1127679 - "ASSERTION: writing-mode mismatch" with inline-flex, rtl

Status: RESOLVED FIXED

(Core :: Layout, defect)

Description

[Jesse Ruderman]

Attachment: testcase

###!!! ASSERTION: writing-mode mismatch: 'aOrigin.GetWritingMode() == GetWritingMode()', file WritingModes.h

Comment 1

[Jesse Ruderman]

Attachment: stack

Comment 2

[Jonathan Kew [:jfkthame]]

Attachment: Reftest for relative positioning of flex items with different directionality

Here's a reftest that not only hits this assertion, but also exhibits incorrect rendering as a result of the incorrect writing mode.

Comment 3

[Jonathan Kew [:jfkthame]]

Attachment: Fix use of incorrect writing mode in flex-item relative positioning

And here's a patch that fixes it.

Comment 4

[Simon Montagu :smontagu]

Comment on attachment 8557335 [details] [diff] [review]
Fix use of incorrect writing mode in flex-item relative positioning

Review of attachment 8557335 [details] [diff] [review]:
-----------------------------------------------------------------

This looks right, but I wonder if the same issue is going to come up elsewhere, in which case it would be worth adding a version of ApplyRelativePositioning which takes a WritingMode and makes this conversion? Or maybe always doing so in the non-static version, since the conversion will be a no-op if the passed-in WritingMode is identical to mWritingMode?

Comment 5

[Jonathan Kew [:jfkthame]]

I wondered that, too, but I'm inclined to start with this and see how things go... if we find we're needing the same pattern in other places, we can consider refactoring into a helper.

Comment 6

[Jonathan Kew [:jfkthame]]

https://hg.mozilla.org/integration/mozilla-inbound/rev/3f097e9f3069
https://hg.mozilla.org/integration/mozilla-inbound/rev/16d608fdd682

Comment 7

[Ryan VanderMeulen [:RyanVM]]

https://hg.mozilla.org/mozilla-central/rev/3f097e9f3069
https://hg.mozilla.org/mozilla-central/rev/16d608fdd682