Previously, process_timers() kept going as long as there were due
timers, which could be for extended periods of time or even forever
if there were slow timer callbacks with either periodic timers or if
timers were added repeatedly.
This prevented dequeuing mpv events, and subsequently, among others,
prevented mpv from quitting until process_timers() completed.
For instance, this caused process_timers() to never return:
function render() <longer than 1/60 s on a slow system> end
mp.add_periodic_timer(1/60, render)
Similarly, it never returned if a timer callback always added a new
one-shot which was already due by the time the callback completed.
This commit ensures that process_timers() only executes callbacks which
were due when it started, so that timers which are added (or repeated)
during process_timers() will wait for the next iteration - after mpv
events are dequeued.
This has no performance impact under normal conditions (when callbacks
complete before the next timer is due).
Additionally, previously idle-observers were executed unconditionally
after the timers because indeed there was nothing due when (if...)
process_timers() completed. However, now process_timers() can return
even if there are due timers, so skip idle-observers on such case.
This fixes a long-standing apparent issue where mpv would display the last
frame with no subtitles at EOF. This is caused by sub rendering switching from
video timestamps to audio timestamps when the video ends, and audio streams
often running past the timestamp of the last video frame. However, authoring
tools (most notably Aegisub) don't tend to provide easy ways to add meaningful
subtitles after the end of the video, so this is rarely actually useful.
This makes cover-art-auto behave more like sub-auto and audio-file-auto:
- load cover art with a language, e.g. if playing foo.mp3, foo.en.jpg
will be loaded with lang=en
- load cover art containing the media filename with fuzzy and all, e.g.
'foo (large).jpg'
- make all/2 load all images in the directory, and make fuzzy/1 the
default
These are all uncommon use cases, but synchronizing the behavior of the
external file options simplifies the code.
The idle-observers block has at least 3 function calls, and it was
entered every time the script went into idle, however, we can save these
calls by checking first that there are indeed such observers.
Remove the check that the external filename is not the same as the
currently playing one, which prevents mpv from loading images again as
external cover art, but this isn't necessary because cover art is only
added when playing standalone audio. I had only added this check because
I would otherwise get a segfault only when compiling with gcc 10.2 with
optimize and changing position within a playlist of multiple images (and
this couldn't even be reproduced by Dudemanguy on the same gcc version),
but this was caused by the uninitialized lang variable which is now
fixed.
Commit 029ff10 added a goto statement which skipped initializing the
`lang' variable. This could crash depending on compiler optimizations
and other factors: if the lang bstr pointer happened to end up NULL
(which is apparently the case with most compilers) then it's validly
empty, but if it pointed to a random and incorrect memory address then
it crashed.
The crash was observed when mpv was compiled using gcc 10.2 with
optimizations enabled, and affected some third party Windows builds.
This commit ensures the goto doesn't skip the initialization.
Thanks to votemp for figuring this out.
Fixes#8922.
Previously tick() was ensured unconditionally only after mouse-move,
but there are other mouse-events which need re-rendering (tick), like
mouse-down (icons may get grayed-out) or mouse-up (icons may change).
For instance: mpv --pause --myclip.mkv
then move the mouse over the volume/mute OSC button, then - without
moving the mouse - press and release the left mouse button.
The osc was not re-rendered because it's paused and the mouse didn't
move, so the volume icon didn't get grayed-out when held down, and the
icon doesn't change on mouse-up (to reflect the new mute state).
Now both these changes are rendered correctly.
This is not a new issue, however, until the last commit - 96b246d
init probably didn't happen much (or at all) between mouse-down and
mouse-up, but after this commit, if there are chapters in a live-stream
then osc_init() is used to re-render the markers at the adjusted
position - which breaks the OSC buttons functionality if init happened
between mouse-down and mouse-up.
Commit 6abb7e3 updates the markers when the chapters change, but it
doesn't update their relative position at the bar when the duration
changes.
This means that adding chapters to a live stream would result in
corresponding chapter markers which were static while the duration
changed and thus their positions became incorrect over time until the
OSC was reinitialized.
This is fixed by observing the duration property if chapters are present
and reinitializing the OSC when the duration changes.
The live_markers user option, which determines whether the duration
property is observed when there are chapters, has been added in order to
allow disabling this behaviour as calling request_init() frequently
might have some impact on low-end systems.
The impact of request_init() on render() was measured to increase from
1-1.5 ms to 2-3 ms on a 2010 MacBook Air, while the impact was neglible
on a 2016 Surface Book (increasing only to an average of 1.4 ms from
1.3 ms for n=1500 render cycles).
The live_markers option is enabled by default.
If a plane is not connected to any displays, we won't set an entry in
the mapping of planes to displays. So ensure these unset entries are
null and skip them.
Fixes#8913
This is the Vulkan equivalent of the drm context for OpenGL, with
the big difference that it's implemented purely in terms of Vulkan
calls and doesn't actually require drm or kms.
The basic idea is to identify a display, mode, and plane on a device,
and then create a display backed surface for the swapchain. In theory,
past that point, everything is the same, and this is in fact the case
on Intel hardware. I can get a video playing on a vt.
On nvidia, naturally, things don't work that way. Instead, nvidia only
implemented the extension for scenarios where a VR application is
stealing a display from a running window system, and not for
standalone scenarios. With additional code, I've got this scenario to
work but that's a separate incremental change.
Other people have tested on AMD, and report roughly the same behaviour
as on Intel.
Note, that in this change, the VT will not be correctly restored after
qutting. The only way to restore the VT is to introduce some drm
specific code which I will illustrate in a separate change.
The VkDisplayKHR context type requires making calls against the
physical device before the libplacebo context is initialised.
That means we can't simply use the physical device object that
libplacebo would create - instead we have to create a separate one,
but make sure it's referring to the same physical device.
To that end, we need the device name that the user may have requested
so we can pass it on.
This moves the JSON parsing above the main youtube-dl error-handling
block and integrates parsing errors into that block. Now, if a parsing
error occurs, youtube-dl's stderr will be logged as it is with other
errors. This also catches errors that cause youtube-dl to output
"null", which would previously be mishandled as a parsing error and
crash ytdl_hook when it attempted to concatenate the error string from
parse_json.
This was originally just a bugfix for a race condition, but the scope
expanded a bit. Currently, the wayland code does a prepare_read ->
dispatch_pending -> display_flush -> read_events -> dispatch_pending
routine that's basically straight from the wayland client API
documentation. This essentially just queues up all the wayland events
mpv has and dispatches them to the compositor. We do this for blocking
purposes on every frame we render. A very similar thing is done for
wait_events from the VO.
This code can pretty easily be unified and split off into a separate
function, vo_wayland_dispatch_events. vo_wayland_wait_frame will call
this function in a while loop (which will break either on timeout or if
we receive frame callback from the compositor). wait_events needs to
just check this in case we get some state change on the wakeup_pipe
(i.e. waking up from the paused state).
As for the actual bugfix part of this, it's a slight regression from
c26d833. The toplevel config event always forced a redraw when a surface
became activated again. This is for something like displaying cover art
on a music file. If the window was originally out of view and then later
brought back into focus, no picture would be rendered (i.e. the window
is just black). That's because something like cover art is just 1 frame
and the VO stops doing any other additional rendering. If you miss that
1 frame, nothing would show up ever again. The fix in this case is to
always just force a redraw when the mpv window comes back into view.
Well with the aforementioned commit, we stopped doing
wl_display_roundtrip calls on every frame. That means we no longer do
roundtrip blocking calls. We just be sure to queue up all of the events
we have and then dispatch them. Because wayland is fundamentally an
asynchronous protocol, there's no guarantee what order these events
would be processed in. This meant that on occasion, a
vo_wayland_wait_frame call (this could occur multiple times depending on
the exact situation) would occur before the compositor would send back
frame callback. That would result in the aforementioned bug of having
just a black window. The fix, in this case, is to just do a
vo_wayland_wait_frame call directly before we force the VO to do a
redraw. Note that merely dispatching events isn't enough because we
specifically need to wait for the compositor to give us frame callback
before doing a new render.
P.S. fix a typo too.
And also change the existing WM_KILLFOCUS handler to return 0 instead
of 'break' (which later calls DefWindowProcW), as MSDN says we should
do for WM_{KILL,SET}FOCUS.
It seems that the 'focused' property is now supported by all main VOs:
x11, macOS, wayland, Windows.
TCT/sixel/caca probably don't support it, and unknown with SDL.
Fixes#8868
When the OSC initializes, it checks whether the current video has
chapters, and if it does not, it disables its chapter functionality
(chapter buttons are grayed out and unusable, chapter indicators don't
show on the seek bar). If another script changed the chapter list
after the video has loaded, those changes would be ignored by the OSC
until some other event causes it to re-initialize, because it did not
observe the chapter list property. This is fixed by adding
observation of chapter-list alongside the other properties that
trigger re-initialization.
Avoiding blindly copying the codec_tag between different formats allows
us to mux packets from, say, mpegts streams to matroska, making the
recorder (dump-cache) much more usable as unsupported codec_tags can
make the muxer reject the streams.
This brings my scaletempo2 benchmark down from ~22s to ~7s on my machine
(-march=native), and down to ~11s with a generic compile.
Guarded behind an appropriate #ifdef to avoid being ableist against
people who have the clinical need to run obscure platforms.
Closes#8848
Not only does this have semantics that make far more sense, it also has
a default that makes far more sense. (Equivalent to the old
`icc-contrast=inf`)
This removes the weird 1000:1 contrast default assumption which
especially broke perceptual profiles and also screws things up for
OLED/CRT/etc.
Should probably close some issues but I honestly can't be bothered to
figure out which of the thousands colorimetry-related issues are
affected.
Take two. f4e89dd went wrong by moving vo_wayland_wait_frame before
start_frame was called. Whether or not this matters depends on the
compositor, but some weird things can happen. Basically, it's a
scheduling issue. vo_wayland_wait_frame queues all events and sends them
to the server to process (with no blocking if presentation time is
available). If mpv changes state while rendering (and this function is
called before every frame is drawn), then that event also gets
dispatched and sent to the compositor. This, in some cases, can cause
some funny behavior because the next frame gets attached to the surface
while the old buffer is getting released. It's safer to call this
function after the swap already happens and well before mpv calls its
next draw. There's no weird scheduling of events, and the compositor log
is more normal.
The second part of this is to fix some stuttering issues. This is mostly
just conjecture, but probably what was happening was this thing called
"composition". The easiest way to see this is to play a video on the
default audio sync mode (probably easiest to see on a typical 23.976
video). Have that in a window and float it over firefox (floating
windows are bloat on a tiling wm anyway). Then in firefox, do some short
bursts of smooth scrolling (likely uses egl). Some stutter in video
rendering could be observed, particularly in panning shots.
Compositors are supposed to prevent tearing so what likely was happening
was that the compositor was simply holding the buffer a wee bit longer
to make sure it happened in sync with the smooth scrolling. Because the
mpv code waits precisely on presentation time, the loop would timeout on
occasion instead of receiving the frame callback. This would then lead
to a skipped frame when rendering and thus causing stuttering.
The fix is simple: just only count consecutive timeouts as not receiving
frame callback. If a compositor holds the mpv buffer slightly longer to
avoid tearing, then we will definitely receive frame callback on the
next round of the render loop. This logic also appears to be sound for
plasma (funfact: Plasma always returns frame callback even when the
window is hidden. Not sure what's up with that, but luckily it doesn't
matter to us.), so get rid of the goofy 1/vblank_time thing and just
keep it a simple > 1 check.
Checking against 1.0 is wrong for this code, because it's in an absolute
luminance scale relative to reference white. We should be normalizing
this by `dst.sig_peak`.
This is actually a very nice simplification that should have been
thought of years ago (sue me). In a nutshell, the story with the
wayland code is that the frame callback and swap buffer behavior doesn't
fit very well with mpv's rendering loop. It's been refactored/changed
quite a few times over the years and works well enough but things could
be better. The current iteration works with an external swapchain to
check if we have frame callback before deciding whether or not to
render. This logic was implemented in both egl and vulkan.
This does have its warts however. There's some hidden state detection
logic which works but is kind of ugly. Since wayland doesn't allow
clients to know if they are actually visible (questionable but
whatever), you can just reasonably assume that if a bunch of callbacks
are missed in a row, you're probably not visible. That's fine, but it is
indeed less than ideal since the threshold is basically entirely
arbitrary and mpv does do a few wasteful renders before it decides that
the window is actually hidden.
The biggest urk in the vo_wayland_wait_frame is the use of
wl_display_roundtrip. Wayland developers would probably be offended by
the way mpv abuses that function, but essentially it was a way to have
semi-blocking behavior needed for display-resample to work. Since the
swap interval must be 0 on wayland (otherwise it will block the entire
player's rendering loop), we need some other way to wait on vsync. The
idea here was to dispatch and poll a bunch of wayland events, wait (with
a timeout) until we get frame callback, and then wait for the compositor
to process it. That pretty much perfectly waits on vsync and lets us
keep all the good timings and all that jazz that we want for mpv. The
problem is that wl_display_roundtrip is conceptually a bad function. It
can internally call wl_display_dispatch which in certain instances,
empty event queue, will block forever. Now strictly speaking, this
probably will never, ever happen (once I was able to to trigger it by
hardcoding an error into a compositor), but ideally
vo_wayland_wait_frame should never infinitely block and stall the
player. Unfortunately, removing that function always lead to problems
with timings and unsteady vsync intervals so it survived many refactors.
Until now, of course. In wayland, the ideal is to never do wasteful
rendering (i.e. don't render if the window isn't visible). Instead of
wrestling around with hidden states and possible missed vblanks, let's
rearrange the wayland rendering logic so we only ever draw a frame when
the frame callback is returned to use (within a reasonable timeout to
avoid blocking forever).
This slight rearrangement of the wait allows for several simplifications
to be made. Namely, wl_display_roundtrip stops being needed. Instead, we
can rely entirely on totally nonblocking calls (dispatch_pending, flush,
and so on). We still need to poll the fd here to actually get the frame
callback event from the compositor, but there's no longer any reason to
do extra waiting. As soon as we get the callback, we immediately draw.
This works quite well and has stable vsync (display-resample and audio).
Additionally, all of the logic about hidden states is no longer needed.
If vo_wayland_wait_frame times out, it's okay to assume immediately that
the window is not visible and skip rendering.
Unfortunately, there's one limitation on this new approach. It will only
work correctly if the compositor implements presentation time. That
means a reduced version of the old way still has to be carried around in
vo_wayland_wait_frame. So if the compositor has no presentation time,
then we are forced to use wl_display_roundtrip and juggle some funny
assumptions about whether or not the window is hidden or not. Plasma is
the only real notable compositor without presentation time at this stage
so perhaps this "legacy" mechanism could be removed in the future.
when using the bundle with activated big enough cache, very slow and
wasteful memory allocations lead to jittery playback and lot of dropped
frames. the cache had to have a certain size so it would constantly
allocate new memory to reproduce this. this never happens when started
from the terminal.
the source of the problem is a different malloc allocation policy,
MALLOC_NANO, that allocated a huge amount of virtual memory without
actually using it. the usage was between 0% to 25% of that virtual
memory. the binaries allocation policy on the other hand used >80% of
that allocated virtual memory and was a lot more efficient, it would use
MALLOC_TINY instead.
this is fixed by setting the MallocNanoZone environment variable to 0
to use the V1 of the allocation policy. when started from the bundle via
launchd this is forced to 1 and V2 policy which causes this problem.
some more info can be found in following file and its comments on the
Apple open source site:
https://opensource.apple.com/source/libmalloc/libmalloc-317.40.8/src/nano_malloc_common.c.auto.htmlFixes#7405
this can cause stutter on remote files because in certain cases this
causes a reconnect to the remote that leads to the file not being read
fast enough. VLC had the same problem and fixes it the same way.
b8b8c438f8Fixes#4434
Shows uploader, channel, description fields. This works only if the
media media is constructed as EDL (for youtube it usually does this),
and if the all_formats option is not set to true (does anyone even use
it?). The former usually happens because youtube serves audio and video
separately, though it will not for live HLS/DASH. The latter uses
delayed media opening, which breaks the global_tags mechanism (see
previous commit).
Add new header which shows up as tags/metadata (associated with
--display-tags). The way this is added means it doesn't always work,
because root->meta (see code) can be NULL for some absurd reason. But it
works for the one case I intended to use it (ytdl_hook, see next
commit), though only in default configurations.
When trying to use completion for mpv binaries specified with some shell
glob, e.g. ~/dev/mpv/build/mpv, the current code doesn't substitute the
homedir prefix into the path name, resulting in runtime errors about
the file '~/dev/mpv/build/mpv' not being found.
The simple fixed is to use $~var instead of $var whenever expanding the
filename, which performs the same globbing that would otherwise be
performed when executing the command.
This reverts commit b8156a9a86.
Apparently there are two problems here. One on Linux (fixed by the
original change and this revert) and one on alternative-medicine-Job's
OS. Since the latter has deprecated OpenGL and OpenGL is a second-class
citizen, I think it's better to prefer the fix for a platform that is
still alive.
When inserting vf_sub, the VO OSD is directed not to draw itself. But
this flag was never unset, even when removing vf_sub.
The fix is pretty shit, but appears to work.
