mirror/eigen
2
0
Fork 0
mirror of https://gitlab.com/libeigen/eigen.git synced 2024-12-27 07:29:52 +08:00
Code Issues Packages Projects Releases Wiki Activity
8,968 Commits 11 Branches 89 Tags 168 MiB
a432fc102d
Commit Graph

8968 Commits

This Branch
This Branch
All Branches
Author SHA1 Message Date
Benoit Steiner
a432fc102d Moved the choice of ThreadPool to unsupported/Eigen/CXX11/ThreadPool 2016-12-12 15:24:16 -08:00
Benoit Steiner
8ae68924ed Made ThreadPoolInterface::Cancel() an optional functionality 2016-12-12 11:58:38 -08:00
Gael Guennebaud
57acb05eef Update and extend doc on alignment issues. 2016-12-11 22:45:32 +01:00
Benoit Steiner
76fca22134 Use a more accurate timer to sleep on Linux systems. 2016-12-09 15:12:24 -08:00
Benoit Steiner
4deafd35b7 Introduce a portable EIGEN_SLEEP macro. 2016-12-09 14:52:15 -08:00
Benoit Steiner
aafa97f4d2 Fixed build error with MSVC 2016-12-09 14:42:32 -08:00
Benoit Steiner
2f5b7a199b Reworked the threadpool cancellation mechanism to not depend on pthread_cancel since it turns out that pthread_cancel doesn't work properly on numerous platforms. 2016-12-09 13:05:14 -08:00
Benoit Steiner
3d59a47720 Added a message to ease the detection of platforms on which thread cancellation isn't supported. 2016-12-08 14:51:46 -08:00
Benoit Steiner
28ee8f42b2 Added a Flush method to the RunQueue 2016-12-08 14:07:56 -08:00
Benoit Steiner
69ef267a77 Added the new threadpool cancel method to the threadpool interface based class. 2016-12-08 14:03:25 -08:00
Benoit Steiner
7bfff85355 Added support for thread cancellation on Linux 2016-12-08 08:12:49 -08:00
Benoit Steiner
6811e6cf49 Merged in srvasude/eigen/fix_cuda_exp (pull request PR-268) 
Fix expm1 CUDA implementation (do not shadow exp CUDA implementation).
 2016-12-08 05:14:11 -08:00
Gael Guennebaud
747202d338 typo 2016-12-08 12:48:15 +01:00
Gael Guennebaud
bb297abb9e make sure we use the right eigen version 2016-12-08 12:00:11 +01:00
Gael Guennebaud
8b4b00d277 fix usage of custom compiler 2016-12-08 11:59:39 +01:00
Gael Guennebaud
7105596899 Add missing include and use -O3 2016-12-07 16:56:08 +01:00
Gael Guennebaud
780f3c1adf Fix call to convert on linux 2016-12-07 16:30:11 +01:00
Gael Guennebaud
3855ab472f Cleanup file structure 2016-12-07 14:23:49 +01:00
Gael Guennebaud
59a59fa8e7 Update perf monitoring scripts to generate html/svg outputs 2016-12-07 13:36:56 +01:00
Gael Guennebaud
f2c506b03d Add a script example to run and upload performance tests 2016-12-06 16:46:52 +01:00
Gael Guennebaud
1b4e085a7f generate png file for web upload 2016-12-06 16:46:22 +01:00
Gael Guennebaud
f725f1cebc Mention the CMAKE_PREFIX_PATH variable. 2016-12-06 15:23:45 +01:00
Gael Guennebaud
f90c4aebc5 Update monitored changeset lists 2016-12-06 15:07:46 +01:00
Gael Guennebaud
eb621413c1 Revert vec/y to vec*(1/y) in row-major TRSM: 
- div is extremely costly
- this is consistent with the column-major case
- this is consistent with all other BLAS implementations
 2016-12-06 15:04:50 +01:00
Gael Guennebaud
8365c2c941 Fix BLAS backend for symmetric rank K updates. 2016-12-06 14:47:09 +01:00
Gael Guennebaud
0c4d05b009 Explain how to choose your favorite Eigen version 2016-12-06 11:34:06 +01:00
Silvio Traversaro
e049a2a72a Added relocatable cmake support also for CMake before 3.0 and after 2.8.8 2016-12-06 10:37:34 +01:00
Srinivas Vasudevan
e6c8b5500c Change comparisons to use Scalar instead of RealScalar. 2016-12-05 14:01:45 -08:00
Srinivas Vasudevan
f7d7c33a28 Fix expm1 CUDA implementation (do not shadow exp CUDA implementation). 2016-12-05 12:19:01 -08:00
Silvio Traversaro
18481b518f Make CMake config file relocatable 2016-12-05 10:39:52 +01:00
Gael Guennebaud
c68c8631e7 fix compilation of BTL's blaze interface 2016-12-05 23:02:16 +01:00
Gael Guennebaud
1ff1d4a124 Add performance monitoring for LLT 2016-12-05 23:01:52 +01:00
Srinivas Vasudevan
09ee7f0c80 Fix small nit where I changed name of plog1p to pexpm1. 2016-12-02 15:30:12 -08:00
Srinivas Vasudevan
a0d3ac760f Sync from Head. 2016-12-02 14:14:45 -08:00
Srinivas Vasudevan
218764ee1f Added support for expm1 in Eigen. 2016-12-02 14:13:01 -08:00
Gael Guennebaud
66f65ccc36 Ease compiler job to generate clean and efficient code in mat*vec. 2016-12-02 22:41:26 +01:00
Gael Guennebaud
fe696022ec Operators += and -= do not resize! 2016-12-02 22:40:25 +01:00
Angelos Mantzaflaris
18de92329e use numext::abs 
(grafted from 0a08d4c60b
)
 2016-12-02 11:48:06 +01:00
Angelos Mantzaflaris
e8a6aa518e 1. Add explicit template to abs2 (resolves deduction for some arithmetic types) 
2. Avoid signed-unsigned conversion in comparison (warning in case Scalar is unsigned)
(grafted from 4086187e49
)
 2016-12-02 11:39:18 +01:00
Gael Guennebaud
a6b971e291 Fix memory leak in Ref<Sparse> 2016-12-05 16:59:30 +01:00
Gael Guennebaud
8640ffac65 Optimize SparseLU::solve for rhs vectors 2016-12-05 15:41:14 +01:00
Gael Guennebaud
62acd67903 remove temporary in SparseLU::solve 2016-12-05 15:11:57 +01:00
Gael Guennebaud
0db6d5b3f4 bug #1356: fix calls to evaluator::coeffRef(0,0) to get the address of the destination 
by adding a dstDataPtr() member to the kernel. This fixes undefined behavior if dst is empty (nullptr).
 2016-12-05 15:08:09 +01:00
Gael Guennebaud
91003f3b86 typo 2016-12-05 13:51:07 +01:00
Gael Guennebaud
445c015751 extend monitoring benchmarks with transpose matrix-vector and triangular matrix-vectors. 2016-12-05 13:36:26 +01:00
Gael Guennebaud
e3f613cbd4 Improve performance of row-major-dense-matrix * vector products for recent CPUs. 
This revised version does not bother about aligned loads/stores,
and rather processes 8 rows at ones for better instruction pipelining.
 2016-12-05 13:02:01 +01:00
Gael Guennebaud
3abc827354 Clean debugging code 2016-12-05 12:59:32 +01:00
Benoit Steiner
462c28e77a Merged in srvasude/eigen (pull request PR-265) 
Add Expm1 support to Eigen.
 2016-12-05 02:31:11 +00:00
Gael Guennebaud
4465d20403 Add missing generic load methods. 2016-12-03 21:25:04 +01:00
Gael Guennebaud
6a5fe86098 Complete rewrite of column-major-matrix * vector product to deliver higher performance of modern CPU. 
The previous code has been optimized for Intel core2 for which unaligned loads/stores were prohibitively expensive.
This new version exhibits much higher instruction independence (better pipelining) and explicitly leverage FMA.
According to my benchmark, on Haswell this new kernel is always faster than the previous one, and sometimes even twice as fast.
Even higher performance could be achieved with a better blocking size heuristic and, perhaps, with explicit prefetching.
We should also check triangular product/solve to optimally exploit this new kernel (working on vertical panel of 4 columns is probably not optimal anymore).
 2016-12-03 21:14:14 +01:00