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

Issue 598614 link

Starred by 1 user

Issue metadata

Status: Archived
Owner: ----
Closed: Sep 2016
Cc:
Components:
EstimatedDays: ----
NextAction: ----
OS: ----
Pri: 3
Type: Bug



Sign in to add a comment

Bisect "Reproduced a change" of 0.00% (+/-0.00%)

Project Member Reported by perezju@chromium.org, Mar 29 2016

Issue description

What steps will reproduce the problem?
(1) Ran a bisect with a metric that did *not* change. (See output below.)
https://build.chromium.org/p/tryserver.chromium.perf/builders/android_nexus5_perf_bisect/builds/3520

What is the expected output?

Bisect should complain that no change was reproduced. And most certainly not blame some random CL in the range.

What do you see instead?

===== BISECT JOB RESULTS =====
Status: Positive: Reproduced a change.

Test Command: tools/perf/run_benchmark -v --browser=android-chromium --output-format=buildbot --also-run-disabled-tests memory.memory_health_plan --pageset-repeat 5
Test Metric: foreground-memory_android_memtrack_gl_gpu_process/http_yahoo_com
Relative Change: 0.00% (+/-0.00%)
Estimated Confidence: 99.90%
Retested CL with revert: Yes

===== SUSPECTED CL(s) =====
Subject : Use task runner instead of message loop in WifiDataProvider.
Author  : mvanouwerkerk@chromium.org
Link    : https://chromium.googlesource.com/chromium/src/+/cb9d633e2f1f1f13881899fa92f356bf8cd79e02
Commit  : cb9d633e2f1f1f13881899fa92f356bf8cd79e02
Date    : Thu, 10 Mar 2016 14:03:01 +0000

===== RETEST RESULTS =====
        ToT                                                                                           18599936.00     +-0.00                 
      Reverted                                                                                        18599936.00     +-0.00                 

===== TESTED COMMITS =====
       Depot          Position                                   SHA                                       Mean      Std. Error     State    
      chromium         380426                  236e6e14f9521698f801c74d34cb0621c3ccf8d7               18599936.00     +-0.00         Bad     
      chromium         380414                  4146624afda978a7f78e81eac9ac613d488f4827               18599936.00     +-0.00         Bad     
      chromium         380408                  fe36ec4d077c3a89979854bf37328dd41bb24ca6               18599936.00     +-0.00         Bad     
      chromium         380405                  ff323ec5588d6e61a4267ab32251cf2310c84533               18599936.00     +-0.00         Bad     
      chromium         380403                  fd7ede807eb0e64820c805e8d5d0caf7ba9c5516               18599936.00     +-0.00         Bad     

      chromium         380402                  cb9d633e2f1f1f13881899fa92f356bf8cd79e02               18599936.00     +-0.00     Suspected CL

      chromium         380401                  ed16d0afb92f998cad2cdd60543037f09e150263               18599936.00     +-0.00         Good    

Average build time : 0:00:51
Average test time  : 1:23:14

| O O | Visit http://www.chromium.org/developers/speed-infra/perf-bug-faq
|  X  | for more information addressing perf regression bugs. For feedback,
| / \ | file a bug with label Cr-Tests-AutoBisect.  Thank you!
 
Cc: -qyears...@chromium.org dtu@chromium.org robert...@chromium.org pras...@chromium.org
Components: Tests>AutoBisect

Comment 2 by dtu@chromium.org, Mar 29 2016

How did you kick off the bisect? Looks like it's going through a deprecated code path (non-recipe script-based bisect)
Yes, that was a manual config file here: https://codereview.chromium.org/1806583002 using the instructions from Prasad here: https://docs.google.com/document/d/1TrhCIxtPUOJjcr-60nJonsNuUUNolpUKEoonlKFrWtU/edit

Not sure if there is an alternative we could use at this time.
Status: Archived (was: Untriaged)
Components: Speed>Bisection

Sign in to add a comment