Permalink
Commits on Jun 5, 2018
  1. Merge pull request #263 from naibaf7/naibaf7/master

    karlrupp committed Jun 5, 2018
    Update OpenCL header files.
  2. Merge pull request #262 from IncinX/master

    karlrupp committed Jun 5, 2018
    Fix FindOpenCL.cmake which doesn't find the OpenCL SDK on Windows
Commits on Apr 20, 2018
Commits on Apr 10, 2018
  1. Armadillo: Added .sync() for Armadillo version 8 and above.

    karlrupp committed Apr 10, 2018
    Resolves #260. Thanks to GitHub user @conradsnicta for reporting.
  2. matrix_slice: Fixed outer_prod bug.

    karlrupp committed Apr 10, 2018
    Problem was due to an overly aggressive clear() on the matrix_base.
    Resolves #258.
Commits on Mar 19, 2018
  1. Merge pull request #246 from cdeterman/master

    karlrupp committed Mar 19, 2018
    allow for SVD and qr_method to be run on CPUs with stringent work size requirements. This corrects for invalid work sizes
  2. Merge pull request #250 from robinchrist/master

    karlrupp committed Mar 19, 2018
    Update documentation in ilu_operations.hpp
Commits on Aug 26, 2017
  1. Merge pull request #242 from marty1885/master

    karlrupp committed Aug 26, 2017
    add platform select to dense_blas
Commits on Aug 24, 2017
  1. uBLAS: Fixed viennacl::copy() from ublas when matrix has empty rows.

    karlrupp committed Aug 24, 2017
    CSR-data structure (row array) for ublas::compressed_matrix is not fully set up if some rows remain empty.
    As a consequence, just copying over the row-array is not enough.
    This commit replaces the plain copy with a manual build of the row array.
Commits on Jul 28, 2017
  1. spGEMM: Fixed unused variable.

    karlrupp committed Jul 28, 2017
    As a positive side effect, this will improve performance a (little) bit.
Commits on Jun 14, 2017
  1. Merge pull request #238 from 5c0t-qi/Fix_doc

    karlrupp committed Jun 14, 2017
    Remove the random_matrix/vector description in types.dox
Commits on Jun 6, 2017
  1. Merge pull request #236 from 5c0t-qi/fix_pow_iter_example

    karlrupp committed Jun 6, 2017
    Fix::a minimal example of Power Iteration does not work
Commits on Apr 25, 2017
  1. compressed_matrix: Implemented copy-CTOR.

    karlrupp committed Apr 25, 2017
    The old behavior of the copy-CTOR was sharing data members rather than providing a clean copy.
    This commit now provides a copy-CTOR, with the same the behavior with viennacl::matrix<T> (i.e. a clean copy, no data sharing).
Commits on Mar 8, 2017
  1. Merge pull request #226 from dividiti/avoid-overwriting-cache

    karlrupp committed Mar 8, 2017
    Include build options into cache hash and avoid overwriting cache
Commits on Feb 28, 2017
  1. Merge pull request #225 from dividiti/compile-before-caching

    karlrupp committed Feb 28, 2017
    Compile all kernels before caching program binary
Commits on Feb 23, 2017
  1. ILUT: Added approximate triangular solves.

    karlrupp committed Feb 23, 2017
    This allows for GPU-accelerated application of ILUT.
    The triangular solves are replaced by a truncated Neumann series with relaxation.
    Enable via member function
     approximate_solves(iter)
    of the viennacl::linalg::ilut_tag to run 'iter' relaxations.
    Values in the range of 1-5 tend to give best results.
Commits on Jan 10, 2017
Commits on Dec 12, 2016
  1. compressed_matrix: Added copy() from host to device accepting CSR raw…

    karlrupp committed Dec 12, 2016
    … arrays.
    
    To better support the use case in #214.
Commits on Oct 25, 2016
  1. Matrix: Added support for element_OP(v, alpha) and element_OP(alpha, v)

    karlrupp committed Oct 25, 2016
    Operations: element_prod(), element_div(), element_pow().
    Only for float and double, no integer types supported (yet).
  2. Vector: Added support for element_OP(v, alpha) and element_OP(alpha, v)

    karlrupp committed Oct 25, 2016
    Operations: element_prod(), element_div(), element_pow().
    Only for float and double, no integer types supported (yet).
Commits on Oct 24, 2016
  1. Unary operations: Added math functions from C99.

    karlrupp committed Oct 24, 2016
    The following new functions are now available:
     element_acosh()
     element_asinh()
     element_atanh()
     element_erf()
     element_erfc()
     element_exp2()
     element_exp10()
     element_log2()
     element_round()
     element_rsqrt()
     element_trunc()
    
    Commit based on patch provided by Matthias Möller, authored by Andrzej Jaeschke <andrzej.jaeschke@gmail.com>
    The original patch is extended to also clean up the CUDA backend -> less code :-)
    C++11 guards are added to the tests.
    Code compiles and tests run without C++11, but more testing desired to make sure everything is in order.
Commits on Oct 21, 2016
Commits on Oct 17, 2016
  1. Merge pull request #206 from smarthi/sundry

    karlrupp committed Oct 17, 2016
    Fix the patch version for 1.7.1
Commits on Oct 5, 2016
Commits on Sep 22, 2016
  1. AMG: Fixed compilation errors and warnings for CUDA and OpenCL handles.

    karlrupp committed Sep 22, 2016
    Problem only shows up with CUDA and OpenCL enabled.
Commits on Sep 14, 2016
  1. Merge pull request #201 from intelfx/device-specific-forwards

    karlrupp committed Sep 14, 2016
    Device-specific: mark functions in headers as inline
Commits on Jun 14, 2016
  1. OpenCL: Fixed missing integer implementation of element_pow()

    karlrupp committed Jun 14, 2016
    Resolves #187. Cudos to cdeterman.
Commits on Apr 6, 2016
  1. Merge pull request #182 from ruby-accel/master

    karlrupp committed Apr 6, 2016
    Eigen: Only include Eigen/SparseCore (sufficient) instead of Eigen/Sparse
Commits on Mar 16, 2016
Commits on Mar 3, 2016
  1. Merge pull request #181 from gongzg/master

    karlrupp committed Mar 3, 2016
    OpenCL: Avoid incorrect flag when create ocl memory object.
Commits on Feb 16, 2016
  1. Reordering: Fixed handling of multiple components in GPS.

    karlrupp committed Feb 16, 2016
    Addresses #179 (i.e. works for simple test cases).
    Reporter should check whether this fixes the problems encountered.