Revert "Use accurate X11 event timestamp computation."
This reverts commit 3225e1d5. Reason for revert: Findit (https://goo.gl/kROfz5) identified CL at revision 594844 as the culprit for flakes in the build cycles as shown on: https://findit-for-me.appspot.com/waterfall/flake/flake-culprit?key=ag9zfmZpbmRpdC1mb3ItbWVyQwsSDEZsYWtlQ3VscHJpdCIxY2hyb21pdW0vMzIyNWUxZDVlYTQ2MjZhNzcwZDgzNWRkMDNhMjViNmMwOTViNThiNQw Sample Failed Build: https://ci.chromium.org/buildbot/chromium.linux/Linux%20Tests/72634 Sample Failed Step: events_unittests Sample Flaky Test: X11EventTest.AutoRepeat Original change's description: > Use accurate X11 event timestamp computation. > > X events have a timestamp which is only well defined relative to the X11 Server > time. The previous computation for timestamp for X11 events was making the > assumption that Server time and Chrome time were the same. This assumption is > not always true -- this is likely the root cause of "bad" timestamps observed in > https://bugs.chromium.org/p/chromium/issues/detail?id=650338#c1 > > This CL changes event timestamp computation to make a roundtrip to the X11 > Server to get an accurate base::TimeTicks. This logic was lifted out of the > responsiveness calculator, which was already doing this computation. The latter > will subsequently be changed to use the computation in this CL. > > Change-Id: I963019cd8bfb8ce14e06b3743a159c9c85f2cb82 > Bug: 859155 > Reviewed-on: https://chromium-review.googlesource.com/1249383 > Reviewed-by: Thomas Anderson <thomasanderson@chromium.org> > Reviewed-by: Avi Drissman <avi@chromium.org> > Commit-Queue: Erik Chen <erikchen@chromium.org> > Cr-Commit-Position: refs/heads/master@{#594844} Change-Id: I911f8bd268739b5e91c550e3f6a6186ae4dfbecb No-Presubmit: true No-Tree-Checks: true No-Try: true Bug: 859155, 890121 Reviewed-on: https://chromium-review.googlesource.com/1250210 Cr-Commit-Position: refs/heads/master@{#594958}
Showing
Please register or sign in to comment