| Age | Commit message (Collapse) | Author | |
|---|---|---|---|
| 2010-10-15 | llvmpipe: use aligned loads/stores for plane values | Keith Whitwell | |
| 2010-10-15 | gallium: move some intrinsics helpers to u_sse.h | Keith Whitwell | |
| 2010-10-15 | llvmpipe: slightly shrink the size of a binned triangle | Keith Whitwell | |
| 2010-10-12 | llvmpipe: make sure intrinsics code is guarded with PIPE_ARCH_SSE | Keith Whitwell | |
| 2010-10-12 | llmvpipe: improve mm_mullo_epi32 | José Fonseca | |
| Apply Jose's suggestions for a small but measurable improvement in isosurf. | |||
| 2010-10-12 | gallium: move sse intrinsics debug helpers to u_sse.h | Keith Whitwell | |
| 2010-10-12 | llvmpipe: Fix MSVC build. | José Fonseca | |
| MSVC doesn't accept more than 3 __m128i arguments. | |||
| 2010-10-12 | llvmpipe: fix typo in last commit | Keith Whitwell | |
| 2010-10-12 | llvmpipe: try to do more of rast_tri_3_16 with intrinsics | Keith Whitwell | |
| There was actually a large quantity of scalar code in these functions previously. This tries to move more into intrinsics. Introduce an sse2 mm_mullo_epi32 replacement to avoid sse4 dependency in the new rasterization code. | |||
| 2010-10-08 | llvmpipe: add rast_tri_4_16 for small lines and points | Keith Whitwell | |
| 2010-09-13 | llvmpipe: Fix non SSE2 builds. | José Fonseca | |
| Should fix fdo 30168. | |||
| 2010-09-12 | llvmpipe: introduce tri_3_4 for tiny triangles | Keith Whitwell | |
| 2010-09-12 | llvmpipe: refactor tri_3_16 | Keith Whitwell | |
| Keep step array as a set of four m128i's and reuse throughout the rasterization. | |||
| 2010-09-12 | llvmpipe: pass linear masks to fragment shader | Keith Whitwell | |
| Fragment shader can extract the correct bits for each quad. | |||
| 2010-08-31 | llvmpipe: slightly simplify build_mask | Keith Whitwell | |
| 2010-08-31 | llvmpipe: combine linear mask calculation | Keith Whitwell | |
| 2010-08-31 | llvmpipe: intrinsics versions of build_mask functions | Keith Whitwell | |
| 2010-08-27 | llvmpipe: native rasterization for lines | Hui Qi Tay | |
| Rasterize lines directly by treating them as 4-sided polygons. Still need to check the exact pixel rasteration. | |||
| 2010-08-15 | llvmpipe: special case triangles which fall in a single 16x16 block | Keith Whitwell | |
| Check for these and route them to a dedicated handler with one fewer levels of recursive rasterization. | |||
| 2010-08-15 | llvmpipe: remove all traces of step arrays, pos_tables | Keith Whitwell | |
| No need to calculate these values any longer, nor to store them in the bin data. Improves isosurf a bit more, 115->123 fps. | |||
| 2010-08-15 | llvmpipe: eliminate last usage of step array in rast_tmp.h | Keith Whitwell | |
| For 16 and 64 pixel levels, calculate a mask which is linear in x and y (ie not in the swizzle layout). When iterating over full and partial masks, figure out position by manipulating the bit number set in the mask, rather than relying on postion arrays. Similarly, calculate the lower-level c values from dcdx, dcdy and the position rather than relying on the step array. | |||
| 2010-08-15 | llvmpipe: version of block4 which doesn't need the full step array | Keith Whitwell | |
| No noticable slowdown with isosurf. | |||
| 2010-08-15 | llvmpipe: reorganize block4 loop, nice speedup | Keith Whitwell | |
| isosurf 95->115 fps just by exchanging the two inner loops in this function... | |||
| 2010-07-13 | llvmpipe: pass mask into fragment shader | Keith Whitwell | |
| Move this code back out to C for now, will generate separately. Shader now takes a mask parameter instead of C0/C1/C2/etc. Shader does not currently use that parameter and rasterizes whole pixel stamps always. | |||
| 2010-02-24 | llvmpipe: more lp_rasterizer_task parameter passing | Brian Paul | |
| 2010-02-24 | llvmpipe: pass fewer parameters to rasterization functions | Brian Paul | |
| 2010-02-24 | llvmpipe: added some assertions | Brian Paul | |
| 2010-02-17 | llvmpipe: use ffs technique for full tiles also | Keith Whitwell | |
| Need to compute two masks here for full and partial 16x16 blocks. Gives a further good improvement for isosurf particularly: isosurf 97 -> 108 gears 597 -> 611 | |||
| 2010-02-17 | llvmpipe: rework do_block_16 to use bitmasks and ffs | Keith Whitwell | |
| Some nice speedups: gears: 547 -> 597 isosurf: 83 -> 98 Others like gloss unchanged. Could do further work in this direction. | |||
| 2010-01-21 | llvmpipe: use some local vars to index step arrays | Brian Paul | |
| Saves a few more cycles. | |||
| 2010-01-21 | llvmpipe: added simple perf/statistics counting facility | Brian Paul | |
| Currently counting number of tris, how many tiles of each size are fully covered, partially covered or empty, etc. Set LP_DEBUG=counters to enable. Results are printed upon context destruction. | |||
| 2010-01-15 | llvmpipe: skip 4x4 in/out test code | Brian Paul | |
| It's a litte faster to just do the in/out testing in the shader jit code. | |||
| 2010-01-15 | llvmpipe: added comment about lookup-tables vs. computation | Brian Paul | |
| 2010-01-15 | llvmpipe: generate two shader varients, one omits triangle in/out testing | Brian Paul | |
| When we know that a 4x4 pixel block is entirely inside of a triangle use the jit function which omits the in/out test code. Results in a few percent speedup in many tests. | |||
| 2009-12-17 | llvmpipe: replace INT_MIN/2 with INT_MIN | Brian Paul | |
| Since changing the in/out test we can just use INT_MIN to be sure the comparison against the step values always passes. | |||
| 2009-12-17 | llvmpipe: improve the in/out test a little | Brian Paul | |
| Instead of: s = c + step m = s > 0 Do: m = step > c (with negated c) | |||
| 2009-12-16 | llvmpipe: do final the pixel in/out triangle test in the fragment shader | Brian Paul | |
| The test to determine which of the pixels in a 2x2 quad is now done in the fragment shader rather than in the calling C code. This is a little faster but there's a few more things to do. Note that the step[] array elements are in a different order now. Rather than being in row-major order for the 4x4 grid, they're in "quad-major" order. The setup of the step arrays is a little more complicated now. So is the course/intermediate tile test code, but some lookup tables help with that. Next steps: - early-cull 2x2 quads which are totally outside the triangle. - skip the in/out test for fully contained quads - make the in/out comparison code tighter/faster. | |||
| 2009-12-07 | llvmpipe: repartition lp_rasterizer state for threading | Brian Paul | |
| Some of the state is per-thread. Put that state in new lp_rasterizer_task struct. | |||
| 2009-12-04 | llvmpipe: use LP_DBG() macro everywhere | Brian Paul | |
| 2009-12-01 | llvmpipe: added assertions | Brian Paul | |
| And remove unused BLOCKSIZE. | |||
| 2009-12-01 | llvmpipe: simplify mask computation | Brian Paul | |
| Make this a little easier to understand. | |||
| 2009-12-01 | llvmpipe: replace shifts with multiplies to be clearer | Brian Paul | |
| The compiler will still do the multiplies with shifts. It's just a bit easier to follow the logic with multiplies. | |||
| 2009-12-01 | llvmpipe: make nr_blocks unsigned | Brian Paul | |
| 2009-12-01 | llvmpipe: comments, reformatting and assertions in tri rast code | Brian Paul | |
| 2009-10-20 | llvmpipe: move block list into rast struct | Keith Whitwell | |
| 2009-10-20 | llvmpipe: build list of 4x4 blocks to be shaded | Keith Whitwell | |
| 2009-10-20 | llvmpipe: recursive rasterization within a tile | Keith Whitwell | |
| 2009-10-20 | llvmpipe: precalculate some offsets | Keith Whitwell | |
| 2009-10-19 | llvmpipe: calculate masks in format desired by shader | Keith Whitwell | |
| Also remove branches calculating masks for quads. | |||
| 2009-10-19 | llvmpipe: pre-multiply some constants by fixed_one | Keith Whitwell | |
 |
index : android-x86-mesa.git | |
| Androïd/x86 port of Mesa drivers | Hugues Hiegel |
| summaryrefslogtreecommitdiff |