Commit graph

1368 commits

Author SHA1 Message Date
Romain Vimont
f1b2d6bbbb Bump version to 2.0 2023-03-12 02:05:37 +01:00
Romain Vimont
05a55e3687 Happy new year 2023! 2023-03-12 02:04:58 +01:00
Romain Vimont
affda26bfa Document audio player
Add some high-level documentation on the audio player implementation.
2023-03-12 02:04:58 +01:00
Romain Vimont
0bf866fa8d Apply new compensation only if it changed
If the compensation is the same (typically when it is 0), do not reapply
it.
2023-03-12 02:04:58 +01:00
Romain Vimont
73727e7fdf Disable clock drift compensation for tiny values
For less than 1ms, the estimated drift is just noise.
2023-03-12 02:04:58 +01:00
Romain Vimont
c22c87eded Fail on deprecated options
Suggest the video and audio specific options instead.
2023-03-12 02:04:58 +01:00
Romain Vimont
426dfbf21d Remove dead code about the deprecated -F option
The -F option was already removed.
2023-03-12 02:04:58 +01:00
Romain Vimont
5512777404 Remove deprecated option --render-expired-frames
This option did nothing since it was deprecated. Totally remove it.
2023-03-12 02:04:58 +01:00
Romain Vimont
cc07f8dac4 Upgrade platform-tools (34.0.1) for Windows
Include the latest version of adb in Windows releases.
2023-03-12 02:04:58 +01:00
Romain Vimont
f5bb9e576d Upgrade SDL (2.26.4) for Windows
Include the latest version of SDL in Windows releases.
2023-03-12 02:04:58 +01:00
Romain Vimont
eca8766545 Compute buffering and compensation without lock
Once underflow has been read with a lock, the buffering and compensation
may be performed without shared variables.
2023-03-12 02:04:58 +01:00
Romain Vimont
0b8a5ca923 Do not read avg_buffering from the player thread
On buffer underflow, the average buffering must be updated, but it is
intended to be accessed only from the receiver thread.

Make the player and the receiver thread communicate the underflow via a
new field (ap->underflow).
2023-03-12 02:04:58 +01:00
Romain Vimont
e06acc1ba2 Simplify bytebuf naming
Rename read_available to can_read and write_available to can_write.
This is more readable.
2023-03-12 02:04:57 +01:00
Romain Vimont
14f9d82fda Add audio sample ring-buffer
Add a thin wrapper around bytebuf to handle samples instead of bytes.
This simplifies the audio player, which mostly handles samples.
2023-03-12 02:04:07 +01:00
Romain Vimont
bb509d9317 Define the audio output buffer in milliseconds
In theory, this buffer must be dimensioned for a target duration, so its
size in bytes should depend on the sample rate.
2023-03-11 15:57:25 +01:00
Romain Vimont
238ab872ba Pass video size as codec metadata
On initial connection, scrcpy sent some device metadata:
 - the device name (to be used as window title)
 - the initial video size (before any frame or even SPS/PPS)

But it is better to provide the initial video size as part as the video
stream, so that it can be demuxed and exposed via AVCodecContext to
sinks.

This avoids to pass an explicit "initial frame size" for the screen, the
recorder and the v4l2 sink.
2023-03-11 15:57:25 +01:00
Romain Vimont
3a72f3fb4d Report errors on screen event error
Make scrcpy fail if an important screen event (like frame update) fails.
2023-03-11 15:57:25 +01:00
Romain Vimont
be985b8242 Copy codec parameters from context
Now that the recorder have access to the codec context, it may
automatically initialize the stream codec parameters.

The V4L2 sink could do the same.
2023-03-11 15:57:25 +01:00
Romain Vimont
a9f6001f51 Simplify recorder
After the refactor performed by the previous commit, the functions to
wait the video stream and the audio stream could be inlined.
2023-03-11 15:57:25 +01:00
Romain Vimont
5052e15f7f Create recorder streams from packet sinks ops
Previously, the packet sink push() implementation just set the codec and
notified a wait condition. Then the recorder thread read the codec and
created the AVStream.

But this was racy: an AVFrame could be pushed before the creation of the
AVStream, causing its video_stream_index or audio_stream_index to be
initialized to -1.

Also, in the future, the AVStream initialization might need data
provided by the packet sink open(), so initialize it there (with a
mutex).
2023-03-11 15:57:25 +01:00
Romain Vimont
4bdf632dfa Pass AVCodecContext to packet sinks
Create the codec context from the demuxer, so that it can fill context
data for the decoder and recorder.
2023-03-11 15:57:25 +01:00
Romain Vimont
4db50ddbb7 Enable log signaling buffering threshold exceeded
It is as important as underflow logs.
2023-03-11 15:55:44 +01:00
Romain Vimont
46f6918179 Stop and join sc_file_pusher only if initialized
The sc_file_pusher is lazy-initialized, but it was stopped and joined in
all cases (accessing uninitialized values).

Detected by poisoning the struct scrcpy instance with ASAN enabled.
2023-03-10 22:45:48 +01:00
chengjian.scj
d93582724d Initialize interrupted field explicitly
The field sc_fps_counter.interrupted was never initialized explicitly.

Signed-off-by: Romain Vimont <rom@rom1v.com>
2023-03-10 22:33:52 +01:00
Romain Vimont
408f458636 Decrease recorder thread priority
Recording is background task, writing the packets to a file is not
urgent.
2023-03-10 22:22:16 +01:00
Romain Vimont
aa450ffc3f Increase audio thread priority
The audio demuxer thread is the one filling the audio buffer read by the
SDL audio thread. It is time critical to avoid buffer underflow.
2023-03-10 22:22:16 +01:00
Romain Vimont
5ee59e0f13 Add thread priority API
Expose an API to change the priority of the current thread.
2023-03-10 22:22:16 +01:00
Romain Vimont
4a25f3e53b Print info logs to stdout
All server logs were printed to stdout, while all client logs were
printed to stderr.

Instead, use stderr for warnings and errors, stdout for the others:
 - stdout: verbose, debug, info
 - stderr: warn, error
2023-03-10 22:22:15 +01:00
Romain Vimont
7da45c246e Warn on ignored audio options
For raw audio codec, some audio options are ignored.

PR #3757 <https://github.com/Genymobile/scrcpy/pull/3757>
2023-03-10 22:22:15 +01:00
Romain Vimont
d2952c7e93 Add --audio-codec=raw option
Add support for raw (PCM S16 LE) audio codec (a raw decoder is included
in FFmpeg).

PR #3757 <https://github.com/Genymobile/scrcpy/pull/3757>
2023-03-10 22:22:15 +01:00
Romain Vimont
02dd1be4a1 Stop on decoder frame push error
On push, frame sinks report downstream errors to stop upstream
components. Do not ignore the error.
2023-03-10 22:22:15 +01:00
Romain Vimont
df55bc2683 Add --audio-buffer
Expose an option to add a buffering delay (in milliseconds) before
playing audio.

This is similar to the options --display-buffer and --v4l2-buffer for
video frames.

PR #3757 <https://github.com/Genymobile/scrcpy/pull/3757>
2023-03-10 22:22:15 +01:00
Romain Vimont
d66b0b3dcc Add compat support for FFmpeg < 5.1
The new chlayout API has been introduced in FFmpeg 5.1. Use the old
channel_layout API on older versions.

PR #3757 <https://github.com/Genymobile/scrcpy/pull/3757>
2023-03-10 22:22:15 +01:00
Romain Vimont
fbe0f951e1 Add audio player
Play the decoded audio using SDL.

The audio player frame sink receives the audio frames, resample them
and write them to a byte buffer (introduced by this commit).

On SDL audio callback (from an internal SDL thread), copy samples from
this byte buffer to the SDL audio buffer.

The byte buffer is protected by the SDL_AudioDeviceLock(), but it has
been designed so that the producer and the consumer may write and read
in parallel, provided that they don't access the same slices of the
ring-buffer buffer.

PR #3757 <https://github.com/Genymobile/scrcpy/pull/3757>

Co-authored-by: Simon Chan <1330321+yume-chan@users.noreply.github.com>
2023-03-10 22:22:15 +01:00
Romain Vimont
e1333f6f3b Optionally do not delay the first frame
A delay buffer delayed all the frames except the first one, to open the
scrcpy window immediately and get a picture.

Make this feature optional, so that the delay buffer might also be used
for audio (especially for simulating a high delay for debugging).

PR #3757 <https://github.com/Genymobile/scrcpy/pull/3757>
2023-03-10 22:22:15 +01:00
Romain Vimont
9b3ca208bf Accept clock estimation with a single point
If there is only one point, assume the slope is 1.

PR #3757 <https://github.com/Genymobile/scrcpy/pull/3757>
2023-03-10 22:22:15 +01:00
Romain Vimont
48a537d45c Remove anonymous struct in delay buffer
For clarity, the fields used only when a delay was set were wrapped in
an anonymous structure.

Now that the delay buffer has been extracted to a separate component,
the delay is necessarily set (it may not be 0), so the fields are always
used.

PR #3757 <https://github.com/Genymobile/scrcpy/pull/3757>
2023-03-10 22:22:15 +01:00
Romain Vimont
1230149fdd Use delay buffer as a frame source/sink
The components needing delayed frames (sc_screen and sc_v4l2_sink)
managed a sc_video_buffer instance, which itself embedded a
sc_frame_buffer instance (to keep only the most recent frame).

In theory, these components should not be aware of delaying: they should
just receive AVFrames later, and only handle a sc_frame_buffer.

Therefore, refactor sc_delay_buffer as a frame source (it consumes)
frames) and a frame sink (it produces frames, after some delay), and
plug an instance in the pipeline only when a delay is requested.

This also removes the need for a specific sc_video_buffer.

PR #3757 <https://github.com/Genymobile/scrcpy/pull/3757>
2023-03-10 22:22:15 +01:00
Romain Vimont
974227a3fc Use frame source trait in decoder 2023-03-10 22:22:15 +01:00
Romain Vimont
6543964f12 Introduce frame source trait
There was a frame sink trait, implemented by components able to receive
AVFrames, but each frame source had to manually send frame to sinks.

In order to mutualise sink management, add a frame sink trait.
2023-03-10 22:22:15 +01:00
Romain Vimont
f3197e178d Use packet source trait in demuxer 2023-03-10 22:22:15 +01:00
Romain Vimont
c39054a63d Introduce packet source trait
There was a packet sink trait, implemented by components able to
receive AVPackets, but each packet source had to manually send packets
to sinks.

In order to mutualise sink management, add a packet source trait.
2023-03-10 22:22:15 +01:00
Romain Vimont
f410f2bdc4 Extract sc_delay_buffer
A video buffer had 2 responsibilities:
 - handle the frame delaying mechanism (queuing packets and pushing them
   after the expected delay);
 - keep only the most recent frame (using a sc_frame_buffer).

In order to be able to reuse only the frame delaying mechanism, extract
it to a separate component, sc_delay_buffer.
2023-03-10 22:22:15 +01:00
Romain Vimont
6379c08012 Fix buffering pts conversion
The mistake had no effect, because tick is also internally expressed in
microseconds.
2023-03-10 22:22:15 +01:00
Romain Vimont
4540f1d69e Report video buffer downstream errors
Make the video buffer stop if its consumer could not receive a frame.
2023-03-10 22:22:15 +01:00
Romain Vimont
ad94ccca0b Stop the video buffer on error
If an error occurs from the video buffer thread (typically an
out-of-memory error), then stop.
2023-03-10 22:22:15 +01:00
Romain Vimont
a3703340fc Fix possible race condition on video_buffer end
The video_buffer thread clears the queue once it is stopped, but new
frames might still be pushed asynchronously.

To avoid the problem, do not push any frame once the video_buffer is
stopped.
2023-03-10 22:22:15 +01:00
Romain Vimont
6f38c6311b Remove sc_queue
All uses have been replaced by VecDeque.
2023-03-10 22:22:15 +01:00
Romain Vimont
338310677e Remove cbuf
All uses have been replaced by VecDeque.
2023-03-10 22:22:15 +01:00
Romain Vimont
f978e4d6de Use VecDeque in aoa_hid
Replace cbuf by VecDeque in aoa_hid
2023-03-10 22:22:15 +01:00