Update Eigen to commit:008ff3483a8c5604639e1c4d204eae30ad737af6
CHANGELOG
=========
008ff3483 - Fix broken tensor executor test, allow tensor packets of size 1.
28c7c1a62 - Log position of first difference for easier debugging.
cf8218641 - Adds new CMake Options for controlling build components.
b2ee235a4 - Split and reduce SVD test sizes.
0ae94456a - Remove duplicate IsRowMajor declaration.
0e6f4e43f - Fix a few confusing comments in psincos_float.
f1b9692d6 - Removed EIGEN_UNUSED decorations from many functions that are in fact used
27d8f29be - Update vectorization_logic tests for all platforms.
c9ff739af - Fix JacobiSVD_LAPACKE bindings
d0b1aef6f - Speed lscg by using .noalias
55c7400db - Fix enum conversion warnings in BooleanRedux.
711803c42 - Skip denormal test if `Cond` is false.
d819a33bf - Remove poor non-convergence checks in NonLinearOptimization.
9c07e201f - Modified sqrt/rsqrt for denormal handling.
1c2690ed2 - Adjust tolerance of matrix_power test for MSVC.
b48922cb5 - Fix SVD for MSVC+CUDA.
bf6726a0c - Fix any/all reduction in the case of row-major layout
f03df0df5 - Fix SVD for MSVC.
19c39bea2 - Fix mixingtypes for g++-11.
2ed4bee78 - Fix frexp packetmath tests for MSVC.
d58e62913 - Disable deprecated warnings for SVD tests on MSVC.
3d7e2d0e3 - Fix packetmath compilation error.
897071977 - Fix gcc-5 packetmath_12 bug.
f0b81fefb - Disable deprecated warnings in SVD tests.
8b875dbef - Changes to fast SQRT/RSQRT
f9b7564fa - E2K: initial support of LCC MCST compiler for the Elbrus 2000 CPU architecture
ae86a146b - Modify test expression to avoid numerical differences (#2402).
cd80e04ab - Add assert for edge case if Thin U Requested at runtime
35727928a - Fix test macro conflicts with STL headers in C++20
2dd879d4b - [SYCL] Fix CMake for SYCL support
550af3938 - Fix for crash bug in SPQRSupport: Initialize pointers to nullptr to avoid free() calls of invalid pointers.
58a90c746 - Use fixed-sized U/V for fixed-sized inputs.
c367ed26a - Make FixedInt constexpr, fix ODR of fix<N>
766087329 - Re-add `svd::compute(Matrix, options)` method to avoid breaking external projects.
a58af20d6 - Add descriptions to Matrix typedefs.
28e008b99 - Fix sqrt/rsqrt for NEON.
23755030c - Fix MSVC+NVCC 9.2 pragma error.
7197b577f - Remove unused macros in AVX packetmath.
bded5028a - Fix ODR failures in TensorRandom.
18eab8f99 - Add convenience method `constexpr std::size_t size() const` to `Eigen::IndexList`
fbc62f7df - Complete doc with MatrixXNt and MatrixNXt
cb5ca1c90 - Cleanup compiler warnings, etc from recent changes in GEMM & GEMV for PowerPC
cec0005c7 - Return alphas() and betas() by const reference
92d0026b7 - Provide a definition for numeric_limits static data members
b94bddcde - Typo in COD's doc: matrixR() -> matrixT()
94bed2b80 - Fix collision with resolve.h.
b88de3f24 - Update MPL2 with https.
9441d94dc - Revert "Make fixed-size Matrix and Array trivially copyable after C++20"
979fdd58a - Add generic fast psqrt and prsqrt impls and make them correct for 0, +Inf, NaN, and negative arguments.
4bffbe84f - Restrict GCC<6.3 maxpd workaround to only gcc.
e7f4a901e - Define EIGEN_HAS_AVX512_MATH in PacketMath.
6b60bd675 - Fix 32-bit arm int issue.
96da541cb - Fix AVX512 math function consistency, enable for ICC.
cafeadffe - Fix ODR violations.
18b50458b - Update SVD Module with Options template parameter
89c6ab238 - removed some documentation referencing c++98 behaviour
66464bd2a - Fix number of block columns to NOT overflow the cache (PowerPC) abnormally in GEMV
PiperOrigin-RevId: 433536960
Change-Id: I6f24e8a4fcc412b68bee72d438aa23afdb1037dc
108 files changed
