Commits · d156a78c64b52113cc97124c5dcb9705c2fe4829 · Kaushik Kulkarni / viennacl-dev · GitLab

Jan 19, 2016
- OpenCL: Fixed work size restriction in sum_inner_prod kernel. · d156a78c
  Karl Rupp authored Jan 19, 2016
  d156a78c
Jan 18, 2016
- Matrix: Added arbitrary expressions for trans() · e73e96d1
  Karl Rupp authored Jan 18, 2016
  e73e96d1
- Sparse matrices: No more temporaries for y += Ax; and y -= Ax; · dc1c0413
  Karl Rupp authored Jan 18, 2016
  dc1c0413
- OpenCL: max and min on vectors now also work with work group size 1. · 26e81488
  Karl Rupp authored Jan 18, 2016
  26e81488
- OpenCL: Removed use of abs() on unsigned integers. · fa046011
  Karl Rupp authored Jan 18, 2016
  fa046011
- OpenCL: Compatibility of summation kernel for work group size 1. · 745e3400
  Karl Rupp authored Jan 18, 2016
  745e3400
Jan 17, 2016
- OpenCL: Improved compatibility of triangular solves for work size 1. · b6cf6eb2
  Karl Rupp authored Jan 17, 2016
  b6cf6eb2
- OpenCL: Clarified error message for 'unknown_error'. · f6754ff3
  Karl Rupp authored Jan 17, 2016
  f6754ff3
- Travis-CI: Removed no longer needed call to viennacl-info · 37c5f507
  Karl Rupp authored Jan 17, 2016
  37c5f507
Jan 16, 2016
- Travis-CI: Removed OpenCL from tests. · 5b8b6011
  Karl Rupp authored Jan 16, 2016
  5b8b6011
- Travis-CI: Attempt to fix now missing fglrx package. · c015d284
  Karl Rupp authored Jan 16, 2016
  c015d284
- CMake: Now setting MACOSX_RPATH to avoid dev-warning. · 7e57a5ea
  Karl Rupp authored Jan 16, 2016
  7e57a5ea
- Device Database: Improved device detection algorithm. · 107c1da8
  Karl Rupp authored Jan 16, 2016
  107c1da8
Jan 15, 2016
- Merge branch 'karlrupp/feature-reduce-generator-usage' · b9e075c6
  Karl Rupp authored Jan 15, 2016
  b9e075c6
- Generator: Major cleanup, only matrix-products remain. · 993889c4
  Karl Rupp authored Jan 15, 2016
  993889c4
- Generator: Removed unused code. · eb2d0d57
  Karl Rupp authored Jan 15, 2016
  eb2d0d57
- OpenCL: Fixed problems observed with Intel OpenCL Runtime 14.2 · d9bed704
  Karl Rupp authored Jan 15, 2016
  d9bed704
Jan 11, 2016
- Typos: Fixed two typos in docs and one in tests. · cf6ebc5d
  Patrick Sanan authored Jan 11, 2016
  
  cf6ebc5d
Oct 20, 2015
- GCC: Fixed unused variable and conversion warnings. · afe92df2
  Karl Rupp authored Oct 20, 2015
  
  afe92df2
Oct 19, 2015
- Merge pull request #164 from Franz-S/request · a72ee2b4
  Karl Rupp authored Oct 19, 2015
  a72ee2b4
Sep 29, 2015
- CG: Fixed incorrect check for absolute tolerance. · 3de1eb05
  Karl Rupp authored Sep 29, 2015
  3de1eb05
Sep 22, 2015
- OpenCL: Removed float condition for ternary operator in kernels. · d46bcd95
  Karl Rupp authored Sep 22, 2015
  d46bcd95
Sep 15, 2015
- trans(): changed order of instructions for better clarity · 62519178
  Franz-S authored Sep 15, 2015
  
  62519178
- trans(): Fixed false calculation of the boundary for the loop · e4fc1599
  Franz-S authored Sep 15, 2015
  
  e4fc1599
- trans(): changed the matrix names to common data_A and data_B · fabc16a7
  Franz-S authored Sep 14, 2015
  
  fabc16a7
- copy_vec(): Fixed a bug and formatting the code · 6a145798
  Franz-S authored Sep 14, 2015
  6a145798
- OpenMp: Introduction on minimal matrix size for matrix operations. · 54c4f5fd
  Franz-S authored Sep 14, 2015
  
  54c4f5fd
- OpenMp: Parallelized scaled_rank_1_update() · 93135ff3
  Franz-S authored Sep 08, 2015
  93135ff3
- OpenMp: Parallelized prod_impl() · fa445fbd
  Franz-S authored Sep 10, 2015
  fa445fbd
- OpenMp: improved performance on the trans() function · 324d3090
  Franz-S authored Sep 09, 2015
  324d3090
- sum_impl(): Fixed datatype in for-loop, · 49fc6755
  Franz-S authored Sep 08, 2015
  49fc6755
- OpenMP:Parallelized index_norm_inf() · 886b2267
  Franz-S authored Sep 02, 2015
  886b2267
- OpenMP:Parallelized sum_impl() · e5776334
  Franz-S authored Aug 30, 2015
  e5776334
- OpenMP:Parallelized max_impl() · d08567ed
  Franz-S authored Aug 21, 2015
  d08567ed
- OpenMP:Parallelized min_impl() · b3df531e
  Franz-S authored Aug 21, 2015
  b3df531e
Sep 14, 2015
- CSR adaptive, OpenCL: Fixed corner case where sparse matrix has empty rows. · 43076cc2
  Karl Rupp authored Sep 14, 2015
  43076cc2
- compressed_matrix: Proper initialization of row_buffer. · e7950174
  Karl Rupp authored Sep 14, 2015
  e7950174
Sep 08, 2015
- Tests: Fixed incorrect reference calculation for transposed operations. · 6a331e03
  Karl Rupp authored Sep 08, 2015
  
  6a331e03
Sep 03, 2015
- Generator: Changed index type from uint to size_t for GEMM. · fd106a68
  Karl Rupp authored Sep 03, 2015
  fd106a68
Sep 02, 2015
- Device DB: Fixed architecture entry for Tesla K20m. · b631a783
  Karl Rupp authored Sep 02, 2015
  b631a783