Releases: ROCm/rpp
Releases · ROCm/rpp
rpp 2.1.0 for ROCm 7.1.1
RPP code for ROCm 7.1.1 did not change. The library was rebuilt for the updated ROCm 7.1.1 stack.
RPP 2.1.0 for ROCm 7.1.0
Added
- Solarize augmentation for HOST and HIP
- Hue and Saturation adjustment augmentations for HOST and HIP
- Find RPP - cmake module
- Posterize augmentation for HOST and HIP
Changed
- HALF - Fix half.hpp path updates
- Box filter - padding updates
Removed
- Packaging - Remove Meta Package dependency for HIP
- SLES 15 SP6 support
Resolved issues
- Test Suite - Fixes for accuracy
- HIP Backend - Check return status warning fixes
- Bugfix - HIP vector types init
rpp 2.0.0 for ROCm 7.0.2
RPP code for ROCm 7.0.2 did not change. The library was rebuilt for the updated ROCm 7.0.2 stack.
rpp 2.0.0 for ROCm 7.0.1
RPP code for ROCm 7.0.1 did not change. The library was rebuilt for the updated ROCm 7.0.1 stack.
RPP 2.0.0 for ROCm 7.0.0
Added
- Bitwise NOT, Bitwise AND, Bitwise OR augmentations on HOST (CPU) and HIP backends. (#520)
- Tensor Concat augmentation on HOST (CPU) and HIP backends. (#530)
- JPEG Compression Distortion augmentation on HIP backend. (#538)
log1p, defined aslog (1 + x), tensor augmentation support on HOST (CPU) and HIP backends.- JPEG Compression Distortion augmentation on HOST (CPU) backend. (#531)
Changed
- All handle creation and destruction APIs have been consolidated to
rppCreate(), for handle initialization, andrppDestroy(), for handle destruction (#513) - RPP function category "logical_operations" more appropriately renamed to "bitwise_operations". (#520)
- TurboJPEG package installation enabled for RPP Test Suite with
sudo apt-get install libturbojpeg0-dev. Instructions updated in utilities/test_suite/README.md. (#518) - Changed API of swap_channels augmentation to be called channel_permute, which now accepts one new argument, "permutationTensor" (pointer to a unsigned int tensor) that provides the permutation order to swap the RGB channels of each input image in the batch in any order. (#547)
- Old API -
RppStatus rppt_swap_channels_host(RppPtr_t srcPtr, RpptDescPtr srcDescPtr, RppPtr_t dstPtr, RpptDescPtr dstDescPtr, rppHandle_t rppHandle); - New API -
RppStatus rppt_channel_permute_host(RppPtr_t srcPtr, RpptDescPtr srcDescPtr, RppPtr_t dstPtr, RpptDescPtr dstDescPtr, Rpp32u *permutationTensor , rppHandle_t rppHandle);
- Old API -
Removed
- Older versions of RPP handle creation inlcuding
rppCreateWithBatchSize(),rppCreateWithStream(), andrppCreateWithStreamAndBatchSize()are now removed and replaced withrppCreate(). - Older versions of RPP handle destruction API including
rppDestroyGPU()andrppDestroyHost()are now removed and replaced withrppDestroy().
Resolved issues
- Test package - debian packages will install required dependencies
rpp 1.9.10 for ROCm 6.4.4
RPP code for ROCm 6.4.4 did not change. The library was rebuilt for the updated ROCm 6.4.4 stack.
rpp 1.9.10 for ROCm 6.4.3
RPP code for ROCm 6.4.3 did not change. The library was rebuilt for the updated ROCm 6.4.3 stack.
rpp 1.9.10 for ROCm 6.4.2
RPP code for ROCm 6.4.2 did not change. The library was rebuilt for the updated ROCm 6.4.2 stack.
rpp 1.9.10 for ROCm 6.4.1
RPP code for ROCm 6.4.1 did not change. The library was rebuilt for the updated ROCm 6.4.1 stack.
RPP 1.9.10 for ROCm 6.4.0
Added
- RPP Tensor Gaussian Filter support on HOST (CPU) backend. (#478)
- RPP Fog augmentation on HOST (CPU) and HIP backends. (#446)
- RPP Rain augmentation on HOST(CPU) and HIP backends. (#463)
- RPP Warp Perspective on HOST (CPU) and HIP backends. (#451)
- RPP Tensor Bitwise-XOR support on HOST (CPU) and HIP backends. (#464)
- RPP Threshold on HOST (CPU) and HIP backends. (#456)
- RPP Tensor Box Filter support on HOST (CPU) backend.(#425)
- RPP Audio Support for Spectrogram on HIP backend. (#433)
- RPP Audio Support for Mel Filter Bank on HIP backend. (#421)
Changed
- AMD Clang is now the default CXX and C compiler
- AMD RPP can now pass HOST (CPU) build with g++ (#517)
- Test Suite case numbers have been replaced with ENUMs for all augmentations to enhance test suite readability (#499)
- Test suite updated to return error codes from RPP API and display them (#483)
- Internal to RPP working - Restructure half.hpp and hip_fp16.h includes in one common header (#459)
Resolved issues
- CXX Compiler: Fixed HOST (CPU) g++ issues. (#517)
- Deprecation warning fixed for the "'sprintf' is deprecated" warning. (#512)
- Test suite build fix - RPP Test Suite Pre-requisite instructions updated to lock to a specific 'nifti_clib' commit as stated in ReadME - https://github.com/ROCm/rpp/tree/develop/utilities/test_suite#prerequisites (#506)
- Fixed broken image links for pixelate and jitter (#461)
- Internal to RPP working - Bugfix for Log Tensor in stride updation in log_recursive function (#479)