Commit message (Collapse) | Author | Age | |
---|---|---|---|
* | Removed executable bit from header files | 2016-03-23 | |
| | |||
* | Fixed contractions of fp16 | 2016-03-23 | |
| | |||
* | Made type conversion explicit | 2016-03-23 | |
| | |||
* | Added the ability to divide a half float by an index | 2016-03-23 | |
| | |||
* | Added more conversion operators for half floats | 2016-03-23 | |
| | |||
* | Fixed contractions of 16 bit floats | 2016-03-22 | |
| | |||
* | Completed the implementation of vectorized type casting of half floats. | 2016-03-18 | |
| | |||
* | Make all the conversions explicit | 2016-03-18 | |
| | |||
* | Implemented some of the missing type casting for half floats | 2016-03-17 | |
| | |||
* | Merged in ebrevdo/eigen (pull request PR-169) | 2016-03-16 | |
|\ | | | | | | | Bugfixes to cuda tests, igamma & igammac implemented, & tests for digamma, igamma, igammac on CPU & GPU. | ||
| * | Change the header guard around certain numext functions to be CUDA specific. | 2016-03-16 | |
| | | |||
* | | Fixed a typo. | 2016-03-14 | |
| | | |||
* | | Properly gate the use of cuda intrinsics in the code | 2016-03-14 | |
| | | |||
* | | Make sure we only use the half float intrinsic when compiling with a version ↵ | 2016-03-14 | |
| | | | | | | | | of CUDA that is recent enough to provide them | ||
* | | Don't mark the cast operator as explicit, since this is a c++11 feature ↵ | 2016-03-12 | |
| | | | | | | | | that's not supported by older compilers. | ||
* | | Also replaced uint32_t with unsigned int to make the code more portable | 2016-03-11 | |
| | | |||
* | | Replaced a couple more uint16_t with unsigned short | 2016-03-11 | |
| | | |||
* | | Use unsigned short instead of uint16_t since they're more portable | 2016-03-11 | |
| | | |||
* | | Made half floats usable on hardware that doesn't support them natively. | 2016-03-11 | |
| | | |||
* | | Fixed the +=, -=, *= and /= operators to return a reference | 2016-03-10 | |
| | | |||
| * | Update MathFunctions/SpecialFunctions with intelligent header guards. | 2016-03-09 | |
| | | |||
| * | Properly fix merge issues. | 2016-03-08 | |
| |\ | |||
| * | | Resolve bad merge. | 2016-03-08 | |
| | | | |||
| * | | Add certain functions to numext (log, exp, tan) because CUDA doesn't support ↵ | 2016-03-08 | |
|/ / | | | | | | | | | | | std:: Use these in SpecialFunctions. | ||
| * | Finishing touches on igamma/igammac for GPU. Tests now pass. | 2016-03-07 | |
| | | |||
| * | Fix Eigen's building of sharded tests that use CUDA & more igamma/igammac ↵ | 2016-03-07 | |
| | | | | | | | | | | | | | | | | | | | | | | | | | | bugfixes. 0. Prior to this PR, not a single sharded CUDA test was actually being *run*. Fixed that. GPU tests are still failing for igamma/igammac. 1. Add calls for igamma/igammac to TensorBase 2. Fix up CUDA-specific calls of igamma/igammac 3. Add unit tests for digamma, igamma, igammac in CUDA. | ||
* | | Turn on some of the cxx11 features when compiling with visual studio 2015 | 2016-03-05 | |
| | | |||
| * | Make igamma and igammac work correctly. | 2016-03-04 | |
| | | | | | | | | | | This required replacing ::abs with std::abs. Modified some unit tests. | ||
| * | Initial implementation of igamma and igammac. | 2016-03-03 | |
| | | |||
* | | Enable partial support for half floats on Kepler GPUs. | 2016-03-03 | |
| | | |||
* | | Enable the conversion between floats and half floats on older GPUs that ↵ | 2016-03-03 | |
| | | | | | | | | support it. | ||
* | | Merged in ebrevdo/eigen (pull request PR-167) | 2016-03-03 | |
|\| | | | | | | | | | | | Add infinity() support to numext::numeric_limits, use it in lgamma. I tested the code on my gtx-titan-black gpu, and it appears to work as expected. | ||
| * | Small bugfix to numeric_limits for CUDA. | 2016-03-02 | |
| | | |||
| * | Add infinity() support to numext::numeric_limits, use it in lgamma. | 2016-03-02 | |
| | | | | | | | | | | This makes the infinity access a __device__ function, removing nvcc warnings. | ||
* | | bug #537: fix compilation with Apples's compiler | 2016-03-02 | |
| | | |||
* | | Compilation fix | 2016-03-01 | |
| | | |||
* | | Compilation fix | 2016-03-01 | |
| | | |||
* | | Fix shortcoming in fixed-value deduction of startRow/startCol | 2016-02-29 | |
| | | |||
* | | bug #1172: make valuePtr and innderIndexPtr properly return null for empty ↵ | 2016-02-27 | |
| | | | | | | | | matrices. | ||
* | | merge | 2016-02-23 | |
|\ \ | |||
* | | | Fix startRow()/startCol() for dense Block with direct access: | 2016-02-23 | |
| | | | | | | | | | | | | the initial implementation failed for empty rows/columns for which are ambiguous. | ||
| * | | Declare the half float type as arithmetic. | 2016-02-22 | |
| | | | |||
| * | | Implemented the ptranspose function on half floats | 2016-02-21 | |
| | | | |||
| * | | Added the ability to compute the absolute value of a half float | 2016-02-21 | |
|/ / | |||
* | | Moved some of the fp16 operators outside the Eigen namespace to workaround ↵ | 2016-02-20 | |
| | | | | | | | | some nvcc limitations. | ||
* | | merge | 2016-02-19 | |
|\ \ | |||
* | | | bug #1170: skip calls to memcpy/memmove for empty imput. | 2016-02-19 | |
| | | | |||
* | | | Fix nesting type and complete reflection methods of Block expressions. | 2016-02-19 | |
| | | | |||
* | | | Add typedefs for the return type of all block methods. | 2016-02-19 | |
| | | | |||
| * | | Added support for tensor reductions on half floats | 2016-02-19 | |
| | | |