Commit message (Collapse) | Author | Age | |
---|---|---|---|
* | Don't put a command at the end of an enumerator list | 2016-04-12 | |
| | |||
* | Pulled latest update from trunk | 2016-04-12 | |
|\ | |||
* | | Fixed the vectorization logic test | 2016-04-12 | |
| | | |||
* | | Turned on the contraction benchmarks for fp16 | 2016-04-12 | |
| | | |||
| * | Enable the use of half-packet in coeff-based product. | 2016-04-12 | |
|/ | | | | For instance, Matrix4f*Vector4f is now vectorized again when using AVX. | ||
* | Fixed the zeta test | 2016-04-12 | |
| | |||
* | Defer the decision to vectorize tensor CUDA code to the meta kernel. This ↵ | 2016-04-12 | |
| | | | | makes it possible to decide to vectorize or not depending on the capability of the target cuda architecture. In particular, this enables us to vectorize the processing of fp16 when running on device of capability >= 5.3 | ||
* | More accurate cost estimates for exp, log, tanh, and sqrt. | 2016-04-11 | |
| | |||
* | Added epsilon, dummy_precision, infinity and quiet_NaN NumTraits for fp16 | 2016-04-11 | |
| | |||
* | Pulled latest update from trunk | 2016-04-11 | |
|\ | |||
| * | Update doc regarding the genericity of EIGEN_USE_BLAS | 2016-04-11 | |
| | | |||
| * | Improve constness of level2 blas API. | 2016-04-11 | |
| | | |||
| * | Move LAPACK declarations from blas.h to lapack.h and fix compatibility with ↵ | 2016-04-11 | |
| | | | | | | | | EIGEN_USE_MKL | ||
| * | Cleanup obsolete assign_scalar_eig2mkl helper. | 2016-04-11 | |
| | | |||
| * | Remove all references to MKL in BLAS wrappers. | 2016-04-11 | |
| | | |||
| * | Fix long to int conversion in BLAS API. | 2016-04-11 | |
| | | |||
| * | Silent unused warning. | 2016-04-11 | |
| | | |||
| * | Relax dependency on MKL for EIGEN_USE_BLAS | 2016-04-11 | |
| | | |||
| * | Improve constness of blas level-3 interface. | 2016-04-11 | |
| | | |||
| * | Fix static/inline keywords order. | 2016-04-11 | |
| | | |||
| * | Typos in doc. | 2016-04-11 | |
| | | |||
| * | Proper handling of domain errors. | 2016-04-10 | |
| | | |||
| * | Merge upstream | 2016-04-09 | |
| |\ | |||
| * | | Added nans to zeta function. | 2016-04-09 | |
| | | | |||
| | * | bug #1197: fix/relax some LM unit tests | 2016-04-09 | |
| | | | |||
| | * | bug #1160: fix and relax some lm unit tests by turning faillures to warnings | 2016-04-09 | |
| | | | |||
* | | | Use numext::abs instead of std::abs in scalar_fuzzy_default_impl to make it ↵ | 2016-04-08 | |
| |/ |/| | | | | | usable inside GPU kernels. | ||
* | | Fixed the implementation of Eigen::numext::isfinite, Eigen::numext::isnan, ↵ | 2016-04-08 | |
| | | | | | | | | andEigen::numext::isinf on CUDA devices | ||
* | | Fixed isfinite_impl: NumTraits<T>::highest() and NumTraits<T>::lowest() are ↵ | 2016-04-08 | |
| | | | | | | | | finite numbers. | ||
* | | Disabled the use of half2 on cuda devices of compute capability < 5.3 | 2016-04-08 | |
| | | |||
* | | Initial support for taking the power of fp16 | 2016-04-08 | |
| | | |||
* | | Fixed the packet_traits for half floats. | 2016-04-08 | |
| | | |||
* | | Created the new EIGEN_TEST_CUDA_CLANG option to compile the CUDA tests using ↵ | 2016-04-08 | |
| | | | | | | | | clang instead of nvcc | ||
* | | Don't test the division by 0 on float16 when compiling with msvc since msvc ↵ | 2016-04-08 | |
| | | | | | | | | detects and errors out on divisions by 0. | ||
* | | Add references to filed LLVM bugs | 2016-04-08 | |
| | | |||
* | | Properly handle complex numbers. | 2016-04-07 | |
| | | |||
* | | Turn on the coeffWise benchmarks on fp16 | 2016-04-07 | |
| | | |||
* | | Fixed the type casting benchmarks for fp16 | 2016-04-07 | |
| | | |||
* | | Merged in parthaEth/eigen (pull request PR-175) | 2016-04-07 | |
|\ \ | | | | | | | | | | Static casting scalar types so as to let chlesky module of eigen work with ceres | ||
* | | | Renamed float16 into cxx11_float16 since the test relies on c++11 features | 2016-04-07 | |
| | | | |||
* | | | Added missing EIGEN_DEVICE_FUNC to the tensor conversion code. | 2016-04-07 | |
| | | | |||
* | | | Fixed the benchmarking of fp16 coefficient wise operations | 2016-04-07 | |
| | | | |||
| * | | Static casting scalar types so as to let chlesky module of eigen work with ceres | 2016-04-08 | |
|/ / | |||
* | | Worked around numerical noise in the test for the zeta function. | 2016-04-07 | |
| | | |||
* | | Fixed a typo in the polygamma test. | 2016-04-07 | |
| | | |||
* | | Updated the unary functors to use the numext implementation of typicall ↵ | 2016-04-07 | |
| | | | | | | | | functions instead of the one provided in the standard library. The standard library functions aren't supported officially by cuda, so we're better off using the numext implementations. | ||
* | | Move the functions operating on fp16 out of the std namespace and into the ↵ | 2016-04-07 | |
| | | | | | | | | Eigen::numext namespace | ||
* | | Added tests for float16 | 2016-04-07 | |
| | | |||
* | | Fixed a typo in a test | 2016-04-07 | |
| | | |||
* | | Updated the isnan, isinf and isfinite functions to make compatible with cuda ↵ | 2016-04-07 | |
| | | | | | | | | devices. |