mirror/eigen
2
0
Fork 0
mirror of https://gitlab.com/libeigen/eigen.git synced 2025-04-18 19:30:38 +08:00
Code Issues Packages Projects Releases Wiki Activity
eigen/Eigen/Core

Ignoring revisions in .git-blame-ignore-revs. Click here to bypass and see the normal blame view.

415 lines
13 KiB
Plaintext
Raw Permalink Normal View History

add copyright on two public headers that are not so trivial...
2009-05-06 15:48:28 +00:00
// This file is part of Eigen, a lightweight C++ template library
remove sentence "Eigen itself is part of the KDE project." it never made very precise sense. but now does it still make any?
2009-05-22 20:25:33 +02:00
// for linear algebra.
add copyright on two public headers that are not so trivial...
2009-05-06 15:48:28 +00:00
//
email change
2010-06-24 23:21:58 +02:00
// Copyright (C) 2008 Gael Guennebaud <gael.guennebaud@inria.fr>
reverse order of testing for eigen2 support stages. Higher stages now have priority. So if your whole project builds with say stage 10, you can manually enable stage 20 for selected files.
2011-01-27 10:34:44 -05:00
// Copyright (C) 2007-2011 Benoit Jacob <jacob.benoit.1@gmail.com>
add copyright on two public headers that are not so trivial...
2009-05-06 15:48:28 +00:00
//
Automatic relicensing to MPL2 using Keirs script. Manual fixup follows.
2012-07-13 14:42:47 -04:00
// This Source Code Form is subject to the terms of the Mozilla
// Public License v. 2.0. If a copy of the MPL was not distributed
// with this file, You can obtain one at http://mozilla.org/MPL/2.0/.
add copyright on two public headers that are not so trivial...
2009-05-06 15:48:28 +00:00
Issue an error in case of direct inclusion of internal headers.
2021-09-10 19:12:26 +00:00
#ifndef EIGEN_CORE_MODULE_H
#define EIGEN_CORE_MODULE_H
* add Gael copyright lines on 2 more files * macro renaming: EIGEN_NDEBUG becomes EIGEN_NO_DEBUG as this is much better (and similar to Qt) and EIGEN_CUSTOM_ASSERT becomes EIGEN_USE_CUSTOM_ASSERT * protect Core header by a EIGEN_CORE_H
2008-03-26 09:13:11 +00:00
Fix #1757: remove the word 'suicide'
2020-06-09 18:15:52 -04:00
// first thing Eigen does: stop the compiler from reporting useless warnings.
Applyiing Benoit's comment to return the missing line back in Eigen/Core
2016-10-14 16:39:41 +01:00
#include "src/Core/util/DisableStupidWarnings.h"
NVCC: EIGEN_NO_DEBUG must be defined before including Macro.h
2013-02-21 19:05:23 +01:00
Cleanup the mess in Eigen/Core by moving CUDA/HIP stuff at more appropriate places (Macros.h), and alignment/vectorization logic is now in util/ConfigureVectorization.h
2018-07-12 16:57:41 +02:00
// then include this file where all our macros are defined. It's really important to do it first because
// it's where we do all the compiler/OS/arch detections and define most defaults.
#include "src/Core/util/Macros.h"
Partial OpenCL support via SYCL compatible with ComputeCpp CE.
2016-09-19 12:44:13 +01:00
Cleanup the mess in Eigen/Core by moving CUDA/HIP stuff at more appropriate places (Macros.h), and alignment/vectorization logic is now in util/ConfigureVectorization.h
2018-07-12 16:57:41 +02:00
// This detects SSE/AVX/NEON/etc. and configure alignment settings
#include "src/Core/util/ConfigureVectorization.h"
Define EIGEN_TRY, EIGEN_CATCH, EIGEN_THROW as suggested by Moritz Klammer. Make it possible to run unit-tests with exceptions disabled via EIGEN_TEST_NO_EXCEPTIONS flag. Enhanced ctorleak unit-test
2014-07-22 13:16:44 +02:00
Cleanup the mess in Eigen/Core by moving CUDA/HIP stuff at more appropriate places (Macros.h), and alignment/vectorization logic is now in util/ConfigureVectorization.h
2018-07-12 16:57:41 +02:00
// We need cuda_runtime.h/hip_runtime.h to ensure that
Drop EIGEN_USING_STD_MATH in favour of EIGEN_USING_STD
2020-10-09 02:05:05 +02:00
// the EIGEN_USING_STD macro works properly on the device side
Cleanup the mess in Eigen/Core by moving CUDA/HIP stuff at more appropriate places (Macros.h), and alignment/vectorization logic is now in util/ConfigureVectorization.h
2018-07-12 16:57:41 +02:00
#if defined(EIGEN_CUDACC)
Extend CUDA support to matrix inversion and selfadjointeigensolver
2018-06-11 18:33:24 +02:00
#include <cuda_runtime.h>
merging updates from upstream
2018-07-11 09:17:33 -04:00
#elif defined(EIGEN_HIPCC)
#include <hip/hip_runtime.h>
Fixing the typo regarding missing #if needed for proper handling of exceptions in Eigen/Core.
2016-10-16 12:52:34 +01:00
#endif
Define EIGEN_TRY, EIGEN_CATCH, EIGEN_THROW as suggested by Moritz Klammer. Make it possible to run unit-tests with exceptions disabled via EIGEN_TEST_NO_EXCEPTIONS flag. Enhanced ctorleak unit-test
2014-07-22 13:16:44 +02:00
#ifdef EIGEN_EXCEPTIONS
#include <new>
#endif
issue #2581: review and cleanup of compiler version checks
2023-01-17 18:58:34 +00:00
// Disable the ipa-cp-clone optimization flag with MinGW 6.x or older (enabled by default with -O3)
bug #556: workaround mingw bug with -O3 or -fipa-cp-clone
2013-07-05 23:47:40 +02:00
// See http://eigen.tuxfamily.org/bz/show_bug.cgi?id=556 for details.
issue #2581: review and cleanup of compiler version checks
2023-01-17 18:58:34 +00:00
#if EIGEN_COMP_MINGW && EIGEN_GNUC_STRICT_LESS_THAN(6, 0, 0)
bug #556: workaround mingw bug with -O3 or -fipa-cp-clone
2013-07-05 23:47:40 +02:00
#pragma GCC optimize("-fno-ipa-cp-clone")
#endif
Fix NVCC+ICC issues. NVCC does not understand `__forceinline`, so we need to use `inline` when compiling for GPU. ICC specializes `std::complex` operators for `float` and `double` by default, which cannot be used on device and conflict with Eigen's workaround in CUDA/Complex.h. This can be prevented by defining `_OVERRIDE_COMPLEX_SPECIALIZATION_` before including `<complex>`. Added this define to the tests and to `Eigen/Core`, but this will not work if the user includes `<complex>` before `<Eigen/Core>`. ICC also seems to generate a duplicate `Map` symbol in `PlainObjectBase`: ``` error: "Map" has already been declared in the current scope static ConstMapType Map(const Scalar *data) ``` I tracked this down to `friend class Eigen::Map`. Putting the `friend` statements at the bottom of the class seems to resolve this issue. Fixes #2180
2021-03-11 11:23:00 -08:00
// Prevent ICC from specializing std::complex operators that silently fail
// on device. This allows us to use our own device-compatible specializations
// instead.
issue #2581: review and cleanup of compiler version checks
2023-01-17 18:58:34 +00:00
#if EIGEN_COMP_ICC && defined(EIGEN_GPU_COMPILE_PHASE) && !defined(_OVERRIDE_COMPLEX_SPECIALIZATION_)
Fix NVCC+ICC issues. NVCC does not understand `__forceinline`, so we need to use `inline` when compiling for GPU. ICC specializes `std::complex` operators for `float` and `double` by default, which cannot be used on device and conflict with Eigen's workaround in CUDA/Complex.h. This can be prevented by defining `_OVERRIDE_COMPLEX_SPECIALIZATION_` before including `<complex>`. Added this define to the tests and to `Eigen/Core`, but this will not work if the user includes `<complex>` before `<Eigen/Core>`. ICC also seems to generate a duplicate `Map` symbol in `PlainObjectBase`: ``` error: "Map" has already been declared in the current scope static ConstMapType Map(const Scalar *data) ``` I tracked this down to `friend class Eigen::Map`. Putting the `friend` statements at the bottom of the class seems to resolve this issue. Fixes #2180
2021-03-11 11:23:00 -08:00
#define _OVERRIDE_COMPLEX_SPECIALIZATION_ 1
#endif
<complex> must be included first
2012-02-10 22:49:09 +01:00
#include <complex>
- split and rename defined tokens to enable the use of BLAS/Lapack/VML/etc - include MKL headers outside the Eigen namespace.
2011-12-09 10:06:49 +01:00
// this include file manages BLAS and MKL related macros
// and inclusion of their respective header files
#include "src/Core/util/MKL_support.h"
updates based on PR feedback There are two major changes (and a few minor ones which are not listed here...see PR discussion for details) 1. Eigen::half implementations for HIP and CUDA have been merged. This means that - `CUDA/Half.h` and `HIP/hcc/Half.h` got merged to a new file `GPU/Half.h` - `CUDA/PacketMathHalf.h` and `HIP/hcc/PacketMathHalf.h` got merged to a new file `GPU/PacketMathHalf.h` - `CUDA/TypeCasting.h` and `HIP/hcc/TypeCasting.h` got merged to a new file `GPU/TypeCasting.h` After this change the `HIP/hcc` directory only contains one file `math_constants.h`. That will go away too once that file becomes a part of the HIP install. 2. new macros EIGEN_GPUCC, EIGEN_GPU_COMPILE_PHASE and EIGEN_HAS_GPU_FP16 have been added and the code has been updated to use them where appropriate. - `EIGEN_GPUCC` is the same as `(EIGEN_CUDACC || EIGEN_HIPCC)` - `EIGEN_GPU_DEVICE_COMPILE` is the same as `(EIGEN_CUDA_ARCH || EIGEN_HIP_DEVICE_COMPILE)` - `EIGEN_HAS_GPU_FP16` is the same as `(EIGEN_HAS_CUDA_FP16 or EIGEN_HAS_HIP_FP16)`
2018-06-14 10:21:54 -04:00
#if defined(EIGEN_HAS_CUDA_FP16) || defined(EIGEN_HAS_HIP_FP16)
#define EIGEN_HAS_GPU_FP16
Made sure all the required header files are included when trying to use fp16
2016-04-19 17:44:12 -07:00
#endif
Support BFloat16 in Eigen
2020-06-20 19:16:24 +00:00
#if defined(EIGEN_HAS_CUDA_BF16) || defined(EIGEN_HAS_HIP_BF16)
#define EIGEN_HAS_GPU_BF16
#endif
Added EIGEN_DONT_PARALLELIZE preprocessor directive
2010-04-26 17:08:54 +02:00
#if (defined _OPENMP) && (!defined EIGEN_DONT_PARALLELIZE)
clean a bit the parallelizer
2010-02-22 11:08:37 +01:00
#define EIGEN_HAS_OPENMP
#endif
#ifdef EIGEN_HAS_OPENMP
Only include <atomic> if needed.
2021-12-02 23:55:25 +00:00
#include <atomic>
clean a bit the parallelizer
2010-02-22 11:08:37 +01:00
#include <omp.h>
#endif
MSVC for windows mobile does not have the errno.h file
2010-10-07 18:09:15 +02:00
// MSVC for windows mobile does not have the errno.h file
Introduce unified macros to identify compiler, OS, and architecture. They are all defined in util/Macros.h and prefixed with EIGEN_COMP_, EIGEN_OS_, and EIGEN_ARCH_ respectively.
2014-11-04 21:58:52 +01:00
#if !(EIGEN_COMP_MSVC && EIGEN_OS_WINCE) && !EIGEN_COMP_ARM
MSVC for windows mobile does not have the errno.h file
2010-10-07 18:09:15 +02:00
#define EIGEN_HAS_ERRNO
#endif
#ifdef EIGEN_HAS_ERRNO
Added a generic reallocation implementation based on ei_aligned_malloc/_free. Rewrote ei_handmade_aligned_realloc such that it is now using std::realloc. Reorganized functions in Memory.h for better readability. Add missing <cerrno> include to Core (it's now required in Memory.h).
2010-02-28 14:32:57 +01:00
#include <cerrno>
MSVC for windows mobile does not have the errno.h file
2010-10-07 18:09:15 +02:00
#endif
fix bug #250: compilation error with gcc 4.6 (STL header files no longer include cstddef)
2011-04-19 16:34:25 +02:00
#include <cstddef>
* temporarily remove doxygen customization, we'll see if that fix api.kde.org but I no hope, that would be too simple ! * added Rotation2D typedefs * remove CoreDeclarations header file
2008-08-28 15:28:23 +00:00
#include <cstdlib>
#include <cmath>
#include <functional>
Collapsed revision (based on pull request PR-325) * Support compiling without IO streams Add the preprocessor definition EIGEN_NO_IO which, if defined, disables all use of the IO streams part of the standard library.
2018-10-22 21:14:40 +02:00
#ifndef EIGEN_NO_IO
Avoid including <sstream> with EIGEN_NO_IO
2022-07-29 18:02:51 +00:00
#include <sstream>
Collapsed revision (based on pull request PR-325) * Support compiling without IO streams Add the preprocessor definition EIGEN_NO_IO which, if defined, disables all use of the IO streams part of the standard library.
2018-10-22 21:14:40 +02:00
#include <iosfwd>
#endif
fix compile errors with gcc 4.3: unresolved func call to ei_cache_friendly_product, and undeclared memcpy
2008-08-03 15:44:06 +00:00
#include <cstring>
* Add the possibility to customize the output of matrices, e.g.: IoFormat OctaveFmt(4, AlignCols, ", ", ";\n", "", "", "[", "]"); cout << mat.format(OctaveFmt); The first "4" is the precision. Documentation missing. * Some compilation fixes
2008-08-21 13:17:21 +00:00
#include <string>
* Added aligned_allocator for using 16-byte aligned types with STL containers. There is still a compile-time problem with STL containers that have a standard-conformant resize() method, but this should resolve the original user issue which was storing aligned objects in a std::map.
2009-01-09 00:55:53 +00:00
#include <limits>
now random<integer types> spans over 0..RAND_MAX, or -RAND_MAX/2..RAND_MAX/2 for signed types, or the most significant bits for smaller integer types.
2011-02-07 10:55:41 -05:00
#include <climits> // for CHAR_BIT
Set of fixes and workaround to make sun studio more happy. Still remains the problem of alignment and vectorization.
2009-07-10 16:10:03 +02:00
// for min/max:
#include <algorithm>
* Patch by Konstantinos Margaritis: AltiVec vectorization. * Fix several warnings, temporarily disable determinant test.
2008-05-03 12:21:23 +00:00
Forward declaring std::array does not work with all std libs, so let's just include <array>
2018-07-13 13:06:44 +02:00
#include <array>
Fix cuda compilation
2023-05-08 16:15:47 +00:00
#include <vector>
Forward declaring std::array does not work with all std libs, so let's just include <array>
2018-07-13 13:06:44 +02:00
bug #725: make move ctor/assignment noexcept.
2016-06-03 14:28:25 +02:00
// for std::is_nothrow_move_assignable
#include <type_traits>
Only include <iosfwd> unless either EIGEN_DEBUG_ASSIGN is defined or we're in eigen2 support mode
2010-02-27 19:04:22 -05:00
// for outputting debug info
#ifdef EIGEN_DEBUG_ASSIGN
Fix bug #89: on GCC <= 4.3, use a custom assert implementation to work around a compiler bug
2011-02-16 08:50:19 -05:00
#include <iostream>
Only include <iosfwd> unless either EIGEN_DEBUG_ASSIGN is defined or we're in eigen2 support mode
2010-02-27 19:04:22 -05:00
#endif
The intrin.h header needs to be included after cmath in order to prevent warnigns. Fixed (hopefully) final Index realted warnings.
2010-06-21 18:39:24 +02:00
// required for __cpuid, needs to be included after cmath
win: include intrin header in Windows on ARM intrin header is needed for _BitScanReverse and _BitScanReverse64
2021-08-27 16:39:24 +02:00
// also required for _BitScanReverse on Windows on ARM
#if EIGEN_COMP_MSVC && (EIGEN_ARCH_i386_OR_x86_64 || EIGEN_ARCH_ARM64) && !EIGEN_OS_WINCE
The intrin.h header needs to be included after cmath in order to prevent warnigns. Fixed (hopefully) final Index realted warnings.
2010-06-21 18:39:24 +02:00
#include <intrin.h>
#endif
[SYCL] This PR adds the minimum modifications to Eigen core required to run Eigen unsupported modules on devices supporting SYCL. * Adding SYCL memory model * Enabling/Disabling SYCL backend in Core * Supporting Vectorization
2019-06-27 12:25:09 +01:00
#if defined(EIGEN_USE_SYCL)
Specialised basic math functions for SYCL device.
2016-11-17 11:47:13 +00:00
#undef min
#undef max
#undef isnan
#undef isinf
#undef isfinite
Eigen moved the `scanLauncehr` function inside the internal namespace. This commit applies the following changes: - Moving the `scamLauncher` specialization inside internal namespace to fix compiler crash on TensorScan for SYCL backend. - Replacing `SYCL/sycl.hpp` to `CL/sycl.hpp` in order to follow SYCL 1.2.1 standard. - minor fixes: commenting out an unused variable to avoid compiler warnings.
2020-05-11 16:10:33 +01:00
#include <CL/sycl.hpp>
[SYCL] This PR adds the minimum modifications to Eigen core required to run Eigen unsupported modules on devices supporting SYCL. * Adding SYCL memory model * Enabling/Disabling SYCL backend in Core * Supporting Vectorization
2019-06-27 12:25:09 +01:00
#include <map>
#include <memory>
#include <utility>
#include <thread>
#ifndef EIGEN_SYCL_LOCAL_THREAD_DIM0
#define EIGEN_SYCL_LOCAL_THREAD_DIM0 16
#endif
#ifndef EIGEN_SYCL_LOCAL_THREAD_DIM1
#define EIGEN_SYCL_LOCAL_THREAD_DIM1 16
Apply clang-format
2023-11-29 11:12:48 +00:00
#endif
[SYCL] This PR adds the minimum modifications to Eigen core required to run Eigen unsupported modules on devices supporting SYCL. * Adding SYCL memory model * Enabling/Disabling SYCL backend in Core * Supporting Vectorization
2019-06-27 12:25:09 +01:00
#endif
Apply clang-format
2023-11-29 11:12:48 +00:00
Removed the deprecated EIGEN2_SUPPORT, as previously announced. A compilation error is raised, if this compile-switch is defined. The documentation references to the corresponding pages from Eigen3.2 now. Also, the Eigen2 testsuite has been removed.
2014-07-01 16:58:11 +02:00
#if defined EIGEN2_SUPPORT_STAGE40_FULL_EIGEN3_STRICTNESS || defined EIGEN2_SUPPORT_STAGE30_FULL_EIGEN3_API || \
defined EIGEN2_SUPPORT_STAGE20_RESOLVE_API_CONFLICTS || defined EIGEN2_SUPPORT_STAGE10_FULL_EIGEN2_API || \
defined EIGEN2_SUPPORT
// This will generate an error message:
#error Eigen2-support is only available up to version 3.2. Please go to "http://eigen.tuxfamily.org/index.php?title=Eigen2" for further information
add minor to Eigen2Support
2010-04-22 20:49:01 -04:00
#endif
bug #1382: move using std::size_t/ptrdiff_t to Eigen's namespace (still better than the global namespace!)
2017-01-23 22:03:57 +01:00
namespace Eigen {
MIsc. source and comment typos Found using `codespell` and `grep` from downstream FreeCAD
2018-03-11 10:01:44 -04:00
// we use size_t frequently and we'll never remember to prepend it with std:: every time just to
this had to be done here, not at the end.
2010-02-12 09:03:16 -05:00
// ensure QNX/QCC support
using std::size_t;
Add matrix condition estimator module that implements the Higham/Hager algorithm from http://www.maths.manchester.ac.uk/~higham/narep/narep135.pdf used in LPACK. Add rcond() methods to FullPivLU and PartialPivLU.
2016-04-01 10:27:59 -07:00
// gcc 4.6.0 wants std:: for ptrdiff_t
better fix for gcc 4.6.0 / ptrdiff_t, as suggested by Benoit
2011-05-05 18:48:18 +02:00
using std::ptrdiff_t;
this had to be done here, not at the end.
2010-02-12 09:03:16 -05:00
bug #1382: move using std::size_t/ptrdiff_t to Eigen's namespace (still better than the global namespace!)
2017-01-23 22:03:57 +01:00
} // namespace Eigen
* move memory related stuff to util/Memory.h * clean ugly doxygen inheritence of expressions * keep improving the documentation... slowly !
2008-08-26 19:12:23 +00:00
/** \defgroup Core_Module Core module
* remove debug code commited by mistake in Assign * keep going on the doc: added a short geometry tutorial
2008-08-26 23:07:33 +00:00
* This is the main module of Eigen providing dense matrix and vector support
* (both fixed and dynamic size) with all the features corresponding to a BLAS library
* and much more...
* move memory related stuff to util/Memory.h * clean ugly doxygen inheritence of expressions * keep improving the documentation... slowly !
2008-08-26 19:12:23 +00:00
*
* \code
* #include <Eigen/Core>
* \endcode
*/
fix lapacke config
2023-02-03 16:40:08 +00:00
#ifdef EIGEN_USE_LAPACKE
#ifdef EIGEN_USE_MKL
#include "mkl_lapacke.h"
#else
#include "src/misc/lapacke.h"
#endif
#endif
Add IWYU export pragmas to top-level headers.
2023-02-08 17:40:31 +00:00
// IWYU pragma: begin_exports
* temporarily remove doxygen customization, we'll see if that fix api.kde.org but I no hope, that would be too simple ! * added Rotation2D typedefs * remove CoreDeclarations header file
2008-08-28 15:28:23 +00:00
#include "src/Core/util/Constants.h"
#include "src/Core/util/Meta.h"
Add custom ODR-safe assert.
2023-01-20 17:38:13 +00:00
#include "src/Core/util/Assert.h"
Move is_diagonal to XprHelper, forward declare Ref
2014-02-18 11:03:59 +01:00
#include "src/Core/util/ForwardDeclarations.h"
* temporarily remove doxygen customization, we'll see if that fix api.kde.org but I no hope, that would be too simple ! * added Rotation2D typedefs * remove CoreDeclarations header file
2008-08-28 15:28:23 +00:00
#include "src/Core/util/StaticAssert.h"
Fix bug #314: move remaining math functions from internal to numext namespace
2013-06-10 23:40:56 +02:00
#include "src/Core/util/XprHelper.h"
* move memory related stuff to util/Memory.h * clean ugly doxygen inheritence of expressions * keep improving the documentation... slowly !
2008-08-26 19:12:23 +00:00
#include "src/Core/util/Memory.h"
Add support for symbolic expressions as arguments of operator()
2017-01-16 22:21:23 +01:00
#include "src/Core/util/IntegralConstant.h"
Add a simple serialization mechanism. The `Serializer<T>` class implements a binary serialization that can write to (`serialize`) and read from (`deserialize`) a byte buffer. Also added convenience routines for serializing a list of arguments. This will mainly be for testing, specifically to transfer data to and from the GPU.
2021-08-26 14:35:12 -07:00
#include "src/Core/util/Serializer.h"
Add support for fixed-value in symbolic expression, c++11 only for now.
2017-01-19 19:25:29 +01:00
#include "src/Core/util/SymbolicIndex.h"
Fix cuda compilation
2023-05-08 16:15:47 +00:00
#include "src/Core/util/EmulateArray.h"
#include "src/Core/util/MoreMeta.h"
Add support for symbolic expressions as arguments of operator()
2017-01-16 22:21:23 +01:00
part 2 of the reorganization. Benefits/changes: 1) Eigen2 co-installable with Eigen1 without conflict, without affecting programs including either. 2) #include<Eigen/Core> without the .h without conflict with the Core/ directory 3) Uniformize coding style of the CMakeLists.
2007-12-28 16:20:00 +00:00
#include "src/Core/NumTraits.h"
#include "src/Core/MathFunctions.h"
* move memory related stuff to util/Memory.h * clean ugly doxygen inheritence of expressions * keep improving the documentation... slowly !
2008-08-26 19:12:23 +00:00
#include "src/Core/GenericPacketMath.h"
Factorize the 4 copies of tanh implementations, make numext::tanh consistent with array::tanh, enable fast tanh in fast-math mode only.
2016-08-23 14:23:08 +02:00
#include "src/Core/MathFunctionsImpl.h"
bug #1436: fix compilation of Jacobi rotations with ARM NEON, some specializations of internal::conj_helper were missing.
2017-06-15 10:16:30 +02:00
#include "src/Core/arch/Default/ConjHelper.h"
Clean up float16 a.k.a. Eigen::half support in Eigen. Move the definition of half to Core/arch/Default and move arch-specific packet ops to their respective sub-directories.
2019-08-27 11:30:31 -07:00
// Generic half float support
#include "src/Core/arch/Default/Half.h"
Support BFloat16 in Eigen
2020-06-20 19:16:24 +00:00
#include "src/Core/arch/Default/BFloat16.h"
Fix a circular dependency regarding pshift* functions and GenericPacketMathFunctions. Another solution would have been to make pshift* fully generic template functions with partial specialization which is always a mess in c++03.
2019-09-06 09:26:04 +02:00
#include "src/Core/arch/Default/GenericPacketMathFunctionsFwd.h"
move arch-specific code to arch/SSE and arch/AltiVec subdirs. rename the noarch PacketMath.h to DummyPacketMath.h
2008-05-12 08:30:42 +00:00
Preliminary support for AVX512
2015-12-10 15:34:57 -08:00
#if defined EIGEN_VECTORIZE_AVX512
Add support for AVX512-FP16 for vectorizing half precision math
2022-08-17 18:15:21 +00:00
#if defined EIGEN_VECTORIZE_AVX512FP16
#include "src/Core/arch/AVX512/PacketMathFP16.h"
#endif
Preliminary support for AVX512
2015-12-10 15:34:57 -08:00
#include "src/Core/arch/SSE/PacketMath.h"
Add missing SSE/AVX type-casting in AVX512 mode
2018-11-28 16:09:08 +01:00
#include "src/Core/arch/SSE/TypeCasting.h"
Implement AVX512 vectorization of std::complex<float/double>
2018-12-06 15:58:06 +01:00
#include "src/Core/arch/SSE/Complex.h"
Preliminary support for AVX512
2015-12-10 15:34:57 -08:00
#include "src/Core/arch/AVX/PacketMath.h"
Add missing SSE/AVX type-casting in AVX512 mode
2018-11-28 16:09:08 +01:00
#include "src/Core/arch/AVX/TypeCasting.h"
Implement AVX512 vectorization of std::complex<float/double>
2018-12-06 15:58:06 +01:00
#include "src/Core/arch/AVX/Complex.h"
Preliminary support for AVX512
2015-12-10 15:34:57 -08:00
#include "src/Core/arch/AVX512/PacketMath.h"
Fix missing header inclusion and colliding definitions for half type casting, which broke build with -march=native on Haswell/Skylake.
2019-08-30 14:03:29 -07:00
#include "src/Core/arch/AVX512/TypeCasting.h"
Implement AVX512 vectorization of std::complex<float/double>
2018-12-06 15:58:06 +01:00
#include "src/Core/arch/AVX512/Complex.h"
Include SSE packet instructions when compiling with avx512 enabled.
2016-12-19 07:32:48 -08:00
#include "src/Core/arch/SSE/MathFunctions.h"
#include "src/Core/arch/AVX/MathFunctions.h"
Fixed indentation
2016-02-04 10:34:10 -08:00
#include "src/Core/arch/AVX512/MathFunctions.h"
Remove AVX512VL dependency in trsm
2022-04-14 11:35:26 -07:00
#include "src/Core/arch/AVX512/TrsmKernel.h"
Preliminary support for AVX512
2015-12-10 15:34:57 -08:00
#elif defined EIGEN_VECTORIZE_AVX
Added support for AVX to Eigen.
2014-01-29 11:43:05 -08:00
// Use AVX for floats and doubles, SSE for integers
Implement pcplflip, palign, predux and the likes from AVC/complexes
2014-03-27 14:47:00 +01:00
#include "src/Core/arch/SSE/PacketMath.h"
Unify SSE/AVX psin functions. It is based on the SSE version which is much more accurate, though very slightly slower. This changeset also includes the following required changes: - add packet-float to packet-int type traits - add packet float<->int reinterpret casts - add faster pselect for AVX based on blendv
2018-11-27 22:41:51 +01:00
#include "src/Core/arch/SSE/TypeCasting.h"
Implement pcplflip, palign, predux and the likes from AVC/complexes
2014-03-27 14:47:00 +01:00
#include "src/Core/arch/SSE/Complex.h"
Added support for AVX to Eigen.
2014-01-29 11:43:05 -08:00
#include "src/Core/arch/AVX/PacketMath.h"
Unify SSE/AVX psin functions. It is based on the SSE version which is much more accurate, though very slightly slower. This changeset also includes the following required changes: - add packet-float to packet-int type traits - add packet float<->int reinterpret casts - add faster pselect for AVX based on blendv
2018-11-27 22:41:51 +01:00
#include "src/Core/arch/AVX/TypeCasting.h"
Added support for AVX to Eigen.
2014-01-29 11:43:05 -08:00
#include "src/Core/arch/AVX/Complex.h"
bug #1631: fix compilation with ARM NEON and clang, and cleanup the weird pshiftright_and_cast and pcast_and_shiftleft functions.
2018-11-27 23:45:00 +01:00
#include "src/Core/arch/SSE/MathFunctions.h"
#include "src/Core/arch/AVX/MathFunctions.h"
Added support for AVX to Eigen.
2014-01-29 11:43:05 -08:00
#elif defined EIGEN_VECTORIZE_SSE
* throw bad_alloc if exceptions are enabled, after patch by Kenneth Riddile * disable vectorization on MSVC 2005, as it doesn't have all the required intrinsics. require 2008.
2008-12-16 15:17:29 +00:00
#include "src/Core/arch/SSE/PacketMath.h"
Unify SSE/AVX psin functions. It is based on the SSE version which is much more accurate, though very slightly slower. This changeset also includes the following required changes: - add packet-float to packet-int type traits - add packet float<->int reinterpret casts - add faster pselect for AVX based on blendv
2018-11-27 22:41:51 +01:00
#include "src/Core/arch/SSE/TypeCasting.h"
add vectorization of sqrt for float
2009-03-27 14:41:46 +00:00
#include "src/Core/arch/SSE/MathFunctions.h"
add intitial support for the vectorization of complex<float>
2010-07-05 16:18:09 +02:00
#include "src/Core/arch/SSE/Complex.h"
First time it compiles, but fails to pass the tests.
2014-09-09 16:58:48 +00:00
#elif defined(EIGEN_VECTORIZE_ALTIVEC) || defined(EIGEN_VECTORIZE_VSX)
Initial VSX commit
2014-08-29 20:03:49 +00:00
#include "src/Core/arch/AltiVec/PacketMath.h"
New BF16 pcast functions and move type casting to TypeCasting.h
2023-04-18 02:38:38 +00:00
#include "src/Core/arch/AltiVec/TypeCasting.h"
Implement plog and pexp for AltiVec.
2015-07-30 11:12:42 -07:00
#include "src/Core/arch/AltiVec/MathFunctions.h"
Initial VSX commit
2014-08-29 20:03:49 +00:00
#include "src/Core/arch/AltiVec/Complex.h"
Added initial NEON support, most tests pass however we had to use some hackish workarounds as gcc on ARM (both CodeSourcery 4.4.1 used and experimental 4.5) fail to ensure proper alignment with __attribute__((aligned(16))). This has to be fixed upstream to remove the workarounds.
2010-03-03 11:25:41 -06:00
#elif defined EIGEN_VECTORIZE_NEON
#include "src/Core/arch/NEON/PacketMath.h"
Unify SSE/AVX psin functions. It is based on the SSE version which is much more accurate, though very slightly slower. This changeset also includes the following required changes: - add packet-float to packet-int type traits - add packet float<->int reinterpret casts - add faster pselect for AVX based on blendv
2018-11-27 22:41:51 +01:00
#include "src/Core/arch/NEON/TypeCasting.h"
Added vectorized implementation of the exponential function for ARM/NEON
2015-02-10 14:02:38 -08:00
#include "src/Core/arch/NEON/MathFunctions.h"
Added NEON/Complex.h, ~3.5x faster than scalar std::complex<float> minor fix in AltiVec Complex.h
2010-07-10 00:09:29 +03:00
#include "src/Core/arch/NEON/Complex.h"
Add support for Arm SVE This patch adds support for Arm's new vector extension SVE (Scalable Vector Extension). In contrast to other vector extensions that are supported by Eigen, SVE types are inherently *sizeless*. For the use in Eigen we fix their size at compile-time (note that this is not necessary in general, SVE is *length agnostic*). During compilation the flag `-msve-vector-bits=N` has to be set where `N` is a power of two in the range of `128`to `2048`, indicating the length of an SVE vector. Since SVE is rather young, we decided to disable it by default even if it would be available. A user has to enable it explicitly by defining `EIGEN_ARM64_USE_SVE`. This patch introduces the packet types `PacketXf` and `PacketXi` for packets of `float` and `int32_t` respectively. The size of these packets depends on the SVE vector length. E.g. if `-msve-vector-bits=512` is set, `PacketXf` will contain `512/32 = 16` elements. This MR is joint work with Miguel Tairum <miguel.tairum@arm.com>.
2021-01-21 21:11:57 +00:00
#elif defined EIGEN_VECTORIZE_SVE
#include "src/Core/arch/SVE/PacketMath.h"
#include "src/Core/arch/SVE/TypeCasting.h"
#include "src/Core/arch/SVE/MathFunctions.h"
some primitives ported, but missing intrinsics and crash with asm() are a problem
2016-03-27 18:47:49 -04:00
#elif defined EIGEN_VECTORIZE_ZVECTOR
#include "src/Core/arch/ZVector/PacketMath.h"
enable the other includes as well
2016-04-05 05:59:30 -04:00
#include "src/Core/arch/ZVector/MathFunctions.h"
#include "src/Core/arch/ZVector/Complex.h"
Add MIPS changes missing from previous merge.
2018-07-18 12:27:50 -07:00
#elif defined EIGEN_VECTORIZE_MSA
#include "src/Core/arch/MSA/PacketMath.h"
#include "src/Core/arch/MSA/MathFunctions.h"
#include "src/Core/arch/MSA/Complex.h"
Add architecture definition files for Qualcomm Hexagon Vector Extension (HVX)
2023-08-01 17:47:57 +00:00
#elif defined EIGEN_VECTORIZE_HVX
#include "src/Core/arch/HVX/PacketMath.h"
* split PacketMath.h to SSE and Altivec specific files * improved the flexibility of the new product implementation, now all sizes seems to be properly handled.
2008-05-05 17:19:47 +00:00
#endif
Updates corresponding to the latest round of PR feedback The major changes are 1. Moving CUDA/PacketMath.h to GPU/PacketMath.h 2. Moving CUDA/MathFunctions.h to GPU/MathFunction.h 3. Moving CUDA/CudaSpecialFunctions.h to GPU/GpuSpecialFunctions.h The above three changes effectively enable the Eigen "Packet" layer for the HIP platform 4. Merging the "hip_basic" and "cuda_basic" unit tests into one ("gpu_basic") 5. Updating the "EIGEN_DEVICE_FUNC" marking in some places The change has been tested on the HIP and CUDA platforms.
2018-07-11 10:39:54 -04:00
#if defined EIGEN_VECTORIZE_GPU
#include "src/Core/arch/GPU/PacketMath.h"
#include "src/Core/arch/GPU/MathFunctions.h"
Clean up float16 a.k.a. Eigen::half support in Eigen. Move the definition of half to Core/arch/Default and move arch-specific packet ops to their respective sub-directories.
2019-08-27 11:30:31 -07:00
#include "src/Core/arch/GPU/TypeCasting.h"
Vector primitives for CUDA
2014-10-03 19:45:19 -07:00
#endif
[SYCL] This PR adds the minimum modifications to Eigen core required to run Eigen unsupported modules on devices supporting SYCL. * Adding SYCL memory model * Enabling/Disabling SYCL backend in Core * Supporting Vectorization
2019-06-27 12:25:09 +01:00
#if defined(EIGEN_USE_SYCL)
Creating separate SYCL required PR for uncontroversial files.
2018-08-03 16:59:15 +01:00
#include "src/Core/arch/SYCL/InteropHeaders.h"
[SYCL] This PR adds the minimum modifications to Eigen core required to run Eigen unsupported modules on devices supporting SYCL. * Adding SYCL memory model * Enabling/Disabling SYCL backend in Core * Supporting Vectorization
2019-06-27 12:25:09 +01:00
#if !defined(EIGEN_DONT_VECTORIZE_SYCL)
Creating separate SYCL required PR for uncontroversial files.
2018-08-03 16:59:15 +01:00
#include "src/Core/arch/SYCL/PacketMath.h"
#include "src/Core/arch/SYCL/MathFunctions.h"
#include "src/Core/arch/SYCL/TypeCasting.h"
#endif
[SYCL] This PR adds the minimum modifications to Eigen core required to run Eigen unsupported modules on devices supporting SYCL. * Adding SYCL memory model * Enabling/Disabling SYCL backend in Core * Supporting Vectorization
2019-06-27 12:25:09 +01:00
#endif
Fix a circular dependency regarding pshift* functions and GenericPacketMathFunctions. Another solution would have been to make pshift* fully generic template functions with partial specialization which is always a mess in c++03.
2019-09-06 09:26:04 +02:00
factorize default performance related settings to a single file included after the architecture specific files such that they can be adapted by each platform.
2010-03-03 18:47:58 +01:00
#include "src/Core/arch/Default/Settings.h"
Fix a circular dependency regarding pshift* functions and GenericPacketMathFunctions. Another solution would have been to make pshift* fully generic template functions with partial specialization which is always a mess in c++03.
2019-09-06 09:26:04 +02:00
// This file provides generic implementations valid for scalar as well
#include "src/Core/arch/Default/GenericPacketMathFunctions.h"
put inline keywords everywhere appropriate. So we don't need anymore to pass -finline-limit=1000 to gcc to get good performance. By the way some cleanup.
2008-05-12 17:34:46 +00:00
Add TernaryFunctors and the betainc SpecialFunction. TernaryFunctors and their executors allow operations on 3-tuples of inputs. API fully implemented for Arrays and Tensors based on binary functors. Ported the cephes betainc function (regularized incomplete beta integral) to Eigen, with support for CPU and GPU, floats, doubles, and half types. Added unit tests in array.cpp and cxx11_tensor_cuda.cu Collapsed revision * Merged helper methods for betainc across floats and doubles. * Added TensorGlobalFunctions with betainc(). Removed betainc() from TensorBase. * Clean up CwiseTernaryOp checks, change igamma_helper to cephes_helper. * betainc: merge incbcf and incbd into incbeta_cfe. and more cleanup. * Update TernaryOp and SpecialFunctions (betainc) based on review comments.
2016-06-02 17:04:19 -07:00
#include "src/Core/functors/TernaryFunctors.h"
Split the huge Functors.h file
2013-11-06 10:36:10 +01:00
#include "src/Core/functors/BinaryFunctors.h"
#include "src/Core/functors/UnaryFunctors.h"
#include "src/Core/functors/NullaryFunctors.h"
#include "src/Core/functors/StlFunctors.h"
Make swap unit test work with evaluators
2013-12-02 15:07:45 +01:00
#include "src/Core/functors/AssignmentFunctors.h"
Split the huge Functors.h file
2013-11-06 10:36:10 +01:00
Move CUDA/Complex.h to GPU/Complex.h, remove TensorReductionCuda.h The `Complex.h` file applies equally to HIP/CUDA, so placing under the generic `GPU` folder. The `TensorReductionCuda.h` has already been deprecated, now removing for the next Eigen version.
2021-10-08 11:36:27 -07:00
// Specialized functors for GPU.
#ifdef EIGEN_GPUCC
#include "src/Core/arch/GPU/Complex.h"
More clearly disable the inclusion of src/Core/arch/CUDA/Complex.h without CUDA
2018-07-18 13:51:36 +02:00
#endif
Added support for constand std::complex numbers on GPU
2016-10-03 11:06:24 -07:00
Vectorize fp16 tanh and logistic functions on Neon Activates vectorization of the Eigen::half versions of the tanh and logistic functions when they run on Neon. Both functions convert their inputs to float before computing the output, and as a result of this commit, the conversions and the computation in float are vectorized.
2021-10-26 12:24:33 -07:00
// Specializations of vectorized activation functions for NEON.
#ifdef EIGEN_VECTORIZE_NEON
#include "src/Core/arch/NEON/UnaryFunctors.h"
#endif
Large code refactoring: - generalize some utilities and move them to Meta (size(), array_size()) - move handling of all and single indices to IndexedViewHelper.h - several cleanup changes
2017-01-11 17:24:02 +01:00
#include "src/Core/util/IndexedViewHelper.h"
Add support for automatic-size deduction in reshaped, e.g.: mat.reshaped(4,AutoSize); <-> mat.reshaped(4,mat.size()/4);
2017-02-21 15:57:25 +01:00
#include "src/Core/util/ReshapedHelper.h"
Initial commit to add a generic indexed-based view of matrices. This version already works as a read-only expression. Numerous refactoring, renaming, extension, tuning passes are expected...
2017-01-06 00:01:44 +01:00
#include "src/Core/ArithmeticSequence.h"
Collapsed revision (based on pull request PR-325) * Support compiling without IO streams Add the preprocessor definition EIGEN_NO_IO which, if defined, disables all use of the IO streams part of the standard library.
2018-10-22 21:14:40 +02:00
#ifndef EIGEN_NO_IO
#include "src/Core/IO.h"
#endif
rename Coeffs.h -> DenseCoeffsBase.h
2010-05-08 16:02:13 -04:00
#include "src/Core/DenseCoeffsBase.h"
add a DenseBase class for MAtrixBase and ArrayBase and more code factorisation
2009-12-04 23:17:14 +01:00
#include "src/Core/DenseBase.h"
part 2 of the reorganization. Benefits/changes: 1) Eigen2 co-installable with Eigen1 without conflict, without affecting programs including either. 2) #include<Eigen/Core> without the .h without conflict with the Core/ directory 3) Uniformize coding style of the CMakeLists.
2007-12-28 16:20:00 +00:00
#include "src/Core/MatrixBase.h"
Renamed AnyMatrixBase to EigenBase.
2010-02-20 15:26:02 +01:00
#include "src/Core/EigenBase.h"
* throw bad_alloc if exceptions are enabled, after patch by Kenneth Riddile * disable vectorization on MSVC 2005, as it doesn't have all the required intrinsics. require 2008.
2008-12-16 15:17:29 +00:00
Add an option to test evaluators globally
2013-11-07 16:38:14 +01:00
#include "src/Core/Product.h"
#include "src/Core/CoreEvaluators.h"
#include "src/Core/AssignEvaluator.h"
work around Doxygen bug triggered by r814874, which caused many classes to disappear from the docs.
2008-06-02 19:29:23 +00:00
#ifndef EIGEN_PARSED_BY_DOXYGEN // work around Doxygen bug triggered by Assign.h r814874
// at least confirmed with Doxygen 1.5.5 and 1.5.6
* throw bad_alloc if exceptions are enabled, after patch by Kenneth Riddile * disable vectorization on MSVC 2005, as it doesn't have all the required intrinsics. require 2008.
2008-12-16 15:17:29 +00:00
#include "src/Core/Assign.h"
work around Doxygen bug triggered by r814874, which caused many classes to disappear from the docs.
2008-06-02 19:29:23 +00:00
#endif
* throw bad_alloc if exceptions are enabled, after patch by Kenneth Riddile * disable vectorization on MSVC 2005, as it doesn't have all the required intrinsics. require 2008.
2008-12-16 15:17:29 +00:00
Make ArrayBase operator+=(scalar) and -=(scalar) use SelfCwiseBinaryOp optimization
2013-08-19 16:40:50 +02:00
#include "src/Core/ArrayBase.h"
add overloads of lazyAssign to detect common aliasing issue with transpose and adjoint
2009-08-15 22:19:29 +02:00
#include "src/Core/util/BlasUtil.h"
renaming: ei_matrix_storage -> DenseStorage DenseStorageBase -> PlainObjectBase
2010-10-20 09:34:13 -04:00
#include "src/Core/DenseStorage.h"
hehe, the complicated nesting scheme in Flagged in the previous commit was a sign that we were doing something wrong. In fact, having NestByValue as a special case of Flagged was wrong, and the previous commit, while not buggy, was inefficient because then when the resulting NestByValue xpr was nested -- hence copied -- the original xpr which was already nested by value was copied again; hence instead of 1 copy we got 3 copies. The solution was to ressuscitate the old Temporary.h (renamed NestByValue.h) as it was the right approach.
2008-05-28 05:14:16 +00:00
#include "src/Core/NestByValue.h"
Remove deprecated code not used by evaluators
2014-09-18 15:15:27 +02:00
// #include "src/Core/ForceAlignedAccess.h"
add ReturnByValue pseudo expression for in-place evaluation with a return-by-value API style (will soon use it for the transform products)
2009-03-04 13:00:00 +00:00
#include "src/Core/ReturnByValue.h"
As proposed on the list: - rename EvalBeforeAssignBit to MayAliasBit - make .lazy() remove the MayAliasBit only, and mark it as deprecated - add a NoAlias pseudo expression, and MatrixBase::noalias() function Todo: - we have to decide whether += and -= assume no aliasing by default ? - once we agree on the API: update the Sparse module and the unit tests respectively.
2009-08-15 18:35:51 +02:00
#include "src/Core/NoAlias.h"
renaming: ei_matrix_storage -> DenseStorage DenseStorageBase -> PlainObjectBase
2010-10-20 09:34:13 -04:00
#include "src/Core/PlainObjectBase.h"
part 2 of the reorganization. Benefits/changes: 1) Eigen2 co-installable with Eigen1 without conflict, without affecting programs including either. 2) #include<Eigen/Core> without the .h without conflict with the Core/ directory 3) Uniformize coding style of the CMakeLists.
2007-12-28 16:20:00 +00:00
#include "src/Core/Matrix.h"
Fixed ctor from const raw data for Matrices and added the missing implementation for Arrays. Fixed a warning regarding the conversion from int to bool in MapBase.
2010-12-15 15:19:51 +01:00
#include "src/Core/Array.h"
Add TernaryFunctors and the betainc SpecialFunction. TernaryFunctors and their executors allow operations on 3-tuples of inputs. API fully implemented for Arrays and Tensors based on binary functors. Ported the cephes betainc function (regularized incomplete beta integral) to Eigen, with support for CPU and GPU, floats, doubles, and half types. Added unit tests in array.cpp and cxx11_tensor_cuda.cu Collapsed revision * Merged helper methods for betainc across floats and doubles. * Added TensorGlobalFunctions with betainc(). Removed betainc() from TensorBase. * Clean up CwiseTernaryOp checks, change igamma_helper to cephes_helper. * betainc: merge incbcf and incbd into incbeta_cfe. and more cleanup. * Update TernaryOp and SpecialFunctions (betainc) based on review comments.
2016-06-02 17:04:19 -07:00
#include "src/Core/CwiseTernaryOp.h"
Patch by Gael Guennebaud: coeff-wise binary operators. This unifies + and - and moreover this patch introduces coeff-wise * and / based on this. Also, corresponding test.
2008-02-29 14:35:14 +00:00
#include "src/Core/CwiseBinaryOp.h"
* Added generic unary operators (replace Opposite and Conjugate) * functor templates are not template template parameter anymore (this allows to make templated functors !) * Main page: extented compiler discussion * A small hack to support gcc 3.4 and 4.0 (see the main page) * Fix a cast type issue in Cast * Various doxygen updates (mainly Cwise stuff and added doxygen groups in MatrixBase to split the huge memeber list, still not perfect though) * Updated Gael's email address
2008-03-03 10:52:44 +00:00
#include "src/Core/CwiseUnaryOp.h"
Fix a couple of issue with the vectorization. In particular, default ei_p* functions are provided to handle not suported types seemlessly. Added a generic null-ary expression with null-ary functors. They replace Zero, Ones, Identity and Random.
2008-04-24 18:35:39 +00:00
#include "src/Core/CwiseNullaryOp.h"
* add a writable generic coeff wise expression (CwiseUnaryView) * add writable .real() and .imag() functions
2009-05-20 15:41:23 +02:00
#include "src/Core/CwiseUnaryView.h"
fix compilation by including file in correct order
2010-07-19 23:32:13 +02:00
#include "src/Core/SelfCwiseBinaryOp.h"
* added optimized paths for matrix-vector and vector-matrix products (using either a cache friendly strategy or re-using dot-product vectorized implementation) * add LinearAccessBit to Transpose
2008-07-09 22:30:18 +00:00
#include "src/Core/Dot.h"
re-implement stableNorm using a homemade blocky and vectorization friendly algorithm (slow if no vectorization)
2009-07-17 16:22:39 +02:00
#include "src/Core/StableNorm.h"
miserable half-working state, commiting to a fork just in case, just to perfect my day, my hard disk would die. Will write a more detailed commit message once it's working.
2010-02-18 20:42:38 -05:00
#include "src/Core/Stride.h"
Move is_diagonal to XprHelper, forward declare Ref
2014-02-18 11:03:59 +01:00
#include "src/Core/MapBase.h"
* Big change in Block and Map: - added a MapBase base xpr on top of which Map and the specialization of Block are implemented - MapBase forces both aligned loads (and aligned stores, see below) in expressions such as "x.block(...) += other_expr" * Significant vectorization improvement: - added a AlignedBit flag meaning the first coeff/packet is aligned, this allows to not generate extra code to deal with the first unaligned part - removed all unaligned stores when no unrolling - removed unaligned loads in Sum when the input as the DirectAccessBit flag * Some code simplification in CacheFriendly product * Some minor documentation improvements
2008-08-09 18:41:24 +00:00
#include "src/Core/Map.h"
Move is_diagonal to XprHelper, forward declare Ref
2014-02-18 11:03:59 +01:00
#include "src/Core/Ref.h"
part 2 of the reorganization. Benefits/changes: 1) Eigen2 co-installable with Eigen1 without conflict, without affecting programs including either. 2) #include<Eigen/Core> without the .h without conflict with the Core/ directory 3) Uniformize coding style of the CMakeLists.
2007-12-28 16:20:00 +00:00
#include "src/Core/Block.h"
add a VectorBlock expr as a specialization of Block
2009-07-05 11:33:55 +02:00
#include "src/Core/VectorBlock.h"
Initial commit to add a generic indexed-based view of matrices. This version already works as a read-only expression. Numerous refactoring, renaming, extension, tuning passes are expected...
2017-01-06 00:01:44 +01:00
#include "src/Core/IndexedView.h"
Use fix<> API to specify compile-time reshaped sizes.
2017-01-29 15:20:35 +01:00
#include "src/Core/Reshaped.h"
part 2 of the reorganization. Benefits/changes: 1) Eigen2 co-installable with Eigen1 without conflict, without affecting programs including either. 2) #include<Eigen/Core> without the .h without conflict with the Core/ directory 3) Uniformize coding style of the CMakeLists.
2007-12-28 16:20:00 +00:00
#include "src/Core/Transpose.h"
#include "src/Core/DiagonalMatrix.h"
constant Diagonal ---> DiagonalBits introduce ei_is_diagonal to check for it DiagonalCoeffs ---> Diagonal and allow Index to by Dynamic -> add MatrixBase::diagonal(int) with unittest and doc
2009-05-10 16:24:39 +00:00
#include "src/Core/Diagonal.h"
new implementation of diagonal matrices and diagonal matrix expressions
2009-06-28 21:27:37 +02:00
#include "src/Core/DiagonalProduct.h"
Feature/skew symmetric matrix3
2022-09-08 20:44:40 +00:00
#include "src/Core/SkewSymmetricMatrix3.h"
* Added a generic *redux* mini framework allowing custom redux operations as well as partial redux (vertical or horizontal redux). Includes shortcuts for: sum, minCoeff and maxCoeff. There is no shortcut for the partial redux. * Added a generic *visitor* mini framework. A visitor is a custom object sequentially applied on each coefficient with knowledge of its value and coordinates. It is currentlly used to implement minCoeff(int*,int*) and maxCoeff(int*,int*). findBiggestCoeff is now a shortcut for "this->cwiseAbs().maxCoeff(i,j)" * Added coeff-wise min and max. * fixed an issue with ei_pow(int,int) and gcc < 4.3 or ICC
2008-03-16 14:36:25 +00:00
#include "src/Core/Redux.h"
#include "src/Core/Visitor.h"
part 2 of the reorganization. Benefits/changes: 1) Eigen2 co-installable with Eigen1 without conflict, without affecting programs including either. 2) #include<Eigen/Core> without the .h without conflict with the Core/ directory 3) Uniformize coding style of the CMakeLists.
2007-12-28 16:20:00 +00:00
#include "src/Core/Fuzzy.h"
alpha 3.1. in this commit: - finally get the Eval stuff right. get back to having Eval as a subclass of Matrix with limited functionality, and then, add a typedef MatrixType to get the actual matrix type. - add swap(), findBiggestCoeff() - bugfix by Ramon in Transpose - new demo: doc/echelon.cpp
2008-01-15 13:55:47 +00:00
#include "src/Core/Swap.h"
* Added support for a comma initializer: mat.block(i,j,2,2) << 1, 2, 3, 4; If the number of coefficients does not match the matrix size, then an assertion is raised. No support for xpr on the right side for the moment. * Added support for assertion checking. This allows to test that an assertion is indeed raised when it should be. * Fixed a mistake in the CwiseUnary example.
2008-03-08 19:02:24 +00:00
#include "src/Core/CommaInitializer.h"
add a stupid Product<A,B> expression produced by prod(a,b), and implement a first version of its evaluator
2011-03-23 16:12:21 +01:00
#include "src/Core/GeneralProduct.h"
Add a Solve expression for uniform treatment of solve() methods.
2014-02-19 11:33:29 +01:00
#include "src/Core/Solve.h"
Split LU/Inverse.h to Core/Inverse.h for the generic Inverse expression, and LU/InverseImpl.h for the dense implementation of dense.inverse()
2014-02-24 11:49:30 +01:00
#include "src/Core/Inverse.h"
Add LU::transpose().solve() and LU::adjoint().solve() API.
2015-12-01 14:38:47 +01:00
#include "src/Core/SolverBase.h"
Clean a bit the implementation of inverse permutations
2015-10-08 18:36:39 +02:00
#include "src/Core/PermutationMatrix.h"
#include "src/Core/Transpositions.h"
* bybye Part, welcome TriangularView and SelfAdjointView. * move solveTriangular*() to TriangularView::solve*() * move .llt() to SelfAdjointView * add a high level wrapper to the efficient selfadjoint * vector product * improve LLT so that we can specify which triangular part is meaningless => there are still many things to do (doc, cleaning, improve the matrix products, etc.)
2009-07-06 23:43:20 +02:00
#include "src/Core/TriangularMatrix.h"
#include "src/Core/SelfAdjointView.h"
bug #466: better fix for the race condition: this new patch add an initParallel() function which must be called at the initialization time of any multi-threaded application calling Eigen from multiple threads.
2012-06-14 14:24:15 +02:00
#include "src/Core/products/GeneralBlockPanelKernel.h"
Generalize parallel GEMM implementation in Core to work with ThreadPool in addition to OpenMP.
2023-11-10 17:42:30 +00:00
#ifdef EIGEN_GEMM_THREADPOOL
#include "ThreadPool"
#endif
clean a bit the parallelizer
2010-02-22 11:08:37 +01:00
#include "src/Core/products/Parallelizer.h"
Get rid of GeneralProduct<> for GemmProduct
2014-02-21 16:43:03 +01:00
#include "src/Core/ProductEvaluators.h"
more product refactoring
2009-08-06 12:20:02 +02:00
#include "src/Core/products/GeneralMatrixVector.h"
#include "src/Core/products/GeneralMatrixMatrix.h"
fix inclusion order
2012-06-12 11:40:33 +02:00
#include "src/Core/SolveTriangular.h"
generalize our internal rank K update routine to support more general A*B product while evaluating only one triangular part and make it available via, e.g.: R.triangularView<Lower>() += s * A * B;
2010-11-10 18:58:55 +01:00
#include "src/Core/products/GeneralMatrixMatrixTriangular.h"
more product refactoring
2009-08-06 12:20:02 +02:00
#include "src/Core/products/SelfadjointMatrixVector.h"
#include "src/Core/products/SelfadjointMatrixMatrix.h"
Implement efficient sefladjoint product (aka SYRK) : C += alpha * U U^T It is currently available via SelfAdjointView::rankKupdate. TODO: allows to write SelfAdjointView += u * u.adjoint()
2009-07-23 19:01:20 +02:00
#include "src/Core/products/SelfadjointProduct.h"
Add an efficient rank2 update function (like the level2 blas xSYR2 routine). Note that it is already used in Tridiagonalization.
2009-07-11 21:14:59 +02:00
#include "src/Core/products/SelfadjointRank2Update.h"
add triangular * vector product
2009-07-13 13:17:55 +02:00
#include "src/Core/products/TriangularMatrixVector.h"
trmm is now working in all storage order configurations
2009-07-27 10:27:01 +02:00
#include "src/Core/products/TriangularMatrixMatrix.h"
more product refactoring
2009-08-06 12:20:02 +02:00
#include "src/Core/products/TriangularSolverMatrix.h"
trsv: add support for inner-stride!=1, reduce code instanciation, move implementation to a new products/XX.h file
2010-11-05 12:43:14 +01:00
#include "src/Core/products/TriangularSolverVector.h"
started an implementation of BandMatrix: at least the read/write access to the main/sub/super diagonals seems to work well.
2009-07-14 23:27:37 +02:00
#include "src/Core/BandMatrix.h"
Move the definition of DenseBase::InnerIterator to Core module. (needed to make blueNorm generic)
2013-01-15 22:03:54 +01:00
#include "src/Core/CoreIterators.h"
Add matrix condition estimator module that implements the Higham/Hager algorithm from http://www.maths.manchester.ac.uk/~higham/narep/narep135.pdf used in LPACK. Add rcond() methods to FullPivLU and PartialPivLU.
2016-04-01 10:27:59 -07:00
#include "src/Core/ConditionEstimator.h"
added an *optional* Eigen2 dynamic library. it allows the possiblity to save some compilation time by linking to it *and* defining the token EIGEN_EXTERN_INSTANCIATIONS
2008-05-31 23:21:49 +00:00
Fix non-VSX PowerPC build
2022-08-08 18:18:17 +00:00
#if defined(EIGEN_VECTORIZE_VSX)
Fix excessive GEBP register spilling for 32-bit NEON. Clang does a poor job of optimizing the GEBP microkernel on 32-bit ARM, leading to excessive 16-byte register spills, slowing down basic f32 matrix multiplication by approx 50%. By specializing `gebp_traits`, we can eliminate the register spills. Volatile inline ASM both acts as a barrier to prevent reordering and enforces strict register use. In a simple f32 matrix multiply example, this modification reduces 16-byte spills from 109 instances to zero, leading to a 1.5x speed increase (search for `16-byte Spill` in the assembly in https://godbolt.org/z/chsPbE). This is a replacement of !379. See there for further discussion. Also moved `gebp_traits` specializations for NEON to `Eigen/src/Core/arch/NEON/GeneralBlockPanelKernel.h` to be alongside other NEON-specific code. Fixes #2138.
2021-02-03 08:18:28 -08:00
#include "src/Core/arch/AltiVec/MatrixProduct.h"
#elif defined EIGEN_VECTORIZE_NEON
#include "src/Core/arch/NEON/GeneralBlockPanelKernel.h"
Add AVX512 s/dgemm optimizations for compute kernel (2nd try)
2022-05-28 02:00:21 +00:00
#endif
#if defined(EIGEN_VECTORIZE_AVX512)
#include "src/Core/arch/AVX512/GemmKernel.h"
Add architecture definition files for Qualcomm Hexagon Vector Extension (HVX)
2023-08-01 17:47:57 +00:00
#endif
#if defined(EIGEN_VECTORIZE_HVX)
#include "src/Core/arch/HVX/GeneralBlockPanelKernel.h"
- Vectorizing MMA packing. - Optimizing MMA kernel. - Adding PacketBlock store to blas_data_mapper.
2020-05-19 19:24:11 +00:00
#endif
finish to merge Array into Core: - mv Array/* into Core/ - merge Functors.h files, and move Norms.h into Dot.h
2010-06-19 23:36:38 +02:00
#include "src/Core/Select.h"
#include "src/Core/VectorwiseOp.h"
bug #65: add vectorization of partial reductions along the outer-dimension, for instance: colmajor_mat.rowwise().mean()
2018-10-09 23:36:50 +02:00
#include "src/Core/PartialReduxEvaluator.h"
finish to merge Array into Core: - mv Array/* into Core/ - merge Functors.h files, and move Norms.h into Dot.h
2010-06-19 23:36:38 +02:00
#include "src/Core/Random.h"
#include "src/Core/Replicate.h"
#include "src/Core/Reverse.h"
#include "src/Core/ArrayWrapper.h"
bug #231: initial implementation of STL iterators for dense expressions
2018-10-01 23:21:37 +02:00
#include "src/Core/StlIterators.h"
Moved the Array module to Core.
2010-01-20 20:51:01 +01:00
- split and rename defined tokens to enable the use of BLAS/Lapack/VML/etc - include MKL headers outside the Eigen namespace.
2011-12-09 10:06:49 +01:00
#ifdef EIGEN_USE_BLAS
Remove all references to MKL in BLAS wrappers.
2016-04-11 16:04:09 +02:00
#include "src/Core/products/GeneralMatrixMatrix_BLAS.h"
#include "src/Core/products/GeneralMatrixVector_BLAS.h"
#include "src/Core/products/GeneralMatrixMatrixTriangular_BLAS.h"
#include "src/Core/products/SelfadjointMatrixMatrix_BLAS.h"
#include "src/Core/products/SelfadjointMatrixVector_BLAS.h"
#include "src/Core/products/TriangularMatrixMatrix_BLAS.h"
#include "src/Core/products/TriangularMatrixVector_BLAS.h"
#include "src/Core/products/TriangularSolverMatrix_BLAS.h"
- split and rename defined tokens to enable the use of BLAS/Lapack/VML/etc - include MKL headers outside the Eigen namespace.
2011-12-09 10:06:49 +01:00
#endif // EIGEN_USE_BLAS
#ifdef EIGEN_USE_MKL_VML
Intel(R) MKL support added. * * * License disclaimer changed to BSD license for MKL_support.h * * * Pardiso support fixed, test added. blas/lapack tests fixed: Scalar parameter was added in Cholesky, product_matrix_vector_triangular remaned to triangular_matrix_vector_product. * * * PARDISO test was added physically.
2011-12-05 14:52:21 +07:00
#include "src/Core/Assign_MKL.h"
#endif
import evaluator works
2011-03-23 11:54:00 +01:00
finish to merge Array into Core: - mv Array/* into Core/ - merge Functors.h files, and move Norms.h into Dot.h
2010-06-19 23:36:38 +02:00
#include "src/Core/GlobalFunctions.h"
Add IWYU export pragmas to top-level headers.
2023-02-08 17:40:31 +00:00
// IWYU pragma: end_exports
add support for global math function for array
2010-01-27 23:23:59 +01:00
shut up stupid ICC warnings
2011-02-22 09:31:22 -05:00
#include "src/Core/util/ReenableStupidWarnings.h"
Patch by Kenneth Riddile: disable MSVC warnings, reenable them outside of Eigen, and add a MSVC-friendly path in StaticAssert.
2008-12-18 20:48:02 +00:00
Issue an error in case of direct inclusion of internal headers.
2021-09-10 19:12:26 +00:00
#endif // EIGEN_CORE_MODULE_H
Reference in New Issue Copy Permalink