In commit c6901a8b, the frame clock reported time was changed from
simply reporting the time we ran the frame clock cycle to reporting a
smoothed value that increased by the frame interval each time it was
called.
However, this change caused some problems, such as:
https://gitlab.gnome.org/GNOME/gtk/-/merge_requests/1415https://gitlab.gnome.org/GNOME/gtk/-/merge_requests/1416https://gitlab.gnome.org/GNOME/gtk/-/merge_requests/1482
I think a lot of this is caused by the fact that we just overwrote the
old frame time with the smoothed, monotonous timestamp, breaking
some things that relied on knowing the actual time something happened.
This is a new approach to doing the smoothing that is more explicit.
The "frame_time" we store is the actual time we ran the update cycle,
and then we separately compute and store the derived smoothed time and
its period, allowing us to easily return a smoothed time at any time
by rounding the time difference to an integer number of frames.
The initial frame_time can be somewhat arbitrary, as it depends on the
first cycle which is not driven by the frame clock. But follow-up
cycles are typically tied to the the compositor sending the drawn
signal. It may happen that the initial frame is exactly in the middle
between two frames where jitter causes us to randomly round in
different directions when rounding to nearest frame. To fix this we
additionally do a quadratic convergence towards the "real" time,
during presentation driven clock cycles (i.e. when the frame times are
small).
On my X11 + nvidia setup gnome-shell doesn't report presentation times.
However it does report refresh rate. We were mostly using this in our
calculation except when computing predicted presentation time, were
it fell back on the default 60Hz.
The marks are averaged based on the name, so this makes more sense.
Also rename the map/unmap marks to have the same capitalization as
everything else.
This drops the marks for before/after-paint as they are internal
things that very rarely use any time, and also flush/resume-events
as any events reported here will get separate marks so will be easy
to see anyway.
Also, we rename the entire frameclock cycle to "frameclock cycle"
rather than "paint_idle" which is rather cryptic.
These don't take a duration, instead they call g_get_monotonic_time() to
and subtract the start time for it.
Almost all our calls are like this, and this makes the callsites clearer
and avoids inlining the clock call into the call site.
When we use if (GDK_PROFILER_IS_RUNNING) this means we get an
inlined if (FALSE) when the compiler support is not compiled in, which
gets rid of all the related code completely.
We also expand to G_UNLIKELY(gdk_profiler_is_running ()) in the supported
case which might cause somewhat better code generation.
usec is the scale of the monotonic timer which is where we get almost
all the times from. The only actual source of nsec is the opengl
GPU time (but who knows what the actual resulution of that is).
Changing this to usec allows us to get rid of " * 1000" in a *lot* of
places all over the codebase, which are ugly and confusing.
Instead of reporting the frame clock phases as defined,
report the duration of the signal emissions, which is more
useful for tracking down what is taking time.
To make a frame clock tick as long as any of the associated surfaces
expect to receive ticks, make the surfaces inhibit freezing the clock,
instead of directly tell the frame clock to freeze itself.
This makes it so that as long as any surface using a certain frame clock
is not frozen (e.g. just received a frame event from the display
server), the frame clock will not be frozen.
With this, the frame clock is initiated as frozen, and won't be thawed
until any surface inhibits freeze. It will be frozen again, when every
surface has that previously inhibited freeze uninhibited freeze.
If we set c_marshaller manually, then g_signal_newv() will not setup a
va_marshaller for us. However, if we provide c_marshaller as NULL, it will
setup both the c_marshaller (to g_cclosure_marshal_VOID__VOID) and
va_marshaller (to g_cclosure_marshal_VOID__VOIDv) for us.
We were adding incomplete frame timings to the
profile, which lead to occasional nonsense
numbers. Instead, only add timings to the profile
once we marked them as complete. This also
gives us an opportunity to add the presentation
time as a marker.
Remove all the old 2.x and 3.x version annotations.
GTK+ 4 is a new start, and from the perspective of a
GTK+ 4 developer all these APIs have been around since
the beginning.
Typically, there won't be any references on old frame timings except for
the most recent timing. So instead of discarding these and re-entering
gslice twice, just steal the old frame timing and reuse it.
https://bugzilla.gnome.org/show_bug.cgi?id=765592
These were showing up higher in Sysprof profiles.
The simple fix is to avoid the emit_by_name() and let the interface emit
the signals directly. No function preconditions are provided since these
are internal API.
The g_print documentation explicitly says not to do this, since
g_print is meant to be redirected by applications. Instead use
g_message for logging that can be triggered via GTK_DEBUG.
Add an API to start or stop continually updating the frame clock.
This is a slight convenience for applcations and avoids the problem
of getting one more frame run after an animation stops, but the
primary motivation for this is because it looks like we might have
to use timeBeginPeriod()/timeEndPeriod() on Windows to get reasonably
accurate timing, and for that we'll need to know if there is an
animation running.
https://bugzilla.gnome.org/show_bug.cgi?id=693934
* remove gdk_frame_clock_get_frame_time_val(); a convenience
function that would rarely be used.
* remove gdk_frame_clock_get_requested() and
::frame-requested signal; while we might want to eventually
be able to track the requested phases for a clock, we don't
have a current use case.
* Make gdk_frame_clock_freeze/thaw() private: they are only
used within GTK+ and have complex semantics.
* Remove gdk_frame_clock_get_last_complete(). Another convenience
function that I don't have a current use case for.
* Rename:
gdk_frame_clock_get_start() => gdk_frame_clock_get_history_start()
gdk_frame_clocK_get_current_frame_timings() => gdk_frame_clock_get_timings()
Since we're not exporting the ability to create your own frame
clock for now, remove the setters for GdkFrameTimings fields.
Also remove all setters and getters for fields that are more
about implementation than about quantities that are meaningful
to the applcation and just access the fields directly within
GDK.
Now that GdkFrameClock is a class, not interface, there's no real advantage
to splitting the frame history into an aggregate object, so directly
merge it into GdkFrameClock.
It's unlikely that anyone will want to have, say, a GtkWidget that
also acts as a GdkFrameClock, so an abstract base class is as
flexible as making GdkFrameClock an interface, but has advantages:
- If we decide to never make implementing your own frame clock
possible, we can remove the virtualization.
- We can put functionality like history into the base class.
- Avoids the oddity of a interface without a public interface
VTable, which may cause problems for language bindings.
Add a very simple GtkWidget function for an "tick" callback, which
is connected to the ::update signal of GdkFrameClock.
Remove:
- GtkTimeline. The consensus is that it is too complex.
- GdkPaintClockTarget. In the rare cases where tick callbacks
aren't sufficient, it's possible to track the
paint clock with ::realize/::unrealize/::hierarchy-changed.
GtkTimeline is kept using ::update directly to allow using a GtkTimeline
with a paint clock but no widget.
For an operation like synchronizing audio to video playback, we need to
be able to predict the time that a frame will be presented. The details
of this depend on the windowing system, so make the backend predict
a presentation time for ::begin-frame and set it on the GdkFrameTimings.
The timing algorithm of GdkFrameClockIdle is adjusted to give predictable
presentation times for frames that are not throttled by the windowing
system.
Helper functions:
gdk_frame_clock_get_current_frame_timings()
gdk_frame_clock_get_refresh_info()
are added for operations that would otherwise be needed multiple times
in different locations.
https://bugzilla.gnome.org/show_bug.cgi?id=685460
In order to be able to track statistics about how well we are drawing,
and in order to be able to do sophisticated things with frame timing
like predicting per-frame latencies and synchronizing audio with video,
we need to be able to track exactly when previous frames were drawn
to the screen.
Information about each frame is stored in a new GdkFrameTimings object.
A new GdkFrameHistory object is added which keeps a queue of recent
GdkFrameTimings (this is added to avoid further complicating the
implementation of GdkFrameClock.)
https://bugzilla.gnome.org/show_bug.cgi?id=685460
When we have pending motion events, instead of delivering them
directly, request the new FLUSH_EVENTS phase of the frame clock.
This allows us to compress repeated motion events sent to the
same window.
In the FLUSH_EVENTS phase, which occur at priority GDK_PRIORITY_EVENTS + 1,
we deliver any pending motion events then turn off event delivery
until the end of the next frame. Turning off event delivery means
that we'll reliably paint the compressed motion events even if more
have arrived.
Add a motion-compression test case which demonstrates behavior when
an application takes too long handle motion events. It is unusable
without this patch but behaves fine with the patch.
https://bugzilla.gnome.org/show_bug.cgi?id=685460
