496293 - loading http://www.mtv3.fi locks up browser

Status: VERIFIED DUPLICATE of bug 494332

(Core :: Layout, defect)

Description

[Olli Pettay [:smaug][[e-mail]]]

Loading or re-loading http://www.mtv3.fi (a Finnish news site) locks up the whole
browser. When looking at the stack, browser seems to be reflowing the page.
This is trunk only.

Comment 1

[Henrik Skupin [:whimboo][⌚️UTC+1]]

Cannot see the lock with Mozilla/5.0 (Macintosh; U; Intel Mac OS X 10.5; en-US; rv:1.9.2a1pre) Gecko/20090603 Minefield/3.6a1pre ID:20090603031250

Comment 2

[Olli Pettay [:smaug][[e-mail]]]

Based on Shark-profile, we're looping somewhere under blockframe reflowing inlines.

Comment 3

[Henrik Skupin [:whimboo][⌚️UTC+1]]

Attachment: Assertion (not a UTF8 string)

Comment 4

[Henrik Skupin [:whimboo][⌚️UTC+1]]

Attachment: Assertion (cannot call GetUsedPadding on a dirty frame)

Comment 5

[Henrik Skupin [:whimboo][⌚️UTC+1]]

Attachment: Assertion (cannot call GetUsedBorder on a dirty frame)

Comment 6

[Henrik Skupin [:whimboo][⌚️UTC+1]]

Attachment: Assertion (Input wasn't UTF8 or incorrect length was calculated)

Comment 7

[David Baron :dbaron: (⌚️UTC-5, no longer working on Mozilla)]

Is this a regression from bug 25888?

Comment 8

[Henrik Skupin [:whimboo][⌚️UTC+1]]

Attachment: Assertion (Adding child where we already have a child?)

This one I see after I have disabled Flash.

Comment 9

[Henrik Skupin [:whimboo][⌚️UTC+1]]

Attachment: Assertion (available space should never grow: 'aOldAvailableSpace.x <= aNewAvailableSpace.x && aOldAvailableSpace.XMost() >= aNewAvailableSpace.XMost())

The following assertion is visible when doing a shift+reload.

Comment 10

[Henrik Skupin [:whimboo][⌚️UTC+1]]

I believe the last added assertion is the one which is causing this problem. We loop in nsBlockFrame::PlaceLine:

http://mxr.mozilla.org/mozilla-central/source/layout/generic/nsBlockFrame.cpp#4042

Comment 11

[Olli Pettay [:smaug][[e-mail]]]

(In reply to comment #7)
> Is this a regression from bug 25888?
Probably. I can't reproduce this using 2009/05/19, but can with 2009/05/21

Comment 12

[Henrik Skupin [:whimboo][⌚️UTC+1]]

I'll bisect: http://hg.mozilla.org/mozilla-central/pushloghtml?fromchange=815856a7144b&tochange=be75e40365ce

Comment 13

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

There's probably not much point in bisecting since we know what bug it was.

Comment 14

[Henrik Skupin [:whimboo][⌚️UTC+1]]

(In reply to comment #13)
> There's probably not much point in bisecting since we know what bug it was.

We only have the date but yes, after running builds from both changesets I can second that it is a regression from bug 25888.

Shall I file separate bugs for the other assertions?

Comment 15

[David Baron :dbaron: (⌚️UTC-5, no longer working on Mozilla)]

(In reply to comment #9)
> Created an attachment (id=381497) [details]
> Assertion (available space should never grow: 'aOldAvailableSpace.x <=
> aNewAvailableSpace.x && aOldAvailableSpace.XMost() >=
> aNewAvailableSpace.XMost())
> 
> The following assertion is visible when doing a shift+reload.

This assertion is what causes the problem in this bug.

The other assertions are other bugs, though may already be filed.

Comment 16

[Henrik Skupin [:whimboo][⌚️UTC+1]]

(In reply to comment #15)
> The other assertions are other bugs, though may already be filed.

I filed bug 496336 for the UTF8 assertions.

Comment 17

[David Baron :dbaron: (⌚️UTC-5, no longer working on Mozilla)]

This works for me, but it wouldn't surprise me if the patch I'm going to attach to bug 494332 fixes it.

Comment 18

[Olli Pettay [:smaug][[e-mail]]]

I believe something has changed in the page, since I'm no longer getting
lockups either.

Comment 19

[Henrik Skupin [:whimboo][⌚️UTC+1]]

It's still there with Mozilla/5.0 (Macintosh; U; Intel Mac OS X 10.5; en-US; rv:1.9.2a1pre) Gecko/20090609 Minefield/3.6a1pre ID:20090609212135.

As I said in comment 9 just do a shift+reload. Only then you will see the assertion.

Comment 21

[David Baron :dbaron: (⌚️UTC-5, no longer working on Mozilla)]

Could somebody test with a build from 2009-06-17 or later?  I suspect this was fixed by bug 494332?

Comment 22

[philippe (part-time)]

(In reply to comment #21)
> Could somebody test with a build from 2009-06-17 or later?  I suspect this was
> fixed by bug 494332?
With the most recent OS X build (20090709), no more problems.

Comment 24

[Henrik Skupin [:whimboo][⌚️UTC+1]]

Looks good for me too. The assertion is not visible anymore.