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

Issue 699130 link

Starred by 1 user

Issue metadata

Status: Fixed
Owner:
Closed: Mar 2017
Cc:
EstimatedDays: ----
NextAction: ----
OS: ----
Pri: 2
Type: Bug-Regression



Sign in to add a comment

4.9% regression in system_health.memory_mobile at 454643:454700

Project Member Reported by kraynov@chromium.org, Mar 7 2017

Issue description

See the link to graphs below.
 
All graphs for this bug:
  https://chromeperf.appspot.com/group_report?bug_id=699130

Original alerts at time of bug-filing:
  https://chromeperf.appspot.com/group_report?keys=agxzfmNocm9tZXBlcmZyFAsSB0Fub21hbHkYgIDglPeKrAkM


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

android-nexus6
Cc: herb@google.com
Owner: herb@google.com

=== Auto-CCing suspected CL author herb@google.com ===

Hi herb@google.com, the bisect results pointed to your CL, please take a look at the
results.


=== BISECT JOB RESULTS ===
Perf regression found with culprit

Suspected Commit
  Author : Herb Derby
  Commit : 01254bcf222a7806afbfbcb9c7fe13cc522448b3
  Date   : Fri Mar 03 21:11:08 2017
  Subject: Use an exponential growth strategy for extra blocks.

Bisect Details
  Configuration: android_nexus6_perf_bisect
  Benchmark    : system_health.memory_mobile
  Metric       : memory:chrome:all_processes:reported_by_chrome:malloc:effective_size_avg/browse_news/browse_news_qq
  Change       : 5.52% | 41555538.6667 -> 43851132.0

Revision                             Result                   N
chromium@454642                      41555539 +- 807078       6      good
chromium@454671                      41531851 +- 660622       6      good
chromium@454686                      41633432 +- 492992       6      good
chromium@454693                      41481127 +- 1057879      6      good
chromium@454693,skia@767c7e7a0b      41584273 +- 639368       6      good
chromium@454693,skia@e459afd6ce      41438009 +- 637426       6      good
chromium@454693,skia@374a0178cf      41623091 +- 483417       6      good
chromium@454693,skia@baf06bc89a      41328849 +- 898581       6      good
chromium@454693,skia@01254bcf22      43792379 +- 482630       6      bad       <--
chromium@454694                      43809869 +- 333162       6      bad
chromium@454695                      43792417 +- 453822       6      bad
chromium@454697                      43636527 +- 486676       6      bad
chromium@454700                      43851132 +- 342791       6      bad

Please refer to the following doc on diagnosing memory regressions:
  https://chromium.googlesource.com/chromium/src/+/master/docs/memory-infra/memory_benchmarks.md

To Run This Test
  src/tools/perf/run_benchmark -v --browser=android-chromium --output-format=chartjson --upload-results --pageset-repeat=1 --also-run-disabled-tests --story-filter=browse.news.qq system_health.memory_mobile

Debug Info
  https://chromeperf.appspot.com/buildbucket_job_status/8985765205880207888

Is this bisect wrong?
  https://chromeperf.appspot.com/bad_bisect?try_job_id=5231041234599936


| 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 Speed>Bisection.  Thank you!
Project Member

Comment 4 by bugdroid1@chromium.org, Mar 8 2017

The following revision refers to this bug:
  https://skia.googlesource.com/skia/+/7dd57b6a936af923a031f21c4ca9dc1031742473

commit 7dd57b6a936af923a031f21c4ca9dc1031742473
Author: Herb Derby <herb@google.com>
Date: Wed Mar 08 19:58:36 2017

Use Fibonacci instead of 2^n for block growth.

Chrome on android showed an increase of 5% memory use when 2^n block
growth was introduced. Use Fibonacci instead.

BUG= chromium:699130 

Change-Id: I228d66385c63d487e72db46356f44e9efb5fa0f3
Reviewed-on: https://skia-review.googlesource.com/9447
Reviewed-by: Ben Wagner <bungeman@google.com>
Commit-Queue: Herb Derby <herb@google.com>

[modify] https://crrev.com/7dd57b6a936af923a031f21c4ca9dc1031742473/src/core/SkArenaAlloc.h
[modify] https://crrev.com/7dd57b6a936af923a031f21c4ca9dc1031742473/src/core/SkArenaAlloc.cpp

Project Member

Comment 5 by bugdroid1@chromium.org, Mar 8 2017

The following revision refers to this bug:
  https://chromium.googlesource.com/chromium/src.git/+/f146073c9468c8339515f32e784444ff59e96495

commit f146073c9468c8339515f32e784444ff59e96495
Author: skia-deps-roller@chromium.org <skia-deps-roller@chromium.org>
Date: Wed Mar 08 22:28:40 2017

Roll src/third_party/skia/ 598524df9..41eac798f (8 commits)

https://skia.googlesource.com/skia.git/+log/598524df99bf..41eac798f8ae

$ git log 598524df9..41eac798f --date=short --no-merges --format='%ad %ae %s'
2017-03-08 bsalomon Add instanced rendering command line flag to viewer.
2017-03-06 herb Remove SkChunckAlloc
2017-03-08 herb Use Fibonacci instead of 2^n for block growth.
2017-03-08 reed remove SkClipVisitor
2017-03-08 djsollen Revert "Optimize mipmap downsample_2_2 in sRGB mode"
2017-03-08 reed remove SkPictureUtils.h
2017-03-08 kjlubick Add jobs for building for Chromecast
2017-03-08 mtklein Revert "Fix SkJumper gcc warning"

Created with:
  roll-dep src/third_party/skia
BUG= 699130 


Documentation for the AutoRoller is here:
https://skia.googlesource.com/buildbot/+/master/autoroll/README.md

If the roll is causing failures, see:
http://www.chromium.org/developers/tree-sheriffs/sheriff-details-chromium#TOC-Failures-due-to-DEPS-rolls


CQ_INCLUDE_TRYBOTS=master.tryserver.blink:linux_trusty_blink_rel
TBR=djsollen@chromium.org

Change-Id: I4d50315003ff7ebbd71d09cf469eadb1ead94a82
Reviewed-on: https://chromium-review.googlesource.com/451612
Reviewed-by: Skia Deps Roller <skia-deps-roller@chromium.org>
Commit-Queue: Skia Deps Roller <skia-deps-roller@chromium.org>
Cr-Commit-Position: refs/heads/master@{#455557}
[modify] https://crrev.com/f146073c9468c8339515f32e784444ff59e96495/DEPS

Comment 6 by herb@google.com, Mar 9 2017

Status: Fixed (was: Untriaged)
This appears to be fixed by commit:
chromium/chromium/src/f146073c9468c8339515f32e784444ff59e96495

which contains the skia: https://skia-review.googlesource.com/c/9447

This should be fixed.

Sign in to add a comment