| - Small fix for clDSPOp::Copy() overlapping case - Support for Intel C++ 8.x |
|
More...
|
over 20 years ago
|
| - Code cleanups - Started AMD x86-64 asm optimizations |
|
More...
|
over 20 years ago
|
| doxygen cfg file updates |
|
More...
|
over 20 years ago
|
| Added dsp/ReBufferT.hh |
|
More...
|
over 20 years ago
|
| - Fixed complex clDSPOp::Reverse() - clFlipBand now works - Removed obsoleted classes/interfaces (clFilter, clDecimator, clReBuffer2) - Removed unmaintained Win32 and QNX ports - Renamed clFilter2 -> clFilter - Renamed clReBuffer3 -> clReBufferT |
|
More...
|
over 20 years ago
|
| In-place clDSPOp::Reverse() corrected, now clFlipBand works correctly |
|
More...
|
over 20 years ago
|
| Added class for flipping odd frequency bands |
|
More...
|
over 20 years ago
|
| - Portability fixes - Less memory allocations - Support for Apple Mac OS-X and Linux on Alpha - Better support for Sun Solaris/SPARC |
|
More...
|
over 20 years ago
|
| - Makefile change to support building rpm packages - .spec file for rpm packages |
|
More...
|
over 20 years ago
|
| Added doxygen configuration for PDF |
|
More...
|
over 20 years ago
|
| - Support for FFTW3 - Generic code cleanups (reduced stack usage on g++) - Some bugfixes |
|
More...
|
over 20 years ago
|
| Small documentation update |
|
More...
|
over 20 years ago
|
| Full support for Intel Integrated Performance Primitives |
|
More...
|
over 20 years ago
|
| *** empty log message *** |
|
More...
|
over 20 years ago
|
| Started support for Intel's Integrated Performance Primitives |
|
More...
|
over 20 years ago
|
| - Makefile fix - New example code |
|
More...
|
almost 21 years ago
|
| - clDSPOp::Sort() and related for type 'long' - Re-enabled ISO standard rounding with nearbyint() - Support for Intel C++ math library |
|
More...
|
almost 21 years ago
|
| Makefile updates |
|
More...
|
almost 21 years ago
|
| Added clDSPOp::Resample() and clDSPOp::ResampleAvg() |
|
More...
|
almost 21 years ago
|
| Intel C++ compiler options tuned |
|
More...
|
about 21 years ago
|
| Fixes for Intel C++ P4 optimization flags |
|
More...
|
about 21 years ago
|
| Makefile updates |
|
More...
|
about 21 years ago
|
| Support for Solaris 8 (on UltraSPARC, 64-bit) |
|
More...
|
over 21 years ago
|
| Makefile updates for the new Intel C++ compiler version 7 |
|
More...
|
over 21 years ago
|
| - Version update for new development release - Prefetching asm is now always enabled |
|
More...
|
over 21 years ago
|
| Added multiply-add |
|
More...
|
over 21 years ago
|
| Some minor optimization |
|
More...
|
over 21 years ago
|
| Fixed a showstopper bug in E3DNow! copy, it was implemented when full register set usage was added. |
|
More...
|
over 21 years ago
|
| - Fixed a prefetch bug in dsp_x86_3dnow_copy*() - Improved dsp_x86_3dnow_copy*() performance by using all registers |
|
More...
|
over 21 years ago
|
| Doxygen config update |
|
More...
|
over 21 years ago
|
