Signed-off-by: Tobias Rapp
---
libavformat/avienc.c | 2 --
libavformat/segment.c | 2 --
2 files changed, 4 deletions(-)
diff --git a/libavformat/avienc.c b/libavformat/avienc.c
index e8c0c71..da3d3de 100644
--- a/libavformat/avienc.c
+++ b/libavformat/avienc.c
@@ -19,8 +19,6 @@
* Foundation
On 7/7/17, Muhammad Faiz wrote:
> Avoid pthread_cond_broadcast that wakes up all workers. Make each of them
> uses distict mutex/cond.
>
What about video filters?
___
ffmpeg-devel mailing list
ffmpeg-devel@ffmpeg.org
http://ffmpeg.org/mailman/listinfo/f
On Thu, Jul 6, 2017 at 4:18 PM, Muhammad Faiz wrote:
> It is redundant with costable. The first half of sintable is
> identical with the second half of costable. The second half
> of sintable is negative value of the first half of sintable.
>
> The computation is changed to handle sign of sin valu
It is redundant with costable. The first half of sintable is
identical with the second half of costable. The second half
of sintable is negative value of the first half of sintable.
The computation is changed to handle sign of sin values, in
C code and ARM assembly code.
Signed-off-by: Muhammad F
On Fri, 7 Jul 2017 08:51:51 +0800
Jun Zhao wrote:
> On 2017/7/6 7:02, Mark Thompson wrote:
> > diff --git a/libavutil/hwcontext.h b/libavutil/hwcontext.h
> > index edf12cc631..fe7613b379 100644
> > --- a/libavutil/hwcontext.h
> > +++ b/libavutil/hwcontext.h
> > @@ -32,6 +32,7 @@ enum AVHWDeviceT
On Thu, 6 Jul 2017 22:59:24 +0100
Mark Thompson wrote:
> Hardware pixel formats do not tell you anything about their actual
> contents, but should still score higher than formats with completely
> unknown properties, which in turn should score higher than invalid
> formats.
>
> Do not return an
On Fri, Jul 7, 2017 at 2:45 PM, Paul B Mahol wrote:
> On 7/7/17, Muhammad Faiz wrote:
>> Avoid pthread_cond_broadcast that wakes up all workers. Make each of them
>> uses distict mutex/cond.
>>
>
> What about video filters?
./ffmpeg -filter_complex_threads 16 -filter_complex
"nullsrc=s=1920x1080
On Fri, Jul 07, 2017 at 09:25:46AM +0700, Muhammad Faiz wrote:
> Avoid pthread_cond_broadcast that wakes up all workers. Make each of them
> uses distict mutex/cond.
>
Please make sure this doesn't make tsan go crazy.
Also, the code is pretty much based on lavc/pthread_slice.c, so you may
want t
When 'out' is an AVFrame that does not have buffers preallocated,
swr_convert_frame tries to allocate buffers of the right size. However
in calculating this size it failed to check for whether 'in' is NULL
(requesting that swr's internal buffers are to be flushed).
---
libswresample/swresample_fra
On Fri, Jun 16, 2017 at 08:26:37PM +0900, DongHoon Kang wrote:
> Signed-off-by: DongHoon Kang
> ---
> libavcodec/htmlsubtitles.c | 14 +++---
> 1 file changed, 7 insertions(+), 7 deletions(-)
>
Rebased, updated for recent change, and applied.
Sorry for the delay, and thanks for the pa
Just some small observations... Even if you postpone this patch,
perhaps it can help improve the next one.
On Thu, Jul 06, 2017 at 17:36:39 -0700, Louis O'Bryan wrote:
> @@ -2,6 +2,7 @@ Entries are sorted chronologically from oldest to youngest
> within each release,
> releases are sorted from
On Thu, 6 Jul 2017 17:36:39 -0700
"Louis O'Bryan" wrote:
> From: Louis O'Bryan
>
> Signed-off-by: Louis O'Bryan
> ---
> Changelog | 1 +
> doc/general.texi| 2 +
> libavcodec/Makefile | 1 +
> libavcodec/allcodecs.c | 3 +
> libavcodec/avcodec.h| 1 +
Avoid pthread_cond_broadcast that wakes up all workers. Make each of them
uses distict mutex/cond. Also let main thread help running jobs.
Benchmark using afir with threads=5 and 4096 taps fir:
channels=1:
old:
1849650 decicycles in afir_execute, 2 runs, 0 skips
1525719 decicycl
On Fri, Jul 7, 2017 at 4:26 PM, Clément Bœsch wrote:
> On Fri, Jul 07, 2017 at 09:25:46AM +0700, Muhammad Faiz wrote:
>> Avoid pthread_cond_broadcast that wakes up all workers. Make each of them
>> uses distict mutex/cond.
>>
>
> Please make sure this doesn't make tsan go crazy.
Fixed. Post new p
On 7/7/2017 3:20 AM, Dave Rice wrote:
> Resolves https://trac.ffmpeg.org/ticket/6145 and writes a clap atom that is
> coincident with the frames width and height.
>
>
> From 23d80d0d47829fed61e817b1e7c3f6d420c9ab5c Mon Sep 17 00:00:00 2001
> From: Dave Rice
> Date: Thu, 6 Jul 2017 21:12:38 -040
On 7/7/2017 8:43 AM, Tobias Rapp wrote:
> Signed-off-by: Tobias Rapp
> ---
> libavformat/avienc.c | 2 --
> libavformat/segment.c | 2 --
> 2 files changed, 4 deletions(-)
LGTM.
- Derek
___
ffmpeg-devel mailing list
ffmpeg-devel@ffmpeg.org
http://ffm
On 7/6/2017 11:38 PM, Steven Liu wrote:
> Maybe there have no video stream, but have audio stream, so just give
> user a warning and continue?
Not an expert on the HDS code, but how can you fragment audio with no
min frag size set, and no video stream?
- Derek
On 7/7/2017 4:32 PM, Derek Buitenhuis wrote:
> Patch itself LGTM.
Just realized, this will change a bunch of FATE tests, won't it?
- Derek
___
ffmpeg-devel mailing list
ffmpeg-devel@ffmpeg.org
http://ffmpeg.org/mailman/listinfo/ffmpeg-devel
Signed-off-by: Paul B Mahol
---
libavcodec/get_bits.h | 205 +++---
1 file changed, 196 insertions(+), 9 deletions(-)
diff --git a/libavcodec/get_bits.h b/libavcodec/get_bits.h
index c530015..8a9021a 100644
--- a/libavcodec/get_bits.h
+++ b/libavcodec/
Signed-off-by: Paul B Mahol
---
libavcodec/utvideodec.c | 1 +
1 file changed, 1 insertion(+)
diff --git a/libavcodec/utvideodec.c b/libavcodec/utvideodec.c
index 40c1277..c57d727 100644
--- a/libavcodec/utvideodec.c
+++ b/libavcodec/utvideodec.c
@@ -27,6 +27,7 @@
#include
#include
+#defin
Signed-off-by: Paul B Mahol
---
doc/libav-merge.txt | 1 -
1 file changed, 1 deletion(-)
diff --git a/doc/libav-merge.txt b/doc/libav-merge.txt
index 690c4ef..56d6109 100644
--- a/doc/libav-merge.txt
+++ b/doc/libav-merge.txt
@@ -96,7 +96,6 @@ Stuff that didn't reach the codebase:
- e7078e842
> Am 07.07.2017 um 20:48 schrieb Paul B Mahol :
>
> Signed-off-by: Paul B Mahol
> ---
> libavcodec/utvideodec.c | 1 +
> 1 file changed, 1 insertion(+)
>
> diff --git a/libavcodec/utvideodec.c b/libavcodec/utvideodec.c
> index 40c1277..c57d727 100644
> --- a/libavcodec/utvideodec.c
> +++ b/lib
Hi,
On Fri, Jul 7, 2017 at 2:48 PM, Paul B Mahol wrote:
> typedef struct GetBitContext {
> const uint8_t *buffer, *buffer_end;
> +#ifdef CACHED_BITSTREAM_READER
> +uint64_t cache;
> +unsigned bits_left;
> +#endif
>
Can you post some stats (from relevant systems, ideally, e.g. 32-b
On 7/7/17, Ronald S. Bultje wrote:
> Hi,
>
> On Fri, Jul 7, 2017 at 2:48 PM, Paul B Mahol wrote:
>
>> typedef struct GetBitContext {
>> const uint8_t *buffer, *buffer_end;
>> +#ifdef CACHED_BITSTREAM_READER
>> +uint64_t cache;
>> +unsigned bits_left;
>> +#endif
>>
>
> Can you post s
Hi,
On Fri, Jul 7, 2017 at 4:43 PM, Paul B Mahol wrote:
> On 7/7/17, Ronald S. Bultje wrote:
> > (I'm assuming the low-level interface no longer works with the cached
> > reader, so can we prevent users from accessing these macros unless
> > cached=1?)
>
> They are not supposed to work, Do you
Add the debug_threads boolean field to FrameThreadContext. The
debug_threads field records whether the FF_DEBUG_THREADS bit is set in
the debug field of the avctx passed to ff_frame_thread_init(). The
debug_threads field is set when avcodec_open2() is called, never changes
thereafter, and applies t
On Wed, Jul 5, 2017 at 3:30 PM Marton Balint wrote:
>
> On Fri, 9 Jun 2017, Devin Heitmueller wrote:
>
> > Hello Marton,
> >
> > On Tue, Jun 6, 2017 at 5:45 PM, Marton Balint wrote:
> >
> >> As far as I remember multiple side data of the same type is not
> something we
> >> wanted to support. Wh
On 7/7/2017 12:32 PM, Derek Buitenhuis wrote:
> On 7/7/2017 3:20 AM, Dave Rice wrote:
>> Resolves https://trac.ffmpeg.org/ticket/6145 and writes a clap atom that is
>> coincident with the frames width and height.
>>
>>
>> From 23d80d0d47829fed61e817b1e7c3f6d420c9ab5c Mon Sep 17 00:00:00 2001
>> Fr
Hi Ronald,
On Wed, Jul 5, 2017 at 7:24 PM, Ronald S. Bultje wrote:
> Hi Wan-Teh,
>
> On Wed, Jul 5, 2017 at 8:08 PM, Wan-Teh Chang wrote:
>>
>> Hi Ronald,
>>
>> On Wed, Jul 5, 2017 at 3:31 PM, Ronald S. Bultje
>> wrote:
>> > Hi Wan-Teh,
>> >
>> > On Wed, Jul 5, 2017 at 6:12 PM, Wan-Teh Chang w
On Thu, Jul 06, 2017 at 07:28:32PM +0100, Derek Buitenhuis wrote:
> Signed-off-by: Derek Buitenhuis
> ---
> libavformat/hdsenc.c | 1 +
> 1 file changed, 1 insertion(+)
>
> diff --git a/libavformat/hdsenc.c b/libavformat/hdsenc.c
> index 347df83a51..c362a75d8b 100644
> --- a/libavformat/hdsenc.c
On Fri, Jul 07, 2017 at 02:25:20AM +0100, Derek Buitenhuis wrote:
> On 7/7/2017 1:36 AM, Louis O'Bryan wrote:
> > I am adding a new encoder for camera sensor metadata. This is an
> > implementation of a not-yet-published open standard for adding camera
> > sensor data to mp4 containers, including t
Note: I suspect we can simply delete the following line from
update_context_from_user() in libavcodec/pthread_frame.c:
dst->debug= src->debug;
That also fixes the tsan warning, but it'll take more time to
investigate whether it is necessary to update the |debug| field from
the user's AVCo
Hi,
On Fri, Jul 7, 2017 at 5:30 PM, Wan-Teh Chang
wrote:
> Note: I suspect we can simply delete the following line from
> update_context_from_user() in libavcodec/pthread_frame.c:
>
> dst->debug= src->debug;
>
> That also fixes the tsan warning, but it'll take more time to
> investigate
On Fri, Jul 07, 2017 at 09:43:22AM +0200, Tobias Rapp wrote:
> Signed-off-by: Tobias Rapp
> ---
> libavformat/avienc.c | 2 --
> libavformat/segment.c | 2 --
> 2 files changed, 4 deletions(-)
>
> diff --git a/libavformat/avienc.c b/libavformat/avienc.c
> index e8c0c71..da3d3de 100644
> --- a/l
I don't know what you are trying to achieve, but an encoder that just copies
things through makes no sense.
If it's just about validating or logging some kind of parser or
(bitstream-)filter probably makes more sense.
Even if doing it as encoder, why should it be a video encoder instead of
somet
Thanks for the feedback.
I don't think AVFrame.pkt_size should be used for any real purpose.
> It's a hack and doesn't really have anything to do with AVFrame data
> (the idea is that it corresponds to the source packet, but it's only a
> hint at best).
The AVFrame format is a raw byte stream wit
On Sat, Jul 8, 2017 at 12:24 AM, Louis O'Bryan
wrote:
>
> The use case is to write data to a new stream in the mp4 container. The
> encoder isn't changing the data. This data would reside in the same mp4
> container as video and audio streams. Are you suggesting there is a way I
> can accomplish t
On 7/7/2017 10:13 PM, James Almer wrote:
> Isn't this necessary only for files with raw video? As is, this box
> would be written on any mov file with a video stream.
This was addressed a previous email:
http://lists.ffmpeg.org/pipermail/ffmpeg-devel/2017-July/213350.html
I guess the spec is
This also add supports for 4K DeckLink cards because they always output the
ancillary data in 10-bit.
v2:
- only try teletext decoding for 576i PAL mode
- some comments as requested by Aaron Levinson
Signed-off-by: Marton Balint
---
doc/indevs.texi | 4 +--
libavdevice/decklink_de
It sounds like my code should be outside an encoder then. I think this
stems from my misunderstanding of how the API can be used. I will rewrite
my code that muxes streams to use an AVPacket directly instead.
On Fri, Jul 7, 2017 at 3:47 PM, Hendrik Leppkes wrote:
> On Sat, Jul 8, 2017 at 12:24 A
v2:
- use uint16_t instead of int to store 10-bit ancillary data
- fix ancillary line numbers for 1080p
- some comments and clarifications as requested by Aaron Levinson
Signed-off-by: Marton Balint
---
doc/indevs.texi | 20 ---
libavdevice/decklink_dec.cpp | 127 ++
On Fri, Jul 07, 2017 at 09:04:37PM +0700, Muhammad Faiz wrote:
> Avoid pthread_cond_broadcast that wakes up all workers. Make each of them
> uses distict mutex/cond. Also let main thread help running jobs.
>
> Benchmark using afir with threads=5 and 4096 taps fir:
> channels=1:
> old:
> 184965
On Fri, Jul 07, 2017 at 08:48:46PM +0200, Paul B Mahol wrote:
> Signed-off-by: Paul B Mahol
> ---
> libavcodec/get_bits.h | 205
> +++---
> 1 file changed, 196 insertions(+), 9 deletions(-)
>
> diff --git a/libavcodec/get_bits.h b/libavcodec/get_bits.
On Fri, Jul 07, 2017 at 10:54:47AM +0100, hexpointer wrote:
> When 'out' is an AVFrame that does not have buffers preallocated,
> swr_convert_frame tries to allocate buffers of the right size. However
> in calculating this size it failed to check for whether 'in' is NULL
> (requesting that swr's in
On Sat, Jul 8, 2017 at 6:45 AM, Michael Niedermayer
wrote:
> On Fri, Jul 07, 2017 at 09:04:37PM +0700, Muhammad Faiz wrote:
>> Avoid pthread_cond_broadcast that wakes up all workers. Make each of them
>> uses distict mutex/cond. Also let main thread help running jobs.
>>
>> Benchmark using afir wi
On Sat, Jul 08, 2017 at 09:06:05AM +0700, Muhammad Faiz wrote:
> On Sat, Jul 8, 2017 at 6:45 AM, Michael Niedermayer
> wrote:
> > On Fri, Jul 07, 2017 at 09:04:37PM +0700, Muhammad Faiz wrote:
> >> Avoid pthread_cond_broadcast that wakes up all workers. Make each of them
> >> uses distict mutex/co
46 matches
Mail list logo
