Remove the extra vf_chain.output field - there's absolutely no need for
it, because there is always a last filter which will buffer the output.
For some reason, vf_chain.last was never set, which we now need to fix
too.
The purpose of temporarily setting stop_play was to make the audio
uninit code to explicitly drain audio if needed. This was the only way
to do it before ao_drain() was made a separate function; now we can just
do it explicitly instead.
We absolutely need to clear the AO reference in the mixer.
The audio_status must be changed to a state where no code assumes that
the AO is available. (It's allowed to do this blindly.)
When initialization failed, vo_lavc may cause an irrecoverable state in
the ffmpeg-related structs. Therefore, we reject additional
initialization attempts at least until we know a better way to clean up
the mess.
ao_lavc currently cannot be initialized more than once, yet it's good to
do consistent changes there as well.
Also, clean up uninit-after-failure handling to be less spammy.
The mp_audio_from_avframe() function requires the AVFrame to be
refcounted, and merely increases its refcount while referencing the same
data. For non-refcounted frames, it simply did nothing and potentially
would make the caller pass around a frame with dangling pointers.
(libavcodec should always return refcounted frames, but it's not clear
what other code does; and also the function should simply work, instead
of having weird requirements on its arguments.)
Revert commit 24e52f66; even though the old beheavior doesn't make sense
(as the commit message assured), it turns out that this works better:
typically, it means preroll will start from the previous video key frame
(the video CUE index will contain clusters with video key frames only),
which often coincides with subtitle changes. Thus the old behavior is
actually better.
Change the code that uses CueDuration elements. Instead of merely
checking whether preroll should be done, find the first cluster that
needs to be read to get all subtitle packets. (The intention is to
compensate for the enlarged preroll cluster-range due to reverting
commit 24e52f66.)
The previous fix breaks another obscure case: if the second vf_sub adds
margins, the image is accidentally not extended, which would return in
an assertion failure when returning the bogus image.
This rewrites the audio decode loop to some degree. Audio filters don't
do refcounted frames yet, so af.c contains a hacky "emulation".
Remove some of the weird heuristic-heavy code in dec_audio.c. Instead of
estimating how much audio we need to filter, we always filter full
frames. Maybe this should be adjusted later: in case filtering increases
the volume of the audio data, we should try not to buffer too much
filter output by reducing the input that is fed at once.
For ad_spdif.c and ad_mpg123.c, we don't avoid extra copying yet - it
doesn't seem worth the trouble.
Use a pseudo-filter when changing speed with resampling, instead of
somehow changing a samplerate somewhere. This uses the same underlying
mechanism, but is a bit more structured and cleaner. It also makes some
of the following changes easier.
Since we now always use filters to change audio speed, move most of the
work set_playback_speed() does to recreate_audio_filters().
A helper to allocate refcounted audio frames from a pool. This will
replace the static buffer many audio filters use (af->data), because
such static buffers are incompatible with refcounting.
Causes the player to reload the demuxer and to relist the found
streams. Probably slightly dangerous/broken, because the demuxer
thread and possibly even the decoders will keep reading data from
the new title before the new demuxer takes over.
Fixes#1250.
A first step towards refcounted audio frames.
Amazingly, the API just does what we want, and the code becomes
simpler. We will need to NIH allocation from a pool, though.
If the audio callback suddenly stops, and the AO provides no "reset"
callback, then reset() could deadlock by waiting on the audio callback
forever.
The waiting was needed to enter a consistent state, where the audio
callback guarantees it won't access the ringbuffer. This in turn is
needed because mp_ring_reset() is not concurrency-safe.
This active waiting is unavoidable. But the way it was implemented, the
audio callback had to call ao_read_data() at least once when reset() is
called. Fix this by making ao_read_data() set a flag upon entering and
leaving, which basically turns p->state into some sort of spinlock.
The audio callback actually never needs to spin, because there are only
2 states: playing audio, or playing silence. This might be a bit
surprising, because usually atomic_compare_exchange_strong() requires a
retry-loop idiom for correct operation.
This commit is needed because ao_wasapi can (or will in the future)
randomly stop the audio callback in certain corner cases. Then the
player would hang forever in reset().
As usual, we use C11 semantics, and emulate it if <stdatomic.h> is not
available.
It's a bit messy with __sync_val_compare_and_swap(). We assume it has
"strong" semantics (it can't fail sporadically), but I'm not sure if
this is really the case. On the other hand, weak semantics don't seem to
be possible, since the builtin can't distinguish between the two failure
cases that could occur. Also, to match the C11 interface, use of gcc
builtins is unavoidable. Add a check to the build system to make sure
the compiler supports them (although I don't think there's any compiler
which supports __sync_*, but not these extensions).
Needed for the following commit.
This is what you would expect. Before this commit, each
ao_request_reload() call would just queue a reload command, and then
recreate the AO for the number of times the function was called.
Instead of sending a command, introduce some sort of event retrieval
mechanism. At least for the reload case, use atomics, because we're too
lazy to setup an extra mutex.
This commit fixes a "cosmetic" user interface issue. Instead of
displaying the interpolated seek time on OSD, show the actual audio
time.
This is rather silly: when seeking in audio-only mode, it takes some
iterations until audio is "ready", but on the other hand, the audio
state machine is rather fickle, and fixing this cosmetic issue would be
intrusive. So just add a hack that paints over the ugly behavior as
perceived by the user. Probably the lesser evil.
It doesn't happen if video is enabled, because that mode sets the
current time immediately to video PTS. (Audio has to be synced to video,
so the code is a bit more complex.)
Fixes#1233.
This reverts commit d859549424.
Going to apply the alternative fix through PR #1256, which came just
some seconds after pushing the reverted commit. The reverted commit
was reported as not actually working.
Found by clang sanitizer. Casting unsigned integers to signed integers
with same size has implementation defined behavior (it's even allowed to
crash), but it seems reasonable to expect that reasonable
implementations do a complement of 2 "conversion".
Following the discussion in #1253.
The events won't be removed for a while, though. (Or maybe never, unless
we run out of bits for the uint64_t event mask.)
This is not a real change (the events still work, and the alternative
mechanisms were established a few API revisions earlier), but for the
sake of notifying API users, update DOCS/client-api-changes.rst.
The values compared here happen to be of unsigned enum types - but the
test is not supposed to break if we somehow force the enum to signed, or
if the compiler happens to use a signed type (as far as I remember, the
exact integer type the compiler can use is implementation-defined).
The main need I see for this is with libmpv - it would be confusing if
some application showed up as "mpv" on whateverthehell PulseAudio uses
it for (generally it does show up on various PA GUI tools).
This silences the warning:
video/out/gl_video.c:1091:51: runtime error: division by zero
when running with clang -fsanitize=undefined. Division by zero is legal
according to IEEE, but I guess clang doesn't care about standard. While
triggering this warning isn't actually avoided in all cases, it's
avoided in the common case and also makes people shut up about it.
Call VOCTRL_GET_DISPLAY_NAMES it when the property is
requested. The vo should return the names of the displays that the mpv
window is covering. For example, with x11 vos, xrandr names LVDS1,
HDMI1, etc.
XRRGetOutputInfo contains a "name" element which corresponds to to the
display names given to the user by the "xrandr" command line
utility. Copy it into the xrandr_display struct for each display.
On VOCTRL_GET_DISPLAY_NAMES, send a copy of the names
of the displays spanned by the mpv window on.
