Commits · d33620ea8448a34994257e35afc85e04f9497b23 · Kaushik Kulkarni / viennacl-dev · GitLab

Apr 29, 2013
- Added missing specification of raw_size() when using shallow_copy(). Added documentation. · d33620ea
  Karl Rupp authored Apr 28, 2013
  
  d33620ea
Apr 28, 2013
- Fixed GMRES: Reported wrong estimated residual, hence terminated prematurely (or never). · c7f962be
  Karl Rupp authored Apr 28, 2013
  
  c7f962be
- Now using const_entry_proxy for const-access for vectors and matrices. · 42088aec
  Karl Rupp authored Apr 28, 2013
  42088aec
- Fixed unused parameter warnings in optimized builds. · b027792d
  Karl Rupp authored Apr 28, 2013
  
  b027792d
Apr 27, 2013
- Prepared manual and changelogs for release. · 2f249749
  Karl Rupp authored Apr 27, 2013
  
  2f249749
- Removed warnings of type -Wextra when using GCC 4.6 · a250619a
  Karl Rupp authored Apr 27, 2013
  
  a250619a
- Removed use of enable_if for matrix operations. Now everything compiles... · 9bc56932
  Karl Rupp authored Apr 27, 2013
  9bc56932
- Simplified function interfaces for vector operations. · 4dd25d38
  Karl Rupp authored Apr 27, 2013
  4dd25d38
- Cleanup: Removed unused predicates and unnecessary overloads. · d1cf1570
  Karl Rupp authored Apr 27, 2013
  
  d1cf1570
- Updated version number to 1.4.2. · 4bda481f
  Karl Rupp authored Apr 27, 2013
  
  4bda481f
Apr 26, 2013
- Everything compiles cleanly again, tests on Linux pass. · edad8198
  Karl Rupp authored Apr 26, 2013
  edad8198
- Everything compiles again. Verification of correct computations still required. · 5e0de7d1
  Karl Rupp authored Apr 25, 2013
  
  5e0de7d1
Apr 25, 2013
- WiP: Introduced matrix_base. blas2-tutorial working again. · 96cafaa6
  Karl Rupp authored Apr 25, 2013
  96cafaa6
Apr 15, 2013
- WiP: Further integrated vector_base<>. vector-test-{cpu,cuda,opencl} passes again. · 4252b47f
  Karl Rupp authored Apr 15, 2013
  4252b47f
Apr 14, 2013
- [WiP] Introduced vector_base. · f82c23c6
  Karl Rupp authored Apr 14, 2013
  f82c23c6
- Fixed issue 3608388 on sourceforge-bugtracker (copying matrix if last entry not set) · bb5871e0
  Karl Rupp authored Apr 14, 2013
  
  bb5871e0
- Improved CPU-based vector loops within the custom kernel generator · 6746adfc
  Karl Rupp authored Mar 31, 2013
  
  6746adfc
- Added improved CSR SpMV kernel for CPUs · 1f125ecc
  Karl Rupp authored Mar 30, 2013
  
  1f125ecc
- Reverted 128x128 work layout for CSR SpMV. Instead, increased search range of... · 1ceb58f2
  Karl Rupp authored Mar 29, 2013
  1ceb58f2
- Reverted to old CPU work size deduction, which is better for simple vector kernels. · 9f683d58
  Karl Rupp authored Mar 28, 2013
  9f683d58
- OpenCL work size now defaults to 128x128 on CPUs. · f9d9ddb8
  Karl Rupp authored Mar 26, 2013
  f9d9ddb8
Feb 22, 2013
- Fixed missing hint on ENABLE_PEDANTIC_FLAGS in README.md · 3f705b04
  Karl Rupp authored Feb 22, 2013
  
  3f705b04
- Added README for developer repository on github. · 7b85e311
  Karl Rupp authored Feb 22, 2013
  
  7b85e311
- Updated changelogs for 1.4.1. Added missing kernel generation file for AMD GEMM kernels. · 7dd9736a
  Karl Rupp authored Feb 21, 2013
  
  7dd9736a
- Improved detection of the Intel OpenCL SDK. · 02d72620
  Karl Rupp authored Feb 21, 2013
  
  02d72620
Feb 21, 2013
- Added (missing?) newline at end of a test file. · 87f0d7e0
  Karl Rupp authored Feb 21, 2013
  
  87f0d7e0
- Added missing preprocessor switch for old Boost.Filesystem in converter.cpp · ec4e0e93
  Karl Rupp authored Feb 21, 2013
  
  ec4e0e93
- Removed erroneous build/auxiliary/converter.cpp. · 50103ae9
  Karl Rupp authored Feb 21, 2013
  
  50103ae9
- Slightly improved Doxygen-comments. Updated manual for release. · 433a09aa
  Karl Rupp authored Feb 21, 2013
  
  433a09aa
- Benchmarks now report performance by counting additions and multiplications separately. · 439d739c
  Karl Rupp authored Feb 21, 2013
  
  439d739c
- Added support for old Boost.Filesystem (prior to 1.44) in converter script. · 3dbfeb9e
  Karl Rupp authored Feb 21, 2013
  
  3dbfeb9e
- Fixed a bug in the fast GEMM kernel for NVIDIA hardware when using strides != 1 · d34c00ba
  Karl Rupp authored Feb 21, 2013
  
  d34c00ba
- * Updated AMD GEMM kernels to include scalars alpha and beta. · f975605b
  Karl Rupp authored Feb 21, 2013
  f975605b
- Fixed remaining problems with inclusion of Philippe's GEMM kernels. Tests pass. :-) · b5d46714
  Karl Rupp authored Feb 20, 2013
  
  b5d46714
Feb 20, 2013
- Added missing kernel generator script for AMD kernels. · 0745369e
  Karl Rupp authored Feb 20, 2013
  
  0745369e
- Added fast GEMM kernels for AMD Tahiti based on input from Philippe's... · 7979fb41
  Karl Rupp authored Feb 20, 2013
  7979fb41
- Updated copyright information to "2010-2013" · a2866d1c
  Karl Rupp authored Feb 20, 2013
  
  a2866d1c
- Fixed a warning (/* in comment) in CL/cl_gl_ext.h (staying with OpenCL 1.1.... · 25856942
  Karl Rupp authored Feb 20, 2013
  25856942
- Replaced all uses of size_t by std::size_t (exception: viennacl/generator/*,... · 249ee179
  Karl Rupp authored Feb 19, 2013
  249ee179
- Ensured defined behavior when resizing an empty compressed_matrix. · 10a5f086
  Karl Rupp authored Feb 19, 2013
  
  10a5f086