587439 - Huge input lag on pages with big dom on FF4 beta 3

Status: RESOLVED WORKSFORME

(Core :: Layout, defect)

Description

[Artur Mlodzinski]

User-Agent:       Mozilla/5.0 (Windows; Windows NT 5.1; rv:2.0b3) Gecko/20100805 Firefox/4.0b3
Build Identifier: Mozilla/5.0 (Windows; Windows NT 5.1; rv:2.0b3) Gecko/20100805 Firefox/4.0b3

In Firefox 4 beta 3 there is a huge input lag when typing to a textarea or textfield on pages with very big dom. Everything is fine when using FireFox 3.6.8.
When i type really fast and stop typing, i can see the letters appear after i typed them.

Reproducible: Always

Steps to Reproduce:
1. Load a page with very big dom
2. Type something fast in textarea or textfield

Actual Results:  
Typed letters appear slowly on screen. Not instantly.

Expected Results:  
Letters should appear instantly.

Comment 1

[(mostly gone) XtC4UaLL [:xtc4uall]]

Can you give an Example or even attach a reduced Testcase?

Comment 2

[Artur Mlodzinski]

I created an example. 
http://mfirm.eu/FF/array-grid.html
The easiest way to see the lag is to just type random letters really fast.
The lag is much worse than in this example in our real application and is easly seen when typing slowly.

Comment 3

[Artur Mlodzinski]

Did you manage to reproduce this bug when opening provided url in FF4.0 Beta3?

Comment 4

[(mostly gone) XtC4UaLL [:xtc4uall]]

Yes, i can reproduce the Issue using Mozilla/5.0 (Windows NT 5.1; rv:2.0b5pre) Gecko/20100819 Minefield/4.0b5pre ID:20100819102200.

I fear the Testcase needs to be reduced more and/or at least be profiled (https://wiki.mozilla.org/Performance:Tools).

Moreover a Regression-Window (Last Build: ok; First Build: not okay) would help by knowing the "Built from" Url from "about:buildconfig" of these two Builds with checking http://ftp.mozilla.org/pub/mozilla.org/firefox/nightly/ ("mozilla-central" flagged Folders).

Comment 5

[(mostly gone) XtC4UaLL [:xtc4uall]]

Attachment: Reporters Testcase (zipped)

Comment 6

[Artur Mlodzinski]

I did a Regression-Window
Last Build OK: Built from http://hg.mozilla.org/mozilla-central/rev/5fda39cd703c
downloaded from http://ftp.mozilla.org/pub/mozilla.org/firefox/nightly/2010/07/
changeset 47677	5fda39cd703c
parent 47676	e63da1dc90b6
child 47679	3f5e9b0e02c5
pushlog:	5fda39cd703c

First Build not ok: Built from http://hg.mozilla.org/mozilla-central/rev/96de199027d7
downloaded from  http://ftp.mozilla.org/pub/mozilla.org/firefox/nightly/2010/07/2010-07-16-04-mozilla-central/
changeset 47790	96de199027d7
parent 47789	bd21c5390faf
child 47791	2d09181ec4cd
child 47805	5575b3579d2f

Comment 7

[Artur Mlodzinski]

Last Build OK i pasted wrong link. Correct link is:
http://ftp.mozilla.org/pub/mozilla.org/firefox/nightly/2010/07/2010-07-15-04-mozilla-central/

Comment 8

[(mostly gone) XtC4UaLL [:xtc4uall]]

Great! Since that Range points to Bug 564991 as the Cause i verified this by checking saved Hourly Builds before and after Checkin, thus the Range is
http://hg.mozilla.org/mozilla-central/pushloghtml?fromchange=92339b84d089&tochange=e1d7fd5255fd

There are several already filed Performance Regression Bugs, like Bug 579260, Bug 579323, Bug 585258. Without Profiling the Testcase, unsure if already covered by those.

Comment 9

[Artur Mlodzinski]

I tested on mozilla 4.0 beta 5 pre and the issue still exists :(
Downloaded from: http://ftp.mozilla.org/pub/mozilla.org/firefox/nightly/2010-08-30-04-mozilla-central/
Version build from: http://hg.mozilla.org/mozilla-central/rev/0487454f4219

Comment 10

[Artur Mlodzinski]

Still the same on 4.0b6pre
http://ftp.mozilla.org/pub/mozilla.org/firefox/nightly/2010-09-06-04-mozilla-central/
Built from http://hg.mozilla.org/mozilla-central/rev/67aef7ffb282

Comment 11

[Robert O'Callahan (:roc) (email my personal email if necessary)]

I'll investigate this for final but it might be difficult to fix completely.

Comment 12

[Robert O'Callahan (:roc) (email my personal email if necessary)]

The reporter's testcase can be loaded at jar:https://bug587439.bugzilla.mozilla.org/attachment.cgi?id=467541!/array-grid.html

Comment 13

[Mike Beltzner [:beltzner, not reading bugmail]]

** PRODUCT DRIVERS PLEASE NOTE **

This bug is one of 7 automatically changed from blocking2.0:final+ to blocking2.0:.x during the endgame of Firefox 4 for the following reasons:

 - it was marked as a soft blocking issue without a requirement for beta coverage

Comment 14

[Artur Mlodzinski]

Any chance to get it fixed soon?. 
Out production application is much slower in Firefox4 than it was in Firefox3 in user experience. All user actions like opening new windows, clicking on list boxes, typing etc. are affected by this issue.

Comment 15

[(mostly gone) XtC4UaLL [:xtc4uall]]

Is this still an Issue?

If I compare running Attachment 467541 [details] locally against Firefox 3.6, Firefox 4 and Firefox 22/Nightly I believe to see that recent Releases are on par with Firefox 3.6, no?

Comment 16

[Artur Mlodzinski]

It ok now.