Commits · ecb1088aa518ae3b22e74881ef42c6547676505d · Kaushik Kulkarni / viennacl-dev · GitLab

Dec 09, 2014
- Solver Bench: Added sliced_ell_matrix and added pipelined runs. · ecb1088a
  Karl Rupp authored Dec 09, 2014
  
  ecb1088a
- FFT: Fixed flag passed to CUDA kernel. · c09831bb
  Karl Rupp authored Dec 09, 2014
  c09831bb
- OpenMP: Fixed unspecified behavior for operations using reductions. · 71e46368
  Karl Rupp authored Dec 09, 2014
  71e46368
- OpenMP: Removed use of private and shared clauses. · 0a8e2999
  Karl Rupp authored Dec 09, 2014
  0a8e2999
- Iterative: Moved local array declaration out of CSR-adaptive kernel. · 40686dbf
  Karl Rupp authored Dec 09, 2014
  40686dbf
- Iterative: Fixed overloads for pipelined iterative solvers. · b825e5d2
  Karl Rupp authored Dec 09, 2014
  b825e5d2
Dec 06, 2014
- Direct solve bench: Removed accidental uBLAS dependency. · 8682205a
  Karl Rupp authored Dec 06, 2014
  
  8682205a
- CUDA: Fixing uses of 'uint' and performance warnings. · 4ebc5c60
  Karl Rupp authored Dec 06, 2014
  
  4ebc5c60
- CUDA: Fixed complaints about destructor in Visual Studio. · 0e2a3758
  Karl Rupp authored Dec 06, 2014
  0e2a3758
Dec 05, 2014
- vector_iterator: Fixed internal handling of smart-pointer. · 55b05254
  Karl Rupp authored Dec 05, 2014
  55b05254
Dec 04, 2014
- GMRES: Reverted to previous kernel in first pipelined stage for non-NVIDIA GPUs. · a4b7354a
  Karl Rupp authored Dec 04, 2014
  a4b7354a
- GMRES: Improved kernel first first stage of pipelined orthogonalization. · 4381e000
  Karl Rupp authored Dec 04, 2014
  4381e000
- compressed_matrix: Fixed missing context switch for CSR-adaptive metainfo. · 9d8bae24
  Karl Rupp authored Dec 04, 2014
  9d8bae24
- Pipelined solvers: Added better parameters for NVIDIA GPUs. · b6758fb9
  Karl Rupp authored Dec 04, 2014
  b6758fb9
- sliced_ell_matrix: Setting defaults for NVIDIA GPUs to 256. · acb1ca0c
  Karl Rupp authored Dec 04, 2014
  acb1ca0c
Nov 20, 2014
- Doxygen: Added symbolic link to changelog, did not work with 1.8.8. · 7a0f5794
  Karl Rupp authored Nov 20, 2014
  
  7a0f5794
- Doxygen: Now taking version number directly from CMakeLists.txt · f6856df3
  Karl Rupp authored Nov 20, 2014
  
  f6856df3
- CUDA: Added CSR-adaptive to pipelined iterative solvers. · e0d55f9e
  Karl Rupp authored Nov 20, 2014
  
  e0d55f9e
- OpenCL: Added CSR-adaptive for pipelined iterative solvers. · 8678f02f
  Karl Rupp authored Nov 20, 2014
  
  8678f02f
- CUDA: Cleanup of CSR-adaptive implemenentation, adjustment of block sizes. · bac9d4ab
  Karl Rupp authored Nov 20, 2014
  bac9d4ab
- Tests: Fixed incorrect test code in libviennacl-blas1. · 9eed5e20
  Karl Rupp authored Nov 20, 2014
  
  9eed5e20
- Changelog: Fixed incorrect co-author name of CSR-adaptive paper. · 61792c08
  Karl Rupp authored Nov 20, 2014
  
  61792c08
- Changelog: Added notes for 1.6.1 release. · bdd0ddd9
  Karl Rupp authored Nov 20, 2014
  
  bdd0ddd9
- Doxygen: Fixed warnings. · 078328ff
  Karl Rupp authored Nov 20, 2014
  
  078328ff
- Updated version to 1.6.1. · 5bf0373b
  Karl Rupp authored Nov 20, 2014
  
  5bf0373b
- Visual Studio 2012: Fixed performance warnings and a test compilation error. · 1f51ee98
  Karl Rupp authored Nov 20, 2014
  1f51ee98
Nov 19, 2014
- Direct solve: Fixed errors obtained after resolution of self-assignment problems. · df29d5f3
  Karl Rupp authored Nov 19, 2014
  
  df29d5f3
- CUDA: Fixed compilation error in triangular solve kernels. · f321e151
  Karl Rupp authored Nov 19, 2014
  f321e151
- compressed_matrix: Implemented CSR-adaptive in CUDA and OpenCL. · 7d212433
  Karl Rupp authored Nov 19, 2014
  7d212433
- Self-assignment: Added test and fixed all bugs found. · 403b7c87
  Karl Rupp authored Nov 19, 2014
  403b7c87
- (sliced_)ell_matrix, hyb_matrix: Added overload for STL-emulated sparse matrix. · e98a3685
  Karl Rupp authored Nov 19, 2014
  e98a3685
Nov 17, 2014
- inplace_solve, dense: Simplified code and improved performance. · 0ba719f3
  Karl Rupp authored Nov 17, 2014
  0ba719f3
Nov 16, 2014
- Merge pull request #110 from d-meiser/disable-coveralls · 8c848791
  Karl Rupp authored Nov 16, 2014
  8c848791
- Cleanup: Removed unused source files. · 45058f42
  Karl Rupp authored Nov 16, 2014
  45058f42
- Disable coveralls. · 49a6a336
  Dominic Meiser authored Nov 16, 2014
  49a6a336
- compressed_matrix: Improved documentation. · 91ca8382
  Karl Rupp authored Nov 16, 2014
  91ca8382
- Tests: Disabled non-symmetric eigenvalue routines. · a249b804
  Karl Rupp authored Nov 16, 2014
  a249b804
- compressed_matrix: Fixed invalid memory access for triangular solves. · 6087a180
  Karl Rupp authored Nov 16, 2014
  6087a180
- Matrix: Fixed kernels for transposition in OpenCL and CUDA. · 1dea3deb
  Karl Rupp authored Nov 16, 2014
  1dea3deb
Nov 15, 2014
- OpenCL: Another attempt to work around the problems with older SDKs on CPUs. · 4a68e0b2
  Karl Rupp authored Nov 15, 2014
  4a68e0b2