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

Issue 635796 link

Starred by 1 user
Status: Duplicate
Merged: issue 636360
Owner:
u...@chromium.org
Closed: Aug 2016
Cc:
█ hpayer@chromium.org
sullivan@chromium.org
rmcilroy@chromium.org
u...@chromium.org
EstimatedDays: ----
NextAction: ----
OS: ----
Pri: 2
Type: Bug-Regression

Blocked on:
issue 636310


Sign in to add a comment

12.5%-40.3% regression in v8.top_25_smooth at 410087:410358

Project Member Reported by rmcilroy@chromium.org, Aug 9 2016 
See the link to graphs below.
Project Member

Comment 3 by 42576172...@developer.gserviceaccount.com, Aug 9 2016 


===== BISECT JOB RESULTS =====
Status: failed


===== TESTED REVISIONS =====
Revision         Mean     Std Dev  N  Good?
chromium@410322  99.1011  3.28355  5  good
chromium@410323  100.015  3.81266  5  good
chromium@410324  117.019  1.00688  5  bad
chromium@410325  117.361  3.42546  5  bad
chromium@410328  117.958  5.85048  5  bad

Bisect job ran on: android_one_perf_bisect
Bug ID: 635796

Test Command: src/tools/perf/run_benchmark -v --browser=android-chromium --output-format=chartjson --upload-results --also-run-disabled-tests v8.key_mobile_sites_smooth
Test Metric: v8_gc_total/v8_gc_total
Relative Change: 19.03%
Score: 0

Buildbot stdio: http://build.chromium.org/p/tryserver.chromium.perf/builders/android_one_perf_bisect/builds/1504
Job details: https://chromeperf.appspot.com/buildbucket_job_status/9004820595303696288


Not what you expected? We'll investigate and get back to you!
  https://chromeperf.appspot.com/bad_bisect?try_job_id=5311283172212736

| 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 component Tests>AutoBisect.  Thank you!
Project Member

Comment 5 by 42576172...@developer.gserviceaccount.com, Aug 9 2016 


===== BISECT JOB RESULTS =====
Status: failed


===== TESTED REVISIONS =====
Revision         Mean     Std Dev  N  Good?
chromium@410322  99.1011  3.28355  5  good
chromium@410323  100.015  3.81266  5  good
chromium@410324  117.019  1.00688  5  bad
chromium@410325  117.361  3.42546  5  bad
chromium@410328  117.958  5.85048  5  bad

Bisect job ran on: android_one_perf_bisect
Bug ID: 635796

Test Command: src/tools/perf/run_benchmark -v --browser=android-chromium --output-format=chartjson --upload-results --also-run-disabled-tests v8.key_mobile_sites_smooth
Test Metric: v8_gc_total/v8_gc_total
Relative Change: 19.03%
Score: 0

Buildbot stdio: http://build.chromium.org/p/tryserver.chromium.perf/builders/android_one_perf_bisect/builds/1504
Job details: https://chromeperf.appspot.com/buildbucket_job_status/9004820595303696288


Not what you expected? We'll investigate and get back to you!
  https://chromeperf.appspot.com/bad_bisect?try_job_id=5311283172212736

| 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 component Tests>AutoBisect.  Thank you!

Comment 6 by rmcilroy@chromium.org, Aug 9 2016 

Looks like this was caused by something in the V8 roll (as expected) but the bisect bot is not descending down into the V8 repo. Annie, is this related to the https://github.com/catapult-project/catapult/issues/2626 or is there another issue?

My suspicion is this might be "Unify memory optimization modes and flags." https://codereview.chromium.org/2218703004" any thoughts Ulan?

Comment 7 by rmcilroy@chromium.org, Aug 9 2016

Cc: sullivan@chromium.org u...@chromium.org rmcilroy@chromium.org
Owner: u...@chromium.org
Looks like this was caused by something in the V8 roll (as expected) but the bisect bot is not descending down into the V8 repo. Annie, is this related to the https://github.com/catapult-project/catapult/issues/2626 or is there another issue?

My suspicion is this might be "Unify memory optimization modes and flags." https://codereview.chromium.org/2218703004" any thoughts Ulan?

Comment 8 by rmcilroy@chromium.org, Aug 9 2016 

 Issue 635795  has been merged into this issue.
Project Member

Comment 10 by 42576172...@developer.gserviceaccount.com, Aug 10 2016 


===== BISECT JOB RESULTS =====
Status: failed


===== TESTED REVISIONS =====
Revision         Mean     Std Dev   N  Good?
chromium@410322  85.9527  1.74197   5  good
chromium@410323  84.3195  0.563767  5  good
chromium@410324  100.923  2.7448    5  bad
chromium@410325  96.9791  4.37677   5  bad
chromium@410328  95.5434  2.30454   5  bad

Bisect job ran on: mac_hdd_perf_bisect
Bug ID: 635796

Test Command: src/tools/perf/run_benchmark -v --browser=release --output-format=chartjson --upload-results --also-run-disabled-tests v8.top_25_smooth
Test Metric: v8_gc_total/v8_gc_total
Relative Change: 11.16%
Score: 0

Buildbot stdio: http://build.chromium.org/p/tryserver.chromium.perf/builders/mac_hdd_perf_bisect/builds/728
Job details: https://chromeperf.appspot.com/buildbucket_job_status/9004820595303696288


Not what you expected? We'll investigate and get back to you!
  https://chromeperf.appspot.com/bad_bisect?try_job_id=5311283172212736

| 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 component Tests>AutoBisect.  Thank you!

Comment 11 by rmcilroy@chromium.org, Aug 10 2016

Blockedon: 636310
Project Member

Comment 13 by 42576172...@developer.gserviceaccount.com, Aug 11 2016 


===== BISECT JOB RESULTS =====
Status: failed


===== TESTED REVISIONS =====
Revision         Mean     Std Dev  N  Good?
chromium@410322  42.4211  2.09161  8  good
chromium@410328  46.5773  2.65857  8  bad

Bisect job ran on: android_nexus6_perf_bisect
Bug ID: 635796

Test Command: src/tools/perf/run_benchmark -v --browser=android-chromium --output-format=chartjson --upload-results --also-run-disabled-tests v8.key_mobile_sites_smooth
Test Metric: v8_gc_total_outside_idle/v8_gc_total_outside_idle
Relative Change: 10.03%
Score: 0

Buildbot stdio: http://build.chromium.org/p/tryserver.chromium.perf/builders/android_nexus6_perf_bisect/builds/2458
Job details: https://chromeperf.appspot.com/buildbucket_job_status/9004633386811579888


Not what you expected? We'll investigate and get back to you!
  https://chromeperf.appspot.com/bad_bisect?try_job_id=5818266010779648

| 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 component Tests>AutoBisect.  Thank you!
Project Member

Comment 15 by 42576172...@developer.gserviceaccount.com, Aug 11 2016 


===== BISECT JOB RESULTS =====
Status: failed


===== TESTED REVISIONS =====
Revision         Mean     Std Dev  N  Good?
chromium@410322  42.4211  2.09161  8  good
chromium@410328  46.5773  2.65857  8  bad

Bisect job ran on: android_nexus6_perf_bisect
Bug ID: 635796

Test Command: src/tools/perf/run_benchmark -v --browser=android-chromium --output-format=chartjson --upload-results --also-run-disabled-tests v8.key_mobile_sites_smooth
Test Metric: v8_gc_total_outside_idle/v8_gc_total_outside_idle
Relative Change: 10.03%
Score: 0

Buildbot stdio: http://build.chromium.org/p/tryserver.chromium.perf/builders/android_nexus6_perf_bisect/builds/2458
Job details: https://chromeperf.appspot.com/buildbucket_job_status/9004633386811579888


Not what you expected? We'll investigate and get back to you!
  https://chromeperf.appspot.com/bad_bisect?try_job_id=5818266010779648

| 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 component Tests>AutoBisect.  Thank you!

Comment 16 by rmcilroy@chromium.org, Aug 11 2016

Cc: hpayer@chromium.org
+Hannes since Ulan is OOO. Still haven't been able to get a clean bisect, I'll try again.
Project Member

Comment 18 by 42576172...@developer.gserviceaccount.com, Aug 12 2016

Mergedinto: 636360
Status: Duplicate (was: Assigned)

===== BISECT JOB RESULTS =====
Status: completed


===== SUSPECTED CL(s) =====
Subject : Unify memory optimization modes and flags.
Author  : ulan
Commit description:
  
This patch folds --optimize-for-size flag and check for low-memory device
into Heap::ShouldOptimizeForMemoryUsage() predicate.

It has the following side effects:
- the heap growing factor for low-memory devices is capped at 1.3 (old value was 2.0).
- the memory reducer will be more aggressive for low-memory devices.

BUG= chromium:634900 

Review-Url: https://codereview.chromium.org/2218703004
Cr-Commit-Position: refs/heads/master@{#38406}
Commit  : 6cebf7d963c18e2ab09f2320c949baf0a4bed721
Date    : Sun Aug 07 15:36:10 2016


===== TESTED REVISIONS =====
Revision                       Mean     Std Dev  N  Good?
chromium@410319                86.3574  3.97748  6  good
chromium@410322                88.2326  4.64061  9  good
chromium@410323                89.3742  4.37259  7  good
chromium@410323,v8@422d14350d  83.41    2.25591  5  good
chromium@410323,v8@6cebf7d963  100.284  12.7474  9  bad    <--
chromium@410323,v8@5e685567ea  96.0386  4.54476  6  bad
chromium@410323,v8@265399ed9a  100.718  2.10733  5  bad
chromium@410324                100.477  7.16029  9  bad
chromium@410325                98.3245  7.0239   8  bad
chromium@410330                99.1376  6.06302  5  bad

Bisect job ran on: mac_hdd_perf_bisect
Bug ID: 635796

Test Command: src/tools/perf/run_benchmark -v --browser=release --output-format=chartjson --upload-results --also-run-disabled-tests v8.top_25_smooth
Test Metric: v8_gc_total/v8_gc_total
Relative Change: 9.71%
Score: 95.0

Buildbot stdio: http://build.chromium.org/p/tryserver.chromium.perf/builders/mac_hdd_perf_bisect/builds/729
Job details: https://chromeperf.appspot.com/buildbucket_job_status/9004618373157109616


Not what you expected? We'll investigate and get back to you!
  https://chromeperf.appspot.com/bad_bisect?try_job_id=5807081110634496

| 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 component Tests>AutoBisect.  Thank you!
►

Sign in to add a comment