New issue
Advanced search Search tips

Issue 771292 link

Starred by 1 user
Status: WontFix
Owner:
eseckler@chromium.org
Closed: Apr 2018
Cc:
m...@chromium.org
dtu@chromium.org
EstimatedDays: ----
NextAction: ----
OS: ----
Pri: 2
Type: Bug-Regression

Blocked on:
issue 768530


Sign in to add a comment

11.8%-30.1% regression in performance_browser_tests at 503949:503956

Project Member Reported by majidvp@chromium.org, Oct 3 2017 
See the link to graphs below.
Project Member

Comment 1 by 42576172...@developer.gserviceaccount.com, Oct 3 2017 

All graphs for this bug:
  https://chromeperf.appspot.com/group_report?bug_id=771292

(For debugging:) Original alerts at time of bug-filing:
  https://chromeperf.appspot.com/group_report?sid=6afcd6813e7c8ebf94c8e54dd46d9d2b3110925574f27cecfa642ff020d364a2


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

chromium-rel-mac11-air

Comment 3 by primiano@chromium.org, Oct 13 2017

Cc: dtu@chromium.org
Not sure why bisect didn't autokick (+dtu)
Kicked manually
https://pinpoint-dot-chromeperf.appspot.com/job/14b655ad780000
Project Member

Comment 4 by 42576172...@developer.gserviceaccount.com, Oct 13 2017 

📍 Found a significant difference after 1 commit.
https://pinpoint-dot-chromeperf.appspot.com/job/14b655ad780000

[viz] Ensure that ClientLTFS acknowledges racy BeginFrames.
By eseckler@chromium.org · Sun Sep 24 07:41:53 2017
chromium @ 1b697fc546af51e355c458bc4e10f821a1c8d551

Understanding performance regressions:
  http://g.co/ChromePerformanceRegressions

Comment 5 by sullivan@chromium.org, Oct 13 2017

Owner: eseckler@chromium.org
Status: Assigned (was: Untriaged)

Comment 6 by dtu@chromium.org, Oct 13 2017 

#3 auto-kick: should be fixed in https://github.com/catapult-project/catapult/issues/3959 !

Comment 7 by eseckler@chromium.org, Oct 16 2017 

I'd imagine that the reason for this regression is similar to that in  bug 727513  - I'll try to get a trace. It seems the perf dashboard's "request debug trace" doesn't work at the moment, filed  bug 774947 .

Comment 8 by eseckler@chromium.org, Nov 14 2017

Blockedon: 762194

Comment 9 by eseckler@chromium.org, Nov 14 2017

Blockedon: -762194 768530

Comment 10 by eseckler@chromium.org, Jan 8 2018 

I haven't been able to retrieve a trace for the regressed benchmark, since the dashboard doesn't seem to support that for this benchmark.

I don't have access to a MacBook Air right now to try to repro and get a trace.

The patch in #4 also improved a similar benchmark on the same device, see  bug 742345 .

Comment 11 by eseckler@chromium.org, Apr 3 2018

Status: WontFix (was: Assigned)
Given that the patch improved a number of similar metrics (see graphs), I'm going to accept the regressions as intended / won't fix. Slight timing changes for frame production are expected for the change.

Comment 12 by majidvp@chromium.org, Apr 3 2018 

cc'ing benchmark owner miu@

The patch certainly improves other test in this benchmark (attached a screenshot)
ALtQSTgp4sD.png
390 KB View Download
►

Sign in to add a comment