1225005 - "Assertion failure: containingBlockWidth >= 0" with huge padding, columns, abs pos

Status: RESOLVED FIXED

(Core :: Layout, defect)

Description

[Jesse Ruderman]

Attachment: testcase

Assertion failure: containingBlockWidth >= 0, at layout/generic/nsFrame.cpp:4620

Comment 1

[Jesse Ruderman]

Attachment: stack

Comment 2

[Bob Clary [:bc] (inactive)]

fyi, Bughunter sees this on http://www.kasa.cz/udirna-rulyt-zahradni-mala-nahradni-obal-silne-deformovany-obal-2000014126/

Comment 3

[Carsten Book [:Tomcat]]

and now https://support.skype.com/ar/faq/FA12097/dlyl-jwdt-almkalmat-sth-mktb-windows?intsrc=client-_-windows-_-7.14-_-go-help.guides.soundsetupguide

Comment 4

[Carsten Book [:Tomcat]]

Attachment: stack from todays tip on m-c trunk (windows 7 debug build)

david is this something for you, or do you know who could take a look at this ?

Comment 5

[Carsten Book [:Tomcat]]

example url from bughunter was http://altapress.ru/story/168911?photo=6#gallery

Comment 6

[David Baron :dbaron: (⌚️UTC-4, no longer working on Mozilla)]

Perhaps Mats (who added the assertion) can look.  (Though I was going to suggest Mats even before checking that he added it...)

Comment 7

[Mats Palmgren (inactive)]

Attachment: fix+test

Comment 8

[Pulsebot]

https://hg.mozilla.org/integration/mozilla-inbound/rev/7c7fe4da388c

Comment 9

[Carsten Book [:Tomcat]]

https://hg.mozilla.org/mozilla-central/rev/7c7fe4da388c

Comment 10

[Ryan VanderMeulen [:RyanVM]]

The crashtest here appears to be permafailing on Aurora since the uplift today.
https://treeherder.mozilla.org/logviewer.html#?job_id=2096724&repo=mozilla-aurora

The two we're hitting are:

###!!! ASSERTION: Can't solve for both start and end: 'NS_AUTOOFFSET != offsets.IEnd(outerWM)', file /builds/slave/m-aurora-and-api-15-d-00000000/build/src/layout/generic/nsAbsoluteContainingBlock.cpp, line 448

###!!! ASSERTION: unexpected max block size: 'aBSize >= 0', file /builds/slave/m-aurora-and-api-15-d-00000000/build/src/layout/generic/nsFloatManager.cpp, line 128

Mats, is this safe to annotate away or did you want to take a look?

Comment 11

[Mats Palmgren (inactive)]

The code after those assertions looks safe regardless of values, so please annotate.  Thanks.

Comment 12

[Wes Kocher (:KWierso) (Not reading bugmail; email directly if needed)]

I've updated the expectations on aurora with https://hg.mozilla.org/releases/mozilla-aurora/rev/9a446438d3a4


I imagine we'll want to land something similar to trunk so we don't hit this every merge going forward, but it's still asserting four times on trunk. Would we want to change it to (Android,2-4) (!Android,4) to cover everything?

Comment 13

[Mats Palmgren (inactive)]

Sure.  Or just 2-4 for all platforms.

Comment 14

[Wes Kocher (:KWierso) (Not reading bugmail; email directly if needed)]

Landed https://hg.mozilla.org/integration/mozilla-inbound/rev/6312f4020b6d to inbound to fix this going forward.

Comment 15

[Wes Kocher (:KWierso) (Not reading bugmail; email directly if needed)]

https://hg.mozilla.org/mozilla-central/rev/6312f4020b6d