Chromium Code Reviews
chromiumcodereview-hr@appspot.gserviceaccount.com (chromiumcodereview-hr) | Please choose your nickname with Settings | Help | Chromium Project | Gerrit Changes | Sign out
(648)

Issue 9817001: Implement active wheel fling transfer via WebCompositorInputHandlerClient (Closed)

Created:
8 years, 9 months ago by nduca
Modified:
8 years, 9 months ago
Reviewers:
jamesr
CC:
chromium-reviews, joi+watch-content_chromium.org, darin-cc_chromium.org, jam, apatrick_chromium, jamesr
Visibility:
Public.

Description

Implement active wheel fling transfer via WebCompositorInputHandlerClient There are cases where we'll process a GestureFlingStart event on the compositor thread, initiate a wheel fling, and then later determine that we can't process the rest of the animation on the compositor thread. One example of where this could happen is if the page registers a mousewheel JS event listener while the fling is in progress. In this case, we need to transfer the animation to the WebView to process so it can generate the correct events. This implements the transfer as a call out via the WebCompositorInputHandlerClient. I think this is a reasonable place to put the interface, since it's intimately related with the input processing stream. BUG=don't have handy, but there is one TEST=wheel fling on CrOS should work with --enable-threaded-compositing

Patch Set 1 #

Patch Set 2 : Really builds #

Total comments: 1
Unified diffs Side-by-side diffs Delta from patch set Stats (+84 lines, -16 lines) Patch
M content/renderer/gpu/compositor_thread.h View 1 2 chunks +7 lines, -1 line 0 comments Download
M content/renderer/gpu/compositor_thread.cc View 1 6 chunks +77 lines, -15 lines 1 comment Download

Messages

Total messages: 3 (0 generated)
nduca
8 years, 9 months ago (2012-03-21 07:42:47 UTC) #1
jamesr
https://chromiumcodereview.appspot.com/9817001/diff/2001/content/renderer/gpu/compositor_thread.cc File content/renderer/gpu/compositor_thread.cc (right): https://chromiumcodereview.appspot.com/9817001/diff/2001/content/renderer/gpu/compositor_thread.cc#newcode92 content/renderer/gpu/compositor_thread.cc:92: // We will always destroy the CompositorWrapper through note: ...
8 years, 9 months ago (2012-03-21 14:38:30 UTC) #2
jamesr
8 years, 9 months ago (2012-03-21 18:05:17 UTC) #3
RenderViewImpl supports WeakPtr<>, so if we can get a hold of one of those then
the shutdown path should be easy. I don't know of any way to do that but it
_should_ be easy to do given a routing_id.  We could also probably do it as part
of the input handler initialization, since that's always initiated from a
RenderView.

I think someone else should review this, I wrote some of it so I'm not a good
objective viewer.

Powered by Google App Engine
This is Rietveld 408576698