61985 - Mozilla uses lots of CPU during "sending request to..."

Status: RESOLVED WORKSFORME

(Core :: Layout, defect)

Description

[Alex Bischoff]

From Bugzilla Helper:
User-Agent: Mozilla/5.0 (Windows; U; Windows NT 5.0; en-US; m18) Gecko/20001204
BuildID:    2000120420

While loading the page, the status bar displays "sending request to
dealnews.com" at several points. This is fine. However, during the "sending
request", the browser locks up -- scroll bars, menus and toolbars don't respond.
After the page is finished loading, things come back to normal.

Reproducible: Always
Steps to Reproduce:
1. Load the page
2. As the page is loading, try to scroll down the page (I think the page is
dynamically generated, so it could take a few seconds to load completely).
3. Notice that as soon as "sending request to dealnews.com" appears in the
status bar, the browser stops responding.

Actual Results:  The browser locks up until the page has finished loading.

Expected Results:  The browser should behave normally during "sending request
to" (menus and scrollbars should work, etc)

I checked the Bug-Query page to look for duplicates, and I didn't find any in my
brief search.

Comment 1

[Matthias Versen [:Matti]]

I can confirm this bug with build 2000113020 / win2k SP1
Mozilla freeze while page loading.
This page is very big (>190k).

Comment 2

[Andreas Franke (gone)]

Not sure if this qualifies for the "freeze" keyword, but adding it together with
"perf" to indicate a "freeze" that goes away after some time.

Comment 3

[Asa Dotzler [:asa]]

I don't see any significant freeze here.  the page seems to load up in about 3
or 4 seconds during which time the CPU is used pretty heavily but not enough to
starve the UI.  tested 022209 mozilla win32 build on NT4

Comment 4

[Alex Bischoff]

Asa: 

Right -- it doesn't freeze anymore, but the Mozilla daily builds of three months
ago (when I first reported the bug) did freeze ;). I'm guessing that one of the
other bugs fixed between now and then resolved the freezing part.

So, perhaps the summary should be reworded to reflect that the current builds
use the CPU heavily on this page but the UI doesn't actually freeze?

Comment 5

[Asa Dotzler [:asa]]

updating summary. A testcase narrowing this down to whatever in the page is
causing the spike would be extremely helpful.

Comment 6

[ohely]

Not sure if this is too helpful, but I see the CPU spike using Linux build
2001031208, and in Netscape 4.75, and also if I'm loading a local copy of the
page, and with image loading turned off.

Comment 7

[Ronald van Kuijk]

I'll try to simplify this page over the weekend. A quick scan however shows that
it uses the width of td tags in a very inconsistent way. e.g.

- Make 3 cells in a row, giv one a fixed width of 9 (absolute) and the other
100%.and the last 9 (absolute again)
- Make 2 rows, each with one cell and set the width of the first at 96% and the
other at 98%
I remember an onld problem where mozilla had problems with a combination of
%,px,pt etc in one row....

Besides this, there are many nested tables (when wil the time come that HTML
designer do not use tables anymore for these kinds of layout issues? I HATE IT ;-) )

Comment 8

[Ronald van Kuijk]

probably related to bug #43039 (and different other 'table' related bugs) So
this one should be probably assigned to [e-mail] as well.

Maybe I should triage these table-freeze bugs? And marking some as dups and
reassinging to one engineer...?

Comment 9

[Ronald van Kuijk]

CC-ing myself, btw. it happens on Linux as well - OS -> ALL ? 

I'll look into the many <FONT> tags as well

Comment 10

[Asa Dotzler [:asa]]

updating component

Comment 11

[karnaze (gone)]

Reassigning to waterson and moving to m1.0

Comment 12

[Asa Dotzler [:asa]]

Bugs targeted at mozilla1.0 without the mozilla1.0 keyword moved to mozilla1.0.1 
(you can query for this string to delete spam or retrieve the list of bugs I've 
moved)

Comment 13

[Ronald van Kuijk]

loads fine for me on linux build 20020426 (800MHz, 256MB system) (even on my 266
MHz system it is ok....)

Comment 14

[Gareth Randall]

This bug is BACK AGAIN for:
Mozilla 1.1
Mozilla/5.0 (X11; U; Linux i586; en-US; rv:1.1) Gecko/20020826
(the standard download)

Platform is:
Linux 2.2.17 i586

(Bug was NOT in Mozilla 1.0)

Can someone go and fix this (again) now?

Comment 15

[Matthias Versen [:Matti]]

URL is dead, marking wfm because there is no testcase