Hajime Hoshi
5afcc65387
shareable: Add comments
2019-04-11 18:15:56 -07:00
Hajime Hoshi
b1b8d0b4b1
graphics: Move texel-region adjustment from shaders to program
...
This reduces GPU burden.
2019-02-16 14:50:03 +09:00
Hajime Hoshi
3f28592700
graphics: Refactoring: Remove PutVertex
2019-02-15 00:22:32 +09:00
Hajime Hoshi
7c506bc5bc
restorable: Refactoring: Add (*Image).QuadVertices/PutVertex
2019-02-15 00:18:12 +09:00
Hajime Hoshi
fd250c8d8c
graphics: Rename NextPowerOf2Int -> InternalImageSize and add image size adjustment
...
There is a minimum internal image size on some system like old iOS
devices. This change adds adjustment of the size.
Issue: #810
2019-02-14 23:06:01 +09:00
Hajime Hoshi
a3eddeb50f
Improve panic messages
2019-02-06 23:43:03 -10:00
Hajime Hoshi
78ed824351
graphics: Make verticesBackend concurrent safe
...
Fixes #789
2019-01-19 02:06:50 +09:00
Hajime Hoshi
189b8a17e9
graphics: Use source-border check even at DrawTriangles
...
DrawTriangles can now take a sub-image.
2018-12-23 03:24:39 +09:00
Hajime Hoshi
4149a56524
graphics: Refactoring: Use 12 floats for each vertex
...
This is a preparation for #761 .
2018-12-23 02:36:28 +09:00
Hajime Hoshi
5f8a0e34bb
graphics: isPowerOf2(0 or negative) should return false
2018-12-16 14:26:16 +09:00
Hajime Hoshi
155ee828b8
Fix misspelling
2018-12-03 18:23:25 +01:00
Hajime Hoshi
31f15bc5ad
Revert "graphics: Change the number of floats for a vertex from 10 to 12"
...
This reverts commit 7586c660d5
2018-11-18 00:07:15 +09:00
Hajime Hoshi
7586c660d5
graphics: Change the number of floats for a vertex from 10 to 12
...
This is a preparation for Metal, that forces 4 floats for a vertex
position.
2018-11-17 15:30:11 +09:00
Hajime Hoshi
09c8516545
graphcis: Move IndicesNum from opengl to graphics
2018-11-07 01:49:45 +09:00
Hajime Hoshi
c9f49efd41
graphics: Add VertexFlaotNum
2018-11-06 02:34:52 +09:00
seebs
74e204d952
graphics: Speed up DrawTriangles ( #723 )
...
DrawTriangles is expensive and slow because of massive memory
allocation and garbage collection costs. This patch moves from ~47TPS
on my laptop (with ~24k triangles) to 60TPS. The first part
is just allocating the right size of vertex buffer up front; that
got to about 55TPS. The second part replaces the frequent
allocations of []float32 in Vertex() calls with writing the
desired values into a provided destination slice.
Time spent in drawing triangles for 1,000 frames:
13.07s baseline
11.09s preallocate whole buffer to avoid resizing
6.13s use new PutVertex function
This might need some cleanup, but I think it's good evidence that
the design change is viable.
2018-10-31 10:53:17 +09:00
Hajime Hoshi
ce1c616f69
Merge graphics and graphicsutil
2018-10-28 23:03:06 +09:00
