Closed Bug 1071675 Opened 10 years ago Closed 10 years ago

2.22% Linux32 Dromaeo CSS regression on M-C (v.35) sept 16 from revision 55b46de164d8

Categories

(Core :: General, defect)

x86
Linux
defect
Not set
normal

Tracking

()

RESOLVED WONTFIX

People

(Reporter: jmaher, Unassigned)

References

Details

(Keywords: perf, regression, Whiteboard: [talos_regression])

Dromaeo CSS (higher is better), took a slight regression last week: http://graphs.mozilla.org/graph.html#tests=%5B%5B72,94,33%5D%5D&sel=none&displayrange=30&datatype=running This appears to only be related to linux32. I did some retriggers: https://tbpl.mozilla.org/?fromchange=02c3915bf742&tochange=b50d767bb3e0&jobname=Ubuntu%20HW%2012.04%20mozilla-central%20talos%20dromaeojs it does appear to be related to bug 1068163. Here is some documentation about dromaeo: https://wiki.mozilla.org/Buildbot/Talos/Tests#Dromaeo_CSS
:bholley, can you weigh on in this?
Flags: needinfo?(bobbyholley)
Bug 1068163 looks like this: https://hg.mozilla.org/mozilla-central/rev/3488fc0bf6ef Basically, changing from: if (x) { crash } to if (x) { return null; } Assuming we weren't crashing before, this definitively cannot be the cause of a regression.
Flags: needinfo?(bobbyholley)
well there is a pattern and the act of retriggering removes doubt that something in the environment changed since I retriggered old jobs. Is there a chance we are changing our linking such that we require more memory paging/swapping?
(In reply to Joel Maher (:jmaher) from comment #3) > well there is a pattern and the act of retriggering removes doubt that > something in the environment changed since I retriggered old jobs. Is there > a chance we are changing our linking such that we require more memory > paging/swapping? I really have no idea, and I'm not really sure what we'd do about it.
v.35 has already left beta, I think we need to close this as wontfix.
Status: NEW → RESOLVED
Closed: 10 years ago
Resolution: --- → WONTFIX
You need to log in before you can comment on or make changes to this bug.