ffmpeg | branch: master | Haihao Xiang | Tue Feb 6
16:17:17 2018 +0800| [00ae5c1d3d87efc7f13b86cce80e139c805db179] | committer:
Mark Thompson
lavc/vaapi_encode: Don't return error if the underlying driver doesn't support
B frames
The underlying driver need not support B frames -
ffmpeg | branch: master | Haihao Xiang | Thu May 3
13:51:48 2018 +0800| [314994051b3d10dc17ed0c8f5669cc20450cd413] | committer:
Mark Thompson
vaapi_encode_h264: Take VAAPIEncodeH264Context::sei_needed as an ORed value
We may check priv->sei_needed only when writing SEI messages.
Signed-
ffmpeg | branch: master | Haihao Xiang | Fri May 4
12:33:39 2018 +0800| [65be65da37eb543e7e79357f04e4cfc8b514d8df] | committer:
Mark Thompson
cbs_h264: Need [] in the name when subscript is required
Otherwise it will hit an assert in the function
ff_cbs_trace_syntax_element() in cbs.c, line
ffmpeg | branch: master | Haihao Xiang | Fri May 4
22:41:07 2018 +0800| [1b0e0578c2ed1bf63e7c014906e02d654f286c0c] | committer:
Mark Thompson
vaapi_encode_vp8: memset the the structure to 0
The structure has reserved bytes, it is required to set the reserved
bytes to 0 for future use
ffmpeg | branch: master | Haihao Xiang | Thu May 10
11:08:22 2018 +0800| [7e78801fa5802dd49712d2b1fdb8bb31d8373ecc] | committer:
Mark Thompson
vaapi_encode: Add an assert in vaapi_encode_truncate_gop()
The flag of input_available must be set when pic_start is not NULL, so
add an assert to
ffmpeg | branch: master | Haihao Xiang | Thu May 10
11:08:23 2018 +0800| [bed670a1de29b58fcb3fe046562d8bd125b1457f] | committer:
Mark Thompson
hwcontext_vaapi: Add an assert in vaapi_map_from_drm()
Every fourcc in vaapi_drm_format_map should be in
vaapi_format_map, so add an assert to ensure
ffmpeg | branch: master | Haihao Xiang | Thu May 10
13:47:45 2018 +0800| [345b6962b68b78bd435d1c931935e86008bf309a] | committer:
Mark Thompson
vaapi_encode_h265: Insert mastering display colour volume
'-sei xxx' is added to control SEI insertion, so far only mastering
display colour
ffmpeg | branch: master | Haihao Xiang | Thu May 10
13:47:44 2018 +0800| [56ed01169269451b7bb327c6bb250eec36ee3de2] | committer:
Mark Thompson
cbs_h265: read/write HEVC PREFIX SEI
Similar to H264, cbs_h265_{read, write}_nal_unit() can handle HEVC
prefix SEI NAL units. Currently mastering
ffmpeg | branch: master | Haihao Xiang | Thu May 10
13:47:46 2018 +0800| [2943dd35b7591b8e84ee8bf2e8dbc78627f36632] | committer:
Mark Thompson
cbs_h265: read/write content light level information SEI message
Signed-off-by: Haihao Xiang
> http://git.videolan.org/gitweb.cgi/ffmpeg.git
ffmpeg | branch: master | Haihao Xiang | Thu May 10
13:47:47 2018 +0800| [0612e29b59d46942f732d6d59dbc863ac1b1e980] | committer:
Mark Thompson
vaapi_encode_h265: Insert content light level information
Signed-off-by: Haihao Xiang
> http://git.videolan.org/gitweb.cgi/ffmpeg.git/?a=commi
ffmpeg | branch: master | Haihao Xiang | Fri Mar 30
10:42:32 2018 +0800| [4c14a25281bb496092e2a6f930e4f1f9e900514f] | committer:
Michael Niedermayer
lavc/hevc: Don't parse NAL unit for a dummy buffer
hevc parser mistakenly reports the following message if a dummy buffer
is padded fo
ffmpeg | branch: master | Haihao Xiang | Tue Jun 11
12:55:47 2024 +0800| [a4630d479a54058f5abfd56fdcfd0b454e5de0ee] | committer:
Lynne
lavu/hwcontext_vulkan: Support write on drm frame
Otherwise nothing is written into the destination when a write mapping
is requested.
For example, a vulkan
ffmpeg | branch: master | Haihao Xiang | Fri Jul 24
15:29:03 2020 +0800| [d951eea6fd7d62acff70640db002997db01315fb] | committer:
Mark Thompson
hwcontext_vaapi: avoid fd leak in vaapi_device_derive
> http://git.videolan.org/gitweb.cgi/ffmpeg.git/?a=commi
ffmpeg | branch: master | Haihao Xiang | Fri Jul 24
10:39:09 2020 +0800| [fa3da243e6ba1567aadfa198bafd32c320e01205] | committer:
Mark Thompson
hwcontext_vaapi: remove duplicate formats from sw_format list
hwcontext_vaapi maps different VA fourcc to the same pix_fmt for U/V
plane swap cases
ffmpeg | branch: master | Haihao Xiang | Wed Sep 16
15:07:54 2020 +0800| [cc25ae5d8ad2cef2dc8a21b828e89e5077b9dae3] | committer:
Zhong Li
lavc/qsvdec: Add QSV AV1 decoder
AV1 decoder is supported on Tiger Lake+ platforms since libmfx 1.34
Signed-off-by: Haihao Xiang
Signed-off-by: Zhong Li
ffmpeg | branch: master | Haihao Xiang | Mon Dec 13
14:50:35 2021 +0800| [50c38e1a4453f8073b980e91b336e7d94e65769a] | committer:
Zhong Li
lavc/qsvenc: define profile array per codec
The SDK defines HEVC, VP9 and AV1 profiles in the same values
e.g.
MFX_PROFILE_HEVC_MAIN =1
ffmpeg | branch: master | Haihao Xiang | Mon Dec 13
14:50:36 2021 +0800| [2544f709ba8ae2b4929183ddf37f3e3acc6cac5a] | committer:
Zhong Li
lavc/qsvenc: add VP9 profiles
Signed-off-by: Zhong Li
> http://git.videolan.org/gitweb.cgi/ffmpeg.git/?a=commi
ffmpeg | branch: master | Haihao Xiang | Mon Dec 13
14:50:34 2021 +0800| [3857ecbe70e81cb6ad7a7f155c311e8522b93b3e] | committer:
Zhong Li
lavc/qsvenc: remove VC1 profiles
The SDK doesn't support VC1 encoding. In addition, both
MFX_PROFILE_VC1_SIMPLE and MFX_PROFILE_HEVC_MAIN are 1 in th
ffmpeg | branch: master | Haihao Xiang | Mon Dec 13
14:50:38 2021 +0800| [bbe9faa01cb1a60ddcfb35c79907bdba4bf24a5c] | committer:
Zhong Li
lavc/qsvenc: dump parameters for mjpeg encoding in verbose mode
Signed-off-by: Zhong Li
> http://git.videolan.org/gitweb.cgi/ffmpeg.git/?a=commi
ffmpeg | branch: master | Haihao Xiang | Mon Dec 13
14:50:37 2021 +0800| [45801af11f29c2448db0b89f81b904b568bcd7dc] | committer:
Zhong Li
lavc/qsvenc: dump parameters for VP9 encoding in verbose mode
Signed-off-by: Zhong Li
> http://git.videolan.org/gitweb.cgi/ffmpeg.git/?a=commi
ffmpeg | branch: master | Haihao Xiang | Wed Dec 22
16:26:02 2021 +0800| [ba44843f405d69ba51332b05a1eb959e93343567] | committer:
Haihao Xiang
MAINTAINERS: add my gpg fingerprint
Signed-off-by: Haihao Xiang
> http://git.videolan.org/gitweb.cgi/ffmpeg.git/?a=commi
ffmpeg | branch: master | Haihao Xiang | Fri Jan 29
08:54:32 2021 +0800| [8d4c0c8823194c71bb825f2ccd28b912f3791e08] | committer:
Haihao Xiang
lavc/qsvdec: needn't free the string for AV_OPT_TYPE_STRING AVOption
The string for AV_OPT_TYPE_STRING AVOption gets freed by av_opt_free()
ffmpeg | branch: master | Haihao Xiang | Mon Dec 13
14:40:55 2021 +0800| [0d82613b7dcdbe9e51b77f06baa72d5715343b8f] | committer:
Haihao Xiang
lavc/qsvenc: set base address for V plane
The SDK checks Data.V when using system memory for VP9 encoding. This
fixed the error below:
$ ffmpeg
ffmpeg | branch: master | Haihao Xiang | Wed Jan 5
11:52:07 2022 +0800| [7c6f9b9d638a9ae6020c8a388d5a6f2a89620663] | committer:
Haihao Xiang
Revert "avutils/hwcontext: When deriving a hwdevice, search for existing device
in both directions"
This reve
ffmpeg | branch: master | Haihao Xiang | Mon Sep 28
14:14:42 2020 +0800| [c4ae6908f28ceacb39b96de9f8761aae1968bf42] | committer:
Haihao Xiang
lavc/qsvenc: add encode support for screen content coding extension
Enables HEVC Screen Content Coding extension support on ICL+ platform
Signed-off
ffmpeg | branch: master | Haihao Xiang | Fri Aug 6
10:10:33 2021 +0800| [641c4346b3d7fa1c023792a260ed0165c6d4a2a4] | committer:
Haihao Xiang
lavc/qsvenc_hevc: add -pic_timing_sei option
The SDK may insert picture timing SEI for hevc and the code to set mfx
parameter has been added in qsvenc
ffmpeg | branch: master | Haihao Xiang | Mon Jan 24
16:24:53 2022 +0800| [8dd507bf0daa826691ea5960ed5634233518c57c] | committer:
Haihao Xiang
lavc/qsv: allow to add more parameter buffers to QSV frame
Signed-off-by: Haihao Xiang
> http://git.videolan.org/gitweb.cgi/ffmpeg.git/?a=commi
ffmpeg | branch: master | Haihao Xiang | Mon Jan 24
16:24:54 2022 +0800| [7c24a768d3e9625183835c1051a72158154a11d8] | committer:
Haihao Xiang
lavc/qsvdec: track the runtime session version
We may check the runtime version for the given features
Signed-off-by: Haihao Xiang
> h
ffmpeg | branch: master | Haihao Xiang | Mon Jan 24
16:24:55 2022 +0800| [8ca06a8148db1b5e8394b2941790fcae29a84f46] | committer:
Haihao Xiang
lavc/qsvdec: export AVFilmGrainParams side data
When AV_CODEC_EXPORT_DATA_FILM_GRAIN is present, AV1 decoder should
disable film grain application and
ffmpeg | branch: master | Haihao Xiang |
Thu Jan 13 13:45:25 2022 +0800| [80801e570566976195f515216de4403cdcf4f7a3] |
committer: Haihao Xiang
lavc/qsvenc: add tile encoding support for VP9
Add -tile_rows and -tile_cols options to specify the number of tile
rows and columns
Signed-off-by
ffmpeg | branch: master | Haihao Xiang | Tue Feb 8
12:50:03 2022 +0800| [2727ff069e856c8fd098ddb8c26f42ec3b54a252] | committer:
Haihao Xiang
lavu/fifo: fix regression
offset_w might be updated after growing the FIFO
Fix ticket #9630
Tested-by: U. Artie Eoff
Reviewed-by: mkver
Reviewed-by
ffmpeg | branch: master | Haihao Xiang |
Thu Apr 7 15:00:53 2022 +0800| [b2dd3e844d9e4d16cdd336157e07d3b2c9240017] |
committer: Haihao Xiang
qsvenc: fix typo
Reviewed-by: U. Artie Eoff
Signed-off-by: Haihao Xiang
> http://git.videolan.org/gitweb.cgi/ffmpeg.git/?a=commi
ffmpeg | branch: master | Haihao Xiang | Mon Nov 27
14:01:50 2023 +0800| [e233f3e75fda8dd60a3dc5f55b7a4bd32b238b6a] | committer:
Haihao Xiang
lavc/qsvdec: return 0 if more data is required
The type of qsv decoders is FF_CODEC_CB_TYPE_DECODE which must not
return AVERROR(EAGAIN). commit
ffmpeg | branch: master | Haihao Xiang | Tue Nov 28
13:38:09 2023 +0800| [d36d9994e4273371c457d37b0a161c887ae9f6b1] | committer:
Haihao Xiang
lavu/hwcontext_vaapi: ignore nonexistent device in default DRM device selection
It is possible that renderD128 doesn't exist but renderD1
ffmpeg | branch: master | Haihao Xiang | Mon Nov 27
16:11:52 2023 +0800| [5717fbbea262a028bcfa751e6c0e6798284dfade] | committer:
Haihao Xiang
configure: don't warn deprecated symbols from libvpl
libvpl deprecated some symbols (e.g. MFX_EXTBUFF_VPP_DENOISE2 is used to
re
ffmpeg | branch: master | Haihao Xiang | Mon Nov 27
14:01:51 2023 +0800| [fc73b372cd43e52d3aa0aefc6d76fd340bf753f1] | committer:
Haihao Xiang
lavc/qsvdec: reduce info message when more data is required
demote the info to AV_LOG_VERBOSE
Signed-off-by: Haihao Xiang
> http://git.videolan.
ffmpeg | branch: master | Haihao Xiang | Mon Nov 27
13:37:22 2023 +0800| [05debdaa5fde6440b0575d7e2a1fd2e357b55f65] | committer:
Haihao Xiang
lavu/hwcontext_qsv: use mfxImplDescription instead of mfxExtendedDeviceId on
Linux
mfxExtendedDeviceId mightn't be supported in certain configura
ffmpeg | branch: master | Haihao Xiang | Fri Nov 24
15:00:55 2023 +0800| [35a555e2b98ecf79d217c421b912549b90d223b0] | committer:
Haihao Xiang
lavfi/vf_vpp_qsv: set the default value of async_depth to 4
Both qsv encoders and decoders use 4 as the default value of
async_depth, let's use
ffmpeg | branch: master | Haihao Xiang | Tue Nov 28
12:42:27 2023 +0800| [f89cff96d080a238e3b7c7020db08c0f8ed796e4] | committer:
Haihao Xiang
lavu/hwcontext_qsv: Make sure hardware vendor is Intel for qsv on d3d11va
When multiple hardwares are available, the default one might not be
Intel
ffmpeg | branch: master | Haihao Xiang | Tue Nov 28
12:42:26 2023 +0800| [e5f8b5313e2993fc53c74a4e58a3b630da7a703e] | committer:
Haihao Xiang
doc/ffmpeg: Update the description about d3d11va
Signed-off-by: Haihao Xiang
> http://git.videolan.org/gitweb.cgi/ffmpeg.git/?a=commi
ffmpeg | branch: release/6.1 | Haihao Xiang | Mon Nov
27 14:01:50 2023 +0800| [d596225a57f1d49f3f1931a073de6e548b9440d9] | committer:
Haihao Xiang
lavc/qsvdec: return 0 if more data is required
The type of qsv decoders is FF_CODEC_CB_TYPE_DECODE which must not
return AVERROR(EAGAIN). commit
ffmpeg | branch: master | Haihao Xiang | Fri Dec 29
12:23:26 2023 +0800| [440e08e4bb463d37de1d67a1fc47ebae542570ef] | committer:
Haihao Xiang
configure: autodetect libdrm
option kernel_driver for vaapi device creation can be used to choose the
desired device on Linux, which is more convenient
ffmpeg | branch: master | Haihao Xiang | Tue Jan 16
13:23:10 2024 +0800| [6414ecd85754af9dc758d067841fdff079d1bd71] | committer:
Stefano Sabatini
doc/ffmpeg: update the documentation about qsv device creation
Signed-off-by: Haihao Xiang
> http://git.videolan.org/gitweb.cgi/ffmpeg.git
ffmpeg | branch: master | Haihao Xiang | Tue Jan 16
13:23:11 2024 +0800| [6e4d85a94ef05655f22a86317a44a42d1ae22d61] | committer:
Stefano Sabatini
doc/ffmpeg: update the documentation about vaapi device creation
Signed-off-by: Haihao Xiang
> http://git.videolan.org/gitweb.cgi/ffmpeg.git
ffmpeg | branch: master | Haihao Xiang | Fri Jan 26
12:28:22 2024 +0800| [cd31eac9997dc49b1d2b8de97b38c51ed8e9f4c9] | committer:
Haihao Xiang
lavc/qsvenc: Add workaround for VP9 keyframe
The runtime doesn't set the frame type to MFX_FRAMETYPE_IDR on the
returned mfx bitstream for a key
ffmpeg | branch: master | Haihao Xiang | Thu Feb 1
16:35:35 2024 +0800| [d263fce2b209e86a5a1e8f1b6aa33430ecc2c187] | committer:
Haihao Xiang
lavc/qsvenc: update the selection of bitrate control method
The default method is changed to CQP
Signed-off-by: Haihao Xiang
> h
ffmpeg | branch: master | Haihao Xiang | Tue Mar 5
09:45:54 2024 +0800| [eb7d019b326b0c34e07f2369f31e8ba266fa666b] | committer:
Haihao Xiang
lavfi/vulkan_filter: fix input format
Otherwise s->input_format is always yuv420p.
This fixes invalid output format for hwframe download in
ffmpeg | branch: master | Haihao Xiang | Mon Mar 11
12:37:57 2024 +0800| [697251bb0c9ea4b66276cf89e1e526535a86d117] | committer:
Haihao Xiang
lavc/qsvdec: Do not print warning when draining cached frames
When all cached frames are drained, the output mfxSyncPoint pointer is
NULL and
ffmpeg | branch: master | Haihao Xiang | Mon Mar 11
12:56:02 2024 +0800| [74a8e080d021b6c88677b35d879d269bfc4f93fb] | committer:
Haihao Xiang
lavu/hwcontext_qsv: Join the download/upload session to the main session
This may reduce the number of internal threads when using hwupload or
ffmpeg | branch: master | Haihao Xiang | Fri Mar 8
09:09:37 2024 +0800| [d296c8689d487e422aeacbb92292b78dac84ff17] | committer:
Haihao Xiang
lavu/hwcontext_vulkan: check PCI ID if possible
Otherwise the derived device and the source device might have different
PCI ID in a multiple-device
ffmpeg | branch: master | Haihao Xiang | Thu Mar 28
14:55:57 2024 +0800| [1590a96adc28a150ad641080262499f58158ee24] | committer:
Haihao Xiang
lavc/vaapi_encode: convert from lambda to qp
When AV_CODEC_FLAG_QSCALE is set, the value of avctx->global_quality is
lambda.
Signed-off-by: Hai
ffmpeg | branch: release/7.0 | Haihao Xiang | Thu Mar
28 14:55:57 2024 +0800| [74e4e900bb52c4a0f5f4fe38e2b7a40239d9e845] | committer:
Haihao Xiang
lavc/vaapi_encode: convert from lambda to qp
When AV_CODEC_FLAG_QSCALE is set, the value of avctx->global_quality is
lambda.
Signed-off
ffmpeg | branch: master | Haihao Xiang | Tue Mar 26
15:19:13 2024 +0800| [41e3d36a39979a5c6ca36198b03be740e14ef7b0] | committer:
Haihao Xiang
lavc/vaapi_decode: Use dynamic frame pool if possible
libva2 doesn't require a fixed surface-array any more, so we may use
dynamic frame poo
ffmpeg | branch: master | Haihao Xiang | Tue Mar 26
15:47:44 2024 +0800| [16616a3d1be07d1b20268df1bd5727bb4ca33c92] | committer:
Haihao Xiang
lavfi/vaapi_vpp: Use dynamic frame pool in outlink if possible
This can avoid to exhaust the buffers within outlink when libva2 is
available.
For
ffmpeg | branch: master | Haihao Xiang | Mon Mar 18
14:06:14 2024 +0800| [42eb10ecc6781b4c2deee180a501156363f2c905] | committer:
Haihao Xiang
lavfi: Add pad_vaapi filter
Signed-off-by: Haihao Xiang
> http://git.videolan.org/gitweb.cgi/ffmpeg.git/?a=commi
ffmpeg | branch: master | Haihao Xiang | Mon Mar 18
14:06:16 2024 +0800| [1b5ca2e63134d50b4a4ea7a8d31d3ffaba145101] | committer:
Haihao Xiang
Changelog: Add pad_vaapi, drawbox_vaapi entry
Signed-off-by: Haihao Xiang
> http://git.videolan.org/gitweb.cgi/ffmpeg.git/?a=commi
ffmpeg | branch: master | Haihao Xiang | Mon Mar 18
16:12:44 2024 +0800| [3172a6722e945bd31c3a71d510471103680a3459] | committer:
Haihao Xiang
lavfi/tonemap_vaapi: By default use bt709 for output frame
By default don't use the color properties from input frame as output
frame properties
ffmpeg | branch: master | Haihao Xiang | Mon Mar 18
14:06:15 2024 +0800| [5c55e4e2975dd6e577fb1b4597e5292496ec8cbb] | committer:
Haihao Xiang
lavfi: Add drawbox_vaapi filter
Signed-off-by: Haihao Xiang
> http://git.videolan.org/gitweb.cgi/ffmpeg.git/?a=commi
ffmpeg | branch: master | Haihao Xiang | Mon Mar 18
16:12:45 2024 +0800| [db22804145a456718b780c5cadbd468aa5e67c68] | committer:
Haihao Xiang
lavfi/tonemap_vaapi: Update the log
demote the message to AV_LOG_VERBOSE.
Signed-off-by: Haihao Xiang
> http://git.videolan.org/gitweb.
ffmpeg | branch: master | Haihao Xiang | Mon Apr 22
20:34:14 2024 +0800| [3032a74cac4f7979d6e70db80360634e48e13827] | committer:
James Almer
configure: fix rpuenc typos
Otherwise there are link errors:
LD ffprobe_g
/usr/bin/ld: libavcodec/libavcodec.so: undefined reference to
ffmpeg | branch: master | Haihao Xiang | Mon Apr 22
14:57:08 2024 +0800| [8c62d77139ca07390414fcfd26b2a4d506fed3b9] | committer:
Haihao Xiang
lavc/vp9: Fix regression introduced in 0ba05857
It is possible that ff_progress_frame_await() is called but
ff_progress_frame_report() isn't c
ffmpeg | branch: master | Haihao Xiang | Tue Apr 23
15:57:52 2024 +0800| [578ac5988769592c3b1d80b58af676e38e45e259] | committer:
Haihao Xiang
lavfi/qsv: Copy metadata fields from the given input
Currently it always copies the metadata fields from the last input when
there are multiple inputs
ffmpeg | branch: master | Haihao Xiang | Thu Feb 29
13:34:33 2024 +0800| [df76c3756d41e909b78565f49928c43c0f84caf0] | committer:
Haihao Xiang
lavc/qsvenc: always allocate the array to store mfxExtBuffer points
This allows us to append mfxExtBuffer per user's settings
Signed-off-by: H
ffmpeg | branch: master | Haihao Xiang | Wed Apr 10
16:17:44 2024 +0800| [7f3ba6bbfaef30839af6996996a713abaec013a7] | committer:
Haihao Xiang
lavc/vaapi_encode_av1: implement write_extra_header callback
This can be used to insert a metadata OBU to the stream later.
Signed-off-by: Haihao
ffmpeg | branch: master | Haihao Xiang | Wed Apr 10
16:24:57 2024 +0800| [ceace488ff801636b5332fd65d9d2c292367ad42] | committer:
Haihao Xiang
lavc/vaapi_encode_av1: insert HDR_CLL metadata if have
Only look for HDR_CLL on key frame on the output.
Signed-off-by: Haihao Xiang
> h
ffmpeg | branch: master | Haihao Xiang | Wed Apr 10
16:24:04 2024 +0800| [0bdf71ada71cd31f5885bcaa5e209adf9a02b5a5] | committer:
Haihao Xiang
lavc/vaapi_encode_av1: Insert HDR_MDCV metadata if have
Only look for HDR_MDVC on key frame on the output.
Signed-off-by: Haihao Xiang
> h
ffmpeg | branch: master | Haihao Xiang | Wed May 8
14:03:08 2024 +0800| [4c0bb7d4a9190a2990bafc4e7938ab06f5d8b3b5] | committer:
Haihao Xiang
lavu/hwcontext_qsv: update AVQSVFramesContext to support dynamic frame pool
Add AVQSVFramesContext.info and update the description.
Signed-off-by
ffmpeg | branch: master | Haihao Xiang | Fri May 17
13:05:07 2024 +0800| [3178c99fa93bd44eabbe3adc5d4a3dd089445dee] | committer:
Haihao Xiang
lavu/version: fix minor version
The latest version should be 59.18.100 since commit 01c5f4ad
$ git diff 01c5f4ad~1..HEAD doc/APIchanges
...
+2024-05
ffmpeg | branch: master | Haihao Xiang | Wed May 8
14:03:11 2024 +0800| [d3cc5ead42dccdf5ae8f2942c629ed7f691f6948] | committer:
Haihao Xiang
lavu/hwcontext_qsv: add support for dynamic frame pool in qsv_map_to
Make it work with the source which has a dynamic frame pool.
Signed-off-by
ffmpeg | branch: master | Haihao Xiang | Wed May 8
14:03:09 2024 +0800| [96db4a62e0ac642441e25b74c4e6fb7233d1350c] | committer:
Haihao Xiang
lavu/hwcontext_qsv: create dynamic frame pool if required
When AVHWFramesContext.initial_pool_size is 0, a dynamic frame pool is
required. We may
ffmpeg | branch: master | Haihao Xiang | Wed May 8
14:03:15 2024 +0800| [5646285f761c4c7f6104bb5c956a393a2f6239c0] | committer:
Haihao Xiang
lavfi/qsvvpp: use the right mfxFrameInfo when dynamic frame pool is used
Signed-off-by: Haihao Xiang
> http://git.videolan.org/gitweb.cgi/ffmpeg.
ffmpeg | branch: master | Haihao Xiang | Wed May 8
14:03:13 2024 +0800| [75015f9b0e0ef203bd051ef01bf2cb18710b8379] | committer:
Haihao Xiang
lavc/qsvenc: use the right info for encoding
Signed-off-by: Haihao Xiang
> http://git.videolan.org/gitweb.cgi/ffmpeg.git/?a=commi
ffmpeg | branch: master | Haihao Xiang | Wed May 8
14:03:10 2024 +0800| [932f78c4e561ad70ee92aa38475a716070c0e57d] | committer:
Haihao Xiang
lavu/hwcontext_qsv: add support for dynamic frame pool in qsv_frames_derive_to
Make it work with the source which has a dynamic frame pool.
Signed-off
ffmpeg | branch: master | Haihao Xiang | Wed May 8
14:03:16 2024 +0800| [127ded507828a1127b0196ffde115171ab5c46f3] | committer:
Haihao Xiang
lavfi/qsvvpp: require a dynamic frame pool for output if possible
Signed-off-by: Haihao Xiang
> http://git.videolan.org/gitweb.cgi/ffmpeg.git
ffmpeg | branch: master | Haihao Xiang | Wed May 8
14:03:14 2024 +0800| [a00cfc6c2461d9a2525b01e64fe140615c5d5f34] | committer:
Haihao Xiang
lavc/qsvdec: require a dynamic frame pool if possible
This allows a downstream element stores more frames from qsv decoders
and fixes error in
ffmpeg | branch: master | Haihao Xiang | Wed May 8
14:03:12 2024 +0800| [cda721e01de571cd4004b8710526e3d2b47f6fd0] | committer:
Haihao Xiang
lavc/qsv: fix the mfx allocator to support dynamic frame pool
When the external allocator is used for dynamic frame allocation, only
video memory is
ffmpeg | branch: master | Haihao Xiang | Thu May 23
14:51:36 2024 +0800| [8155808ce6030a6aed3201cf426a256776a36392] | committer:
James Almer
libavcodec/x86/vvc/vvc_sad: fix assembler error
X86ASMlibavcodec/x86/vvc/vvc_sad.o
libavcodec/x86/vvc/vvc_sad.asm:85: error: invalid number of
ffmpeg | branch: master | Haihao Xiang | Wed May 22
12:16:00 2024 +0800| [a72e9aeabcbe04765e3340ae9e5022b46cbfff56] | committer:
Haihao Xiang
lavc/qsvenc_av1: accept HDR metadata if have
The sdk av1 encoder can accept HDR metadata via mfxEncodeCtrl::ExtParam.
Signed-off-by: Haihao Xiang
ffmpeg | branch: master | Haihao Xiang | Wed May 22
12:15:59 2024 +0800| [473e84ad62a05e83bba5e00b4073980ee171a5bd] | committer:
Haihao Xiang
lavc/qsvdec: update HDR side data on output AVFrame for AV1 decoding
The SDK may provide HDR metadata for HDR streams via mfxExtBuffer
attached on
ffmpeg | branch: master | Haihao Xiang | Thu May 23
10:24:36 2024 +0800| [c872ba5899eb520bac313e063336c007852499ea] | committer:
Haihao Xiang
lavc/qsvenc: respect user's setting for keyframes
For example:
./ffmpeg -hwaccel qsv -i input.mp4 -force_key_frames:v source -c:v
hevc_qsv -f
ffmpeg | branch: master | Haihao Xiang | Wed May 22
12:31:53 2024 +0800| [dbdd9ccded9018718bad6df14b56bdd4b7e327f1] | committer:
Haihao Xiang
lavc/qsvdec: fix keyframes
MFX_FRAMETYPE_IDR is ORed to the frame type for AVC and HEVC keyframes,
and MFX_FRAMETYPE_I is taken as keyframe flag for
ffmpeg | branch: master | Haihao Xiang | Tue Jan 5
15:02:55 2021 +0800| [db28e9c1d98c956328dd106885b198c8b187202f] | committer:
Linjie Fu
lavf/qsv: Add functions to print mfx iopattern, warning and error
It is a copy of the relevant part in lavc/qsv but use different function
names to avoid
ffmpeg | branch: master | Haihao Xiang | Tue Jan 5
15:02:56 2021 +0800| [b8cd37a59fe249daf9d84cb0d4ea0a6f42d208af] | committer:
Linjie Fu
qsv: dump more info in error, debug and verbose mode
Dump iopattern mode and the SDK error/warning desciptions for qsv based
filters and iopattern mode
ffmpeg | branch: master | Haihao Xiang | Fri Jun 11
10:19:46 2021 +0800| [efc22e63e5899f5eda8ee50fc92ecd752aedce87] | committer:
Zhong Li
lavfi/vf_vpp_qsv: fix the time_base for outlink
Since commit 89ffcd1, the status pts of the output link is set to a
value in the input link time base, not
ffmpeg | branch: master | Haihao Xiang | Fri Jun 11
10:19:45 2021 +0800| [971b4ac733630e8d913e17151ce8e01d3c75d1a1] | committer:
Zhong Li
lavc/qsvdec: fix pts
The time base used for compressed bitstream and video frame in the SDK
is { 1, 9 }. [1][2]
This can avoid the error message below
ffmpeg | branch: master | Haihao Xiang | Thu Nov 5
15:20:10 2020 +0800| [9583e66ea0232043353bec7bda5f82554e04232c] | committer:
Zhong Li
qsvenc_hevc: allow user set more coding options
The SDK supports NalHrdConformance, RecoveryPointSEI and AUDelimiter for
hevc encoder, so we may allow user
ffmpeg | branch: master | Haihao Xiang | Thu Nov 5
15:08:23 2020 +0800| [f2feb31b78be9f862867d567f12758b635b32b74] | committer:
Zhong Li
qsvenc: Value '0' is not a valid value for parameter GopOptFlag
The accepted values for GopOptFlag are MFX_GOP_CLOSED (1) and
MFX_GOP_STRICT (2
ffmpeg | branch: master | Haihao Xiang | Wed Jul 28
16:15:47 2021 +0800| [65fdc0e589551179f793f69528a975b2253699d2] | committer:
James Almer
lavc/qsvenc: pass the color properties to the SDK
Otherwise the color properties won't be encoded into the bitstream
header
Reviewed-by: Xiang, H
ffmpeg | branch: master | Haihao Xiang | Wed Jul 28
16:15:46 2021 +0800| [c8cfe676948c1d97c8d6d50564065d0f2d52a568] | committer:
James Almer
lavc/qsvdec: update color properties in codec context
User may get color properties from the SDK via VIDEO_SIGNAL_INFO extbuf
Reviewed-by: Xiang
ffmpeg | branch: master | Haihao Xiang | Fri Jul 30
10:39:24 2021 +0800| [43aeeab764c6fd89aea67be9c0b16cedd78a] | committer:
James Almer
lavfi/qsvvpp: do not mix up FFmpeg and SDK error code
The function ff_qsvvpp_filter_frame should return a FFmpeg error code if
there is an error
ffmpeg | branch: master | Haihao Xiang | Wed Aug 11
14:01:55 2021 +0800| [c3ef9e027769b8fb272ff2794b3e84c98bddea7d] | committer:
James Almer
doc/examples/qsvdec: simplify this example via hw_device_ctx interface
Signed-off-by: James Almer
> http://git.videolan.org/gitweb.cgi/ffmpeg.git
ffmpeg | branch: master | Haihao Xiang | Wed Aug 11
14:01:54 2021 +0800| [ecee3b07cde23e05bcc6b4eaa55d860b62dbd2dc] | committer:
James Almer
qsvdec: add support for HW_DEVICE_CTX method
This allows user set hw_device_ctx instead of hw_frames_ctx for QSV
decoders, hence we may remove the ad
ffmpeg | branch: master | Haihao Xiang | Thu Aug 12
10:33:56 2021 +0800| [115f5e803502451ac64fa698730923299e5acc4e] | committer:
James Almer
lavc/qsvenc: allows the SDK runtime to choose LowPower/non-LowPower modes
The SDK supports LowPower and non-LowPower modes, but some features are
ffmpeg | branch: master | Haihao Xiang | Wed Aug 11
14:44:05 2021 +0800| [51a80aacce1a7bd20823798dc9e5ec5f23a3b62d] | committer:
James Almer
ffmpeg_hw: Don't ignore key parameters when initializing a hw device
Currently user may use '-init_hw_device type=name' to initial
ffmpeg | branch: master | Haihao Xiang | Tue Jun 29
09:45:38 2021 +0800| [6803d8b4403ecea18fc288475b4d7cdedd0fb744] | committer:
James Almer
ffmpeg_hw: make hardware selection for filters more user friendly
When a device is derived from a source device, there are at least 2
devices, and
ffmpeg | branch: master | Haihao Xiang | Wed Sep 28
09:39:32 2022 +0800| [fce8b908511e57d9c77434280398296ad539f8b2] | committer:
Haihao Xiang
lavc/cbs_av1: restore CodedBitstreamAV1Context when AVERROR(ENOSPC)
The current pbc might be small for an obu frame, so a new pbc is
required then
ffmpeg | branch: master | Haihao Xiang | Thu Oct 6
15:35:34 2022 +0800| [3f28116ea26a87fbe1c807248effaf449008d979] | committer:
Haihao Xiang
lavc/qsv: specify Shift for each format too
Signed-off-by: Haihao Xiang
> http://git.videolan.org/gitweb.cgi/ffmpeg.git/?a=commi
ffmpeg | branch: master | Haihao Xiang | Thu Oct 6
15:35:35 2022 +0800| [aba25b391c7f68cea456d1b862a71bcd105b46f3] | committer:
Haihao Xiang
lavu/hwcontext_qsv: add support for 10bit 4:4:4 content on Linux
XV30 is used for 10bit 4:4:4 content in FFmpeg VAAPI, so XV30 should be
used for 10bit
ffmpeg | branch: master | Haihao Xiang | Thu Oct 6
15:35:33 2022 +0800| [1496e7c1732f6f2e2e8d9c6a1866c13756e54ac4] | committer:
Haihao Xiang
lavu/hwcontext_qsv: specify Shift for each format
We can't get Shift from bit depth for some formats in the SDK. For
example, bit depth is 10, ho
ffmpeg | branch: master | Haihao Xiang | Fri Oct 7
10:49:12 2022 +0800| [f3b5277057ad84071721f01419fe4badeceaff08] | committer:
Haihao Xiang
lavc/qsvenc_hevc: use open GOP by default
HEVC spec has CRA frame which allows random access with open GOP, hence
it can achieve higher compression
1 - 100 of 163 matches
Mail list logo
