Files
rocm-systems/CHANGELOG.md
T
Aryan Salmanpour 08003e0fd7 Add an environment variable to control the VCN hardware color space conversion enablement (#167)
* Add an environment variable to control the VCN hardware color space conversion enablement

* Update version

* optimize the mem_alignment logic for hip memory allocation
2025-07-24 22:05:31 -04:00

2.0 KiB
Исходник Ответственный История

rocJPEG changelog

Documentation for rocJPEG is available at https://rocm.docs.amd.com/projects/rocJPEG/en/latest/

rocjpeg 1.1.0 for ROCm 7.0.0

Added

  • cmake config files
  • CTEST - New tests were introduced for JPEG batch decoding using various output formats, such as yuv_planar, y, rgb, and rgb_planar, both with and without region-of-interest (ROI).

Changed

  • Readme - cleanup and updates to pre-reqs
  • The decode_params argument of the rocJpegDecodeBatched API is now an array of RocJpegDecodeParams structs representing the decode parameters for the batch of JPEG images.
  • libdrm_amdgpu is now explicitly linked with rocjpeg.

Removed

  • Dev Package - No longer installs pkg-config

Resolved issues

  • Fixed a bug that prevented copying the decoded image into the output buffer when the output buffer is larger than the input image.
  • Resolved an issue with resizing the internal memory pool by utilizing the explicit constructor of the vector's type during the resizing process.
  • Addressed and resolved CMake configuration warnings.

rocJPEG 0.8.0 for ROCm 6.4

Changed

  • AMD Clang++ is now the default CXX compiler.
  • The jpegDecodeMultiThreads sample has been renamed to jpegDecodePerf, and batch decoding has been added to this sample instead of single image decoding for improved performance.

rocJPEG 0.6.0 for ROCm 6.3.0

Changes

  • Supported initial enablement of the rocJPEG library
  • Supported JPEG chroma subsampling:
    • YUV 4:4:4
    • YUV 4:4:0
    • YUV 4:2:2
    • YUV 4:2:0
    • YUV 4:0:0
  • Supported various output image format:
    • Native (i.e., native unchanged output from VCN Hardware, it can be either packed YUV or planar YUV)
    • YUV planar
    • Y only
    • RGB
    • RGB planar
  • Supported single-image and batch-image decoding
  • Supported Different partition modes on MI300 series
  • Supported region-of-interest (ROI) decoding
  • Supported color space conversion from YUV to RGB