Commit message (Collapse) | Author | Age | ||
---|---|---|---|---|
... | ||||
* | | Marked unused variables as such | Benoit Steiner | 2016-05-25 | |
| | | ||||
* | | Specify the rounding mode in the correct location | Benoit Steiner | 2016-05-25 | |
| | | ||||
* | | Updated the README file for the tensor benchmarks | Benoit Steiner | 2016-05-25 | |
| | | ||||
* | | Explicitly specify the rounding mode when converting floats to fp16 | Benoit Steiner | 2016-05-25 | |
| | | ||||
* | | Made the IndexPair code compile in non cxx11 mode | Benoit Steiner | 2016-05-25 | |
| | | ||||
* | | Made the index pair list code more portable accross various compilers | Benoit Steiner | 2016-05-25 | |
| | | ||||
* | | Improved the performance of tensor padding | Benoit Steiner | 2016-05-25 | |
| | | ||||
* | | Added support for statically known lists of pairs of indices | Benoit Steiner | 2016-05-25 | |
| | | ||||
* | | Disable the use of MMX instructions on x86_64 since too many compilers only ↵ | Benoit Steiner | 2016-05-25 | |
| | | | | | | | | support them in 32bit mode | |||
* | | Use numext:: instead of std:: functions. | Benoit Steiner | 2016-05-25 | |
| | | ||||
* | | Fix typo in dont_over_optimize | Gael Guennebaud | 2016-05-25 | |
| | | ||||
* | | Fix warning. | Gael Guennebaud | 2016-05-25 | |
| | | ||||
* | | Fix compilation with ICC. | Gael Guennebaud | 2016-05-25 | |
| | | ||||
* | | Fix static/inline order. | Gael Guennebaud | 2016-05-25 | |
| | | ||||
* | | There is no need to make the fp16 full reduction kernel a static function. | Benoit Steiner | 2016-05-24 | |
| | | ||||
* | | Fixed compilation warning | Benoit Steiner | 2016-05-24 | |
| | | ||||
* | | Cleaned up the fp16 code a little more | Benoit Steiner | 2016-05-24 | |
| | | ||||
* | | Pulled latest updates from trunk | Benoit Steiner | 2016-05-24 | |
|\ \ | ||||
* | | | Cleaned up the fp16 code | Benoit Steiner | 2016-05-24 | |
| | | | ||||
| * | | update doc wrt to unaligned vectorization | Gael Guennebaud | 2016-05-24 | |
| | | | ||||
| * | | Document EIGEN_UNALIGNED_VECTORIZE. | Gael Guennebaud | 2016-05-24 | |
| | | | ||||
| * | | Workaround clang/llvm bug in code generation. | Gael Guennebaud | 2016-05-24 | |
| | | | ||||
| * | | bug #256: enable vectorization with unaligned loads/stores. | Gael Guennebaud | 2016-05-24 | |
| | | | | | | | | | | | | | | | This concerns all architectures and all sizes. This new behavior can be disabled by defining EIGEN_UNALIGNED_VECTORIZE=0 | |||
| * | | Block<> should not disable vectorization based on inner-size, this is the ↵ | Gael Guennebaud | 2016-05-24 | |
| | | | | | | | | | | | | responsibilty of the assignment logic. | |||
| * | | Clean propagation of Dest/Src alignments. | Gael Guennebaud | 2016-05-24 | |
|/ / | ||||
* | | Remove now-unused protate PacketMath func | Benoit Jacob | 2016-05-24 | |
| | | ||||
* | | Remove the rotating kernel. It was only useful on some ARM CPUs (Qualcomm ↵ | Benoit Jacob | 2016-05-24 | |
| | | | | | | | | Krait) that are not as ubiquitous today as they were when I introduced it. | |||
* | | Don't attempt to use MMX instructions with visualstudio since they're only ↵ | Benoit Steiner | 2016-05-24 | |
| | | | | | | | | partially supported. | |||
* | | Worked around missing clang intrinsic | Benoit Steiner | 2016-05-24 | |
| | | ||||
* | | Use the generic ploadquad intrinsics since it does the job | Benoit Steiner | 2016-05-24 | |
| | | ||||
* | | Worked around missing clang intrinsics | Benoit Steiner | 2016-05-24 | |
| | | ||||
* | | Added missing EIGEN_DEVICE_FUNC qualifier | Benoit Steiner | 2016-05-23 | |
| | | ||||
* | | Include mmintrin.h to make it possible to use mmx instructions when needed. ↵ | Benoit Steiner | 2016-05-23 | |
| | | | | | | | | For example, this will enable the definition of a half packet for the Packet4f type. | |||
* | | Use the Index type instead of integers to specify the strides in ↵ | Benoit Steiner | 2016-05-23 | |
| | | | | | | | | pgather/pscatter | |||
* | | Added missing alignment in the fp16 packet traits | Benoit Steiner | 2016-05-23 | |
| | | ||||
* | | ptranspose is not a template. | Benoit Steiner | 2016-05-23 | |
| | | ||||
* | | Avoid unnecessary float to double conversions | Benoit Steiner | 2016-05-23 | |
| | | ||||
* | | Avoid unnecessary float to double conversion. | Benoit Steiner | 2016-05-23 | |
| | | ||||
* | | Started to vectorize the processing of 16bit floats on CPU. | Benoit Steiner | 2016-05-23 | |
| | | ||||
* | | Don't optimize the processing of the last rows of a matrix matrix product in ↵ | Benoit Steiner | 2016-05-23 | |
| | | | | | | | | cases that violate the assumptions made by the optimized code path. | |||
* | | Fixed a typo in the array.cpp test | Benoit Steiner | 2016-05-23 | |
| | | ||||
* | | Merged in rmlarsen/eigen (pull request PR-188) | Benoit Steiner | 2016-05-23 | |
|\ \ | | | | | | | | | | Minor cleanups: 1. Get rid of a few unused variables. 2. Get rid of last uses of EIGEN_USE_COST_MODEL. | |||
* | | | Replace multiple constructors of half-type by a generic/templated ↵ | Christoph Hertzberg | 2016-05-23 | |
| | | | | | | | | | | | | constructor. This fixes an incompatibility with long double, exposed by the previous commit. | |||
* | | | Silenced several double-promotion warnings | Christoph Hertzberg | 2016-05-22 | |
| | | | ||||
* | | | fixed macro name | Christoph Hertzberg | 2016-05-22 | |
| | | | ||||
* | | | Fix some sign-compare warnings | Christoph Hertzberg | 2016-05-22 | |
| | | | ||||
* | | | Identify clang++ even if it is not named llvm-clang++ | Christoph Hertzberg | 2016-05-22 | |
| | | | ||||
* | | | Document EIGEN_MAX_CPP_VER and user controllable compiler features. | Gael Guennebaud | 2016-05-20 | |
| | | | ||||
* | | | Make EIGEN_HAS_CONSTEXPR user configurable | Gael Guennebaud | 2016-05-20 | |
| | | | ||||
* | | | Make EIGEN_HAS_VARIADIC_TEMPLATES user configurable | Gael Guennebaud | 2016-05-20 | |
| | | |