Signed-off-by: Josh Dekker
---
configure | 2 +
libavutil/Makefile| 1 +
libavutil/macos_kperf.c | 140 ++
libavutil/macos_kperf.h | 23 +++
libavutil/timer.h | 17 -
tests/checkasm/checkasm.c | 14 +++-
tests
This is a patch originally, submitted in 2017 (author/date info left
intact). At the time, it didn't get much attention I assume due to the
sheer size of it. I have split the patch into only its QPEL/EPEL parts,
rebasing, and doing some cleaning of the patches as much is reasonable
for a 9001 line
Set applied.
--
Josh
___
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".
Signed-off-by: Josh Dekker
---
configure| 2 +
tests/checkasm/Makefile | 1 +
tests/checkasm/checkasm.c| 19 -
tests/checkasm/checkasm.h| 10 ++-
tests/checkasm/macos_kperf.c | 143 +++
tests/checkasm/macos_kperf.h | 23
Set pushed with all Martin's changes implemented. More NEON & updates soon.
--
Josh
On 2021-02-04 12:32, Josh Dekker wrote:
Hi,
Rebases the unpushed part of my patches on top of Reimar's set. Also
implements Martin's suggestions except 'unrolling the loop' for S
Only works for 8x8.
Signed-off-by: Josh Dekker
---
libavcodec/aarch64/Makefile | 3 +-
libavcodec/aarch64/hevcdsp_init_aarch64.c | 7 ++
libavcodec/aarch64/hevcdsp_sao_neon.S | 87 +++
3 files changed, 96 insertions(+), 1 deletion(-)
create mode 100644
the first 300 frames of "LG 4K HDR Demo - New York.ts",
running on Apple M1.
Signed-off-by: Josh Dekker
---
libavcodec/aarch64/Makefile | 2 +
libavcodec/aarch64/hevcdsp_idct_neon.S| 380 ++
libavcodec/aarch64/hevcdsp_init_aarch64.c | 45 +++
From: Reimar Döffinger
Speedup is fairly small, around 1.5%, but these are fairly simple.
Signed-off-by: Josh Dekker
---
libavcodec/aarch64/hevcdsp_idct_neon.S| 190 ++
libavcodec/aarch64/hevcdsp_init_aarch64.c | 24 +++
2 files changed, 214 insertions(+)
diff --git
Signed-off-by: Josh Dekker
---
libavcodec/aarch64/hevcdsp_idct_neon.S| 54 +++
libavcodec/aarch64/hevcdsp_init_aarch64.c | 16 +++
2 files changed, 70 insertions(+)
diff --git a/libavcodec/aarch64/hevcdsp_idct_neon.S
b/libavcodec/aarch64/hevcdsp_idct_neon.S
index
Hi,
Rebases the unpushed part of my patches on top of Reimar's set. Also
implements Martin's suggestions except 'unrolling the loop' for SAO band
function, will update the band function when I fix non 8x8 cases.
--
Josh
___
ffmpeg-devel mailing list
On 2021-01-13 17:06, Robin Cooksey wrote:
I’ve attached a patch which makes avformat handle the 308 Permanent Redirect
HTTP status code – which is more recently defined in
https://tools.ietf.org/html/rfc7538
The change just treats 308 in the same way as the other 30x status codes.
Thanks. Ap
On 2021-01-07 13:10, Josh Dekker wrote:
Co-authored-by: Niklas Haas
Signed-off-by: Josh Dekker
---
tests/checkasm/Makefile | 2 +-
tests/checkasm/checkasm.c | 10 +
tests/checkasm/checkasm.h | 10 +
tests/checkasm/hevc_pel.c | 523 ++
4 files
On 2021-01-23 14:14, Martin Storsjö wrote:
On Sat, 23 Jan 2021, Reimar Döffinger wrote:
Setting the defaults for $arch happens only later, so
the current code would not set AS correctly if --arch
was not specified on the command-line.
Fix it by adding an explicit fallback to $arch_default.
---
Hi,
On 2021-01-08 21:36, reimar.doeffin...@gmx.de wrote:
From: Reimar Döffinger
Makes SIMD-optimized 8x8 and 16x16 idcts for 8 and 10 bit depth
available on aarch64.
For a UHD HDR (10 bit) sample video these were consuming the most time
and this optimization reduced overall decode time from 19
Co-authored-by: Niklas Haas
Signed-off-by: Josh Dekker
---
tests/checkasm/Makefile | 2 +-
tests/checkasm/checkasm.c | 10 +
tests/checkasm/checkasm.h | 10 +
tests/checkasm/hevc_pel.c | 523 ++
4 files changed, 544 insertions(+), 1 deletion(-)
create
Only works for 8x8.
Signed-off-by: Josh Dekker
---
libavcodec/aarch64/Makefile | 3 +-
libavcodec/aarch64/hevcdsp_init.c | 7 +++
libavcodec/aarch64/hevcdsp_sao_neon.S | 87 +++
3 files changed, 96 insertions(+), 1 deletion(-)
create mode 100644
Signed-off-by: Josh Dekker
---
libavcodec/aarch64/Makefile| 3 +-
libavcodec/aarch64/hevcdsp_idct_neon.S | 74 ++
libavcodec/aarch64/hevcdsp_init.c | 19 +++
3 files changed, 95 insertions(+), 1 deletion(-)
create mode 100644 libavcodec/aarch64
Signed-off-by: Josh Dekker
---
libavcodec/aarch64/Makefile | 2 +
libavcodec/aarch64/hevcdsp_add_res_neon.S | 298 ++
libavcodec/aarch64/hevcdsp_init.c | 59 +
libavcodec/hevcdsp.c | 2 +
libavcodec/hevcdsp.h
checkasm: all 657 tests passed
hevc_add_res_4x4_8_c: 49.7
hevc_add_res_4x4_8_neon: 20.5
hevc_add_res_4x4_10_c: 45.7
hevc_add_res_4x4_10_neon: 18.7
hevc_add_res_8x8_8_c: 211.0
hevc_add_res_8x8_8_neon: 24.5
hevc_add_res_8x8_10_c: 195.7
hevc_add_res_8x8_10_neon: 24.0
hevc_add_res_16x16_8_c: 787.2
hevc
Signed-off-by: Josh Dekker
---
checkasm: all 648 tests passed
hevc_add_res_4x4_8_c: 49.7
hevc_add_res_4x4_8_neon: 20.5
hevc_add_res_4x4_10_c: 46.0
hevc_add_res_4x4_10_neon: 19.0
hevc_add_res_8x8_8_c: 209.0
hevc_add_res_8x8_8_neon: 24.5
hevc_add_res_8x8_10_c: 192.7
hevc_add_res_8x8_10_neon: 27.0
On 2021/01/03 20:18, Michael Niedermayer wrote:
On Sun, Jan 03, 2021 at 06:32:11PM +0100, Kieran Kunhya wrote:
Hello,
As it's 2021 I would like to propose FFmpeg purchase one or more (e.g
FATE + development) Apple M1 Mac Minis and provide access to developers.
This is something I have done a fe
On 2020/12/09 11:19, Alan Kelly wrote:
---
Activates avx2 version of yuv2yuvX
Adds checkasm for yuv2yuvX
Modifies ff_yuv2yuvX_* signature to match yuv2yuvX_*
Replaces non-temporal stores with temporal stores
libswscale/x86/Makefile | 1 +
libswscale/x86/swscale.c| 106 +++
Hi,
As discussed in the meeting, I'm starting a RFC for Machines & Platforms of
interest for testing, developer access and FATE. These would be funded by SPI.
The two platforms mentioned were a Mac Mini (M1 Apple Silicon platform) and a
TALOS II (POWER9 platform). My personal suggestion would be
vio)
- Deprecating libpostproc
- Writing down development rules
- Switching to a merge request-like system
- Propose FFmpeg/SPI purchase a Mac Mini ARM machine for development (and FATE
if required).
## People Present (15)
- Jean-Baptiste Kempf
- Josh Dekker (Illya)
- Jan Ekström
- Michael Niederma
24 matches
Mail list logo
