Commit message (Collapse) | Author | Age | |
---|---|---|---|
* | Rename predux_downto4 to be more accurate on its semantic. | 2018-04-03 | |
| | |||
* | Fix unit testing of predux_downto4 (bad name), and add unit testing of prsqrt | 2018-04-03 | |
| | |||
* | MIsc. source and comment typos | 2018-03-11 | |
| | | | | Found using `codespell` and `grep` from downstream FreeCAD | ||
* | Added support for expm1 in Eigen. | 2016-12-02 | |
| | |||
* | replace sizeof(Packet) with PacketSize else it breaks for ZVector.Packet4f | 2016-11-17 | |
| | |||
* | Merged eigen/eigen into default | 2016-11-03 | |
|\ | |||
| * | Add pinsertfirst function and implement pinsertlast for complex on SSE/AVX. | 2016-11-02 | |
| | | |||
| * | Add a pinsertlast function replacing the last entry of a packet by a scalar. | 2016-10-25 | |
| | | | | | | | | (useful to vectorize LinSpaced) | ||
* | | Merged eigen/eigen into default | 2016-10-12 | |
|\| | |||
* | | Renamed predux_half into predux_downto4 | 2016-10-06 | |
| | | |||
* | | Merged latest updates from trunk | 2016-10-05 | |
|\ \ | |||
| | * | Fix a bug in the implementation of Carmack's fast sqrt algorithm in Eigen ↵ | 2016-10-04 | |
| |/ | | | | | | | | | | | | | | | | | | | | | | | | | | | (enabled by EIGEN_FAST_MATH), which causes the vectorized parts of the computation to return -0.0 instead of NaN for negative arguments. Benchmark speed in Giga-sqrts/s Intel(R) Xeon(R) CPU E5-1650 v3 @ 3.50GHz ----------------------------------------- SSE AVX Fast=1 2.529G 4.380G Fast=0 1.944G 1.898G Fast=1 fixed 2.214G 3.739G This table illustrates the worst case in terms speed impact: It was measured by repeatedly computing the sqrt of an n=4096 float vector that fits in L1 cache. For large vectors the operation becomes memory bound and the differences between the different versions almost negligible. | ||
| * | Add a note regarding gcc bug #72867 | 2016-09-22 | |
| | | |||
| * | Fix compilation in non C++11 mode. | 2016-08-23 | |
| | | |||
| * | Add log1p support for CUDA and half floats | 2016-08-08 | |
| | | |||
| * | Made the packetmath test compile again. A better fix would be to move the ↵ | 2016-07-11 | |
| | | | | | | | | special function tests to the unsupported directory where the code now resides. | ||
| * | Disabled GCC6's ignored-attributes warning in packetmath unit test. | 2016-05-26 | |
| | | |||
| * | Silenced several double-promotion warnings | 2016-05-22 | |
| | | |||
| * | Make EIGEN_HAS_C99_MATH user configurable | 2016-05-20 | |
| | | |||
| * | Extended the tests for ptanh | 2016-05-10 | |
| | | |||
| * | Enable and fix -Wdouble-conversion warnings | 2016-05-05 | |
| | | |||
* | | Extended the packetmath test to cover all the alignments made possible by ↵ | 2016-04-29 | |
| | | | | | | | | avx512 instructions. | ||
* | | Pull latest updates from upstream | 2016-04-11 | |
|\| | |||
| * | enable all tests again | 2016-04-05 | |
| | | |||
| * | actually include ZVector files, passes most basic tests (float still fails) | 2016-03-28 | |
| | | |||
| * | some primitives ported, but missing intrinsics and crash with asm() are a ↵ | 2016-03-27 | |
| | | | | | | | | problem | ||
| * | Added a regression test for tanh | 2016-02-10 | |
| | | |||
* | | Updated the packetmath test to call predux_half instead of predux4 | 2016-02-01 | |
| | | |||
| * | Merged in ebrevdo/eigen (pull request PR-148) | 2015-12-11 | |
|/| | | | | | | | Add special functions to eigen: lgamma, erf, erfc. | ||
| * | Fixed compilation error triggered by MSVC 2008 | 2015-12-10 | |
| | | |||
| * | Only implement the lgamma, erf, and erfc functions when using a compiler ↵ | 2015-12-10 | |
| | | | | | | | | compliant with the C99 specification. | ||
* | | Workaround gcc issue with -O3 and the i387 FPU. | 2015-12-10 | |
| | | |||
| * | Only disable the erf, erfc, and lgamma tests for older versions of c++. | 2015-12-07 | |
| | | |||
| * | Fixed a couple of typos | 2015-12-07 | |
| | | | | | | | | Cleaned up the code a bit. | ||
| * | Add special functions to Eigen: lgamma, erf, erfc. | 2015-12-07 | |
|/ | | | | Includes CUDA support and unit tests. | ||
* | Fix AVX round/ceil/floor, and fix respective unit test | 2015-11-04 | |
| | |||
* | Add round, ceil and floor for SSE4.1/AVX (Bug #70) | 2015-11-01 | |
| | |||
* | Fix packetmath unit test for pdiv not being always defined | 2015-10-13 | |
| | |||
* | Fix packetmath unit test for exp and log | 2015-09-02 | |
| | |||
* | merge | 2015-08-16 | |
|\ | |||
| * | Protect further isnan/isfinite/isinf calls | 2015-08-16 | |
| | | |||
* | | Fix prototype of plset and generalize linspace functor. | 2015-08-07 | |
|/ | |||
* | Many files were missing in previous changeset. | 2015-07-29 | |
| | |||
* | Moved some utilities to TensorMeta.h to make it easier to reuse them accross ↵ | 2015-06-29 | |
| | | | | | | several tensor operations. Created the TensorDimensionList class to encode the list of all the dimensions of a tensor of rank n. This could be done using TensorIndexList, however TensorIndexList require cxx11 which isn't yet supported as widely as we'd like. | ||
* | std::isnan is c++11 only | 2015-06-24 | |
| | |||
* | Added more checks to test the correctness of the pexp implementation | 2015-06-23 | |
| | |||
* | Rename free functions isFinite, isInf, isNaN to be compatible with c++11 | 2015-06-10 | |
| | |||
* | Remove ambiguity with recent numext methods isNaN and isInf | 2015-03-17 | |
| | |||
* | Pulled the latest changes from the trunk | 2015-02-06 | |
|\ | |||
| * | Addendum to bug #859: pexp(NaN) for double did not return NaN, also, ↵ | 2014-10-20 | |
| | | | | | | | | | | | | plog(NaN) did not return NaN. psqrt(NaN) and psqrt(-1) shall return NaN if EIGEN_FAST_MATH==0 |