On 2014-08-13 14:05, Michael Niedermayer wrote:
> On Wed, Aug 13, 2014 at 10:46:45AM +0200, James Darnley wrote:
>> On 2014-08-13 04:50, Michael Niedermayer wrote:
>>> On Tue, Aug 12, 2014 at 11:22:07PM +0200, James Darnley wrote:
>>>> ---
>>>> libavco
State that the maximum value of order is 32. This limit is used in both C and
x86 assebly code.
---
libavcodec/flacdsp.h |7 +++
1 files changed, 7 insertions(+), 0 deletions(-)
diff --git a/libavcodec/flacdsp.h b/libavcodec/flacdsp.h
index 272cf2a..e8487c8 100644
--- a/libavcodec/flacds
The LPC encoder is not used at compression levels below 3 but the basic acodec
test uses -compression_level 2. This adds a test using -compression_level 5,
which is the current default.
---
tests/fate/acodec.mak|4 +++-
tests/ref/acodec/flac-comp-5 |4
2 files changed, 7 inse
This should help to clarify the API.
---
libavcodec/flacdsp.h |2 +-
1 files changed, 1 insertions(+), 1 deletions(-)
diff --git a/libavcodec/flacdsp.h b/libavcodec/flacdsp.h
index e8487c8..23d1ae4 100644
--- a/libavcodec/flacdsp.h
+++ b/libavcodec/flacdsp.h
@@ -35,7 +35,7 @@ typedef struct F
On 2014-08-14 18:03, Christophe Gisquet wrote:
> Hi,
>
> 2014-08-14 17:15 GMT+02:00 James Darnley :
>> The LPC encoder is not used at compression levels below 3 but the basic
>> acodec
>> test uses -compression_level 2. This adds a test using -compression_level
After further searching and understanding the makefile I see that the
lpc code is tested by the complete fate sample suite. This patch is
unnecessary.
Anyway to answer the questions...
On 2014-08-14 18:14, Christophe Gisquet wrote:
> Hi,
>
> 2014-08-14 18:08 GMT+02:00 Jame
On 2014-08-16 01:15, Michael Niedermayer wrote:
> Signed-off-by: Michael Niedermayer
> ---
> libavcodec/dvdsubdec.c |4 ++--
> 1 file changed, 2 insertions(+), 2 deletions(-)
Fine, clearly a pointer.
signature.asc
Description: OpenPGP digital signature
__
On 2014-08-16 01:15, Michael Niedermayer wrote:
> Signed-off-by: Michael Niedermayer
> ---
> libavcodec/dvbsubdec.c |4 ++--
> 1 file changed, 2 insertions(+), 2 deletions(-)
Fine, both pointers.
signature.asc
Description: OpenPGP digital signature
___
On 2014-08-16 01:15, Michael Niedermayer wrote:
> Signed-off-by: Michael Niedermayer
> ---
> libavformat/udp.c |4 ++--
> 1 file changed, 2 insertions(+), 2 deletions(-)
>
> diff --git a/libavformat/udp.c b/libavformat/udp.c
> index a8baa5b..8412eeb 100644
> --- a/libavformat/udp.c
> +++ b/l
On 2014-08-16 01:15, Michael Niedermayer wrote:
> Signed-off-by: Michael Niedermayer
> ---
> libavcodec/hevc_mvs.c |2 +-
> 1 file changed, 1 insertion(+), 1 deletion(-)
>
fine
signature.asc
Description: OpenPGP digital signature
___
ffmpeg-de
On 2014-08-16 01:15, Michael Niedermayer wrote:
> Signed-off-by: Michael Niedermayer
> ---
> libavcodec/mips/compute_antialias_float.h |3 ---
> 1 file changed, 3 deletions(-)
It is unused so that is fine.
signature.asc
Description: OpenPGP digital signature
On 2014-08-18 14:53, Stefano Sabatini wrote:
> +@item whole_len
> +Set the target number of sample in the audio stream. After the value
^^
samples, plural
signature.asc
Description: OpenPGP digital signature
___
ffmpeg-
On 2014-08-20 10:10, Christophe Gisquet wrote:
> It was added per pixel instead of per line.
> ---
> libavcodec/dpxenc.c | 4 ++--
> 1 file changed, 2 insertions(+), 2 deletions(-)
>
> diff --git a/libavcodec/dpxenc.c b/libavcodec/dpxenc.c
> index 059d8c6..aca745b 100644
> --- a/libavcodec/dpxenc
---
libavutil/frame.h |2 +-
1 files changed, 1 insertions(+), 1 deletions(-)
diff --git a/libavutil/frame.h b/libavutil/frame.h
index 2391d3e..d6d017c 100644
--- a/libavutil/frame.h
+++ b/libavutil/frame.h
@@ -162,7 +162,7 @@ typedef struct AVFrame {
* For audio, only linesize[0] may b
On 2014-08-27 22:18, Timothy Gu wrote:
> On Wed, Aug 27, 2014 at 12:38 PM, James Darnley
> wrote:
>> ---
>> libavutil/frame.h |2 +-
>> 1 files changed, 1 insertions(+), 1 deletions(-)
>
> OK
Thanks but don't apply this patch just yet. Similar typos wer
Also replace the plural form "multiplies" with "multiples".
---
This drops the addition of the apostrophe from my earlier patch.
doc/fftools-common-opts.texi |2 +-
libavcodec/dirac.c |2 +-
libavcodec/mpeg12enc.c |2 +-
libavcodec/msvideo1enc.c |2 +-
libavut
I am looking for someone with an XOP capable CPU to compile and run two
basic fate tests for me. All you have to do is:
- get the ffmpeg source from my git repo (link below)
- checkout the flac_enc_lpc branch: git checkout flac_enc_lpc
- configure with --enable-gpl
- build ffmpeg
- run: make fate-
On 2014-09-02 20:39, Kelly Anderson wrote:
> TESTacodec-flac-comp-5
> TESTacodec-flac-24-comp-5
> tests/Makefile:192: recipe for target 'fate-acodec-flac-24-comp-5' failed
> make: *** [fate-acodec-flac-24-comp-5] Error 1
Thanks. Good to know that the 16-bit version works. I am probably
On 2014-09-02 21:25, James Darnley wrote:
> On 2014-09-02 20:39, Kelly Anderson wrote:
>> TESTacodec-flac-comp-5
>> TESTacodec-flac-24-comp-5
>
>> tests/Makefile:192: recipe for target 'fate-acodec-flac-24-comp-5' failed
>> make: *** [fate-acodec-fl
On 2014-09-02 17:13, Pascal Massimino wrote:
> %macro DECLARE_VAR 2
> %define %1 %2
> %define %1d %2d
> %endmacro
>
> ;; Mappings for common variables
> DECLARE_VAR index, r5
> DECLARE_VAR total, r4
> DECLARE_VAR parallel_width, r6
I suggest you leave these up to the cglobal macro
> DE
On 2014-09-03 02:13, Michael Niedermayer wrote:
> On Tue, Sep 02, 2014 at 11:49:47PM +0200, Reimar Döffinger wrote:
> [...]
>
>> reason that configure is already too slow.
>
> ccache + --tempprefix + dash instead of bash should help a bit
Tell that to Windows!
Okay trying dash did reduce the 10
cpuflags
Henrik Gramner (2):
x86inc: Make ym# behave the same way as xm#
x86inc: Make INIT_CPUFLAGS support an arbitrary number of cpuflags
James Darnley (2):
lavc/x86/v210: give cpuflag to INIT macro
lavu/x86/x86inc: deprecate INIT_AVX
Loren Merritt (1):
x86inc: free up variable nam
From: Anton Mitrofanov
---
libavutil/x86/x86inc.asm | 581 +++---
1 files changed, 295 insertions(+), 286 deletions(-)
diff --git a/libavutil/x86/x86inc.asm b/libavutil/x86/x86inc.asm
index 1cfb055..87f3af7 100644
--- a/libavutil/x86/x86inc.asm
+++ b/lib
This lets the cglobal macro automatically append a suffix to the function name.
This means that INIT_XMM avx must be used rather than INIT_AVX.
---
libavcodec/x86/v210.asm | 22 --
1 files changed, 12 insertions(+), 10 deletions(-)
diff --git a/libavcodec/x86/v210.asm b/liba
From: Henrik Gramner
This makes more sense for future implementations of templates with zmm
registers.
---
libavutil/x86/x86inc.asm |8
1 files changed, 4 insertions(+), 4 deletions(-)
diff --git a/libavutil/x86/x86inc.asm b/libavutil/x86/x86inc.asm
index 87f3af7..65a23d8 100644
-
The same can be done with INIT_XMM avx
---
libavutil/x86/x86inc.asm |8
1 files changed, 0 insertions(+), 8 deletions(-)
diff --git a/libavutil/x86/x86inc.asm b/libavutil/x86/x86inc.asm
index 65a23d8..c028c0b 100644
--- a/libavutil/x86/x86inc.asm
+++ b/libavutil/x86/x86inc.asm
@@ -85
From: Loren Merritt
---
libavutil/x86/x86inc.asm | 18 +-
1 files changed, 9 insertions(+), 9 deletions(-)
diff --git a/libavutil/x86/x86inc.asm b/libavutil/x86/x86inc.asm
index c028c0b..4cc6b0e 100644
--- a/libavutil/x86/x86inc.asm
+++ b/libavutil/x86/x86inc.asm
@@ -821,12 +8
From: Henrik Gramner
Previously there was a limit of two cpuflags.
---
libavutil/x86/x86inc.asm | 41 ++---
1 files changed, 22 insertions(+), 19 deletions(-)
diff --git a/libavutil/x86/x86inc.asm b/libavutil/x86/x86inc.asm
index 4cc6b0e..fa6a6ab 100644
---
paddq is an SSE2 instruction so it cannot be used for MMX.
This was probably just a typo because the sums are dwords anyway.
---
We should have picked this up in review but we didn't...
libavfilter/x86/vf_idet.asm |2 +-
1 files changed, 1 insertions(+), 1 deletions(-)
diff --git a/libavfi
On 2014-09-05 00:31, Michael Niedermayer wrote:
> This avoids several failures on fate.ffmpeg.org, and thus makes real
> bugs easier to spot
A good reason.
> +volatile uint8_t in[1000]; // volatile to workaround
> http://llvm.org/bugs/show_bug.cgi?id=20849
Perhaps add a FIXME to the comment
On 2014-09-05 00:13, James Darnley wrote:
> From: Anton Mitrofanov
>
> ---
> libavutil/x86/x86inc.asm | 581
> +++---
> 1 files changed, 295 insertions(+), 286 deletions(-)
If I didn't make it clear, this has not yet been added t
On 2014-09-08 14:17, Gaullier Nicolas wrote:
> [FFmpeg-devel] MXF : default fied dominance is TFF
That should be *field* dominance. If you're going to submit another
patch please correct the commit message.
signature.asc
Description: OpenPGP digital signature
_
On 2014-09-11 23:00, Timothy Gu wrote:
> On Thu, Sep 11, 2014 at 1:57 PM, Carl Eugen Hoyos wrote:
>> Lou Logan lrcd.com> writes:
>>
>>> - http://www.mplayerhq.hu/DOCS/tech/patches.txt";>
>>> MPlayer patch guidelines
>>
>> I understand that this is not ideal but ...
>>
>>> + Using git to develop F
On 2014-09-11 23:10, James Darnley wrote:
> I support the change and I am willing to look at adding and adapting
> some relevant points from the mplayer text.
Point 0, "diffs please", is covered by our request for git format-patch
in the patch submission checklist[1] point 2 an
It gets rendered inline in HTML and becomes not very clear.
---
doc/developer.texi | 3 +++
1 file changed, 3 insertions(+)
diff --git a/doc/developer.texi b/doc/developer.texi
index e51e97a..f1fed5b 100644
--- a/doc/developer.texi
+++ b/doc/developer.texi
@@ -323,9 +323,12 @@ Always fill out the
This adds some text covering points from that mplayer text which are missing in
our docs. I welcome your comments, particularly on the language used.
The first two patches are really just cosmetic changes but I noticed them while
reading the file.
James Darnley (6):
docs: add missing code
---
doc/developer.texi | 6 +++---
1 file changed, 3 insertions(+), 3 deletions(-)
diff --git a/doc/developer.texi b/doc/developer.texi
index 1e1d3b8..e51e97a 100644
--- a/doc/developer.texi
+++ b/doc/developer.texi
@@ -328,7 +328,7 @@ area changed: Short 1 line description
details describing wh
This addresses point 12 from the mplayer patches text file. Discussion about
removing a link to this document can be found in the thread on ffmpeg-devel
titled: [PATCH] web/contact: replace MPlayer guidelines link
---
doc/developer.texi | 3 +++
1 file changed, 3 insertions(+)
diff --git a/doc/d
This addresses point 11 from the mplayer patches text file. Discussion about
removing a link to this document can be found in the thread on ffmpeg-devel
titled: [PATCH] web/contact: replace MPlayer guidelines link
---
doc/developer.texi | 5 +
1 file changed, 5 insertions(+)
diff --git a/doc
This addresses point 13 from the mplayer patches text file. Discussion about
removing a link to this document can be found in the thread on ffmpeg-devel
titled: [PATCH] web/contact: replace MPlayer guidelines link
---
doc/developer.texi | 5 +
1 file changed, 5 insertions(+)
diff --git a/doc
This addresses point 6 from the mplayer patches text file. Discussion about
removing a link to this document can be found in the thread on ffmpeg-devel
titled: [PATCH] web/contact: replace MPlayer guidelines link
---
doc/developer.texi | 7 +++
1 file changed, 7 insertions(+)
diff --git a/do
On 2014-09-12 01:33, Timothy Gu wrote:
> On Thu, Sep 11, 2014 at 4:31 PM, James Darnley
> wrote:
>> ---
>> doc/developer.texi | 6 +++---
>> 1 file changed, 3 insertions(+), 3 deletions(-)
>
> I would use @command{}.
I will change it to command.
Should I go
On 2014-09-12 16:52, compn wrote:
> On Fri, 12 Sep 2014 11:42:01 +0200
> Stefano Sabatini wrote:
>
>> On date Friday 2014-09-12 03:12:06 +0200, Michael Niedermayer encoded:
>>> On Fri, Sep 12, 2014 at 01:31:29AM +0200, James Darnley wrote:
>> [...]
>>>
On 2014-09-12 01:34, Timothy Gu wrote:
> On Thu, Sep 11, 2014 at 4:31 PM, James Darnley
> wrote:
>> This addresses point 12 from the mplayer patches text file. Discussion about
>> removing a link to this document can be found in the thread on ffmpeg-devel
>> titled: [PAT
On 2014-09-12 00:15, Carl Eugen Hoyos wrote:
> James Darnley gmail.com> writes:
>
>> Point 2, "unified diffs", is covered by out request
>> for patches. Are we doing to allow git diff dumps?
>
> While I always send unified diffs, some people (still)
field, vf);
diff --git a/libavfilter/vf_eq.c b/libavfilter/vf_eq.c
new file mode 100644
index 000..2dc21dc
--- /dev/null
+++ b/libavfilter/vf_eq.c
@@ -0,0 +1,157 @@
+/*
+ * Copyright (c) 2014 James Darnley
+ *
+ * Original MPlayer filter by Richard Felker.
+ *
+ * This file is part o
On 2014-09-15 13:17, Clément Bœsch wrote:
> On Mon, Sep 15, 2014 at 12:46:14AM +0200, James Darnley wrote:
>> ---
>>
>> Still a couple of things on the todo list (see the top of the new file) but I
>> thought I would send the basics along for people to
On 2014-09-15 02:03, Timothy Gu wrote:
> On Sep 14, 2014 3:46 PM, "James Darnley" wrote:
>>
>> ---
>>
>> Still a couple of things on the todo list (see the top of the new file)
> but I
>> thought I would send the basics along for people to n
On 2014-09-15 00:46, James Darnley wrote:
> ...
Update: I've got eq2 merged into the one file and working. I would send
another WIP patch but I've made several commits, all of which should be
merged together.
One question I do have though is: should the user options for contrast
On 2014-09-21 12:51, wm4 wrote:
> it's easy to prevent at least early EOF. Playback will happen to work,
> even if it's slightly broken.
Do we want to print a warning in the "slightly broken" case? Does it
just mean that the middle cannot be gapless between file1 and file2?
signature.asc
Descr
On 2014-10-01 03:24, James Almer wrote:
> ~15 faster.
15 what faster? 15 cycles? 15%? 15 times?
signature.asc
Description: OpenPGP digital signature
___
ffmpeg-devel mailing list
ffmpeg-devel@ffmpeg.org
http://ffmpeg.org/mailman/listinfo/ffmpeg-de
On 2014-09-30 20:50, James Almer wrote:
> +pshufdm2, m0, 0xb1
> +pshufdm3, m1, 0xb1
> +pshufdm3, m1, 0xb1
> +pshufdm4, m2, 0xb1
> +pshufd xm1, xm0, 0x4e
> +pshufd xm2, xm0, 0xb1
> +pshufd xm3, xm1, 0xb1
> +pshufd
Since we have a few newbies on the list with the start of OPW and I've
seen a few other cases recently, I thought this would be a good time to
remind people of the feature of Git called Hooks.
Simply put: a git hook is an appropriately named executable file in the
.git/hook directory in your sourc
On 2014-10-11 19:59, Clément Bœsch wrote:
> ---
> tests/fate/subtitles.mak | 4 ++--
> 1 file changed, 2 insertions(+), 2 deletions(-)
Looks okay.
signature.asc
Description: OpenPGP digital signature
___
ffmpeg-devel mailing list
ffmpeg-devel@ffmpeg
On 2014-10-11 19:59, Clément Bœsch wrote:
> ---
> libavformat/assenc.c | 11 +--
> 1 file changed, 9 insertions(+), 2 deletions(-)
The diff looks fine. So if the code does what you want it to do, push.
Just one comment below.
> +ass->ssa_mode = !strstr(avctx->extradata, "\n[V4+
On 2014-10-11 19:59, Clément Bœsch wrote:
> Fixes Ticket #3207
> ---
> libavcodec/ass.c| 29 +--
> libavcodec/version.h| 2 +-
> tests/ref/fate/sub-aqtitle | 72 +--
> tests/ref/fate/sub-charenc | 106
>
On 2014-10-11 19:59, Clément Bœsch wrote:
> diff --git a/libavcodec/ass_split.h b/libavcodec/ass_split.h
> index 06c1ce3..c912252 100644
> --- a/libavcodec/ass_split.h
> +++ b/libavcodec/ass_split.h
> @@ -41,13 +41,28 @@ typedef struct {
> char *font_name; /**< font face (case sensitive)
On 2014-10-15 10:49, Thomas Volkert wrote:
> Please, vote for a Facebook url here:
> http://goo.gl/forms/AHVFvPxp8X
>
> This will show us the community opinion about this.
I can't. It requires a google account.
signature.asc
Description: OpenPGP digital signature
On 2014-10-17 23:09, Nicholas Robbins wrote:
>> So is it possible to make a filter that is basically just a wrapper around
>> another filter? This hypothetical filter (call it vf_keystone) would just
>> call
>> vf_perspective with crazy options. Clearly I could just copy the
>> vf_perspective
On 2018-04-26 13:15, Daniel Oberhoff wrote:
> I was wondering if there is any chance to move development to github?
> I.e. not just mirror, but as primary development repo, with issues and
> pull requests? Would make collaboration a *lot* easier (think of
> submitting a pr instead of having to gene
On 2018-01-16 22:26, Martin Vignali wrote:
> diff --git a/libavutil/x86/x86util.asm b/libavutil/x86/x86util.asm
> index d7cd996842..9db2d90e57 100644
> --- a/libavutil/x86/x86util.asm
> +++ b/libavutil/x86/x86util.asm
> @@ -335,7 +335,7 @@
> %endmacro
>
> %macro ABS2 4
> -%if cpuflag(ssse3)
> +
From: Rostislav Pehlivanov
The rounding caused by the size scaler wasn't compensated for and the
slice sizes grew beyond what is allowed per frame.
Signed-off-by: Rostislav Pehlivanov
---
libavcodec/vc2enc.c | 22 ++
1 file changed, 14 insertions(+), 8 deletions(-)
diff --
On 2018-01-31 15:56, James Darnley wrote:
> From: Rostislav Pehlivanov
>
> The rounding caused by the size scaler wasn't compensated for and the
> slice sizes grew beyond what is allowed per frame.
>
> Signed-off-by: Rostislav Pehlivanov
> ---
I forgot to say that
On 2018-02-10 14:17, Rostislav Pehlivanov wrote:
> Signed-off-by: Rostislav Pehlivanov
> ---
> libavcodec/x86/h264_idct.asm | 6 +-
> 1 file changed, 5 insertions(+), 1 deletion(-)
>
> diff --git a/libavcodec/x86/h264_idct.asm b/libavcodec/x86/h264_idct.asm
> index 8804638091..c54f9f1a68 100
A lax or tolerant decoder may support an End Sequence between fields of the same
frame but the spec. says this is not allowed.
2012 spec. 10.3.2, 2017 spec. 10.4.3:
Where pictures are fields, a sequence shall comprise a whole number of
frames (i.e., an even number of fields) and shall begin and en
James Darnley (2):
avcodec/vc2enc: do not write an End Sequence after the first field of
field-separated pictures
avcodec/vc2enc: do not reset the last_parse_code variable every frame
libavcodec/vc2enc.c | 9 +++--
tests/ref/vsynth/vsynth1-vc2-420p | 2
It should be kept between frames so that the encoder can actually know
whether the previous parse_code was an End Sequence.
---
libavcodec/vc2enc.c | 1 -
tests/ref/vsynth/vsynth1-vc2-420p | 2 +-
tests/ref/vsynth/vsynth1-vc2-420p10 | 2 +-
tests/ref/vsynth/vsynth1-vc
While cherry-picking some stuff for avx512 I have noticed that ffmpeg
has a discrepancy in the comments for the two avx512 flags.
Lets start with the public header
libavutil/cpu.h
56│ #define AV_CPU_FLAG_AVX512 0x10 ///< AVX-512 functions: requires
OS support even if YMM/ZMM register
I guess it could also be scaled to ymm if you're a big Skylake fan :P
(in which case you'd probably want to reorder the shuffle indices so
---
.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"
---
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
+%else
+pand m1, m6, m1
+pandn m0, m6, m0
+porm0, m0, m1
+%endif
Isn't that pattern a vpblendb or some such ?
I think Kieran already responded to this on IRC but I will too.
Unfortunately not. This blend is at the bit lev
---
tests/checkasm/v210enc.c | 6 --
1 file changed, 4 insertions(+), 2 deletions(-)
diff --git a/tests/checkasm/v210enc.c b/tests/checkasm/v210enc.c
index 9942e08137..9fb8321c25 100644
--- a/tests/checkasm/v210enc.c
+++ b/tests/checkasm/v210enc.c
@@ -72,8 +72,10 @@
randomize_buf
---
libavcodec/x86/v210enc.asm | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/libavcodec/x86/v210enc.asm b/libavcodec/x86/v210enc.asm
index afac238ede..c2ad3d72c0 100644
--- a/libavcodec/x86/v210enc.asm
+++ b/libavcodec/x86/v210enc.asm
@@ -62,7 +62,7 @@ SECTION .text
; v210
avx512 on Skylake-X (Xeon D-2123IT):
1.19x faster (970±91.2 vs. 817±104.4 decicycles) compared with avx2
avx512icl on Ice Lake (Xeon Silver 4316):
2.52x faster (1350±5.3 vs. 535±9.5 decicycles) compared with avx2
---
libavcodec/x86/v210enc.asm| 99 +++
libavcod
ARCH_X86_64 is always defined. So checks of this type need to check with #if.
Thanks. I forgot the ffmpeg convention there.
___
ffmpeg-devel mailing list
ffmpeg-devel@ffmpeg.org
https://ffmpeg.org/mailman/listinfo/ffmpeg-devel
To unsubscribe, visit l
---
tests/checkasm/v210enc.c | 6 --
1 file changed, 4 insertions(+), 2 deletions(-)
diff --git a/tests/checkasm/v210enc.c b/tests/checkasm/v210enc.c
index 9942e08137..9fb8321c25 100644
--- a/tests/checkasm/v210enc.c
+++ b/tests/checkasm/v210enc.c
@@ -72,8 +72,10 @@
randomize_buf
---
libavcodec/x86/v210enc.asm | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/libavcodec/x86/v210enc.asm b/libavcodec/x86/v210enc.asm
index afac238ede..c2ad3d72c0 100644
--- a/libavcodec/x86/v210enc.asm
+++ b/libavcodec/x86/v210enc.asm
@@ -62,7 +62,7 @@ SECTION .text
; v210
avx512 on Skylake-X (Xeon D-2123IT):
1.19x faster (970±91.2 vs. 817±104.4 decicycles) compared with avx2
avx512icl on Ice Lake (Xeon Silver 4316):
2.52x faster (1350±5.3 vs. 535±9.5 decicycles) compared with avx2
---
libavcodec/x86/v210enc.asm| 99 +++
libavcod
---
libavcodec/x86/v210enc.asm | 8
1 file changed, 4 insertions(+), 4 deletions(-)
diff --git a/libavcodec/x86/v210enc.asm b/libavcodec/x86/v210enc.asm
index 552164a8be..d3639cd440 100644
--- a/libavcodec/x86/v210enc.asm
+++ b/libavcodec/x86/v210enc.asm
@@ -314,7 +314,7 @@ cglobal v210_
---
libavcodec/x86/v210enc.asm | 1 -
1 file changed, 1 deletion(-)
diff --git a/libavcodec/x86/v210enc.asm b/libavcodec/x86/v210enc.asm
index d3639cd440..daf5f2ab81 100644
--- a/libavcodec/x86/v210enc.asm
+++ b/libavcodec/x86/v210enc.asm
@@ -331,7 +331,6 @@ cglobal v210_planar_pack_8, 5, 5, 7+no
---
configure | 5 +
1 file changed, 5 insertions(+)
diff --git a/configure b/configure
index f4eedfc207..eaa5ef6b20 100755
--- a/configure
+++ b/configure
@@ -4315,6 +4315,11 @@ case "$toolchain" in
add_cflags -fsanitize=address
add_ldflags -fsanitize=address
;;
+
On 12/7/22 17:08, James Darnley wrote:
---
configure | 5 +
1 file changed, 5 insertions(+)
diff --git a/configure b/configure
index f4eedfc207..eaa5ef6b20 100755
--- a/configure
+++ b/configure
@@ -4315,6 +4315,11 @@ case "$toolchain" in
add_cflags -fsaniti
---
libavcodec/x86/v210.asm | 12 ++--
1 file changed, 6 insertions(+), 6 deletions(-)
diff --git a/libavcodec/x86/v210.asm b/libavcodec/x86/v210.asm
index 3b9e0761df..600a4ddc5f 100644
--- a/libavcodec/x86/v210.asm
+++ b/libavcodec/x86/v210.asm
@@ -65,18 +65,18 @@ cglobal v210_planar_unp
Ice Lake (Xeon Silver 4316): 2.01x faster (1147±36.8 vs. 571±38.2 decicycles)
compared with avx2
---
I think I can merge this with the existing macro without it being too ugly.
That might allow a plain avx512 version too but I can't say if that would be any
faster.
libavcodec/x86/v210-init.c |
On 2014-11-10 15:01, Nicolas George wrote:
>> + fread(&c, 1, 1, in);
>> + fread(&Y, 1, 1, in);
>> + fread(&Cr, 1, 1, in);
>> + fread(&Cb, 1, 1, in);
>
> You could replace that with a single read into an array; that would not
> allow to call the variables R, Cr, Cb, but that d
On 2015-07-05 20:14, Paul B Mahol wrote:
> + * TODO: add SIMD
If you and nobody else is working assembly then I am volunteering (and
have literally just started).
signature.asc
Description: OpenPGP digital signature
___
ffmpeg-devel mailing list
ffmpe
On 2015-07-05 20:14, Paul B Mahol wrote:
> +static int mode01(int c, int a1, int a2, int a3, int a4, int a5, int a6, int
> a7, int a8)
> +{
> +const int mi = FFMIN(FFMIN(FFMIN(a1, a2), FFMIN(a3, a4)),
> FFMIN(FFMIN(a5, a6), FFMIN(a7, a8)));
> +const int ma = FFMAX(FFMAX(FFMAX(a1, a2), FFM
/libavfilter/removegrain.h
new file mode 100644
index 000..60401fb
--- /dev/null
+++ b/libavfilter/removegrain.h
@@ -0,0 +1,40 @@
+/*
+ * Copyright (c) 2015 Paul B Mahol
+ * Copyright (c) 2015 James Darnley
+ *
+ * This file is part of FFmpeg.
+ *
+ * FFmpeg is free software; you can redistribute it and
On 2015-07-11 15:45, Carl Eugen Hoyos wrote:
> James Darnley gmail.com> writes:
>
>> +;* TODO: gpl text goes here.
>
> Yes, please copy it from another asm file.
>
> And please make sure that the file does
> not get compiled if --enable-gpl was not
> specif
James Darnley
+ *
+ * This file is part of FFmpeg.
+ *
+ * FFmpeg is free software; you can redistribute it and/or
+ * modify it under the terms of the GNU Lesser General Public
+ * License as published by the Free Software Foundation; either
+ * version 2.1 of the License, or (at your option) any
On 2015-07-13 01:34, Ronald S. Bultje wrote:
> Hi,
>
> On Sun, Jul 12, 2015 at 5:54 PM, Paul B Mahol wrote:
>
>> On 7/12/15, Ronald S. Bultje wrote:
>>> +typedef struct PSNRDSPContext {
>>> +uint64_t (*sse_line)(const uint8_t *buf, const uint8_t *ref, int w);
>>
>> Besides naming of functio
On 2015-07-14 19:50, Michael Niedermayer wrote:
> From: Michael Niedermayer
>
> Signed-off-by: Michael Niedermayer
> ---
> libavfilter/internal.h | 17 +
> 1 file changed, 17 insertions(+)
>
> diff --git a/libavfilter/internal.h b/libavfilter/internal.h
> index a7ec751..fcd1a
On 2015-07-11 18:34, James Almer wrote:
> On 11/07/15 10:40 AM, James Darnley wrote:
>> @@ -566,7 +554,19 @@ static int filter_slice(AVFilterContext *ctx, void
>> *arg, int jobnr, int nb_jobs)
>> }
>>
>> *dst++ = *src++;
>> -
On 2015-07-14 21:47, Michael Niedermayer wrote:
> On Tue, Jul 14, 2015 at 08:42:40PM +0200, James Darnley wrote:
>> On 2015-07-14 19:50, Michael Niedermayer wrote:
>>> + * One example of a source returning AVERROR(EAGAIN) is a buffer into
>>> which the
>>> + *
On 2015-07-15 00:15, Paul B Mahol wrote:
> diff --git a/doc/filters.texi b/doc/filters.texi
> index 9f7b976..d4d8bad 100644
> --- a/doc/filters.texi
> +++ b/doc/filters.texi
> @@ -3718,6 +3718,24 @@ Set whether or not chroma is considered in the metric
> calculations. Default is
> @code{1}.
> @e
On 2015-07-14 23:23, James Almer wrote:
> On 14/07/15 3:54 PM, James Darnley wrote:
>> On 2015-07-11 18:34, James Almer wrote:
>>> On 11/07/15 10:40 AM, James Darnley wrote:
>>>> new file mode 100644
>>>> index 000..5e1feea
>>>> --- /
/null
+++ b/libavfilter/removegrain.h
@@ -0,0 +1,40 @@
+/*
+ * Copyright (c) 2015 Paul B Mahol
+ * Copyright (c) 2015 James Darnley
+ *
+ * This file is part of FFmpeg.
+ *
+ * FFmpeg is free software; you can redistribute it and/or
+ * modify it under the terms of the GNU Lesser General Public
301 - 400 of 517 matches
Mail list logo
