DwM: optimize mesh batch conversion
authorCampbell Barton <ideasman42@gmail.com>
Thu, 29 Jun 2017 10:11:16 +0000 (20:11 +1000)
committerCampbell Barton <ideasman42@gmail.com>
Thu, 29 Jun 2017 10:11:16 +0000 (20:11 +1000)
commitfd3589e4c969d2e08b864e0e60c8c25f23e50145
tree9be9f85d2de50f32dd736ded6a6bc15cfb520f4a
parent34566aa96910e29eceb45d29c9d50471452f8d73
DwM: optimize mesh batch conversion

- Replace GWN_vertbuf_attr_set with Gwn_VertBufRaw & GWN_vertbuf_raw_step
  to avoid intermediate copy.
- Avoid extra conversion step with: float[3] -> short[3] -> Gwn_PackedNormal.
  We can skip the short[3].

Gives approx 6% speedup here.
source/blender/draw/intern/draw_cache_impl_mesh.c