- Jun 06, 2022
-
-
This is needed in the next commit.
-
RTP plugin weren't included
-
- Jun 05, 2022
-
-
- Jun 04, 2022
-
-
The PTS is a fraction of the length but should always be re-normalized by adding VLC_TICK_0 to be a valid timestamp. test_input_decoder_mock decoder: Wait for the picture to be flushed from the vout main input debug: control type=2 main input debug: ES_OUT_RESET_PCR called main input error: Invalid PCR value in ES_OUT_SET_(GROUP_)PCR !
-
The typical case is an application that use vgl callback with a GLX context, vaapi decoder will initiliaze properly and provide a hw_frames_ctx context, but due to the lack of interop, the decoder won't be usable and fallback to the sw decoder, but hw_frames_ctx will still be set and will cause a crash later on.
-
If the va module succeed to open but is identified as unusable before the end of FFmpeg getFormat callback, the module may have to do some cleanups in the AVCodecContext. the close callback will be called with a NULL AVCodecContext outside the getFormat callback as the AVCodecContext ownership is transferred to FFmpeg
-
In order to solve bad convergence of the initial jitter. This will soften the aout->time_get imprecision at the beginning. Thanks Denis Charmet for the idea. Ref #27023
-
Refs #27023
-
We can safely go down the code, it will return the same value as we did: - next is NULL - sys->displayed.date is VLC_TICK_INVALID since it's the first picture
-
They all return vlc_tick_now() + VOUT_REDISPLAY_DELAY eventually.
-
-
We can do the actual test directly
-
And keep the comment for now.
-
-
-
-
-
All other code pathes return before reaching this code.
-
And always return as render_now is always true here.
-
next cannot be NULL here.
-
It's in each section before.
-
To be removed next.
-
It's most likely wrong to display a forced picture right away. But for now the code is equivalent.
-
It's already true by default.
-
We return with VOUT_REDISPLAY_DELAY before that.
-
- We can't get in this code with next != NULL. - We can't get in this code with first, because next is NULL, we exited earlier
-
The code is logically equivalent. It will allow some simplifications later.
-
We can only reach this code with next = NULL, therefore first can only be false and refresh has its default value "false". So this test is always true.
-
date_refresh was always VLC_TICK_MAX
-
This will allow some simplifications
-
Fix a possible interrupt miss if vlc_credential_get() receive an interruption and if access don't check vlc_killed() before doing I/O. Refs #27027
-
Refs #27027
-
No functional changes. Refs #27027
-
- Jun 03, 2022
-
-
This refernces uint32_t without include, breaking this header if included before others.
-
Straightforward. Since pl_icc_profile_compute_signature is fast, we can call it on every frame without real worry. This is usually only relevant for still images anyways.
-
This is mainly designed to enable the use case brought forth by demux/image.c, which forwards this metadata to the decoded blocks, which in turn get forwarded through the rawvideo decoder.
-
It might be cleaner to somehow enumerate all ancillaries and forward them to the block. But for now, just do it for ICC profiles specifically.
-
Current libpng always has, and hopefully, always will return uncompressed ICC profiles, denoted with PNG_COMPRESSION_TYPE_BASE.
-
This implementation relies on the helper function jpeg_read_icc_profile which was introduced in libjpeg-turbo v1.5.90. Its absence from other versions of libjpeg means we may wish to copy this helper function into our own code in the future. But for now, keep it simple and avoid unnecessary bloat.
-
-