[FFmpeg-devel] [PATCH] avcodec/jpeg2000: use powf instead of pow

It may be entirely possible to optimize further by exploiting the fact that it is always an integer exponent. However, care must be taken since the base itself is not an integer, i.e a naive repeated squaring or similar such algorithm may be inaccurate. Range information can only be obtained by kn

Re: [FFmpeg-devel] [PATCH] swscale/utils: replace pow by exp2

On Tue, Dec 8, 2015 at 1:43 PM, Michael Niedermayer wrote: > On Mon, Dec 07, 2015 at 11:07:17PM -0500, Ganesh Ajjanagadde wrote: >> exp2 is a faster function. >> >> Signed-off-by: Ganesh Ajjanagadde >> --- >> libswscale/utils.c | 3 ++- >> 1 file changed, 2 insertions(+), 1 deletion(-) > > LGTM

Re: [FFmpeg-devel] [PATCH] avcodec/jpeg2000: replace naive pow call with smarter exp2fi

On Tue, Dec 8, 2015 at 1:49 PM, Michael Niedermayer wrote: > On Mon, Dec 07, 2015 at 09:50:49PM -0500, Ganesh Ajjanagadde wrote: >> pow is a very wasteful function for this purpose. A low hanging fruit >> would be simply to replace with exp2f, and that does yield some speedup. >> However, there ar

Re: [FFmpeg-devel] [PATCH] avfilter/formats: fix segfault when allocation fails

On Sat, Dec 5, 2015 at 2:43 PM, Ganesh Ajjanagadde wrote: > This is a somewhat subtle failure that can occur when the realloc_array > fails in FORMATS_REF. > > Signed-off-by: Ganesh Ajjanagadde > --- > libavfilter/formats.c | 2 +- > 1 file changed, 1 insertion(+), 1 deletion(-) > > diff --git a

Re: [FFmpeg-devel] Ideas to replace the options system

I like the idea of JSON syntax. To avoid escaping problems simply read the arguments out of a file instead of off the command line. Suggestion -vf '@/path/to/filter-syntax' to read filter syntax from a file -vf '-' to read filter syntax from STDIN ___

Re: [FFmpeg-devel] [PATCH 03/11] diracdec: Rename lowdelay_subband to decode_subband because it is shared with HQ profile

On Wed, Dec 09, 2015 at 12:05:29AM +, Kieran Kunhya wrote: > --- > libavcodec/diracdec.c | 18 -- > 1 file changed, 8 insertions(+), 10 deletions(-) should be ok [...] -- Michael GnuPG fingerprint: 9FF2128B147EF6730BADF133611EC787040B0FAB The real ebay dictionary, page

Re: [FFmpeg-devel] [PATCH 09/11] diracdec: Store version major/minor flags

On Wed, Dec 09, 2015 at 12:05:35AM +, Kieran Kunhya wrote: > --- > libavcodec/dirac.c | 15 +++ > 1 file changed, 7 insertions(+), 8 deletions(-) LGTM thanks [...] -- Michael GnuPG fingerprint: 9FF2128B147EF6730BADF133611EC787040B0FAB No human being will ever know the Trut

[FFmpeg-devel] [PATCH] [RFC] avcodec/h264_slice: Duplicate addition to avoid ubsan warning

From: Michael Niedermayer Fixes mozilla bug 1230247 Signed-off-by: Michael Niedermayer --- libavcodec/h264_slice.c | 21 - 1 file changed, 12 insertions(+), 9 deletions(-) diff --git a/libavcodec/h264_slice.c b/libavcodec/h264_slice.c index bbadfc9..cc82875 100644 --- a/

[FFmpeg-devel] [PATCH] diracdec: Fix FPE on invalid low_delay data

--- libavcodec/diracdec.c | 5 + 1 file changed, 5 insertions(+) diff --git a/libavcodec/diracdec.c b/libavcodec/diracdec.c index 6a53f38..0542ad7 100644 --- a/libavcodec/diracdec.c +++ b/libavcodec/diracdec.c @@ -2043,6 +2043,11 @@ static int dirac_decode_data_unit(AVCodecContext *avctx, co

[FFmpeg-devel] Bug #3823 (yuvj422p jpeg rtp enc) request for help

Hi! Could please anybody look at MJPEG RTP encoder issue with yuvj422p https://trac.ffmpeg.org/ticket/3823#comment:17 ? This pixel format becomes popular for IP cameras having no RTSP, and there are constantly issues with RTPizing it. I consider donation for fixing this. Thanks. -- Bluecherry de

[FFmpeg-devel] [PATCH 08/11] diracdec: Support new extended quantiser range

--- libavcodec/diracdec.c | 58 --- 1 file changed, 37 insertions(+), 21 deletions(-) diff --git a/libavcodec/diracdec.c b/libavcodec/diracdec.c index c42360d..fdcca01 100644 --- a/libavcodec/diracdec.c +++ b/libavcodec/diracdec.c @@ -57,7 +57,7 @@

[FFmpeg-devel] [PATCH 10/11] diracdec: Read picture types by using parse_code

--- libavcodec/diracdec.c | 57 +-- 1 file changed, 42 insertions(+), 15 deletions(-) diff --git a/libavcodec/diracdec.c b/libavcodec/diracdec.c index fdcca01..9c09d51 100644 --- a/libavcodec/diracdec.c +++ b/libavcodec/diracdec.c @@ -154,7 +154,11

[FFmpeg-devel] [PATCH 11/11] diracdec: Add support for HQ profile

--- libavcodec/diracdec.c | 119 +- 1 file changed, 80 insertions(+), 39 deletions(-) diff --git a/libavcodec/diracdec.c b/libavcodec/diracdec.c index 9c09d51..6a53f38 100644 --- a/libavcodec/diracdec.c +++ b/libavcodec/diracdec.c @@ -187,6 +187,11

[FFmpeg-devel] [PATCH 06/11] diracdec: Add 10-bits to pix_fmt table

--- libavcodec/dirac.c | 15 ++- 1 file changed, 6 insertions(+), 9 deletions(-) diff --git a/libavcodec/dirac.c b/libavcodec/dirac.c index fd6e869..6b1a93b 100644 --- a/libavcodec/dirac.c +++ b/libavcodec/dirac.c @@ -109,10 +109,11 @@ static const struct { { AVCOL_PRI_BT709,

[FFmpeg-devel] [PATCH 05/11] diracdec: Extract version parameters

--- libavcodec/dirac.c | 1 + libavcodec/dirac.h | 6 ++ libavcodec/diracdec.c | 3 ++- libavformat/oggparsedirac.c | 3 ++- 4 files changed, 11 insertions(+), 2 deletions(-) diff --git a/libavcodec/dirac.c b/libavcodec/dirac.c index aa82dd9..fd6e869 100644 --- a/libav

[FFmpeg-devel] [PATCH 07/11] diracdec: Replace dirac parse codes with better ones

--- libavcodec/dirac.h| 22 ++ libavcodec/diracdec.c | 18 +++--- 2 files changed, 25 insertions(+), 15 deletions(-) diff --git a/libavcodec/dirac.h b/libavcodec/dirac.h index 6852e82..b4982f8 100644 --- a/libavcodec/dirac.h +++ b/libavcodec/dirac.h @@ -34,6 +3

[FFmpeg-devel] [PATCH 09/11] diracdec: Store version major/minor flags

--- libavcodec/dirac.c | 15 +++ 1 file changed, 7 insertions(+), 8 deletions(-) diff --git a/libavcodec/dirac.c b/libavcodec/dirac.c index 6b1a93b..9bf0fd4 100644 --- a/libavcodec/dirac.c +++ b/libavcodec/dirac.c @@ -294,22 +294,21 @@ int avpriv_dirac_parse_sequence_header(AVCodecCon

[FFmpeg-devel] [PATCH 02/11] diracdec: Template DSP functions adding 10-bit versions

--- libavcodec/dirac_dwt.c | 530 ++- libavcodec/dirac_dwt.h | 20 +- libavcodec/diracdec.c| 8 +- libavcodec/diracdsp.c| 25 +- libavcodec/diracdsp.h| 4 +- libavcodec/x86/dirac_dwt.c | 65 +++

[FFmpeg-devel] [PATCH 01/11] diracdec: Move strides to bytes, and pointer types to uint8_t.

Start templating functions for move to support 10-bit Parts of this patch were written by Rostislav Pehlivanov --- libavcodec/dirac.c | 10 +- libavcodec/dirac.h | 3 +- libavcodec/diracdec.c | 253 +++- libavformat/oggparsedirac.c

[FFmpeg-devel] [PATCH 04/11] diracdec: Make slice parameters common between lowdelay and future hq profile

--- libavcodec/diracdec.c | 36 +++- 1 file changed, 19 insertions(+), 17 deletions(-) diff --git a/libavcodec/diracdec.c b/libavcodec/diracdec.c index d2a4bac..fca722d 100644 --- a/libavcodec/diracdec.c +++ b/libavcodec/diracdec.c @@ -168,14 +168,15 @@ typedef str

[FFmpeg-devel] [PATCH 03/11] diracdec: Rename lowdelay_subband to decode_subband because it is shared with HQ profile

--- libavcodec/diracdec.c | 18 -- 1 file changed, 8 insertions(+), 10 deletions(-) diff --git a/libavcodec/diracdec.c b/libavcodec/diracdec.c index 4c02259..d2a4bac 100644 --- a/libavcodec/diracdec.c +++ b/libavcodec/diracdec.c @@ -719,11 +719,9 @@ static void decode_component(Di

Re: [FFmpeg-devel] [PATCH 2/2] diracdec: Add HQ Profile support

On Tue, Dec 08, 2015 at 07:27:42PM +, Kieran Kunhya wrote: > From: Rostislav Pehlivanov > > --- > libavcodec/dirac.c | 31 ++-- > libavcodec/dirac.h | 28 > libavcodec/dirac_dwt.h | 2 +- > libavcodec/diracdec.c | 357 > +++--

Re: [FFmpeg-devel] [PATCH] aacenc: update max_sfb when num_swb changes

On 08.12.2015 22:00, Rostislav Pehlivanov wrote: > LGTM > > Thanks, sorry I took so long to see the email > Push when you can Pushed. Maybe you can also have a look at the other problem I mentioned [1]. Best regards, Andreas 1: https://ffmpeg.org/pipermail/ffmpeg-devel/2015-December/184683.htm

Re: [FFmpeg-devel] [PATCH 1/2] diracdec: Move strides to bytes, and pointer types to uint8_t. Also template functions for move to support 10-bit Parts of this patch were written by Rostislav Pehlivano

On Tue, Dec 08, 2015 at 09:49:19PM +, Kieran Kunhya wrote: > > Is this intended to be in this or a subsequent patch ? > > Meant to be in this patch because 10-bit isn't specific to HQ profile. > Or am I missing something? no, just thought this one was just intended to be cosmetic and restruct

Re: [FFmpeg-devel] [PATCH 2/2] diracdec: Add HQ Profile support

> Can this patch be split into independant changes ? > > it seems some parts are reindented differntly and some fields are > moved like s->lowdelay.num_* -> s->num_* > > this should make it easier to review I could do but the majority of the patch is still the addition of HQ profile... Kieran ___

Re: [FFmpeg-devel] [PATCH 1/2] diracdec: Move strides to bytes, and pointer types to uint8_t. Also template functions for move to support 10-bit Parts of this patch were written by Rostislav Pehlivano

> Is this intended to be in this or a subsequent patch ? Meant to be in this patch because 10-bit isn't specific to HQ profile. Or am I missing something? Kieran ___ ffmpeg-devel mailing list ffmpeg-devel@ffmpeg.org http://ffmpeg.org/mailman/listinfo/ffm

Re: [FFmpeg-devel] [PATCH 1/2] diracdec: Move strides to bytes, and pointer types to uint8_t. Also template functions for move to support 10-bit Parts of this patch were written by Rostislav Pehlivano

On Tue, Dec 08, 2015 at 07:27:41PM +, Kieran Kunhya wrote: [...] > +if (bit_depth == 8) > +ret = ff_spatial_idwt_init2_8bit(d, buffer, width, height, stride, > type, decomposition_count, temp); > +else if (bit_depth == 10) > +ret = ff_spatial_idwt_init2_10bit(d, buffer

Re: [FFmpeg-devel] [PATCH 2/2] diracdec: Add HQ Profile support

On Tue, Dec 08, 2015 at 07:27:42PM +, Kieran Kunhya wrote: [...] > @@ -1057,29 +1129,19 @@ static int dirac_unpack_idwt_params(DiracContext *s) > > CHECKEDREAD(s->wavelet_depth, tmp > MAX_DWT_LEVELS || tmp < 1, "invalid > number of DWT decompositions\n") > > -if (!s->low_delay) {

Re: [FFmpeg-devel] [PATCH] aacenc: update max_sfb when num_swb changes

LGTM Thanks, sorry I took so long to see the email Push when you can On 6 December 2015 at 21:33, Andreas Cadhalpun < andreas.cadhal...@googlemail.com> wrote: > This fixes out-of-bounds reads in avoid_clipping. > > Signed-off-by: Andreas Cadhalpun > --- > libavcodec/aacenc.c | 1 + > 1 file ch

Re: [FFmpeg-devel] [PATCH 08/10] avfilter/vf_colorkey: avoid unnecessary floating point multiply

On Tue, Dec 8, 2015 at 1:52 PM, Michael Niedermayer wrote: > On Mon, Dec 07, 2015 at 06:07:03PM -0500, Ganesh Ajjanagadde wrote: >> On Sun, Nov 22, 2015 at 12:05 PM, Ganesh Ajjanagadde >> wrote: >> > Normalization can be pulled outside the sqrt. >> > >> > Signed-off-by: Ganesh Ajjanagadde >> > -

Re: [FFmpeg-devel] [PATCH 3/3] fate/api-jpeg-codec-param: rename to api-mjpeg-codec-param

On Tue, Dec 08, 2015 at 07:00:43PM +0100, Michael Niedermayer wrote: > On Tue, Dec 08, 2015 at 08:10:11AM +0100, Matthieu Bouron wrote: > > On Mon, Dec 7, 2015 at 11:41 PM, Michael Niedermayer > > wrote: > > > > > On Mon, Dec 07, 2015 at 10:42:39AM +0100, Matthieu Bouron wrote: > > > > From: Matt

[FFmpeg-devel] [PATCH 1/2] diracdec: Move strides to bytes, and pointer types to uint8_t. Also template functions for move to support 10-bit Parts of this patch were written by Rostislav Pehlivanov

--- libavcodec/dirac.c | 10 +- libavcodec/dirac.h | 3 +- libavcodec/dirac_dwt.c | 530 ++--- libavcodec/dirac_dwt.h | 20 +- libavcodec/dirac_dwt_template.c | 615 +++ libavcodec/

[FFmpeg-devel] [PATCH 2/2] diracdec: Add HQ Profile support

From: Rostislav Pehlivanov --- libavcodec/dirac.c | 31 ++-- libavcodec/dirac.h | 28 libavcodec/dirac_dwt.h | 2 +- libavcodec/diracdec.c | 357 +++- libavformat/oggparsedirac.c | 3 +- 5 files changed, 265 insertio

Re: [FFmpeg-devel] [PATCH 08/10] avfilter/vf_colorkey: avoid unnecessary floating point multiply

On Mon, Dec 07, 2015 at 06:07:03PM -0500, Ganesh Ajjanagadde wrote: > On Sun, Nov 22, 2015 at 12:05 PM, Ganesh Ajjanagadde > wrote: > > Normalization can be pulled outside the sqrt. > > > > Signed-off-by: Ganesh Ajjanagadde > > --- > > libavfilter/vf_colorkey.c | 2 +- > > 1 file changed, 1 inse

Re: [FFmpeg-devel] [PATCH] avcodec/jpeg2000: replace naive pow call with smarter exp2fi

On Mon, Dec 07, 2015 at 09:50:49PM -0500, Ganesh Ajjanagadde wrote: > pow is a very wasteful function for this purpose. A low hanging fruit > would be simply to replace with exp2f, and that does yield some speedup. > However, there are 2 drawbacks of this: > 1. It does not exploit the integer natur

Re: [FFmpeg-devel] [PATCH] swscale/utils: replace pow by exp2

On Mon, Dec 07, 2015 at 11:07:17PM -0500, Ganesh Ajjanagadde wrote: > exp2 is a faster function. > > Signed-off-by: Ganesh Ajjanagadde > --- > libswscale/utils.c | 3 ++- > 1 file changed, 2 insertions(+), 1 deletion(-) LGTM thx [...] -- Michael GnuPG fingerprint: 9FF2128B147EF6730BADF13

Re: [FFmpeg-devel] [PATCH] lavf/img2enc: add atomic_writing option

On Tue, Dec 08, 2015 at 10:47:23AM +0100, Clément Bœsch wrote: > From: Clément Bœsch > > This behaviour change caused a regression on our side recently, we might > want to disable the option by default. > --- > libavformat/img2enc.c | 4 +++- > 1 file changed, 3 insertions(+), 1 deletion(-) sho

Re: [FFmpeg-devel] [PATCH 3/3] fate/api-jpeg-codec-param: rename to api-mjpeg-codec-param

On Tue, Dec 08, 2015 at 08:10:11AM +0100, Matthieu Bouron wrote: > On Mon, Dec 7, 2015 at 11:41 PM, Michael Niedermayer > wrote: > > > On Mon, Dec 07, 2015 at 10:42:39AM +0100, Matthieu Bouron wrote: > > > From: Matthieu Bouron > > > > > > --- > > > tests/fate/api.mak | 6 +++--- > > > 1 file c

Re: [FFmpeg-devel] [PATCH 3/3] fate/api-jpeg-codec-param: rename to api-mjpeg-codec-param

On 12/8/2015 4:10 AM, Matthieu Bouron wrote: > On Mon, Dec 7, 2015 at 11:41 PM, Michael Niedermayer > wrote: > >> On Mon, Dec 07, 2015 at 10:42:39AM +0100, Matthieu Bouron wrote: >>> From: Matthieu Bouron >>> >>> --- >>> tests/fate/api.mak | 6 +++--- >>> 1 file changed, 3 insertions(+), 3 dele

Re: [FFmpeg-devel] [PATCHv2] avcodec/h264: Set CORRUPT flag on output frames that are not fully recovered

On Tue, Dec 08, 2015 at 10:24:09AM +0200, Sebastian Dröge wrote: > From: Sebastian Dröge > > In the merge commit 78265fcfeee153e5e26ad4dbc7831a84ade447d6 this behaviour > was broken and the CORRUPT flag would never ever be set on a frame. However > the flag on the AVCodecContext was taken into ac

Re: [FFmpeg-devel] [PATCH 2/2] configure: libbsd support for arc4random()

On Tue, Dec 8, 2015 at 9:01 AM, Ganesh Ajjanagadde wrote: > On Tue, Dec 8, 2015 at 8:16 AM, Clément Bœsch wrote: >> On Tue, Dec 08, 2015 at 07:34:51AM -0500, Ganesh Ajjanagadde wrote: >>> On Tue, Dec 8, 2015 at 7:27 AM, wm4 wrote: >>> > On Sun, 6 Dec 2015 22:56:33 -0500 >>> > Ganesh Ajjanagadde

Re: [FFmpeg-devel] [PATCH 2/2] configure: libbsd support for arc4random()

On Tue, Dec 8, 2015 at 8:16 AM, Clément Bœsch wrote: > On Tue, Dec 08, 2015 at 07:34:51AM -0500, Ganesh Ajjanagadde wrote: >> On Tue, Dec 8, 2015 at 7:27 AM, wm4 wrote: >> > On Sun, 6 Dec 2015 22:56:33 -0500 >> > Ganesh Ajjanagadde wrote: >> > >> >> On non-BSD machines, there exists a package l

Re: [FFmpeg-devel] [PATCH 2/2] configure: libbsd support for arc4random()

On Tue, Dec 08, 2015 at 07:34:51AM -0500, Ganesh Ajjanagadde wrote: > On Tue, Dec 8, 2015 at 7:27 AM, wm4 wrote: > > On Sun, 6 Dec 2015 22:56:33 -0500 > > Ganesh Ajjanagadde wrote: > > > >> On non-BSD machines, there exists a package libbsd for providing BSD > >> functionality. This can be used

Re: [FFmpeg-devel] [PATCH] lavf/segment: add option to write empty filler segments as needed

On date Monday 2015-12-07 14:24:44 -0600, Rodger Combs encoded: > --- > libavformat/segment.c | 22 +++--- > 1 file changed, 15 insertions(+), 7 deletions(-) Missing docs. > diff --git a/libavformat/segment.c b/libavformat/segment.c > index 8432d0f..acd266b 100644 > --- a/libavf

Re: [FFmpeg-devel] [PATCH 2/2] configure: libbsd support for arc4random()

On Tue, Dec 8, 2015 at 7:27 AM, wm4 wrote: > On Sun, 6 Dec 2015 22:56:33 -0500 > Ganesh Ajjanagadde wrote: > >> On non-BSD machines, there exists a package libbsd for providing BSD >> functionality. This can be used to get support for arc4random. >> >> Thus, an opt-in --enable-libbsd is added to

Re: [FFmpeg-devel] [PATCH 2/2] configure: libbsd support for arc4random()

On Sun, 6 Dec 2015 22:56:33 -0500 Ganesh Ajjanagadde wrote: > On non-BSD machines, there exists a package libbsd for providing BSD > functionality. This can be used to get support for arc4random. > > Thus, an opt-in --enable-libbsd is added to configure for this > functionality. > > Tested on

[FFmpeg-devel] [PATCH 2/3] rtpdec_jpeg: use framesize from SDP if present

This enables us to process pics larger than 2040 pixels in dimensions, overcoming the limitation of RFC 2435. Signed-off-by: Andrey Utkin --- libavformat/rtpdec_jpeg.c | 46 +++--- 1 file changed, 43 insertions(+), 3 deletions(-) diff --git a/libavformat/

[FFmpeg-devel] [PATCH 3/3] Don't forget to run SDP parsing callback for static RTP types

Signed-off-by: Andrey Utkin --- libavformat/rtsp.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/libavformat/rtsp.c b/libavformat/rtsp.c index 39539e9..4c0a057 100644 --- a/libavformat/rtsp.c +++ b/libavformat/rtsp.c @@ -567,7 +567,7 @@ static void sdp_parse_line(AVFormatCon

[FFmpeg-devel] [PATCH 1/3] sdp: print a=framesize for MJPEG

Signed-off-by: Andrey Utkin --- libavformat/sdp.c | 2 ++ 1 file changed, 2 insertions(+) diff --git a/libavformat/sdp.c b/libavformat/sdp.c index 2ab37a8..27112ef 100644 --- a/libavformat/sdp.c +++ b/libavformat/sdp.c @@ -657,6 +657,8 @@ static char *sdp_write_media_attributes(char *buff, int

[FFmpeg-devel] [PATCH 0/3] Workaround for large JPEG over RTP

RFCs for JPEG over RTP don't support dimensions larger than 2040 pixels, but it is possible to work around if we stick with information from SDP. Andrey Utkin (3): sdp: print a=framesize for MJPEG rtpdec_jpeg: use framesize from SDP if present Don't forget to run SDP parsing callback for sta

[FFmpeg-devel] [PATCH] avfilter: add swaprect filter

Signed-off-by: Paul B Mahol --- doc/filters.texi | 54 ++ libavfilter/Makefile | 1 + libavfilter/allfilters.c | 1 + libavfilter/vf_swaprect.c | 267 ++ 4 files changed, 323 insertions(+) create mode 100644 libavfilter/vf_s

Re: [FFmpeg-devel] Memory leaks in libavformat/segment.c

On 12/8/15 7:05 PM, Hendrik Leppkes wrote: On Tue, Dec 8, 2015 at 11:46 AM, Lvqier wrote: On 12/8/15 6:26 PM, Hendrik Leppkes wrote: On Tue, Dec 8, 2015 at 11:19 AM, Lvqier wrote: Hi Hendrik, The attached is another patch which uses av_reallocp to fix this issue. Thanks, that looks

Re: [FFmpeg-devel] Memory leaks in libavformat/segment.c

On Tue, Dec 8, 2015 at 11:46 AM, Lvqier wrote: > On 12/8/15 6:26 PM, Hendrik Leppkes wrote: >> >> On Tue, Dec 8, 2015 at 11:19 AM, Lvqier wrote: >>> >>> Hi Hendrik, >>> >>> The attached is another patch which uses av_reallocp to fix this >>> issue. >>> >> Thanks, that looks almost good. But

Re: [FFmpeg-devel] Memory leaks in libavformat/segment.c

On 12/8/15 6:26 PM, Hendrik Leppkes wrote: On Tue, Dec 8, 2015 at 11:19 AM, Lvqier wrote: Hi Hendrik, The attached is another patch which uses av_reallocp to fix this issue. Thanks, that looks almost good. But you should check for a negative return value, as all errors are negative. Do

Re: [FFmpeg-devel] Ideas to replace the options system

On 12/4/15, Nicolas George wrote: > Hi. > > This is a rather long explanation on ideas I have to replace the options > system with something better. I will not work on it before I have made > significant progress on de-recursiving lavfi, but I can still think of it > and mature its design while wa

Re: [FFmpeg-devel] Memory leaks in libavformat/segment.c

On Tue, Dec 8, 2015 at 11:19 AM, Lvqier wrote: > Hi Hendrik, > > The attached is another patch which uses av_reallocp to fix this issue. > Thanks, that looks almost good. But you should check for a negative return value, as all errors are negative. - Hendrik _

Re: [FFmpeg-devel] Memory leaks in libavformat/segment.c

Hi Hendrik, The attached is another patch which uses av_reallocp to fix this issue. /Best Regards, lvqier - lvq...@gmail.com / ** 青春如烟，唱一首笑忘歌 On 12/5/15 11:23 PM, Lvqier wrote: Ok, I will try av_reallocp /Best Regards, lvq

[FFmpeg-devel] [PATCH] lavf/img2enc: add atomic_writing option

From: Clément Bœsch This behaviour change caused a regression on our side recently, we might want to disable the option by default. --- libavformat/img2enc.c | 4 +++- 1 file changed, 3 insertions(+), 1 deletion(-) diff --git a/libavformat/img2enc.c b/libavformat/img2enc.c index c2e6e4d..929180

[FFmpeg-devel] HEVC DSP and x86 MC SIMD improvements from Libav

Hey guys, Libav pushed some changes to the HEVC DSP and a new set of x86 MC SIMD, which of course conflicts with our version of it. From a quick cursory look, their implementation seems to have a bunch of advantages, namely 32-bit support in many functions, and SSE2 implementations while ours is

Re: [FFmpeg-devel] [PATCH] avcodec/h264: Set CORRUPT flag on output frames that are not fully recovered

On Di, 2015-12-08 at 01:28 +0100, Michael Niedermayer wrote: >  > this breaks "-flags2 showall", many (ugly) frames are no longer shown > one sample showing this is: Indeed, I changed behaviour here. It should work the same as before now. Before and after my patch there is no difference between A

[FFmpeg-devel] [PATCHv2] avcodec/h264: Set CORRUPT flag on output frames that are not fully recovered

From: Sebastian Dröge In the merge commit 78265fcfeee153e5e26ad4dbc7831a84ade447d6 this behaviour was broken and the CORRUPT flag would never ever be set on a frame. However the flag on the AVCodecContext was taken into account properly, including AV_CODEC_FLAG2_SHOW_ALL. The reason for this was

Re: [FFmpeg-devel] Ideas to replace the options system

On 05.12.2015 00:54, James Darnley wrote: On 2015-12-04 15:33, Nicolas George wrote: Why do we need a new options system? Most importantly: escaping hell OMG yes! I have seen several times the amount of backslashes Windows users are forced to use to provide a path to some of the filters.