1118051 - layout/reftests/border-radius/clipping-6.html fails on Android when event-regions are enabled

Status: RESOLVED FIXED

(Core :: Layout, defect)

Description

[Kartikaya Gupta (email:[e-mail])]

See the try run at https://treeherder.mozilla.org/ui/#/jobs?repo=try&revision=2f2eafcb696d

There is a R2 failure on Android 4.0 API10+ debug as a result of turning on event-regions. This needs to be investigated and resolved before we can turn on event-regions across all platforms.

Comment 1

[Timothy Nikkel (:tnikkel)]

With event regions enabled we get a bunch of empty extra layers for the event regions, but I'm not sure if that's causing the difference. The event regions items increase the size of the transform layer's visible region. I think this might be the reason. Even if it's not we don't want to increase the size of it (in this case the affected layer has an intermediate surface so we would definitely be using more resources then we need to). As long as the hit testing code can handle having hit regions be larger than the visible region of the layer.

Comment 2

[Kartikaya Gupta (email:[e-mail])]

Yeah, the event-regions-based hit testing code doesn't check the visible region so it should be fine if the hit region is larger than the visible region.

Comment 3

[Timothy Nikkel (:tnikkel)]

Attachment: patch

I checked that we still get event regions on layers, and even layers that are only created to hold event regions (and so have an empty visible region) are still created and get event regions.

Comment 4

[Timothy Nikkel (:tnikkel)]

https://hg.mozilla.org/integration/mozilla-inbound/rev/653ead886844

Comment 5

[Carsten Book [:Tomcat]]

https://hg.mozilla.org/mozilla-central/rev/653ead886844

Comment 6

[Timothy Nikkel (:tnikkel)]

Backed out for causing bug 1130067

https://hg.mozilla.org/integration/mozilla-inbound/rev/ecd613d5d87c

Comment 7

[Timothy Nikkel (:tnikkel)]

The problem in bug 1130067 appears to be that when we call FindPaintedLayerFor we always choose the bottom most suitable painted layer because the vis rect passed in is empty (so doesn't intersect the draw rect of any of the painted layers in the stack). I hacked up a patch to use the non-empty way of calculating bounds only for FindPaintedLayerFor for layer event regions items, but empty GetBounds for everything else and that fixes it.

Comment 8

[Timothy Nikkel (:tnikkel)]

Attachment: patch v2

Comment 9

[Timothy Nikkel (:tnikkel)]

https://hg.mozilla.org/integration/mozilla-inbound/rev/82b5134173f4

Comment 10

[Carsten Book [:Tomcat]]

https://hg.mozilla.org/mozilla-central/rev/82b5134173f4

Comment 11

[Kartikaya Gupta (email:[e-mail])]

tn, can you request uplift to b2g37 for this bug? It is needed in order to fix bug 1139541 on 2.2. Thanks!

Comment 12

[Timothy Nikkel (:tnikkel)]

Comment on attachment 8562021 [details] [diff] [review]
patch v2

NOTE: Please see https://wiki.mozilla.org/Release_Management/B2G_Landing to better understand the B2G approval process and landings.

[Approval Request Comment]
Bug caused by (feature/regressing bug #): need this for bug 1139541
User impact if declined: bug 1139541
Testing completed: been landed for a while now, no issues
Risk to taking this patch (and alternatives if risky): pretty safe
String or UUID changes made by this patch: none

Comment 13

[Ryan VanderMeulen [:RyanVM]]

https://hg.mozilla.org/releases/mozilla-b2g37_v2_2/rev/c7eb8258adf8