pool/libmfx-gen
SHA256
19
0
Fork 1
Code Issues Pull Requests Activity
Files
slfo-main
libmfx-gen/libmfx-gen.changes
Stefan Dirsch 09f03780d2 Update to 26.1.0 (jsc#PED-14029)
2026-01-20 00:13:21 +01:00

260 lines
11 KiB
Plaintext
Raw Permalink Blame History

This file contains ambiguous Unicode characters
This file contains Unicode characters that might be confused with other characters. If you think that this is intentional, you can safely ignore this warning. Use the Escape button to reveal them.
-------------------------------------------------------------------
Sun Jan 18 18:37:10 UTC 2026 - Stefan Dirsch <sndirsch@suse.com>
- Update to 26.1.0; changes since 25.3.4
* Update descriptions of dispatcher API query funcs
* VPL Encode Log Enhancement using AI.
* Changed LA_STATS_SIZE from calculated value (32 * 2) to
hardcoded 480. This removes the comment about enlarging size
for VdencHevcLaStats future usage and replaces the expression
with its evaluated result. The change maintains the same buffer
size (480 bytes) but makes the constant more explicit and
removes dependency on the previous calculation rationale.
* Add new member and enum for dec csc and scale caps
-------------------------------------------------------------------
Mon Nov 10 20:01:34 UTC 2025 - Stefan Dirsch <sndirsch@suse.com>
- Update to 25.3.4; changes since 25.2.4
* Upstream PTL encoding support
* Added VVC Decoding level 6.3 and profile
MFX_PROFILE_VVC_MAIN10_STILL_PICTURE
* Fixed XYUV color format handling. Export more capability
information in mfxImplDescription for HEVC/AVC/AV1 encoder
* Use VPP for YUV422H to RGB CSC if VD-SFC not available
-------------------------------------------------------------------
Tue Jun 10 15:25:23 UTC 2025 - Stefan Dirsch <sndirsch@suse.com>
- Update to version 25.2.4; changes since 25.1.4
* [Encode] Support negative QP for AVC
* Update Shift description for Decoder
* [Media Common] Bump up version to 25.2.0
* [RT Common] Upstream new BMG DIDs
* Define MFX_VARIANT_TYPE_QUERY as bitmask
* Add VVC_DEC STILL_PICTURE profile and Level 6.3
* [Decode] Fix VVC conformance window setting issue
* [Decode] Add latest VVC level 6.3 and MAIN10_STILL
* [RT Common] Add BMG Device ID for upstream
* Revert "[Decode] Add latest VVC level 6.3 and MAIN10_STILL "
* [RT Common] Update property-based query implementation
* Remove double copy to/from GPU in hwupload and hwdownload
* Refine for commit 618b2ed (#7828)
* Add new capability query information to VPL
* [Encode] Rename the feature block for close
* [RT Common] Remove duplicated codec IDs or VPP filters for query
* Revert "Remove double copy to/from GPU in hwupload and hwdownload (#9488)"
* [RT Common] Export MFXQueryImplsProperties as global symbol for Linux
* [Media Common] Bump up version to 25.2.1
* [Decode] Add latest VVC level 6.3 and Main10_STILL
* [VP] Add SR 444 420 Format Support
* [Decode] Reduce hevc status array size
* [Encode] Update Caps for AVC
* Fix XYUV color format handling
* remove fastpass for ip scan
* [RT Common] Mutex optimization for fast copy
* Updates from editor
* Update action runner to 22.04
* Update API version to 2.15
* [HEVCe] Export more information in QueryImplsDescription
* Remove clang12 from the github workflow
* [Media Common] Bump up version to 25.2.2
* [RT Common] Refine revenue build
* [RT Common] Refine BuildOptionsPlatform.cmake
* Fix bug when renderID not start from 128 will return error
* [Decode] Add R16 format into shift option
* [Media Common] Bump up version to 25.2.3
* Add AI assisted Encoder Control API
* Update FOURCC RGB4/BGR4/ARGB16/ABGR16 description
* [RT Common] Manage platform string in set for different build plan
* Add Panther Lake platform support in readme
* [RT Common] Minor style updates
* [Encode] 4k Low QP corruption issue fix.
* [Media Common] Bump up version to 25.2.4
* [encode] Upstream av1ehw/hevcehw for PTL
* [Encode] Enable HEVCe Saliency Control
* [RT Common] Upstream more device IDs for PTL
* [Common] Reclassify status message
-------------------------------------------------------------------
Mon May 12 13:43:15 UTC 2025 - Bjørn Lie <bjorn.lie@gmail.com>
- Update to version 25.1.4
* VPL GPU Runtime version 25.1.4 is the final release to include
the MediaSDK package within the MediaStack.tar.gz file.
Starting with version 25.2, the MediaSDK package will no longer
be available. If you require the MediaSDK package, you can
still download it from earlier releases.
* What’s New
- Enhanced AV1 decoding performance when all decode frame
surface is busy.
- Fixed AV1/AVC/HEVC decoding crash issue for surface creasion
when resolution change.
- Enabled property-based query.
-------------------------------------------------------------------
Mon May 12 06:12:49 UTC 2025 - Bernhard Wiedemann <bwiedemann@suse.com>
- Stop storing build machine kernel version (boo#1101107)
-------------------------------------------------------------------
Tue Feb 11 10:44:04 UTC 2025 - Stefan Dirsch <sndirsch@suse.com>
- renamed tarball to make buildservice happy
-------------------------------------------------------------------
Mon Feb 10 15:16:54 UTC 2025 - Matthias Brugger <mbrugger@suse.com>
- Update spec URL
-------------------------------------------------------------------
Mon Feb 10 12:38:39 UTC 2025 - Stefan Dirsch <sndirsch@suse.com>
- Update to version 25.1.1
* bugfixes and support for new hardware
-------------------------------------------------------------------
Thu Sep 12 12:05:49 UTC 2024 - Stefan Dirsch <sndirsch@suse.com>
- Update to version 24.3.3 - changes since 24.1.5 (related: jira#PED-10348)
* Added VVC decode API
* Enhanced DPB management for AV1 decoding
* Supported dynamic resolution change for HEVC/AVC/AV1 decoding in
VPL memory 2.0
* Restored AV1 filmgrain decoding
* Enabled API trace for remaining API methods (Reset, GetVideoParam,
GetEncodeStat)
* Fixed bitstream corruption when adding HDR header for AV1 encoding
with ffmpeg
* Refined reference frame structure for video conference to improve
quality
* Enabled buffer caching's control for GPU accelerated copying when
CM copy is used
* bugfixes
-------------------------------------------------------------------
Sat Jun 29 20:55:07 UTC 2024 - Bjørn Lie <bjorn.lie@gmail.com>
- Update to version 24.1.5
* Fixed stuck issue for VP9 decoding
* Fixed MPEG2 and JPEG decoding crash issue
* Refined AV1 encoding reference frame structure for low delay
frames to improve encoding quality
* Enabled AV1 encoding additional forward reference for P/GPB
frames on TU1
* Fixed reset API returns buffer not enough in AV1 CQP mode
* Added IDR flag to VP9 intra frames to support bitstream packing
- Use full url for source tarball.
- Use ldconfig_scriptlets macro for post(un) handling.
-------------------------------------------------------------------
Sat Feb 3 12:18:23 UTC 2024 - Stefan Dirsch <sndirsch@suse.com>
- updated hardware Supplements
-------------------------------------------------------------------
Sat Feb 3 11:47:40 UTC 2024 - Stefan Dirsch <sndirsch@suse.com>
- Update to version 24.1.3
* bugfix release for 24.1
-------------------------------------------------------------------
Mon Nov 7 12:49:19 UTC 2022 - Stefan Dirsch <sndirsch@suse.com>
- needed for jira#PED-1174 (Video decoding/encoding support
(VA-API, ...) for Intel GPUs is outside of Mesa)
-------------------------------------------------------------------
Thu Nov 3 18:27:22 UTC 2022 - Stefan Dirsch <sndirsch@suse.com>
- Update to version 22.6.1
* latest bugfix release
- supersedes u_libva-2.16.0.patch
-------------------------------------------------------------------
Fri Sep 30 13:42:35 UTC 2022 - Stefan Dirsch <sndirsch@suse.com>
- u_libva-2.16.0.patch
* fixes build against libva 2.16.0
* culprit: https://github.com/intel/libva/commit/8682f9e30f2fabf2ccc6f7609db035ed1af44703
-------------------------------------------------------------------
Mon Jul 11 09:15:33 UTC 2022 - Stefan Dirsch <sndirsch@suse.com>
- Update to version 22.4.4
* Decode:
+ Fix VC1 Decode assertion failure when frame type is NONE_PICTURE
+ Fix AV1 Decode hang issue in error clips that OBU info size is
larger than bitstream size.
* part of Intel oneVPL GPU Runtime 2022Q2 Release 22.4.4
- supersedes U_Removed-proprietary-mentions-from-file-headers-1023.patch
- filelist: added {_libdir}/libmfx-gen/enctools.so
-------------------------------------------------------------------
Tue Oct 12 08:15:05 UTC 2021 - Stefan Dirsch <sndirsch@suse.com>
- U_Removed-proprietary-mentions-from-file-headers-1023.patch
* upstream patch, which replaces proprietary Intel license with
OpenSource MIT license (github issue#183)
- this addresses the legal issues we had with checkin for sle15-sp4/
Leap 15.4
-------------------------------------------------------------------
Mon Oct 4 15:13:49 UTC 2021 - Stefan Dirsch <sndirsch@suse.com>
- fixed JIRA number in the initial changelog
-------------------------------------------------------------------
Wed Sep 29 09:07:33 UTC 2021 - Stefan Dirsch <sndirsch@suse.com>
- Update to version 21.3.4
* [vp9d] fix unsigned qindex issue for vp9 segmentation case (#569)
* [EncTools] Add closed GOP support for LPLA (#585)
* [Jpegd] Modify the resolution judgment condition of the current jpeg (#522)
* [LPLA] Zero sao_enabled flag for LPLA (#538)
* [AVCe] Fix typos in NumRefActive (#595)
* Refactor cmake (#466)
* Refactor root cmake (#467)
* cm: set correct timeout (#604)
* [EncTools] Make IDRDist U32 (#596)
* Refine for commit 2bc6fab
* Revert "[Jpegd] Modify the resolution judgment condition of the current jpeg (#522)" (#619)
* Revert "Fix vpp_reset and mfxvideovpp_query beh test (#573)" (#620)
* [Jpegd] modify the resolution judgment condition of the current jpeg (#623)
* Fix exception handling issue (#581)
* [VPP] Enabled Denoise2(as known as HVS) on Linux (#592)
* core: removed VPL / MSDK1.x RT switch (#603)
* umc allocator adapter: decoders will not accept locked surface (#629)
* [MPEG2d] fix bug when the bitstream end with slice start code (#598)
* [VPP] Fix increasing num of input channels doesn't work (#647)
* CM: refactored CM integration (#413)
* Refine for commit 8221d4a
* [UMC] remove vm_shared_object file (#492)
* Revert "CM: refactored CM integration" (#654)
* Revert "umc allocator adapter: decoders will not accept locked surface (#629)" (#655)
* umc allocator adapter: decoders will not accept locked surface (#3114) (#657)
* [Jpeg]modify color range on Linux (#653)
* Fixes. Code review 1 (#640)
* Revert "[Jpegd] modify the resolution judgment condition of the current jpeg (#623)" (#663)
* [Encode] MFX_EXT_BRC_ENABLE rename to MFX_ENABLE_EXT_BRC (#627)
* vaapi core: clean up unused interface (#667)
* [VPP] Add surface support(in CPU memory) for HVS denoise (#601)
* [SCD] Fix sys mem allocate issue (#690)
* Disable SCD (#708)
* [core] Move CloneSession to RT (#695)
* [Encode] Support main10 still picture profile (#713)
* Refine for commit 0a3c8e7
* [Decode]Enable AVC CSC via sfc pipeline (#712)
* [Decode]Enable VP9 CSC via sfc pipeline (#722)
* Removed multiple calls vaGetConfigAttributes (#716)
* Remove assert in ConvertRateControlMFX2VAAPI (#717)
-------------------------------------------------------------------
Thu Sep 23 13:03:55 UTC 2021 - Stefan Dirsch <sndirsch@suse.com>
- created package libmfx-gen (jira#SLE/SLE-19361)
Intel oneVPL GPU Runtime is a Runtime implementation of oneVPL
API for Intel Gen GPUs: Runtime provides access to
hardware-accelerated video decode, encode and filtering.
Reference in New Issue View Git Blame Copy Permalink