Closed Bug 741212 Opened 12 years ago Closed 2 years ago

Firefox freezes almost completely and uses 25% of CPU when opening a specific page with a few flash objects.

Categories

(Core :: JavaScript Engine, defect)

11 Branch
x86_64
Windows 7
defect
Not set
normal

Tracking

()

RESOLVED INCOMPLETE
Tracking Status
firefox12 + wontfix
firefox13 + ---
firefox14 - ---

People

(Reporter: vxnano, Unassigned, NeedInfo)

References

Details

(Keywords: perf, regression, Whiteboard: [js:p2][js:bumped:2][js:ni])

User Agent: Mozilla/5.0 (Windows NT 6.1; WOW64; rv:11.0) Gecko/20100101 Firefox/11.0
Build ID: 20120312181643

Steps to reproduce:

Open a new browser window, then type:
http://www.manager-tools.com/manager-tools-basics/  in the address bar and press Enter.





Actual results:

Firefox freezes almost completely. It responds very slowly, about 15-20 seconds after I click anything. CPU usage is 25% constantly. I waited for few minutes, but it did not end.




Expected results:

The page should be displayed and the broser should be responsive.

I tried doing it on a new windows account and in safe mode. Neither did help. 
On IE9 the page loads almost instantly..
Summary: Firefox freezes completely and uses 25% of CPU when opening a specific page with a few flash objects. → Firefox freezes almost completely and uses 25% of CPU when opening a specific page with a few flash objects.
I can reproduce on Firefox11-Nightly14.0a1 with clean profile + only Flash 10.3.183.18.
Open the url and wait.
CPU usage never settles down (15-25% of Core2Quad@2.5GHz).

In Firefox 10, CPU usage becomes settle down(0-2% of Core2Quad@2.5GHz).

Disabled Javascript helps.

Regression window(m-c)
CPU usage settles down:
http://hg.mozilla.org/mozilla-central/rev/a68c96c1d8e0
Mozilla/5.0 (Windows NT 6.1; WOW64; rv:11.0a1) Gecko/20111203 Firefox/11.0a1 ID:20111203031117
CPU usage never settles down:
http://hg.mozilla.org/mozilla-central/rev/13afcd4c097c
Mozilla/5.0 (Windows NT 6.1; WOW64; rv:11.0a1) Gecko/20111203 Firefox/11.0a1 ID:20111203122618
Pushlog:
http://hg.mozilla.org/mozilla-central/pushloghtml?fromchange=a68c96c1d8e0&tochange=13afcd4c097c


Regression window(jm)
CPU usage settles down:
http://hg.mozilla.org/projects/jaegermonkey/rev/f6b97927b0ea
Mozilla/5.0 (Windows NT 6.1; WOW64; rv:11.0a1) Gecko/20111116 Firefox/11.0a1 ID:20111116040319
CPU usage never settles down:
http://hg.mozilla.org/projects/jaegermonkey/rev/a7094fec1764
Mozilla/5.0 (Windows NT 6.1; WOW64; rv:11.0a1) Gecko/20111118 Firefox/11.0a1 ID:20111118150124
Pushlog:
http://hg.mozilla.org/projects/jaegermonkey/pushloghtml?fromchange=f6b97927b0ea&tochange=a7094fec1764

Suspected: bug 703047
Assignee: nobody → general
Blocks: 703047
Status: UNCONFIRMED → NEW
Component: Untriaged → JavaScript Engine
Ever confirmed: true
Keywords: regression
Product: Firefox → Core
QA Contact: untriaged → general
Keywords: perf
This does not appear to be critical enough to take in our final beta before release, since we've already shipped with this bug in FF11 and have not had other user complaints. Let's shoot for fixing this in FF13.
Brian, can you have a look at this? The regression range points at ObjShrink work.

I was able to repro this on Win7 in Firefox 12. The profiler results are puzzling: both plugin-container and firefox use up most of a core, but nothing is very hot--JS isn't very high in the list, and no function is particularly hot, so it doesn't look like a JS performance fault. I'm wondering if there's some slight behavioral change that is causing ilooping-type behavior with the plugins. cc'ing Luke as well because he's fixed one or two bugs like this before, I think.

What's also interesting is that Chrome has similar behavior, but Safari doesn't.
This page loads fine for me in a Mac debug build, Flash 10.1.102.64
What else can be done to get some movement towards a possible fix here?  Brian will you be testing out other OSes?  We've still got time to get something landed on Firefox 13.
(In reply to Brian Hackett (:bhackett) from comment #4)
> This page loads fine for me in a Mac debug build, Flash 10.1.102.64

You should be able to repro this on Windows.
I happened upon this bug by way of reading the meeting notes from yesterday (38 bugs remaining to be fixed in Fx13) and it bears a *little* bit of similarity to CPU spiking (and disk thrashing) I reported in bug 741803.
Whiteboard: [js:p2:fx16]
Just wanted to post an update since Comment 7. Re bug 741803, it seems AdBlock Plus might be playing a role in the behavior I saw. Although the specific reason why it was causing the behavior in 741803 is unknown, I did submit a report to the devs of AdBlock Plus.
This regression has now been shipped in 2 releases without an explosion in negative feedback. Untracking.
Whiteboard: [js:p2:fx16] → [js:p2:fx17][js:bumped:1]
Whiteboard: [js:p2:fx17][js:bumped:1] → [js:p2:fx18][js:bumped:2]
Whiteboard: [js:p2:fx18][js:bumped:2] → [js:p2][js:bumped:2][js:ni]
Assignee: general → nobody
20160502172042 Mozilla/5.0 (Windows NT 6.1; rv:46.0) Gecko/20100101 Firefox/46.0
20160602030220 Mozilla/5.0 (Windows NT 6.1; rv:49.0) Gecko/20100101 Firefox/49.0

I have tested your issue on latest FF release and latest Nightly build and could not reproduce it.
Is this still reproducible on your end ? If yes, can you please retest this using latest FF release and latest Nightly build (https://nightly.mozilla.org/) and report back the results ? When doing this, please use a new clean Firefox profile, maybe even safe mode, to eliminate custom settings as a possible cause (https://goo.gl/PNe90E).
Flags: needinfo?(vxnano)

Is this Bug flash related or fixed anyway and can be closed now?

Flags: needinfo?(thee.chicago.wolf)

Since Flash has been killed off since the end of 2020, this is probably no longer a thing? Probably safe to close.

Flags: needinfo?(thee.chicago.wolf)
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.