deff795908
Rename some internal buffer enums to better match what they do
2021-01-17 20:59:16 -04:00
b7400892a4
Merge branch 'master' into 12.0-development
2020-12-22 19:35:06 -04:00
81272ba5b5
Only use pinned memory on AMD GPUs.
...
As per in Discord, for some reason Intel driver in Linux(?) expose
pinned memory extension and causes driver crash in one test case.
2020-12-15 10:25:38 +08:00
a0a933a1b9
Merge branch 'master' into 12.0
2020-02-15 13:51:12 -04:00
15f0e4109b
Fix a typo in a comment
2020-02-15 13:50:47 -04:00
813b1e0e69
Remove debug print
2020-02-05 18:03:34 -04:00
ca84b6db48
More robust error handling for AMD Pinned Memory buffers (issue #1540 )
...
Also print out any errors that are produced from that (for now).
2020-02-05 07:49:38 -04:00
51db32cfdf
Merge default into minor.
...
--HG--
branch : minor
2020-01-06 22:41:01 -04:00
eab3b83a2e
Fix a hard freeze on AMD+Windows systems when the internal buffers used for automatic batching are resized.
2020-01-06 19:58:22 -04:00
5f9b5a21d0
Merge default into minor
...
--HG--
branch : minor
2020-01-03 22:45:36 -04:00
fc4847c69d
Update copyright for the new year
2020-01-03 22:40:36 -04:00
bc12312f0f
macOS: rename most internal uses of "macosx" to "macos".
...
--HG--
branch : minor
2019-12-22 17:54:20 -04:00
8e5b71e8e7
Improve automatic batching performance slightly on AMD gpus by removing some redundant/useless opengl calls (there is no mapped buffer to flush).
2019-06-01 15:58:42 -03:00
ddef70e141
Potentially improve automatic batching performance slightly on nvidia/intel GPUs.
2019-06-01 15:06:19 -03:00
0752f27bdf
Update copyright year for 2019
2019-01-03 21:09:05 -04:00
3af66da51e
Automatic batching: improve the consistency of performance and draw calls for the vertex buffer upload approach taken on older Windows and Linux GPUs which are only capable of GL3.
2018-12-27 23:52:54 -04:00
43ee232897
macOS: Use a (non-persistent) map/sync approach for stream buffers, instead of glBufferSubData. This usually won't affect performance either way, but avoids some worst-case performance cliffs on my Intel GPU in my testing.
2018-12-16 15:15:32 -04:00
bcc5e77632
Simplify OpenGL 3+ stream buffer syncing code used with automatic batching.
2018-12-12 17:51:03 -04:00
535be96f03
Optimize StreamBuffer::getUsableSize to not be a virtual method.
2018-09-29 14:50:49 -03:00
3857d09132
Don't use persistently mapped buffers on Intel Ivy Bridge GPUs on Windows. glClientWaitSync can hang with those drivers, apparently.
2018-01-06 00:42:16 -04:00
e80247c191
Happy new year! ✨
2018-01-03 19:47:35 +01:00
8672ff2272
Use a more efficient method of streaming automatically batched vertices on AMD Windows and Linux systems.
...
--HG--
branch : minor
2017-11-11 20:48:02 -04:00
b4f8f08b70
Enable (when available) and improve performance of persistently mapped buffer path for the vertex/index buffers used with automatic batching.
...
--HG--
branch : minor
2017-09-03 18:23:54 -03:00
f2597168a2
Fix text object drawing and a bug in stream buffers used in autobatching
...
--HG--
branch : minor
2017-01-30 22:11:26 -04:00
7d7f8e0a92
Mark classes as ‘final’
...
--HG--
branch : minor
2017-01-29 02:06:17 -04:00
13af754d3d
Move most of Text and Video love.graphics classes out of the opengl subfolder.
...
--HG--
branch : minor
2017-01-28 21:37:09 -04:00
731b5d5cf0
Added support for GLSL 3 shaders (GLSL 3.30 and GLSL ES 3.00).
...
To use GLSL 3 in a shader, the first line of the shader has to be: #pragma language glsl3
glsl1 is the default language.
Added “glsl3” graphics feature as part of the table returned by love.graphics.getSupported().
Added love.graphics.validateShader(boolean gles, shadercode). It returns a boolean along with an error string if the shader code has errors for its target language and platform (gles or desktop).
Implemented support for Core Profile OpenGL 3.3 in love.graphics.
--HG--
branch : minor
2017-01-22 18:54:04 -04:00
Alex Szpakowski
Miku AuahDark
Bart van Strien