New issue
Advanced search Search tips

Issue 795358 link

Starred by 2 users

Issue metadata

Status: Fixed
Owner:
Closed: Mar 2018
Cc:
EstimatedDays: ----
NextAction: ----
OS: ----
Pri: 2
Type: Bug-Regression



Sign in to add a comment

2872.6%-11161% regression in system_health.memory_mobile at 524202:524270

Project Member Reported by chiniforooshan@chromium.org, Dec 15 2017

Issue description

See the link to graphs below.
 
Project Member

Comment 1 by 42576172...@developer.gserviceaccount.com, Dec 15 2017

All graphs for this bug:
  https://chromeperf.appspot.com/group_report?bug_id=795358

(For debugging:) Original alerts at time of bug-filing:
  https://chromeperf.appspot.com/group_report?sid=90fae75b1efb720cbb4156d0436e3f8604d9db6db4cc1c13a33cc3eec05eb213


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

android-webview-nexus6
Project Member

Comment 3 by 42576172...@developer.gserviceaccount.com, Dec 16 2017

Cc: ericrk@chromium.org
Owner: ericrk@chromium.org
Status: Assigned (was: Untriaged)

=== Auto-CCing suspected CL author ericrk@chromium.org ===

Hi ericrk@chromium.org, 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 : Eric Karl
  Commit : 4310c895f6295c06f367e8484092ca9012a40fb5
  Date   : Thu Dec 14 22:47:29 2017
  Subject: Log deleted GPU discardable entries as 0 size

Bisect Details
  Configuration: android_webview_nexus6_aosp_perf_bisect
  Benchmark    : system_health.memory_mobile
  Metric       : memory:webview:all_processes:reported_by_chrome:skia:effective_size_avg/browse_social/browse_social_facebook_infinite_scroll
  Change       : 11216.11% | 973232.0 -> 110131988.0

Revision             Result                     N
chromium@524201      973232 +- 0.0              6      good
chromium@524210      973232 +- 0.0              6      good
chromium@524215      973232 +- 0.0              6      good
chromium@524216      113603068 +- 24862939      6      bad       <--
chromium@524217      110477820 +- 18883603      6      bad
chromium@524219      107889148 +- 16363925      6      bad
chromium@524236      114225660 +- 14425522      6      bad
chromium@524270      110131988 +- 23222425      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-webview --output-format=chartjson --upload-results --pageset-repeat=1 --also-run-disabled-tests --story-filter=browse.social.facebook.infinite.scroll system_health.memory_mobile

More information on addressing performance regressions:
  http://g.co/ChromePerformanceRegressions

Debug information about this bisect:
  https://chromeperf.appspot.com/buildbucket_job_status/8960114258635234048


For feedback, file a bug with component Speed>Bisection
Project Member

Comment 4 by 42576172...@developer.gserviceaccount.com, Dec 16 2017

Issue 795361 has been merged into this issue.
Project Member

Comment 5 by 42576172...@developer.gserviceaccount.com, Dec 18 2017

Issue 795763 has been merged into this issue.

Comment 6 by ericrk@chromium.org, Dec 18 2017

Status: Started (was: Assigned)
This is purely a logging issue (not a real regression). Forgot that this memory was logged from Skia as well. We need to update the Skia memory dump provider to log 0-size for discarded entries as well. I'll put together a CL.
Project Member

Comment 7 by 42576172...@developer.gserviceaccount.com, Dec 20 2017

Issue 795774 has been merged into this issue.
ericrk: Thanks for the explanation in #18! Are you still planning to put together a CL?

Comment 9 by ssid@chromium.org, Mar 13 2018

Cc: ssid@chromium.org
Sorry for the delay - patch is in progress:
https://skia-review.googlesource.com/c/skia/+/114681
Project Member

Comment 12 by bugdroid1@chromium.org, Mar 20 2018

The following revision refers to this bug:
  https://chromium.googlesource.com/chromium/src.git/+/120a10693da6c61782cf52e5c81fb30730362b37

commit 120a10693da6c61782cf52e5c81fb30730362b37
Author: skia-chromium-autoroll@skia-buildbots.google.com.iam.gserviceaccount.com <skia-chromium-autoroll@skia-buildbots.google.com.iam.gserviceaccount.com>
Date: Tue Mar 20 03:14:38 2018

Roll src/third_party/skia/ cb88470d8..ec77963bf (10 commits)

https://skia.googlesource.com/skia.git/+log/cb88470d82d6..ec77963bf32e

$ git log cb88470d8..ec77963bf --date=short --no-merges --format='%ad %ae %s'
2018-03-19 angle-skia-autoroll Roll skia/third_party/externals/angle2/ 3761dc7bf..daedf4d23 (3 commits)
2018-03-19 liyuqian Make consistent isRadiusIncreasing decision
2018-03-19 caryclark remove imageinfo flatten altogether
2018-03-19 bsalomon Revert "Revert "New read pixels implementation that is simpler but does all conversions on CPU.""
2018-03-19 bsalomon Remove deprecated versions of YUV texture SkImage factories
2018-03-19 ericrk Allow SkTraceMemoryDump to exclude wrapped objects
2018-03-19 hcm Fix go syntax for Debian compile
2018-03-19 bsalomon Remove SkSurface::MakeFromBackendTexture variant that doesn't take SkColorType
2018-03-19 ethannicholas fix switch fallthrough warning
2018-03-19 liyuqian Fix concentric 2pt conical gradient in GPU where r1 < r0

Created with:
  roll-dep src/third_party/skia
BUG= chromium:795358 


The AutoRoll server is located here: https://autoroll.skia.org

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

If the roll is causing failures, please contact the current sheriff, who should
be CC'd on the roll, and stop the roller if necessary.


CQ_INCLUDE_TRYBOTS=master.tryserver.blink:linux_trusty_blink_rel;luci.chromium.try:linux_optional_gpu_tests_rel;master.tryserver.chromium.mac:mac_optional_gpu_tests_rel;master.tryserver.chromium.win:win_optional_gpu_tests_rel
TBR=jvanverth@chromium.org

Change-Id: I83f77a4beed619b2bc27becaafe1224131abb3d1
Reviewed-on: https://chromium-review.googlesource.com/969800
Reviewed-by: skia-chromium-autoroll <skia-chromium-autoroll@skia-buildbots.google.com.iam.gserviceaccount.com>
Commit-Queue: skia-chromium-autoroll <skia-chromium-autoroll@skia-buildbots.google.com.iam.gserviceaccount.com>
Cr-Commit-Position: refs/heads/master@{#544276}
[modify] https://crrev.com/120a10693da6c61782cf52e5c81fb30730362b37/DEPS

Project Member

Comment 13 by bugdroid1@chromium.org, Mar 21 2018

The following revision refers to this bug:
  https://chromium.googlesource.com/chromium/src.git/+/958358e66f1dfca7b8f1b71bce9485a5d4058739

commit 958358e66f1dfca7b8f1b71bce9485a5d4058739
Author: Eric Karl <ericrk@chromium.org>
Date: Wed Mar 21 18:52:10 2018

Tell Skia memory dump not to dump wrapped/external objects

Skia currently dumps wrapped objects, and Chrome tries to alias these
objects with Chrome's own dumps. This isn't always possible, so instead
this change uses a new Skia API to ask Skia to skip dumping these.

Bug:  795358 
Change-Id: I143082aa340495f083a651532c8bbce621516490
Reviewed-on: https://chromium-review.googlesource.com/969709
Reviewed-by: Brian Salomon <bsalomon@chromium.org>
Commit-Queue: Eric Karl <ericrk@chromium.org>
Cr-Commit-Position: refs/heads/master@{#544785}
[modify] https://crrev.com/958358e66f1dfca7b8f1b71bce9485a5d4058739/skia/ext/skia_trace_memory_dump_impl.cc
[modify] https://crrev.com/958358e66f1dfca7b8f1b71bce9485a5d4058739/skia/ext/skia_trace_memory_dump_impl.h

Project Member

Comment 14 by bugdroid1@chromium.org, Mar 26 2018

The following revision refers to this bug:
  https://chromium.googlesource.com/chromium/src.git/+/63bccd57a9798839dafb874554578ce1b6f3b42d

commit 63bccd57a9798839dafb874554578ce1b6f3b42d
Author: Eric Karl <ericrk@chromium.org>
Date: Mon Mar 26 16:50:40 2018

Tell Skia memory dump not to dump wrapped/external GPU objects

Skia currently dumps wrapped objects, and Chrome tries to alias these
objects with Chrome's own dumps. This isn't always possible, so instead
this change uses a new Skia API to ask Skia to skip dumping these.

Bug:  795358 
Change-Id: Iee3dbbb09037cb84fa8a33f0512cb2a54ef42f32
Reviewed-on: https://chromium-review.googlesource.com/979773
Reviewed-by: Sadrul Chowdhury <sadrul@chromium.org>
Commit-Queue: Eric Karl <ericrk@chromium.org>
Cr-Commit-Position: refs/heads/master@{#545820}
[modify] https://crrev.com/63bccd57a9798839dafb874554578ce1b6f3b42d/services/ui/public/cpp/gpu/context_provider_command_buffer.cc

Status: Fixed (was: Started)
Project Member

Comment 16 by bugdroid1@chromium.org, Mar 29 2018

Labels: merge-merged-next-release
The following revision refers to this bug:
  https://skia.googlesource.com/skia/+/6ab4e8f2e499169071bd4168082e8b6cdcb4a107

commit 6ab4e8f2e499169071bd4168082e8b6cdcb4a107
Author: Eric Karl <ericrk@chromium.org>
Date: Thu Mar 29 19:47:39 2018

Allow SkTraceMemoryDump to exclude wrapped objects

Allow SkTraceMemoryDump to exclude wrapped objects from dumps. This
helps avoid duplicate dumping when Skia is wrapping an external object
which is already dumped externally.

Bug:  795358 
Change-Id: Icbda96b564c81b958d40f74693280ac7d5ba7332
Reviewed-on: https://skia-review.googlesource.com/114681
Reviewed-by: Brian Salomon <bsalomon@google.com>
Commit-Queue: Eric Karl <ericrk@chromium.org>
(cherry picked from commit af7700265b74123d8ad3de6dde0c21545453140b)
Reviewed-on: https://skia-review.googlesource.com/117020
Reviewed-by: Robert Phillips <robertphillips@google.com>

[modify] https://crrev.com/6ab4e8f2e499169071bd4168082e8b6cdcb4a107/src/gpu/gl/GrGLTextureRenderTarget.cpp
[modify] https://crrev.com/6ab4e8f2e499169071bd4168082e8b6cdcb4a107/include/core/SkTraceMemoryDump.h
[modify] https://crrev.com/6ab4e8f2e499169071bd4168082e8b6cdcb4a107/src/gpu/gl/GrGLTexture.cpp
[modify] https://crrev.com/6ab4e8f2e499169071bd4168082e8b6cdcb4a107/src/gpu/gl/GrGLRenderTarget.cpp
[modify] https://crrev.com/6ab4e8f2e499169071bd4168082e8b6cdcb4a107/src/gpu/gl/GrGLTexture.h
[modify] https://crrev.com/6ab4e8f2e499169071bd4168082e8b6cdcb4a107/src/gpu/GrGpuResource.cpp
[modify] https://crrev.com/6ab4e8f2e499169071bd4168082e8b6cdcb4a107/tests/TraceMemoryDumpTest.cpp

Issue 787179 has been merged into this issue.

Sign in to add a comment