Closed Bug 793957 Opened 12 years ago Closed 11 years ago

Investigate robohornet "innerHTML table" benchmark

Categories

(Core :: Layout, defect)

defect
Not set
normal

Tracking

()

RESOLVED WONTFIX

People

(Reporter: khuey, Unassigned)

References

Details

Us: 752.05ms Chrome: 272.09ms
Need to profile where the time is spent. Is it js, or html parser, or DOM or layout.
Need to look at more closely what the test is actually timing, but at least the first profile showed almost all the time spent in layout, creating frames and reflowing. Setting innerHTML/parsing was like <5% of the total cpu time.
Component: DOM → Layout
Hmm, back to DOM. If I read the testing code correctly, it is actually testing the sync innerHTML performance, not async framecreation.
Component: Layout → DOM
Hmm, or not. If I make the test to do innerHTML on display:none element, we run the test 10x faster.
Component: DOM → Layout
Whether or not the test is testing the right thing [1], reflow is too slow here. [1] https://github.com/robohornet/robohornet/issues/66
AFAIK, robohornet is not a thing anymore (if it ever was), and its tests are too broken to be useful. We can find better tests and profile those.
Status: NEW → RESOLVED
Closed: 11 years ago
Resolution: --- → WONTFIX
You need to log in before you can comment on or make changes to this bug.