New issue
Advanced search Search tips

Issue 872592 link

Starred by 1 user

Issue metadata

Status: Untriaged
Owner: ----
Cc:
EstimatedDays: ----
NextAction: ----
OS: ----
Pri: 2
Type: Bug-Regression



Sign in to add a comment

A zero-to-nonzero regression in rendering.mobile at 580084:580278

Project Member Reported by toyoshim@chromium.org, Aug 9

Issue description

See the link to graphs below.
 
All graphs for this bug:
  https://chromeperf.appspot.com/group_report?bug_id=872592

(For debugging:) Original alerts at time of bug-filing:
  https://chromeperf.appspot.com/group_report?sid=0a961105f5adb2e17233f68b62c4c176ab66d6b23cdf0c15d6016841a06886f6


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

Android Nexus5 Perf
📍 Couldn't reproduce a difference.
https://pinpoint-dot-chromeperf.appspot.com/job/12cd3f94640000
In the bisect, the last good revision showed a bad result.
I started another bisect against an older revision range that should show good results. It they show bad results, this may depend on running environment, and is not a real regression.
📍 Couldn't reproduce a difference.
https://pinpoint-dot-chromeperf.appspot.com/job/14cdc028640000
ok, they show good results... let me try another bisect :/
r579416 good (#c6)
r579513 good (#c6)
r580083 bad (#c3)
r580278 bad (#c3)

so the new bisect run in [579513:580083]
Cc: eirage@chromium.org
Owner: eirage@chromium.org
Status: Assigned (was: Untriaged)
📍 Found a significant difference after 1 commit.
https://pinpoint-dot-chromeperf.appspot.com/job/1182e464640000

Enable use-zoom-for-dsf on Android on M70 by eirage@chromium.org
https://chromium.googlesource.com/chromium/src/+/d0b8bbd2dbf9c06dcbbd01a4b1285cf7a1bc2ba0
0.01604 → 1.506 (+1.49)

Understanding performance regressions:
  http://g.co/ChromePerformanceRegressions
Owner: ----
Status: Untriaged (was: Assigned)
There is no significant difference between disable/enable --use-zoom-for-dsf for 	thread_raster_cpu_time_per_frame on the text_hover test. My change isn't the cause of this regression.
Attach a local run result of text_hover_60000_pixels_per_second_desktop_gpu_raster. Left is with flag disable, same as without my change), right is with flag enable (with my change).

Enabling the flag didn't cause the value to go up but down a little bit.


results_60000.html
1.4 MB View Download

Sign in to add a comment