New issue
Advanced search Search tips

Issue 805365 link

Starred by 1 user

Issue metadata

Status: WontFix
Owner:
Closed: Jan 2018
Cc:
Components:
EstimatedDays: ----
NextAction: ----
OS: ----
Pri: 2
Type: Bug-Regression



Sign in to add a comment

140% regression in v8.browsing_desktop at 529563:530031

Project Member Reported by petermarshall@chromium.org, Jan 24 2018

Issue description

See the link to graphs below.
 
Project Member

Comment 1 by 42576172...@developer.gserviceaccount.com, Jan 24 2018

All graphs for this bug:
  https://chromeperf.appspot.com/group_report?bug_id=805365

(For debugging:) Original alerts at time of bug-filing:
  https://chromeperf.appspot.com/group_report?sid=33e04ec50fe1507db130de09cff1db153d7f3964ea70a3841f9e42c54477d8ca


Bot(s) for this bug's original alert(s):

chromium-rel-mac11
Project Member

Comment 2 by 42576172...@developer.gserviceaccount.com, Jan 24 2018

๐Ÿ“ Pinpoint job started.
https://pinpoint-dot-chromeperf.appspot.com/job/13c3eb54840000

Comment 3 by dtu@chromium.org, Jan 25 2018

The Pinpoint issue was fixed in https://github.com/catapult-project/catapult/issues/4185
I'll kick off a new bisect.
Project Member

Comment 4 by 42576172...@developer.gserviceaccount.com, Jan 25 2018

๐Ÿ“ Pinpoint job started.
https://pinpoint-dot-chromeperf.appspot.com/job/148e92bc840000
Project Member

Comment 5 by 42576172...@developer.gserviceaccount.com, Jan 28 2018

Cc: mlippautz@chromium.org dgozman@chromium.org jbroman@chromium.org pfeldman@chromium.org
Owner: pfeldman@chromium.org
Status: Assigned (was: Untriaged)
๐Ÿ“ Found significant differences after each of 2 commits.
https://pinpoint-dot-chromeperf.appspot.com/job/148e92bc840000

[bindings] Fix external memory accounting for SerializedScriptValue
By mlippautz@chromium.org ยท Wed Jan 17 14:37:16 2018
chromium @ da7248b830fcf503ab0fd0e4292c367bb5244ab9

DevTools: swapping from local frame to remote does not clear iframe contents in DOM domain.
By pfeldman@chromium.org ยท Wed Jan 17 17:49:25 2018
chromium @ 8f4c57cb0db64b9dc0c5aadcf50e0b1e457c612d

Understanding performance regressions:
  http://g.co/ChromePerformanceRegressions
Owner: mlippautz@chromium.org
My change is a noop w/o inspector open.
Cc: -mlippautz@chromium.org gab@chromium.org
Status: WontFix (was: Assigned)
My change can shift GC timing as we correctly report memory on the worker threads. In the regressed case I can see a MajorGC on a dedicated worker where we spend 44ms waiting for tasks to finish. Main thread wall time is 1.45ms. 

gab@ is currently looking into scheduling problems of V8 tasks in general and this might be related to the issues we have seen where we only make marginal progress on the main thread.

Comment 8 by gab@chromium.org, Feb 2 2018

Components: Speed>Bisection
@Speed>Bisection : I can't seem to get anything out of pinpoint..

Clicking on " Analyze benchmark results"
@ https://pinpoint-dot-chromeperf.appspot.com/job/148e92bc840000
results in "HTTP response was too large: 50504270. The limit is: 33554432."

And I have no idea how I'm supposed to interpret the rest of the UI...
Cc: simonhatch@chromium.org
Related: https://github.com/catapult-project/catapult/issues/4168
Cc: dtu@chromium.org
+dtu

Thanks, looks like a bunch of tasks expired, filed a bug about the weird results: https://github.com/catapult-project/catapult/issues/4227

Sign in to add a comment