1054216 - Slow Mouse Scrolling

Status: RESOLVED INCOMPLETE

(Core :: General, defect)

Description

[mdew]

User Agent: Mozilla/5.0 (Windows NT 6.1; WOW64; rv:29.0) Gecko/20100101 Firefox/29.0 (Beta/Release)
Build ID: 20140813133002

Steps to reproduce:

Open http://multimedia.scmp.com/typhoons/ and use the scroll wheel.


Actual results:

Very slow mouse scrolling, unusable.


Expected results:

Normal mouse scrolling, smooth page movement.

Chrome performs well on the site.

Comment 1

[Jim Jeffery not reading bug-mail 1/2/11]

Setting to NEW,  confirmed 'jank', slow scrolling on site.

win7 x64 running latest hourly m-c build based on cset:
https://hg.mozilla.org/mozilla-central/rev/c9f8cc9ce89c

Site has a lot of images, and I suspect the slowness/jank are due to decoding of images when scrolling into the viewport.

Comment 2

[Alice0775 White]

WORKAROUND general.smoothScroll.mouseWheel = false

Comment 3

[:Gijs (he/him)]

(In reply to Jim Jeffery not reading bug-mail 1/2/11 from comment #1)
> Setting to NEW,  confirmed 'jank', slow scrolling on site.
> 
> win7 x64 running latest hourly m-c build based on cset:
> https://hg.mozilla.org/mozilla-central/rev/c9f8cc9ce89c
> 
> Site has a lot of images, and I suspect the slowness/jank are due to
> decoding of images when scrolling into the viewport.

Maybe, but it also has scroll and resize handlers... on my mbp, I don't see much jank, but I can reproduce the CPU being hit hard nevertheless. Could you try profiling this using the gecko profiler add-on? (see  https://developer.mozilla.org/en-US/docs/Performance/Profiling_with_the_Built-in_Profiler )

Comment 4

[Jim Jeffery not reading bug-mail 1/2/11]

Attachment: profile run

Here is requested profile run, hope I did it right, never done this before.

Comment 5

[Jim Jeffery not reading bug-mail 1/2/11]

Attachment: 2nd try -

Not sure first profile attachment is correct, another try using app/octetstream

Comment 6

[:Gijs (he/him)]

(In reply to Jim Jeffery not reading bug-mail 1/2/11 from comment #5)
> Created attachment 8475937 [details]
> 2nd try -
> 
> Not sure first profile attachment is correct, another try using
> app/octetstream

You can just copy/paste the cleopatra URL. :-)

Comment 7

[Jim Jeffery not reading bug-mail 1/2/11]

The URL is just giving the Cleopatra interface, no info if displayed until I browse to the local file.

http://people.mozilla.org/~bgirard/cleopatra/#

If I use the 2nd try and paste it into the interface I get profile.. Hope this helps.

Comment 8

[Jim Jeffery not reading bug-mail 1/2/11]

Here is the URL to the full profile run:

http://people.mozilla.org/~bgirard/cleopatra/#report=5d771044076b9fe6031f15ca2c077807bfa8674b

Comment 9

[:Gijs (he/him)]

AFAICT this is caused by the skrollr.min.js library and how it's invoked from a scroll handler (s.refresh call in checkScrollChapter). At the moment it's not clear to me why this is, but I would bet on our scroll events being different from Chrome's and that causing "issues" with the aforementioned library. Basically, the code manually sets scrollTop when it gets scroll events. Without more details pointing to how we pass "wrong" values (which I doubt, but hey, theoretically possible) that cause this, I don't really know what we can do here without spending a full day trying to unminimize the script and how the page invokes it, and why it breaks in Firefox but not webkit/blink.

Comment 10

[Firefox Bug Husbandry Bot]

Moving from Core::Untriaged to Core::General https://bugzilla.mozilla.org/show_bug.cgi?id=1407598