This removes the limitation of the sculpt overlays not being visible
with modifiers active.
Details
Details
Diff Detail
Diff Detail
- Repository
- rB Blender
- Branch
- sculpt-mask-modifiers (branched from master)
- Build Status
Buildable 9732 Build 9732: arc lint + arc unit
Event Timeline
| source/blender/blenkernel/BKE_paint.h | ||
|---|---|---|
| 207 ↗ | (On Diff #28060) | Style: Use uchar r_color[4] to show how many element are needed. |
| source/blender/draw/intern/draw_cache_extract.h | ||
| 107 | I don't really like the name overlays. Prefer sculpt_data. Keep sculpt_overlay for the batch. | |
| source/blender/draw/intern/draw_cache_extract_mesh.c | ||
| 2579 | You are falling into a pitfall not guarded by any assert (this is my fault). Using the vbo->data (as in gpuSculptOverlay *vbo_data = (gpuSculptOverlay *)vbo->data;) buffer dirrectly is dangerous and needs to be done ONLY if the format has no padding. But yours have. See attr_align(const GPUVertAttr *a) definition for more information. | |
| 2655 | Can't this use threading? I don't see any data dependency here. | |
| source/blender/blenkernel/BKE_paint.h | ||
|---|---|---|
| 207 ↗ | (On Diff #29008) | Declaration is not updated. |
| source/blender/draw/intern/draw_cache_extract_mesh.c | ||
| 2655 | What I meant here, is to make use of the range extract functions. This is not mandatory but if you don't do it, add a TODO comment so someone can pick it up. (and don't set use_threading = true if you are not actually doing it) | |