d994e56b60
Changes/cleanup to TextureCache::Load and other mipmap related code.
...
The significant change is what is now line 520 of TextureCacheBase.cpp:
((std::max(mipWidth, bsw) * std::max(mipHeight, bsh) * bsdepth) >> 1)
to
TexDecoder_GetTextureSizeInBytes(expanded_mip_width, expanded_mip_height, texformat);
Fixes issue 5328.
Fixes issue 5461.
2013-02-15 22:56:29 -06:00
e259343275
fix efb copy shaders
2013-02-15 13:13:45 +01:00
9f4a616f2e
build fix
2013-02-14 19:04:52 +01:00
a0ef58418b
workaround for buggy intel windows driver. revert this as soon as there are newer driver
2013-02-14 18:59:45 +01:00
0325e37bfb
merge glsl headers into one place
2013-02-13 21:34:48 +01:00
3392562501
rewrite efb copy shader, copied from dx11
2013-02-13 18:01:06 +01:00
21b83b436c
create shader cache directoy
2013-02-13 16:50:56 +01:00
b3675d15dc
enable shader cache again
2013-02-13 16:30:15 +01:00
398b37f371
fix ValidateShaderIDs
2013-02-13 15:16:32 +01:00
bbc292c210
merge Vertex and PixelShaderCache into ProgramShaderCache
...
this is the first step, uniform handling is still missing
2013-02-13 13:12:19 +01:00
ed0abc9dc5
Merge branch 'mipmap_fixes'.
2013-02-07 20:40:33 +01:00
6870c1fdd5
Fix some potential issues when blending on EFB formats without alpha.
...
Clean up state transition tables.
2013-02-07 18:01:24 +01:00
f079b04a5d
also check for glsl errors on linking, activation GLSL_DEBUG again (slowdown on shader compilation)
2013-02-07 12:47:41 +01:00
9987765b1f
build fix
2013-02-07 10:48:42 +01:00
376a807dea
first try of osx fix, mostly by pauldachz
2013-02-07 10:34:29 +01:00
e7a8d3f854
check for GL_ARB_sync, it's in ogl 3.2
2013-02-06 01:05:19 +01:00
ef602fd643
workaround for glGetUniformLocation in efb2ram
2013-02-05 18:22:23 +01:00
76adc77fa6
bigger buffers
2013-02-05 18:01:27 +01:00
1d86dafbe2
evil hack: map the buffer once and use this pointer after unmaping
...
This isn't allowed by gl and may _only_ work, if the driver uses cpu bound buffer.
But it may speed up :-)
2013-02-05 00:53:14 +01:00
f677b8cba3
update to glew1.9, win binaries also from official release
2013-02-04 21:11:12 +01:00
2108529120
also align the size of buffers for pinned memory
2013-02-04 20:48:28 +01:00
6e0b0192a8
correct spelling of align
2013-02-02 19:39:26 +01:00
0c7aa4ff4f
build fix
2013-02-02 19:32:40 +01:00
6e20af9ee9
align memory, use pinned memory if available
2013-02-02 19:26:29 +01:00
f0795bad12
fix glew fix
2013-02-01 17:12:05 +01:00
cb4cb1713d
workaround older glew versions
2013-02-01 17:09:50 +01:00
2a1f592d82
first try for pinned memory, complete untested
2013-02-01 17:04:27 +01:00
3bf4ffe7fa
usable sync of ringbuffer
2013-02-01 16:43:08 +01:00
3af9840a4c
stream by map and sync
...
but not working perfectly, so disabled
2013-02-01 15:15:25 +01:00
878bd7f26c
implement streaming by bufferSubData, split upload and allocation in ringbuffer
2013-02-01 12:30:08 +01:00
30170575c8
create StreamBuffer class for ogl upload
2013-01-31 23:11:53 +01:00
011e326698
remove glIsEnabled(GL_BLEND), we do restore state after it, so it doesn't matter
2013-01-31 13:49:20 +01:00
4a463f4588
also add glBufferSubData for ubo upload
...
so osx start game, but everything is black
2013-01-31 13:00:15 +01:00
c05f66bdd0
first try for osx core profile
...
now it doesn't crash, but it still hangs on startup
2013-01-31 12:20:17 +01:00
01d8c21e1d
don't stream in rasterfont
...
mapping blocks on nvidia workstation, so use glBufferData
2013-01-31 11:30:44 +01:00
0e04e0c305
Fix some shadowing warnings.
2013-01-29 16:43:22 -06:00
afb5be10d9
OSX likes to put unordered_map in the weirdest of places...
2013-01-28 15:36:55 -06:00
fa45403557
Got to love that my server's cmake is half retarded.
2013-01-28 15:32:38 -06:00
d94f3c4155
Stop using std::pair and std::map. Switch over to u64 and std::unordered_map. Provides a very small speed boost.
2013-01-28 15:18:54 -06:00
0fdeb81038
Add some code for generating a shader constant usage profile.
2013-01-28 22:15:23 +01:00
67be1e939a
PixelShaderGen: Some more work...
2013-01-28 21:59:49 +01:00
dc0f470215
Added new shader cache uids for pixel shader gen.
2013-01-28 21:44:39 +01:00
3c8df842bb
Moved some of the new shader uid stuff to a common header file.
2013-01-28 21:05:29 +01:00
ca0e292dd4
Replace the shader uid system with a new one which quasi-automatically generates uids for shaders.
...
Currently used in the vertex shader only (had to fork lighting shaders for now).
2013-01-28 21:05:29 +01:00
c5fa3e0f3d
move RestoreAPIState and ResetAPIState into backends
...
it's a backend specific hack, so it should be here
should give a small speedup in dx11 efb2tex
2013-01-28 18:16:03 +01:00
fd06342a97
set hint GL_STREAM_READ
...
it's wrong, but so we are guaranteed to get pinned memory.
it's slower for rendering, but faster for mapping.
2013-01-28 13:03:31 +01:00
4fadb65259
Only check shader compilation status in debug or with DEBUG_GLSL defined. invalidate range on bindbufferrange since we don't care about what is in the range. Both give a performance boost for me.
2013-01-26 11:24:24 -06:00
3a52b37c3b
only create shaders after shadercaches are initialized
2013-01-26 16:33:54 +01:00
13e9bb56f8
fix bug when rasterfont disappears sometimes
2013-01-25 20:39:19 +01:00
41b1128fdd
orphan vbo also with glBufferData
2013-01-25 13:28:05 +01:00
a526fc5f92
ubo streaming
2013-01-25 13:20:42 +01:00
73eb98ed8e
Move swap control to the host specific GLInterface files.
2013-01-24 10:31:08 -06:00
e0ffdda26e
Merge branch 'immediate-removal' into GLSL-master
...
Conflicts:
Source/Core/VideoCommon/Src/PixelShaderGen.cpp
Source/Plugins/Plugin_VideoSoftware/Src/SWRenderer.cpp
immediate-removal is a new created branch seperated from master but reverted the revert of immediate-removal
so we get less conflicts by merging
2013-01-24 16:58:28 +01:00
d60cc373d1
Revert "Revert 30dd9c2 e9d00bf db5f4c8 and bff0fae"
...
This reverts commit d0301ca89d
2013-01-24 16:11:07 +01:00
6afc30240a
partial revert of 8a6f747408
...
texture_rect isn't in core 3.0, so the long texture function must be used
2013-01-22 00:18:42 +01:00
da0ec6618f
always init ogl configs
2013-01-21 18:44:33 +01:00
cf11f0ea72
clear uniform cache on startup
2013-01-21 15:43:04 +01:00
ebaaf6d468
glShadeModel is also obsolete
2013-01-21 10:37:16 +01:00
8a6f747408
glsl: remove usage of old texture2D* function
2013-01-19 11:07:06 +01:00
d73c950623
cache texture bindings
2013-01-19 01:06:34 +01:00
e7d5b274c0
add stage parameter for texture load, so ogl can bind to the correct sampler
2013-01-19 00:47:48 +01:00
074f73c641
move utils texture to sampler 8+9
...
rasterfont tex will aways bound to samp8.
efb copy utils will be done in samp9.
2013-01-19 00:39:31 +01:00
6401a18143
remove texture sampler from uniform list
2013-01-19 00:12:02 +01:00
2c010f530e
report glsl shader warings as errors
2013-01-18 12:49:48 +01:00
714ff50fdf
set blending if dual source might be triggered
2013-01-18 00:44:35 +01:00
a87fd89fce
disable ogl shader cache
...
it doesn't check the shader itself only the id of the shaders.
So there are cache mismatch atm
2013-01-17 12:36:20 +01:00
3054d36534
use glBindFragDataLocation instead of glDrawBuffer
2013-01-17 11:16:12 +01:00
0ef3bd9c77
Revert "Made several variables/parameters unsigned in the DX9, DX11 and OGL plugins. They make more sense like this (given their names)."
...
Turns out I was wrong in my previous commit. My bad.
This reverts commit 8743166663
2013-01-16 15:46:11 -05:00
b778b86e26
improve texture mipmap count
...
GL_TEXTURE_MAX_LEVEL set how many mipmaps should be allocated, but all of them must be created.
GL_TEXTURE_MAX_LOD set how many mipmaps may be used.
2013-01-16 21:15:07 +01:00
8a7f1e9264
workaround for older ati drivers
2013-01-16 19:37:57 +01:00
8743166663
Made several variables/parameters unsigned in the DX9, DX11 and OGL plugins. They make more sense like this (given their names).
...
This also gets rid of some more typecasts in some cases.
2013-01-16 09:42:51 -05:00
b401131af2
small texcache cleanup
2013-01-16 15:12:39 +01:00
681272d65d
use renderbuf fo xfb
2013-01-16 01:37:00 +01:00
bb200acdd8
also don't search in the shader cache if the shader isn't changed
2013-01-15 23:42:24 +01:00
687d74e6c2
only change shader if needed
2013-01-15 14:22:40 +01:00
85200db6a2
Revert "fix byte order in osd rasterfont"
...
This reverts commit db5f4c8c20
2013-01-15 12:25:06 +01:00
1cfb5a21fa
only use rgba for opengl
2013-01-14 23:43:51 +01:00
ff889c0e65
use attrib pointers in nativeVertexFormat
2013-01-14 22:59:08 +01:00
2f78986e2c
Merge branch 'Graphic_Update' into GLSL-master
...
Conflicts:
Source/Core/VideoCommon/Src/VertexManagerBase.cpp
Source/Plugins/Plugin_VideoOGL/Src/NativeVertexFormat.cpp
Source/Plugins/Plugin_VideoOGL/Src/Render.cpp
Source/Plugins/Plugin_VideoOGL/Src/VertexManager.cpp
2013-01-14 21:36:31 +01:00
ff5ac8c505
fix win compiler errors
2013-01-14 20:08:53 +01:00
5fe3def64c
videoConfig cleanup
2013-01-14 20:00:33 +01:00
adeac6f4a5
remove old deprecated calls and fixed functions parameters
2013-01-14 17:48:38 +01:00
c3aafc77b3
upload complete uniform buffer at once
...
this is the way of dx11. it would upload more per draw, but uses less calls.
will be faster if many uniforms are changed, but slower else
2013-01-14 13:58:11 +01:00
b18c0a5d1b
remove ubo for efb2ram
2013-01-14 12:37:31 +01:00
f3d663c2f4
efb2tex without ubo
2013-01-12 16:29:16 +01:00
2838077313
fix warnings
2013-01-11 21:24:59 +01:00
8651b34e22
also add vertex shader for efb2tex
...
goodbye fixed-function shaders
2013-01-11 15:41:42 +01:00
7c01c2bf11
move efb2ram shaders to textureCache
2013-01-11 15:24:43 +01:00
60b8e4fb1c
vertex shader for texture converter
2013-01-11 11:59:42 +01:00
092a99cbd2
glsl shaders for showEfbCopyRegions
2013-01-09 20:08:55 +01:00
bff02b3b73
Merge branch 'arb_framebuffer' into GLSL-master
...
Conflicts:
Source/Plugins/Plugin_VideoOGL/Src/FramebufferManager.cpp
Source/Plugins/Plugin_VideoOGL/Src/RasterFont.cpp
Source/Plugins/Plugin_VideoOGL/Src/Render.cpp
Source/Plugins/Plugin_VideoOGL/Src/TextureCache.cpp
Source/Plugins/Plugin_VideoOGL/Src/TextureConverter.cpp
2013-01-09 11:00:26 +01:00
84debc17d5
fix msaa
...
blitting between msaa fbos may not scale. So we need to resolve the efb before swapping.
2013-01-09 01:50:52 +01:00
be706a3977
Disable color writing when alpha test always fails.
2013-01-08 18:56:01 +01:00
d26bcb0847
Move alpha pretest to BPMemory.h and rename a bunch of alpha testing related stuff
2013-01-08 18:56:01 +01:00
e979b2d4a2
DX11: Fix a small mistake.
...
Remove some incorrect and/or confusing comments.
Fixes issue 5251. Thanks to limburgerite :)
2013-01-08 18:16:48 +01:00
d0301ca89d
Revert 30dd9c2 e9d00bf db5f4c8 and bff0fae
2013-01-07 13:47:34 -06:00
b7906b49c9
fix byte order in osd rasterfont
2013-01-07 20:11:26 +01:00
ef32231431
fix screenshots
2013-01-03 21:40:54 +01:00
fc02427d54
xfb draw by blit - disable postprocessing
...
blitting can't do postprocessing - so for postprocessing, we have to add a new stage
2013-01-03 20:44:47 +01:00
b67b1c376d
fix virtual xfb - remove glCopyTexImage2D
...
glCopyTexImage2D isn't in core (and so not supported everywhere) and it cannot store into GL_TEXTURE_RECTANGLE
2013-01-03 18:36:19 +01:00
363ffd3a5f
swap() by blitting
2013-01-03 12:33:18 +01:00
ecedf56eb4
uses the ARB_framebuffer_object syntax
...
also require this extention (OGL3.0), but it have one _realy_ big advantage:
- now it's possible to blit between different texture sizes, so all util draw calls can be implemented as blit
2013-01-03 12:06:47 +01:00
a1ca288555
fix some AMD issues
...
This is only a hack. The correct solution would be:
1. don't use GL_TEXTURE_RECT ever. GL_TEXTURE_2D should also be faster
2. use GL_TEXTURE8+ for util textures
3. both
2013-01-03 09:45:12 +01:00
b38b62afc6
remove glsl binding support. convert every shader to version 130
2013-01-02 16:56:08 +01:00
c2898748ac
bind locations _before_ link program, fix dest alpha
2012-12-31 02:34:27 +01:00
30dd9c2e17
always calls glBindBuffer(0) after disabling vao
2012-12-29 12:50:42 +01:00
e9d00bfeeb
revert RasterFont optimization
...
isn't needed and also not tested and buggy
2012-12-29 00:37:50 +01:00
d9f3fc4495
fix ubo for intel cards (also amd?)
2012-12-28 17:06:08 +01:00
48ede4dd30
switch to glsl 130, more than OpenGL 3.1 isn't needed
2012-12-28 16:05:14 +01:00
70c63ce6cf
fix dual pass alpha
2012-12-28 14:24:12 +01:00
82b33471ae
cleanup rasterfont
2012-12-28 11:46:00 +01:00
7d93834cd8
Bit more cleanup from removing Nvidia CG
2012-12-27 22:53:07 -06:00
d7f43c7a68
remove cg from windows project files
2012-12-28 01:08:29 +01:00
193056493a
also use shaderCaches in rasterFont
2012-12-28 00:52:44 +01:00
316a33d1e6
Merge branch 'master' into GLSL-master
...
Conflicts:
Source/Core/DolphinWX/Src/VideoConfigDiag.h
Source/Plugins/Plugin_VideoOGL/Src/GLUtil.h
Source/Plugins/Plugin_VideoOGL/Src/Render.cpp
Source/Plugins/Plugin_VideoOGL/Src/TextureCache.cpp
Source/Plugins/Plugin_VideoOGL/Src/TextureConverter.cpp
2012-12-27 10:36:54 +01:00
db5f4c8c20
fix byte order in osd rasterfont
2012-12-26 23:48:33 +01:00
01cdcc83a3
Merge Fail: Make sure we are using the interface file to get backbuffer dimensions. GL helper function was duplicated.
2012-12-26 13:15:19 -06:00
67c1c6b062
Merge branch 'GLES-software'
2012-12-26 12:54:58 -06:00
e2489196e4
Got overzealous. We aren't removing CG quite yet with this branch.
2012-12-26 00:41:25 -06:00
9a58f9a3bb
Rename another function...
2012-12-26 00:34:09 -06:00
f4e7a280a9
Remove cg.lib from some files for Windows and renaming a function.
2012-12-26 00:07:43 -06:00
01953ff64a
Rename the interface files and hopefully fix OSX building.
2012-12-25 19:08:24 -06:00
bff0faea2d
Merge branch 'immediate-removal'
...
Conflicts:
Source/Plugins/Plugin_VideoOGL/Src/Render.cpp
2012-12-25 12:39:00 +01:00
399a6af66d
revert update the rasterfont shader to GLSL3.3 - OGL 2.0 should be enough
2012-12-25 12:22:02 +01:00
2fedab3bad
OGL: Make OSD messages fade out properly instead of having them disappear spontaneously.
2012-12-24 19:50:16 +01:00
9209253e0d
Initial removal of Nvidia CG. Still some more cleanup to go
2012-12-24 11:09:52 -06:00
296b9b1c16
Small bit of clean up in GLES-software.
2012-12-24 09:37:12 -06:00
85ac76e8e2
fix gcc compilation
2012-12-19 23:42:22 +01:00
b78f5debe6
Initial push of GLES and GLUtil file breakup.
2012-12-17 14:54:20 -06:00
1919a458e8
only use one buffer, orphaning should do the rest
2012-12-15 17:28:58 +01:00
ba8264c2ac
use VAO in VertexManager
...
to use VAO, we must use VBO, so some legency code was removed:
- ARB_map_buffer_range must be available (OGL 3.0), don't call glBufferSubData if not
- ARB_draw_elements_base_vertex also (OGL 3.2), else we have to set the pointers every time
- USE_JIT was removed, it was broken and it isn't needed any more
And the index and vertex buffers are now synchronized, so that there will be one VAO per
NativeVertexFormat and Buffer.
2012-12-15 14:43:01 +01:00
79a7ce4827
move glBindBuffer and glBindVertexArray out of VertexManager
...
Signed-off-by: Ryan Houdek <Sonicadvance1@gmail.com>
2012-12-13 15:28:52 -06:00
09274e2483
Check texture params before updateing them
...
Signed-off-by: Ryan Houdek <Sonicadvance1@gmail.com>
2012-12-13 15:28:48 -06:00
61836f8c51
helper for compiling glsl
...
Signed-off-by: Ryan Houdek <Sonicadvance1@gmail.com>
2012-12-13 15:28:44 -06:00
48812f7bf7
update the rasterfont shader to GLSL3.3 - so OGL 3.3 is needed
...
Signed-off-by: Ryan Houdek <Sonicadvance1@gmail.com>
2012-12-13 15:28:39 -06:00
1bd21f44b2
custom shader for RasterFont, fix color support
...
OpenGL2.0 compatible
Signed-off-by: Ryan Houdek <Sonicadvance1@gmail.com>
2012-12-13 15:28:36 -06:00
23a3336f9a
increase hash size to u64
...
Signed-off-by: Ryan Houdek <Sonicadvance1@gmail.com>
2012-12-13 15:28:32 -06:00
75dd48247f
initial release of new RasterFont without color support
...
this new design will once create a texture for all chars.
while rendering a string, a list of polygons (position on screen + texture)
for this string is generated on the fly and print at once by glDrawArrays.
atm, there is no support for colors, so everything will display white.
Signed-off-by: Ryan Houdek <Sonicadvance1@gmail.com>
2012-12-13 15:28:28 -06:00
c207422987
using of vao, warning: ARB_vertex_array_object is needed
...
Signed-off-by: Ryan Houdek <Sonicadvance1@gmail.com>
2012-12-13 15:28:25 -06:00
5c8800968a
also check for vbo updates in EncodeToRamUsingShader
...
Signed-off-by: Ryan Houdek <Sonicadvance1@gmail.com>
2012-12-13 15:28:22 -06:00
1f1b32b663
only update vbo on changes in XFBSource::Draw
...
Signed-off-by: Ryan Houdek <Sonicadvance1@gmail.com>
2012-12-13 15:28:19 -06:00
34b1451fbe
cache vbos in TextureCache::TCacheEntry::FromRenderTarget
...
Signed-off-by: Ryan Houdek <Sonicadvance1@gmail.com>
2012-12-13 15:28:15 -06:00
d0c4332d99
don't update vbo, if there are no changes in Renderer::Swap
...
Signed-off-by: Ryan Houdek <Sonicadvance1@gmail.com>
2012-12-13 15:28:13 -06:00
bbcb442983
in the end, also Renderer::Swap in vbo
...
two last draw-calls are missing:
VertexManager::Draw (see Graphic_Update branch)
RasterFont::printString (perhaps reimplement with an texture)
Signed-off-by: Ryan Houdek <Sonicadvance1@gmail.com>
2012-12-13 15:28:10 -06:00
b41c06a9ac
EncodeToRamUsingShader in vbo
...
Signed-off-by: Ryan Houdek <Sonicadvance1@gmail.com>
2012-12-13 15:28:05 -06:00
8ea2ddbc50
XFBSource::Draw in vbo
...
Signed-off-by: Ryan Houdek <Sonicadvance1@gmail.com>
2012-12-13 15:27:59 -06:00
d44228f1b7
TextureCache::TCacheEntry::FromRenderTarget as vbo
...
Signed-off-by: Ryan Houdek <Sonicadvance1@gmail.com>
2012-12-13 15:27:56 -06:00
6b3125728b
use vbo for ShowEFBCopyRegions
...
Signed-off-by: Ryan Houdek <Sonicadvance1@gmail.com>
2012-12-13 15:27:53 -06:00
8b62be2025
fix ShowEFBCopyRegions
...
those fancy colors were the result of the usage of non-allocated heap
Signed-off-by: Ryan Houdek <Sonicadvance1@gmail.com>
2012-12-13 15:27:49 -06:00
6864b40e26
reset glEnableClientState befor every draw
...
should be done with VAO, but atm, this is not possible :-(
this also partial revert the fix in fb92c338af
2012-12-13 15:27:46 -06:00
888b5fb061
remove usage of glMultiTexCoord2f
...
Signed-off-by: Ryan Houdek <Sonicadvance1@gmail.com>
2012-12-13 15:27:38 -06:00
30718230b2
Fix some warnings.
2012-12-06 20:25:07 -05:00
97f5b1665f
what is this even...
2012-11-26 02:11:52 -05:00
4ff9e03509
Merge branch 'efb_scaling_fixes'.
2012-11-19 13:18:57 +01:00
a38bb488d2
Remove Renderer::xScale and Renderer::yScale.
2012-11-16 14:29:09 +01:00
78031c2d54
Move ComputeDrawRectangle() to Renderer::UpdateDrawRectangle().
2012-11-16 14:25:49 +01:00
14aa7150d9
Add option for author name for movies. Also, minor cleanup for previous options i've added.
2012-11-14 16:23:20 -05:00
3936c06ee8
more fixes for my last commit, player problem in twin snakes is fixed
2012-11-11 22:39:27 -03:00
a5d210129d
Add an on screen lag counter.
2012-11-11 17:57:06 -05:00
53b62ab169
This should fix the bugs introduced by my last commit please retest the games that where showing graphical glitches.
2012-11-10 11:37:08 -03:00
fb92c338af
Bit of cleanup and fixing of one issue that was noticeable in SMS with Mario's shadow.
2012-10-27 22:50:06 -05:00
dfb3c44d1a
Remove glMatrixMode and glLoadIdentity calls in Render.cpp which do absolutely nothing.
2012-10-27 18:27:16 -05:00
8c1d104ca3
Change EFB copy location rectangles over to a Vertex array
2012-10-27 18:10:31 -05:00
c1fd640964
Rebase immediate-removal on master
2012-10-27 17:22:48 -05:00
ee72852491
implement some code to reduce the amounts of calls to setup vertex format, in d3d9 it gives no noticeable speedup, in opengl it still does not work right.
...
thanks to neobrain for the idea
2012-10-26 23:18:09 -03:00
eaa1ea71c1
Implement the new buffer approach in opengl. sadly in my machine it gives my only 2 more fps and if your hardware does not support ARB_map_buffer_range is even slower than plain vertex arrays.
...
change naming in all the backends vertex managers to make more easy to continue with the merge an some future improvements.
please test this as i'm interested in knowing the performance in linux and windows with the different hardware platforms.
2012-10-26 11:34:02 -03:00
52963584f0
[bugfix] Make sure to specify the correct max lod level.
2012-10-20 21:07:03 +02:00
9d5fc3d7f0
[bugfix] Properly decode lod_bias.
2012-10-20 21:07:03 +02:00
0706050b74
[cleanup] Vastly clean up backend-specific sampler state logic.
2012-10-20 21:07:02 +02:00
069d949a57
[cleanup] TextureCache: Kill the autogen_mipmap parameter
2012-10-20 21:07:02 +02:00
5230146c73
Hey, long time no commits :).
...
So to compensate lets bring back some speed to the emulation.
change a little the way the vertex are send to the gpu,
This first implementation changes dx9 a lot and dx11 a little to increase the parallelism between the cpu and gpu.
ogl: is my next step in ogl is a little more trickier so i have to take a little more time.
the original concept is Marcos idea, with my little touch to make it even more faster.
what to look for: SPEEEEEDDD :).
please test it a lot and let me know if you see any problem.
in dx9 the code is prepared to fall back to the previous implementation if your card does not support the amount of buffers needed.
So if you did not experience any speed gains you know where is the problem :).
for the ones with more experience and compression of the code please test changing the amount and size of the buffers to tune this for your specific machine.
The current values are the sweet spot for my machine.
All must Thanks Marcos, I hate him for giving good ideas when I'm full of work.
2012-10-20 10:22:15 -03:00
57426ee726
Forced commit
2012-10-13 13:00:04 -05:00
1a8005d948
set ProgramShaderCache program format correctly.
2012-10-09 23:56:00 -05:00
eb7a0c485a
More "stuff" for SS, also a small fix in the program shader cache cache setup."
2012-10-09 23:56:00 -05:00
210ecad15f
let us try normal c++ static init instead...
2012-10-09 23:56:00 -05:00
03b09bed5d
Get the program binary type correctly or else ATI makes massive (~400MB) shader caches. Also, don't need the line in the PixelShaderGen.
2012-10-09 23:56:00 -05:00
2e15440896
Add support for Dual source blending to older ATI cards that don't support 420pack but do support GL_ARB_blend_func_extended. This is more proper as well anyways.
2012-10-09 23:56:00 -05:00
5085cebaf3
These checks aren't needed when using CG.
2012-10-09 23:55:59 -05:00
108722bed0
oops, forgot to make sure the gl program is actually free'd.
2012-10-09 23:54:18 -05:00
e3854ded73
Woops, better not forget the ing
2012-10-09 23:54:18 -05:00
4cd748bbec
Remove some warnings in ProgramShadercache, Was using wrong variable for checking dual source blending.
2012-10-09 23:54:18 -05:00
4c1fef8f98
refactor ProgramShaderCache::PCacheEntry
2012-10-09 23:54:18 -05:00
aaa405c973
Checking GLEW_VERSION_4_0 is superfluous since we check GL_ARB_get_program_binary, and it's a runtime variable anyways.
2012-10-09 23:54:18 -05:00
31a8424bcc
fix formatting uglies introduced in glsl-master branch
2012-10-09 23:54:17 -05:00
c4e7a288e5
Add in the Windows fix.
2012-10-09 23:43:22 -05:00
ee529b7125
Only use explicit attribute locations when we are supporting GLSL > 1.2 since we need in/out instead of attribute variable types. This was brought to my attention from MESA. MESA supports GL_ARB_explicit_attrib_location, but yets to support GLSL 1.3, so basically useless extension to MESA right now?
2012-10-09 23:43:22 -05:00
5c486587a6
Write all shaders to disk on emulator stop instead of constantly. Also change pair from u64 to u32.
2012-10-09 23:43:22 -05:00
4eb227bd4c
If user doesn't compile with glew 1.6, cut out program binaries so everything still compiles.
2012-10-09 23:43:22 -05:00
189d12b61c
If CG isn't available, still compile and fallback on GLSL
2012-10-09 23:43:22 -05:00
7c91476650
Missed a if, don't want people crashing now do we?
2012-10-09 23:42:41 -05:00
9996f27120
Give OSX users more of a chance of supporting Single pass DSB in the future.
2012-10-09 23:42:41 -05:00
d012c75005
Implement Program shaders cache. Seems to reduce a small amount of stuttering when F-Zero starts. Did it because I can :|
2012-10-09 23:42:41 -05:00
ef1e157786
Go back to using glGetString with GL_EXTENSIONS because glew 1.7 is required to check for newer 4.2 extensions that way.
2012-10-09 23:42:40 -05:00
ac77bbd47b
Move the GLSL extension checks over here so it is nicer.
2012-10-09 23:42:40 -05:00
7cb3d86e31
Add a define back so we work without GLEW 1.6, Also I missed a line when I was fixing binding sampler locations last night when I was tired.
2012-10-09 23:42:40 -05:00
1aad2e9e56
There we go, actually found the issue.
2012-10-09 23:42:40 -05:00
6759ee701d
This thing fails when we don't support binding. Not 100% sure why I need to set this multiple times.
2012-10-09 23:42:40 -05:00
66a245dd68
Remove this silliness
2012-10-09 23:42:40 -05:00
9c09e1973a
Some shader programs don't come with Vertex shaders attached, don't try to bind a Vertex Uniform block then.
2012-10-09 23:42:40 -05:00
f8d0c28e53
Set Sampler values at program make time instead of every frame. Fix an issue when The user had UBO support but not Binding support.
2012-10-09 23:42:40 -05:00
Jordan Woyak