Commit Graph

828 Commits

Author SHA1 Message Date
Benoit Steiner
5266ff8966 Cleaned up a regression test 2016-10-08 19:12:44 +00:00
RJ Ryan
bfc264abe8 Add a test that GPU complex product reductions match CPU reductions. 2016-10-06 11:10:14 -07:00
Benoit Steiner
4387433acf Increased the robustness of the reduction tests on fp16 2016-10-05 10:42:41 -07:00
Benoit Steiner
aad20d700d Increase the tolerance to numerical noise. 2016-10-05 10:39:24 -07:00
Benoit Steiner
616a7a1912 Improved support for compiling CUDA code with clang as the host compiler 2016-10-03 17:09:33 -07:00
Benoit Steiner
2bda1b0d93 Updated the tensor sum and mean reducer to enable them to process complex numbers on cuda gpus. 2016-09-28 17:08:41 -07:00
RJ Ryan
608b1acd6d Don't use c++11 features and fix include. 2016-09-20 07:49:05 -07:00
RJ Ryan
b2c6dc48d9 Add CUDA-specific std::complex<T> specializations for scalar_sum_op, scalar_difference_op, scalar_product_op, and scalar_quotient_op. 2016-09-20 07:18:20 -07:00
Emil Fresk
6edd2e2851 Made AutoDiffJacobian more intuitive to use and updated for C++11
Changes:
* Removed unnecessary types from the Functor by inferring from its types
* Removed inputs() function reference, replaced with .rows()
* Updated the forward constructor to use variadic templates
* Added optional parameters to the Fuctor for passing parameters,
  control signals, etc
* Has been tested with fixed size and dynamic matricies

Ammendment by chtz: overload operator() for compatibility with not fully conforming compilers
2016-09-16 14:03:55 +02:00
Benoit Steiner
e4d4d15588 Register the cxx11_tensor_device only for recent cuda architectures (i.e. >= 3.0) since the test instantiate contractions that require a modern gpu. 2016-09-12 19:01:52 -07:00
Benoit Steiner
4dfd888c92 CUDA contractions require arch >= 3.0: don't compile the cuda contraction tests on older architectures. 2016-09-12 18:49:01 -07:00
Benoit Steiner
028e299577 Fixed a bug impacting some outer reductions on GPU 2016-09-12 18:36:52 -07:00
Benoit Steiner
5f50f12d2c Added the ability to compute the absolute value of a complex number on GPU, as well as a test to catch the problem. 2016-09-12 13:46:13 -07:00
Benoit Steiner
8321dcce76 Merged latest updates from trunk 2016-09-12 10:33:05 -07:00
Benoit Steiner
eb6ba00cc8 Properly size the list of waiters 2016-09-12 10:31:55 -07:00
Gael Guennebaud
dabc81751f Fix compilation when cuda_fp16.h does not exist. 2016-09-05 17:14:20 +02:00
Benoit Steiner
87a8a1975e Fixed a regression test 2016-09-02 19:29:33 -07:00
Benoit Steiner
6c05c3dd49 Fix the cxx11_tensor_cuda.cu test on 32bit platforms. 2016-09-02 11:12:16 -07:00
Benoit Steiner
039e225f7f Added a test for nullary expressions on CUDA
Also check that we can mix 64 and 32 bit indices in the same compilation unit
2016-09-01 13:28:12 -07:00
Benoit Steiner
c53f783705 Updated the contraction code to support constant inputs. 2016-09-01 11:41:27 -07:00
Gael Guennebaud
72a4d49315 Fix compilation with CUDA 8 2016-09-01 13:39:33 +02:00
Gael Guennebaud
1f84f0d33a merge EulerAngles module 2016-08-30 10:01:53 +02:00
Gael Guennebaud
e074f720c7 Include missing forward declaration of SparseMatrix 2016-08-29 18:56:46 +02:00
Gael Guennebaud
6cd7b9ea6b Fix compilation with cuda 8 2016-08-29 11:06:08 +02:00
Gael Guennebaud
0f56b5a6de enable vectorization path when testing half on cuda, and add test for log1p 2016-08-26 14:55:51 +02:00
Igor Babuschkin
0425118e2a Merge upstream changes 2016-08-05 14:34:57 +01:00
Benoit Steiner
5eea1c7f97 Fixed cut and paste bug in debud message 2016-08-04 17:34:13 -07:00
Benoit Steiner
b50d8f8c4a Extended a regression test to validate that we basic fp16 support works with cuda 7.0 2016-08-03 16:50:13 -07:00
Benoit Steiner
fad9828769 Deleted redundant regression test. 2016-08-03 16:08:37 -07:00
Benoit Steiner
d92df04ce8 Cleaned up the new float16 test a bit 2016-08-03 11:50:07 -07:00
Benoit Steiner
81099ef482 Added a test for fp16 2016-08-03 11:41:17 -07:00
Gael Guennebaud
cc2f6d68b1 bug #1264: fix compilation 2016-07-27 23:30:47 +02:00
Gael Guennebaud
8972323c08 Big 1261: add missing max(ADS,ADS) overload (same for min) 2016-07-27 14:52:48 +02:00
Gael Guennebaud
5d94dc85e5 bug #1260: add regression test 2016-07-27 14:38:30 +02:00
Gael Guennebaud
0d7039319c bug #1260: remove doubtful specializations of ScalarBinaryOpTraits 2016-07-27 14:35:52 +02:00
Gael Guennebaud
fd1117f2be Implement digits10 for mpreal 2016-07-25 14:38:55 +02:00
Benoit Steiner
c6b0de2c21 Improved partial reductions in more cases 2016-07-22 17:18:20 -07:00
Gael Guennebaud
32d95e86c9 merge 2016-07-22 16:43:12 +02:00
Gael Guennebaud
d7a0e52478 Fix testing of log nearby 1 2016-07-22 15:44:26 +02:00
Gael Guennebaud
7acf23c14c Truely split unit test. 2016-07-22 15:41:23 +02:00
Gael Guennebaud
d075d122ea Move half unit test from unsupported to main tests 2016-07-22 14:34:19 +02:00
Gael Guennebaud
82798162c0 Extend unit testing of half with ADL and arrays. 2016-07-21 15:47:21 +02:00
Gael Guennebaud
c98bac2966 Manually add -stdd=c++11 to nvcc for old cmake versions 2016-07-12 09:29:18 +02:00
Benoit Steiner
40eb97516c reverted unintended change. 2016-07-11 14:28:03 -07:00
Benoit Steiner
03b71c273e Made the packetmath test compile again. A better fix would be to move the special function tests to the unsupported directory where the code now resides. 2016-07-11 13:50:24 -07:00
Gael Guennebaud
fd60966310 merge 2016-07-11 18:11:47 +02:00
Gael Guennebaud
7d636349dc Fix configuration of CUDA:
- preserve user defined CUDA_NVCC_FLAGS
 - remove the -ansi flag that conflicts with -std=c++11
 - do not add -std=c++11 if already there
2016-07-11 18:09:04 +02:00
Gael Guennebaud
131ee4bb8e Split test_slice_in_expr which seems to be huge for visual 2016-07-11 11:46:55 +02:00
Gael Guennebaud
544935101a Fix warnings 2016-07-08 11:38:52 +02:00
Gael Guennebaud
59bf2774a3 Fix warnings 2016-07-08 11:38:11 +02:00