Age | Commit message (Collapse) | Author | |
---|---|---|---|
2000-11-05 | - Changes for new software rasterizer modules | Keith Whitwell | |
- Remove support for choosing software fallbacks from core code - Remove partial fallback code from vbrender.c -- drivers are now expected to be able to find a triangle/quad function for every state, even if they have to use _swsetup_Triangle or _swsetup_Quad. - Marked derived variables in the GLcontext struct with a leading underscore '_'. | |||
2000-11-05 | Reworked to use the new software rasterizer. Optimized line/tri functions | Keith Whitwell | |
are hooked into the software rasterizer. | |||
2000-11-05 | Use the new software rasterizer. Reworked optimized line,tri,point | Keith Whitwell | |
functions to fit into the framework provided for extending the software rasterizer. | |||
2000-11-05 | Reorganized software rasterizer as a module which manages its own state, | Keith Whitwell | |
with tighter interfaces with the rest of the world. Proper documentation to come. | |||
2000-11-05 | A new module to provide RasterSetup and advanced triangle/line/point | Keith Whitwell | |
functionality layered on top of the software rasterizer. An example entrypoint: void _swsetup_Triangle( GLcontext, GLuint, GLuint, GLuint, GLuint ) will coerce the software rasterizer to draw flat, twoside-lit, unfilled and offset triangles (including decomposition to points or lines). | |||
2000-11-05 | Lots of changes: | Keith Whitwell | |
- use the new interface to the software rasterizer. - manage all fallbacks internally, hooking in swrast or swrast_setup if necessary. - removed lots of marginal code no longer appropriate in a maturing driver. - reworked the vertex-setup and triangle routines, including drawing unfilled triangles from within the driver. | |||
2000-11-01 | Committing in . | Jouk Jansen | |
Modified Files: Mesa/src/descrip.mms Updated Makefiles for VMS ---------------------------------------------------------------------- | |||
2000-10-31 | Always call Driver.SetReadBuffer() in copy pixels functions, just to be safe | Brian Paul | |
2000-10-31 | Moved the software rasterizer to a new directory. | Keith Whitwell | |
2000-10-31 | Moved software rasterizer functionality to new directory. | Keith Whitwell | |
2000-10-31 | Remove dead code in FX driver. | Keith Whitwell | |
2000-10-30 | Rearrange VERT_* flags to support 4 texture units. | Keith Whitwell | |
2000-10-30 | Removed UPDATE_IMAGE_TRANSFER_STATE, use _NEW_PIXEL flag instead. | Brian Paul | |
More minor GLchan changes. Silence some compiler warnings in cva.[ch] | |||
2000-10-30 | Removed references to gcc's __FUNCTION__ macro. | Keith Whitwell | |
Removed dead file fog_tmp.h | |||
2000-10-30 | set *t->flag = enabled in gl_extensions_add(), otherwise, extension booleans ↵ | Brian Paul | |
weren't set | |||
2000-10-30 | Replace the flags Mesa was using for ctx->NewState with a new set | Keith Whitwell | |
based on the GL attribute groups. Introduced constants describing the circumstances under which some key derived values can change: _SWRAST_NEW_RASTERMASK -- ctx->RasterMask _SWRAST_NEW_TRIANGLE -- The software rasterizer's triangle function _DD_NEW_FEEDBACK -- the 'DD_FEEDBACK' bit in ctx->TriangleCaps These are helpful in deciding whether you need to recalculate state if your recalculation involves reference to a derived value. | |||
2000-10-30 | Committing in . | Jouk Jansen | |
Modified Files: Mesa/src/X/xm_api.c included macros.h ---------------------------------------------------------------------- | |||
2000-10-29 | renamed Current.ByteColor Current.Color | Brian Paul | |
2000-10-29 | more minor header file re-org (moved CONST, ASSERT, INLINE to config.h) | Brian Paul | |
2000-10-29 | More GLchan color channel changes. | Brian Paul | |
Some header file re-org: Move matrix, vertex buffer structs to types.h to fix #include mess. Remove typedef, extern declarations from config.h | |||
2000-10-28 | New colormac.h file for color-related macros. | Brian Paul | |
Lot's of clean-up in macros.h and mmath.h | |||
2000-10-28 | Basic work to support deep color channels: | Brian Paul | |
Replace GLubyte with GLchan Replace 255 with CHAN_MAX | |||
2000-10-28 | Do per-pixel fog when GL_FOG_HINT is GL_NICEST. | Keith Whitwell | |
Fix scaling bug in per-pixel fog calculations. | |||
2000-10-28 | Lots of misc. bugfixes to bring GGIMesa in sync with the latest Mesa sources. | Jon Taylor | |
2000-10-27 | restored test for __alpha in _mesa_win_fog_coords_from_z() | Brian Paul | |
2000-10-27 | Enabled GL_EXT_secondary_color. Fixed a bunch of typos in the dlist.c | Brian Paul | |
and state.c file for plugging those functions into the dispatch table. Don't use Mesa 3.5 for DRI until SGI approves the new dispatch offsets. Commented-out references to ctx->FogMode and VB->Specular in FX driver. Minor clean-up in extensions.c Removed unused prototype in fog.h | |||
2000-10-27 | Implement EXT_fog_coord and EXT_secondary_color. | Keith Whitwell | |
EXT_secondary_color is disabled until we get some dispatch offsets from SGI. | |||
2000-10-26 | added GL_BGR case to bytes_per_pixel() | Brian Paul | |
2000-10-26 | clarified meaning of DEFAULT_SOFTWARE_DEPTH_BITS value | Brian Paul | |
2000-10-24 | added incomplete() debug function/macro | Brian Paul | |
2000-10-24 | signal NEW_TEXTURING if there's a Driver.BindTexture() function | Brian Paul | |
2000-10-23 | generate separate libOSMesa library ala XFree86/DRI | Brian Paul | |
2000-10-23 | Major audit of all Mesa's x86 assembly code. This round is basically | Gareth Hughes | |
general cleanups - more to come. Added P6 architecture timing to debug_xform routines. Still need to add test_all_vertex_functions test for the v16 asm. Dynamic reconfiguration of counter overhead for more accurate benchmarking. | |||
2000-10-21 | fixed a texture enable bug | Brian Paul | |
2000-10-21 | polygon offset changes; added Visual->MRD field | Brian Paul | |
2000-10-20 | Changes for multitexture > 3, code clean-ups. | Brian Paul | |
Added GLboolean ctx->Texture.MultiTextureEnabled to determine when multitexture is enabled. Eventually ctx->Texture.ReallyEnabled may become a boolean. | |||
2000-10-19 | Rewrote get_static_proc_address(). It made mistakes in some situations | Brian Paul | |
2000-10-19 | glPixelStore(GL_PACK/UNPACK_SKIP_IMAGES) wasn't handled correctly | Brian Paul | |
2000-10-19 | call Driver.SetReadBuffer() in clear_color_buffers() to fix colormasked ↵ | Brian Paul | |
clear bug | |||
2000-10-19 | improved blending accuracy to fix Glean test failures | Brian Paul | |
2000-10-18 | More fixes for MAX_TEXTURE_UNITS > 2. Seems to work now. | Brian Paul | |
Misc code clean-ups, minor bug fixes. MAX_TEXTURE_UNITS now defaults to 3. | |||
2000-10-17 | ported render begin/end calls from Mesa 3.4 | Brian Paul | |
2000-10-16 | added _mesa_get_teximages_from_driver() | Brian Paul | |
2000-10-09 | if depthBits == 0, set DepthMask = 2^16 to fix per-fragment fog problems | Brian Paul | |
2000-10-05 | update TriangleCaps in GL_POINT/POLYGON_SMOOTH cases | Brian Paul | |
2000-10-05 | better handling of GLX_DEPTH_SIZE in glXChooseVisual() | Brian Paul | |
2000-10-05 | drawing color-index image to RGB window was broken | Brian Paul | |
2000-10-02 | more work to support N texture units | Brian Paul | |
2000-09-30 | support for N texture units | Brian Paul | |
2000-09-29 | generalized support for N texure units | Brian Paul | |