Move the loop counter decrement further from the branch instruction,
this hides the latency of the decrement.
In loops that first load, then store (the horizontal prediction cases),
do the decrement after the load (where the next instruction would
stall a bit anyway, waiting for the result of the
From: Kevin Wheatley
This fix moves the potential definition of _GNU_SOURCE prior to
any includes of system header files as required by the documentation
https://www.gnu.org/software/libc/manual/html_node/Feature-Test-Macros.html
This corrects the CPU_COUNT macro availability, resulting in
sched
Apr 12, 2021, 10:07 by mar...@martin.st:
> Move the loop counter decrement further from the branch instruction,
> this hides the latency of the decrement.
>
> In loops that first load, then store (the horizontal prediction cases),
> do the decrement after the load (where the next instruction would
This prevents bogus values being reported on frame multithreaded decoding
scenarios.
Signed-off-by: James Almer
---
libavcodec/mpeg4videodec.c | 13 +
1 file changed, 13 insertions(+)
diff --git a/libavcodec/mpeg4videodec.c b/libavcodec/mpeg4videodec.c
index 2c440a5026..de66fe8b83 1
1. Refined function get_cabac_inline_mips.
2. Optimize function get_cabac_bypass and get_cabac_bypass_sign.
Speed of decoding h264: 4.89x ==> 5.05x(tested on 3A4000).
---
libavcodec/mips/cabac.h | 140 ++--
1 file changed, 112 insertions(+), 28 deletion
The MSA optimization has been refined in commit 93218c2 and ce0a52e.
It is better than MMI version now.
Speed of decoding H264: 4.83x ==> 4.89x (tested on 3A4000).
---
libavcodec/mips/h264chroma_init_mips.c | 19 +--
1 file changed, 9 insertions(+), 10 deletions(-)
diff --git a/li
v2: Fixed a build error in [PATCH 2/5].
v3: add patch 4/5.
v4: Fix bug in 2/5 caused by instruction 'lhu' on BIGENDIAN environment.
[PATCH v4 1/5] avcodec/mips: Restore the initialization sequence of
[PATCH v4 2/5] avcodec/mips: Refine get_cabac_inline_mips.
[PATCH v4 3/5] avcodec/mips: Optimize f
From: gxw
Using mask to avoid judgment, H264 4K decoding speed
improved about 0.1fps tested on 3A4000
Signed-off-by: Shiyou Yin
---
libavcodec/mips/h264dsp_msa.c | 465 --
1 file changed, 171 insertions(+), 294 deletions(-)
diff --git a/libavcodec/mips/
From: gxw
Speed of decoding H264 1080P: 5.05x ==> 5.13x
Signed-off-by: Shiyou Yin
---
libavcodec/mips/Makefile| 3 +-
libavcodec/mips/h264_deblock_msa.c | 153
libavcodec/mips/h264dsp_init_mips.c | 2 +
libavcodec/mips/h264dsp_mips.h |
MSA2 optimizations are attached to MSA macros in generic_macros_msa.h.
It's difficult to do runtime check for them. Remove this part of code
can make it more robust. H264 1080p decoding: 5.13x==>5.12x.
---
configure | 7 +--
libavutil/mips/generic_macros_msa.h | 37 -
Ping.
On 2021-04-10 20:00, Gyan Doshi wrote:
It was added in 51ac1f616f due to ticket #4218, in order to show a single
image via ffserver. With ffserver long gone, it serves no purpose.
---
libavformat/Makefile | 1 -
libavformat/allformats.c | 1 -
libavformat/rawenc.c | 13 --
> 2021年3月31日 下午10:39，Michael Niedermayer 写道：
>
> On Tue, Mar 30, 2021 at 08:51:52PM +0800, Shiyou Yin wrote:
>> 1. Refined function get_cabac_inline_mips.
>> 2. Optimize function get_cabac_bypass and get_cabac_bypass_sign.
>>
>> Speed of decoding h264: 4.89x ==> 5.05x(tested on 3A4000).
>> ---
This encoder has AVCodec.pix_fmts set, so ff_encode_preinit() already
checks for this.
Signed-off-by: Andreas Rheinhardt
---
Will apply tomorrow unless there are objections.
libavcodec/jpeglsenc.c | 8
1 file changed, 8 deletions(-)
diff --git a/libavcodec/jpeglsenc.c b/libavcodec/jpe
Benchmarks:
pred16x16_dc_10_c: 124.0
pred16x16_dc_10_neon: 97.2
pred16x16_horizontal_10_c: 71.7
pred16x16_horizontal_10_neon: 66.2
pred16x16_top_dc_10_c: 90.7
pred16x16_top_dc_10_neon: 71.5
pred16x16_vertical_10_c: 64.7
pred16x16_vertical_10_neon: 61.7
Some functions work slower than C and are lef
Signed-off-by: Dhanish Vijayan
---
libavcodec/cuviddec.c | 199 ++
1 file changed, 199 insertions(+)
diff --git a/libavcodec/cuviddec.c b/libavcodec/cuviddec.c
index ec57afdefe..3b07d0a874 100644
--- a/libavcodec/cuviddec.c
+++ b/libavcodec/cuviddec.c
@@ -
On Sat, Mar 20, 2021 at 5:34 PM <1160386...@qq.com> wrote:
>
> From: He Yang <1160386...@qq.com>
>
> Signed-off-by: He Yang <1160386...@qq.com>
Sorry for taking a while to respond, and thank you for the
contribution. I have verified that this conversion and FindWindowW
usage indeed fixes issues wi
On 4/12/2021 2:07 PM, Andreas Rheinhardt wrote:
This encoder has AVCodec.pix_fmts set, so ff_encode_preinit() already
checks for this.
Signed-off-by: Andreas Rheinhardt
---
Will apply tomorrow unless there are objections.
libavcodec/jpeglsenc.c | 8
1 file changed, 8 deletions(-)
On 4/12/2021 5:21 PM, Dhanish Vijayan wrote:
Signed-off-by: Dhanish Vijayan
---
libavcodec/cuviddec.c | 199 ++
1 file changed, 199 insertions(+)
diff --git a/libavcodec/cuviddec.c b/libavcodec/cuviddec.c
index ec57afdefe..3b07d0a874 100644
--- a/libav
James Almer:
> On 4/12/2021 2:07 PM, Andreas Rheinhardt wrote:
>> This encoder has AVCodec.pix_fmts set, so ff_encode_preinit() already
>> checks for this.
>>
>> Signed-off-by: Andreas Rheinhardt
>> ---
>> Will apply tomorrow unless there are objections.
>>
>> libavcodec/jpeglsenc.c | 8
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 +
Apr 13, 2021, 02:45 by j...@itanimul.li:
> 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 +
From: "Chen,Wenbin"
Add mbbrc to hevc_qsv
For detailed description, please see "mbbrc" part in:
https://github.com/Intel-Media-SDK/MediaSDK/blob/master/doc/mediasdk-man.md#mfxextcodingoption2
Signed-off-by: Wenbin Chen
---
libavcodec/qsvenc.c | 5 +++--
1 file changed, 3 insertions(+), 2 delet
On Tue, 2021-04-13 at 10:22 +0800, wenbin.c...@intel.com wrote:
> From: "Chen,Wenbin"
>
> Add mbbrc to hevc_qsv
> For detailed description, please see "mbbrc" part in:
>
https://github.com/Intel-Media-SDK/MediaSDK/blob/master/doc/mediasdk-man.md#mfxextcodingoption2
>
> Signed-off-by: Wenbin Che
---
doc/APIchanges | 2 +
libavutil/Makefile | 2 +
libavutil/detection_bbox.c | 73 +
libavutil/detection_bbox.h | 107 +
libavutil/frame.c | 1 +
libavutil/frame.h | 6 +++
6 files changed,
---
libavfilter/f_sidedata.c | 2 ++
libavfilter/vf_showinfo.c | 29 +
2 files changed, 31 insertions(+)
diff --git a/libavfilter/f_sidedata.c b/libavfilter/f_sidedata.c
index 3757723375..6f25d2b311 100644
--- a/libavfilter/f_sidedata.c
+++ b/libavfilter/f_sidedata.c
Below are the example steps to do object detection:
1. download and install l_openvino_toolkit_p_2021.1.110.tgz from
https://software.intel.com/content/www/us/en/develop/tools/openvino-toolkit/download.html
or, we can get source code (tag 2021.1), build and install.
2. export LD_LIBRARY_PATH wit
On 31.03.2021 12:13, Tobias Rapp wrote:
The "packets_and_frames" element has been added to ffprobe.xsd in
0c9f0da0f7656059e9bd41931d250aafddf35ea3 but apparently removing the
check in ffprobe.c has been forgotten.
Signed-off-by: Tobias Rapp
---
fftools/ffprobe.c | 7 ---
1 file changed,
27 matches
Mail list logo
