archive/blender-archive
Archived
1
1
Fork 0
Code Pull Requests Packages Activity
65,488 Commits 660 Branches 0 Tags
7bc9353957f495750445492dee8137321693de4e
Commit Graph

2979 Commits

Author SHA1 Message Date
Bastien Montagne a67ca5e30c Cleanup unused return value (was making coverity sad). 2016-08-16 18:22:03 +02:00
Sergey Sharybin a74dab86ee Fix redundant declarations after recent changes in GPU debug 2016-08-16 14:40:23 +02:00
Sergey Sharybin b1d0432bb7 Fix strict compiler flags with older GCC 2016-08-01 18:01:27 +02:00
Campbell Barton 710ab5be36 Cleanup: spelling, style 2016-07-31 17:41:05 +10:00
Campbell Barton 4e5ee5b75e Fix own error in recent heap update 2016-07-21 00:35:42 +10:00
Campbell Barton bd59206b5c Cleanup: style, spelling 2016-07-19 09:16:38 +10:00
Campbell Barton c974def837 BLI_heap: replace memarena w/ local allocator 
- Since element size its known it's less work to do inline.
- In test with high-poly model, gave ~9% overall speedup for decimate modifier.
 2016-07-17 19:29:30 +10:00
Campbell Barton c57d823683 Cleanup: minor edits to BLI_heap 2016-07-17 19:29:30 +10:00
Campbell Barton b32408eff8 BLI_math: move interp_*_cubic to its own function 2016-07-11 17:16:35 +10:00
Campbell Barton 7a3ea87bbf Cleanup: use normalize_v#_length 2016-07-08 10:14:49 +10:00
Campbell Barton 6035cf05bf BLI_math: add normalize functions which fit to a length 
Convenient since its common to normalize then scale,
since these are inlined, use for regular normalize w/ 1.0 length.
 2016-07-08 09:52:29 +10:00
Campbell Barton f36741c2eb Fix T48793: Bilinear filter clamps at edge pixels 2016-07-08 02:14:54 +10:00
Brecht Van Lommel 1ba90582f5 Fix a few compiler warnings on OS X / clang. 
Two were actual bugs, though they existed only in unused code:
* In Freestyle it was unintentionally copying a scene rather than referencing it.
* In BLI_array_store_is_valid there was use of uninitialized memory.
 2016-07-04 18:22:10 +02:00
Campbell Barton 9f5621bb4a Cleanup: comment blocks 2016-07-02 10:08:33 +10:00
Germano Cavalcante b01a56ee5c Transform Snap: Optimize edge-snap using BVH tree 
changes in BLI_kdopbvh:

- `BLI_bvhtree_find_nearest_to_ray` now takes is_ray_normalized and scale argument.
- `BLI_bvhtree_find_nearest_to_ray_angle` has been added (use for perspective view).

changes in BLI_bvhutils:

- `bvhtree_from_editmesh_edges_ex` was added.

changes in math_geom:

- `dist_squared_ray_to_seg_v3` was added.

other changes:

- `do_ray_start_correction` is no longer necessary to snap to verts.
- the way in which the test of depth was done before is being simulated in callbacks.
 2016-06-30 17:52:03 +10:00
Campbell Barton 0971749f4c Cleanup: spelling, indentation 2016-06-29 20:37:54 +10:00
Joshua Leung 14f056144e Bendy Bones Instability Fix - Second Attempt 
So the error seems to be in cubic_tangent_factor_circle_v3(),
which was introduced with D2001.

I've tweaked the most obvious culprit here - the epsilon factor.
It used to be 10^-7, but I've reduced it down to 10^-5 now,
and it's looking a lot more stable now :)

---------

BTW, about the derivation of the magic 0.390464 factor I briefly subbed back
as a workaround for this bug, see:
    http://www.whizkidtech.redprince.net/bezier/circle/
 2016-06-28 02:52:20 +12:00
Campbell Barton c7a6ff0981 Docs: arg names 2016-06-27 13:21:14 +10:00
Campbell Barton 475b7f5a2c BLI_array_utils: add BLI_array_rfindindex 
Array search from back to front.
 2016-06-23 11:36:49 +10:00
Campbell Barton 493c6b622f Cleanup: style 2016-06-22 14:02:51 +10:00
Campbell Barton 2465bd90d5 Cleanup: style, whitespace, doxy filepaths 2016-06-19 06:33:29 +10:00
Campbell Barton 5ea27bec1f BMesh Decimate: use doubles to calculate optimized position 
This allows the error threshold for calculating the optimized location to be much lower.

Resolves visible artifacts w/ 1m-tri happy-buddha example.
 2016-06-16 19:25:02 +10:00
Campbell Barton 47a5d7d1bc BLI_math: Add double versions of functions 
- mul_v3_m3v3_db
- mul_m3_v3_db
- negate_v3_db
 2016-06-16 19:20:08 +10:00
Campbell Barton 66b12ef4ab BLI_math: cleanup arg names 
project functions arg naming made it hard to tell which vector was projected onto.
 2016-06-12 15:39:04 +10:00
Campbell Barton dccf5afbef BLI_rand: add BLI_rng_get_char_n 
Use to fill an array of bytes to random values.
 2016-06-12 00:41:02 +10:00
Bastien Montagne d733826708 Fix T48614: Blender from buildbot crash when Separate selection in this particular scene. 
Regression from recent rB2c5dc66d5effd4072f438afb, if last item of last chunk of a mempool was valid,
it would not be returned by mempool iterator step, which would always return NULL in that case.
 2016-06-09 17:53:51 +02:00
Sergey Sharybin 37fc4b575f Fix FPE exception happening when converting linear<->srgb using SIMD 2016-06-08 16:00:34 +02:00
Campbell Barton 6ca6d3c4fd Cleanup: typo 2016-06-08 22:31:35 +10:00
Campbell Barton e02679f71e Cleanup: typos 2016-06-08 22:25:23 +10:00
Campbell Barton 0a029e3dd1 BLI_array_store: move helper functions into their own API 2016-06-08 19:12:23 +10:00
Campbell Barton 64663b1f73 Cleanup: warnings in previous commit 2016-06-02 19:48:45 +10:00
Campbell Barton cc7b817099 Minor edits to last commit 
Failed with chunk merging disabled
 2016-06-02 18:42:09 +10:00
Campbell Barton 0ce98b1ffb BLI_array_store: Move writing many chunks into a function 
Minor optimization, avoid some checks each iteration.
 2016-06-02 18:13:13 +10:00
Campbell Barton 7980c7c10f BLI_array_store: store max size in BArrayInfo 2016-06-02 18:05:11 +10:00
Campbell Barton 0bd8d6d194 Add extra validation checks to array-store 2016-06-02 16:41:41 +10:00
Bastien Montagne 57d5ddc251 Revert "BLI_ghash: Fix initial over-allocation of mempool chunks." 
Useless change in fact, sorry for the noise.

This reverts commit b08473680e.
 2016-06-01 17:38:50 +02:00
Campbell Barton 2c5dc66d5e Optimize mempool iteration 
Around ~10% improvement in own tests.
 2016-06-02 00:07:18 +10:00
Campbell Barton 8cf8679b53 Revert "Correct invalid pointer-pair compare check" 
This reverts commit d5e0e681ce.

Tsk, these functions return false on a match.
 2016-06-01 23:08:40 +10:00
Bastien Montagne b08473680e BLI_ghash: Fix initial over-allocation of mempool chunks. 
Code intended to create only one pool by default here, but code in `mempool_maxchunks()` would make it two.
 2016-06-01 12:58:59 +02:00
Campbell Barton aedeca7d1c BLI_mempool: Use an 'odd' FREEWORD for big/little endian 
This also changes freeword to an intptr_t to ensure
not only the first 4 bits of a pointer are tested on 64bit systems.
 2016-06-01 02:54:47 +10:00
Campbell Barton 3df30c1a6e Cleanup: parenthesize defines 2016-06-01 00:19:01 +10:00
Campbell Barton 665cb1b291 Change the hash-table to be 3x total items to hash 2016-05-30 18:00:03 +10:00
Campbell Barton bd6a64ced7 Remove accidental static var 2016-05-30 17:27:06 +10:00
Campbell Barton 53b60eed45 Add BLI_array_store copy-on-write API 
This supported in-memory de-duplication,
useful to avoid in-efficient memory use when storing multiple, similar arrays.
 2016-05-30 16:18:24 +10:00
Campbell Barton d5e0e681ce Correct invalid pointer-pair compare check 2016-05-26 22:20:12 +10:00
Campbell Barton a17cba339c BLI_math: Add function to calculate circular cubic curve tangents 2016-05-23 21:35:54 +10:00
Bastien Montagne 22ff9c5568 Fix T48497: Stupid typo in recent own BLI_task forloop work that broke non-parallelized case. 2016-05-22 18:35:44 +02:00
Brecht Van Lommel 2630207ada Fix GCC/Linux build error after finite/isfinite changes. 2016-05-17 23:40:25 +02:00
Brecht Van Lommel 21fddf7d1c C99/C++11: replace deprecated finite() by isfinite(). 2016-05-17 21:39:16 +02:00
Bastien Montagne 688858d3a8 BLI_task: Add new 'BLI_task_parallel_range_finalize()'. 
Together with the extended loop callback and userdata_chunk, this allows to perform
cumulative tasks (like aggregation) in a lockfree way using local userdata_chunk to store temp data,
and once all workers have finished, to merge those userdata_chunks in the finalize callback
(from calling thread, so no need to lock here either).

Note that this changes how userdata_chunk is handled (now fully from 'main' thread,
which means a given worker thread will always get the same userdata_chunk, without
being re-initialized anymore to init value at start of each iter chunk).
 2016-05-16 17:15:18 +02:00