Always fall back to root scroll layer for otherwise ignored scrolls

Always fall back to root scroll layer for otherwise ignored scrolls

Rather than special case when the scrolling layer should default to the root
layer upon an ignored ScrollBegin, always set it to the root layer.  This
avoids platform differences and eliminates corner case logic for setting
the currently scrolling layer.

BUG=299126
Committed: https://src.chromium.org/viewvc/chrome?view=rev&revision=244043

[jdduke (slow), 2013-12-05 17:25:20]

aelias@: Is this about what you had in mind?

[mkosiba (inactive), 2013-12-09 16:20:11]

cool! Thanks Jared! I'm happy as long as the tests continue to pass :)

[aelias_OOO_until_Jul13, 2013-12-26 20:38:00]

Sorry for the review delay.

https://codereview.chromium.org/105393003/diff/110001/cc/trees/layer_tree_hos...
File cc/trees/layer_tree_host_impl.cc (right):

https://codereview.chromium.org/105393003/diff/110001/cc/trees/layer_tree_hos...
cc/trees/layer_tree_host_impl.cc:2154: gfx::ScalePoint(viewport_point,
device_scale_factor_);
This would be incorrect when the transformation is more complex than this (for
example, in some WebView and devtools scenarios).  We should be reusing the
logic in ScrollLayerWithViewportSpaceDelta for this.  So probably the top
controls consumption logic needs to move to that method as well.

[jdduke (slow), 2014-01-09 20:37:28]

https://codereview.chromium.org/105393003/diff/110001/cc/trees/layer_tree_hos...
File cc/trees/layer_tree_host_impl.cc (right):

https://codereview.chromium.org/105393003/diff/110001/cc/trees/layer_tree_hos...
cc/trees/layer_tree_host_impl.cc:2154: gfx::ScalePoint(viewport_point,
device_scale_factor_);
On 2013/12/26 20:38:00, aelias wrote:
> This would be incorrect when the transformation is more complex than this (for
> example, in some WebView and devtools scenarios).  We should be reusing the
> logic in ScrollLayerWithViewportSpaceDelta for this.  So probably the top
> controls consumption logic needs to move to that method as well.

What if we simply use the |max_scroll_offset()|?  I tested this with the latest
patch set, and it seems to work really well in practice.

[aelias_OOO_until_Jul13, 2014-01-09 22:21:42]

lgtm

[jdduke (slow), 2014-01-09 22:40:56]

piman@: Review for content/renderer/gpu/? Thanks.

[piman, 2014-01-09 23:14:54]

lgtm

[commit-bot: I haz the power, 2014-01-09 23:20:28]

CQ is trying da patch. Follow status at
https://chromium-status.appspot.com/cq/jdduke@chromium.org/105393003/270001

[commit-bot: I haz the power, 2014-01-10 01:23:08]

Retried try job too often on linux_aura for step(s) interactive_ui_tests
http://build.chromium.org/p/tryserver.chromium/buildstatus?builder=linux_aura...

[commit-bot: I haz the power, 2014-01-10 01:46:47]

CQ is trying da patch. Follow status at
https://chromium-status.appspot.com/cq/jdduke@chromium.org/105393003/270001

[commit-bot: I haz the power, 2014-01-10 02:13:57]

Change committed as 244043