Chromium Code Reviews
chromiumcodereview-hr@appspot.gserviceaccount.com (chromiumcodereview-hr) | Please choose your nickname with Settings | Help | Chromium Project | Gerrit Changes | Sign out
(619)

Issue 750643003: Android: Improve Composite-with-no-GPU-channel handling (Closed)

Created:
6 years, 1 month ago by no sievers
Modified:
6 years, 1 month ago
Reviewers:
jdduke (slow)
CC:
chromium-reviews, darin-cc_chromium.org, jam
Base URL:
https://chromium.googlesource.com/chromium/src.git@master
Project:
chromium
Visibility:
Public.

Description

Android: Improve Composite-with-no-GPU-channel handling The current code simply bails out of Composite() when the channel is not established yet (or lost). This seems to work in practice, but following the code it looks like we would - generally still think that the Composite task is pending - in OnVSync think that we missed the window and post again - further attempts to schedule, including when the channel eventually gets established, will think it's already pending and going to happen without the need to post a task - but it will really only happen if OnVSync gets called again (somebody happened to request vsync) and we think we missed the frame and schedule again. Improve this by - tracking when we deferred Composite because we are waiting for the channel - bail out of PostComposite if we are in this condition - when established unset and post Composite Committed: https://crrev.com/a0e92afeb69eb7cca3fddeb86aa1238d51d672b3 Cr-Commit-Position: refs/heads/master@{#305298}

Patch Set 1 #

Total comments: 2

Patch Set 2 : #

Unified diffs Side-by-side diffs Delta from patch set Stats (+26 lines, -8 lines) Patch
M content/browser/renderer_host/compositor_impl_android.h View 2 chunks +7 lines, -1 line 0 comments Download
M content/browser/renderer_host/compositor_impl_android.cc View 1 5 chunks +19 lines, -7 lines 0 comments Download

Messages

Total messages: 8 (2 generated)
no sievers
Jared, can you take a look?
6 years, 1 month ago (2014-11-21 21:00:17 UTC) #2
jdduke (slow)
lgtm https://codereview.chromium.org/750643003/diff/1/content/browser/renderer_host/compositor_impl_android.cc File content/browser/renderer_host/compositor_impl_android.cc (right): https://codereview.chromium.org/750643003/diff/1/content/browser/renderer_host/compositor_impl_android.cc#newcode281 content/browser/renderer_host/compositor_impl_android.cc:281: weak_factory_.GetWeakPtr())); Maybe |DCHECK(!defer_composite_for_gpu_channel_);| here?
6 years, 1 month ago (2014-11-21 21:16:35 UTC) #3
no sievers
https://codereview.chromium.org/750643003/diff/1/content/browser/renderer_host/compositor_impl_android.cc File content/browser/renderer_host/compositor_impl_android.cc (right): https://codereview.chromium.org/750643003/diff/1/content/browser/renderer_host/compositor_impl_android.cc#newcode281 content/browser/renderer_host/compositor_impl_android.cc:281: weak_factory_.GetWeakPtr())); On 2014/11/21 21:16:34, jdduke wrote: > Maybe |DCHECK(!defer_composite_for_gpu_channel_);| ...
6 years, 1 month ago (2014-11-21 21:19:19 UTC) #4
commit-bot: I haz the power
CQ is trying da patch. Follow status at https://chromium-cq-status.appspot.com/patch-status/750643003/20001
6 years, 1 month ago (2014-11-21 21:20:15 UTC) #6
commit-bot: I haz the power
Committed patchset #2 (id:20001)
6 years, 1 month ago (2014-11-21 22:13:00 UTC) #7
commit-bot: I haz the power
6 years, 1 month ago (2014-11-21 22:13:38 UTC) #8
Message was sent while issue was closed.
Patchset 2 (id:??) landed as
https://crrev.com/a0e92afeb69eb7cca3fddeb86aa1238d51d672b3
Cr-Commit-Position: refs/heads/master@{#305298}

Powered by Google App Engine
This is Rietveld 408576698