The following command is on how to apply vflip_vulkan filter:
ffmpeg -init_hw_device vulkan -i input.264 -vf
hwupload=extra_hw_frames=16,vflip_vulkan,hwdownload,format=yuv420p output.264
Signed-off-by: Wu Jianhua
---
configure | 1 +
libavfilter/Makefile | 1 +
The following command is on how to apply hflip_vulkan filter:
ffmpeg -init_hw_device vulkan -i input.264 -vf
hwupload=extra_hw_frames=16,hflip_vulkan,hwdownload,format=yuv420p output.264
Signed-off-by: Wu Jianhua
---
configure | 1 +
libavfilter/Makefile | 1 +
17 Nov 2021, 06:04 by lance.lmw...@gmail.com:
> From: Limin Wang
>
> Signed-off-by: Limin Wang
> ---
> libavfilter/vf_gblur_vulkan.c | 2 +-
>
Pushed, thanks.
___
ffmpeg-devel mailing list
ffmpeg-devel@ffmpeg.org
https://ffmpeg.org/mailman/listinfo/f
From: Limin Wang
Signed-off-by: Limin Wang
---
libavfilter/vf_gblur_vulkan.c | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/libavfilter/vf_gblur_vulkan.c b/libavfilter/vf_gblur_vulkan.c
index d4f391b..16c8bbb 100644
--- a/libavfilter/vf_gblur_vulkan.c
+++ b/libavfilter/vf_g
Signed-off-by: J. Dekker
---
libavcodec/aarch64/hevcdsp_sao_neon.S | 195 --
1 file changed, 90 insertions(+), 105 deletions(-)
Now matches the 9,25 indentation like other ASM.
diff --git a/libavcodec/aarch64/hevcdsp_sao_neon.S
b/libavcodec/aarch64/hevcdsp_sao_neon.S
i
--bench on AWS Graviton:
hevc_sao_band_8x8_8_c: 317.5
hevc_sao_band_8x8_8_neon: 97.5
hevc_sao_band_16x16_8_c: 1115.0
hevc_sao_band_16x16_8_neon: 322.7
hevc_sao_band_32x32_8_c: 4599.2
hevc_sao_band_32x32_8_neon: 1246.2
hevc_sao_band_48x48_8_c: 10021.7
hevc_sao_band_48x48_8_neon: 2740.5
hevc_sao_ban
--bench on AWS Graviton:
hevc_sao_edge_8x8_8_c: 516.0
hevc_sao_edge_8x8_8_neon: 81.0
Signed-off-by: J. Dekker
---
libavcodec/aarch64/hevcdsp_init_aarch64.c | 3 ++
libavcodec/aarch64/hevcdsp_sao_neon.S | 51 +++
2 files changed, 54 insertions(+)
Used inline shift for
--bench on AWS Graviton:
hevc_sao_edge_16x16_8_c: 1857.0
hevc_sao_edge_16x16_8_neon: 211.0
hevc_sao_edge_32x32_8_c: 7802.2
hevc_sao_edge_32x32_8_neon: 808.2
hevc_sao_edge_48x48_8_c: 16764.2
hevc_sao_edge_48x48_8_neon: 1796.5
hevc_sao_edge_64x64_8_c: 32647.5
hevc_sao_edge_64x64_8_neon: 3118.5
Sign
This reverts commit 2589060b92eeeb944c6e2b50e38412c0c5fabcf4 which was
originally to fix the FATE test. The real cause of the test breakage was
fixed in 8dc8f04036eb27c8ad419839d4ed3bc67c44fe7a.
Signed-off-by: J. Dekker
---
libavcodec/arm/hevcdsp_qpel_neon.S | 18 +-
1 file chang
The assembly is written assuming that the width is a multiple of 8.
However the real issue is the functions were errorneously assigned to
the 2, 4, 6 & 12 widths. This behaviour never broke the decoder as
samples which trigger the functions for these widths have not been found
in the wild. This re
17 Nov 2021, 04:04 by jianhua...@intel.com:
> Whether failed or not, the block of codes labeled with fail should
> be always run to ensure the av_free(kernel_def) is executed.
>
> Signed-off-by: Wu Jianhua
> ---
> libavfilter/vf_gblur_vulkan.c | 2 --
> 1 file changed, 2 deletions(-)
>
> diff --
Whether failed or not, the block of codes labeled with fail should
be always run to ensure the av_free(kernel_def) is executed.
Signed-off-by: Wu Jianhua
---
libavfilter/vf_gblur_vulkan.c | 2 --
1 file changed, 2 deletions(-)
diff --git a/libavfilter/vf_gblur_vulkan.c b/libavfilter/vf_gblur_vu
On Tue, Nov 16, 2021 at 08:11:41PM +0100, Anton Khirnov wrote:
> Quoting Michael Niedermayer (2021-11-16 14:46:43)
> > On Mon, Nov 15, 2021 at 04:23:59PM +0100, Anton Khirnov wrote:
> > > bits_per_raw_sample is already set in new_video_stream(), so this code
> > > has no effect.
> > > ---
> > > ff
Quoting Michael Niedermayer (2021-11-16 14:46:43)
> On Mon, Nov 15, 2021 at 04:23:59PM +0100, Anton Khirnov wrote:
> > bits_per_raw_sample is already set in new_video_stream(), so this code
> > has no effect.
> > ---
> > fftools/ffmpeg.c | 7 ---
> > fftools/ffmpeg.h | 1 -
> > fftools
On Wed, Nov 3, 2021 at 9:28 PM wrote:
> From: Mark Reid
>
> changes since v1:
> * remove vex intruction on sse4 path
> * some load/pack marcos use less intructions
> * fixed some typos
>
> yuv2gbrp_full_X_4_512_c: 12757.6
> yuv2gbrp_full_X_4_512_sse2: 8946.6
> yuv2gbrp_full_X_4_512_sse4: 5138.6
On 11/16/2021 2:57 PM, James Almer wrote:
> Micro.
>
> LGTM otherwise if tested.
Fixed locally.
- Derek
___
ffmpeg-devel mailing list
ffmpeg-devel@ffmpeg.org
https://ffmpeg.org/mailman/listinfo/ffmpeg-devel
To unsubscribe, visit link above, or email
f
On 11/16/2021 11:52 AM, Derek Buitenhuis wrote:
In 45bfe8b838275235412777dd430206d9a24eb3ee, short_seek_threshold was removed
from the public AVIO struct. Although this option was private and not intended
to be used by public API users, it was nonetheless, because it provided
functionality
that
On 11/16/2021 3:30 AM, Ridley Combs wrote:
> Because it was faster to write a trivial implementation than to track down an
> existing one.
Not a good reason...
> Even now that I've found a couple, one's a static in videotoolboxenc and is
> somewhat heavier than is warranted here, and one's in
In 45bfe8b838275235412777dd430206d9a24eb3ee, short_seek_threshold was removed
from the public AVIO struct. Although this option was private and not intended
to be used by public API users, it was nonetheless, because it provided
functionality
that could otherwise not be gained via public API.
Thi
On Tue, Nov 16, 2021 at 1:57 PM Anton Khirnov wrote:
> Quoting mindm...@gmail.com (2021-11-15 07:22:20)
> > From: Mark Reid
> >
> > Changes av_clipf to return amin if a is nan.
> > Before if a is nan av_clipf_c returned nan and
> > av_clipf_sse would return amax. Now the both
> > should behave t
As per 8.6.1.2.2 of ISO/IEC 14496-12:2015(E), STTS sample offsets
are to be always stored as uint32_t. So far, they have been signed ints
which led to desync in files with very large offsets.
The MOVStts struct was used to store CTTS offsets as well. These can be
negative in version 1. So a new st
On Mon, Nov 15, 2021 at 04:23:59PM +0100, Anton Khirnov wrote:
> bits_per_raw_sample is already set in new_video_stream(), so this code
> has no effect.
> ---
> fftools/ffmpeg.c | 7 ---
> fftools/ffmpeg.h | 1 -
> fftools/ffmpeg_opt.c | 2 +-
> 3 files changed, 1 insertion(+), 9 delet
From: Limin Wang
Signed-off-by: Limin Wang
---
Changelog | 1 +
libavcodec/Makefile| 1 +
libavcodec/allcodecs.c | 1 +
libavcodec/s210enc.c | 90 ++
libavcodec/version.h | 2 +-
5 files changed, 94 insertions(+), 1 deleti
Hi,
RFC: gaussian_blur() has fixed 5x5 kernel with sigma 1.4. Should this also go
into a common place with patch 1/2 or somewhere else?
-Thilo
From cb9f4caee52f74942606c733308f11e51bbca649 Mon Sep 17 00:00:00 2001
From: Thilo Borgmann
Date: Tue, 16 Nov 2021 13:37:58 +0100
Subject: [PATCH 2/2] l
Quoting mindm...@gmail.com (2021-11-15 07:22:20)
> From: Mark Reid
>
> Changes av_clipf to return amin if a is nan.
> Before if a is nan av_clipf_c returned nan and
> av_clipf_sse would return amax. Now the both
> should behave the same.
>
> This works because nan > amin is false.
> The max(nan,
Hi,
shared code with vf_blurriness.c in patch 2/2.
RFC: is there a better place to put these?
-Thilo
From b4e3701b2ed034b8d60aeff5d5ed75e4d4332a9f Mon Sep 17 00:00:00 2001
From: Thilo Borgmann
Date: Tue, 16 Nov 2021 13:35:02 +0100
Subject: [PATCH 1/2] lafi/vf_edgedetect: Move some functions to l
From: Limin Wang
Signed-off-by: Limin Wang
---
Changelog| 1 +
libavformat/Makefile | 1 +
libavformat/rtpenc.c | 4 ++
libavformat/rtpenc.h | 1 +
libavformat/rtpenc_rfc4175.c | 116 +++
libavformat/sdp.c
From: Limin Wang
Signed-off-by: Limin Wang
---
Changelog| 1 +
libavformat/Makefile | 1 +
libavformat/rtpenc.c | 4 ++
libavformat/rtpenc.h | 1 +
libavformat/rtpenc_rfc4175.c | 116 +++
libavformat/sdp.c
> +#if (GNU_PROPERTY_AARCH64_BTI != 0)
> +.pushsection .note.gnu.property, "a";
> +.balign 8;
> +.long 4;
> +.long 0x10;
> +.long 0x5;
> +.asciz "GNU";
> +.long 0xc000; /* GNU_PROPERTY_AARCH64_FEATURE_1_AND */
> +.long 4;
> +.long GNU_PROPERTY_AARCH64_BTI;
no major comments, so set pushed
--
Anton Khirnov
___
ffmpeg-devel mailing list
ffmpeg-devel@ffmpeg.org
https://ffmpeg.org/mailman/listinfo/ffmpeg-devel
To unsubscribe, visit link above, or email
ffmpeg-devel-requ...@ffmpeg.org with subject "unsubscrib
16 Nov 2021, 08:35 by jianhua...@intel.com:
> This commit adds a powerful and customizable gblur Vulkan filter,
> which provides a maximum 127x127 kernel size of Gaussian Filter.
> The size could be adjusted by requirements on quality or performance.
>
> The following command is on how to apply gb
On Sun, 7 Nov 2021 at 4:38 am, Dylan Fernando wrote:
>
>
> On Sat, Nov 6, 2021 at 10:34 AM Dylan Fernando
> wrote:
>
>> nlmeans CUDA filter
>>
>>
>>
>> removed query_formats
>>
> Anybody have any feedback for this?
___
ffmpeg-devel mailing list
ffmpeg-
On Fri, 12 Nov 2021, Jonathan Wright wrote:
Hi,
This patch adds Branch Target Identifiers (BTIs) to all functions defined
in AArch64 assembly files. Most of the BTI landing pads are added
automatically by the 'function' macro.
BTI support is turned on or off at compile time based on the presen
will apply immediately
___
ffmpeg-devel mailing list
ffmpeg-devel@ffmpeg.org
https://ffmpeg.org/mailman/listinfo/ffmpeg-devel
To unsubscribe, visit link above, or email
ffmpeg-devel-requ...@ffmpeg.org with subject "unsubscribe".
On Fri, 12 Nov 2021, Jonathan Wright wrote:
Hi,
This patch changes all AArch64 assembly code to use:
ret x
instead of:
br x
"ret x" is already used in a lot of places so this patch makes it
consistent across the code base. This does not change behavior or
performance.
In addition,
From: nyanmisaka
mfxHDLPair was added to qsv, so modify qsv->opencl map function as well.
Now the following commandline works:
ffmpeg -v verbose -init_hw_device vaapi=va:/dev/dri/renderD128 \
-init_hw_device qsv=qs@va -init_hw_device opencl=ocl@va -filter_hw_device ocl \
-hwaccel qsv -hwaccel_ou
From: nyanmisaka
The data stored in data[3] in VAAPI AVFrame is VASurfaceID while
the data stored in pair->first is the pointer of VASurfaceID, so
we need to do cast to make following commandline works:
ffmpeg -hwaccel vaapi -hwaccel_device /dev/dri/renderD128 \
-hwaccel_output_format vaapi -i i
For vaapi if the init_pool_size is not zero, the pool size is fixed.
This means max surfaces is init_pool_size, but when mapping vaapi
frame to qsv frame, the init_pool_size < nb_surface. The cause is that
vaapi_decode_make_config() config the init_pool_size and it is called
twice. The first time i
38 matches
Mail list logo