Age | Commit message (Collapse) | Author | |
---|---|---|---|
2008-01-28 | gallium: add a couple of hardwired vertex fetch functions | Keith Whitwell | |
2008-01-28 | gallium: only call vertex/prim queue flush when there is something to flush | Keith Whitwell | |
2008-01-26 | gallium: restructure vertex fetch code slightly | Keith Whitwell | |
2008-01-25 | gallium: better flush logic in draw module | Brian | |
This is the other half of Keith's draw/flush patch. There are now 5 flush flags to control what's flushed (post-xform vertex cache, prim cache, vbuf, etc). The gears slow-down in this part of the patch was due to the cull stage not getting invoked. It was unconditional before, but is now gated by 'need_det'. But it also needs to be gated by draw->rasterizer->cull_mode. Gears uses back-face culling. | |||
2008-01-23 | gallium: overhaul usage of vertex_info in draw module. | Brian | |
Remove all dependencies on vertex_info, except for draw_vbuf. Drawing stages now strictly operate on post-transformed vertices and don't know anything about hw vertices. Use vertex program output info for two-side/flat/etc stages. Temporarily disable vbuf module in softpipe driver. | |||
2008-01-04 | llvm: if llvm is enabled don't even try to use sse for vs | Zack Rusin | |
2008-01-03 | clean-ups, silence warnings | Brian | |
2008-01-03 | replace void * with struct draw_vertex_shader opaque type | Brian | |
2007-12-11 | include pipe/tgsi/exec/tgsi_sse2.h if needed. Silences warning. | Brian | |
2007-12-10 | gallium: use hardcoded cliptest code for fixed planes | Keith Whitwell | |
2007-11-23 | gallium: reorg tgsi directories. | Michal Krol | |
2007-11-21 | Simplify draw module's vertex_info. | Brian | |
No longer store the vertex header and clip pos info in the draw module's vertex_info. The vertex_info just describes the data[] elements. This simplifies the code in several places. | |||
2007-11-08 | Reorganize user-space vertex fields in draw_context into 'user' group. | Brian | |
This sub-struct collects the incoming user-provided data/pointers in one place. Ex: draw->mapped_vbuffer becomes draw->user.vbuffer, etc. | |||
2007-11-02 | Add basic entry points for fragment shaders. | Zack Rusin | |
2007-11-02 | Renaming llvmtgsi to gallivm. Taking first steps on the way to supporting | Zack Rusin | |
fragment shaders through llvm. | |||
2007-10-29 | Make gallium compile in win32. | Michal Krol | |
Use FREE, MALLOC, CALLOC, GETENV wrappers. Silence compiler warnings. Add proper copyrights. | |||
2007-10-29 | Refactor the LLVM code a bit. | Zack Rusin | |
Move the CPU vertex shader execution code to the draw module, remove traces of LLVM from the state tracker, abstract execution engine for the purposes of the draw module. | |||
2007-10-27 | Fix SSE bug. | michal | |
2007-10-27 | Respect use_sse flag. | michal | |
2007-10-27 | Enable SSE2 for vertex shaders. | michal | |
2007-10-24 | better debug output | Brian | |
2007-10-24 | Cleanup some code. | Zack Rusin | |
2007-10-24 | Cleanup some of the debugging output code | Zack Rusin | |
2007-10-24 | Cleanup some of the testing code. Implement first pass at actually | Zack Rusin | |
running shaders in llvm. | |||
2007-10-24 | Implement the conversion and do the initial execution pass. | Zack Rusin | |
2007-10-18 | Change type of shader->executable field from void * to generic function pointer. | Brian | |
Fix warnings in draw_create_vertex_shader() | |||
2007-10-18 | better debug code, silence some warnings | Brian | |
2007-10-03 | Move XSTDCALL definition to p_compiler.h | Brian | |
2007-10-03 | Track fragment and vertex shader code generation via pipe shader state objects. | Michel Dänzer | |
Unfortunately, the generated fragment shader code is effectively unusable until it handles quad->mask. | |||
2007-10-03 | Unify the definitions of the 4 component dot product into one | Zack Rusin | |
location. | |||
2007-10-03 | Unify handling of userplanes and regular planes to simplify | Zack Rusin | |
the clipping code. (really done by Keith) | |||
2007-10-02 | remove dead code | Brian | |
2007-10-02 | Move tgsi machine state init/allocations so they're done less frequently. | Brian | |
This, plus expanding all instructions ahead of time, seems to have improved the performance of program execution by 8x or so. | |||
2007-09-28 | Use sse only if GALLIUM_SSE is defined | Zack Rusin | |
2007-09-28 | Redoing the way we handle vertex shaders for the draw module. | Zack Rusin | |
2007-09-28 | Revert "Redoing the way we handle vertex shaders for the draw module." | Zack Rusin | |
This reverts commit 6dcfddb8e2ec2bfb6187b912807fa65f28da2c5e. | |||
2007-09-28 | Revert "Use sse only if GALLIUM_SSE is defined" | Zack Rusin | |
This reverts commit 57b5d3605745c96ddc2b6de7d50c93db65ba1257. | |||
2007-09-28 | Use sse only if GALLIUM_SSE is defined | Zack Rusin | |
2007-09-28 | Redoing the way we handle vertex shaders for the draw module. | Zack Rusin | |
2007-09-27 | Enable codegen based whenever __i386__ is defined. | Keith Whitwell | |
2007-09-27 | disable debug | Keith Whitwell | |
2007-09-27 | Make flushing more lazy in the draw module. | Keith Whitwell | |
2007-09-25 | First attempt at building vertex buffers post-clip. | Keith Whitwell | |
Build a buffer of contigous vertices and indices at the backend of our software transformation/clipping path. This will become the mechanism for emitting buffers of vertices to rasterization hardware. This is similar to but not the same as the post-transform vertex cache. In particular, these vertices are subject to clipping, culling, poly offset, etc. The vertices emitted will all be used by hardware. TODOs include the actual transformation to hardware vertex formats, moving this out of softpipe to somewhere more useful and allowing >1 primitive to share the generated VB. | |||
2007-09-20 | Clean-up the TGSI_SEMANTIC tokens, introduce semantic indexes. | Brian | |
Still need to produce decl instructions for vertex shaders... | |||
2007-09-20 | remove old/used code | Brian | |
2007-09-19 | Checkpoint: replacement of TGSI_ATTRIB_x tokens with input/output semantics. | Brian | |
TGSI_ATTRIB_x tokens still present and used in a few places. Expanded set of TGSI_SEMANTIC_x tokens for describing the meaning of inputs/outputs. These tokens are in a crude state ATM. Lots of #if 0 / disabled code to be removed yet, etc... Softpipe and i915 drivers should be in working condition but not heavily tested. | |||
2007-09-18 | Checkpoint: rework shader input/output register mapping. | Brian | |
This is a step toward removing TGSI_ATTRIB_ tokens. Basically, when translating Mesa programs to TGSI programs, pass in input and output register re-maps, plus interpolation info. There's some known breakage (cubemap.c) so more to be done... | |||
2007-08-27 | Split out vertex shader/cache/fetch functionality from draw_prim.c | Keith Whitwell | |