New issue
Advanced search Search tips
Note: Color blocks (like █ or █) mean that a user may not be available. Tooltip shows the reason.

Issue 804315 link

Starred by 1 user
Status: WontFix
Owner:
ericrk@chromium.org
Closed: Jan 2018
Cc:
█ pmeenan@chromium.org
vmp...@chromium.org
ericrk@chromium.org
EstimatedDays: ----
NextAction: ----
OS: ----
Pri: 2
Type: Bug-Regression



Sign in to add a comment

14.9%-72.6% regression in rasterize_and_record_micro.top_25 at 529803:530042

Project Member Reported by pmeenan@chromium.org, Jan 22 2018 
See the link to graphs below.
Project Member

Comment 1 by 42576172...@developer.gserviceaccount.com, Jan 22 2018 

All graphs for this bug:
  https://chromeperf.appspot.com/group_report?bug_id=804315

(For debugging:) Original alerts at time of bug-filing:
  https://chromeperf.appspot.com/group_report?sid=b21c0e124f540e8d71c2227b192da598fb05a947bdc5470a75c85a301bcb2786


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

chromium-rel-win7-gpu-ati
chromium-rel-win7-gpu-intel
chromium-rel-win7-gpu-nvidia
win-high-dpi
Project Member

Comment 3 by 42576172...@developer.gserviceaccount.com, Jan 23 2018

Cc: vmp...@chromium.org ericrk@chromium.org
Owner: ericrk@chromium.org
Status: Assigned (was: Untriaged)
📍 Found a significant difference after 1 commit.
https://pinpoint-dot-chromeperf.appspot.com/job/149882a4840000

Move tile half-width optimization after all rounding
By ericrk@chromium.org · Wed Jan 17 23:49:37 2018
chromium @ a78535de3d2153706ef21bf918333353e59ea496

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

Comment 5 by 42576172...@developer.gserviceaccount.com, Jan 26 2018 

📍 Couldn't reproduce a difference.
https://pinpoint-dot-chromeperf.appspot.com/job/12f5e392840000

Comment 6 by ericrk@chromium.org, Jan 29 2018

Status: WontFix (was: Assigned)
In almost all cases, my change takes us back to our previous perf levels before https://chromium-review.googlesource.com/c/chromium/src/+/806994 landed. This is expected, as my change reduces the impact of /806994, as parts of that were causing other regressions. Overall I think everything's behaving as expected.

There was one graph where this pattern wasn't clear, but we also regressed in the ref graph, and a targeted check (c#4) shows that things are behaving as expected.
►

Sign in to add a comment