Changelog for libavif-devel-1.1.1-65.3.x86_64.rpm :

* Tue Sep 10 2024 Carsten Ziepke - Enable building with aom and libyuv for Leap
* Sat Aug 31 2024 Dirk Müller - update to 1.1.1:
* In avif.h, change \"AVIF_API AVIF_NODISCARD\" back to \"AVIF_NODISCARD AVIF_API\" to fix clang-cl and MSVC compilation errors in the shared library build on Windows.
* Fix -DAVIF_GTEST=SYSTEM
* Fix infe_type and codec_config_type wrongly read as byte- aligned fields in the
* experimental feature AVIF_ENABLE_EXPERIMENTAL_METAV1.
* When building aom as a local dependency, runtime CPU detection (`CONFIG_RUNTIME_CPU_DETECT`) is now always `ON`;
* Fix CMake config shared library leaks
* Update gain map metadata to current ISO 21496-1 draft.
* cmake: Only search for ASM_NASM language on x86_64 platforms.
* Fix \"No known features for CXX compiler\" CMake error.
* Fix aom link flags so that transitive library link flags are included when aom is a static library
* Fix out-of-order \'dimg\' grid associations
* Report files with an item used in multiple \'dimg\' boxes with
* AVIF_RESULT_NOT_IMPLEMENTED instead of AVIF_RESULT_INVALID_IMAGE_GRID.
* Add experimental API for reading and writing gain maps in AVIF files.
* If enabled at compile time, add `gainMap` field to `avifImage`,
* add `qualityGainMap` field to `avifEncoder`, add `gainMapPresent`, `enableDecodingGainMap`, `enableParsingGainMapMetadata` and `ignoreColorAndAlpha` to `avifDecoder`.
* Utility functions for working with gain maps are also added.
* Gain maps allow readers that support them to display HDR images that look good on both HDR and SDR displays.
* Add experimental support for converting jpeg files with gain maps to AVIF files with gain maps. Requires libxml2, and the AVIF_ENABLE_EXPERIMENTAL_GAIN_MAP compilation flag.
* Add a --qgain-map flag to control the gain map quality in avifenc.
* Add the headerFormat member of new type avifHeaderFormat to avifEncoder.
* Add experimental API for reading and writing \"mif3\"-branded AVIF files behind the compilation flag AVIF_ENABLE_EXPERIMENTAL_METAV1.
* Implement avifImageScale() fallback when libyuv is not available.
* Partial import of libyuv to third_party/libyuv (new LICENSE).
* Add avifenc flag suffixes \":update\" and \":u\". Quality- relative, tiling-relative and codec-specific flags can now be positional, relative to input files.
* Add experimental support for layered AVIF encoding in avifenc.
* Use the --layered flag to enable layered AVIF encoding.
* Layered AVIF has multiple layers, which works like frame of animated AVIF, and layers can be rendered in progressive manner on supported viewers
* Only aom supports layered AVIF encoding at the time of writing.
* Add --scaling-mode flag to set scaling mode of each layer.
* This part of AV1 encoder is not as thoroughly tested, so there are higher possibility encoder may crash when given certain configuration or input.
* Add imageSequenceTrackPresent flag to the avifDecoder struct.
* avifImageScale() function was made part of the public ABI.
* Add avif_cxx.h as a C++ header with basic functionality.
* Add enum aliases AVIF_COLOR_PRIMARIES_SRGB, AVIF_COLOR_PRIMARIES_BT2100,
* AVIF_COLOR_PRIMARIES_DCI_P3, AVIF_TRANSFER_CHARACTERISTICS_PQ.
* Add avifResult enum entry AVIF_RESULT_INTERNAL_ERROR.
* Require libyuv by default (but it can still be disabled with
* -DAVIF_LIBYUV=OFF).
* Add avifdec --icc flag to override the output color profile.
* Add experimental API for reading and writing 16-bit AVIF files behind the
* compilation flag AVIF_ENABLE_EXPERIMENTAL_SAMPLE_TRANSFORM.
* Add AVIF_CHROMA_SAMPLE_POSITION_RESERVED to avifChromaSamplePosition enum.
* Sun Mar 17 2024 Dirk Müller - update to 1.0.4:
* AVIF_ENABLE_WERROR is set to OFF by default.
* Fix wrong alpha plane deallocation when decoded tile pixel format does not match reconstructed output image pixel format
* Fix identical chunk skipping optimization when writing animation data
* Fix ID selection for artificial grid alpha item when decoding a grid of tiles which each have an associated auxiliary alpha image item (https://crbug.com/oss-fuzz/65657).
* Thu Dec 21 2023 Andreas Stieger - update to 1.0.3:
* Rewrite the fix for memory errors fixed in 1.0.2
* CVE-2023-6704: Fix use-after-free errors (boo#1218303)
* src/reformat.c: Allocate the threadData array directly
* Tue Nov 28 2023 Andreas Stieger - update to 1.0.2:
* Update avifCropRectConvertCleanApertureBox() to the revised requirements in ISO/IEC 23000-22:2019/Amd. 2:2021 Section 7.3.6.7.
* CVE-2023-6350: Out of bounds memory to alphaItemIndices (boo#1217614)
* CVE-2023-6351: use-after-free in colorProperties (boo#1217615)- drop fix-gdkpixbuf.patch
* Tue Aug 29 2023 Paolo Stivanin - Update to 1.0.0:
* Incompatible changes: + The clli member was added to the avifImage struct. + The repetitionCount member was added to the avifEncoder and avifDecoder structs. + The quality and qualityAlpha members were added to the avifEncoder struct. + Check that functions returning pointers do not return NULL before accessing those pointers. + Check the return value of avifEncoderSetCodecSpecificOption(). + The maxThreads member was added to the avifRGBImage struct. + Check the return value of avifRGBImageAllocatePixels(), avifRWDataRealloc(), avifRWDataSet(), avifImageSetProfileICC(), avifImageSetMetadataExif() and avifImageSetMetadataXMP(). + The meaning of the keyframeInterval member of avifEncoder struct has changed slightly. When set to a value of \"n\", Before: It forces a keyframe on every nth frame. After: Any set of \"n\" consecutive frame will have at least one keyframe (every nth frame may or may not be a keyframe).
* Added: + Add clli metadata read and write support + Add repetitionCount member to avifEncoder and avifDecoder structs to specify the number of repetitions for animated image sequences. + Add quality and qualityAlpha to avifEncoder. Note: minQuantizer, maxQuantizer, minQuantizerAlpha, and maxQuantizerAlpha are deprecated. Code should be updated to set quality (and qualityAlpha if applicable) and leave minQuantizer, maxQuantizer, minQuantizerAlpha, and maxQuantizerAlpha initialized to the default values. + The --target-size flag in avifenc was added to adapt the quality so that the output file size is as close to the given number of bytes as possible. + Add the public API function avifImageIsOpaque() in avif.h. + Add the public API functions avifImagePlane(), avifImagePlaneRowBytes(), avifImagePlaneWidth(), and avifImagePlaneHeight() in avif.h. + Add API for multi-threaded YUV to RGB color conversion. + Allow lossless 4:0:0 on grayscale input. + Add avifenc --no-overwrite flag to avoid overwriting output file. + Add avifenc --clli flag to set clli. + Add support for all transfer functions when using libsharpyuv.
* Changed: + Exif and XMP metadata is exported to PNG and JPEG files by default, except XMP payloads larger than 65502 bytes in JPEG. + The --grid flag in avifenc can be used for images that are not evenly divided into cells. + Change the encoder to write the boxes within the \"stbl\" box in the order of stsd, stts, stsc, stsz, stco, stss. + avifImageRGBToYUV() and avifImageYUVToRGB() handle avifImage bit depths 8, 10, 12 and now also 16. Files read by apps/shared/ can output 16-bit avifImage instances. + avifImageCreate(), avifImageCreateEmpty(), avifEncoderCreate() and other internal functions now return NULL if a memory allocation failed. + avifEncoderSetCodecSpecificOption() now returns avifResult instead of void to report memory allocation failures.- Add fix-gdkpixbuf.patch.
* Sun May 07 2023 Arjen de Korte - Add BuildRequires pkgconfig(libwebp) to enable libsharpyuv
* Mon Nov 14 2022 ecsos - Update to version 0.11.1:
* Changed: - avifincrtest_helpers: Cast 64-bit offset to size_t - avifmetadatatest: don\'t include avif/internal.h - avifrgbtoyuvtest: skip if no libsharpyuv - Disable tests that may fail if the codec is not aom (#1176)
* Thu Oct 27 2022 Cristian Rodríguez - Remove unused BuildRequires on nasm- Remove indirect/incorrect Buildrequires on zlib- add direct glib Buildrequires
* Sat Oct 15 2022 Bjørn Lie - Update to version 0.11.0:
* There are incompatible ABI changes in this release. The alphaRange member was removed from the avifImage struct. The chromaDownsampling and avoidLibYUV members were added to the avifRGBImage struct. The imageDimensionLimit member was added to the avifDecoder struct. avifImageCopy() and avifImageAllocatePlanes() signatures changed. It is necessary to recompile your code. Also check the return values of avifImageCopy() and avifImageAllocatePlanes().
* Added: - Add man pages for avifenc and avifdec - Add the avifChannelIndex type alias for enum avifChannelIndex - Add avifChromaDownsampling enum - Add chromaDownsampling field to avifRGBImage struct - Add support for AVIF_RGB_FORMAT_RGB_565 - Add imageDimensionLimit field to avifDecoder struct - Add autoTiling field to avifEncoder struct - Add new avifResult codes AVIF_RESULT_CANNOT_CHANGE_SETTING and AVIF_RESULT_INCOMPATIBLE_IMAGE - Add new enum constants AVIF_PIXEL_FORMAT_COUNT and AVIF_RGB_FORMAT_COUNT - avifdec: Add --dimension-limit, which specifies the image dimension limit (width or height) that should be tolerated - avifenc: Add --sharpyuv, which enables \"sharp\" RGB to YUV420 conversion, which reduces artifacts caused by 420 chroma downsampling. Needs libsharpyuv (part of the libwebp repository) at compile time. - avifenc: Add --ignore-exif and --ignore-xmp flags. - avifenc: Add --autotiling, which sets --tilerowslog2 and - -tilecolslog2 automatically. - avifenc: Input Exif orientation is converted to irot/imir by default.
* Changed: - Fix memory leaks of metadata on avifenc exit - Update the handling of \'lsel\' and progressive decoding to AVIF spec v1.1.0 - Treat an absent lsel and layer_id == 0xFFFF equivalently for backward compatibility with earlier drafts of AVIF spec v1.1.0 - Set libavif\'s own default value of cfg.rc_end_usage for libaom - Set the libaom-specific option -a tune=ssim by default - Bump cmake_minimum_required from 3.5 to 3.13 - Fix https://crbug.com/oss-fuzz/48135 - Use several new libyuv functions in reformat_libyuv.c - Fix SVT-AV1\'s issue 1957 related to uninitialized variables crashing the encoder - Update aom.cmd: v3.5.0 - Update rav1e.cmd: v0.5.1 - Update svt.cmd/svt.sh: v1.2.1 - Update libgav1.cmd: v0.18.0 - Update libyuv.cmd: f9fda6e7 (version 1844) - avifImageCopy() and avifImageAllocatePlanes() now return avifResult instead of void to report invalid parameters or memory allocation failures. - avifImageRGBToYUV() now uses libyuv fast paths by default. It may slightly change conversion results. The old behavior can be restored by setting avifRGBImage::chromaDownsampling to AVIF_CHROMA_DOWNSAMPLING_BEST_QUALITY and avifRGBImage::avoidLibYUV to AVIF_TRUE. - avifRGBImage::chromaUpsampling now only applies to conversions that need upsampling chroma from 4:2:0 or 4:2:2 and has no impact on the use of libyuv. - Set avifRGBImage::avoidLibYUV accordingly to control the use of libyuv. - avifenc: Set the YUV format to 4:0:0 for grayscale PNGs - Support updating encoder settings and codec-specific options during encoding - Disable AVIF_STRICT_CLAP_VALID and AVIF_STRICT_PIXI_REQUIRED in the JNI wrapper - avifdec: Return proper exit code in \"info\" mode - In avifenc and avifdec, treat all arguments that start with \'-\' as options - Exif and XMP metadata is imported from PNG and JPEG files. - avifImageSetMetadataExif() parses the Exif metadata and converts any Exif orientation found into transformFlags, irot and imir values. - Write \'auxi\' box for animated images with alpha channel - Write \'auxv\' as handler_type for alpha channel track - Use PNG_COLOR_TYPE_GRAY for 8-bit grayscale output - Replace repeated subtraction by modulo in calcGCD - Change avifImageCreate to take uint32_t instead of int parameters - When writing an image sequence, check if it\'s safe to cast width and height to uint16_t - Allow clamped grid cells in avifEncoderAddImageGrid()
* Removed: - alphaRange field was removed from the avifImage struct. It it presumed that alpha plane is always full range. - The avifCodecConfigurationBox struct becomes a private type for libavif internal use- Bump lib_soversion global (also in baselibs.conf) to 15 following upstream change.- Use ldconfig_scriptlets for post(un) handling.
* Sat Jul 30 2022 Callum Farmer - Enable libyuv on TW
* Sun Apr 24 2022 Dirk Müller - update to 0.10.1:
* tests/docker/build.sh: Build SVT-AV1 using cmake and ninja directly
* Fix a Visual Studio 2017 compiler warning in src\\reformat.c: warning C4204: nonstandard extension used: non-constant aggregate initializer
* Fix the help message of avifdec: --index takes a value
* Fri Apr 08 2022 Andreas Schneider - Update to version 0.10.0
* See https://github.com/AOMediaCodec/libavif/blob/v0.10.0/CHANGELOG.md for a complete changelog or check the CHANGELOG.md in the doc directory.
* Sat Dec 25 2021 Andreas Schneider - Disable libaom and use rav1e by default
* As PHP 8.1 has AVIF support, use rav1e by default as it is more secure being written in Rust.
* Thu Oct 21 2021 Andreas Schneider - Update to version 0.9.3
* See https://github.com/AOMediaCodec/libavif/blob/v0.9.3/CHANGELOG.md for a complete changelog.
* Thu Jun 24 2021 Andreas Schneider - Update to version 0.9.2
* See https://github.com/AOMediaCodec/libavif/blob/v0.9.2/CHANGELOG.md for a complete changelog.
* Thu May 20 2021 Andreas Schneider - Update to version 0.9.1
* See https://github.com/AOMediaCodec/libavif/blob/v0.9.1/CHANGELOG.md for a complete changelog.
* Tue Apr 13 2021 Dominique Leuenberger - Add baselibs.conf: produce libavif10-32bit as a new dependency to libgd3-32bit (support avif in libgd).
* Mon Mar 15 2021 Andreas Schneider - Update to version 0.9.0
* See https://github.com/AOMediaCodec/libavif/blob/v0.9.0/CHANGELOG.md for a complete changelog.
* Wed Jan 13 2021 Bjørn Lie - Build the gdk-pixbuf module: Add pkgconfig(gdk-pixbuf-2.0) BuildRequires and pass DAVIF_BUILD_GDK_PIXBUF=ON to cmake. Package the resulting loader in a new sub package.- Use cmake_build macro.