Commit message (Collapse) | Author | Age | |
---|---|---|---|
* | Make vectorized compute_inverse_size4 compile with AVX. | Rasmus Munk Larsen | 2021-04-22 |
| | |||
* | Replace `-2147483648` by `-0.0f` or `-0.0` constants (this should fix #2189). | Christoph Hertzberg | 2021-04-07 |
| | | | | Also, remove unnecessary `pgather` operations. | ||
* | EOF newline added to InverseSize4. | Antonio Sanchez | 2020-11-18 |
| | | | | | Causing build breakages due to `-Wnewline-eof -Werror` that seems to be common across Google. | ||
* | Unify Inverse_SSE.h and Inverse_NEON.h into a single generic implementation ↵ | Guoqiang QI | 2020-11-17 |
| | | | | using PacketMath. | ||
* | Use reinterpret_cast instead of C-style cast in Inverse_NEON.h | David Tellenbach | 2020-10-04 |
| | |||
* | Don't cast away const in Inverse_NEON.h. | Rasmus Munk Larsen | 2020-10-02 |
| | |||
* | Add missing newline at the end of Inverse_NEON.h | Rasmus Munk Larsen | 2020-09-29 |
| | |||
* | Disable double version of compute_inverse_size4 on Inverse_NEON.h if ↵ | Rasmus Munk Larsen | 2020-09-17 |
| | | | | Packet2d is not supported. | ||
* | Add Inverse_NEON.h | Stephen Zheng | 2020-09-04 |
| | | | | | | | | | | | Implemented fast size-4 matrix inverse (mimicking Inverse_SSE.h) using NEON intrinsics. ``` Benchmark Time CPU Time Old Time New CPU Old CPU New -------------------------------------------------------------------------------------------------------- BM_float -0.1285 -0.1275 568 495 572 499 BM_double -0.2265 -0.2254 638 494 641 496 ``` | ||
* | bug #1644: fix warning | Gael Guennebaud | 2018-12-11 |
| | |||
* | Fix 4x4 inverse with non-linear destination | Gael Guennebaud | 2016-08-30 |
| | |||
* | bug #1167: simplify installation of header files using cmake's ↵ | Gael Guennebaud | 2016-08-29 |
| | | | | install(DIRECTORY ...) command. | ||
* | First part of a big refactoring of alignment control to enable the handling ↵ | Gael Guennebaud | 2015-08-06 |
| | | | | | | | | | of arbitrarily aligned buffers. It includes: - AlignedBit flag is deprecated. Alignment is now specified by the evaluator through the 'Alignment' enum, e.g., evaluator<Xpr>::Alignment. Its value is in Bytes. - Add several enums to specify alignment: Aligned8, Aligned16, Aligned32, Aligned64, Aligned128. AlignedMax corresponds to EIGEN_MAX_ALIGN_BYTES. Such enums are used to define the above Alignment value, and as the 'Options' template parameter of Map<> and Ref<>. - The Aligned enum is now deprecated. It is now an alias for Aligned16. - Currently, traits<Matrix<>>, traits<Array<>>, traits<Ref<>>, traits<Map<>>, and traits<Block<>> also expose the Alignment enum. | ||
* | Fix 4x4 inverse via SSE for submatrices | Gael Guennebaud | 2014-07-31 |
| | |||
* | Automatic relicensing to MPL2 using Keirs script. Manual fixup follows. | Benoit Jacob | 2012-07-13 |
| | |||
* | Get rid of include directives inside namespace blocks (bug #339). | Jitse Niesen | 2012-04-15 |
| | |||
* | Fix out-of-range int constant in 4x4 inverse. | Keir Mierle | 2012-01-05 |
| | | | | | (transplanted from 45bcad41b444a44d28558472ff27d5cd1207d41c ) | ||
* | remove the use of non standard long long | Gael Guennebaud | 2011-06-14 |
| | |||
* | bug #86 : use internal:: namespace instead of ei_ prefix | Benoit Jacob | 2010-10-25 |
| | |||
* | fix 4x4 SSE inversion when storage orders don't match | Gael Guennebaud | 2010-08-24 |
| | |||
* | fix inversion of 4x4 unaligned matrices | Gael Guennebaud | 2010-08-24 |
| | |||
* | email change | Gael Guennebaud | 2010-06-24 |
| | |||
* | fix compilation on 32bits systems | Gael Guennebaud | 2010-02-01 |
| | |||
* | add SSE code (from Intel) for the fast inversion of 4x4 matrices of double | Gael Guennebaud | 2010-01-19 |
| | |||
* | update the fast 4x4 SSE inversion code from more recent Intel's code | Gael Guennebaud | 2010-01-19 |
| | |||
* | suppress unused variable warnings | Gael Guennebaud | 2009-12-15 |
| | |||
* | no, this wasn't equivalent to ei_pload at all, after all! | Benoit Jacob | 2009-12-15 |
| | |||
* | Gael, who is a man of few words^Winstructions, is right, as usual. | Benoit Jacob | 2009-12-15 |
| | |||
* | *use scalar instructions, packet not needed here | Benoit Jacob | 2009-12-14 |
| | | | | *remove unused var warning | ||
* | add SSE path for Matrix4f inverse, taken from Intel except that we do a kosher | Benoit Jacob | 2009-12-14 |
division instead of RCPPS-followed-by-Newton-Raphson. The rationale for that is that elsewhere in Eigen we dont allow ourselves this approximation (which throws 2 bits of mantissa), so there's no reason we should allow it here. |