qf/quakeforge
0
0
Fork 0
mirror of https://git.code.sf.net/p/quake/quakeforge synced 2024-11-29 23:52:22 +00:00
Code Issues Projects Releases Packages Wiki Activity
quakeforge/libs/models/iqm/vulkan_model_iqm.c

461 lines
14 KiB
C
Raw Normal View History

[vulkan] Add support for IQM models Despite the base IQM specification not supporting blend-shapes, I think IQM will become the basis for QF's generic model representation (at least for the more advanced renderers). After my experience with .mu models (KSP) and unity mesh objects (both normal and skinned), and reviewing the IQM spec, it looks like with the addition of support for blend-shapes, IQM is actually pretty good. This is just the preliminary work to get standard IQM models loading in vulkan (seems to work, along with unloading), and they very basics into the renderer (most likely not working: not tested yet). The rest of the renderer seems to be unaffected, though, which is good.
2022-05-04 05:07:27 +00:00
/*
vulkan_model_iqm.c
iqm model processing for Vulkan
Copyright (C) 2011 Bill Currie <bill@taniwha.org>
Author: Bill Currie <bill@taniwha.org>
Date: 2022/05/03
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 <stdlib.h>
#include "QF/dstring.h"
#include "QF/image.h"
#include "QF/quakefs.h"
#include "QF/va.h"
#include "QF/Vulkan/barrier.h"
#include "QF/Vulkan/device.h"
#include "QF/Vulkan/image.h"
#include "QF/Vulkan/resource.h"
#include "QF/Vulkan/staging.h"
#include "QF/Vulkan/qf_iqm.h"
#include "QF/Vulkan/qf_texture.h"
#include "mod_internal.h"
#include "r_shared.h"
#include "vid_vulkan.h"
static byte null_texture[] = {
204, 204, 204, 255,
204, 204, 204, 255,
204, 204, 204, 255,
204, 204, 204, 255,
};
#if 0
static byte null_normmap[] = {
127, 127, 255, 255,
127, 127, 255, 255,
127, 127, 255, 255,
127, 127, 255, 255,
};
#endif
static void
vulkan_iqm_clear (model_t *mod, void *data)
{
vulkan_ctx_t *ctx = data;
qfv_device_t *device = ctx->device;
iqm_t *iqm = (iqm_t *) mod->aliashdr;
qfv_iqm_t *mesh = iqm->extra_data;
mod->needload = true;
[vulkan] Get IQM rendering working The bones aren't animated yet (and I realized I made the mistake of thinking the bone buffer was per-model when it's really per-instance (I think this mistake is in the rest of QF, too)), skin rendering is a mess, need to default vertex attributes that aren't in the model... Still, it's quite satisfying seeing Mr Fixit on screen again :) I wound up moving the pipeline spec in with the rest of the pipelines as the system isn't really ready for separating them.
2022-05-07 01:08:46 +00:00
for (int i = 0; i < iqm->num_meshes; i++) {
Vulkan_IQMRemoveSkin (ctx, &mesh->skins[i]);
}
Vulkan_IQMRemoveBones (ctx, iqm);//FIXME doesn't belong here (per-instance)
[vulkan] Add support for IQM models Despite the base IQM specification not supporting blend-shapes, I think IQM will become the basis for QF's generic model representation (at least for the more advanced renderers). After my experience with .mu models (KSP) and unity mesh objects (both normal and skinned), and reviewing the IQM spec, it looks like with the addition of support for blend-shapes, IQM is actually pretty good. This is just the preliminary work to get standard IQM models loading in vulkan (seems to work, along with unloading), and they very basics into the renderer (most likely not working: not tested yet). The rest of the renderer seems to be unaffected, though, which is good.
2022-05-04 05:07:27 +00:00
QFV_DestroyResource (device, mesh->bones);
QFV_DestroyResource (device, mesh->mesh);
free (mesh);
Mod_FreeIQM (iqm);
}
static void
vulkan_iqm_init_image (iqm_t *iqm, int meshnum, qfv_resobj_t *image)
{
const char *material = iqm->text + iqm->meshes[meshnum].material;
dstring_t *str = dstring_new ();
dstring_copystr (str, material);
QFS_StripExtension (str->str, str->str);
[vulkan] Initialize resource image from tex_t This moves the qfv_resobj_t image initialization code from the IQM loader into the resource management code. This will allow me to reuse the code for setting up glyph data. As a bonus, it cleans up the IQM code nicely.
2022-10-02 06:24:38 +00:00
tex_t dummy_tex;
[vulkan] Add support for IQM models Despite the base IQM specification not supporting blend-shapes, I think IQM will become the basis for QF's generic model representation (at least for the more advanced renderers). After my experience with .mu models (KSP) and unity mesh objects (both normal and skinned), and reviewing the IQM spec, it looks like with the addition of support for blend-shapes, IQM is actually pretty good. This is just the preliminary work to get standard IQM models loading in vulkan (seems to work, along with unloading), and they very basics into the renderer (most likely not working: not tested yet). The rest of the renderer seems to be unaffected, though, which is good.
2022-05-04 05:07:27 +00:00
tex_t *tex;
[vulkan] Initialize resource image from tex_t This moves the qfv_resobj_t image initialization code from the IQM loader into the resource management code. This will allow me to reuse the code for setting up glyph data. As a bonus, it cleans up the IQM code nicely.
2022-10-02 06:24:38 +00:00
if (!(tex = LoadImage (va (0, "textures/%s", str->str), 0))) {
dummy_tex = (tex_t) {
.width = 2,
.height = 2,
.format = tex_rgba,
[vulkan] Add support for IQM models Despite the base IQM specification not supporting blend-shapes, I think IQM will become the basis for QF's generic model representation (at least for the more advanced renderers). After my experience with .mu models (KSP) and unity mesh objects (both normal and skinned), and reviewing the IQM spec, it looks like with the addition of support for blend-shapes, IQM is actually pretty good. This is just the preliminary work to get standard IQM models loading in vulkan (seems to work, along with unloading), and they very basics into the renderer (most likely not working: not tested yet). The rest of the renderer seems to be unaffected, though, which is good.
2022-05-04 05:07:27 +00:00
};
[vulkan] Initialize resource image from tex_t This moves the qfv_resobj_t image initialization code from the IQM loader into the resource management code. This will allow me to reuse the code for setting up glyph data. As a bonus, it cleans up the IQM code nicely.
2022-10-02 06:24:38 +00:00
tex = &dummy_tex;
[vulkan] Add support for IQM models Despite the base IQM specification not supporting blend-shapes, I think IQM will become the basis for QF's generic model representation (at least for the more advanced renderers). After my experience with .mu models (KSP) and unity mesh objects (both normal and skinned), and reviewing the IQM spec, it looks like with the addition of support for blend-shapes, IQM is actually pretty good. This is just the preliminary work to get standard IQM models loading in vulkan (seems to work, along with unloading), and they very basics into the renderer (most likely not working: not tested yet). The rest of the renderer seems to be unaffected, though, which is good.
2022-05-04 05:07:27 +00:00
}
[vulkan] Initialize resource image from tex_t This moves the qfv_resobj_t image initialization code from the IQM loader into the resource management code. This will allow me to reuse the code for setting up glyph data. As a bonus, it cleans up the IQM code nicely.
2022-10-02 06:24:38 +00:00
QFV_ResourceInitTexImage (image, material, 1, tex);
[vulkan] Add support for IQM models Despite the base IQM specification not supporting blend-shapes, I think IQM will become the basis for QF's generic model representation (at least for the more advanced renderers). After my experience with .mu models (KSP) and unity mesh objects (both normal and skinned), and reviewing the IQM spec, it looks like with the addition of support for blend-shapes, IQM is actually pretty good. This is just the preliminary work to get standard IQM models loading in vulkan (seems to work, along with unloading), and they very basics into the renderer (most likely not working: not tested yet). The rest of the renderer seems to be unaffected, though, which is good.
2022-05-04 05:07:27 +00:00
dstring_delete (str);
}
static void
iqm_transfer_texture (tex_t *tex, VkImage image, qfv_stagebuf_t *stage,
qfv_device_t *device)
{
qfv_devfuncs_t *dfunc = device->funcs;
if (tex->format != tex_rgb && tex->format != tex_rgba) {
Sys_Error ("can't transfer iqm image");
}
// FIXME correct only for rgb and rgba
size_t layer_size = tex->width * tex->height * tex->format;
qfv_packet_t *packet = QFV_PacketAcquire (stage);
byte *dst = QFV_PacketExtend (packet, layer_size);
qfv_imagebarrier_t ib = imageBarriers[qfv_LT_Undefined_to_TransferDst];
ib.barrier.image = image;
ib.barrier.subresourceRange.levelCount = VK_REMAINING_MIP_LEVELS;
ib.barrier.subresourceRange.layerCount = VK_REMAINING_ARRAY_LAYERS;
dfunc->vkCmdPipelineBarrier (packet->cmd, ib.srcStages, ib.dstStages,
0, 0, 0, 0, 0,
1, &ib.barrier);
memcpy (dst, tex->data, layer_size);
[vulkan] Get IQM rendering working The bones aren't animated yet (and I realized I made the mistake of thinking the bone buffer was per-model when it's really per-instance (I think this mistake is in the rest of QF, too)), skin rendering is a mess, need to default vertex attributes that aren't in the model... Still, it's quite satisfying seeing Mr Fixit on screen again :) I wound up moving the pipeline spec in with the rest of the pipelines as the system isn't really ready for separating them.
2022-05-07 01:08:46 +00:00
VkBufferImageCopy copy = {
packet->offset, 0, 0,
{VK_IMAGE_ASPECT_COLOR_BIT, 0, 0, 1},
{0, 0, 0}, {tex->width, tex->height, 1},
};
dfunc->vkCmdCopyBufferToImage (packet->cmd, packet->stage->buffer,
image, VK_IMAGE_LAYOUT_TRANSFER_DST_OPTIMAL,
1, &copy);
[vulkan] Add support for IQM models Despite the base IQM specification not supporting blend-shapes, I think IQM will become the basis for QF's generic model representation (at least for the more advanced renderers). After my experience with .mu models (KSP) and unity mesh objects (both normal and skinned), and reviewing the IQM spec, it looks like with the addition of support for blend-shapes, IQM is actually pretty good. This is just the preliminary work to get standard IQM models loading in vulkan (seems to work, along with unloading), and they very basics into the renderer (most likely not working: not tested yet). The rest of the renderer seems to be unaffected, though, which is good.
2022-05-04 05:07:27 +00:00
int mipLevels = QFV_MipLevels (tex->width, tex->height);
if (mipLevels == 1) {
ib = imageBarriers[qfv_LT_TransferDst_to_ShaderReadOnly];
ib.barrier.image = image;
ib.barrier.subresourceRange.levelCount = VK_REMAINING_MIP_LEVELS;
ib.barrier.subresourceRange.layerCount = VK_REMAINING_ARRAY_LAYERS;
dfunc->vkCmdPipelineBarrier (packet->cmd, ib.srcStages, ib.dstStages,
0, 0, 0, 0, 0,
1, &ib.barrier);
} else {
QFV_GenerateMipMaps (device, packet->cmd, image, mipLevels,
tex->width, tex->height, 1);
}
QFV_PacketSubmit (packet);
}
static void
vulkan_iqm_load_textures (model_t *mod, iqm_t *iqm, qfv_iqm_t *mesh,
vulkan_ctx_t *ctx)
{
qfv_device_t *device = ctx->device;
dstring_t *str = dstring_new ();
tex_t *tex;
size_t buff_size = 0;
qfv_resobj_t *objects = mesh->mesh->objects;
for (int i = 0; i < iqm->num_meshes; i++) {
int image_ind = 3 + 2 * i;
VkExtent3D extent = objects[image_ind].image.extent;
// probably 3 or 4 bytes per pixel FIXME
buff_size = max (buff_size, extent.width * extent.height * 4);
}
qfv_stagebuf_t *stage = QFV_CreateStagingBuffer (device,
va (ctx->va_ctx, "iqm:%s",
mod->name),
4 * buff_size,
ctx->cmdpool);
for (int i = 0; i < iqm->num_meshes; i++) {
int image_ind = 3 + 2 * i;
__auto_type image = &objects[image_ind].image;
__auto_type view = &objects[image_ind + 1].image_view;
qfv_iqm_skin_t *skin = &mesh->skins[i];
*skin = (qfv_iqm_skin_t) {
.view = view->view,
.colora = { 255, 255, 255, 255 },
.colorb = { 255, 255, 255, 255 },
};
dstring_copystr (str, iqm->text + iqm->meshes[i].material);
QFS_StripExtension (str->str, str->str);
if (!(tex = LoadImage (va (0, "textures/%s", str->str), 1))) {
tex_t null_tex = {
.width = 2,
.height = 2,
.format = tex_rgba,
.data = null_texture,
};
tex = &null_tex;
}
iqm_transfer_texture (tex, image->image, stage, device);
[vulkan] Get IQM rendering working The bones aren't animated yet (and I realized I made the mistake of thinking the bone buffer was per-model when it's really per-instance (I think this mistake is in the rest of QF, too)), skin rendering is a mess, need to default vertex attributes that aren't in the model... Still, it's quite satisfying seeing Mr Fixit on screen again :) I wound up moving the pipeline spec in with the rest of the pipelines as the system isn't really ready for separating them.
2022-05-07 01:08:46 +00:00
Vulkan_IQMAddSkin (ctx, skin);
[vulkan] Add support for IQM models Despite the base IQM specification not supporting blend-shapes, I think IQM will become the basis for QF's generic model representation (at least for the more advanced renderers). After my experience with .mu models (KSP) and unity mesh objects (both normal and skinned), and reviewing the IQM spec, it looks like with the addition of support for blend-shapes, IQM is actually pretty good. This is just the preliminary work to get standard IQM models loading in vulkan (seems to work, along with unloading), and they very basics into the renderer (most likely not working: not tested yet). The rest of the renderer seems to be unaffected, though, which is good.
2022-05-04 05:07:27 +00:00
}
dstring_delete (str);
QFV_DestroyStagingBuffer (stage);
}
static void
vulkan_iqm_load_arrays (model_t *mod, iqm_t *iqm, qfv_iqm_t *mesh,
vulkan_ctx_t *ctx)
{
qfv_device_t *device = ctx->device;
qfv_devfuncs_t *dfunc = device->funcs;
[vulkan] Get IQM rendering working The bones aren't animated yet (and I realized I made the mistake of thinking the bone buffer was per-model when it's really per-instance (I think this mistake is in the rest of QF, too)), skin rendering is a mess, need to default vertex attributes that aren't in the model... Still, it's quite satisfying seeing Mr Fixit on screen again :) I wound up moving the pipeline spec in with the rest of the pipelines as the system isn't really ready for separating them.
2022-05-07 01:08:46 +00:00
iqmctx_t *ictx = ctx->iqm_context;
[vulkan] Add support for IQM models Despite the base IQM specification not supporting blend-shapes, I think IQM will become the basis for QF's generic model representation (at least for the more advanced renderers). After my experience with .mu models (KSP) and unity mesh objects (both normal and skinned), and reviewing the IQM spec, it looks like with the addition of support for blend-shapes, IQM is actually pretty good. This is just the preliminary work to get standard IQM models loading in vulkan (seems to work, along with unloading), and they very basics into the renderer (most likely not working: not tested yet). The rest of the renderer seems to be unaffected, though, which is good.
2022-05-04 05:07:27 +00:00
size_t geom_size = iqm->num_verts * sizeof (iqmgvert_t);
size_t rend_size = iqm->num_verts * sizeof (iqmrvert_t);
size_t elem_size = iqm->num_elements * sizeof (uint16_t);
size_t buff_size = geom_size + rend_size + elem_size + 1024;
qfv_stagebuf_t *stage = QFV_CreateStagingBuffer (device,
va (ctx->va_ctx, "iqm:%s",
mod->name),
buff_size, ctx->cmdpool);
qfv_packet_t *gpacket = QFV_PacketAcquire (stage);
iqmgvert_t *gverts = QFV_PacketExtend (gpacket, geom_size);
qfv_packet_t *rpacket = QFV_PacketAcquire (stage);
iqmrvert_t *rverts = QFV_PacketExtend (rpacket, rend_size);
qfv_packet_t *epacket = QFV_PacketAcquire (stage);
uint16_t *elements = QFV_PacketExtend (epacket, elem_size);
//FIXME this whole thing is silly, but some person went and interleaved
//all the vertex data prematurely
for (int i = 0; i < iqm->num_verts; i++) {
byte *data = iqm->vertices + i * iqm->stride;
iqmgvert_t *gv = gverts + i;
iqmrvert_t *rv = rverts + i;
for (int j = 0; j < iqm->num_arrays; j++) {
__auto_type va = &iqm->vertexarrays[j];
// FIXME assumes standard iqm sizes
size_t size = 0;
switch (va->type) {
case IQM_POSITION:
size = sizeof (gv->vertex);
memcpy (gv->vertex, data, size);
break;
case IQM_TEXCOORD:
size = sizeof (rv->uv);
memcpy (rv->uv, data, size);
break;
case IQM_NORMAL:
size = sizeof (rv->normal);
memcpy (rv->normal, data, size);
break;
case IQM_TANGENT:
size = sizeof (rv->tangent);
memcpy (rv->tangent, data, size);
break;
case IQM_BLENDINDEXES:
size = sizeof (gv->bones);
memcpy (gv->bones, data, size);
break;
case IQM_BLENDWEIGHTS:
size = sizeof (gv->weights);
memcpy (gv->weights, data, size);
break;
case IQM_COLOR:
size = sizeof (rv->color);
memcpy (rv->color, data, size);
break;
case IQM_CUSTOM:
// FIXME model loader doesn't handle these, so nothing to do
break;
}
data += size;
}
}
memcpy (elements, iqm->elements, elem_size);
qfv_bufferbarrier_t bb[] = {
bufferBarriers[qfv_BB_Unknown_to_TransferWrite],
bufferBarriers[qfv_BB_Unknown_to_TransferWrite],
bufferBarriers[qfv_BB_Unknown_to_TransferWrite],
};
bb[0].barrier.buffer = mesh->geom_buffer;
bb[0].barrier.size = geom_size;
bb[1].barrier.buffer = mesh->rend_buffer;
bb[1].barrier.size = rend_size;
bb[2].barrier.buffer = mesh->index_buffer;
bb[2].barrier.size = elem_size;
VkBufferCopy copy_region[] = {
{ gpacket->offset, 0, geom_size },
{ rpacket->offset, 0, rend_size },
{ epacket->offset, 0, elem_size },
};
dfunc->vkCmdPipelineBarrier (gpacket->cmd, bb[0].srcStages, bb[0].dstStages,
0, 0, 0, 1, &bb[0].barrier, 0, 0);
dfunc->vkCmdPipelineBarrier (rpacket->cmd, bb[0].srcStages, bb[0].dstStages,
0, 0, 0, 1, &bb[1].barrier, 0, 0);
dfunc->vkCmdPipelineBarrier (epacket->cmd, bb[0].srcStages, bb[0].dstStages,
0, 0, 0, 1, &bb[2].barrier, 0, 0);
dfunc->vkCmdCopyBuffer (gpacket->cmd, stage->buffer,
mesh->geom_buffer, 1, &copy_region[0]);
dfunc->vkCmdCopyBuffer (rpacket->cmd, stage->buffer,
mesh->rend_buffer, 1, &copy_region[1]);
dfunc->vkCmdCopyBuffer (epacket->cmd, stage->buffer,
mesh->index_buffer, 1, &copy_region[2]);
bb[0] = bufferBarriers[qfv_BB_TransferWrite_to_VertexAttrRead];
bb[1] = bufferBarriers[qfv_BB_TransferWrite_to_VertexAttrRead];
bb[2] = bufferBarriers[qfv_BB_TransferWrite_to_VertexAttrRead];
bb[0].barrier.buffer = mesh->geom_buffer;
bb[0].barrier.size = geom_size;
bb[1].barrier.buffer = mesh->rend_buffer;
bb[1].barrier.size = rend_size;
bb[2].barrier.buffer = mesh->index_buffer;
bb[2].barrier.size = elem_size;
dfunc->vkCmdPipelineBarrier (gpacket->cmd, bb[0].srcStages, bb[0].dstStages,
0, 0, 0, 1, &bb[0].barrier, 0, 0);
dfunc->vkCmdPipelineBarrier (rpacket->cmd, bb[0].srcStages, bb[0].dstStages,
0, 0, 0, 1, &bb[1].barrier, 0, 0);
dfunc->vkCmdPipelineBarrier (epacket->cmd, bb[0].srcStages, bb[0].dstStages,
0, 0, 0, 1, &bb[2].barrier, 0, 0);
QFV_PacketSubmit (gpacket);
QFV_PacketSubmit (rpacket);
QFV_PacketSubmit (epacket);
QFV_DestroyStagingBuffer (stage);
vec4f_t *bone_data;
dfunc->vkMapMemory (device->dev, mesh->bones->memory, 0, VK_WHOLE_SIZE,
0, (void **)&bone_data);
[vulkan] Get IQM rendering working The bones aren't animated yet (and I realized I made the mistake of thinking the bone buffer was per-model when it's really per-instance (I think this mistake is in the rest of QF, too)), skin rendering is a mess, need to default vertex attributes that aren't in the model... Still, it's quite satisfying seeing Mr Fixit on screen again :) I wound up moving the pipeline spec in with the rest of the pipelines as the system isn't really ready for separating them.
2022-05-07 01:08:46 +00:00
for (size_t i = 0; i < ictx->frames.size * iqm->num_joints; i++) {
[vulkan] Add support for IQM models Despite the base IQM specification not supporting blend-shapes, I think IQM will become the basis for QF's generic model representation (at least for the more advanced renderers). After my experience with .mu models (KSP) and unity mesh objects (both normal and skinned), and reviewing the IQM spec, it looks like with the addition of support for blend-shapes, IQM is actually pretty good. This is just the preliminary work to get standard IQM models loading in vulkan (seems to work, along with unloading), and they very basics into the renderer (most likely not working: not tested yet). The rest of the renderer seems to be unaffected, though, which is good.
2022-05-04 05:07:27 +00:00
vec4f_t *bone = bone_data + i * 3;
bone[0] = (vec4f_t) {1, 0, 0, 0};
bone[1] = (vec4f_t) {0, 1, 0, 0};
bone[2] = (vec4f_t) {0, 0, 1, 0};
}
VkMappedMemoryRange range = {
VK_STRUCTURE_TYPE_MAPPED_MEMORY_RANGE, 0,
mesh->bones->memory, 0, VK_WHOLE_SIZE,
};
dfunc->vkFlushMappedMemoryRanges (device->dev, 1, &range);
dfunc->vkUnmapMemory (device->dev, mesh->bones->memory);
[vulkan] Get IQM rendering working The bones aren't animated yet (and I realized I made the mistake of thinking the bone buffer was per-model when it's really per-instance (I think this mistake is in the rest of QF, too)), skin rendering is a mess, need to default vertex attributes that aren't in the model... Still, it's quite satisfying seeing Mr Fixit on screen again :) I wound up moving the pipeline spec in with the rest of the pipelines as the system isn't really ready for separating them.
2022-05-07 01:08:46 +00:00
Vulkan_IQMAddBones (ctx, iqm); //FIXME doesn't belong here (per-instance)
[vulkan] Add support for IQM models Despite the base IQM specification not supporting blend-shapes, I think IQM will become the basis for QF's generic model representation (at least for the more advanced renderers). After my experience with .mu models (KSP) and unity mesh objects (both normal and skinned), and reviewing the IQM spec, it looks like with the addition of support for blend-shapes, IQM is actually pretty good. This is just the preliminary work to get standard IQM models loading in vulkan (seems to work, along with unloading), and they very basics into the renderer (most likely not working: not tested yet). The rest of the renderer seems to be unaffected, though, which is good.
2022-05-04 05:07:27 +00:00
}
void
Vulkan_Mod_IQMFinish (model_t *mod, vulkan_ctx_t *ctx)
{
qfv_device_t *device = ctx->device;
[vulkan] Get IQM rendering working The bones aren't animated yet (and I realized I made the mistake of thinking the bone buffer was per-model when it's really per-instance (I think this mistake is in the rest of QF, too)), skin rendering is a mess, need to default vertex attributes that aren't in the model... Still, it's quite satisfying seeing Mr Fixit on screen again :) I wound up moving the pipeline spec in with the rest of the pipelines as the system isn't really ready for separating them.
2022-05-07 01:08:46 +00:00
iqmctx_t *ictx = ctx->iqm_context;
[vulkan] Add support for IQM models Despite the base IQM specification not supporting blend-shapes, I think IQM will become the basis for QF's generic model representation (at least for the more advanced renderers). After my experience with .mu models (KSP) and unity mesh objects (both normal and skinned), and reviewing the IQM spec, it looks like with the addition of support for blend-shapes, IQM is actually pretty good. This is just the preliminary work to get standard IQM models loading in vulkan (seems to work, along with unloading), and they very basics into the renderer (most likely not working: not tested yet). The rest of the renderer seems to be unaffected, though, which is good.
2022-05-04 05:07:27 +00:00
iqm_t *iqm = (iqm_t *) mod->aliashdr;
mod->clear = vulkan_iqm_clear;
mod->data = ctx;
// FIXME assumes only one texture per mesh (currently the case, but
// when materials are added...)
// 2 is for image + image view
int num_objects = 4 + 2 * iqm->num_meshes;
qfv_iqm_t *mesh = calloc (1, sizeof (qfv_iqm_t)
[vulkan] Get IQM rendering working The bones aren't animated yet (and I realized I made the mistake of thinking the bone buffer was per-model when it's really per-instance (I think this mistake is in the rest of QF, too)), skin rendering is a mess, need to default vertex attributes that aren't in the model... Still, it's quite satisfying seeing Mr Fixit on screen again :) I wound up moving the pipeline spec in with the rest of the pipelines as the system isn't really ready for separating them.
2022-05-07 01:08:46 +00:00
+ ictx->frames.size * sizeof (VkDescriptorSet)
[vulkan] Add support for IQM models Despite the base IQM specification not supporting blend-shapes, I think IQM will become the basis for QF's generic model representation (at least for the more advanced renderers). After my experience with .mu models (KSP) and unity mesh objects (both normal and skinned), and reviewing the IQM spec, it looks like with the addition of support for blend-shapes, IQM is actually pretty good. This is just the preliminary work to get standard IQM models loading in vulkan (seems to work, along with unloading), and they very basics into the renderer (most likely not working: not tested yet). The rest of the renderer seems to be unaffected, though, which is good.
2022-05-04 05:07:27 +00:00
+ 2 * sizeof (qfv_resource_t)
+ num_objects * sizeof (qfv_resobj_t)
+ iqm->num_meshes * sizeof (qfv_iqm_skin_t));
[vulkan] Get IQM rendering working The bones aren't animated yet (and I realized I made the mistake of thinking the bone buffer was per-model when it's really per-instance (I think this mistake is in the rest of QF, too)), skin rendering is a mess, need to default vertex attributes that aren't in the model... Still, it's quite satisfying seeing Mr Fixit on screen again :) I wound up moving the pipeline spec in with the rest of the pipelines as the system isn't really ready for separating them.
2022-05-07 01:08:46 +00:00
mesh->bones_descriptors = (VkDescriptorSet *) &mesh[1];
mesh->bones = (qfv_resource_t *)&mesh->bones_descriptors[ictx->frames.size];
[vulkan] Add support for IQM models Despite the base IQM specification not supporting blend-shapes, I think IQM will become the basis for QF's generic model representation (at least for the more advanced renderers). After my experience with .mu models (KSP) and unity mesh objects (both normal and skinned), and reviewing the IQM spec, it looks like with the addition of support for blend-shapes, IQM is actually pretty good. This is just the preliminary work to get standard IQM models loading in vulkan (seems to work, along with unloading), and they very basics into the renderer (most likely not working: not tested yet). The rest of the renderer seems to be unaffected, though, which is good.
2022-05-04 05:07:27 +00:00
mesh->mesh = &mesh->bones[1];
mesh->bones[0] = (qfv_resource_t) {
.name = mod->name,
.va_ctx = ctx->va_ctx,
.memory_properties = VK_MEMORY_PROPERTY_HOST_CACHED_BIT,
.num_objects = 1,
.objects = (qfv_resobj_t *) &mesh->bones[2],
};
mesh->bones->objects[0] = (qfv_resobj_t) {
.name = "bones",
.type = qfv_res_buffer,
.buffer = {
[vulkan] Get IQM rendering working The bones aren't animated yet (and I realized I made the mistake of thinking the bone buffer was per-model when it's really per-instance (I think this mistake is in the rest of QF, too)), skin rendering is a mess, need to default vertex attributes that aren't in the model... Still, it's quite satisfying seeing Mr Fixit on screen again :) I wound up moving the pipeline spec in with the rest of the pipelines as the system isn't really ready for separating them.
2022-05-07 01:08:46 +00:00
.size = ictx->frames.size * iqm->num_joints * 3 * sizeof (vec4f_t),
[vulkan] Add support for IQM models Despite the base IQM specification not supporting blend-shapes, I think IQM will become the basis for QF's generic model representation (at least for the more advanced renderers). After my experience with .mu models (KSP) and unity mesh objects (both normal and skinned), and reviewing the IQM spec, it looks like with the addition of support for blend-shapes, IQM is actually pretty good. This is just the preliminary work to get standard IQM models loading in vulkan (seems to work, along with unloading), and they very basics into the renderer (most likely not working: not tested yet). The rest of the renderer seems to be unaffected, though, which is good.
2022-05-04 05:07:27 +00:00
.usage = VK_BUFFER_USAGE_TRANSFER_DST_BIT
[vulkan] Get IQM rendering working The bones aren't animated yet (and I realized I made the mistake of thinking the bone buffer was per-model when it's really per-instance (I think this mistake is in the rest of QF, too)), skin rendering is a mess, need to default vertex attributes that aren't in the model... Still, it's quite satisfying seeing Mr Fixit on screen again :) I wound up moving the pipeline spec in with the rest of the pipelines as the system isn't really ready for separating them.
2022-05-07 01:08:46 +00:00
| VK_BUFFER_USAGE_STORAGE_BUFFER_BIT,
[vulkan] Add support for IQM models Despite the base IQM specification not supporting blend-shapes, I think IQM will become the basis for QF's generic model representation (at least for the more advanced renderers). After my experience with .mu models (KSP) and unity mesh objects (both normal and skinned), and reviewing the IQM spec, it looks like with the addition of support for blend-shapes, IQM is actually pretty good. This is just the preliminary work to get standard IQM models loading in vulkan (seems to work, along with unloading), and they very basics into the renderer (most likely not working: not tested yet). The rest of the renderer seems to be unaffected, though, which is good.
2022-05-04 05:07:27 +00:00
},
};
mesh->mesh[0] = (qfv_resource_t) {
.name = "mesh",
.va_ctx = ctx->va_ctx,
.memory_properties = VK_MEMORY_PROPERTY_DEVICE_LOCAL_BIT,
.num_objects = num_objects - 1,
.objects = mesh->bones->objects + 1,
};
mesh->mesh->objects[0] = (qfv_resobj_t) {
.name = "geom",
.type = qfv_res_buffer,
.buffer = {
.size = iqm->num_verts * sizeof (iqmgvert_t),
.usage = VK_BUFFER_USAGE_TRANSFER_DST_BIT
| VK_BUFFER_USAGE_VERTEX_BUFFER_BIT,
},
};
mesh->mesh->objects[1] = (qfv_resobj_t) {
.name = "rend",
.type = qfv_res_buffer,
.buffer = {
.size = iqm->num_verts * sizeof (iqmrvert_t),
.usage = VK_BUFFER_USAGE_TRANSFER_DST_BIT
| VK_BUFFER_USAGE_VERTEX_BUFFER_BIT,
},
};
mesh->mesh->objects[2] = (qfv_resobj_t) {
.name = "index",
.type = qfv_res_buffer,
.buffer = {
.size = iqm->num_elements * sizeof (uint16_t),
.usage = VK_BUFFER_USAGE_TRANSFER_DST_BIT
| VK_BUFFER_USAGE_INDEX_BUFFER_BIT,
},
};
for (int i = 0; i < iqm->num_meshes; i++) {
int image_ind = 3 + 2 * i;
__auto_type image = &mesh->mesh->objects[image_ind];
vulkan_iqm_init_image (iqm, i, image);
mesh->mesh->objects[image_ind + 1] = (qfv_resobj_t) {
.name = "view",
.type = qfv_res_image_view,
.image_view = {
.image = image_ind,
.type = VK_IMAGE_VIEW_TYPE_2D,
.format = mesh->mesh->objects[image_ind].image.format,
.aspect = VK_IMAGE_ASPECT_COLOR_BIT,
},
};
}
mesh->skins = (qfv_iqm_skin_t *) &mesh->bones->objects[num_objects];
QFV_CreateResource (device, mesh->mesh);
QFV_CreateResource (device, mesh->bones);
mesh->geom_buffer = mesh->mesh->objects[0].buffer.buffer;
mesh->rend_buffer = mesh->mesh->objects[1].buffer.buffer;
mesh->index_buffer = mesh->mesh->objects[2].buffer.buffer;
[vulkan] Get IQM rendering working The bones aren't animated yet (and I realized I made the mistake of thinking the bone buffer was per-model when it's really per-instance (I think this mistake is in the rest of QF, too)), skin rendering is a mess, need to default vertex attributes that aren't in the model... Still, it's quite satisfying seeing Mr Fixit on screen again :) I wound up moving the pipeline spec in with the rest of the pipelines as the system isn't really ready for separating them.
2022-05-07 01:08:46 +00:00
mesh->bones_buffer = mesh->bones->objects[0].buffer.buffer;
iqm->extra_data = mesh;
[vulkan] Add support for IQM models Despite the base IQM specification not supporting blend-shapes, I think IQM will become the basis for QF's generic model representation (at least for the more advanced renderers). After my experience with .mu models (KSP) and unity mesh objects (both normal and skinned), and reviewing the IQM spec, it looks like with the addition of support for blend-shapes, IQM is actually pretty good. This is just the preliminary work to get standard IQM models loading in vulkan (seems to work, along with unloading), and they very basics into the renderer (most likely not working: not tested yet). The rest of the renderer seems to be unaffected, though, which is good.
2022-05-04 05:07:27 +00:00
vulkan_iqm_load_textures (mod, iqm, mesh, ctx);
vulkan_iqm_load_arrays (mod, iqm, mesh, ctx);
}
Reference in a new issue Copy permalink