d099dde7a9
wined3d: Track shader constants in the shader backend.
2008-12-18 13:17:02 +01:00
a14dab4937
wined3d: Don't mark shader constants dirty when we're recording a stateblock.
2008-12-18 13:16:52 +01:00
0968cb9b6c
wined3d: Fix constant setting for the ARB backend.
2008-12-18 13:16:46 +01:00
cf7c8a2425
wined3d: Fix double const.
2008-12-17 15:28:39 +01:00
73823ef5f6
wined3d: Remove the shader_cleanup() method from the shader backend.
2008-12-17 14:02:24 +01:00
78ad5de8da
wined3d: Record floating point pixel shader constants in EndStateBlock() as well.
2008-12-17 14:02:09 +01:00
4adb342327
wined3d: Remove GetFVF() and SetFVF() from IWineD3DDevice.
2008-12-17 14:02:03 +01:00
704f9ea5d1
wined3d: Get rid of IWineD3DPixelShaderImpl_GenerateShader.
2008-12-16 14:36:25 +01:00
2aa7a7fe16
wined3d: Avoid loading opengl32.dll.
...
This confuses applications like Steam, which hook d3d9 and opengl
functions. It sees that the application uses opengl32, but it doesn't
realize that d3d9 is wrapped to opengl. Thus it starts messing around
with wined3d's wgl context. It usually tries to draw geometry with the
context, but cannot deal with some of the obscure extensions we have
activated.
2008-12-16 14:36:17 +01:00
4bc5e760b6
wined3d: Initialize max_combined_samplers properly.
...
Otherwise it will be 0 if GL_ARB_multitexture is not available.
2008-12-16 14:36:02 +01:00
5315b7992d
wined3d: Correct a stage number.
2008-12-16 14:34:20 +01:00
31da3c0578
wined3d: Emulate GL_EXT_fog_coord.
2008-12-16 14:34:11 +01:00
d707f018df
wined3d: Remove some needless initializations.
...
They're not needed, they are handled by the state management anyway.
2008-12-16 14:32:43 +01:00
4b268c79d7
wined3d: Some GL_EXT_fog_coord prototype fixes.
2008-12-16 14:32:29 +01:00
0dc0444c14
wined3d: Reinstall the G16R16F format surface load fixup.
...
We cannot remove this because we still have to load the surface as
RGB. The shader may take care of setting the blue channel to 1.0 now,
but we still get the red and green channels loaded incorrectly if we
don't insert a blue channel before loading.
2008-12-16 14:32:11 +01:00
6ec741e766
wined3d: Set up the shaders when delaying fixed func applying.
2008-12-16 14:31:39 +01:00
199a3468bf
wined3d: Fake GL_ARB_multitexture.
2008-12-16 14:31:25 +01:00
143fb431e0
wined3d: Fix use_vs() usage in vertexdeclaration().
2008-12-16 12:52:10 +01:00
f3743fb402
wined3d: Shaders will never have a NULL function.
2008-12-16 12:52:04 +01:00
46572f682c
wined3d: Refuse to create shaders with a NULL function.
2008-12-16 12:51:49 +01:00
9011452243
wined3d: Simplify shader_generate_main() a bit.
2008-12-15 14:00:29 +01:00
8553665cb1
wined3d: Move the shader version to reg_maps.
2008-12-15 14:00:26 +01:00
dd094fe69d
wined3d: Explicitly pass the shader version to some more functions.
2008-12-15 14:00:21 +01:00
c0a715977d
wined3d: Remove a blank line in state.c.
2008-12-15 12:24:52 +01:00
c937278262
wined3d: Give ATIFS a pixel shader handler.
2008-12-15 12:24:38 +01:00
3e035ddf05
wined3d: Don't leak memory on failure in allocate_shader_constants().
2008-12-12 11:40:44 +01:00
7d29aecd6f
wined3d: Use a simpler way to calculate the number of texture levels.
...
It's probably slightly faster too.
2008-12-12 11:40:39 +01:00
70ed814b95
wined3d: Get rid of the textureDimensions field in the state block.
2008-12-12 11:40:32 +01:00
f36c377d0e
wined3d: We only care is a sampler is sampled at all in device_map_vsamplers(), not the specific type.
2008-12-12 11:40:26 +01:00
fb475c7c29
wined3d: Don't reparse the entire shader just to update the sampler types.
2008-12-12 11:40:20 +01:00
f2f90b6965
wined3d: Fix ARG_UNUSED and use it properly.
...
That bug was uncovered by Henri's change of the bitfield sizes.
2008-12-11 14:07:31 +01:00
fef73782ce
WineD3D: Abstract some ATIfs op debug TRACEs.
...
It makes the code more readable and will help us to merge the ALPHAOP
and COLOROP code.
2008-12-11 14:07:21 +01:00
4916cd5478
wined3d: Don't load INT and BOOL constants needlessly.
2008-12-11 14:06:04 +01:00
441f515d08
wined3d: Don't trace the shader if nobody's looking.
2008-12-11 12:14:55 +01:00
8124f5e14a
wined3d: shader_trace_init() shouldn't have side effects.
...
This moves setting baseShader.hex_version and baseShader.functionLength to
shader_get_registers_used(), where it's more appropriate.
2008-12-11 12:14:50 +01:00
9541c346b3
wined3d: Remove some more unused code.
2008-12-11 12:14:40 +01:00
15d497cb26
wined3d: baseShader.functionLength should be 0 when a NULL pFunction is passed.
2008-12-11 12:14:36 +01:00
a2f259afff
wined3d: Get rid of the "len" variable in shader_trace_init().
2008-12-11 12:14:30 +01:00
b409061337
wined3d: Make some constant arrays also static.
...
As pointed out by Dan Kegel.
2008-12-10 10:50:37 +01:00
bcf0b4b128
wined3d: Compare & copy with the correct clipplane size in IWineD3DStateBlockImpl_Capture().
2008-12-10 10:50:30 +01:00
4aa00e8a21
wined3d: Rename texture_stage_op.color_correction to texture_stage_op.color_fixup.
...
This is consistent with other uses of struct color_fixup_desc.
2008-12-10 10:50:24 +01:00
4aeaba0d29
wined3d: Use the actual fixup for the stage in gen_ati_shader().
...
This fixes a regression introduced by commit
89139b7589
2008-12-10 10:50:18 +01:00
ce4d03318b
wined3d: Fix some spelling errors.
2008-12-09 11:27:51 +01:00
50a87e23e7
wined3d: Fix the IWineD3DVertexShader and IWineD3DPixelShader IUnknown methods.
...
IWineD3DVertexShader isn't supposed to implement IWineD3DPixelShader.
2008-12-09 11:27:32 +01:00
4997bee1bf
wined3d: Add functions to initialize and free shader buffers.
2008-12-09 11:27:13 +01:00
acbdd807b6
wined3d: Don't leak memory in IWineD3DPixelShaderImpl_GenerateShader().
...
This fixes a regression introduced by commit
0bf32b12f5
2008-12-09 11:27:02 +01:00
f1f62fb5ce
wined3d: Remove some unused code.
2008-12-09 11:26:49 +01:00
c3cdc671a4
wined3d: Remove unneeded address-of operators from array names.
2008-12-08 22:03:06 +01:00
5ad3bbabda
wined3d: Avoid disabling numbered arrays if we're going to enable them afterwards anyway.
2008-12-08 12:28:31 +01:00
9887d66dad
wined3d: Return early from shader_trace_init() when a NULL pFunction is passed.
...
This is mainly to reduce the size and indentation level of the if-block.
2008-12-08 12:28:27 +01:00
e18629bedd
wined3d: Explicitly pass the shader version to shader_skip_unrecognized().
2008-12-08 12:28:20 +01:00
6e677775ce
wined3d: Explicitly pass the shader version to shader_get_param().
2008-12-08 12:28:17 +01:00
afd5a53bd0
wined3d: Load texture coordinates from the correct streams.
...
This fixes a regression introduced by commit
a2febdd200
2008-12-08 12:28:11 +01:00
0c954ef743
wined3d: Explicitly pass the shader version to shader_dump_param() and shader_dump_arr_entry().
2008-12-05 11:59:09 +01:00
71a8a77cf9
wined3d: Explicitly pass the shader version to shader_dump_decl_usage().
2008-12-05 11:59:06 +01:00
a13df0e4ef
wined3d: Explicitly pass the version and instruction table to shader_get_opcode().
2008-12-05 11:59:02 +01:00
af1bc3d9ed
wined3d: Remove a redundant condition in shader_get_opcode().
2008-12-05 11:58:58 +01:00
a2febdd200
wined3d: Slightly improve drawStridedSlow() performance.
...
This moves a couple of checks outside of the main drawing loop, since
they're not going to change anyway.
2008-12-05 11:58:53 +01:00
74acbf9c16
wined3d: Get rid of vcheckGLcall.
...
Everywhere it's used a regular checkGLcall will do just as well.
2008-12-04 12:00:29 +01:00
e04556bb16
wined3d: Fix D3DFMT_R32F, R16F, G16R16F, ... in the pixel shader.
...
This allows us to drop the load time conversion and the clear
readback hack and replaces it with a color fixup in the fixed
function pipeline replacement.
2008-12-04 12:00:15 +01:00
89139b7589
wined3d: Make shader texture format fixups more generic.
...
Based on a patch by Stefan Dösinger. This is more flexible, and allows
the shader backend implementation to be simpler, since it doesn't have
to know about specific formats. The next patch makes use of this.
2008-12-04 11:59:58 +01:00
9f2fa8ba25
wined3d: Remove sampled_format from IWineD3DBaseShaderClass, it isn't used anywhere.
2008-12-04 11:59:14 +01:00
b451048eb7
wined3d: Move GlPixelFormatDesc to wined3d_private.h where it belongs.
...
Also remove the silly typedef.
2008-12-04 11:59:06 +01:00
7806e4d85a
wined3d: Make sure a context is set for query object generation.
2008-12-03 10:59:49 +01:00
6800d3dbb7
wined3d: Fix some macros definitions to make them proper C statements.
2008-12-03 10:03:45 +01:00
5284fce1ac
wined3d: Disable checkGLcall when WINE_NO_DEBUG_MSGS is specified.
2008-12-03 09:49:14 +01:00
70968e69ac
wined3d: Use a bitmask to store which bool and int constants are set.
...
Note that constants_set was never actually NULL for
shader_glsl_load_constantsB() and shader_glsl_load_constantsI().
2008-12-03 09:49:10 +01:00
0ae60765de
wined3d: Make functions static where possible.
2008-12-03 09:47:35 +01:00
2acf8d7947
wined3d: Rename resource functions to reflect that they're not COM functions.
2008-12-03 09:47:26 +01:00
4d60b6633c
wined3d: Rename basetexture functions to reflect that they're not COM functions.
2008-12-03 09:47:16 +01:00
9c306e9871
wined3d: Make some functions and variables static.
2008-12-02 15:31:17 +01:00
33e0d1610a
wined3d: Sign-compare warnings fix.
2008-12-02 14:12:39 +01:00
92b93172df
wined3d: Remove some unused code.
2008-12-02 14:04:40 +01:00
89241081f6
wined3d: IWineD3DBaseTexture and IWineD3DResource don't need Vtbl's.
2008-12-02 14:04:27 +01:00
5532c990b6
wined3d: Const correctness fixes.
2008-12-02 14:04:15 +01:00
7c0cb8c0c3
wined3d: Const correctness fixes for state.c.
2008-12-02 14:04:08 +01:00
fec9820e56
wined3d: Handle a few "silent" FIXMEs more consistent with the rest of the code.
...
Also avoids needlessly initializing static variables.
2008-12-02 14:04:00 +01:00
89509eb628
wined3d: Sign-compare warnings fix.
2008-12-01 14:25:52 +01:00
f1271f846b
wined3d: Output the rectangles in the debug message.
2008-12-01 14:01:21 +01:00
6f98f19703
wined3d: Const correctness fixes for drawprim.c.
2008-12-01 12:45:29 +01:00
21b8535902
wined3d: Make the lpData member of WineDirect3DStridedData const.
2008-12-01 12:45:23 +01:00
657a016ac5
wined3d: Fix the glWeightPointerARB prototype.
2008-12-01 12:45:18 +01:00
50ebf26894
wined3d: Const correctness fixes for directx.c.
2008-12-01 12:45:10 +01:00
942f82100a
wined3d: Remove some superfluous casts.
2008-12-01 12:45:01 +01:00
f40afa87c6
wined3d: Sign-compare warnings fix.
2008-11-28 12:57:25 +01:00
5326038aa8
wined3d: Const correctness fixes for surface_base.c.
2008-11-28 12:54:02 +01:00
c7880e8916
wined3d: Make minMipLookup_noFilter and magLookup_noFilter const.
...
Note that minMipLookup and magLookup aren't particularly safe to use,
they're global arrays initialized from IWineD3DImpl_FillGLCaps(). The same
goes for the other global dynamic lookup tables.
2008-11-28 12:53:57 +01:00
449c219eb0
wined3d: Const correctness fixes for device.c.
2008-11-28 12:53:41 +01:00
f2b34c97cc
wined3d: Remove some unnecessary casts in process_vertices_strided().
...
lpData is already a pointer to BYTE. Note that the cast for the normal
data was actually wrong.
2008-11-28 12:53:33 +01:00
31d8b27625
wined3d: Const correctness fixes for ati_fragment_shader.c.
2008-11-28 12:53:27 +01:00
4cea13dec7
wined3d: Const correctness fixes for surface.c.
2008-11-26 12:05:27 +01:00
fecfddfb3a
wined3d: Const correctness fixes for glsl_shader.c.
2008-11-26 12:05:21 +01:00
d8f6e63541
wined3d: Const correctness fixes for arb_program_shader.c.
2008-11-26 12:05:14 +01:00
cfb3bea895
wined3d: Make the SHADER_OPCODE_ARG parameter to shader handlers const.
2008-11-26 12:05:06 +01:00
eb78c2a082
wined3d: Const correctness fixes for utils.c.
2008-11-26 12:04:56 +01:00
f552c90c30
wined3d: Make some functions static.
2008-11-25 13:37:55 +01:00
f0a816e8cb
wined3d: Slightly reduce the scope of fragment_palette_conversion.
2008-11-25 13:37:50 +01:00
49b55f677e
wined3d: Don't call IWineD3DSurfaceImpl_GetData() from surface_gdi.c.
2008-11-25 13:37:45 +01:00
b4f0b5bdd0
wined3d: Const correctness fixes.
2008-11-25 13:37:39 +01:00
a6917b143d
wined3d: Use IDL to generate wined3d.h.
2008-11-25 13:37:22 +01:00
2f98fce9cf
wined3d: Put the ps compile parameters into the glsl program hashmap.
...
This avoids the double search for a pixel shader. The pixel shader
compilation parameter structure is recorded in the GLSL program
hashmap, together with the WineD3D pixel shader.
2008-11-25 13:07:23 +01:00
33482a732e
wined3d: Enable blue = 1.0 fixup for D3DFMT_V8U8.
...
The dx7 sdk demos need this.
2008-11-25 13:07:14 +01:00
0bf32b12f5
wined3d: Add the ability to duplicate GL pixel shaders.
...
Some stateblock parameters have to be compiled into the GL pixel
shader code, like lines for pixelformat fixups. This leads to problems
when applications switch those settings, requiring a recompilation of
the shader. This patch enables wined3d to have multiple GL shaders for
a D3D shader(pixel shaders only so far) to handle this more
efficiently.
2008-11-25 13:07:03 +01:00
aed9305c78
wined3d: Store shader IDs in the vs and ps impl structures.
2008-11-25 13:06:45 +01:00
dd8905543b
wined3d: Pass some stateblock values around directly.
...
This was suggested by Ivan quite a while ago, and we need it to better
handle conflicting texture format corrections and similar stateblock
value changes which until now required a recompilation of the entire
shader
2008-11-25 13:06:35 +01:00
ff767f4984
wined3d: Make the shader backend call CompileShader.
...
A number of considerations contribute to this:
1) The shader backend knows best which shader(s) it needs. GLSL needs
both, arb only one
2) The shader backend may pass some parameters to the compilation
code(e.g. which pixel format fixup to use)
3) The structures used in (2) are different in vs and ps, so a
baseshader::Compile won't work
4) The structures in (2) are wined3d-private structures, so
having a public method in the vtable won't work(its a bad idea
anyway).
2008-11-25 13:06:23 +01:00
c0e271c4de
wined3d: Remove WINAPI on static functions where not needed.
2008-11-25 11:49:57 +01:00
273357264a
wined3d: Add registry key for overriding the pci vendor id.
2008-11-24 14:52:32 +01:00
52d59718c7
wined3d: Add registry key for overriding pci device id.
2008-11-24 14:52:25 +01:00
f442bafb8a
wined3d: Add detection for NVIDIA's GTX 280.
2008-11-24 14:41:58 +01:00
632d34651f
wined3d: Implement CONVERT_CK_RGB24 in d3dfmt_convert_surface.
2008-11-24 14:18:41 +01:00
fdba8b8593
wined3d: Sign-compare warnings fix.
2008-11-21 11:29:12 +01:00
ba78f072e2
wined3d: Remove some unused code.
2008-11-20 14:25:11 +01:00
8899f3412b
wined3d: wined3d_gl.h should be internal to wined3d.
2008-11-20 14:24:56 +01:00
5c79a9f437
wined3d: Rename the fragment ffp desc structures.
...
This reflects the fact that they describe the fragment pipeline.
The vertex pipeline will use its own structures.
2008-11-20 12:17:45 +01:00
cd011661e6
d3d9: Test psize_min vs psize_max vs psize behavior.
...
Add a test that checks what happens if D3DRS_POINTSIZE, D3DRS_POINTSIZE_MIN
and POINTSIZE_MAX have conflicting values. D3DRS_POINTSIZE_MAX trumps
D3DRS_POINTSIZE_MIN, and both MIN and MAX clamp the D3DRS_POINTSIZE value if it is
outside of their range.
2008-11-20 12:17:18 +01:00
234e995bdc
wined3d: Make sure the arbfp pipeline replacement constants are loaded.
...
The code here skipped constant loading when a pixel shader was in use,
and only reloaded them on ffp use if the shader implementation used
ARB too. This way a e.g. texfactor change could get lost if GLSL
shaders are used, and the texfactor changed while a pixel shader was
in use.
2008-11-20 12:16:57 +01:00
495fcc72a6
wined3d: Avoid using a stack buffer in a few places in shader_hw_map2gl().
2008-11-19 14:22:49 +01:00
9b118cc345
wined3d: Give mov & mova their own handler.
2008-11-19 14:22:41 +01:00
bd427f58a0
wined3d: Make it more obvious the pshader path in shader_hw_map2gl() doesn't handle opcodes without parameters.
2008-11-19 14:22:34 +01:00
29f2fefc2b
wined3d: Fix the cube map coordinates in surface_blt_to_drawable().
...
I don't see how this can ever have worked properly.
2008-11-19 14:22:29 +01:00
f571ecd23c
wined3d: Add two missing D3D8/D3D9 capabilities.
2008-11-10 12:57:30 +01:00
620a423b22
wined3d: Kill the GL_ATI_envmap_bumpmap code.
...
GL_ATI_envmap_bumpmap provides two things: Signed V8U8 pixel formats,
and bump mapping. The extension is only supported on fglrx, and this
driver also supports GL_ARB_fragment_program. Thus the bump mapping
code is never used on any driver out there. Furthermore, if it is
used, it tends to crash the driver
The signed pixel format is used, as it can be used by pixel shaders or
the ARBfp replacement. However, the format is broken in fglrx, and
negative values are clamped to 0.0. This results in test
failures. WineD3D has an alternative codepath using scale+bias to
enable V8U8 using a standard signed RGB which works correctly on
fglrx.
2008-11-04 11:41:14 +01:00
a8497a5545
wined3d: Fix the num blend values -> type match.
2008-11-04 11:40:37 +01:00
efeab9d200
wined3d: Use the integer variant of zero instead of casting "0.0".
2008-11-04 11:26:05 +01:00
93af8d3631
d3d: Do not restore the display mode in ddraw.
2008-11-03 13:34:13 +01:00
8b38533f69
wined3d: Restore the fragment replacement prog after depth_blt.
...
The current code properly enabled/disabled GL_ARB_fragment_program
after a depth blit, but it did not restore the bound fragment program
properly. This leads to problems if a depth blit was done between two
draws without any change of the fragment processing settings.
2008-11-03 13:34:13 +01:00
b11dd54258
wined3d: Remove another redundant NULL check (LLVM/Clang).
2008-11-03 11:49:46 +01:00
dfde4c91e0
wined3d: vertexshader should never be NULL in generate_param_reorder_function() (LLVM/Clang).
2008-11-03 11:49:41 +01:00
373a43442d
wined3d: Remove a FIXME that doesn't apply anymore.
2008-11-03 11:49:20 +01:00
6e1b7333e6
wined3d: Print an error when drawStridedSlowVs() is called with 0 idxSize and non-NULL idxData (LLVM/Clang).
2008-10-31 12:58:43 +01:00
c090471a49
wined3d: Properly break in get_argreg() (LLVM/Clang).
2008-10-31 12:58:37 +01:00
795c4a77aa
wined3d: Handle projected cube textures.
2008-10-29 11:52:11 +01:00
d262752aea
wined3d: Handle lack of NPOT support for depth blts.
2008-10-29 11:52:06 +01:00
80818bfc8d
wined3d: Support some more depth blt texture types in surface_depth_blt().
2008-10-29 11:52:01 +01:00
17ab7f6b83
wined3d: Support some more depth blt texture types for arb programs.
2008-10-29 11:51:54 +01:00
f4e0dec42e
wined3d: Fix typo in baseshader.c.
2008-10-28 13:26:01 +01:00
c52a55d724
wined3d: Fix some indentation.
2008-10-28 12:46:51 +01:00
092b51e82a
wined3d: Rename CreateAdditionalSwapChain to CreateSwapChain.
2008-10-28 12:46:46 +01:00
b33ddbde6a
wined3d: Support some more texture types for GLSL depth blts.
2008-10-28 12:33:36 +01:00
437b342424
wined3d: Pass explicit texcoords to depth blt.
2008-10-28 12:33:27 +01:00
57401fcab2
wined3d: Pass the texture type to the shader depth blt function.
2008-10-28 12:33:18 +01:00
4034a29f19
wined3d: Move depth_blt to surface.c.
2008-10-28 12:33:13 +01:00
69553c2514
wined3d: Fix the WineDirect3DCreate() export.
2008-10-27 12:37:27 +01:00
7142754820
wined3d: Remove the SDKVersion parameter to WineDirect3DCreate().
2008-10-24 14:24:59 +02:00
e79d89ff40
wined3d: Fix some typos.
2008-10-21 11:53:40 +02:00
bf187ed1e0
wined3d: Set the sampler only once in shader_glsl_select_depth_blt().
...
It (obviously) never changes.
2008-10-20 12:23:26 +02:00
2188aa87d6
wined3d: Add support for EXT_framebuffer_multisample.
2008-10-20 12:23:16 +02:00
728e61e78b
wined3d: Remove ENTER_GL/LEAVE_GL from VolumeTexture_PreLoad because it doesn't make any GL calls.
2008-10-20 11:39:01 +02:00
3e3141da34
wined3d: Protect gl calls in Volume_LoadTexture with ENTER_GL/LEAVE_GL.
2008-10-20 11:38:54 +02:00
Henri Verbeet