_3d, view, sizeof (view));
[vulkan] Make a lot of progress for brush models Light maps are maybe updating, but as nothing is actually rendered yet, it's hard to tell.
2021-01-19 16:25:54 +00:00
}
static void
[vulkan] Support multiple render passes Multiple render passes are needed for supporting shadow mapping, and this is a huge step towards breaking the Vulkan render free of Quake, and hopefully will lead the way for breaking the GL renderers free as well.
2021-12-02 12:58:29 +00:00
Vulkan_RenderEntities (qfv_renderframe_t *rFrame)
[vulkan] Make a lot of progress for brush models Light maps are maybe updating, but as nothing is actually rendered yet, it's hard to tell.
2021-01-19 16:25:54 +00:00
{
if (!r_drawentities->int_val)
return;
#define RE_LOOP(type_name, Type) \
do { \
[vulkan] Create the lights buffer It's a tad bogus as it's the lights close to the camera, but it should at least be a good start once things are working. There's currently something very wrong with the state of things.
2021-01-27 07:16:28 +00:00
entity_t *ent; \
int begun = 0; \
[vulkan] Make a lot of progress for brush models Light maps are maybe updating, but as nothing is actually rendered yet, it's hard to tell.
2021-01-19 16:25:54 +00:00
for (ent = r_ent_queue; ent; ent = ent->next) { \
[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
if (ent->renderer.model->type != mod_##type_name) \
[vulkan] Make a lot of progress for brush models Light maps are maybe updating, but as nothing is actually rendered yet, it's hard to tell.
2021-01-19 16:25:54 +00:00
continue; \
if (!begun) { \
[vulkan] Support multiple render passes Multiple render passes are needed for supporting shadow mapping, and this is a huge step towards breaking the Vulkan render free of Quake, and hopefully will lead the way for breaking the GL renderers free as well.
2021-12-02 12:58:29 +00:00
Vulkan_##Type##Begin (rFrame); \
[vulkan] Make a lot of progress for brush models Light maps are maybe updating, but as nothing is actually rendered yet, it's hard to tell.
2021-01-19 16:25:54 +00:00
begun = 1; \
} \
[vulkan] Implement view model rendering Quake just looked wrong without the view model. I can't say I like the way the depth range is hacked, but it was necessary because the view model needs to be processed along with the rest of the alias models (didn't feel like adding more command buffers, which I imagine would be expensive with the pipeline switching).
2021-07-25 05:03:25 +00:00
/* hack the depth range to prevent view model */\
/* from poking into walls */\
if (ent == vr_data.view_model) { \
[vulkan] Support multiple render passes Multiple render passes are needed for supporting shadow mapping, and this is a huge step towards breaking the Vulkan render free of Quake, and hopefully will lead the way for breaking the GL renderers free as well.
2021-12-02 12:58:29 +00:00
Vulkan_AliasDepthRange (rFrame, 0, 0.3); \
[vulkan] Implement view model rendering Quake just looked wrong without the view model. I can't say I like the way the depth range is hacked, but it was necessary because the view model needs to be processed along with the rest of the alias models (didn't feel like adding more command buffers, which I imagine would be expensive with the pipeline switching).
2021-07-25 05:03:25 +00:00
} \
[vulkan] Support multiple render passes Multiple render passes are needed for supporting shadow mapping, and this is a huge step towards breaking the Vulkan render free of Quake, and hopefully will lead the way for breaking the GL renderers free as well.
2021-12-02 12:58:29 +00:00
Vulkan_Draw##Type (ent, rFrame); \
[vulkan] Implement view model rendering Quake just looked wrong without the view model. I can't say I like the way the depth range is hacked, but it was necessary because the view model needs to be processed along with the rest of the alias models (didn't feel like adding more command buffers, which I imagine would be expensive with the pipeline switching).
2021-07-25 05:03:25 +00:00
/* unhack in case the view_model is not the last */\
if (ent == vr_data.view_model) { \
[vulkan] Support multiple render passes Multiple render passes are needed for supporting shadow mapping, and this is a huge step towards breaking the Vulkan render free of Quake, and hopefully will lead the way for breaking the GL renderers free as well.
2021-12-02 12:58:29 +00:00
Vulkan_AliasDepthRange (rFrame, 0, 1); \
[vulkan] Implement view model rendering Quake just looked wrong without the view model. I can't say I like the way the depth range is hacked, but it was necessary because the view model needs to be processed along with the rest of the alias models (didn't feel like adding more command buffers, which I imagine would be expensive with the pipeline switching).
2021-07-25 05:03:25 +00:00
} \
[vulkan] Make a lot of progress for brush models Light maps are maybe updating, but as nothing is actually rendered yet, it's hard to tell.
2021-01-19 16:25:54 +00:00
} \
if (begun) \
[vulkan] Support multiple render passes Multiple render passes are needed for supporting shadow mapping, and this is a huge step towards breaking the Vulkan render free of Quake, and hopefully will lead the way for breaking the GL renderers free as well.
2021-12-02 12:58:29 +00:00
Vulkan_##Type##End (rFrame); \
[vulkan] Make a lot of progress for brush models Light maps are maybe updating, but as nothing is actually rendered yet, it's hard to tell.
2021-01-19 16:25:54 +00:00
} while (0)
[vulkan] Start work on alias model rendering
2021-01-26 11:58:24 +00:00
RE_LOOP (alias, Alias);
[vulkan] Make a lot of progress for brush models Light maps are maybe updating, but as nothing is actually rendered yet, it's hard to tell.
2021-01-19 16:25:54 +00:00
//RE_LOOP (iqm, IQM);
//RE_LOOP (sprite, Sprite);
}
static void
[vulkan] Implement view model rendering Quake just looked wrong without the view model. I can't say I like the way the depth range is hacked, but it was necessary because the view model needs to be processed along with the rest of the alias models (didn't feel like adding more command buffers, which I imagine would be expensive with the pipeline switching).
2021-07-25 05:03:25 +00:00
Vulkan_DrawViewModel (vulkan_ctx_t *ctx)
[vulkan] Make a lot of progress for brush models Light maps are maybe updating, but as nothing is actually rendered yet, it's hard to tell.
2021-01-19 16:25:54 +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;
[vulkan] Make a lot of progress for brush models Light maps are maybe updating, but as nothing is actually rendered yet, it's hard to tell.
2021-01-19 16:25:54 +00:00
if (vr_data.inhibit_viewmodel
|| !r_drawviewmodel->int_val
|| !r_drawentities->int_val
[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)
[vulkan] Make a lot of progress for brush models Light maps are maybe updating, but as nothing is actually rendered yet, it's hard to tell.
2021-01-19 16:25:54 +00:00
return;
[vulkan] Implement view model rendering Quake just looked wrong without the view model. I can't say I like the way the depth range is hacked, but it was necessary because the view model needs to be processed along with the rest of the alias models (didn't feel like adding more command buffers, which I imagine would be expensive with the pipeline switching).
2021-07-25 05:03:25 +00:00
R_EnqueueEntity (ent);
[vulkan] Make a lot of progress for brush models Light maps are maybe updating, but as nothing is actually rendered yet, it's hard to tell.
2021-01-19 16:25:54 +00:00
}
void
[vulkan] Support multiple render passes Multiple render passes are needed for supporting shadow mapping, and this is a huge step towards breaking the Vulkan render free of Quake, and hopefully will lead the way for breaking the GL renderers free as well.
2021-12-02 12:58:29 +00:00
Vulkan_RenderView (qfv_renderframe_t *rFrame)
[vulkan] Make a lot of progress for brush models Light maps are maybe updating, but as nothing is actually rendered yet, it's hard to tell.
2021-01-19 16:25:54 +00:00
{
[vulkan] Support multiple render passes Multiple render passes are needed for supporting shadow mapping, and this is a huge step towards breaking the Vulkan render free of Quake, and hopefully will lead the way for breaking the GL renderers free as well.
2021-12-02 12:58:29 +00:00
vulkan_ctx_t *ctx = rFrame->vulkan_ctx;
[vulkan] Implement view model rendering Quake just looked wrong without the view model. I can't say I like the way the depth range is hacked, but it was necessary because the view model needs to be processed along with the rest of the alias models (didn't feel like adding more command buffers, which I imagine would be expensive with the pipeline switching).
2021-07-25 05:03:25 +00:00
double t[9] = {};
[vulkan] Make a lot of progress for brush models Light maps are maybe updating, but as nothing is actually rendered yet, it's hard to tell.
2021-01-19 16:25:54 +00:00
int speeds = r_speeds->int_val;
[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
if (!r_worldentity.renderer.model) {
return;
}
[vulkan] Make a lot of progress for brush models Light maps are maybe updating, but as nothing is actually rendered yet, it's hard to tell.
2021-01-19 16:25:54 +00:00
if (speeds)
t[0] = Sys_DoubleTime ();
setup_frame (ctx);
setup_view (ctx);
if (speeds)
t[1] = Sys_DoubleTime ();
R_MarkLeaves ();
if (speeds)
t[2] = Sys_DoubleTime ();
R_PushDlights (vec3_origin);
if (speeds)
t[3] = Sys_DoubleTime ();
[vulkan] Support multiple render passes Multiple render passes are needed for supporting shadow mapping, and this is a huge step towards breaking the Vulkan render free of Quake, and hopefully will lead the way for breaking the GL renderers free as well.
2021-12-02 12:58:29 +00:00
Vulkan_DrawWorld (rFrame);
[vulkan] Make a lot of progress for brush models Light maps are maybe updating, but as nothing is actually rendered yet, it's hard to tell.
2021-01-19 16:25:54 +00:00
if (speeds)
t[4] = Sys_DoubleTime ();
[vulkan] Support multiple render passes Multiple render passes are needed for supporting shadow mapping, and this is a huge step towards breaking the Vulkan render free of Quake, and hopefully will lead the way for breaking the GL renderers free as well.
2021-12-02 12:58:29 +00:00
Vulkan_DrawSky (rFrame);
[vulkan] Make a lot of progress for brush models Light maps are maybe updating, but as nothing is actually rendered yet, it's hard to tell.
2021-01-19 16:25:54 +00:00
if (speeds)
t[5] = Sys_DoubleTime ();
[vulkan] Implement view model rendering Quake just looked wrong without the view model. I can't say I like the way the depth range is hacked, but it was necessary because the view model needs to be processed along with the rest of the alias models (didn't feel like adding more command buffers, which I imagine would be expensive with the pipeline switching).
2021-07-25 05:03:25 +00:00
Vulkan_DrawViewModel (ctx);
[vulkan] Support multiple render passes Multiple render passes are needed for supporting shadow mapping, and this is a huge step towards breaking the Vulkan render free of Quake, and hopefully will lead the way for breaking the GL renderers free as well.
2021-12-02 12:58:29 +00:00
Vulkan_RenderEntities (rFrame);
[vulkan] Make a lot of progress for brush models Light maps are maybe updating, but as nothing is actually rendered yet, it's hard to tell.
2021-01-19 16:25:54 +00:00
if (speeds)
t[6] = Sys_DoubleTime ();
[vulkan] Support multiple render passes Multiple render passes are needed for supporting shadow mapping, and this is a huge step towards breaking the Vulkan render free of Quake, and hopefully will lead the way for breaking the GL renderers free as well.
2021-12-02 12:58:29 +00:00
Vulkan_DrawWaterSurfaces (rFrame);
[vulkan] Make a lot of progress for brush models Light maps are maybe updating, but as nothing is actually rendered yet, it's hard to tell.
2021-01-19 16:25:54 +00:00
if (speeds)
t[7] = Sys_DoubleTime ();
Vulkan_DrawParticles (ctx);
if (speeds)
t[8] = Sys_DoubleTime ();
if (speeds) {
Sys_Printf ("frame: %g, setup: %g, mark: %g, pushdl: %g, world: %g,"
[vulkan] Implement view model rendering Quake just looked wrong without the view model. I can't say I like the way the depth range is hacked, but it was necessary because the view model needs to be processed along with the rest of the alias models (didn't feel like adding more command buffers, which I imagine would be expensive with the pipeline switching).
2021-07-25 05:03:25 +00:00
" sky: %g, ents: %g, water: %g, part: %g\n",
[vulkan] Fix a missed array index edit Caused by the view model implementation, and missed because gcc needs optimization to cache such errors (and my own error, of course).
2021-07-26 02:43:31 +00:00
(t[8] - t[0]) * 1000, (t[1] - t[0]) * 1000,
[vulkan] Make a lot of progress for brush models Light maps are maybe updating, but as nothing is actually rendered yet, it's hard to tell.
2021-01-19 16:25:54 +00:00
(t[2] - t[1]) * 1000, (t[3] - t[2]) * 1000,
(t[4] - t[3]) * 1000, (t[5] - t[4]) * 1000,
(t[6] - t[5]) * 1000, (t[7] - t[6]) * 1000,
[vulkan] Implement view model rendering Quake just looked wrong without the view model. I can't say I like the way the depth range is hacked, but it was necessary because the view model needs to be processed along with the rest of the alias models (didn't feel like adding more command buffers, which I imagine would be expensive with the pipeline switching).
2021-07-25 05:03:25 +00:00
(t[8] - t[7]) * 1000);
[vulkan] Make a lot of progress for brush models Light maps are maybe updating, but as nothing is actually rendered yet, it's hard to tell.
2021-01-19 16:25:54 +00:00
}
}
void
Vulkan_NewMap (model_t *worldmodel, struct model_s **models, int num_models,
vulkan_ctx_t *ctx)
{
int i;
for (i = 0; i < 256; i++) {
d_lightstylevalue[i] = 264; // normal light value
}
memset (&r_worldentity, 0, sizeof (r_worldentity));
[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
r_worldentity.renderer.model = worldmodel;
[vulkan] Make a lot of progress for brush models Light maps are maybe updating, but as nothing is actually rendered yet, it's hard to tell.
2021-01-19 16:25:54 +00:00
// Force a vis update
r_viewleaf = NULL;
R_MarkLeaves ();
R_FreeAllEntities ();
Vulkan_ClearParticles (ctx);
Vulkan_RegisterTextures (models, num_models, ctx);
[vulkan] Disable lightmap support Hopefully, it won't be needed.
2021-03-20 09:02:27 +00:00
//Vulkan_BuildLightmaps (models, num_models, ctx);
[vulkan] Make a lot of progress for brush models Light maps are maybe updating, but as nothing is actually rendered yet, it's hard to tell.
2021-01-19 16:25:54 +00:00
Vulkan_BuildDisplayLists (models, num_models, ctx);
[vulkan] Implement real-time lighting Light styles and shadows aren't implemented yet. The map's entities are used to create the lights, and the PVS used to determine which lights might be visible (ie, the surfaces they light). That could do with some more improvements (eg, checking if a leaf is outside a spotlight's cone), but the concept seems to work.
2021-03-20 07:08:44 +00:00
Vulkan_LoadLights (worldmodel, worldmodel->brush.entities, ctx);
[vulkan] Make a lot of progress for brush models Light maps are maybe updating, but as nothing is actually rendered yet, it's hard to tell.
2021-01-19 16:25:54 +00:00
}
/*void
glsl_R_LineGraph (int x, int y, int *h_vals, int count)
{
}*/
/*void
glsl_R_TimeRefresh_f (void)
{
double start, stop, time;
int i;
glsl_ctx->end_rendering ();
start = Sys_DoubleTime ();
for (i = 0; i < 128; i++) {
r_refdef.viewangles[1] = i * (360.0 / 128.0);
Vulkan_RenderView (ctx);
glsl_ctx->end_rendering ();
}
stop = Sys_DoubleTime ();
time = stop - start;
Sys_Printf ("%g seconds (%g fps)\n", time, 128 / time);
}*/
Reference in a new issue Copy permalink