| Commit message (Collapse) | Author | Age |
|---|
| ... | |
| | |
| |
| |
| |
| |
| |
| | |
https://bitbucket.org/eigen/eigen/commits/77b447c24e3344e43ff64eb932d4bb35a2db01ce
While providing a 50% speedup on Haswell+ processors, the large relative error outside [-18, 18] in this approximation causes problems, e.g., when computing gradients of activation functions like softplus in neural networks.
|
| |\ \
| | |
| | |
| | | |
Fix shadow warnings in AlignedBox and SparseBlock
|
| | | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | | |
Recent changes have introduced the following build error when compiling with HIPCC
---------
unsupported/test/../../Eigen/src/Core/GenericPacketMath.h:254:58: error: 'ldexp': no overloaded function has restriction specifiers that are compatible with the ambient context 'pldexp'
---------
The fix for the error is to pick the math function(s) from the global namespace (where they are declared as device functions in the HIP header files) when compiling with HIPCC.
|
| |\ \ \
| | | |
| | | |
| | | |
| | | |
| | | | |
SYCL Backend
Approved-by: Rasmus Larsen <rmlarsen@google.com>
|
| |/ / /
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | | |
branch.
* Unifying all loadLocalTile from lhs and rhs to an extract_block function.
* Adding get_tensor operation which was missing in TensorContractionMapper.
* Adding the -D method missing from cmake for Disable_Skinny Contraction operation.
* Wrapping all the indices in TensorScanSycl into Scan parameter struct.
* Fixing typo in Device SYCL
* Unifying load to private register for tall/skinny no shared
* Unifying load to vector tile for tensor-vector/vector-tensor operation
* Removing all the LHS/RHS class for extracting data from global
* Removing Outputfunction from TensorContractionSkinnyNoshared.
* Combining the local memory version of tall/skinny and normal tensor contraction into one kernel.
* Combining the no-local memory version of tall/skinny and normal tensor contraction into one kernel.
* Combining General Tensor-Vector and VectorTensor contraction into one kernel.
* Making double buffering optional for Tensor contraction when local memory is version is used.
* Modifying benchmark to accept custom Reduction Sizes
* Disabling AVX optimization for SYCL backend on the host to allow SSE optimization to the host
* Adding Test for SYCL
* Modifying SYCL CMake
|
| |/ / |
|
| | |
| |
| |
| | |
to fix GPU build.
|
| | |
| |
| |
| | |
std::lexp.
|
| |\ \
| | |
| | |
| | | |
Fix warnings
|
| | | | |
|
| | | | |
|
| | | | |
|
| | | |
| | |
| | |
| | | |
with cuda.
|
| |/ / |
|
| | | |
|
| | |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| | |
Confirm that install directory is identical
before and after this simplifying patch.
```bash
hg clone <<Eigen>>
mkdir eigen-bld
cd eigen-bld
cmake ../Eigen -DCMAKE_INSTALL_PREFIX:PATH=/tmp/bef
make install
find /tmp/pre_eigen_modernize >/tmp/bef
# Apply this patch
cmake ../Eigen -DCMAKE_INSTALL_PREFIX:PATH=/tmp/aft
make install
find /tmp/post_eigen_modernize |sed 's/post_e/pre_e/g' >/tmp/aft
diff /tmp/bef /tmp/aft
```
|
| | |
| |
| |
| |
| |
| |
| |
| |
| | |
Features committed in 2016 have required cmake verison 2.8.11.
`sergiu Tue Nov 22 12:25:06 2016 +0100: target_compile_definitions`
Set the minimum cmake version to the minimum version that
is capable of compiling or installing the code base.
|
| | | |
|
| | |
| |
| |
| |
| | |
Ancient CMake versions required upper-case commands. Later command names
became case-insensitive. Now the preferred style is lower-case.
|
| | |
| |
| |
| |
| |
| | |
Ancient versions of CMake required else(), endif(), and similar block
termination commands to have arguments matching the command starting the block.
This is no longer the preferred style.
|
| | |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| | |
2. Simplify handling of special cases by taking advantage of the fact that the
builtin vrsqrt approximation handles negative, zero and +inf arguments correctly.
This speeds up the SSE and AVX implementations by ~20%.
3. Make the Newton-Raphson formula used for rsqrt more numerically robust:
Before: y = y * (1.5 - x/2 * y^2)
After: y = y * (1.5 - y * (x/2) * y)
Forming y^2 can overflow for very large or very small (denormalized) values of x, while x*y ~= 1. For AVX512, this makes it possible to compute accurate results for denormal inputs down to ~1e-42 in single precision.
4. Add a faster double precision implementation for Knights Landing using the vrsqrt28 instruction and a single Newton-Raphson iteration.
Benchmark results: https://bitbucket.org/snippets/rmlarsen/5LBq9o
|
| | |
| |
| |
| | |
plog/pexp, but the later was disabled on some compilers
|
| | | |
|
| | |
| |
| |
| | |
constant ADs.
|
| | | |
|
| | | |
|
| | | |
|
| | |
| |
| |
| | |
fixed-size matrices.
|
| | | |
|
| | | |
|
| | |
| |
| |
| | |
Patch adapted from Hans Johnson's PR 748.
|
| | |
| |
| |
| | |
device.deallocate() / evaluator.cleanup() complete, since the device might be destroyed after on_done() runs.
|
| | | |
|
| |/ |
|
| |
|
|
| |
Add a new EIGEN_HAS_INTRINSIC_INT128 macro, and use this instead of __SIZEOF_INT128__. This fixes related issues with TensorIntDiv.h when building with Clang for Windows, where support for 128-bit integer arithmetic is advertised but broken in practice.
|
| |
|
|
|
|
|
|
| |
https://bitbucket.org/eigen/eigen/commits/668ab3fc474e54c7919eda4fbaf11f3a99246494
.
std::array is still not supported in CUDA device code on Windows.
|
| |\
| |
| |
| | |
Remove internal::smart_copy and replace with std::copy
|
| | | |
|
| |/ |
|
| | |
|
| | |
|
| |
|
|
|
| |
* The specialization of array class in the different namespace for GCC<=6.4
* The implicit call to `std::array` constructor using the initializer list for GCC <=6.1
|
| |\
| |
| |
| | |
Fix for the HIP build+test errors.
|
| | |
| |
| |
| | |
Add async evaluation to a number of ops.
|
| | |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| | |
The errors were introduced by this commit :
After the above mentioned commit, some of the tests started failing with the following error
```
Built target cxx11_tensor_reduction
Building HIPCC object unsupported/test/CMakeFiles/cxx11_tensor_reduction_gpu_5.dir/cxx11_tensor_reduction_gpu_5_generated_cxx11_tensor_reduction_gpu.cu.o
In file included from /home/rocm-user/eigen/unsupported/test/cxx11_tensor_reduction_gpu.cu:16:
In file included from /home/rocm-user/eigen/unsupported/Eigen/CXX11/Tensor:117:
/home/rocm-user/eigen/unsupported/Eigen/CXX11/src/Tensor/TensorBlockV2.h:155:5: error: the field type is not amp-compatible
DestinationBufferKind m_kind;
^
/home/rocm-user/eigen/unsupported/Eigen/CXX11/src/Tensor/TensorBlockV2.h:211:3: error: the field type is not amp-compatible
DestinationBuffer m_destination;
^
```
For some reason HIPCC does not like device code to contain enum types which do not have the base-type explicitly declared. The fix is trivial, explicitly state "int" as the basetype
|
| |/
|
|
| |
c++11 functionality with older compilers.
|
| | |
|
| | |
|
| | |
|
| | |
|
Rasmus Munk Larsen
Deven Desai
Mehdi Goli
Eugene Zhulenev
Joel Holdsworth
Anshul Jaiswal
Gael Guennebaud
Hans Johnson
Sakshi Goynar