mbedtls_x509_crt_parse_file returns an error with negative numbers, and
positive numbers indicate the number of failed certificates to load from
certificate specific issues, such as critical extensions.
This would fix ticket #11079.
Signed-off-by: Mohit Gupta
---
libavformat/tls_mbedtls.c |
On Tue, Jul 16, 2024 at 07:11:42PM +0200, Anton Khirnov wrote:
> Frame threading in the FFV1 decoder works in a very unusual way - the
> state that needs to be propagated from the previous frame is not decoded
> pixels(¹), but each slice's entropy coder state after decoding the slice.
>
> For that
On Tue, Jul 16, 2024 at 07:11:41PM +0200, Anton Khirnov wrote:
> It is now only called from a single place, so there is no point in it
> being a separate function.
> ---
> libavcodec/ffv1dec.c | 38 --
> 1 file changed, 16 insertions(+), 22 deletions(-)
ok
thx
On Tue, Jul 16, 2024 at 07:11:40PM +0200, Anton Khirnov wrote:
> All remaining accesses to them are for fields that have the same value
> in the main encoder context.
>
> Drop now-unused FFV1Context.slice_contexts.
> ---
> libavcodec/ffv1.c | 18 +++-
> libavcodec/ffv1.h
On Tue, Jul 16, 2024 at 07:11:39PM +0200, Anton Khirnov wrote:
> ---
> libavcodec/ffv1.h| 2 +-
> libavcodec/ffv1dec.c | 18 --
> 2 files changed, 9 insertions(+), 11 deletions(-)
LGTM
thx
[...]
--
Michael GnuPG fingerprint: 9FF2128B147EF6730BADF133611EC787040B0FAB
Ob
On Tue, Jul 16, 2024 at 07:11:38PM +0200, Anton Khirnov wrote:
> ---
> libavcodec/ffv1.h| 16 +---
> libavcodec/ffv1dec.c | 4 ++--
> 2 files changed, 15 insertions(+), 5 deletions(-)
LGTM
thx
[...]
--
Michael GnuPG fingerprint: 9FF2128B147EF6730BADF133611EC787040B0FAB
I
On Tue, Jul 16, 2024 at 07:11:37PM +0200, Anton Khirnov wrote:
> All remaining accesses to them are for fields that have the same value
> in the main encoder context.
> ---
> libavcodec/ffv1enc.c | 57 ---
> libavcodec/ffv1enc_template.c | 24 +++---
Le keskiviikkona 24. heinäkuuta 2024, 16.08.17 EEST Michael Niedermayer a
écrit :
> The question is not if its "warning-worthy", the question is what action
> creates the best FFmpeg. And also an enviroment we are all happy to work in.
That is complete nonsense. It is more obvious than patently o
On Tue, Jul 16, 2024 at 07:11:36PM +0200, Anton Khirnov wrote:
> ---
> libavcodec/ffv1.h| 2 +-
> libavcodec/ffv1dec.c | 6 +++---
> libavcodec/ffv1enc.c | 6 +++---
> 3 files changed, 7 insertions(+), 7 deletions(-)
LGTM
thx
[...]
--
Michael GnuPG fingerprint: 9FF2128B147EF6730BADF133
On Tue, Jul 16, 2024 at 07:11:35PM +0200, Anton Khirnov wrote:
> Instead of the per-slice FFV1Context, which will be removed in future
> commits.
> ---
> libavcodec/ffv1.c | 6 ++
> libavcodec/ffv1.h | 3 +++
> libavcodec/ffv1enc.c | 20 ++--
>
On Tue, Jul 16, 2024 at 07:11:34PM +0200, Anton Khirnov wrote:
> ---
> libavcodec/ffv1.c | 9 ++--
> libavcodec/ffv1.h | 5 +--
> libavcodec/ffv1dec.c | 72
> libavcodec/ffv1dec_template.c | 2 +-
> libavcodec/ffv1enc.c
On Tue, Jul 16, 2024 at 07:11:32PM +0200, Anton Khirnov wrote:
> It cannot change between slices.
that is true currently but considering that its used in some of the inner loops
i dont think carrying an additional pointer around into these loops makes
sense
either leave it or move the ac value to
On Tue, Jul 16, 2024 at 07:11:31PM +0200, Anton Khirnov wrote:
> ---
> libavcodec/ffv1.h | 6 +++---
> libavcodec/ffv1dec.c | 14 +++---
> libavcodec/ffv1dec_template.c | 10 +-
> libavcodec/ffv1enc.c | 26 +-
> libavcodec/ffv1
On Tue, Jul 16, 2024 at 07:11:30PM +0200, Anton Khirnov wrote:
> They cannot change between slices.
> ---
> libavcodec/ffv1.c | 2 --
> libavcodec/ffv1dec.c | 5 ++---
> libavcodec/ffv1dec_template.c | 2 +-
> libavcodec/ffv1enc.c | 2 +-
> libavcodec/ffv1enc_template
On Tue, Jul 16, 2024 at 07:11:29PM +0200, Anton Khirnov wrote:
> ---
> libavcodec/ffv1.c | 3 ---
> libavcodec/ffv1.h | 1 -
> 2 files changed, 4 deletions(-)
ok
thx
[...]
--
Michael GnuPG fingerprint: 9FF2128B147EF6730BADF133611EC787040B0FAB
Let us carefully observe those good qualities
On Tue, Jul 16, 2024 at 07:11:26PM +0200, Anton Khirnov wrote:
> ---
> libavcodec/ffv1.h | 3 ++-
> libavcodec/ffv1enc.c | 7 ---
> libavcodec/ffv1enc_template.c | 14 +++---
> 3 files changed, 13 insertions(+), 11 deletions(-)
LGTM
thx
[...]
--
Michael G
On Tue, Jul 16, 2024 at 07:11:23PM +0200, Anton Khirnov wrote:
> ---
> libavcodec/ffv1.c | 35 ---
> libavcodec/ffv1.h | 5 +++--
> libavcodec/ffv1dec.c | 27 +--
> libavcodec/ffv1dec_template.c | 9 +---
On Tue, Jul 16, 2024 at 07:11:22PM +0200, Anton Khirnov wrote:
> FFV1 decoder and encoder currently use the same struct - FFV1Context -
> both as codec private data and per-slice context. For this purpose
> FFV1Context contains an array of pointers to per-slice FFV1Context
> instances.
>
> This pa
On Tue, Jul 16, 2024 at 07:11:21PM +0200, Anton Khirnov wrote:
> It is unused otherwise
> ---
> libavcodec/ffv1dec.c | 2 +-
> 1 file changed, 1 insertion(+), 1 deletion(-)
ok
thx
[...]
--
Michael GnuPG fingerprint: 9FF2128B147EF6730BADF133611EC787040B0FAB
Frequently ignored answer#1 FFmp
On Tue, Jul 16, 2024 at 07:11:20PM +0200, Anton Khirnov wrote:
> fsdst is by construction always equal to fs, there is even an
> av_assert1() checking that. Just use fs directly.
> ---
> libavcodec/ffv1dec.c | 8 +++-
> 1 file changed, 3 insertions(+), 5 deletions(-)
probably ok
thx
[...]
-
On Tue, Jul 16, 2024 at 07:11:48PM +0200, Anton Khirnov wrote:
> Reorganize the code such that the frame threading code does not call the
> decoders directly, but instead calls back into the generic decoding
> code. This avoids duplicating the logic that wraps the decoder
> invocation and allows re
On Tue, Jul 16, 2024 at 07:11:19PM +0200, Anton Khirnov wrote:
> It is merely a pointer to FFV1Context.picture.f, which can just as well
> be used directly.
> ---
> libavcodec/ffv1.h| 1 -
> libavcodec/ffv1dec.c | 33 +
> 2 files changed, 17 insertions(+), 17 d
On Tue, Jul 16, 2024 at 07:11:18PM +0200, Anton Khirnov wrote:
> ---
> libavcodec/ffv1dec.c | 5 +
> 1 file changed, 1 insertion(+), 4 deletions(-)
ok
thx
[...]
--
Michael GnuPG fingerprint: 9FF2128B147EF6730BADF133611EC787040B0FAB
Complexity theory is the science of finding the exact
On Tue, Jul 16, 2024 at 07:11:17PM +0200, Anton Khirnov wrote:
> ---
> libavcodec/ffv1dec.c | 45 ++--
> 1 file changed, 22 insertions(+), 23 deletions(-)
LGTM
[...]
--
Michael GnuPG fingerprint: 9FF2128B147EF6730BADF133611EC787040B0FAB
Into a blind
On Tue, Jul 16, 2024 at 07:11:16PM +0200, Anton Khirnov wrote:
> ---
> tests/fate/vcodec.mak| 3 ++-
> tests/ref/vsynth/vsynth1-ffv1-v2 | 4
> tests/ref/vsynth/vsynth2-ffv1-v2 | 4
> tests/ref/vsynth/vsynth3-ffv1-v2 | 4
> tests/ref/vsynth/vsynth_lena-ffv
Callers of ff_framesync_get_frame() generally do not expect the result
to be writable, those that do (e.g. ff_framesync_dualinput_get_writable())
ensure writability themselves.
Significantly reduces memory consumption in complex graphs with
framesync-based filters (e.g. scale, ssim).
Reported-By:
On Wed, Jul 24, 2024 at 06:49:04AM +0200, Marvin Scholz (ePirat) wrote:
>
> > On 24. Jul 2024, at 00:10, Michael Niedermayer
> > wrote:
> >
> > On Tue, Jul 23, 2024 at 12:17:43PM -0300, James Almer wrote:
> >>> On 7/19/2024 12:31 PM, Paul B Mahol wrote:
> >>> Internal/private filter structures
On Wed, Jul 24, 2024 at 07:09:45AM +, patches via ffmpeg-devel wrote:
>
>
> -Original Message-
> From: Michael Niedermayer
> Sent: Tuesday, July 23, 2024 10:22 PM
> To: FFmpeg development discussions and patches
> Cc: patches
> Subject: Re: [FFmpeg-devel] [PATCH] avdevice/dshow: D
On Tue, 23 Jul 2024, Michael Niedermayer wrote:
On Wed, Jul 24, 2024 at 12:01:32AM +0300, Martin Storsjö via ffmpeg-devel wrote:
Increase the tolerance from 10 ulp to 11 ulp. This fixes occasional
errors for some inputs; the errors could be reproduced on
aarch64/neon builds, with "checkasm --te
-Original Message-
From: Michael Niedermayer
Sent: Tuesday, July 23, 2024 10:22 PM
To: FFmpeg development discussions and patches
Cc: patches
Subject: Re: [FFmpeg-devel] [PATCH] avdevice/dshow: Don't skip audio devices if
no video device is present
On Tue, Jul 23, 2024 at 08:21:31AM
30 matches
Mail list logo
