695964 - "ASSERTION: Can't get a delta for an untransformed frame!" with transform, svg:foreignObject

Status: RESOLVED WORKSFORME

(Core :: Layout, defect)

Description

[Jesse Ruderman]

Attachment: testcase

###!!! ASSERTION: Can't get a delta for an untransformed frame!: 'aFrame->GetStyleDisplay()->HasTransform()', file layout/base/nsDisplayList.cpp, line 2370

###!!! ASSERTION: If we don't have a transform, then we must be at least attempting to preserve the transforms of our children: 'aFrame->GetStyleDisplay()->mTransformStyle == NS_STYLE_TRANSFORM_STYLE_PRESERVE_3D', file layout/base/nsDisplayList.cpp, line 2518

Comment 1

[Jesse Ruderman]

Attachment: stack trace

Comment 2

[Matt Woodrow (:mattwoodrow)]

nsSVGForeignObjectFrame::IsTransformed is overridden to always return true, which is what we we're checking in nsFrame::FinishAndStoreOverflow.

nsDisplayTransform::GetResultingTransformMatrix requires GetStyleDisplay()->HasTransform() to be true, which it isn't

I'm not sure what the proper behaviour here is, why are foreign objects returning true for transformed when they don't have a transform?

Comment 3

[Timothy Nikkel (:tnikkel)]

If it's not actually transformed maybe we can make it return false?

Comment 4

[Matt Woodrow (:mattwoodrow)]

http://mxr.mozilla.org/mozilla-central/source/layout/svg/base/src/nsSVGForeignObjectFrame.h#84

This looks like an intentional change?

The default implementation would be returning false here, I'm not sure why it has an override.

Comment 5

[Timothy Nikkel (:tnikkel)]

Perhaps it was just simpler than checking if it actually had a non-identity transform?

Comment 6

[Jesse Ruderman]

Modified assertion still happens on trunk:

###!!! ASSERTION: If we don't have a transform, then we must have another reason to have an nsDisplayTransform created: 'aFrame->GetStyleDisplay()->mTransformStyle == NS_STYLE_TRANSFORM_STYLE_PRESERVE_3D || aFrame->GetStyleDisplay()->mBackfaceVisibility == NS_STYLE_BACKFACE_VISIBILITY_HIDDEN', file layout/base/nsDisplayList.cpp, line 2557

Comment 7

[Ryan VanderMeulen [:RyanVM]]

No longer asserts on trunk.

Comment 8

[Pulsebot]

https://hg.mozilla.org/integration/mozilla-inbound/rev/8d523868cf0c

Comment 9

[Carsten Book [:Tomcat]]

https://hg.mozilla.org/mozilla-central/rev/8d523868cf0c

Comment 10

[Carsten Book [:Tomcat]]

https://hg.mozilla.org/releases/mozilla-b2g44_v2_5/rev/8d523868cf0c