aboutsummaryrefslogtreecommitdiffstats
path: root/doc/multithreading.txt
blob: 83849deacca3f378241a6cfa42b5def373368234 (plain) (blame)
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
FFmpeg multithreading methods
==============================================

FFmpeg provides two methods for multithreading codecs.

Slice threading decodes multiple parts of a frame at the same time, using
AVCodecContext execute() and execute2().

Frame threading decodes multiple frames at the same time.
It accepts N future frames and delays decoded pictures by N-1 frames.
The later frames are decoded in separate threads while the user is
displaying the current one.

Restrictions on clients
==============================================

Slice threading -
* The client's draw_horiz_band() must be thread-safe according to the comment
  in avcodec.h.

Frame threading -
* Restrictions with slice threading also apply.
* For best performance, the client should set thread_safe_callbacks if it
  provides a thread-safe get_buffer() callback.
* There is one frame of delay added for every thread beyond the first one.
  Clients must be able to handle this; the pkt_dts and pkt_pts fields in
  AVFrame will work as usual.

Restrictions on codec implementations
==============================================

Slice threading -
 None except that there must be something worth executing in parallel.

Frame threading -
* Codecs can only accept entire pictures per packet.
* Codecs similar to ffv1, whose streams don't reset across frames,
  will not work because their bitstreams cannot be decoded in parallel.

* The contents of buffers must not be read before ff_thread_await_progress()
  has been called on them. reget_buffer() and buffer age optimizations no longer work.
* The contents of buffers must not be written to after ff_thread_report_progress()
  has been called on them. This includes draw_edges().

Porting codecs to frame threading
==============================================

Find all context variables that are needed by the next frame. Move all
code changing them, as well as code calling get_buffer(), up to before
the decode process starts. Call ff_thread_finish_setup() afterwards. If
some code can't be moved, have update_thread_context() run it in the next
thread.

If the codec allocates writable tables in its init(), add an init_thread_copy()
which re-allocates them for other threads.

Add AV_CODEC_CAP_FRAME_THREADS to the codec capabilities. There will be very little
speed gain at this point but it should work.

If there are inter-frame dependencies, so the codec calls
ff_thread_report/await_progress(), set AVCodecInternal.allocate_progress. The
frames must then be freed with ff_thread_release_buffer().
Otherwise leave it at zero and decode directly into the user-supplied frames.

Call ff_thread_report_progress() after some part of the current picture has decoded.
A good place to put this is where draw_horiz_band() is called - add this if it isn't
called anywhere, as it's useful too and the implementation is trivial when you're
doing this. Note that draw_edges() needs to be called before reporting progress.

Before accessing a reference frame or its MVs, call ff_thread_await_progress().