Account for DIP when setting the preferred size for tab capture.
This is an optimization for systems where there is more than one physical pixel per logical pixel (e.g., Chromebook Pixel, Mac Retina). On these systems, the render widget should be "preferably" sized such that its physical size will match that of the tab capture frame size. Because the entire graphics pipeline is handling N^2 less data, and also will no longer need to down-scale images, users should see huge tab capture performance improvements (e.g., higher frame rates, lower CPU utilization and fan noise). Expected side effect: Fullscreen-Within-Tab widgets will appear shrunk. However, since content renderers are still DIP-aware, any UI controls within the widget will appear right-sized. In other words, user interaction will not be diminished by this change. Testing: Confirmed Fullscreen-Within-Tab and tab capture output is correct by varying --force-device-scale-factor=X. Review URL: https://codereview.chromium.org/461443003 Cr-Commit-Position: refs/heads/master@{#289040} git-svn-id: svn://svn.chromium.org/chrome/trunk/src@289040 0039d316-1c4b-4281-b951-d872f2087c98
Showing
Please register or sign in to comment