bug:30895941
Prevents a race where frame work could interleave between frames,
causing SurfaceView position updates to be delivered out of order.
Change-Id: I01e4cc557b69dcf33e877a0e16c0d115ec95e4cc
Bug: 30442298
We already do this for initialize(), fix
it so that update() is parallel with the
UI thread as well.
Shaves ~7ms off of the 99th percentile on
NotificationShade open & close
Change-Id: I1791df495453fb9e1e12362c68e3d20e837e62be
Bug: 30342017
Upload recents thumbnails in the
dead gaps between frames instead of
at the start of a frame. This eliminates
jank caused by the large texture
upload.
Change-Id: I507cd286d199109c7a9a1511d68ba5ab5d28069f
When we stop scheduling for new frames on RenderThread, we should
put the running animations on pause, rather than purge the list
of the running animations, such that in the next full
sync, the animations that were paused will continue to run.
BUG: 30226711
Change-Id: I36ff6f5d26ffa7999f60ca0ff676a35157577dc2
Bug: 30440166
If we are using HWC2, there's a change in timing
when in triple buffering with the pipelined offsets.
This changes JankTracker to recognize that and silently
erase that from the total duration
Change-Id: Ib1fd4209070f17dbd2baed707c8cf73fb11c3cf2
Bug: 30342762
Frame dropping was too frequent and would trigger
during normal triple buffering steady state. Bump
the threshold to drop from 3ms stall to 6ms stall.
Change-Id: I5c1faeaabf0d02323a28e697a4af4105fbcf1c53
bug:29771461
bug:29413700
bug:30181577
Changes frame interval gap detection to look for wider gaps, as they
were incorrectly firing all the time.
Also adds a 500ms minimum gap between frames dropped because of stuffed
swap chain, to prevent dropping too often.
Change-Id: If16ed637d54bf37015704be102c5c2e3731a0824
Add a new mode, controlled by sys.use_fifo_ui property, that enables the
top app's UI and RenderThread to be SCHED_FIFO. This eliminates almost
all jank due to scheduling competition with non-UI critical
threads. This mode may not be suitable for all devices.
bug 24503801
Change-Id: I7b8a31830ad80f7efa00236928d5476998ed4e00
Bug: 29628138
1: Make windowPositionLost synchronous as that's
what the Java side was expecting
2: Make the listener ref counted as otherwise
there's a race condition with the GC, which could
end up with use-after-frees
3: Ensure that all position updates are invoked
prior to frame completion
Change-Id: Iedbc017f611ba2878a49b4586612f79249ca2fe3
Fixes: 29072773
By using computeFrameTime AnimationContext would
potentially end up modifying the latest vsync if
a very-slow frame was received from the UI thread.
This could potentially desync animations that were
RT & UI thread 'synchronized', but more significantly
it would confuse the swap chain which tries to only
draw one frame per vsync causing unneccessary frame
drops.
Change-Id: Ibd2ec3157ce32fee1eec8d56837c45a35e622895
This CL changes the target of VD specific animators to VectorDrawable,
instead of RenderNode. The benefit of doing so is that animators can
now detect whether the animation is meaningful by checking whether
their VD target is in the display list. If not, that means the VD is
not drawing for the current frame, in which case we can be smarter
and more power efficient by removing the animator from the list and
posting a delayed onFinished listener callback.
By setting VD as the animation target, when an ImageView decides to
update its drawable from one AVD to something else, we'll be able
to detect that the previous AVD is no longer in the display list,
and stop providing animation pulse to the stale AVD, which is
something we couldn't do previously. This change also
handles the case where one AVD instance could be drawn in two
different views.
Bug: 27441375
Change-Id: Iaad1ed09cfd526276b95db0dd695275c28e074e8
Fixes: 28218991
If a draw() happens while we are stopped, don't report
that the surface is lost because this will prompt
a tear-down of the surface which isn't desired. It can
result in ViewRootImpl ending up in an internally-bad state
in this case.
Change-Id: If3eb8c6bc8702299e5330bc0917952624dce3b7e
Fixes: 28125010
Restructures 'scene defer', to implement window backdrop overdraw
avoidance in new render pipeline, and disable clipping to content draw
bounds.
Also restructures FrameBuilder's constructors, to separate out into
multiple defer methods.
Change-Id: I53facb904c1a4a4acc493d8a489921a79a50494e
Bug: 27286867
WindowManager has committed to stopped state
controlling the lifecycle of the Surface, so
make that a first-class thing in HWUI as well.
This makes it more resistent to things like
a rogue updateSurface() happening while mStopped=true,
leading to bad things down the line. Instead let
the surface be changed/updated as often as desired,
and just block any attempt to draw on that surface.
Also removes some unnecessary makeCurrent()s, as
EglManager ensures that we *always* have a valid
GL context now (using a pbuffer surface if there is
no window surface set)
Change-Id: Iead78ddebc7997e8fdb0c9534836352f5e54b9bd
Bug: 27709981
This desperately needs a refactor, but to keep
the current (really needed & nice) behavior of
dispatching after sync finishes would be difficult
to handle cleanly without lots of ripping so... #yolo
Change-Id: I831a06c6ae7412a062720d68ecbe3085190f0258
Bug: 27922347
* Dump the full histogram
* Expand the histogram to have a slow-frame section with
large 50ms buckets to raise the cap to 5s to give more
insight into system-health
* Stop excluding first-frame metrics as we want to include
those in our global tracking. Automated tests already filter
these out by doing resets before running anyway.
Change-Id: Idaba8aad591f59d10a6477b11efc0767ff715083
