ffmpeg | branch: master | Lynne | Tue Dec 17 14:56:33 2019
+| [377a095dc37222a373fba00579145e77629a5e81] | committer: Lynne
vc2enc: zero padding of the coefficient buffer
Wavelet types with large amounts of overreading/writing like 9_7 would
write into the padding at high wavelet depths
ffmpeg | branch: master | Lynne | Fri Jan 10 21:55:19 2020
+| [9e01f171f3b7d408f93ec0409d8f9be1d8b291d8] | committer: Lynne
tiffdec: support embedded ICC profiles
> http://git.videolan.org/gitweb.cgi/ffmpeg.git/?a=commit;h=9e01f171f3b7d408f93ec0409d8f9be1d8b291d8
---
libavcodec/tif
ffmpeg | branch: master | Lynne | Sun Oct 27 14:44:00 2019
+| [6fca61bbc917678b3e517be4f6594ce52a16a93c] | committer: Lynne
lavfi: add Vulkan filtering framework
This commit adds a Vulkan filtering infrastructure for libavfilter.
It attempts to abstract as much as possible of the Vulkan
ffmpeg | branch: master | Lynne | Sun Oct 27 14:47:18 2019
+| [a2db7343e02fe4c78c6d301550849a0634f0ac38] | committer: Lynne
lavfi: add an avgblur_vulkan filter
This commit adds a fast avgblur Vulkan filter.
This will reset Intel GPUs on Linux due to a known, two-year-old driver bug
(!834
ffmpeg | branch: master | Lynne | Sun Oct 27 14:46:16 2019
+| [7bb443137c656857449d56d1212d09a7018819b0] | committer: Lynne
lavfi: add an overlay_vulkan filter
This commit adds a basic, non-converting overlay filter for Vulkan.
> http://git.videolan.org/gitweb.cgi/ffmpeg.git/?a=commi
ffmpeg | branch: master | Lynne | Sun Oct 27 14:45:36 2019
+| [d95c509cc64372f8b37d89310250785224751a90] | committer: Lynne
lavfi: add an scale_vulkan filter
This commit adds a basic, non-converting Vulkan scaling filter.
> http://git.videolan.org/gitweb.cgi/ffmpeg.git/?a=commi
ffmpeg | branch: master | Lynne | Sun Oct 27 14:48:16 2019
+| [907ae87d6eb702eb12f073df2eac6da2560eae6f] | committer: Lynne
lavfi: add an chromaber_vulkan filter
This commit adds a chromatic aberration filter for Vulkan that attempts to
emulate a lens chromatic aberration effect.
For a YUV
ffmpeg | branch: master | Lynne | Tue Feb 4 23:49:56 2020
+| [73a8c8e6e47097059faaf758b7056d6657b6] | committer: Lynne
lavu: bump minor version for the Vulkan patchset
> http://git.videolan.org/gitweb.cgi/ffmpeg.git/?a=commit;h=73a8c8e6e47097059faaf758b7056d6657b6
---
libavu
ffmpeg | branch: master | Lynne | Tue Feb 4 23:52:29 2020
+| [a71a5d9214eb191394b3104c8b5d7766654e323a] | committer: Lynne
lavfi: bump minor version for the Vulkan filters
> http://git.videolan.org/gitweb.cgi/ffmpeg.git/?a=commit;h=a71a5d9214eb191394b3104c8b5d7766654e3
ffmpeg | branch: master | Lynne | Tue Feb 4 23:51:55 2020
+| [ee81713fe39548be62b699d934fbe4a48cb6b142] | committer: Lynne
doc/APIchanges: update with Vulkan commit info
> http://git.videolan.org/gitweb.cgi/ffmpeg.git/?a=commit;h=ee81713fe39548be62b699d934fbe4a48cb6b142
---
ffmpeg | branch: master | Lynne | Sat Feb 8 23:06:09 2020
+| [d500eff3cce7ab0c6f7101860b633ca955a9f85e] | committer: Lynne
lavu/tx: mention FFT output is not normalized
Not even FFTW's output is normalized.
This should prevent at least some users from complaining that doing a fo
ffmpeg | branch: master | Lynne | Sat Feb 8 23:32:46 2020
+| [9b9f441ab39d12086bfae044094dbcd8979b8d90] | committer: Lynne
doc/APIchanges: fix vulkan hwcontext date
> http://git.videolan.org/gitweb.cgi/ffmpeg.git/?a=commit;h=9b9f441ab39d12086bfae044094dbcd8979b8d90
---
doc/APIchan
ffmpeg | branch: master | Lynne | Mon Feb 10 17:37:34 2020
+| [223b58c74b4070b5726ed29ddd8df7b343c6b78a] | committer: Lynne
lavu/tx: slightly optimize fft15
Saves 2 additions.
> http://git.videolan.org/gitweb.cgi/ffmpeg.git/?a=commit;h=223b58c74b4070b5726ed29ddd8df7b343c6b
ffmpeg | branch: master | Lynne | Mon Feb 10 17:36:45 2020
+| [a38c6f47c978187c37c0154ced84f1ea4810c7ab] | committer: Lynne
lavu/tx: undef the correct macro
It was renamed and no warning was given for undeffing a nonexisting one.
> http://git.videolan.org/gitweb.cgi/ffmpeg.git/?a=commi
ffmpeg | branch: master | Lynne | Sat Feb 8 23:13:28 2020
+| [e8f054b095baa194623b3852f06fc507ae697503] | committer: Lynne
lavu/tx: implement 32 bit fixed point FFT and MDCT
Required minimal changes to the code so made sense to implement.
FFT and MDCT tested, the output of both was
ffmpeg | branch: master | Lynne | Fri Feb 14 19:55:00 2020
+| [e1c84856bb7d804e74904ba117a2ca9700211082] | committer: Lynne
lavu/tx: improve 3-point fixed precision
There's just no reason not to when its so easy (albeit messy) and its also
reducing the precision of all non-power-o
ffmpeg | branch: master | Lynne | Thu Mar 12 18:01:07 2020
+| [501bd57bdbc488db93c95d40682ef0b4f01ccec5] | committer: Lynne
hwcontext_vulkan: initialize semaphores of DMABUF imports
There was a recent change in Intel's driver that triggered a driver-internal
error if the semaphore giv
ffmpeg | branch: master | Lynne | Thu Mar 12 18:59:12 2020
+| [6353b9e4abd8e3e0c9ae743701f38510939fe6f9] | committer: Lynne
hwcontext_vulkan: support more than one plane per DMABUF layer
Requires the dmabuf modifiers extension.
Allows for importing of compressed images with a second plane
ffmpeg | branch: master | Lynne | Thu Mar 12 18:03:43 2020
+| [b31959d776e6da3ff8519121bafae9753f2be20f] | committer: Lynne
hwcontext_vulkan: duplicate DMABUF objects before importing them
The specifications are very vague about who has ownership, and in this case,
Vulkan takes ownership
ffmpeg | branch: master | Lynne | Thu Mar 12 16:46:01 2020
+| [08d0a8992d75bbcffe9e7f3392cc94a7f84fc9df] | committer: Lynne
hwcontext_vulkan: minor corrections for DMABUF mapping
We need to consider the amount of layers instead of the image's planes.
> http://git.videolan.org/gi
ffmpeg | branch: master | Lynne | Thu Mar 12 16:57:14 2020
+| [9086af2a0a590c7f576b72379d1708392cd96d5c] | committer: Lynne
hwcontext_vulkan: only convert image layout for transfers if necessary
> http://git.videolan.org/gitweb.cgi/ffmpeg.git/?a=commi
ffmpeg | branch: master | Lynne | Thu Mar 12 20:20:22 2020
+| [d778be6e4a0565e9a96adec57339e4c8a2464664] | committer: Lynne
changelog: add entry for the Vulkan hwcontext and filters
> http://git.videolan.org/gitweb.cgi/ffmpeg.git/?a=commit;h=d778be6e4a0565e9a96adec57339e4c8a2464
ffmpeg | branch: master | Lynne | Sun Mar 1 11:23:53 2020
+| [675bb1f4f9de76f2c0eb1c8b1be6781a2cd52d29] | committer: Lynne
diracdec: rewrite golomb reader
This version is able to output multiple coefficients at a time and
is able to altogether remove actual golomb code parsing.
Its also
ffmpeg | branch: master | Lynne | Tue Mar 17 13:08:06 2020
+| [ecc3dceff44b8612c80014d45a514df882883940] | committer: Lynne
hwcontext_vulkan: fix imported image bitmask
> http://git.videolan.org/gitweb.cgi/ffmpeg.git/?a=commit;h=ecc3dceff44b8612c80014d45a514df882883940
---
libavu
ffmpeg | branch: master | Lynne | Sun Mar 15 10:30:34 2020
+| [1a5e9ae4d8e5ee34c983f16a280c9a262a3fbf8d] | committer: Lynne
scale_vulkan: add support for RGB->YUV conversions
Only top-left chroma position supported for now.
> http://git.videolan.org/gitweb.cgi/ffmpeg.git/?a=co
ffmpeg | branch: master | Lynne | Tue Mar 17 22:56:03 2020
+| [2e611cd9c042009fae49ef0f11f1086a5efd7021] | committer: Lynne
MAINTAINERS: add myself and my gpg key
> http://git.videolan.org/gitweb.cgi/ffmpeg.git/?a=commit;h=2e611cd9c042009fae49ef0f11f1086a5efd7021
---
MAINTAINERS
ffmpeg | branch: master | Lynne | Sat Mar 14 16:10:06 2020
+| [9f494d1397d123efd23cf772eb381e194785f4ea] | committer: Lynne
lavu/tx: improve documentation
> http://git.videolan.org/gitweb.cgi/ffmpeg.git/?a=commit;h=9f494d1397d123efd23cf772eb381e194785f4ea
---
libavutil/tx.h |
ffmpeg | branch: master | Lynne | Tue Feb 25 06:49:26 2020
+| [2465fe1302f2c87b37496a00ab490061a2a7ce03] | committer: Lynne
lavu/tx: add 2-point FFT transform
By itself, this allows 6-point, 10-point and 30-point transforms.
When the 9-point transform is added it allows for 18-point FFT
ffmpeg | branch: master | Lynne | Mon Mar 23 22:03:24 2020
+| [ca7a192d104f08a4dbfda31a6c4f29a1f05074ad] | committer: Lynne
movenc: mark Opus encapsulation as stable
The specifications are de-facto frozen now as they've already been used in
production for years, the author has indi
ffmpeg | branch: master | Lynne | Mon Apr 6 19:15:51 2020
+0100| [ca76a5ba1a35e317d0fb327fffe3ad9add6b7e61] | committer: Lynne
scale_vulkan: correctly copy the colormatrix
> http://git.videolan.org/gitweb.cgi/ffmpeg.git/?a=commit;h=ca76a5ba1a35e317d0fb327fffe3ad9add6b7e61
---
libavfil
ffmpeg | branch: master | Lynne | Mon Apr 6 18:18:50 2020
+0100| [97b526c192add6f252b327245fd9223546867352] | committer: Lynne
hwcontext_vulkan: only use one semaphore per image
The idea was to allow separate planes to be filtered independently, however,
in hindsight, literaly nothing uses
ffmpeg | branch: master | Lynne | Tue Apr 21 18:55:24 2020
+0100| [e3c7b22451799a2890062b756b645cc5f2e8752e] | committer: Lynne
hwcontext_vulkan: correctly download and upload flipped images
We derive the destination buffer stride from the input stride,
which meant if the image was flipped
ffmpeg | branch: master | Lynne | Thu Apr 23 18:08:14 2020
+0100| [f66ac83c22d8c088832ce8df922b08a0caa11833] | committer: Lynne
overlay_vulkan: add support for overlaying images with an alpha channel
> http://git.videolan.org/gitweb.cgi/ffmpeg.git/?a=commi
ffmpeg | branch: master | Lynne | Thu Apr 23 18:09:00 2020
+0100| [b136a983035e41058804fdab59078fdaaf5c5e86] | committer: Lynne
scale_vulkan: take frame cropping parameters in account when scaling
Then sample_aspect_ratio line at the bottom was cargo-culted from the vaapi
scaling filter, but
ffmpeg | branch: master | Lynne | Tue Apr 28 12:25:46 2020
+0100| [8296443a70f052a6f5c9a867d28b83a5eb7d304d] | committer: Lynne
oggdec: add support for proper demuxing of chained Opus files and streams
Part of this patch is based on Paul B Mahol's patch from last year.
This also allow
ffmpeg | branch: master | Lynne | Tue Apr 28 12:55:17 2020
+0100| [e983197cbc93420b67aa7e811be47d7278c2c8a2] | committer: Lynne
oggdec: use ffio_ensure_seekback() to seek back on incorrect data
This cleans up the code and simplifies it.
It also speeds up parsing since the old pb position was
ffmpeg | branch: master | Lynne | Tue Apr 28 12:41:34 2020
+0100| [f619e1ec66b89215582eff4404b681b760540b4f] | committer: Lynne
oggdec: eliminate copies and extra buffers
This also makes implementing CRC checking far simpler and more robust.
> http://git.videolan.org/gitweb.cgi/ffmpeg.git
ffmpeg | branch: master | Lynne | Tue Apr 28 12:52:11 2020
+0100| [9ad47762c17d2c6d06595aa17b88112baa91b72c] | committer: Lynne
oggdec: verify page checksum
This makes decoding far more robust, since OggS, the ogg magic,
can be commonly found randomly in streams, which previously made
the
ffmpeg | branch: master | Lynne | Fri May 1 01:36:43 2020
+0100| [4fd0559b77cfdd38d37a94a45cc830d90454194e] | committer: Lynne
fate/oggopus-demux: fix fate failure
Failure was due to the extra comment printed by libavcodec/utils.c since
side data is used to signal the skipped samples.
> h
ffmpeg | branch: master | Lynne | Sun May 3 21:17:33 2020
+0100| [bdd57e2a371f70ee75f70bfde5a9a162c76b48ba] | committer: Lynne
lavc/bsf: add an Opus metadata bitstream filter
The only adjustable field is the gain. Some ripping/transcoding programs
have started to use it.
> h
ffmpeg | branch: master | Lynne | Sun May 10 11:21:52 2020
+0100| [b69f5a72cec0f604b8160c5d18a56036522635b8] | committer: Lynne
hwcontext_vulkan: optionally enable the VK_KHR_surface extension if available
This allows any phys_device derived to be used as a display rendering device.
> h
ffmpeg | branch: master | Lynne | Wed Jun 5 04:53:18 2024
+0200| [8a2fe8a5b964e8ccb0cba642f7aefe256a0fb356] | committer: Lynne
mpeg4audio: rename AOT_USAC_NOSBR to AOT_USAC
The issue is that AOT 45 isn't defined anywhere, and looking at the git
blame, it seems to have sprung up thro
ffmpeg | branch: master | Lynne | Wed Jun 5 05:00:18 2024
+0200| [ee419804da2a6a44a4af5d949869f0e98306d2fc] | committer: Lynne
mpeg4audio: explicitly define each AOT
This makes it far easier to figure out which AOT belongs to which
profile.
Also, explicitly highlight the holes.
> h
ffmpeg | branch: master | Lynne | Wed Jun 5 18:23:12 2024
+0200| [5c328e6c1e56fdb12fcf3a89b18079ad983cbe5b] | committer: Lynne
aacdec: increase MAX_ELEM_ID to 64
In USAC, we set the max to 64.
> http://git.videolan.org/gitweb.cgi/ffmpeg.git/?a=commi
ffmpeg | branch: master | Lynne | Wed Jun 5 05:01:42 2024
+0200| [1c066867df5b786d486cfa1070af101b38542177] | committer: Lynne
aac: define a new profile for USAC
This allows users to determine whether a stream is USAC or not.
> http://git.videolan.org/gitweb.cgi/ffmpeg.git/?a=commi
ffmpeg | branch: master | Lynne | Wed Jun 5 05:04:58 2024
+0200| [91fd6ca000cb87302f5c689de50fcd626939a362] | committer: Lynne
lavc: bump minor and add APIchanges entry for new USAC profile
> http://git.videolan.org/gitweb.cgi/ffmpeg.git/?a=commit;h=91fd6ca000cb87302f5c689de50fcd626939a
ffmpeg | branch: master | Lynne | Wed Jun 5 18:25:00 2024
+0200| [62cd6d9e596fd4e4f5ade39de36203b547b5250b] | committer: Lynne
aacdec_usac: clean up nb_elems on error
Require that there is a valid layout with a valid number of channels
before accepting nb_elems.
The value is required when
ffmpeg | branch: master | Lynne | Wed Jun 5 19:40:54 2024
+0200| [25b848a0bd08a96f4d8e8242cd049c3295b730ef] | committer: Lynne
aacdec_usac: correctly set and use the layout map
> http://git.videolan.org/gitweb.cgi/ffmpeg.git/?a=commit;h=25b848a0bd08a96f4d8e8242cd049c3295b73
ffmpeg | branch: master | Lynne | Wed Jun 5 18:31:52 2024
+0200| [91ab17e2fe57d5c093f9a2c63342e58df6cf024b] | committer: Lynne
aacdec_usac: tag LFE channels as such in the channel map
Missed.
> http://git.videolan.org/gitweb.cgi/ffmpeg.git/?a=commit;h=91ab17e2fe57d5c093f9a2c63342e58df6cf0
ffmpeg | branch: master | Lynne | Thu Jun 6 02:25:03 2024
+0200| [8ecaa64b9bceb68d9107ba5a5e8d1dfaa40bf21f] | committer: Lynne
aacdec_usac: respect tns_on_lr flag
This was left out, and due to av_unused, forgotten about.
> http://git.videolan.org/gitweb.cgi/ffmpeg.git/?a=commi
ffmpeg | branch: master | Lynne | Wed Jun 5 18:43:08 2024
+0200| [ae495b56ffc9a80fffe608e5ef42720ea968952d] | committer: Lynne
aacdec_usac: remove fallback for custom maps with invalid position
Not needed as every possible index is mapped.
> http://git.videolan.org/gitweb.cgi/ffmpeg.git
ffmpeg | branch: master | Lynne | Thu Jun 6 03:50:44 2024
+0200| [1ad9a4008bbb4fbcf3691bc56851d877c955b9e4] | committer: Lynne
aacdec_usac: decouple TNS active from TNS data present flag
The issue was that in case of common TNS parameters, TNS was
entirely skipped, as tns.present was set to 0
ffmpeg | branch: master | Lynne | Thu Jun 6 04:38:47 2024
+0200| [a18d0659f403cf4b93a3c5a293a42e0ff7348e90] | committer: Lynne
aacdec_usac: skip coeff decoding if the number to be decoded is 0
Yet another thing not mentioned in the spec.
> http://git.videolan.org/gitweb.cgi/ffmpeg.git
ffmpeg | branch: master | Lynne | Thu Jun 6 02:51:51 2024
+0200| [c0fdb0cdfdbd9aff82a18402341827bf23c425f1] | committer: Lynne
aacdec_usac: do not continue parsing bitstream on core_mode == 1
Although LPD is not functional yet, the bitstream ends at that point.
> http://git.videolan.
ffmpeg | branch: master | Lynne | Thu Jun 6 17:23:47 2024
+0200| [9b41cc04300e8d00ae3a6326639e975712e21bb6] | committer: Lynne
aacdec_usac: do not round noise amplitude values
Use floating point division instead of integer division.
> http://git.videolan.org/gitweb.cgi/ffmpeg.git/?a=commi
ffmpeg | branch: master | Lynne | Thu Jun 6 17:33:18 2024
+0200| [722352333524cb2047d09e2548b495f511de4331] | committer: Lynne
aacdec_usac: use correct TNS values
The standard slightly modified the maximum TNS bands allowed.
> http://git.videolan.org/gitweb.cgi/ffmpeg.git/?a=commi
ffmpeg | branch: master | Lynne | Fri Jun 7 04:13:54 2024
+0200| [c2d459cb516d05deb388248f2972b456840c9022] | committer: Lynne
aacdec_usac: fix stereo alpha values for transients
Typo.
Also added comments and fixed the branch underneath.
> http://git.videolan.org/gitweb.cgi/ffmpeg.git
ffmpeg | branch: master | Lynne | Fri Jun 7 04:17:13 2024
+0200| [714596bcbf990ec3e82f8fde110954755cc1bf6e] | committer: Lynne
aacdec_usac: zero out alpha values for the current frame
> http://git.videolan.org/gitweb.cgi/ffmpeg.git/?a=commit;h=714596bcbf990ec3e82f8fde110954755cc1b
ffmpeg | branch: master | Lynne | Wed Jun 12 03:20:56 2024
+0200| [11a8e0a4e585c0a5acf09397e39bccf45e1ba7ae] | committer: Lynne
aacdec_usac: dequantize scalefactors after noise synthesis
The issue here is that the spec implied that the offset is done
on the dequantized scalefactor, but in fact
ffmpeg | branch: master | Lynne | Wed Jun 12 03:25:07 2024
+0200| [d79fbad366896873d9d4d219cce9578797944dea] | committer: Lynne
aacdec_usac: rename spectrum decode function and remove unused arg
The LC part of the decoder combines scalefactor application with
spectrum decoding, and this was
ffmpeg | branch: master | Lynne | Wed Jun 12 03:31:58 2024
+0200| [bdd3c6ca5015e29d12fa3d317a6d8b6958156a93] | committer: Lynne
aacdec_usac: always zero out alpha_q values for stereo streams
The issue is that if a frame has no complex stereo prediction,
the alpha values must all be assumed to
ffmpeg | branch: release/5.1 | Lynne | Mon Jan 1 00:00:00 2024
+| [68f2794354fda2675187fd9414532845fce2bd7a] | committer: Michael
Niedermayer
configure: update copyright year
(cherry picked from commit b95ee2ec5f84054de8bf6db9fe1b1119d569f269)
Signed-off-by: Michael Niedermayer
> h
ffmpeg | branch: release/4.3 | Lynne | Mon Jan 1 00:00:00 2024
+| [41a5eae142c8f00980ae6d58bf3cf8a869e5231a] | committer: Michael
Niedermayer
configure: update copyright year
(cherry picked from commit b95ee2ec5f84054de8bf6db9fe1b1119d569f269)
Signed-off-by: Michael Niedermayer
> h
ffmpeg | branch: master | Lynne | Sun Jun 16 10:28:05 2024
+0200| [d45e20c37b1144d9c4ff08732a94fee0786dc0b5] | committer: Lynne
aacdec_usac: remove custom rate_idx and use standard variable for it
m4ac.sampling_index is what aacdec.c uses.
> http://git.videolan.org/gitweb.cgi/ffmpeg.git
ffmpeg | branch: master | Lynne | Sun Jun 16 10:20:56 2024
+0200| [50e5b78b79f5b908e249ca5aca23be7be23bedad] | committer: Lynne
aac: expose ff_aac_sample_rate_idx() in aac.h
The rate index is a value important to both encoders and decoders.
USAC needs it as well, so put it into the shared main
ffmpeg | branch: master | Lynne | Sun Jun 16 10:22:25 2024
+0200| [89de2f0de1a41349fe827c00c8f52ca3c12594ad] | committer: Lynne
aacdec_ac: fix an overread
Fixes reading state->last[i + 1] in ff_aac_ac_get_context for the
last array member.
> http://git.videolan.org/gitweb.cgi/ffmpeg.
ffmpeg | branch: master | Lynne | Sun Jun 16 10:25:26 2024
+0200| [a381cbc7c77bc9477d9c42607a4e849aa20cdd06] | committer: Lynne
aacdec_usac: rename noise_scale to noise_bands
This was a typo.
> http://git.videolan.org/gitweb.cgi/ffmpeg.git/?a=commit;h=a381cbc7c77bc9477d9c42607a4e849aa20cd
ffmpeg | branch: master | Lynne | Wed Jun 12 14:32:22 2024
+0200| [1c3545f053bea8d919608f47d02bc095a006411f] | committer: Lynne
aacdec_usac: apply specification fix M55715
> http://git.videolan.org/gitweb.cgi/ffmpeg.git/?a=commit;h=1c3545f053bea8d919608f47d02bc095a006411f
---
libavcodec/
ffmpeg | branch: master | Lynne | Sun Jun 16 10:52:13 2024
+0200| [0b67c83b2eadf6350587ae7c4a63a8f9bba67cae] | committer: Lynne
aacdec_usac, aacsbr: implement SBR support for USAC
Currently, no eSBR features are supported.
Thankfully, no encoders exist for it yet.
> http://git.videolan.
ffmpeg | branch: master | Lynne | Sun Jun 23 15:58:28 2024
+0200| [dae12ddb2e5d4ec8cf5a965f4ac793edfa3802d6] | committer: Lynne
lavu/stereo3d: change the horizontal FOV field to a rational
This avoids hardcoding any implementation-specific limitiations as
part of the API, and allows for future
ffmpeg | branch: master | Lynne | Sun Jul 14 18:30:26 2024
+0200| [80ddc727178837a5ea5e6e463f28fedf6c1051dd] | committer: Lynne
vulkan: rename read_only to singular
There's nothing stopping users from writing to such buffers.
Its more accurate to say they are singular, i.e. not dupli
ffmpeg | branch: master | Lynne | Sun Jul 14 18:27:23 2024
+0200| [e11087b1629ed9df6c5810120b94597267684b17] | committer: Lynne
vulkan: set VkDescriptorAddressInfoEXT.sType
This was not done, resulting in validation issues, and potential
driver issues.
> http://git.videolan.org/gitweb.
ffmpeg | branch: master | Lynne | Mon Jul 22 03:20:32 2024
+0200| [b1b69ccbc0b2043e60b95735acced292413c44a5] | committer: Lynne
aacdec: set ac->output_elements upon channel element free
The issue is that ac->output_elements is populated from
ac->che, which may be freed, leaving
ffmpeg | branch: master | Lynne | Wed Jul 10 01:52:42 2024
+0200| [bedfabc4373d1cf84e75107c42a56657f2c26520] | committer: Lynne
vulkan: use the new queue family mechanism
> http://git.videolan.org/gitweb.cgi/ffmpeg.git/?a=commit;h=bedfabc4373d1cf84e75107c42a56657f2c26520
---
libavu
ffmpeg | branch: master | Lynne | Tue Jul 9 03:03:19 2024
+0200| [13489c8a2154a2e0e8fd3c3c45f7856b4c3110b0] | committer: Lynne
hwcontext_vulkan: add a new mechanism to expose used queue families
The issue with the old mechanism is that we had to introduce new
API each time we needed a new
ffmpeg | branch: master | Lynne | Tue Jul 16 17:14:42 2024
+0200| [8790a30882d9e48c20b61b3a7ac1ff3225bf649c] | committer: Lynne
hwcontext_vulkan: rewrite queue picking system for the new API
This allows us to support different video ops on different queues,
as well as any other arbitrary
ffmpeg | branch: master | Lynne | Tue Jul 16 17:16:34 2024
+0200| [a30b7c0158fe744678ca9e4c53b3f343d9f387ff] | committer: Lynne
hwcontext_vulkan: initialize optical flow queues if available
Lets us implement FPS conversion.
> http://git.videolan.org/gitweb.cgi/ffmpeg.git/?a=commi
ffmpeg | branch: master | Lynne | Thu Jul 18 10:12:09 2024
+0200| [6757cdb5350fdad381adfe42b305457897f86c8e] | committer: Lynne
vulkan_video: remove NIH pooled buffer implementation
The code predates ff_vk_get_pooled_buffer().
> http://git.videolan.org/gitweb.cgi/ffmpeg.git/?a=commi
ffmpeg | branch: master | Lynne | Thu Jul 18 10:16:27 2024
+0200| [81c5d4ea0e5e90bc030d31ab9e7d0cf090988b96] | committer: Lynne
hwcontext_vulkan: remove unused struct
> http://git.videolan.org/gitweb.cgi/ffmpeg.git/?a=commit;h=81c5d4ea0e5e90bc030d31ab9e7d0cf090988b96
---
libavu
ffmpeg | branch: master | Lynne | Fri Jul 19 00:10:06 2024
+0200| [aea4d4b423c62aecf326ef3ae1578710faa3eca6] | committer: Lynne
hwcontext_vulkan: rewrite upload/download
This commit was long overdue. The old transfer dubiously tried to
merge as much code as possible, and had very little in the
ffmpeg | branch: master | Lynne | Fri Aug 9 01:42:06 2024
+0200| [9e606b33a8e25fa8df383bcf3229e0203f79fae3] | committer: Lynne
hwcontext_vulkan: add HOST_CACHED flag to transfer buffer
Significantly speeds up downloads on devices without host mapping.
> http://git.videolan.org/gitweb.
ffmpeg | branch: master | Lynne | Wed Jul 24 11:40:05 2024
+0200| [c19af16f8db4eac1d603d0b8e439c03946a262b7] | committer: Lynne
hwcontext_vulkan: enable storageBuffer16BitAccess if available
> http://git.videolan.org/gitweb.cgi/ffmpeg.git/?a=commit;h=c19af16f8db4eac1d603d0b8e439c03946a26
ffmpeg | branch: master | Lynne | Wed Jul 24 00:53:28 2024
+0200| [2f7dfb0d157b680bca68bc9e57f7c3529fbfc0fe] | committer: Lynne
vulkan_shaderc: fix error reporting for certain errors
The issue is that shaderc_result_get_num_errors may sometime
return 0 even when
ffmpeg | branch: master | Lynne | Wed Jul 24 20:19:15 2024
+0200| [8eac11105b0addc6f9ff53d4f082171db459e172] | committer: Lynne
vulkan: use allocator callback for buffer creation
This would've let to a segfault if custom allocators were used.
> http://git.videolan.org/gitweb.cgi/ffmpe
ffmpeg | branch: master | Lynne | Wed Jul 24 00:39:31 2024
+0200| [957d34784ad98cb342cd8b23f114b17094f92f5c] | committer: Lynne
hwcontext_vulkan: constify validation layer features table
The struct data seem to get corrupted otherwise.
Possibly a validation layer or libvulkan issue.
> h
ffmpeg | branch: master | Lynne | Wed Jul 24 11:44:06 2024
+0200| [12f868cab558af8e32829eb5ed38189ce9b93f72] | committer: Lynne
vulkan_shaderc: add debug information to shaders
> http://git.videolan.org/gitweb.cgi/ffmpeg.git/?a=commit;h=12f868cab558af8e32829eb5ed38189ce9b93
ffmpeg | branch: master | Lynne | Wed Jul 24 12:58:34 2024
+0200| [55adcb4fc516eb3f14a24a497c81dd1e39e42777] | committer: Lynne
hwcontext_vulkan: add support for VK_EXT_shader_object
We'd like to use it eventually, and its already covered by
the minimum version of the headers we re
ffmpeg | branch: master | Lynne | Sat Aug 10 17:13:57 2024
+0200| [ca591e6b50ee616fc36711cc66d594e51209dddb] | committer: Lynne
vulkan_decode: force layered_dpb to 0 when dedicated_dpb is 0
layered_dpb only makes sense when dedicated_dpb is set to 1.
For some mysterious reason, some Nvidia
ffmpeg | branch: master | Lynne | Sun Aug 11 01:01:35 2024
+0200| [0b25f0bc1d17a7759120e0ac1bcf901258a93cf9] | committer: Lynne
hwcontext_vulkan: correct comment in header
> http://git.videolan.org/gitweb.cgi/ffmpeg.git/?a=commit;h=0b25f0bc1d17a7759120e0ac1bcf901258a93cf9
---
libavu
ffmpeg | branch: master | Lynne | Wed Dec 28 06:31:11 2022
+0100| [2ce0e51503d9178b190660ede8e716f6a134c2ec] | committer: Lynne
hwcontext_vulkan: add support for Vulkan encoding
> http://git.videolan.org/gitweb.cgi/ffmpeg.git/?a=commit;h=2ce0e51503d9178b190660ede8e716f6a134c2ec
---
libavu
ffmpeg | branch: master | Lynne | Sat Aug 10 17:46:04 2024
+0200| [5f0f1f7b7a6e6dde16101f919c18de710a51ca64] | committer: Lynne
libavutil: deprecate the old Vulkan queue API, add doc/APIchanges entries
> http://git.videolan.org/gitweb.cgi/ffmpeg.git/?a=commi
ffmpeg | branch: master | Lynne | Sat Aug 10 21:51:32 2024
+0200| [d6c08a41cb576ad2e160761a0bfd44cf9e3b6232] | committer: Lynne
vulkan: load queue families upon loading properties
Avoids the need to call ff_vk_qf_init if manually filling in
a queue family structure.
> http://git.videolan.
ffmpeg | branch: master | Lynne | Sun Aug 11 03:27:46 2024
+0200| [680d969a305c0927480573a1b455024088b51aeb] | committer: Lynne
vulkan_decode: port to the new queue family API
> http://git.videolan.org/gitweb.cgi/ffmpeg.git/?a=commit;h=680d969a305c0927480573a1b455024088b51aeb
---
libavco
ffmpeg | branch: master | Lynne | Sat Aug 10 18:20:58 2024
+0200| [1c05661ec4e83aecd1b07037e5cff6100234a3b6] | committer: Lynne
vulkan_decode: add \n to error message
> http://git.videolan.org/gitweb.cgi/ffmpeg.git/?a=commit;h=1c05661ec4e83aecd1b07037e5cff6100234a3b6
---
libavco
ffmpeg | branch: master | Lynne | Fri Aug 9 13:36:01 2024
+0200| [83cd77563fe008e34b70dcf1dd4d94c048dbf4fa] | committer: Lynne
vulkan: add support for encode feedback queries
> http://git.videolan.org/gitweb.cgi/ffmpeg.git/?a=commit;h=83cd77563fe008e34b70dcf1dd4d94c048dbf4fa
---
libavu
ffmpeg | branch: master | Lynne | Sun Aug 11 04:27:20 2024
+0200| [12080ff0404059847df8bc30cc5023a9010f60b2] | committer: Lynne
nlmeans_vulkan: fix uninitialized reads
> http://git.videolan.org/gitweb.cgi/ffmpeg.git/?a=commit;h=12080ff0404059847df8bc30cc5023a9010f60b2
---
libavfil
ffmpeg | branch: master | Lynne | Sun Aug 11 04:27:52 2024
+0200| [ef11a6456d40de57bbda9369ecedf5f242020703] | committer: Lynne
hwcontext_vulkan: do not chain structs of unsupported extensions in
vkCreateDevice
Fixes:
vkCreateDevice(): pCreateInfo->pNext
includes a
pointer t
ffmpeg | branch: master | Lynne | Sun Aug 11 05:01:43 2024
+0200| [e25667f9f14fa71d655265905fa91aa39953a17e] | committer: Lynne
hwcontext_vulkan: ignore false positive validation errors
Issue ref:
https://github.com/KhronosGroup/Vulkan-ValidationLayers/issues/6627
> http://git.videolan.
ffmpeg | branch: master | Lynne | Tue Aug 13 19:03:43 2024
+0200| [d138d7a595b7ff3aa1808ad36e5695a5546b12ec] | committer: Lynne
vulkan: make sure descriptor buffers are always DEVICE_LOCAL
Implementations are required to list memory heaps in the most optimal
order. But its better to be
ffmpeg | branch: master | Lynne | Wed Aug 14 14:22:28 2024
+0200| [869f4aec484be200ea676d7d45c5b12c9b914434] | committer: Lynne
vulkan_decode: use the correct queue family for decoding ops
In 680d969a305c0927480573a1b455024088b51aeb, the new API was
used to find a queue family for dispatch
1 - 100 of 801 matches
Mail list logo
