Thanks for your returns.
I'll split that into 2 patches and only enable it for vaapi indeed.
Mathieu
Le mar. 29 nov. 2016 à 04:11, Ronald S. Bultje a
écrit :
> Hi,
>
> On Mon, Nov 28, 2016 at 7:26 PM, Mark Thompson wrote:
>
> > On 28/11/16 21:22, Mathieu Velten wrote:
> > > ---
> > > libavco
As the title says: new assembly for the H.264 decoder. Many thanks to the
authors of the 4:2:0 functions. They were fairly easy to adapt after I
saw the pattern in the C, I just had to find it in the asm.
James Darnley (3):
avcodec/h264: mmxext 4:2:2 chroma intra deblock/loop filter
avcodec/
sse2:
complex: 4.13x faster (1514 vs. 367 cycles)
simple: 4.38x faster (1836 vs. 419 cycles)
avx:
complex: 1.07x faster (260 vs. 244 cycles)
simple: 1.03x faster (284 vs. 274 cycles)
---
libavcodec/x86/h264_idct_10bit.asm | 53 ++
libavcodec/x86/h264dsp_init.
2.87 times faster (1830 vs. 638 cycles)
---
libavcodec/x86/h264_idct.asm | 32
libavcodec/x86/h264dsp_init.c | 7 ++-
2 files changed, 38 insertions(+), 1 deletion(-)
diff --git a/libavcodec/x86/h264_idct.asm b/libavcodec/x86/h264_idct.asm
index 27222cb..c36
2.1 times faster (401 vs. 194 cycles)
---
libavcodec/x86/h264_deblock.asm | 14 ++
libavcodec/x86/h264dsp_init.c | 2 ++
2 files changed, 16 insertions(+)
diff --git a/libavcodec/x86/h264_deblock.asm b/libavcodec/x86/h264_deblock.asm
index 4aabbc0..fe0ab20 100644
--- a/libavcodec/x
On Mon, 28 Nov 2016 15:29:53 -0800
Wan-Teh Chang wrote:
> From: Anton Khirnov
>
> Adapted from the code by Rémi Denis-Courmont from VLC
>
> This merges libav commit c2755864afadfbaa349e8d583665c86fe99fa90b.
>
> Signed-off-by: Wan-Teh Chang
> ---
Patchset seems ok. No changes to the actual f
On Mon, Nov 28, 2016 at 16:29:01 -0500, Gregory J. Wolfe wrote:
> +#if OPENH264_VER_AT_LEAST(1, 6)
> +{ "load_balancing", "enable/disable load balancing",
> OFFSET(load_balancing), AV_OPT_TYPE_BOOL, { .i64 = 1 }, 0, 1, VE },
> +#endif
This could use a documentation update.
And since document
On Mon, Nov 28, 2016 at 10:03:37PM -0500, Vittorio Giovara wrote:
> On Mon, Nov 28, 2016 at 9:06 PM, Michael Niedermayer
> wrote:
> > On Tue, Nov 08, 2016 at 05:03:27PM -0500, Vittorio Giovara wrote:
> >> Signed-off-by: Vittorio Giovara
> >> ---
> >> Applied review.
> >> Please CC.
> >> Vittorio
2016-11-29 12:52 GMT+01:00 James Darnley :
> sse2:
> complex: 4.13x faster (1514 vs. 367 cycles)
> simple: 4.38x faster (1836 vs. 419 cycles)
>
> avx:
> complex: 1.07x faster (260 vs. 244 cycles)
> simple: 1.03x faster (284 vs. 274 cycles)
What are you comparing?
Carl Eugen
2016-11-29 0:29 GMT+01:00 Wan-Teh Chang :
> Adapted from the code by Rémi Denis-Courmont from VLC
Missing copyright statement, please do not commit as-is!
Carl Eugen
___
ffmpeg-devel mailing list
ffmpeg-devel@ffmpeg.org
http://ffmpeg.org/mailman/listin
On Tue, Nov 29, 2016 at 6:46 AM, Carl Eugen Hoyos wrote:
> 2016-11-29 0:29 GMT+01:00 Wan-Teh Chang :
>
>> Adapted from the code by Rémi Denis-Courmont from VLC
>
> Missing copyright statement, please do not commit as-is!
Hi Carl,
I made only the following changes to the Libav commits:
1. I chan
On Tue, Nov 29, 2016 at 7:20 AM, Michael Niedermayer
wrote:
> On Mon, Nov 28, 2016 at 10:03:37PM -0500, Vittorio Giovara wrote:
>> On Mon, Nov 28, 2016 at 9:06 PM, Michael Niedermayer
>> wrote:
>> > On Tue, Nov 08, 2016 at 05:03:27PM -0500, Vittorio Giovara wrote:
>> >> Signed-off-by: Vittorio Gi
On Tue, 29 Nov 2016 07:55:18 -0800
Wan-Teh Chang wrote:
> On Tue, Nov 29, 2016 at 6:46 AM, Carl Eugen Hoyos wrote:
> > 2016-11-29 0:29 GMT+01:00 Wan-Teh Chang :
> >
> >> Adapted from the code by Rémi Denis-Courmont from VLC
> >
> > Missing copyright statement, please do not commit as-is!
>
I'm not technically the maintainer of scale_npp, but this LGTM to me.
Will push with my next batch.
___
ffmpeg-devel mailing list
ffmpeg-devel@ffmpeg.org
http://ffmpeg.org/mailman/listinfo/ffmpeg-devel
(avctx->sample_aspect_ratio.num != 1 || avctx->sample_aspect_ratio.num
!= 1)) {
Damn, never noticed that typo.
Just fixing the typo should be fine as well, but I like the new logic
better so this LGTM and will push soon as well.
___
ffmpeg-devel mailing
On Sun, 27 Nov 2016, Nicolas George wrote:
Le tridi 3 frimaire, an CCXXV, Marton Balint a écrit :
I thought we are trying to move away from workarounds introduced only to be
able to be compatible with libav API. So this is clearly libav dirving
ffmpeg development, which is not fortunate IMHO.
On 2016-11-29 15:30, Carl Eugen Hoyos wrote:
> 2016-11-29 12:52 GMT+01:00 James Darnley :
>> sse2:
>> complex: 4.13x faster (1514 vs. 367 cycles)
>> simple: 4.38x faster (1836 vs. 419 cycles)
>>
>> avx:
>> complex: 1.07x faster (260 vs. 244 cycles)
>> simple: 1.03x faster (284 vs. 274 cycles)
>
On Tue, Nov 29, 2016 at 10:53:15AM -0500, Vittorio Giovara wrote:
> On Tue, Nov 29, 2016 at 7:20 AM, Michael Niedermayer
> wrote:
> > On Mon, Nov 28, 2016 at 10:03:37PM -0500, Vittorio Giovara wrote:
> >> On Mon, Nov 28, 2016 at 9:06 PM, Michael Niedermayer
> >> wrote:
> >> > On Tue, Nov 08, 2016
Both OK. Thanks.
--
Reynaldo H. Verdejo Pinochet
Open Source Group - Samsung Research America
___
ffmpeg-devel mailing list
ffmpeg-devel@ffmpeg.org
http://ffmpeg.org/mailman/listinfo/ffmpeg-devel
- This patch contains the changes to interface the Turing codec
(http://turingcodec.org/) to ffmpeg. The patch was modified to address
the comments in the review as follows:
- Added a pkg-config file to list all dependencies required by
libturing. This should address the issue pointed out b
Hi,
If there are no more comments can we look at merging this please?
Thanks
___
ffmpeg-devel mailing list
ffmpeg-devel@ffmpeg.org
http://ffmpeg.org/mailman/listinfo/ffmpeg-devel
-BEGIN PGP SIGNED MESSAGE-
Hash: SHA512
On 29/11/2016 17:18, Matteo Naccari wrote:
> [...]
Please send any later versions of this patch as a reply to the same
thread with a v3 (or later). For example (for this email): [PATCH v3]
Thanks,
- --
Josh
PGP fingerprint: A93A602D7A6D3C5388D79
I studied the history of vlc/include/vlc_atomic.h and compared it wth
libav/compat/atomics/win32/stdatomic.h.
1. vlc/include/vlc_atomic.h was initially written by Rémi
Denis-Courmont. This is the most substantial commit by Rémi
Denis-Courmont:
http://git.videolan.org/gitweb.cgi/vlc.git/?a=commit;
On Tue, 29 Nov 2016 17:44:15 +
Franklin Phillips wrote:
> Hi,
>
> If there are no more comments can we look at merging this please?
I think your previous post implied that subtitles are actually timed
correctly, right?
If so, I don't mind it being applied as-is.
___
On Tue, 29 Nov 2016 17:18:49 +
Matteo Naccari wrote:
> - This patch contains the changes to interface the Turing codec
> (http://turingcodec.org/) to ffmpeg. The patch was modified to address
> the comments in the review as follows:
> - Added a pkg-config file to list all dependencies r
On Tue, Nov 29, 2016 at 08:51:59AM -0800, Reynaldo H. Verdejo Pinochet wrote:
> Both OK. Thanks.
will push in a moment
thx
[...]
--
Michael GnuPG fingerprint: 9FF2128B147EF6730BADF133611EC787040B0FAB
Opposition brings concord. Out of discord comes the fairest harmony.
-- Heraclitus
signa
Hi,
On Mon, Nov 28, 2016 at 1:15 PM, Nicolas George wrote:
> Deadline: 2016-12-06 00:00 UTC.
>
> I propose, and put to the discussion, that the decision to drop ffserver
> is revoked, conditioned to the fixing of the technical issues that lead
> to it.
>
> In other words, if the technical proble
On Mon, Nov 28, 2016, at 07:29 AM, Andrey Utkin wrote:
> Dimensions of canvas drawtext produces vary depending on symbols in
> text, so add example for printing multiple texts aligned horizontally.
I don't really understand the use case. Can you explain?
___
On date Monday 2016-11-28 19:15:28 +0100, Nicolas George encoded:
> Deadline: 2016-12-06 00:00 UTC.
>
> I propose, and put to the discussion, that the decision to drop ffserver
> is revoked, conditioned to the fixing of the technical issues that lead
> to it.
>
> In other words, if the technical
Hi Stefano,
On Tue, Nov 29, 2016 at 1:46 PM, Stefano Sabatini
wrote:
> On date Monday 2016-11-28 19:15:28 +0100, Nicolas George encoded:
> > Deadline: 2016-12-06 00:00 UTC.
> >
> > I propose, and put to the discussion, that the decision to drop ffserver
> > is revoked, conditioned to the fixing
On Tue, Nov 29, 2016 at 09:31:08AM -0900, l...@lrcd.com wrote:
> On Mon, Nov 28, 2016, at 07:29 AM, Andrey Utkin wrote:
> > Dimensions of canvas drawtext produces vary depending on symbols in
> > text, so add example for printing multiple texts aligned horizontally.
>
> I don't really understand t
On 11/29/2016 3:46 PM, Stefano Sabatini wrote:
> On date Monday 2016-11-28 19:15:28 +0100, Nicolas George encoded:
>> Deadline: 2016-12-06 00:00 UTC.
>>
>> I propose, and put to the discussion, that the decision to drop ffserver
>> is revoked, conditioned to the fixing of the technical issues that
On Mon, 28 Nov 2016 19:15:28 +0100
Nicolas George wrote:
> Deadline: 2016-12-06 00:00 UTC.
>
> I propose, and put to the discussion, that the decision to drop ffserver
> is revoked, conditioned to the fixing of the technical issues that lead
> to it.
>
> In other words, if the technical problem
On Tue, Nov 29, 2016 at 07:10:52PM +0100, wm4 wrote:
> On Tue, 29 Nov 2016 17:44:15 +
> Franklin Phillips wrote:
>
> > Hi,
> >
> > If there are no more comments can we look at merging this please?
>
> I think your previous post implied that subtitles are actually timed
> correctly, right?
T
On Tue, Nov 29, 2016, at 09:55 AM, Andrey Utkin wrote:
>
> The point is that with more naive approach to printing parts of text,
> using just same vertical offset (code given below), it will look fine in
> some cases (e.g. all capital letters), but case of printing part with
> capital letter(s) and
On 11/28/2016 5:52 PM, Clément Bœsch wrote:
> On Mon, Nov 28, 2016 at 07:15:28PM +0100, Nicolas George wrote:
>> Deadline: 2016-12-06 00:00 UTC.
>>
>> I propose, and put to the discussion, that the decision to drop ffserver
>> is revoked, conditioned to the fixing of the technical issues that lead
On 11/29/16, James Almer wrote:
> On 11/28/2016 5:52 PM, Clement Boesch wrote:
>> On Mon, Nov 28, 2016 at 07:15:28PM +0100, Nicolas George wrote:
>>> Deadline: 2016-12-06 00:00 UTC.
>>>
>>> I propose, and put to the discussion, that the decision to drop ffserver
>>> is revoked, conditioned to the
2016-11-29 20:38 GMT+01:00 James Almer :
> Seeing Nicolas is apparently very invested in ffserver, can we expect him to
> maintain it, improve and extend it if it were to remain in the tree?
How is this related?
For which part of FFmpeg can we "expect" anybody to maintain it?
> Or is he just fig
2016-11-29 17:14 GMT+01:00 James Darnley :
> On 2016-11-29 15:30, Carl Eugen Hoyos wrote:
>> 2016-11-29 12:52 GMT+01:00 James Darnley :
>>> sse2:
>>> complex: 4.13x faster (1514 vs. 367 cycles)
>>> simple: 4.38x faster (1836 vs. 419 cycles)
>>>
>>> avx:
>>> complex: 1.07x faster (260 vs. 244 cycle
2016-11-28 19:15 GMT+01:00 Nicolas George :
> Deadline: 2016-12-06 00:00 UTC.
>
> I propose, and put to the discussion, that the decision to drop ffserver
> is revoked, conditioned to the fixing of the technical issues that lead
> to it.
Could you point me to this decision?
What were the reasons a
On 11/29/2016 5:04 PM, Carl Eugen Hoyos wrote:
> 2016-11-29 20:38 GMT+01:00 James Almer :
>
>> Seeing Nicolas is apparently very invested in ffserver, can we expect him to
>> maintain it, improve and extend it if it were to remain in the tree?
>
> How is this related?
> For which part of FFmpeg c
2016-11-29 21:11 GMT+01:00 James Almer :
> On 11/29/2016 5:04 PM, Carl Eugen Hoyos wrote:
>> 2016-11-29 20:38 GMT+01:00 James Almer :
>>
>>> Seeing Nicolas is apparently very invested in ffserver, can we expect him to
>>> maintain it, improve and extend it if it were to remain in the tree?
>>
>> Ho
On 11/29/2016 5:41 PM, Carl Eugen Hoyos wrote:
> 2016-11-29 21:11 GMT+01:00 James Almer :
>> On 11/29/2016 5:04 PM, Carl Eugen Hoyos wrote:
>>> 2016-11-29 20:38 GMT+01:00 James Almer :
>>>
Seeing Nicolas is apparently very invested in ffserver, can we expect him
to
maintain it, impr
On 11/29/16, Carl Eugen Hoyos wrote:
> 2016-11-29 21:11 GMT+01:00 James Almer :
>> On 11/29/2016 5:04 PM, Carl Eugen Hoyos wrote:
>>> 2016-11-29 20:38 GMT+01:00 James Almer :
>>>
Seeing Nicolas is apparently very invested in ffserver, can we expect
him to
maintain it, improve and ex
Hi!
Attached patch fixes aix compilation, similar to 33f5d70d
Please comment, Carl Eugen
From a959ae4347edb37dc342ef03c178a621a1c3a74a Mon Sep 17 00:00:00 2001
From: Carl Eugen Hoyos
Date: Tue, 29 Nov 2016 21:49:22 +0100
Subject: [PATCH] lavfi: Fix aix compilation.
Rename hz in af_apulsator.c a
Hi!
Attached patch fixes ticket #5982.
Please comment, Carl Eugen
From add7f5d51491152af6d0431331543212c2c21ca4 Mon Sep 17 00:00:00 2001
From: Carl Eugen Hoyos
Date: Tue, 29 Nov 2016 22:09:21 +0100
Subject: [PATCH] lavf/mov: Accept multiple fourcc for AVID 1:1.
Fixes ticket #5982.
---
libavfor
On 2016-11-29 21:09, Carl Eugen Hoyos wrote:
> 2016-11-29 17:14 GMT+01:00 James Darnley :
>> On 2016-11-29 15:30, Carl Eugen Hoyos wrote:
>>> 2016-11-29 12:52 GMT+01:00 James Darnley :
sse2:
complex: 4.13x faster (1514 vs. 367 cycles)
simple: 4.38x faster (1836 vs. 419 cycles)
Fixes Ticket3791
Signed-off-by: Michael Niedermayer
---
ffserver.c| 38 +++---
tests/ffserver-regression.sh | 2 +-
tests/ffserver.regression.ref | 2 +-
3 files changed, 37 insertions(+), 5 deletions(-)
diff --git a/ffserver.c b/ffserver.c
On 11/29/16, Carl Eugen Hoyos wrote:
> Hi!
>
> Attached patch fixes aix compilation, similar to 33f5d70d
>
> Please comment, Carl Eugen
>
should be ok
___
ffmpeg-devel mailing list
ffmpeg-devel@ffmpeg.org
http://ffmpeg.org/mailman/listinfo/ffmpeg-devel
On 29.11.2016 02:49, Michael Niedermayer wrote:
> On Tue, Nov 29, 2016 at 12:33:17AM +0100, Andreas Cadhalpun wrote:
>> Signed-off-by: Andreas Cadhalpun
>> ---
>> libavformat/avidec.c | 2 ++
>> 1 file changed, 2 insertions(+)
>
> If previous extradata is freed it should probably print a warning
On 29.11.2016 02:57, Michael Niedermayer wrote:
> On Tue, Nov 29, 2016 at 12:46:11AM +0100, Andreas Cadhalpun wrote:
>> The code calls av_new_packet a few lines above and the allocated memory
>> has to be freed in case of an error.
>
> should be ok
Pushed.
Best regards,
Andreas
On 28.11.2016 07:40, Schenk, Michael wrote:
> When using the decode interrupt feature of ffmpeg may causing crashes by
> accessing previous freed pointers in matroska_read_close.
>
> The attached patch will reset nb_elem to zero after freeing the elements
> because ffmpeg normally tests for nb_ele
On Wed, Nov 30, 2016 at 12:21:41AM +0100, Andreas Cadhalpun wrote:
> On 29.11.2016 02:49, Michael Niedermayer wrote:
> > On Tue, Nov 29, 2016 at 12:33:17AM +0100, Andreas Cadhalpun wrote:
> >> Signed-off-by: Andreas Cadhalpun
> >> ---
> >> libavformat/avidec.c | 2 ++
> >> 1 file changed, 2 inser
On 30.11.2016 01:29, Michael Niedermayer wrote:
> On Wed, Nov 30, 2016 at 12:21:41AM +0100, Andreas Cadhalpun wrote:
>> avidec.c |8
>> 1 file changed, 8 insertions(+)
>> 88b8c27a72e7ec4519e92eee5a3a2764a657a823
>> 0001-avidec-fix-leaking-extradata.patch
>> From e6a4fc92a99e2e378973
It randomly causes failures with an error like:
"Failed to set value '-f' for option 'd': Error number -920332800 occurred"
(The error number is different every time.)
Signed-off-by: Andreas Cadhalpun
---
tests/ffserver-regression.sh | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff -
This fixes producing swf and rm files as done by ffservertest.
Signed-off-by: Andreas Cadhalpun
---
ffserver.c| 2 +-
tests/ffserver.regression.ref | 8
2 files changed, 5 insertions(+), 5 deletions(-)
diff --git a/ffserver.c b/ffserver.c
index 20f5995..ffd75d4 1006
Not doing so makes debugging unnecessarily hard.
Signed-off-by: Andreas Cadhalpun
---
tests/ffserver-regression.sh | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/tests/ffserver-regression.sh b/tests/ffserver-regression.sh
index b776bad..47d6016 100755
--- a/tests/ffserver-re
From 20bedd18213420c77d5e8a26fbe741d8d204ac10 Mon Sep 17 00:00:00 2001
From: Jun Zhao
Date: Tue, 29 Nov 2016 14:14:25 +0800
Subject: [PATCH] lavc/vaapi_h26[45]: add crop info support in vaapi_h26[4,5]
add crop information support in vaapi_h26[4,5] hwaccel decode,
and align h264/hevc software deco
On Wed, Nov 30, 2016 at 02:27:12AM +0100, Andreas Cadhalpun wrote:
> Not doing so makes debugging unnecessarily hard.
>
> Signed-off-by: Andreas Cadhalpun
> ---
> tests/ffserver-regression.sh | 2 +-
> 1 file changed, 1 insertion(+), 1 deletion(-)
i agree, i did the same mysef locally
[...]
--
The entire set looks good to go. Please feel free to push.
Thank you!
On 11/29/2016 05:27 PM, Andreas Cadhalpun wrote:
Not doing so makes debugging unnecessarily hard.
Signed-off-by: Andreas Cadhalpun
---
tests/ffserver-regression.sh | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
Looks good to go. Thank you!
On 11/29/2016 01:38 PM, Michael Niedermayer wrote:
Fixes Ticket3791
Signed-off-by: Michael Niedermayer
---
ffserver.c| 38 +++---
tests/ffserver-regression.sh | 2 +-
tests/ffserver.regression.ref | 2 +-
61 matches
Mail list logo
