1352918 - Be finer grained about hit testing tree updates

Status: RESOLVED FIXED

(Core :: Panning and Zooming, enhancement, P3)

Description

[David Anderson [:dvander] - inactive, e-mail if emergency]

Currently we update the hit testing tree when any layer property changes, but really we only need to care about properties related to hit testing.

Comment 1

[David Anderson [:dvander] - inactive, e-mail if emergency]

Attachment: patch

This limits updates to relevant property changes, and also makes it easy to add diagnostics for why hit testing updates happen.

Comment 2

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

Comment on attachment 8853841 [details] [diff] [review]
patch

Review of attachment 8853841 [details] [diff] [review]:
-----------------------------------------------------------------

::: gfx/layers/LayerAttributes.h
@@ +180,5 @@
> +    if (mScrollbarDirection != aOther.mScrollbarDirection) {
> +      return false;
> +    }
> +    if (mScrollbarThumbRatio != aOther.mScrollbarThumbRatio) {
> +      return false;

I don't think we need thumb ratio here. It seems to be used in AsyncCompositionManager but not in the APZ hit testing. APZ uses LayerMetricsWrapper::GetScrollThumbLength which is just the visible region.

::: gfx/layers/ipc/LayerTransactionParent.cpp
@@ +449,5 @@
> +    const SimpleLayerAttributes& attrs = op.attrs();
> +    const SimpleLayerAttributes& orig = layer->GetSimpleAttributes();
> +    if (attrs.Transform() != orig.Transform()) {
> +      UpdateHitTestingTree(layer, "transform changed");
> +    } else if (!attrs.ScrollingInfoIsEqual(orig)) {

Any particular reason you're checking the transform separately, and not just putting it in the ScrollingInfoIsEqual check? Doesn't matter to me either way, just curious.

Comment 3

[David Anderson [:dvander] - inactive, e-mail if emergency]

I can't remember, but it makes more sense to fold it into one function anyway.

Comment 4

[Pulsebot]

Pushed by [e-mail]:
https://hg.mozilla.org/integration/mozilla-inbound/rev/56071dd2f7fa
Make finer-grained decisions about whether to update hit testing trees. (bug 1352918, r=kats)

Comment 5

[Sebastian Hengst [:aryx] (needinfo me if it's about an intermittent or backout)]

https://hg.mozilla.org/mozilla-central/rev/56071dd2f7fa