Commit Graph

3762 Commits

Author SHA1 Message Date
Henri Verbeet 48e03bc189 wined3d: Introduce a separate structure for various OpenGL implementation limits. 2009-10-29 13:09:27 +01:00
Henri Verbeet d7b087618b wined3d: Remove some unused GLINFO_LOCATION definitions. 2009-10-29 13:09:20 +01:00
Henri Verbeet e6ca176a48 wined3d: Rename DestroyContext() to context_destroy(). 2009-10-28 11:16:05 +01:00
Henri Verbeet c0050b8ec9 wined3d: Restore the thread's GL context on context_release() if it wasn't created by wined3d.
The background on this is that some applications use OpenGL and DirectDraw/Direct3D on
the same thread, typically using OpenGL for rendering and DirectDraw to get
the amount of available video memory. This means that we might replace the
application's current GL context with one with a completely different state,
but also that our current GL context might not be what we think it is.
2009-10-28 11:15:59 +01:00
Henri Verbeet 98027cb535 wined3d: Call context_release() for wined3d contexts made current by context_create(). 2009-10-28 11:15:27 +01:00
Henri Verbeet eeb54b9922 wined3d: Keep better track of where we're using wined3d contexts.
The idea here is that we can restore the thread's current GL context on
context_release() if it doesn't correspond to the current wined3d context on
context_acquire().
2009-10-28 11:15:19 +01:00
Henri Verbeet 401173ffde wined3d: Restore the thread's previous GL context in context_destroy_gl_resources(). 2009-10-28 11:15:00 +01:00
Henri Verbeet 30d2660d99 wined3d: Restore the thread's previous GL context in WineD3D_ReleaseFakeGLContext(). 2009-10-27 10:55:35 +01:00
Henri Verbeet 94509f85b5 wined3d: Restore the wglMakeCurrent() call in context_destroy_gl_resources().
e3ca576576 removed this call because we don't
need it to tell if the context is valid, but we still need it to actually make
the context current.
2009-10-27 10:55:27 +01:00
Henri Verbeet 13c4fe72ad wined3d: Set a driver name appropriate for the vendor in init_driver_info(). 2009-10-27 10:55:09 +01:00
Henri Verbeet 26ba212407 wined3d: Apply PCI ID overrides in init_driver_info() instead of IWineD3DImpl_GetAdapterIdentifier(). 2009-10-27 10:54:57 +01:00
Henri Verbeet 2447c8bc3e wined3d: Store PCI IDs in wined3d_driver_info instead of wined3d_gl_info. 2009-10-27 10:54:50 +01:00
Stefan Dösinger 0c72464abf wined3d: Use the native constant limit for fitting private constants. 2009-10-26 11:42:36 +01:00
Stefan Dösinger 9ee6508132 wined3d: Remove dead code. 2009-10-26 11:42:29 +01:00
Stefan Dösinger f15fc8f41c wined3d: Take native constants into account when reporting ARB limits. 2009-10-26 11:42:20 +01:00
Stefan Dösinger 641d6681d4 wined3d: Use the ARB native constant limit in the GLSL-ARB constants quirk.
Geforce 8 and Geforce 9 cards support 256 env and 1024 native
constants. 1024 is the hardware limit we want to use if we overwrite
the GLSL reported value with the ARB value on Mac OS (because Mac OS
reports more constants than supported by HW on some cards).
2009-10-26 11:42:10 +01:00
Stefan Dösinger f9c4d442a9 wined3d: Get rid of the general shader constant limit. 2009-10-26 11:41:51 +01:00
Stefan Dösinger 70d6304a21 wined3d: Use the shader backend reported constant limit outside the shader backend. 2009-10-26 11:41:43 +01:00
Henri Verbeet 483d5d0a0d wined3d: Print an ERR instead of crashing in context_check_fbo_status().
Some version of Mesa will claim the window system provided framebuffer is
incomplete when using indirect rendering, although the framebuffer objects
spec says this can never happen. Printing an ERR instead of crashing should
make it clearer that this isn't a wined3d problem.
2009-10-26 11:20:52 +01:00
Henri Verbeet d03a105c1a wined3d: Fix a typo. 2009-10-26 11:20:44 +01:00
Henri Verbeet de82b55b20 wined3d: Refuse to parse shaders with an invalid version token. 2009-10-26 11:20:39 +01:00
Henri Verbeet 5fbe5ee027 wined3d: Add a separate structure for driver info. 2009-10-26 11:20:33 +01:00
Henri Verbeet 69bf130476 wined3d: Remove a redundant ActivateContext() call. 2009-10-26 11:20:26 +01:00
Henri Verbeet ade350b4d4 wined3d: Compare and copy the actual palette in palette9_changed() (Valgrind). 2009-10-23 12:03:37 +02:00
Henri Verbeet 53104ac1e7 wined3d: Don't use the stack as cursor mask (Valgrind). 2009-10-23 12:03:30 +02:00
Henri Verbeet 342a71b281 wined3d: Replace the "stateLookup" tables with a single "wrap_lookup" table. 2009-10-23 12:03:15 +02:00
Henri Verbeet a927573395 wined3d: Free the context array in IWineGDISwapChainImpl_Destroy() (Valgrind). 2009-10-22 10:46:20 +02:00
Henri Verbeet 8d2512317f wined3d: Free the blitter private data in arbfp_blit_free() (Valgrind). 2009-10-22 10:46:15 +02:00
Henri Verbeet 6be41fd59a wined3d: Get rid of GL_LIMITS. 2009-10-22 10:46:05 +02:00
Henri Verbeet 55047f811a wined3d: Fix some GLINFO_LOCATION abuse. 2009-10-22 10:45:57 +02:00
Henri Verbeet fa548414f9 wined3d: Free the shader buffer in shader_glsl_free() (Valgrind). 2009-10-21 16:05:57 +02:00
Henri Verbeet 42eb516d96 wined3d: Free the adapter's format info in IWineD3DImpl_Release() (Valgrind). 2009-10-21 16:05:57 +02:00
Henri Verbeet f2aaca2f67 wined3d: Rename WineD3DAdapter to wined3d_adapter. 2009-10-21 16:05:57 +02:00
Stefan Dösinger 122ed40c2b wined3d: Use ARB constant limits in ARB. 2009-10-20 14:46:16 +02:00
Stefan Dösinger 435646550b wined3d: Use GLSL constant limits in GLSL. 2009-10-20 14:46:13 +02:00
Stefan Dösinger 2d802af4d9 wined3d: Make shader specific constants accessible via GL_LIMITS. 2009-10-20 14:46:08 +02:00
Henri Verbeet 94d33d3e86 wined3d: Destroy FBO entries from the context that created them.
EXT_framebuffer_object doesn't specify if FBOs are shareable between GL
contexts, but ARB_framebuffer_object explicitly prohibits it.
2009-10-20 14:00:37 +02:00
Henri Verbeet e3ca576576 wined3d: Use context->valid in context_destroy_gl_resources().
Instead of relying on wglMakeCurrent() to do what we want.
2009-10-20 14:00:36 +02:00
Henri Verbeet 43aaaa8a4b wined3d: Validate the D3D context in FindContext().
Apparently it's valid to use a D3D device after its window is destroyed.
OpenGL isn't always so forgiving, so this patch is a first step to avoid
making GL calls on a context without window.
2009-10-20 14:00:36 +02:00
Henri Verbeet 0cb6f28751 wined3d: Get rid of the useless "isPBuffer" bitfield in struct wined3d_context. 2009-10-20 14:00:36 +02:00
Henri Verbeet adf5bb5f6a wined3d: Cleanup IWineD3DDeviceImpl_UpdateTexture().
The root problem here is that with the original error handling a NULL
dereference occurs (IWineD3DBaseTexture_GetType()) when either the source or
destination texture is NULL. Rewriting the error handling changes the
indentation of almost the entire function though, so this patch ends up
rewriting the entire function.
2009-10-19 11:45:45 +02:00
Henri Verbeet 9dd78e13d5 wined3d: Printing a FIXME once only works if the relevant BOOL is static (LLVM/Clang). 2009-10-16 13:37:21 +02:00
Henri Verbeet 594497f4bb wined3d: Simplify IWineD3DStateBlockImpl_Apply().
Consistently use the stateblock's contained states to determine what to apply.
2009-10-15 12:06:18 +02:00
Henri Verbeet ce10d5f186 wined3d: Don't apply more than num_contained_render_states in IWineD3DStateBlockImpl_Apply(). 2009-10-15 12:06:18 +02:00
Henri Verbeet bac6486520 wined3d: Merge some bitfields in struct SAVEDSTATES. 2009-10-14 11:26:44 +02:00
Henri Verbeet 1f029a30db wined3d: Simplify IWineD3DStateBlockImpl_Capture().
Consistently use the stateblock's contained states to determine what to
capture.
2009-10-14 11:26:38 +02:00
Henri Verbeet a24d14796a wined3d: Remove the useless "changed" and "enabledChanged" fields from struct wined3d_light_info. 2009-10-14 11:26:26 +02:00
Henri Verbeet 3bc9d1894d wined3d: Rename PLIGHTINFOEL to something less awful. 2009-10-14 11:26:17 +02:00
Henri Verbeet 580cd8c5bb wined3d: Get rid of stateblock_copy_values().
Use IWineD3DStateBlockImpl_Capture() for the initial stateblock capture
instead.
2009-10-14 11:26:02 +02:00
Henri Verbeet 8949a2c1ad wined3d: Remove some code that doesn't make sense.
This was introduced by abb11f30cd, but I don't
see how it could ever do something useful. Perhaps that's a lack of
imagination on my part, but the only situation in which device->stateBlock
should be NULL is while creating the initial stateblock, and in that case we
just returned in the line above. Setting the stream frequency of all streams
to 0x01010101 doesn't make sense either.
2009-10-13 11:38:28 +02:00