[Android] Suppress overscroll for unscrollable axes in the browser

[Android] Suppress overscroll for unscrollable axes in the browser

Now that overscroll updates are bundled with event ack IPCs, they're
much less expensive. Accordingly, move overscroll filtering logic for
unscrollable axes to the browser, avoiding unnecessary complexity in the
renderer for both Blink and cc.

BUG=506193
CQ_INCLUDE_TRYBOTS=tryserver.blink:linux_blink_rel

Committed: https://crrev.com/3367dde4ac60c00d422cdd087525ce2f55e5a6e4
Cr-Commit-Position: refs/heads/master@{#340391}

[jdduke (slow), 2015-07-10 18:12:51]

jdduke@chromium.org changed reviewers:
+ tdresser@chromium.org

[jdduke (slow), 2015-07-10 18:12:52]

tdresser@: PTAL. sataya.m@ had made a similar change earlier, but there were
issues with also removing the small epsilon check (which we still need in the
renderer).

[tdresser, 2015-07-10 18:41:30]

https://codereview.chromium.org/1236643003/diff/1/cc/layers/viewport.cc
File cc/layers/viewport.cc (left):

https://codereview.chromium.org/1236643003/diff/1/cc/layers/viewport.cc#oldco...
cc/layers/viewport.cc:161: adjusted.set_y(0.0f);
Why can't these be moved?

https://codereview.chromium.org/1236643003/diff/1/content/browser/android/ove...
File content/browser/android/overscroll_glow.cc (right):

https://codereview.chromium.org/1236643003/diff/1/content/browser/android/ove...
content/browser/android/overscroll_glow.cc:201:
std::ceil(viewport_size_.height()) < content_size.height();
We don't need an epsilon check here?

https://codereview.chromium.org/1236643003/diff/1/content/browser/android/ove...
content/browser/android/overscroll_glow.cc:268: }
Fix ;

[jdduke (slow), 2015-07-10 19:12:50]

https://codereview.chromium.org/1236643003/diff/1/cc/layers/viewport.cc
File cc/layers/viewport.cc (left):

https://codereview.chromium.org/1236643003/diff/1/cc/layers/viewport.cc#oldco...
cc/layers/viewport.cc:161: adjusted.set_y(0.0f);
On 2015/07/10 18:41:30, tdresser wrote:
> Why can't these be moved?

So, the problem is that we're often seeing small residual overscroll deltas
after nearly ever scroll (~.01 or less) even when the "full" scroll delta has
been consume. Ideally we'd fix that, but it seems to also happen on the Blink
thread so maybe there's something more fundamental. In any case, we don't want
to report overscroll when we've logically consumed the entire scroll delta.

https://codereview.chromium.org/1236643003/diff/1/content/browser/android/ove...
File content/browser/android/overscroll_glow.cc (right):

https://codereview.chromium.org/1236643003/diff/1/content/browser/android/ove...
content/browser/android/overscroll_glow.cc:201:
std::ceil(viewport_size_.height()) < content_size.height();
On 2015/07/10 18:41:30, tdresser wrote:
> We don't need an epsilon check here?

This is the logic we had before the check got moved to the renderer (see
https://codereview.chromium.org/230223006/diff/40001/content/browser/renderer...).
I can switch it to an epsilon check if you like.

https://codereview.chromium.org/1236643003/diff/1/content/browser/android/ove...
content/browser/android/overscroll_glow.cc:268: }
On 2015/07/10 18:41:30, tdresser wrote:
> Fix ;

Whoops, done.

[tdresser, 2015-07-10 19:20:13]

LGTM, thanks.

[jdduke (slow), 2015-07-10 23:20:55]

jdduke@chromium.org changed reviewers:
+ enne@chromium.org

[jdduke (slow), 2015-07-10 23:20:56]

+enne@ for cc/ and content/renderer/gpu.

[jdduke (slow), 2015-07-16 15:23:57]

On 2015/07/10 23:20:56, jdduke wrote:
> +enne@ for cc/ and content/renderer/gpu.

Friendly ping? enne@ any objections here?

[jdduke (slow), 2015-07-23 16:46:48]

jdduke@chromium.org changed reviewers:
+ aelias@chromium.org, sievers@chromium.org
- enne@chromium.org

[jdduke (slow), 2015-07-23 16:46:48]

+aelias for cc/
+sievers for content/renderer/gpu

[aelias_OOO_until_Jul13, 2015-07-23 18:27:16]

lgtm

[no sievers, 2015-07-23 22:17:53]

lgtm

[jdduke (slow), 2015-07-23 22:22:33]

The CQ bit was checked by jdduke@chromium.org

[commit-bot: I haz the power, 2015-07-23 22:23:36]

CQ is trying da patch. Follow status at
 https://chromium-cq-status.appspot.com/patch-status/1236643003/20001
View timeline at
 https://chromium-cq-status.appspot.com/patch-timeline/1236643003/20001

[commit-bot: I haz the power, 2015-07-23 23:22:30]

The CQ bit was unchecked by commit-bot@chromium.org

[commit-bot: I haz the power, 2015-07-23 23:22:30]

Try jobs failed on following builders:
  win_chromium_x64_rel_ng on tryserver.chromium.win (JOB_FAILED,
http://build.chromium.org/p/tryserver.chromium.win/builders/win_chromium_x64_...)

[jdduke (slow), 2015-07-25 00:19:06]

The CQ bit was checked by jdduke@chromium.org

[jdduke (slow), 2015-07-25 00:19:06]

The patchset sent to the CQ was uploaded after l-g-t-m from aelias@chromium.org,
sievers@chromium.org, tdresser@chromium.org
Link to the patchset: https://codereview.chromium.org/1236643003/#ps40001
(title: "Remove old test")

[commit-bot: I haz the power, 2015-07-25 00:19:49]

CQ is trying da patch. Follow status at
 https://chromium-cq-status.appspot.com/patch-status/1236643003/40001
View timeline at
 https://chromium-cq-status.appspot.com/patch-timeline/1236643003/40001

[commit-bot: I haz the power, 2015-07-25 01:23:54]

Committed patchset #3 (id:40001)

[commit-bot: I haz the power, 2015-07-25 01:24:47]

Patchset 3 (id:??) landed as
https://crrev.com/3367dde4ac60c00d422cdd087525ce2f55e5a6e4
Cr-Commit-Position: refs/heads/master@{#340391}