Tracking bug for lazy handler deserialization results |
||
Issue descriptionPerf & stability results for the lazy handler deserialization trial run will be collected here. Design doc: http://goo.gl/QxZBL2 Tracking bug: v8:6624
,
Nov 10 2017
Convenient link to chromiumdash: https://chromiumdash-staging.googleplex.com/commit/9664d017f7e219878d4b1a6c25be6c6f9b1b91bd
,
Nov 10 2017
The following revision refers to this bug: https://chromium.googlesource.com/v8/v8.git/+/b0996927fb63f1e2277c8d1cddf4cae2435bdedb commit b0996927fb63f1e2277c8d1cddf4cae2435bdedb Author: Jakob Gruber <jgruber@chromium.org> Date: Fri Nov 10 10:46:33 2017 Revert "[snapshot] Flip --lazy-handler-deserialization" This reverts commit 9664d017f7e219878d4b1a6c25be6c6f9b1b91bd. Reason for revert: Planned revert, this is on Canary 3264 and V8 branch 6.4.245. Original change's description: > [snapshot] Flip --lazy-handler-deserialization > > Flip this for a day or two to gather perf bot results & stability > issues. Results will be collected in https://crbug.com/782637 . > > TBR=hablich@chromium.org > > Bug: v8:6624 , chromium:782637 > Change-Id: I70eb5d6287d4723f935715ef80d9ba129dc282f0 > Reviewed-on: https://chromium-review.googlesource.com/758375 > Reviewed-by: Jakob Gruber <jgruber@chromium.org> > Commit-Queue: Jakob Gruber <jgruber@chromium.org> > Cr-Commit-Position: refs/heads/master@{#49222} TBR=hablich@chromium.org,jgruber@chromium.org # Not skipping CQ checks because original CL landed > 1 day ago. Bug: v8:6624 , chromium:782637 Change-Id: I2a6ee63f7191d186810d7352ee63b6b664e74157 Reviewed-on: https://chromium-review.googlesource.com/763427 Commit-Queue: Jakob Gruber <jgruber@chromium.org> Reviewed-by: Jakob Gruber <jgruber@chromium.org> Cr-Commit-Position: refs/heads/master@{#49290} [modify] https://crrev.com/b0996927fb63f1e2277c8d1cddf4cae2435bdedb/src/flag-definitions.h
,
Nov 14 2017
A quick analysis of chromeperf/ alerts (https://chromeperf.appspot.com/group_report?rev=515142): Memory improvements across the board, all arches, all platforms. Usually at least 250K but often reaching 600K+ for all_processes. e.g.: https://chromeperf.appspot.com/report?sid=9c8093256c372f4343dec644a057c745085b2f4ef9d4fd695066f25a631f0ded&rev=515186 https://chromeperf.appspot.com/report?sid=c3035e159ec2260ba8630319313e42bca14d50488f054ff37dfc444d70a6cbca&rev=515182 https://chromeperf.appspot.com/report?sid=f40593d9a3eb5b4d11da7bf3d6497779019572c54b0f51b73374fa016583f577&rev=515213 https://chromeperf.appspot.com/report?sid=895bf731b6a8eb32b5070835702388841a02f8a2b7e01372c2f5a94a6f08e86a&rev=515213 https://chromeperf.appspot.com/report?sid=a7c848d40c54176d387af082037f679d0f9eda08576bfdf7acf5962d7e96d8d1&rev=515213 RCS: Parse-time changes are unrelated. gc-full-mark-compactor: This tanks and should be investigated. https://chromeperf.appspot.com/report?sid=336583d608473394c0dbb0504f1bfd7d65ecc519b0673d0ca1f69a8b61bd5628&rev=515164 https://chromeperf.appspot.com/report?sid=9ebaabd28135acf094fc37a097593c16a4b74ee82f0ee09f6dc791171915f8fc&rev=515524
,
Nov 22 2017
,
Nov 22 2017
Issue 787817 has been merged into this issue.
,
Nov 28 2017
Experiment is over and the flag has been landed. gc-full-mark-compactor regressions from #4 seem to come from a shift in gc timing (thanks ulan for the investigation). |
||
►
Sign in to add a comment |
||
Comment 1 by bugdroid1@chromium.org
, Nov 8 2017