Commit message (Collapse) | Author | Age | Files | Lines | |
---|---|---|---|---|---|
* | Remove fences from the i915simple winsys interface. | José Fonseca | 2007-11-22 | 5 | -35/+8 |
| | | | | | | | | | Fences will be part of the pipe winsys interface, so remove this to avoid merge conflicts later on. This reverts commit ca7f68a7cf25a51f382bba8c42d8c6ab7db57b5d. This reverts commit dec60d33b2570cf2bdce72a00a1539ee93133f91. This reverts commit 90dd0cb822f2fe14258c786e5c37da69472b7d17. | ||||
* | more simplification, clean-up in draw_set_vertex_info() | Brian | 2007-11-21 | 1 | -8/+0 |
| | |||||
* | Replace draw_set_vertex_attributes() with simpler draw_set_vertex_info(). | Brian | 2007-11-21 | 4 | -32/+21 |
| | | | | Just pass in the vertex_info object and make a copy of it. | ||||
* | Simplify draw module's vertex_info. | Brian | 2007-11-21 | 6 | -34/+20 |
| | | | | | | 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. | ||||
* | add PIPE_FORMAT_Z24_S8 support to softpipe patsh | Brian | 2007-11-20 | 4 | -0/+93 |
| | |||||
* | initial support for PIPE_FORMAT_Z24_S8 | Brian | 2007-11-20 | 1 | -0/+2 |
| | |||||
* | optimize earlyz_quad(), add comments, remove unneeded #includes | Brian | 2007-11-19 | 1 | -21/+13 |
| | |||||
* | optimize linear_interpolation(), perspective_interpolation() functions | Brian | 2007-11-19 | 1 | -25/+18 |
| | |||||
* | Implement early depth test. | Michal Krol | 2007-11-18 | 6 | -25/+150 |
| | | | | | | | | | | | | Early depth test is enabled when depth test is enabled and alpha test is disabled and fragment shader does not write depth. The early-z is implemented by moving the depth test stage just before the fragment shader stage and prepending it with an earlyz stage, introduced with this commit. The earlyz stage prepares the quad->outputs.depth for the following depth test stage by interpolating Z position, just as the fragment shader would do. | ||||
* | Proper fence reference counting. | José Fonseca | 2007-11-17 | 3 | -2/+11 |
| | |||||
* | more convenient debug code | Brian | 2007-11-16 | 1 | -2/+10 |
| | |||||
* | note problem with fragment coord Y orientation | Brian | 2007-11-16 | 1 | -0/+4 |
| | |||||
* | added a clip-bypass flag to rasterizer state | Brian | 2007-11-16 | 2 | -0/+2 |
| | |||||
* | Include fences in the i915simple winsys interface. | José Fonseca | 2007-11-15 | 4 | -8/+26 |
| | |||||
* | comments, assertions | Brian | 2007-11-09 | 2 | -4/+8 |
| | |||||
* | minor simplifcation in vertex_fetch() | Brian | 2007-11-09 | 1 | -8/+1 |
| | |||||
* | Reorganize user-space vertex fields in draw_context into 'user' group. | Brian | 2007-11-08 | 8 | -32/+45 |
| | | | | | This sub-struct collects the incoming user-provided data/pointers in one place. Ex: draw->mapped_vbuffer becomes draw->user.vbuffer, etc. | ||||
* | remove unneeded #includes, added assertion in draw_compute_vertex_size() | Brian | 2007-11-08 | 1 | -4/+2 |
| | |||||
* | Remove unneeded header, draw_prim.h | Brian | 2007-11-08 | 1 | -11/+0 |
| | |||||
* | Fix minor logic error in get_queued_prim() which caused flushing when there ↵ | Brian | 2007-11-08 | 1 | -1/+7 |
| | | | | was still room in the prim queue. | ||||
* | new comments | Brian | 2007-11-08 | 1 | -2/+2 |
| | |||||
* | remove unneeded #includes | Brian | 2007-11-08 | 1 | -2/+0 |
| | |||||
* | comments, remove unneeded #includes | Brian | 2007-11-08 | 1 | -4/+12 |
| | |||||
* | For TGSI_TEXTURE_2D/RECT, need to provide 3 coords (STP) in case shadow ↵ | Brian | 2007-11-08 | 1 | -7/+9 |
| | | | | compare mode is enabled. | ||||
* | Implement shadow comparisons. | Brian | 2007-11-08 | 1 | -1/+67 |
| | |||||
* | use pipe_surface_reference() in sp_tile_cache_set_surface() | Brian | 2007-11-08 | 1 | -1/+1 |
| | |||||
* | fix bad width/height code in softpipe_clear() | Brian | 2007-11-08 | 1 | -3/+2 |
| | |||||
* | Remove TGSI_OPCODE_TXP | Brian | 2007-11-08 | 2 | -9/+5 |
| | | | | Use TGSI_OPCODE_TEX with ExtDivide=TGSI_EXTSWIZZLE_W instead. | ||||
* | check for divide by W flag when translating TGSI_OPCODE_TEX (fixes projected ↵ | Brian | 2007-11-08 | 1 | -1/+7 |
| | | | | textures) | ||||
* | tweak anisotropic filtering code | Brian | 2007-11-08 | 2 | -1/+4 |
| | |||||
* | enable PIPE_CAP_TEXTURE_SHADOW_MAP (but not quite working yet) | Brian | 2007-11-08 | 1 | -1/+1 |
| | |||||
* | remove translate_compare_func(), use i915_translate_compare_func() | Brian | 2007-11-08 | 1 | -19/+1 |
| | |||||
* | Fix some issues hit when rendering to a depth-only renderbuffer (shadowtex). | Brian | 2007-11-08 | 1 | -7/+48 |
| | | | | | Added 16bpp case to translate_depth_format(). Added framebuffer_size() to determine framebuffer size. | ||||
* | remove unneeded #include | Brian | 2007-11-08 | 1 | -1/+0 |
| | |||||
* | Added pipe->get_paramf() to query float limits. | Brian | 2007-11-08 | 5 | -1/+61 |
| | | | | So far max point size, line width, texture anistopy and lod bias. | ||||
* | Rename struct field 'private' to 'priv'. | Michel Dänzer | 2007-11-08 | 1 | -1/+1 |
| | | | | This broke the LLVM build because 'private' is a C++ keyword. | ||||
* | fix warning | Brian | 2007-11-07 | 1 | -1/+2 |
| | |||||
* | update flush/fence comments for i915_flush() | Brian | 2007-11-07 | 1 | -3/+3 |
| | |||||
* | Remove winsys->wait_idle(), subsumed by pipe->flush(). | Brian | 2007-11-07 | 4 | -10/+2 |
| | | | | | | Connect intel_i915_batch_finish() into i915_winsys, just like intel_i915_batch_flush(). Call i915_winsys->batch_finish() in response to pipe->flush(PIPE_FLUSH_WAIT). Now all the batchbuffer/fence code is in one place and a little cleaner. | ||||
* | New PIPE_FLUSH_WAIT flag for pipe->flush(). | Brian | 2007-11-07 | 2 | -2/+7 |
| | | | | | The state tracker doesn't have to directly call winsys->wait_idle() anymore. glFlush and glFinish both go through pipe->flush() now. | ||||
* | Remove context dependencies in winsys layer. | Brian | 2007-11-07 | 3 | -3/+6 |
| | | | | | | The winsys object is now per-screen and shared by multiple contexts. The regionPool is now part of the i915 winsys layer. The winsys wait_idle() and flush_frontbuffer() funcs will get more attention... | ||||
* | just reorder a few members of pipe_winsys, update comments | Brian | 2007-11-07 | 1 | -7/+6 |
| | |||||
* | more dead code removal | Brian | 2007-11-07 | 1 | -63/+1 |
| | |||||
* | re-enable PIPE_FORMAT_U_Z32 support | Brian | 2007-11-07 | 1 | -1/+1 |
| | |||||
* | better depth buffer selection | Brian | 2007-11-07 | 1 | -1/+11 |
| | |||||
* | disable/remove dead code | Brian | 2007-11-07 | 1 | -8/+3 |
| | |||||
* | Disable the call to driBOReference() in i915_update_texture() | Brian | 2007-11-07 | 1 | -1/+5 |
| | | | | | It doesn't seem to be needed and disabling it fixes a big memory leak with some programs like xdemos/wincopy.c | ||||
* | redo xm_surface_release() to match i915 | Brian | 2007-11-07 | 1 | -3/+6 |
| | |||||
* | fix bad varname | Brian | 2007-11-07 | 1 | -1/+1 |
| | |||||
* | Add winsys->surface_release() to complement winsys->surface_alloc(). | Brian | 2007-11-07 | 8 | -45/+45 |
| | | | | | pipe_surface now has a pointer to the winsys which create/owns the surface. This allows clean surface deallocation w/out a rendering context. |