summaryrefslogtreecommitdiffstats
path: root/libdimension/bench
Commit message (Collapse)AuthorAgeFilesLines
* math: Make vectors have an array instead of different fields.Tavian Barnes2015-10-252-14/+5
|
* Modularize the libdimension codebase.Tavian Barnes2015-10-255-37/+39
|
* geometry: Faster AABB transformations.Tavian Barnes2014-06-121-2/+16
|
* Use // comments when possible.Tavian Barnes2014-06-076-46/+46
|
* objects: Refactor how bounding and initialization work.Tavian Barnes2014-06-072-10/+12
|
* triangles: Don't forget to transform normals.Tavian Barnes2014-06-041-1/+1
| | | | This fixes the artifacts seen when rendering meshes.
* triangles: Better API.Tavian Barnes2014-06-041-4/+4
|
* bench: Run the future benchmark last becuase it's the slowest.Tavian Barnes2014-05-311-2/+2
|
* object: Use pool.Tavian Barnes2014-05-312-16/+15
|
* array: Add DMNSN_NEW_ARRAY macro.Tavian Barnes2014-05-311-2/+2
|
* bench: Add a triangle intersection benchmark.Tavian Barnes2014-05-222-2/+70
|
* future: Add to benchmark.Tavian Barnes2014-04-261-24/+69
|
* future: Add a race-free way to examine a partial computation.Tavian Barnes2014-04-231-2/+3
| | | | | | This allows safe OpenGL previews, for example. dmnsn_future* learned the dmnsn_future_{pause,resume}() functions which cause all worker threads to block. render.test now survives Helgrind with no errors.
* Use strerror_r if available instead of sys_errlist.Tavian Barnes2014-02-031-3/+3
|
* prtree: Sort large workloads in parallel.Tavian Barnes2014-02-011-0/+1
| | | | Performance benefit is around 33% for more than 1000 objects.
* Fix a race in the future benchmark.Tavian Barnes2013-03-221-1/+5
|
* Also benchmark cycles/increment for futures.Tavian Barnes2013-03-201-3/+14
|
* Add a benchmark for dmnsn_future.Tavian Barnes2013-03-192-1/+71
|
* Allow other BVH implementations to be used.Tavian Barnes2012-12-171-13/+14
| | | | | dmnsn_bvh is now a generic API, which could potentially support octrees, etc, in addition to PR-trees.
* Replace INCLUDES with AM_CFLAGS.Tavian Barnes2012-12-161-3/+1
|
* Add a non-cached case to the PR-tree benchmark.Tavian Barnes2011-11-171-0/+5
|
* Make API more consistent.Tavian Barnes2011-10-302-3/+3
| | | | Object methods should be dmnsn_<object>_<fn>().
* Rename progress objects to future objects.Tavian Barnes2011-10-191-1/+1
|
* Transform normals as pseudovectors, not vectors.Tavian Barnes2011-09-141-4/+16
| | | | | | | | Also clarify the vector transformation API. Instead of dmnsn_transform_vector(), we have: - dmnsn_transform_point() - dmnsn_transform_direction() - dmnsn_transform_normal()
* Update copyright years.Tavian Barnes2011-08-311-1/+1
|
* Remove .bench files on make clean.Tavian Barnes2011-08-261-0/+3
|
* Use DMNSN_INTERNAL for prtree functions.Tavian Barnes2011-08-252-1/+4
|
* Rename dimension-impl.h to dimension-internal.hTavian Barnes2011-08-251-1/+1
|
* Use propper extensions for tests.Tavian Barnes2011-06-141-12/+12
|
* Vast libdimension API and internals improvements.Tavian Barnes2011-06-131-5/+7
| | | | | Couldn't really do these while I was trying to be POV-Ray compatible, 'cause they would've broken compatibility.
* Rename to dmnsn_vector_normalized().Tavian Barnes2011-05-181-3/+3
|
* Group tests and benchmarks with the corresponding source code.Tavian Barnes2011-05-155-0/+521