summaryrefslogtreecommitdiff
path: root/src/gallium/drivers/r300
AgeCommit message (Collapse)Author
2010-06-13r300g: move two-sided stencilref fallback to its own fileMarek Olšák
2010-06-13r300g: move index buffer translate functions to their new homeMarek Olšák
2010-06-13r300g: add fallback for unaligned/unsupported vertex stride/offset/formatMarek Olšák
There is a problem though, the translate module cannot emit half float vertices.
2010-06-13r300g: upload only vertex buffers referenced by vertex elementsMarek Olšák
2010-06-11r300/compiler: Implement simple loop emulationTom Stellard
The loop emulation unrolls loops as may times as possbile while still keeping the shader program below the maximum instruction limit. At this point, there are no checks for constant conditionals. This is only enabled for fragment shaders.
2010-06-10r300g: Fix memory leak on error path.Vinson Lee
2010-06-09r300g: try harder to create a transfer objectMarek Olšák
I was told this wouldn't help to fix the FDO bug #28443, but still, it's a harmless last resort. Also, linear textures safely fallback to an unpipelined transfer here.
2010-06-09r300g: debug output texture sizeMarek Olšák
2010-06-09r300g: return NULL texture early if buffer_create failsMarek Olšák
2010-06-09r300g: save vertex buffers in blitterMarek Olšák
2010-06-09r300g: report that transfers are always supportedMarek Olšák
2010-06-09r300g: fix loading compressed texturesMarek Olšák
2010-06-08r300g: skip rendering for unsupported vertex formats instead of abortingMarek Olšák
The next step is to replace skipping by an actual fallback.
2010-06-08gallium: adjust the query interface to support custom typesZack Rusin
we need to change it to support composite types
2010-06-07r300g: fix hardlock caused by an incompatible vertex formatMarek Olšák
The drivers aborts instead, I am working on a real fix...
2010-06-07r300g: fix 24-bit depth texturingMarek Olšák
This is a secret texture format, do not tell anybody. And it looks like we can read stencil in the shader too.
2010-06-06r300g: workaround index bias on chipsets that do not support itMarek Olšák
2010-06-05mesa/st: add support for EXT_texture_swizzle.Dave Airlie
This passes on r300g, the only bit I'm not really sure about is the handling of the sampler_view in st_atom_texture.c, I unreference it there if the swizzle value changes and I also have to create a new set of functions to create a new one since the u_sampler.c ones don't handle swizzle so much. adds r300g + softpipe enables, I think other drivers could pass easily enough. Signed-off-by: Dave Airlie <airlied@redhat.com>
2010-06-03Merge branch 'gallium-newclear'Roland Scheidegger
Conflicts: src/gallium/state_trackers/python/p_context.i
2010-06-03gallium: rename clearRT / clearDS to clear_render_target / clear_depth_stencilRoland Scheidegger
more consistent with rest of gallium naming conventions. Also rename driver-internal names for these the same.
2010-06-03r300g: use util/u_debugJoakim Sindholt
2010-06-03r300g: set the minimum texture stride to 64 bytes on RS600, RS690, and RS740Marek Olšák
I hope I've done it right. No way to test it here...
2010-06-02r300g: use the stored GEM domain in add_buffer/write_relocMarek Olšák
I admit blitting to/from GTT is actually enforced in this commit, texture transfers are now damned fast!
2010-06-02r300g: store the GEM domain in buffer and texture structsMarek Olšák
2010-06-02r300g: use r300_buffer_domain everywhereMarek Olšák
2010-06-02r300g: blit directly to/from GTT when transfering texturesMarek Olšák
2010-06-02r300g: let the driver determine the GEM domain for buffer_createMarek Olšák
2010-06-02r300g: make texture transfers more pipelinedMarek Olšák
Mostly based on Michel Dänzer's patch.
2010-06-02r300g: better describe another flush and move it closer to the real problemMarek Olšák
I haven't been able to get rid of this one, see the comments in the code.
2010-06-02r300g: remove one flushMarek Olšák
Looks like it's not needed now.
2010-05-31r300g: set correct pitch in libdrmMarek Olšák
to match what DDX sets. This prevents some flushes from occuring.
2010-05-31r300g: remove texture usage transfer flagMarek Olšák
2010-05-29r300g: adapt to clear interface changesRoland Scheidegger
should support separate depth/stencil clears just fine with changed u_blitter.
2010-05-28r300g: remove unnecessary breaksMarek Olšák
2010-05-28r300g: report vertex format support in is_format_supportedMarek Olšák
2010-05-28r300g/swtcl: fix WPOSMarek Olšák
There you are! r300->viewport_state is undefined when using SW TCL. Piglit score (r300g.tests): HW TCL: 231/275 SW TCL: 233/275
2010-05-28r300g/swtcl: avoid an infinite loop if an occl. query doesn't return anythingMarek Olšák
2010-05-27r300g/swtcl: fix secondary color and back-face color outputsMarek Olšák
These piglit tests have been fixed: - bgra-sec-color-pointer - glsl-routing See comments at the beginning of r300_vs_draw.c WPOS is implemented too but it doesn't work yet. I'm still working on it.
2010-05-27r300g: decouple drawing code and two-sided stencil refvalue fallbackMarek Olšák
It's now more separate from the rest of the driver and it can be disabled by commenting out just 1 line. Well, I couldn't make the previous version work with SW TCL reliably, that's the reason of this little rework.
2010-05-26r300g/swtcl: handle large index count properlyMarek Olšák
This fixes piglit/draw-elements.
2010-05-26r300g/swtcl: force vertex prefetching for non-indexed primitivesMarek Olšák
2010-05-26r300g/swtcl: move emitting AOS to prepare_for_renderingMarek Olšák
2010-05-26r300g/swtcl: vbo transfer sanity checksMarek Olšák
2010-05-26r300g: add get_cs_info winsys entrypoint, abandon check_csMarek Olšák
2010-05-26r300g/swtcl: do not use u_upload_mgr and do not compute max_indexMarek Olšák
2010-05-26r300g: remove unnecessary includesMarek Olšák
2010-05-26r300g: fix fence referencingMarek Olšák
2010-05-26r300/compiler: move hardware caps to the radeon_compiler base structMarek Olšák
Needed for vertex shaders too.
2010-05-26r300g: more efficient finish + fix commentsMarek Olšák
2010-05-26r300g: implement fake but compliant fencesMarek Olšák