Commit Graph

  • baf601a0e3 Fix for issue with static global variables in TensorDeviceGpu.h Turing Eret 2021-04-23 07:43:35 -06:00
  • d216764f46 WIP Everton Constantino 2021-04-23 17:28:17 +00:00
  • 646d92c7f1 WIP Everton Constantino 2021-04-23 15:39:04 +00:00
  • 3804ca0d90 Fix for issue with static global variables in TensorDeviceGpu.h Turing Eret 2021-04-23 07:43:35 -06:00
  • 587a691516 Check existence of BSD random before use. Antonio Sanchez 2021-04-21 11:03:23 -07:00
  • 8830d66c02 DenseStorage safely copy/swap. Antonio Sanchez 2021-04-21 15:45:31 -07:00
  • c62ed9b214 WIP Everton Constantino 2021-04-22 20:42:44 +00:00
  • 045c0609b5 Check existence of BSD random before use. Antonio Sanchez 2021-04-21 11:03:23 -07:00
  • d213a0bcea DenseStorage safely copy/swap. Antonio Sanchez 2021-04-21 15:45:31 -07:00
  • 82a7715b01 WIP Everton Constantino 2021-04-22 18:11:53 +00:00
  • 43ce8e9d2d WIP Everton Constantino 2021-04-22 17:43:22 +00:00
  • 54425a39b2 Make vectorized compute_inverse_size4 compile with AVX. Rasmus Munk Larsen 2021-04-22 15:21:01 +00:00
  • 85a76a16ea Make vectorized compute_inverse_size4 compile with AVX. Rasmus Munk Larsen 2021-04-22 15:21:01 +00:00
  • ca0d3f92d7 WIP Everton Constantino 2021-04-22 14:48:44 +00:00
  • 5bffe09624 WIP Everton Constantino 2021-04-22 13:14:00 +00:00
  • 421891e1db WIP Everton Constantino 2021-04-21 17:58:55 +00:00
  • 34d0be9ec1 Compilation of basicbenchmark fixed Jakub Lichman 2021-04-19 13:46:38 +00:00
  • 42a8bdd4d7 HasExp added for AVX512 Packet8d Jakub Lichman 2021-04-20 11:14:56 +00:00
  • d72c794ccd Compilation of basicbenchmark fixed Jakub Lichman 2021-04-19 13:46:38 +00:00
  • 28564957ac Fix taking address of rvalue compiler issue with TensorFlow (plus other warnings). Chip-Kerchner 2021-04-20 12:01:45 -05:00
  • 06c2760bd1 Fix taking address of rvalue compiler issue with TensorFlow (plus other warnings). Chip-Kerchner 2021-04-20 12:01:45 -05:00
  • ab7fe215f9 Fix ldexp for AVX512 (#2215) Antonio Sanchez 2021-04-14 13:54:11 -07:00
  • f826663a3a WIP Everton Constantino 2021-04-20 20:10:21 +00:00
  • 2b1dfd1ba0 HasExp added for AVX512 Packet8d Jakub Lichman 2021-04-20 11:14:56 +00:00
  • 1d79c68ba0 Fix ldexp for AVX512 (#2215) Antonio Sanchez 2021-04-14 13:54:11 -07:00
  • 1f4c0311cd Bump to 3.3.91 (3.4-rc1) 3.4-rc1 David Tellenbach 2021-04-18 23:43:12 +02:00
  • 3e819d83bf Before 3.4 branch before-3.4 David Tellenbach 2021-04-18 23:36:14 +02:00
  • 69adf26aa3 Modify googlehash use to account for namespace issues. Antonio Sanchez 2021-04-09 19:50:19 -07:00
  • 9357feedc7 Avoid using uninitialized inputs and if available, use slightly more efficient movsd instruction for pset1<Packet2cf>. Christoph Hertzberg 2021-04-13 00:52:30 +02:00
  • a2c0542010 Fix typo in TensorDimensions.h Rasmus Munk Larsen 2021-04-12 18:59:56 +00:00
  • dfd6720d82 Fix for float16 GPU unit test. Rohit Santhanam 2021-04-11 14:54:18 +00:00
  • 1e1c8a735c Use EIGEN_HAS_CXX11 and EIGEN_COMP_CXXVER macros to detect C++ version for std::result_of and std::invoke_result. Fixes #2209 Christoph Hertzberg 2021-04-11 16:51:48 +02:00
  • f6fc66aa75 fixed doxygen for unsupported iterative solver module Jens Wehner 2021-04-11 16:26:14 +00:00
  • d58678069c Make iterators default constructible and assignable, by making... Christoph Hertzberg 2021-04-09 17:03:28 +00:00
  • 2859db0220 This fixes an issue where the compiler was not choosing the GPU specific specialization of ScanLauncher. Rohit Santhanam 2021-04-08 15:14:48 +00:00
  • fcb5106c6e Scaled epsilon the wrong way. Antonio Sanchez 2021-04-07 15:08:39 -07:00
  • 6197ce1a35 Replace -2147483648 by -0.0f or -0.0 constants (this should fix #2189). Also, remove unnecessary pgather operations. Christoph Hertzberg 2021-04-06 16:49:57 +02:00
  • 22edb46823 Align local arrays to Packet boundary. Rasmus Munk Larsen 2021-04-06 16:22:36 +00:00
  • ace7f132ed Fix clang tidy warnings in AnnoyingScalar. Antonio Sanchez 2021-04-05 12:19:27 -07:00
  • 90187a33e1 Fix SelfAdjoingEigenSolver (#2191) Antonio Sanchez 2021-03-25 11:08:19 +00:00
  • 3ddc0974ce Fix two bugs in commit Rasmus Munk Larsen 2021-04-02 22:06:27 +00:00
  • c24bee6120 Fix address of temporary object errors in clang11. Chip Kerchner 2021-04-02 16:27:08 +00:00
  • e4233b6e3d Add CI infrastructure for pre-merge smoke tests. David Tellenbach 2021-04-01 00:08:37 +00:00
  • ae95b74af9 Add CMake infrastructure for smoke testing David Tellenbach 2021-03-31 22:09:00 +00:00
  • 5bbc9cea93 Add an info() method to the SVDBase class to make it possible to tell the user that the computation failed, possibly due to invalid input. Make Jacobi and divide-and-conquer fail fast and return info() == InvalidInput if the matrix contains NaN or +/-Inf. Rasmus Munk Larsen 2021-03-31 21:09:19 +00:00
  • b5a926a0f6 Add GitLab templates for issues and merge requests Guoqiang QI 2021-03-31 16:01:12 +00:00
  • 78ee3d6261 Fix CUDA constexpr issues for numeric_limits. Antonio Sanchez 2021-03-25 12:37:25 -07:00
  • af1247fbc1 Use Index type in loop over coefficients. Antonio Sanchez 2021-03-29 09:25:39 -07:00
  • 87729ea39f Eliminate round_impl double-promotion warnings for c++03. Antonio Sanchez 2021-03-24 10:12:02 -07:00
  • 748489ef9c Un-defining EIGEN_HAS_CONSTEXPR on the HIP platform Deven Desai 2021-03-25 13:45:52 +00:00
  • d59ef212e1 Fixed performance issues for complex VSX and P10 MMA in gebp_kernel (level 3). Chip Kerchner 2021-03-25 11:08:19 +00:00
  • e7b8643d70 Revert "Revert "Adds EIGEN_CONSTEXPR and EIGEN_NOEXCEPT to rows(), cols(), innerStride(), outerStride(), and size()"" Steve Bronder 2021-03-08 12:39:11 -05:00
  • 5521c65afb Eliminate mixingtypes_7 warning. Antonio Sanchez 2021-03-24 11:05:41 -07:00
  • 69a4f70956 Revert "Uses _mm512_abs_pd for Packet8d pabs" Christoph Hertzberg 2021-03-23 18:52:19 +00:00
  • 824272cde8 Re-enable CI for Power David Tellenbach 2021-03-22 19:28:25 +01:00
  • 4811e81966 Remove yet another comma at end of enum David Tellenbach 2021-03-18 23:30:00 +01:00
  • f019b97aca Uses _mm512_abs_pd for Packet8d pabs Steve Bronder 2021-03-18 15:47:52 +00:00
  • 0cc9b5eb40 Split test commainitializer into two substests David Tellenbach 2021-03-18 13:28:51 +01:00
  • c3fbc6cec7 Use singleton pattern for static registered tests. Antonio Sanchez 2021-03-17 14:18:50 -07:00
  • ed964ba3f1 Proposed fix for issue #2187 Niek Bouman 2021-03-18 00:55:36 +00:00
  • 8dfe1029a5 Augment NumTraits with min/max_exponent() again. Antonio Sanchez 2021-03-16 20:12:46 -07:00
  • eb71e5db98 Fix another warning on missing commas David Tellenbach 2021-03-17 03:07:04 +01:00
  • df4bc2731c Revert "Augment NumTraits with min/max_exponent()." David Tellenbach 2021-03-17 03:06:08 +01:00
  • 75ce9cd2a7 Augment NumTraits with min/max_exponent(). Antonio Sanchez 2021-02-12 13:14:05 -08:00
  • 9fb7062440 Silence warning on comma at end of enumerator list David Tellenbach 2021-03-17 01:46:52 +01:00
  • b8502a9dd6 Updated SelfAdjointEigenSolver documentation to include that the eigenvectors matrix is unitary. Theo Fletcher 2021-03-16 03:12:42 +00:00
  • 2e83cbbba9 Add NaN propagation options to minCoeff/maxCoeff visitors. Rasmus Munk Larsen 2021-03-16 17:02:50 +00:00
  • c0a889890f Fixed output of complex matrices Jens Wehner 2021-03-15 21:51:55 +00:00
  • f612df2736 Add fmod(half, half). Antonio Sanchez 2021-03-15 13:00:32 -07:00
  • 14b7ebea11 Fix numext::round pre c++11 for large inputs. Antonio Sanchez 2021-03-10 21:27:35 -08:00
  • c9d4367fa4 Fix pround and add print Chip Kerchner 2021-03-15 19:07:43 +00:00
  • d24f9f9b55 Fix NVCC+ICC issues. Antonio Sanchez 2021-03-11 11:23:00 -08:00
  • 14487ed14e Add increment/decrement operators to Eigen::half. Antonio Sanchez 2021-03-15 10:50:37 -07:00
  • b271110788 Bump up rand histogram threshold. Antonio Sanchez 2021-03-10 22:16:14 -08:00
  • d098c4d64c Disable EIGEN_OPTIMIZATION_BARRIER for PPC clang. Antonio Sanchez 2021-03-10 16:04:02 -08:00
  • 543e34ab9d Re-implement move assignments. Antonio Sanchez 2021-03-05 12:54:26 -08:00
  • b8d1857f0d [MSVC-specific] Define EIGEN_ARCH_x86_64 for native x64 (_M_X64 is defined and _M_ARM64EC is not), and define EIGEN_ARCH_ARM64 for both the native ARM64 (_M_ARM64 is defined) or ARM64EC (_M_ARM64EC is defined). _M_ARM64EC is defined when the code is compiled by MSVC for ARM64EC, a new ARM64 ABI designed to be compatible with x64 application emulation on ARM64. If _M_ARM64EC is defined, _M_X64 and _M_AMD64 are also defined, so x64-specific code (especially intrinsics) is also compiled to ARM64 instructions (compliant with the ARM64EC ABI) for maximum x64 compatibility. Although a majority of x64-specific intrinsics can emulated by ARM64 instructions, it is still a good to simply recompile the native ARM64 code paths to ARM64EC for pure computation tasks, for performance reasons. Ben Niu 2021-03-08 19:43:38 +00:00
  • 853a5c4b84 Fix ambiguous call to CUDA __half constructor. Antonio Sanchez 2021-03-08 21:06:28 -08:00
  • 94327dbfba Fix typo: DEVICE -> GPU Antonio Sanchez 2021-03-08 11:21:00 -08:00
  • 1296abdf82 Fix non-trivial Half constructor for CUDA. Antonio Sanchez 2021-03-08 07:32:54 -08:00
  • 6045243141 Revert stack allocation limit change that crept in. Antonio Sanchez 2021-03-05 14:29:37 -08:00
  • 1a96d49afe Changing the Eigen::half implementation for HIP Deven Desai 2021-03-05 19:27:13 +00:00
  • 2468253c9a Define EIGEN_CPLUSPLUS and replace most __cplusplus checks. Antonio Sanchez 2021-03-01 15:57:22 -08:00
  • 82d61af3a4 Fix rint SSE/NEON again, using optimization barrier. Antonio Sanchez 2021-03-03 19:22:15 -08:00
  • 5f0b4a4010 Revert "Adds EIGEN_CONSTEXPR and EIGEN_NOEXCEPT to rows(), cols(), innerStride(), outerStride(), and size()" David Tellenbach 2021-03-05 13:16:43 +01:00
  • 6cbb3038ac Adds EIGEN_CONSTEXPR and EIGEN_NOEXCEPT to rows(), cols(), innerStride(), outerStride(), and size() Steve Bronder 2021-03-04 18:58:08 +00:00
  • 5bfc67f9e7 Deactive CI for Power due to problems with GitLab runner David Tellenbach 2021-03-04 17:33:40 +01:00
  • a6601070f2 Add log2 operation to TensorBase Eugene Zhulenev 2021-03-04 00:13:36 +00:00
  • 9a663973b4 Revert "Fix rint for SSE/NEON." Antonio Sánchez 2021-03-03 18:51:51 +00:00
  • e72dfeb8b9 Fix rint for SSE/NEON. Antonio Sanchez 2021-03-03 09:41:46 -08:00
  • 199c5f2b47 geo_alignedbox_5 was failing with AVX enabled, due to storing Vector4d in a std::vector without using an aligned allocator. Got rid of using std::vector and simplified the code. Avoid leading _ Christoph Hertzberg 2021-03-01 03:59:21 +01:00
  • 1e0c7d4f49 Add print for SSE/NEON, use NEON rounding intrinsics if available. Antonio Sanchez 2021-02-25 14:29:49 -08:00
  • 976ae0ca6f Document that using raw function pointers doesn't work with unaryExpr. David Tellenbach 2021-02-27 22:58:42 +01:00
  • c65c2b31d4 Make half/bfloat16 constructor take inputs by value, fix powerpc test. Antonio Sanchez 2021-02-26 15:25:00 -08:00
  • 39a590dfb6 Remove unused include Christoph Hertzberg 2021-02-27 19:02:33 +01:00
  • 8f686ac4ec clang 10 aggressively warns about precision loss when converting int to float (or long to double) Christoph Hertzberg 2021-02-27 18:30:47 +01:00
  • 2660d01fa7 Inherit from no_assignment_operator to avoid implicit copy constructor warnings Christoph Hertzberg 2021-02-27 18:25:48 +01:00
  • a3521d743c Fix some enum-enum conversion warnings Christoph Hertzberg 2021-02-27 17:51:16 +01:00
  • ca528593f4 Fixed/masked more implicit copy constructor warnings Christoph Hertzberg 2021-02-27 16:34:42 +01:00
  • 81b5fe2f0a ReturnByValue is already non-copyable Christoph Hertzberg 2021-02-27 16:16:35 +01:00