blender
/
blender
136
406
Fork 590
Code Issues 6.2k Pull Requests 764 Projects 19 Wiki Activity

Docs: add additional notes on tessellation, update comments

This commit is contained in:
Campbell Barton 2021-06-20 17:01:09 +10:00
parent 2d60c496a2
commit 3462b4c7ae
2 changed files with 43 additions and 39 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

69
source/blender/blenkernel/intern/mesh_tessellate.c
View File

@ -49,6 +49,8 @@
/* -------------------------------------------------------------------- */
/** \name MFace Tessellation
*
* #MFace is a legacy data-structure that should be avoided, use #MLoopTri instead.
* \{ */
/**
@ -67,11 +69,10 @@ void BKE_mesh_loops_to_tessdata(CustomData *fdata,
uint (*loopindices)[4],
const int num_faces)
{
/* Note: performances are sub-optimal when we get a NULL mface,
* we could be ~25% quicker with dedicated code...
* Issue is, unless having two different functions with nearly the same code,
* there's not much ways to solve this. Better imho to live with it for now. :/ --mont29
*/
/* NOTE(mont29): performances are sub-optimal when we get a NULL #MFace,
* we could be ~25% quicker with dedicated code.
* The issue is, unless having two different functions with nearly the same code,
* there's not much ways to solve this. Better IMHO to live with it for now (sigh). */
const int numUV = CustomData_number_of_layers(ldata, CD_MLOOPUV);
const int numCol = CustomData_number_of_layers(ldata, CD_MLOOPCOL);
const bool hasPCol = CustomData_has_layer(ldata, CD_PREVIEW_MLOOPCOL);
@ -139,7 +140,7 @@ void BKE_mesh_loops_to_tessdata(CustomData *fdata,
}
if (hasLoopTangent) {
/* need to do for all uv maps at some point */
/* Need to do for all UV maps at some point. */
float(*ftangents)[4] = CustomData_get_layer(fdata, CD_TANGENT);
float(*ltangents)[4] = CustomData_get_layer(ldata, CD_TANGENT);
@ -154,12 +155,15 @@ void BKE_mesh_loops_to_tessdata(CustomData *fdata,
}
/**
* Recreate tessellation.
* Recreate #MFace Tessellation.
*
* \param do_face_nor_copy: Controls whether the normals from the poly
* are copied to the tessellated faces.
*
* \return number of tessellation faces.
*
* \note This doesn't use multi-threading like #BKE_mesh_recalc_looptri since
* it's not used in many places and #MFace should be phased out.
*/
int BKE_mesh_tessface_calc_ex(CustomData *fdata,
CustomData *ldata,
@ -170,13 +174,10 @@ int BKE_mesh_tessface_calc_ex(CustomData *fdata,
int totpoly,
const bool do_face_nor_copy)
{
/* use this to avoid locking pthread for _every_ polygon
* and calling the fill function */
#define USE_TESSFACE_SPEEDUP
#define USE_TESSFACE_QUADS /* NEEDS FURTHER TESTING */
#define USE_TESSFACE_QUADS
/* We abuse MFace->edcode to tag quad faces. See below for details. */
/* We abuse #MFace.edcode to tag quad faces. See below for details. */
#define TESSFACE_IS_QUAD 1
const int looptri_num = poly_to_tri_count(totpoly, totloop);
@ -193,9 +194,9 @@ int BKE_mesh_tessface_calc_ex(CustomData *fdata,
mpoly = CustomData_get_layer(pdata, CD_MPOLY);
mloop = CustomData_get_layer(ldata, CD_MLOOP);
/* allocate the length of totfaces, avoid many small reallocs,
* if all faces are tri's it will be correct, quads == 2x allocs */
/* take care. we are _not_ calloc'ing so be sure to initialize each field */
/* Allocate the length of `totfaces`, avoid many small reallocation's,
* if all faces are triangles it will be correct, `quads == 2x` allocations. */
/* Take care since memory is _not_ zeroed so be sure to initialize each field. */
mface_to_poly_map = MEM_malloc_arrayN((size_t)looptri_num, sizeof(*mface_to_poly_map), __func__);
mface = MEM_malloc_arrayN((size_t)looptri_num, sizeof(*mface), __func__);
lindices = MEM_malloc_arrayN((size_t)looptri_num, sizeof(*lindices), __func__);
@ -208,7 +209,7 @@ int BKE_mesh_tessface_calc_ex(CustomData *fdata,
uint l1, l2, l3, l4;
uint *lidx;
if (mp_totloop < 3) {
/* do nothing */
/* Do nothing. */
}
#ifdef USE_TESSFACE_SPEEDUP
@ -217,7 +218,7 @@ int BKE_mesh_tessface_calc_ex(CustomData *fdata,
mface_to_poly_map[mface_index] = poly_index; \
mf = &mface[mface_index]; \
lidx = lindices[mface_index]; \
/* set loop indices, transformed to vert indices later */ \
/* Set loop indices, transformed to vert indices later. */ \
l1 = mp_loopstart + i1; \
l2 = mp_loopstart + i2; \
l3 = mp_loopstart + i3; \
@ -239,7 +240,7 @@ int BKE_mesh_tessface_calc_ex(CustomData *fdata,
mface_to_poly_map[mface_index] = poly_index; \
mf = &mface[mface_index]; \
lidx = lindices[mface_index]; \
/* set loop indices, transformed to vert indices later */ \
/* Set loop indices, transformed to vert indices later. */ \
l1 = mp_loopstart + 0; /* EXCEPTION */ \
l2 = mp_loopstart + 1; /* EXCEPTION */ \
l3 = mp_loopstart + 2; /* EXCEPTION */ \
@ -293,7 +294,7 @@ int BKE_mesh_tessface_calc_ex(CustomData *fdata,
zero_v3(normal);
/* calc normal, flipped: to get a positive 2d cross product */
/* Calculate the normal, flipped: to get a positive 2D cross product. */
ml = mloop + mp_loopstart;
co_prev = mvert[ml[mp_totloop - 1].v].co;
for (j = 0; j < mp_totloop; j++, ml++) {
@ -305,7 +306,7 @@ int BKE_mesh_tessface_calc_ex(CustomData *fdata,
normal[2] = 1.0f;
}
/* project verts to 2d */
/* Project verts to 2D. */
axis_dominant_v3_to_m3_negate(axis_mat, normal);
ml = mloop + mp_loopstart;
@ -315,7 +316,7 @@ int BKE_mesh_tessface_calc_ex(CustomData *fdata,
BLI_polyfill_calc_arena(projverts, mp_totloop, 1, tris, arena);
/* apply fill */
/* Apply fill. */
for (j = 0; j < totfilltri; j++) {
uint *tri = tris[j];
lidx = lindices[mface_index];
@ -323,7 +324,7 @@ int BKE_mesh_tessface_calc_ex(CustomData *fdata,
mface_to_poly_map[mface_index] = poly_index;
mf = &mface[mface_index];
/* set loop indices, transformed to vert indices later */
/* Set loop indices, transformed to vert indices later. */
l1 = mp_loopstart + tri[0];
l2 = mp_loopstart + tri[1];
l3 = mp_loopstart + tri[2];
@ -359,7 +360,7 @@ int BKE_mesh_tessface_calc_ex(CustomData *fdata,
BLI_assert(totface <= looptri_num);
/* not essential but without this we store over-alloc'd memory in the CustomData layers */
/* Not essential but without this we store over-allocated memory in the #CustomData layers. */
if (LIKELY(looptri_num != totface)) {
mface = MEM_reallocN(mface, sizeof(*mface) * (size_t)totface);
mface_to_poly_map = MEM_reallocN(mface_to_poly_map,
@ -368,14 +369,14 @@ int BKE_mesh_tessface_calc_ex(CustomData *fdata,
CustomData_add_layer(fdata, CD_MFACE, CD_ASSIGN, mface, totface);
/* CD_ORIGINDEX will contain an array of indices from tessfaces to the polygons
* they are directly tessellated from */
/* #CD_ORIGINDEX will contain an array of indices from tessellation-faces to the polygons
* they are directly tessellated from. */
CustomData_add_layer(fdata, CD_ORIGINDEX, CD_ASSIGN, mface_to_poly_map, totface);
CustomData_from_bmeshpoly(fdata, ldata, totface);
if (do_face_nor_copy) {
/* If polys have a normals layer, copying that to faces can help
* avoid the need to recalculate normals later */
* avoid the need to recalculate normals later. */
if (CustomData_has_layer(pdata, CD_NORMAL)) {
float(*pnors)[3] = CustomData_get_layer(pdata, CD_NORMAL);
float(*fnors)[3] = CustomData_add_layer(fdata, CD_NORMAL, CD_CALLOC, NULL, totface);
@ -387,13 +388,11 @@ int BKE_mesh_tessface_calc_ex(CustomData *fdata,
/* NOTE: quad detection issue - fourth vertidx vs fourth loopidx:
* Polygons take care of their loops ordering, hence not of their vertices ordering.
* Currently, our tfaces' fourth vertex index might be 0 even for a quad. However,
* we know our fourth loop index is never 0 for quads (because they are sorted for polygons,
* and our quads are still mere copies of their polygons).
* So we pass NULL as MFace pointer, and BKE_mesh_loops_to_tessdata
* will use the fourth loop index as quad test.
* ...
*/
* Currently, our tfaces' fourth vertex index might be 0 even for a quad.
* However, we know our fourth loop index is never 0 for quads
* (because they are sorted for polygons, and our quads are still mere copies of their polygons).
* So we pass NULL as MFace pointer, and #BKE_mesh_loops_to_tessdata
* will use the fourth loop index as quad test. */
BKE_mesh_loops_to_tessdata(fdata, ldata, NULL, mface_to_poly_map, lindices, totface);
/* NOTE: quad detection issue - fourth vertidx vs fourth loopidx:
@ -431,7 +430,7 @@ void BKE_mesh_tessface_calc(Mesh *mesh)
mesh->totface,
mesh->totloop,
mesh->totpoly,
/* calc normals right after, don't copy from polys here */
/* Calculate normals right after, don't copy from polys here. */
false);
BKE_mesh_update_customdata_pointers(mesh, true);
@ -441,6 +440,8 @@ void BKE_mesh_tessface_calc(Mesh *mesh)
/* -------------------------------------------------------------------- */
/** \name Loop Tessellation
*
* Fill in #MLoopTri data-structure.
* \{ */
/**

13
source/blender/draw/intern/draw_cache_extract_mesh_render_data.c
View File

@ -67,8 +67,8 @@ static void mesh_render_data_loose_geom_load(MeshRenderData *mr, MeshBufferExtra
static void mesh_render_data_loose_geom_ensure(const MeshRenderData *mr,
MeshBufferExtractionCache *cache)
{
/* Early exit: Are loose geometry already available. Only checking for loose verts as loose edges
* and verts are calculated at the same time.*/
/* Early exit: Are loose geometry already available.
* Only checking for loose verts as loose edges and verts are calculated at the same time. */
if (cache->loose_geom.verts) {
return;
}
@ -230,7 +230,7 @@ static void mesh_render_data_mat_offset_build(MeshRenderData *mr, MeshBufferExtr
typedef struct MatOffsetUserData {
MeshRenderData *mr;
/* struct is extended during allocation to hold mat_tri_len for each material. */
/** This struct is extended during allocation to hold mat_tri_len for each material. */
int mat_tri_len[0];
} MatOffsetUserData;
@ -252,7 +252,7 @@ static void mesh_render_data_mat_offset_build_threaded(MeshRenderData *mr,
int face_len,
TaskParallelRangeFunc range_func)
{
/* Extending the MatOffsetUserData with an int per material slot. */
/* Extending the #MatOffsetUserData with an int per material slot. */
size_t userdata_size = sizeof(MatOffsetUserData) +
(mr->mat_len) * sizeof(*cache->mat_offsets.tri);
MatOffsetUserData *userdata = MEM_callocN(userdata_size, __func__);
@ -348,6 +348,9 @@ void mesh_render_data_update_looptris(MeshRenderData *mr,
if (mr->extract_type != MR_EXTRACT_BMESH) {
/* Mesh */
if ((iter_type & MR_ITER_LOOPTRI) || (data_flag & MR_DATA_LOOPTRI)) {
/* NOTE(campbell): It's possible to skip allocating tessellation,
* the tessellation can be calculated as part of the iterator, see: P2188.
* The overall advantage is small (around 1%), so keep this as-is. */
mr->mlooptri = MEM_mallocN(sizeof(*mr->mlooptri) * mr->tri_len, "MR_DATATYPE_LOOPTRI");
if (mr->poly_normals != NULL) {
BKE_mesh_recalc_looptri_with_normals(me->mloop,
@ -575,7 +578,7 @@ void mesh_render_data_free(MeshRenderData *mr)
MEM_SAFE_FREE(mr->mlooptri);
MEM_SAFE_FREE(mr->loop_normals);
/* Loose geometry are owned by MeshBufferExtractionCache. */
/* Loose geometry are owned by #MeshBufferExtractionCache. */
mr->ledges = NULL;
mr->lverts = NULL;