This patchset contains arm64 neon implementation of hscale functions.
Fixed minor style issues and declared C function wrappers as static.
This patchset do not contain the patch for checkasm tool, as the
previous one did. The reason behind it was failing tests on x86 arch
but not on aarch64 or loon
Add arm64 neon implementations for hscale 8 to 19 with filter
sizes 4, 4X and 8. Both implementations are based on very similar ones
dedicated to hscale 8 to 15. The major changes refer to saving
the data - instead of writing the result as int16_t it is done
with int32_t.
These functions are heavi
Add arm64 neon implementations for hscale 16 to 15 with filter
sizes 4, 8 and X4.
The tests and benchmarks run on AWS Graviton 2 instances.
The results from a checkasm tool are shown below.
hscale_16_to_15__fs_4_dstW_512_c: 6703.5
hscale_16_to_15__fs_4_dstW_512_neon: 2298.0
hscale_16_to_15__fs_8_
Provide arm64 neon optimized implementations for hscale16To19 with
filter sizes 4, 8 and X4.
The tests and benchmarks run on AWS Graviton 2 instances.
The results from a checkasm tool are shown below.
hscale_16_to_19__fs_4_dstW_512_c: 6216.0
hscale_16_to_19__fs_4_dstW_512_neon: 2257.0
hscale_16_t
Set the correct amount of tags in tags_per_config[] and ensure the side
channels (Surround in the spec) are used.
Signed-off-by: James Almer
---
libavcodec/aacdec_template.c | 4 +---
libavcodec/aacdectab.h | 4 ++--
2 files changed, 3 insertions(+), 5 deletions(-)
diff --git a/libavcodec
Signed-off-by: James Almer
---
doc/APIchanges| 3 +++
doc/utils.texi| 2 ++
libavutil/channel_layout.c| 1 +
libavutil/channel_layout.h| 2 ++
tests/ref/fate/channel_layout | 1 +
5 files changed, 9 insertions(+)
diff --git a/doc/APIchanges b/doc/APIchange
It corresponds to the 7.1(top) layout.
Signed-off-by: James Almer
---
libavcodec/aacdec_template.c | 23 ++-
libavcodec/aacdectab.h | 6 +++---
libavcodec/mpeg4audio.c | 5 +++--
libavcodec/mpeg4audio.h | 2 +-
4 files changed, 29 insertions(+), 7 deletions
Signed-off-by: James Almer
---
libavformat/mov_chan.c | 5 +
libavformat/mov_chan.h | 2 ++
2 files changed, 7 insertions(+)
diff --git a/libavformat/mov_chan.c b/libavformat/mov_chan.c
index 5b757c6a8a..19da01996c 100644
--- a/libavformat/mov_chan.c
+++ b/libavformat/mov_chan.c
@@ -186,6 +1
James Almer:
> It corresponds to the 7.1(top) layout.
>
> Signed-off-by: James Almer
> ---
> libavcodec/aacdec_template.c | 23 ++-
> libavcodec/aacdectab.h | 6 +++---
> libavcodec/mpeg4audio.c | 5 +++--
> libavcodec/mpeg4audio.h | 2 +-
> 4 files changed
On 10/28/2022 9:31 AM, Andreas Rheinhardt wrote:
James Almer:
It corresponds to the 7.1(top) layout.
Signed-off-by: James Almer
---
libavcodec/aacdec_template.c | 23 ++-
libavcodec/aacdectab.h | 6 +++---
libavcodec/mpeg4audio.c | 5 +++--
libavcodec/mpeg
James Almer:
> On 10/28/2022 9:31 AM, Andreas Rheinhardt wrote:
>> James Almer:
>>> It corresponds to the 7.1(top) layout.
>>>
>>> Signed-off-by: James Almer
>>> ---
>>> libavcodec/aacdec_template.c | 23 ++-
>>> libavcodec/aacdectab.h | 6 +++---
>>> libavcodec/mpeg
On 10/28/2022 9:50 AM, Andreas Rheinhardt wrote:
James Almer:
On 10/28/2022 9:31 AM, Andreas Rheinhardt wrote:
James Almer:
It corresponds to the 7.1(top) layout.
Signed-off-by: James Almer
---
libavcodec/aacdec_template.c | 23 ++-
libavcodec/aacdectab.h | 6
Andreas Rheinhardt:
> Only encoders need two sets of int16_t [12][64]
> (one to save the current best state and one for the current
> working state); decoders need only one. This saves 1.5KiB
> per slice context for a decoder.
>
> Signed-off-by: Andreas Rheinhardt
> ---
> libavcodec/mpegvideo.c
These tables are only accessed in ff_set_qscale()
which only accesses values 1..31 as well as in
encode_picture() in mpegvideo_enc.c, accessing
the value with index 8. So make these tables smaller.
Signed-off-by: Andreas Rheinhardt
---
libavcodec/mpegvideodata.c | 32
Avoids relocations.
Signed-off-by: Andreas Rheinhardt
---
libavcodec/mpeg12enc.c | 2 +-
libavcodec/mpegvideo_enc.c | 4 ++--
libavcodec/mpegvideodata.c | 26 +-
libavcodec/mpegvideodata.h | 4 ++--
libavcodec/speedhqenc.c| 2 +-
5 files changed, 15 insertions
This e.g. allows compilers to bake the offset implied
by using ff_mpeg12_dc_scale_table[3] (as the SpeedHQ encoder
does) into the general offset; for certain arches this is
also necessary in order to avoid building suboptimal code.
Signed-off-by: Andreas Rheinhardt
---
libavcodec/mpegvideodata.h
ff_mpeg1_dc_scale_table is the default value for
[yc]_dc_scale_table (as set by ff_mpv_common_defaults()).
Signed-off-by: Andreas Rheinhardt
---
libavcodec/flvdec.c | 2 --
libavcodec/intelh263dec.c | 3 ---
2 files changed, 5 deletions(-)
diff --git a/libavcodec/flvdec.c b/libavcodec/flv
low_delay is always zero for rv34.
Signed-off-by: Andreas Rheinhardt
---
libavcodec/rv34.c | 4 ++--
1 file changed, 2 insertions(+), 2 deletions(-)
diff --git a/libavcodec/rv34.c b/libavcodec/rv34.c
index acf77d103d..be188edc47 100644
--- a/libavcodec/rv34.c
+++ b/libavcodec/rv34.c
@@ -1568,7
This e.g. allows compilers to bake the offset implied
by using ff_vc1_b_field_mvpred_scales[3] into the
general offset; for certain arches this is also necessary
in order to avoid building suboptimal code.
Signed-off-by: Andreas Rheinhardt
---
libavcodec/vc1data.h | 3 +++
1 file changed, 3 inse
---
.mailmap | 1 -
1 file changed, 1 deletion(-)
diff --git a/.mailmap b/.mailmap
index ba072f38c8..af60290f77 100644
--- a/.mailmap
+++ b/.mailmap
@@ -1,4 +1,3 @@
-
--
2.38.0
___
ffmpeg-devel mailing list
ffmpeg-devel@ffmpeg.org
https://ff
---
libavutil/tests/cpu.c | 1 +
1 file changed, 1 insertion(+)
diff --git a/libavutil/tests/cpu.c b/libavutil/tests/cpu.c
index 5bec742b2b..dadadb31dc 100644
--- a/libavutil/tests/cpu.c
+++ b/libavutil/tests/cpu.c
@@ -77,6 +77,7 @@ static const struct {
{ AV_CPU_FLAG_BMI2, "bmi2"
Hi Zane et al.,
Quick ping on the revised patchset below.
It addresses https://trac.ffmpeg.org/ticket/9842.
Best,
-- Pierre
On Sun, Oct 2, 2022 at 9:28 AM wrote:
>
> From: Pierre-Anthony Lemieux
>
> The IMF CPL contains an optional timecode start address. This patch reads the
> latter, if pr
---
tests/checkasm/checkasm.c | 1 +
tests/checkasm/checkasm.h | 1 +
2 files changed, 2 insertions(+)
diff --git a/tests/checkasm/checkasm.c b/tests/checkasm/checkasm.c
index 421bd096c5..c3d77cb6af 100644
--- a/tests/checkasm/checkasm.c
+++ b/tests/checkasm/checkasm.c
@@ -918,5 +918,6 @@ int che
Negligible speed difference for avx2 on Zen 2 (Ryzen 5700X) and
Broadwell (Xeon E5-2620 v4):
1690±4.3 decicycles vs. 1693±78.4
1439±31.1 decicycles vs 1429±16.7
Moderate speedup with avx512 on Skylake-X (Xeon D-2123IT):
1.22x faster (793±0.8 vs. 649±5.5 decicycles) compared with avx2
Bett
On 28.10.2022 20:23, Michael Niedermayer wrote:
Hi
According to our
https://trac.ffmpeg.org/wiki/Downstreams
Noone and nothing is using 5.0
should i make another release of 5.0 ?
should i move 5.0 to olddownloads ?
does anyone use it ? plan to use it or know of someone using it ?
thx
I see
Op di 11 okt. 2022 om 19:24 schreef Martijn van Beurden :
>
> Recently libFLAC gained the ability (first released in FLAC 1.4.0)
> to create FLAC files containing 32-bit int PCM samples. To
> keep complexity reasonable, the choice was made to limit residuals
> to 32-bit integers, which the encoder
+1 for moving it to the old downloads
Best regards, Reto
___
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 "unsu
> -Original Message-
> From: ffmpeg-devel On Behalf Of
> Soft Works
> Sent: Tuesday, October 25, 2022 11:59 AM
> To: FFmpeg development discussions and patches de...@ffmpeg.org>
> Subject: Re: [FFmpeg-devel] [PATCH v9 02/25] avutil/frame: Prepare
> AVFrame for subtitle handling
>
>
>
Oct 27, 2022, 18:45 by mich...@niedermayer.cc:
> On Tue, Oct 25, 2022 at 01:17:15PM +0200, Lynne wrote:
>
>>
>>
>>
>> Oct 24, 2022, 18:29 by jamr...@gmail.com:
>>
>> > On 10/24/2022 12:56 PM, Lynne wrote:
>> >
>> >> Oct 24, 2022, 09:42 by d.kozin...@samsung.com:
>> >>
>> >>> - Changelog update
>>
This patch replaces the transform used in AAC with lavu/tx and removes
the limitation on only being able to decode 960-sample files
with the float decoder.
This commit also removes a whole bunch of unnecessary and slow
lifting steps the decoder did to compensate for the poor accuracy
of the old int
On Fri, 28 Oct 2022 at 19:57, James Darnley wrote:
> Negligible speed difference for avx2 on Zen 2 (Ryzen 5700X) and
> Broadwell (Xeon E5-2620 v4):
> 1690±4.3 decicycles vs. 1693±78.4
> 1439±31.1 decicycles vs 1429±16.7
>
Just to avoid confusion for anyone who decides to review this over
This also fixes not checking the return values on transform init.
Total decoder speedup on Zen 3: 9%
Patch attached.
>From efe3006093cd80182b293f01aa98fb75733a8188 Mon Sep 17 00:00:00 2001
From: Lynne
Date: Sat, 29 Oct 2022 08:30:56 +0200
Subject: [PATCH] vorbisdec: convert to lavu/tx
This als
32 matches
Mail list logo
