Re: [libav-devel] [PATCH 05/14] cbs_h265: Use helper macro for maximum values of fixed-width elements

2019-05-13 Thread Alexandra Hájková
On Sat, May 4, 2019 at 2:15 PM Luca Barbato wrote: > > From: Mark Thompson > > Apply the same logic as the previous patch to H.265. There are no cases > which currently overflow here, but this is still more consistent. > --- > libavcodec/cbs_h265_syntax_template.c | 16 > 1

Re: [libav-devel] [PATCH 04/14] avcodec/cbs: use a reference to the assembled CodedBitstreamFragment buffer when writing packets

2019-05-13 Thread Alexandra Hájková
On Sat, May 4, 2019 at 2:15 PM Luca Barbato wrote: > > From: James Almer > > This saves one malloc + memcpy per packet > > The CodedBitstreamFragment buffer is padded to follow the requirements > of AVPacket. > > Reviewed-by: jkqxz > Signed-off-by: James Almer > --- > libavcodec/cbs.c |

Re: [libav-devel] [PATCH 07/14] avcodec/cbs_mpeg2: use existing buffer reference when splitting fragments

2019-05-13 Thread Alexandra Hájková
On Sat, May 4, 2019 at 2:15 PM Luca Barbato wrote: > > From: James Almer > > Reviewed-by: Mark Thompson > Signed-off-by: James Almer > --- > libavcodec/cbs_mpeg2.c | 12 +++- > 1 file changed, 3 insertions(+), 9 deletions(-) > > diff --git a/libavcodec/cbs_mpeg2.c

Re: [libav-devel] [PATCH 08/14] avcodec/cbs_h2645: create a reference to the existing buffer when decomposing slice units

2019-05-13 Thread Alexandra Hájková
On Sat, May 4, 2019 at 2:15 PM Luca Barbato wrote: > > From: James Almer > > Reviewed-by: Mark Thompson > Signed-off-by: James Almer > --- > libavcodec/cbs_h2645.c | 18 -- > 1 file changed, 4 insertions(+), 14 deletions(-) > > diff --git a/libavcodec/cbs_h2645.c

Re: [libav-devel] [PATCH 02/14] cbs_h264: Fix overflow in shifts

2019-05-13 Thread Alexandra Hájková
On Sat, May 4, 2019 at 2:14 PM Luca Barbato wrote: > > From: Mark Thompson > > The type of the result of a shift operation is unaffected by the type of > the right operand, so some existing code overflows with undefined behaviour > when the element length is 32. Add a helper macro to calculate

Re: [libav-devel] [PATCH 03/14] cbs: Add a table of all supported codec IDs

2019-05-13 Thread Alexandra Hájková
On Sat, May 4, 2019 at 2:14 PM Luca Barbato wrote: > > From: Mark Thompson > > Use it as the set of codec IDs supported by the trace_headers BSF. > --- > configure | 2 +- > libavcodec/cbs.c | 13 + > libavcodec/cbs.h | 8 >

Re: [libav-devel] [PATCH 01/14] maint: Ignore dot dirs

2019-05-13 Thread Alexandra Hájková
On Sat, May 4, 2019 at 2:14 PM Luca Barbato wrote: > > They are usually created by tools and editors. > --- > .gitignore | 1 + > 1 file changed, 1 insertion(+) > > diff --git a/.gitignore b/.gitignore > index 1a08fd15c5..693fa5636a 100644 > --- a/.gitignore > +++ b/.gitignore > @@ -29,3 +29,4

Re: [libav-devel] [PATCH 06/14] avcodec/cbs_mpeg2: use memcpy when assembling fragments

2019-05-13 Thread Alexandra Hájková
On Sat, May 4, 2019 at 2:15 PM Luca Barbato wrote: > > From: James Almer > > before: > 419022 decicycles in assemble_fragment,2047 runs, 1 skips > > after: > 104621 decicycles in assemble_fragment,2045 runs, 3 skips > > Benched with a 2 minutes long 720x480 DVD mpeg2 sample. >

Re: [libav-devel] [PATCH 14/14] avcodec/cbs_h2645: use AVBufferRef to store list of active parameter sets

2019-05-13 Thread Alexandra Hájková
On Sat, May 4, 2019 at 2:16 PM Luca Barbato wrote: > > From: James Almer > > Removes unnecessary data copies, and partially fixes potential issues > with dangling references held in said lists. > > Reviewed-by: Mark Thompson > Signed-off-by: James Almer > --- > libavcodec/cbs_h264.h | 2 ++

Re: [libav-devel] [PATCH 13/14] cbs_h264: Need [] in the name when subscript is required

2019-05-13 Thread Alexandra Hájková
On Sat, May 4, 2019 at 2:15 PM Luca Barbato wrote: > > From: Haihao Xiang > > Otherwise it will hit an assert in the function > ff_cbs_trace_syntax_element() in cbs.c, line 400. > > Signed-off-by: Haihao Xiang > --- > libavcodec/cbs_h264_syntax_template.c | 2 +- > 1 file changed, 1

Re: [libav-devel] [PATCH 01/14] maint: Ignore dot dirs

2019-05-13 Thread Luca Barbato
On 13/05/2019 10:55, Diego Biurrun wrote: On Sat, May 04, 2019 at 02:14:14PM +0200, Luca Barbato wrote: They are usually created by tools and editors. --- a/.gitignore +++ b/.gitignore @@ -29,3 +29,4 @@ /coverage.info /lcov/ /mapfile +.*/ This ignores dotdirs everywhere, not just in the

Re: [libav-devel] [PATCH 10/14] cbs: Fragment/unit data is always reference counted

2019-05-13 Thread Alexandra Hájková
On Sat, May 4, 2019 at 2:15 PM Luca Barbato wrote: > > From: Mark Thompson > > Make this clear in the documentation and add some asserts to ensure > that it is always true. > --- > libavcodec/cbs.c | 19 --- > libavcodec/cbs.h | 10 ++ > 2 files changed, 18

Re: [libav-devel] [PATCH 09/14] avcodec/cbs_mpeg2: create a reference to the existing buffer when decomposing slice units

2019-05-13 Thread Alexandra Hájková
On Sat, May 4, 2019 at 2:15 PM Luca Barbato wrote: > > From: James Almer > > Reviewed-by: Mark Thompson > Signed-off-by: James Almer > --- > libavcodec/cbs_mpeg2.c | 9 ++--- > 1 file changed, 2 insertions(+), 7 deletions(-) > > diff --git a/libavcodec/cbs_mpeg2.c b/libavcodec/cbs_mpeg2.c

Re: [libav-devel] [PATCH 12/14] cbs: Add support for array subscripts in trace output

2019-05-13 Thread Alexandra Hájková
On Sat, May 4, 2019 at 2:15 PM Luca Barbato wrote: > > From: Mark Thompson > > This makes the trace output for arrays significantly nicer. > --- > libavcodec/cbs.c | 44 ++-- > libavcodec/cbs_h2645.c | 83 --- >

Re: [libav-devel] [PATCH 11/14] cbs_h2645: Simplify representation of fixed values

2019-05-13 Thread Alexandra Hájková
On Sat, May 4, 2019 at 2:15 PM Luca Barbato wrote: > > From: Mark Thompson > > --- > libavcodec/cbs_h2645.c| 5 + > libavcodec/cbs_h264_syntax_template.c | 30 --- > libavcodec/cbs_h265_syntax_template.c | 38 > +++ >

Re: [libav-devel] [PATCH 01/14] maint: Ignore dot dirs

2019-05-13 Thread Diego Biurrun
On Sat, May 04, 2019 at 02:14:14PM +0200, Luca Barbato wrote: > They are usually created by tools and editors. > --- a/.gitignore > +++ b/.gitignore > @@ -29,3 +29,4 @@ > /coverage.info > /lcov/ > /mapfile > +.*/ This ignores dotdirs everywhere, not just in the top-level directory, which seems