Closed Bug 1302544 Opened 8 years ago Closed 6 years ago

Investigate an upper bound on how much performance problem pre-emption could fix

Categories

(Core :: DOM: Core & HTML, defect, P3)

defect

Tracking

()

RESOLVED WONTFIX

People

(Reporter: ehsan.akhgari, Assigned: farre)

References

(Blocks 1 open bug)

Details

We'd like to get a sense of how much time we spend in places in the code where we could benefit from pre-emption.  This will require us to look at call stacks where we're running some JS code in a third-party iframe or in a background tab and whether the refresh driver is active in that case (aka whether we'd have liked to paint) or whether there's an unprocessed user input event in the queue (aka whether we'd have liked to process the user input event.)

Once we have this measurement, we can get a sense of an upper bound on how much we could make things better if we were pre-empting the long running task in that case.
Priority: -- → P3
mystor decided to own the data collection stuff here.
Assignee: nobody → michael
Assignee: michael → afarre
Status: NEW → RESOLVED
Closed: 6 years ago
Resolution: --- → WONTFIX
why wontfix? obsoleted by e10s?
Obsolated by bug 1351021.
Component: DOM → DOM: Core & HTML
You need to log in before you can comment on or make changes to this bug.