Closed Bug 1598789 Opened 5 years ago Closed 2 years ago

Terrible scrolling performance in portions of taskcluster log

Categories

(Core :: Graphics: WebRender, defect, P3)

x86_64
All
defect

Tracking

()

RESOLVED INCOMPLETE

People

(Reporter: jdm, Unassigned)

References

(Blocks 1 open bug, )

Details

Scrolling performance on this taskcluster log is fine until you start scrolling near the bottom of the page where log lines start extending out of the viewport horizontally. Once those lines are back out of the viewport, the scrolling is fluid as expected. The profile I captured (https://profiler.firefox.com/public/772a0bd5bdc813884e6a9d97dee0d92941ef873e/calltree/?globalTrackOrder=0-1-2-3-4-5-6-7-8-9&hiddenGlobalTracks=1-2-3-4-5-6-7-8&localTrackOrderByPid=54790-22-23-0-1-2-3-4-5-6-7-8-9-10-11-12-13-14-15-16-17-18-19-20-21~54793-0~54791-0~62892-0~11816-0~93019-0~10376-0~85119-0~67500-0~85345-0~&thread=31&v=4) shows 70% of time in the Renderer thread is spent under Gl::buffer_data_untyped, with 19% of time spent waiting for the GPU to finish with the previous frame.

Hardware: Unspecified → x86_64
Version: unspecified → Trunk

Curious if this is a regression...

I can see this on Linux too, for what is worth.

OS: macOS → All

The priority flag is not set for this bug.
:jbonisteel, could you have a look please?

For more information, please visit auto_nag documentation.

Flags: needinfo?(jbonisteel)
Blocks: wr-perf
Flags: needinfo?(jbonisteel)
Priority: -- → P3
No longer blocks: wr-perf

Referenced page no longer exists.

Status: NEW → RESOLVED
Closed: 2 years ago
Resolution: --- → INCOMPLETE
You need to log in before you can comment on or make changes to this bug.