This is a meta bug tracking Talos suites running on Try server including performance, crash and Time out relative issues currently I saw. The platform were focused on Linux x64 opt and Windows 10 x64 opt build. About the performance comparison, the way I did for these issues were sending a patch to enable/disable wr on the same platform like Linux x64 opt or Windows 10 x64 opt build to make sure the only difference was about to enable/disable wr. 1. On Linux x64 opt, the patch I modified to make sure the comparison would be disable_wr/enable_hw_acceler v.s. enable_wr/enable_hw_acceler. 1.1 Comparison link on Linux x64 opt https://treeherder.mozilla.org/perf.html#/compare?originalProject=try&originalRevision=226893bbeb3e&newProject=try&newRevision=624d8ef848c2d76eed6a2854e03e3c54950ffd12&framework=1 1.2 try link for disable_wr/enable_hw_acceler on Linux x64 opt https://treeherder.mozilla.org/#/jobs?repo=try&revision=226893bbeb3e 1.3 try link for enable_wr/enable_hw_acceler on Linux x64 opt https://treeherder.mozilla.org/#/jobs?repo=try&revision=624d8ef848c2d76eed6a2854e03e3c54950ffd12 1.4 try command: try: -b o -p linux64 -u none -t all[x64] --rebuild-talos 10 2. On Windows 10 x64 opt, 2.1 Comparison link on Windows 10 x64 opt https://treeherder.mozilla.org/perf.html#/compare?originalProject=try&originalRevision=e20b42426413&newProject=try&newRevision=ec5b6c878d2abaf1c98a5cdc1284fd38625f1822&framework=1 2.2 try link for disable_wr on Windows 10 x64 opt https://treeherder.mozilla.org/#/jobs?repo=try&revision=e20b42426413&selectedJob=142904881 2.3 try link for enable_wr on Windows 10 x64 opt https://treeherder.mozilla.org/#/jobs?repo=try&revision=ec5b6c878d2abaf1c98a5cdc1284fd38625f1822 3.3 try command: try: -b o -p win32,win64 -u none -t all[Windows 7,Windows 10] --rebuild-talos 10 Based on the above results, I will file bugs the delta for enabling wr is above 10% worser than disabling wr and depend on this. About the time out and crash relative bugs, bug 1409289 Comment 6 has an known to figure out. Besides this, there were some hit on Windows 10 x64 opt. For these, I will also file bugs and depend on this.

Here to attach the Comparison results on from today's rebase, which were both Linux x64 opt[1] and Windows 10 x64 opt[2] build. This result also force enabling gfx.webrender.blob-images to compare. [1]: https://treeherder.mozilla.org/perf.html#/compare?originalProject=try&originalRevision=03107377dac6&newProject=try&newRevision=26095b988f06e8cb097118dd9ba051966fe9b195&framework=1 [2]: https://treeherder.mozilla.org/perf.html#/compare?originalProject=try&originalRevision=051d048108f2&newProject=try&newRevision=52ee54a2130bcea167e32308374a6ccd0d5950b4&framework=1 I will update other performance bugs linked to this meta bug based on this result.

Now that all the talos tests are running on m-c/integration branches, I'm going to close this bug and use bug 1416635 for tracking the talos tests that are "regressing" compared to non-WR. We don't need two metabugs that are basically tracking the same thing.

Actually I'm going to use bug 1450997. We have too many metabugs...