Use viewport coords in gpuBenchmarking gestures
This is a follow-up CL to: https://crrev.com/c/690314 Currently, gpuBenchmarking takes coordinates in the root-frame coordinate space. Since input events are in the (visual) viewport space, gpuBenchmarking was trying to convert them in the presence of pinch-zoom by applying the pageScaleFactor. This is incorrect, as the transformation must also include the visual viewport offset. However, it makes a lot more sense for gpuBenchmarking to take coordinates in the viewport space to begin with, leaving the conversion to the caller. The above CL converts telemetry to pass these coordinates in viewport space. It's gated on the presence of the `gesturesExpectedInViewportCoordinates` attribute so that we can make the transition without breaking the waterfall. Bug: 610021 Change-Id: Ic6a4229e230b77427e6f1aa7ef71933c21d93106 Reviewed-on: https://chromium-review.googlesource.com/668376 Commit-Queue: David Bokan <bokan@chromium.org> Reviewed-by:Eric Karl <ericrk@chromium.org> Reviewed-by:
Alexandre Elias <aelias@chromium.org> Cr-Commit-Position: refs/heads/master@{#506804}
Showing
Please register or sign in to comment