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

Issue 874090 link

Starred by 1 user

Issue metadata

Status: Fixed
Owner:
Last visit > 30 days ago
Closed: Sep 11
Cc:
Components:
EstimatedDays: ----
NextAction: ----
OS: Chrome
Pri: 2
Type: Bug

Blocked on:
issue 880584
issue 880595



Sign in to add a comment

Update chromeos related tests on waterfall for single-process-mash

Project Member Reported by sky@chromium.org, Aug 14

Issue description

We need to do the following:

Put mash_browsertests on FYI bot with current filter file.

Have mash_browsertests on main waterfall run a single test. This test should be something that exercises login. This is just to make sure multi-process still works.

Put single_process_mash_browsertests on main waterfall with filter file.

Add single_process_mash_content_browsertests on main waterfall (874089).

Add single_process_mash_content_unittests on main waterfall (866942).

Ryan, I'm passing this your way. Mike and myself can help if you've never done this sort of thing. The README.md in https://chromium.googlesource.com/chromium/src/+/master/testing/buildbot/ has lots of details.

Xiyuan, could you suggest a login test to run for the --mash case?
 
Add single_process_ash_unittests to this as well.
James added a sanity LoginUtilsTest.MashLogin as part of browser_tests.
  https://chromium-review.googlesource.com/c/chromium/src/+/624191

It exercises the login using stubs and should catch failures caused by browser code.


Cc: jamescook@chromium.org
Seeing a number of failures when running browser_tests locally with --enable-features=SingleProcessMash.  Attaching a log of the test results.  Command-line used:

python testing/xvfb.py out/Release/browser_tests --brave-new-test-launcher --test-launcher-bot-mode --cfi-diag=0 --enable-features=SingleProcessMash --override-use-software-gl-for-tests --test-launcher-filter-file=testing/buildbot/filters/mash.browser_tests.filter

Two common failures seen are:

[33535:33535:0817/120937.932464:FATAL:window.cc(375)] Check failed: env_ == child->env_ (0x20315af349a0 vs. 0x20315ad8e6e0)All windows in a hierarchy must share the  same Env.
#0 0x000004c4ae7c base::debug::StackTrace::StackTrace()
#1 0x000004bb3f3b logging::LogMessage::~LogMessage()
#2 0x000006c5ea02 aura::Window::AddChild()
#3 0x0000058bb568 views::internal::NativeWidgetPrivate::ReparentNativeView()
#4 0x0000058b44ef views::NativeViewHostAura::AttachNativeView()
#5 0x0000058745bd views::NativeViewHost::Attach()
#6 0x0000081b5c59 views::WebView::AttachWebContents()
#7 0x00000589ac98 views::View::PropagateAddNotifications()
#8 0x00000589ac4d views::View::PropagateAddNotifications()
#9 0x00000589a015 views::View::AddChildViewAt()
#10 0x0000058a4f5a views::internal::RootView::SetContentsView()
#11 0x0000058a89a4 views::Widget::SetContentsView()
#12 0x0000020a15a7 chromeos::LoginDisplayHostWebUI::InitLoginWindowAndView()
#13 0x00000209f985 chromeos::LoginDisplayHostWebUI::StartWizard()
#14 0x0000020a3039 (anonymous namespace)::ShowLoginWizardFinish()
#15 0x0000020a2688 chromeos::ShowLoginWizard()
#16 0x00000207f985 chromeos::ChromeSessionManager::Initialize()
#17 0x000001f427f6 chromeos::ChromeBrowserMainPartsChromeos::PostProfileInit()
#18 0x000004d27b6b ChromeBrowserMainParts::PreMainMessageLoopRunImpl()
#19 0x000004d26f1a ChromeBrowserMainParts::PreMainMessageLoopRun()
#20 0x000001f41978 chromeos::ChromeBrowserMainPartsChromeos::PreMainMessageLoopRun()
#21 0x000003231d11 content::BrowserMainLoop::PreMainMessageLoopRun()
#22 0x0000036c78b5 content::StartupTaskRunner::RunAllTasksNow()
#23 0x000003230438 content::BrowserMainLoop::CreateStartupTasks()
#24 0x000003234573 content::BrowserMainRunnerImpl::Initialize()
#25 0x00000322dfb2 content::BrowserMain()
#26 0x000004b1074d content::ContentMainRunnerImpl::Run()
#27 0x000006ead415 service_manager::Main()
#28 0x000004b0ebf4 content::ContentMain()
#29 0x00000524901c content::BrowserTestBase::SetUp()
#30 0x000004cc87ab InProcessBrowserTest::SetUp()
#31 0x000001a1d7ac policy::DevicePolicyCrosBrowserTest::SetUp()
#32 0x00000240c20d testing::Test::Run()
#33 0x00000240ce30 testing::TestInfo::Run()
#34 0x00000240d347 testing::TestCase::Run()
#35 0x0000024187f7 testing::internal::UnitTestImpl::RunAllTests()
#36 0x00000241836d testing::UnitTest::Run()
#37 0x000004cdff81 base::TestSuite::Run()
#38 0x000004b96835 ChromeTestSuiteRunner::RunTestSuite()
#39 0x00000526ccdc content::LaunchTests()
#40 0x000004b96cf3 LaunchChromeTests()
#41 0x000004b967ce main
#42 0x7f9ca45412b1 __libc_start_main
#43 0x00000090e7ea _start

and

[28883:28883:0817/120736.336165:ERROR:layer_tree_host_impl.cc(3054)] Forcing zero-copy tile initialization as worker context is missing
../../chrome/browser/ui/test/test_browser_ui.cc:32: Failure
Value of: VerifyUi()
  Actual: false
Expected: true
Stack trace:
#0 0x000002406bac testing::internal::UnitTestImpl::CurrentOsStackTraceExceptTop()
#1 0x000002406589 testing::internal::AssertHelper::operator=()
#2 0x00000113be2e TestBrowserUi::ShowAndVerifyUi()
#3 0x000005249452 content::BrowserTestBase::ProxyRunTestOnMainThreadLoop()
#4 0x000004d2812b ChromeBrowserMainParts::PreMainMessageLoopRunImpl()
#5 0x000004d26f1a ChromeBrowserMainParts::PreMainMessageLoopRun()
#6 0x000001f41978 chromeos::ChromeBrowserMainPartsChromeos::PreMainMessageLoopRun()
#7 0x000003231d11 content::BrowserMainLoop::PreMainMessageLoopRun()
#8 0x0000036c78b5 content::StartupTaskRunner::RunAllTasksNow()
#9 0x000003230438 content::BrowserMainLoop::CreateStartupTasks()
#10 0x000003234573 content::BrowserMainRunnerImpl::Initialize()
#11 0x00000322dfb2 content::BrowserMain()
#12 0x000004b1074d content::ContentMainRunnerImpl::Run()
#13 0x000006ead415 service_manager::Main()
#14 0x000004b0ebf4 content::ContentMain()
#15 0x00000524901c content::BrowserTestBase::SetUp()
#16 0x000004cc87ab InProcessBrowserTest::SetUp()

Seeing similar results when running the tests on trybots.  Logs at: https://ci.chromium.org/p/chromium/builders/luci.chromium.try/linux-chromeos-rel/71650.  Search for 'single_process_mash_browser_tests'.

SingleProcessMash_browser_tests_failures
10.0 MB View Download
Xiyuan's patch @ https://chromium-review.googlesource.com/c/chromium/src/+/1180414 should tests failing with the first stack. Hopefully it also addresses the second stack too.
Not sure https://chromium-review.googlesource.com/c/chromium/src/+/1180414 would fix the issue. AuraTestSuiteSetup seems only used by a bunch of unittests.

The first stack looks like complains the login widget [1] and RWHVAura [2] has different Env. The problem might be related to creating a widget without specifying "parent" or "context". The login widget should be put under LockScreenContainer but this is done by specifying a mus property when using window service.

But without a "parent"/"context", it should use the primary Env. This should be the same Env used by RWHVAura. Not sure why the two does not match.

[1] https://cs.chromium.org/chromium/src/chrome/browser/chromeos/login/ui/login_display_host_webui.cc?rcl=458df2b38c17234780a8ea78d43eb85c5cfe7772&l=1047

[2] https://cs.chromium.org/chromium/src/content/browser/renderer_host/render_widget_host_view_aura.cc?rcl=458df2b38c17234780a8ea78d43eb85c5cfe7772&l=1909
The first crash seems no longer happening if you sync to ToT.
Confirmed not seeing the env dcheck failures on TOT.  Still seeing the VerifyUi() failure however, and also with jamescook@'s cl at https://chromium-review.googlesource.com/c/chromium/src/+/1184115 as well.
(Also seeing a bunch of failures when running ash_unittests with --enable-features=SingleProcessMash)
Here's a SingleProcessMash filter file for browser_tests. We should try it one or twice on the bots to see if there are flaky tests I didn't catch. I ran it after applying all my //chrome/browser IsAshInBrowserProcess() fixes.

The good news is that it only blacklists 3% of the tests!

single_process_mash.browser_tests.filter
22.0 KB Download
Thanks James, https://chromium-review.googlesource.com/c/chromium/src/+/1175222 adds a basic suite of single_process_mash_browser_tests.  I'll experiment with the filter file and send a new CL to add the tests to the FYI bot, and another one to add ash unit tests for S-P-Mash.
Project Member

Comment 11 by bugdroid1@chromium.org, Aug 24

The following revision refers to this bug:
  https://chromium.googlesource.com/chromium/src.git/+/42ad79bbc3c56ab601508070eec07276fea65f6a

commit 42ad79bbc3c56ab601508070eec07276fea65f6a
Author: Ryan Cui <rcui@chromium.org>
Date: Fri Aug 24 00:01:52 2018

Add tests for single process mash

Moves plain Mash tests to FYI waterfall, while retaining a login test on the
main waterfall as a sanity check.

Bug:  874090 
Change-Id: I86c60344610c2a4f99b35f9116c5366771a16c64
Reviewed-on: https://chromium-review.googlesource.com/1175222
Commit-Queue: Ryan Cui <rcui@chromium.org>
Reviewed-by: Scott Violet <sky@chromium.org>
Reviewed-by: James Cook <jamescook@chromium.org>
Cr-Commit-Position: refs/heads/master@{#585665}
[modify] https://crrev.com/42ad79bbc3c56ab601508070eec07276fea65f6a/testing/buildbot/chromium.chromiumos.json
[modify] https://crrev.com/42ad79bbc3c56ab601508070eec07276fea65f6a/testing/buildbot/chromium.fyi.json
[modify] https://crrev.com/42ad79bbc3c56ab601508070eec07276fea65f6a/testing/buildbot/chromium.memory.json
[modify] https://crrev.com/42ad79bbc3c56ab601508070eec07276fea65f6a/testing/buildbot/filters/BUILD.gn
[add] https://crrev.com/42ad79bbc3c56ab601508070eec07276fea65f6a/testing/buildbot/filters/chromeos.mash.browser_tests.filter
[rename] https://crrev.com/42ad79bbc3c56ab601508070eec07276fea65f6a/testing/buildbot/filters/chromeos.mash.fyi.browser_tests.filter
[add] https://crrev.com/42ad79bbc3c56ab601508070eec07276fea65f6a/testing/buildbot/filters/chromeos.single_process_mash.browser_tests.filter
[modify] https://crrev.com/42ad79bbc3c56ab601508070eec07276fea65f6a/testing/buildbot/test_suites.pyl

Project Member

Comment 12 by bugdroid1@chromium.org, Aug 24

The following revision refers to this bug:
  https://chromium.googlesource.com/chromium/src//+/42ad79bbc3c56ab601508070eec07276fea65f6a

commit 42ad79bbc3c56ab601508070eec07276fea65f6a
Author: Ryan Cui <rcui@chromium.org>
Date: Fri Aug 24 00:01:52 2018

Add tests for single process mash

Moves plain Mash tests to FYI waterfall, while retaining a login test on the
main waterfall as a sanity check.

Bug:  874090 
Change-Id: I86c60344610c2a4f99b35f9116c5366771a16c64
Reviewed-on: https://chromium-review.googlesource.com/1175222
Commit-Queue: Ryan Cui <rcui@chromium.org>
Reviewed-by: Scott Violet <sky@chromium.org>
Reviewed-by: James Cook <jamescook@chromium.org>
Cr-Commit-Position: refs/heads/master@{#585665}
[rename] https://crrev.com/42ad79bbc3c56ab601508070eec07276fea65f6a/testing/buildbot/filters/chromeos.mash.fyi.browser_tests.filter
[modify] https://crrev.com/42ad79bbc3c56ab601508070eec07276fea65f6a/testing/buildbot/filters/BUILD.gn
[modify] https://crrev.com/42ad79bbc3c56ab601508070eec07276fea65f6a/testing/buildbot/chromium.fyi.json
[modify] https://crrev.com/42ad79bbc3c56ab601508070eec07276fea65f6a/testing/buildbot/chromium.memory.json
[add] https://crrev.com/42ad79bbc3c56ab601508070eec07276fea65f6a/testing/buildbot/filters/chromeos.mash.browser_tests.filter
[modify] https://crrev.com/42ad79bbc3c56ab601508070eec07276fea65f6a/testing/buildbot/test_suites.pyl
[modify] https://crrev.com/42ad79bbc3c56ab601508070eec07276fea65f6a/testing/buildbot/chromium.chromiumos.json
[add] https://crrev.com/42ad79bbc3c56ab601508070eec07276fea65f6a/testing/buildbot/filters/chromeos.single_process_mash.browser_tests.filter

Project Member

Comment 13 by bugdroid1@chromium.org, Aug 24

The following revision refers to this bug:
  https://chromium.googlesource.com/chromium/src.git/+/a6c0ed2abc77387ccf61e3350d0c4549701d00d8

commit a6c0ed2abc77387ccf61e3350d0c4549701d00d8
Author: Hayato Ito <hayato@chromium.org>
Date: Fri Aug 24 10:09:12 2018

Revert "Add tests for single process mash"

This reverts commit 42ad79bbc3c56ab601508070eec07276fea65f6a.

Reason for revert: [chromium sheriff] A culprit of test failures:
https://ci.chromium.org/p/chromium/builders/luci.chromium.ci/Linux%20Chromium%20OS%20ASan%20LSan%20Tests%20%281%29/28724

single_process_mash_browser_tests Run on OS: 'Ubuntu-14.04'
Max shard duration: 0:03:53.214620 (shard #3)
Min shard duration: 0:01:18.857780 (shard #0)
failures:
LoginPromptBrowserTest.LoginPromptForXHRWithoutCredentialsCancelled
LoginPromptBrowserTest.TestTwoAuths
LoginPromptBrowserTest.ShowCorrectUrlForCrossOriginMainFrameRedirects
LoginPromptBrowserTest.CancelRedundantAuths
LoginPromptBrowserTest.CancelLoginInterstitialOnRedirect
LoginPromptBrowserTest.ShouldReplaceExistingInterstitialWhenNavigated
LoginUtilsTest.MashLogin
LoginPromptBrowserTest.IncorrectConfirmation
LoginPromptBrowserTest.TestBasicAuth
LoginPromptBrowserTest.TestCancelAuth_Manual
LoginPromptBrowserTest.NoLoginPromptForXHRWithBadCredentials
LoginPromptBrowserTest.SupplyRedundantAuths
LoginPromptBrowserTest.SupplyRedundantAuthsMultiProfile
LoginPromptBrowserTest.LoginPromptForXHRWithoutCredentials
LoginPromptBrowserTest.TestDigestAuth
LoginPromptBrowserTest.AllowCrossdomainPromptForSubframes
LoginPromptBrowserTest.ShouldNotProceedExistingInterstitial
LoginPromptBrowserTest.TestCancelAuth_OnBack
LoginPromptBrowserTest.TestCancelAuth_OnForward
LoginPromptBrowserTest.BlockCrossdomainPromptForSubresources
LoginPromptBrowserTest.TestCancelAuth_OnNavigation
LoginPromptBrowserTest.ShowCorrectUrlForCrossOriginMainFrameRequests

Original change's description:
> Add tests for single process mash
> 
> Moves plain Mash tests to FYI waterfall, while retaining a login test on the
> main waterfall as a sanity check.
> 
> Bug:  874090 
> Change-Id: I86c60344610c2a4f99b35f9116c5366771a16c64
> Reviewed-on: https://chromium-review.googlesource.com/1175222
> Commit-Queue: Ryan Cui <rcui@chromium.org>
> Reviewed-by: Scott Violet <sky@chromium.org>
> Reviewed-by: James Cook <jamescook@chromium.org>
> Cr-Commit-Position: refs/heads/master@{#585665}

TBR=jamescook@chromium.org,sky@chromium.org,rcui@chromium.org

Change-Id: Ib9b9258b47e766d43ac7f3dc6b3c65aacf1230f4
No-Presubmit: true
No-Tree-Checks: true
No-Try: true
Bug:  874090 
Reviewed-on: https://chromium-review.googlesource.com/1188165
Reviewed-by: Hayato Ito <hayato@chromium.org>
Commit-Queue: Hayato Ito <hayato@chromium.org>
Cr-Commit-Position: refs/heads/master@{#585781}
[modify] https://crrev.com/a6c0ed2abc77387ccf61e3350d0c4549701d00d8/testing/buildbot/chromium.chromiumos.json
[modify] https://crrev.com/a6c0ed2abc77387ccf61e3350d0c4549701d00d8/testing/buildbot/chromium.fyi.json
[modify] https://crrev.com/a6c0ed2abc77387ccf61e3350d0c4549701d00d8/testing/buildbot/chromium.memory.json
[modify] https://crrev.com/a6c0ed2abc77387ccf61e3350d0c4549701d00d8/testing/buildbot/filters/BUILD.gn
[delete] https://crrev.com/eb2da37d1553a39e94cea0b84a289460e698c3bd/testing/buildbot/filters/chromeos.mash.browser_tests.filter
[delete] https://crrev.com/eb2da37d1553a39e94cea0b84a289460e698c3bd/testing/buildbot/filters/chromeos.single_process_mash.browser_tests.filter
[rename] https://crrev.com/a6c0ed2abc77387ccf61e3350d0c4549701d00d8/testing/buildbot/filters/mash.browser_tests.filter
[modify] https://crrev.com/a6c0ed2abc77387ccf61e3350d0c4549701d00d8/testing/buildbot/test_suites.pyl

Project Member

Comment 14 by bugdroid1@chromium.org, Aug 24

The following revision refers to this bug:
  https://chromium.googlesource.com/chromium/src.git/+/20126be26f6a30d7abead004e738b11bfce88213

commit 20126be26f6a30d7abead004e738b11bfce88213
Author: Scott Violet <sky@chromium.org>
Date: Fri Aug 24 19:59:52 2018

Reland "Add tests for single process mash"

This reverts commit a6c0ed2abc77387ccf61e3350d0c4549701d00d8.

Reason for revert: reverting as fix for leak landed here: https://chromium-review.googlesource.com/c/chromium/src/+/1188584

Original change's description:
> Revert "Add tests for single process mash"
> 
> This reverts commit 42ad79bbc3c56ab601508070eec07276fea65f6a.
> 
> Reason for revert: [chromium sheriff] A culprit of test failures:
> https://ci.chromium.org/p/chromium/builders/luci.chromium.ci/Linux%20Chromium%20OS%20ASan%20LSan%20Tests%20%281%29/28724
> 
> single_process_mash_browser_tests Run on OS: 'Ubuntu-14.04'
> Max shard duration: 0:03:53.214620 (shard #3)
> Min shard duration: 0:01:18.857780 (shard #0)
> failures:
> LoginPromptBrowserTest.LoginPromptForXHRWithoutCredentialsCancelled
> LoginPromptBrowserTest.TestTwoAuths
> LoginPromptBrowserTest.ShowCorrectUrlForCrossOriginMainFrameRedirects
> LoginPromptBrowserTest.CancelRedundantAuths
> LoginPromptBrowserTest.CancelLoginInterstitialOnRedirect
> LoginPromptBrowserTest.ShouldReplaceExistingInterstitialWhenNavigated
> LoginUtilsTest.MashLogin
> LoginPromptBrowserTest.IncorrectConfirmation
> LoginPromptBrowserTest.TestBasicAuth
> LoginPromptBrowserTest.TestCancelAuth_Manual
> LoginPromptBrowserTest.NoLoginPromptForXHRWithBadCredentials
> LoginPromptBrowserTest.SupplyRedundantAuths
> LoginPromptBrowserTest.SupplyRedundantAuthsMultiProfile
> LoginPromptBrowserTest.LoginPromptForXHRWithoutCredentials
> LoginPromptBrowserTest.TestDigestAuth
> LoginPromptBrowserTest.AllowCrossdomainPromptForSubframes
> LoginPromptBrowserTest.ShouldNotProceedExistingInterstitial
> LoginPromptBrowserTest.TestCancelAuth_OnBack
> LoginPromptBrowserTest.TestCancelAuth_OnForward
> LoginPromptBrowserTest.BlockCrossdomainPromptForSubresources
> LoginPromptBrowserTest.TestCancelAuth_OnNavigation
> LoginPromptBrowserTest.ShowCorrectUrlForCrossOriginMainFrameRequests
> 
> Original change's description:
> > Add tests for single process mash
> > 
> > Moves plain Mash tests to FYI waterfall, while retaining a login test on the
> > main waterfall as a sanity check.
> > 
> > Bug:  874090 
> > Change-Id: I86c60344610c2a4f99b35f9116c5366771a16c64
> > Reviewed-on: https://chromium-review.googlesource.com/1175222
> > Commit-Queue: Ryan Cui <rcui@chromium.org>
> > Reviewed-by: Scott Violet <sky@chromium.org>
> > Reviewed-by: James Cook <jamescook@chromium.org>
> > Cr-Commit-Position: refs/heads/master@{#585665}
> 
> TBR=jamescook@chromium.org,sky@chromium.org,rcui@chromium.org
> 
> Change-Id: Ib9b9258b47e766d43ac7f3dc6b3c65aacf1230f4
> No-Presubmit: true
> No-Tree-Checks: true
> No-Try: true
> Bug:  874090 
> Reviewed-on: https://chromium-review.googlesource.com/1188165
> Reviewed-by: Hayato Ito <hayato@chromium.org>
> Commit-Queue: Hayato Ito <hayato@chromium.org>
> Cr-Commit-Position: refs/heads/master@{#585781}

TBR=jamescook@chromium.org,sky@chromium.org,hayato@chromium.org,rcui@chromium.org

Change-Id: I90f8aff5115e528b95ba80095389f61d07974962
No-Presubmit: true
No-Tree-Checks: true
No-Try: true
Bug:  874090 
Reviewed-on: https://chromium-review.googlesource.com/1188779
Reviewed-by: Scott Violet <sky@chromium.org>
Commit-Queue: Scott Violet <sky@chromium.org>
Cr-Commit-Position: refs/heads/master@{#585958}
[modify] https://crrev.com/20126be26f6a30d7abead004e738b11bfce88213/testing/buildbot/chromium.chromiumos.json
[modify] https://crrev.com/20126be26f6a30d7abead004e738b11bfce88213/testing/buildbot/chromium.fyi.json
[modify] https://crrev.com/20126be26f6a30d7abead004e738b11bfce88213/testing/buildbot/chromium.memory.json
[modify] https://crrev.com/20126be26f6a30d7abead004e738b11bfce88213/testing/buildbot/filters/BUILD.gn
[add] https://crrev.com/20126be26f6a30d7abead004e738b11bfce88213/testing/buildbot/filters/chromeos.mash.browser_tests.filter
[rename] https://crrev.com/20126be26f6a30d7abead004e738b11bfce88213/testing/buildbot/filters/chromeos.mash.fyi.browser_tests.filter
[add] https://crrev.com/20126be26f6a30d7abead004e738b11bfce88213/testing/buildbot/filters/chromeos.single_process_mash.browser_tests.filter
[modify] https://crrev.com/20126be26f6a30d7abead004e738b11bfce88213/testing/buildbot/test_suites.pyl

Project Member

Comment 15 by bugdroid1@chromium.org, Aug 27

The following revision refers to this bug:
  https://chromium.googlesource.com/chromium/src.git/+/023022060eebc9faf74e4fdf088dab1a44349cd5

commit 023022060eebc9faf74e4fdf088dab1a44349cd5
Author: Ryan Cui <rcui@chromium.org>
Date: Mon Aug 27 18:21:15 2018

Add single-process-mash browser tests to FYI waterfall

The plan is to monitor the tests for flakiness before moving them to the
main chromium waterfall.

Bug:  874090 
Change-Id: I2c0bddd189c89b5485d0e14a90afbf0eb5ac0e44
Reviewed-on: https://chromium-review.googlesource.com/1185637
Commit-Queue: Ryan Cui <rcui@chromium.org>
Reviewed-by: Scott Violet <sky@chromium.org>
Reviewed-by: James Cook <jamescook@chromium.org>
Cr-Commit-Position: refs/heads/master@{#586320}
[modify] https://crrev.com/023022060eebc9faf74e4fdf088dab1a44349cd5/testing/buildbot/chromium.fyi.json
[modify] https://crrev.com/023022060eebc9faf74e4fdf088dab1a44349cd5/testing/buildbot/filters/BUILD.gn
[add] https://crrev.com/023022060eebc9faf74e4fdf088dab1a44349cd5/testing/buildbot/filters/chromeos.single_process_mash.fyi.browser_tests.filter
[modify] https://crrev.com/023022060eebc9faf74e4fdf088dab1a44349cd5/testing/buildbot/test_suites.pyl

Project Member

Comment 17 by bugdroid1@chromium.org, Aug 28

The following revision refers to this bug:
  https://chromium.googlesource.com/chromium/src.git/+/43644c944b757f791de5b12b7d7ced59a91d407f

commit 43644c944b757f791de5b12b7d7ced59a91d407f
Author: Ryan Cui <rcui@chromium.org>
Date: Tue Aug 28 17:33:51 2018

Update filter file for single-process-mash browser_tests

Updated filter file from jamescook@

Also adds PictureInPictureWindowControllerBrowserTest\
.OpenInFrameWithDevToolsDoesNotCrash to blacklist to fix failure on FYI bot.

Bug:  874090 
Change-Id: I737511e2cb0214a586326ac68889e71ced9e79cf
Reviewed-on: https://chromium-review.googlesource.com/1192334
Commit-Queue: Ryan Cui <rcui@chromium.org>
Reviewed-by: James Cook <jamescook@chromium.org>
Reviewed-by: Scott Violet <sky@chromium.org>
Cr-Commit-Position: refs/heads/master@{#586771}
[modify] https://crrev.com/43644c944b757f791de5b12b7d7ced59a91d407f/testing/buildbot/filters/chromeos.single_process_mash.fyi.browser_tests.filter

Project Member

Comment 18 by bugdroid1@chromium.org, Aug 30

The following revision refers to this bug:
  https://chromium.googlesource.com/chromium/src.git/+/a37b0bc2e31611597140b8e3256978b993bff6ec

commit a37b0bc2e31611597140b8e3256978b993bff6ec
Author: Scott Violet <sky@chromium.org>
Date: Thu Aug 30 17:46:44 2018

ash: make ArcNotificationManagerTest delete objects in TearDown()

To do otherwise means some state lasts longer than the objects created by
Ash. This causes problems with single-process-ash.

BUG= 874090 
TEST=test only changes

Change-Id: I359dbd5b3be90c1733bc9c94c6ea1c57f4e39513
Reviewed-on: https://chromium-review.googlesource.com/1196921
Reviewed-by: Xiyuan Xia <xiyuan@chromium.org>
Commit-Queue: Scott Violet <sky@chromium.org>
Cr-Commit-Position: refs/heads/master@{#587650}
[modify] https://crrev.com/a37b0bc2e31611597140b8e3256978b993bff6ec/ash/system/message_center/arc/arc_notification_view_unittest.cc
[modify] https://crrev.com/a37b0bc2e31611597140b8e3256978b993bff6ec/testing/buildbot/filters/chromeos.single_process_mash.fyi.ash_unittests.filter

Components: Internals>Services>Ash Tests
Blockedon: 880595 880584
rcui@ Just a heads-up: I'm currently working on issue 827532 and is going to add Network Service related tests to the Mojo ChromiumOS bot[1].

I noticed that the bot is currently failing on mash_fyi_browser_tests, single_process_mash_fyi_browser_tests, and sometimes single_process_mash_fyi_ash_unittests, and it would be really helpful if we could add those failures to the filter and keep the bot green.

Hopefully we could move the bot to CQ soon.

Thanks!

[1] https://ci.chromium.org/p/chromium/builders/luci.chromium.ci/Mojo%20ChromiumOS

The recent ash_unittest failures are fixed now ( crbug.com/880595 ).  The browser_test failures are being investigated ( crbug.com/880584 ).  In the meantime, I'll add them to the filter file to get the builder green again.  If all goes well I plan to move the majority of the tests over to the CQ by EOW.  Cheers, Ryan
Project Member

Comment 23 by bugdroid1@chromium.org, Sep 7

The following revision refers to this bug:
  https://chromium.googlesource.com/chromium/src.git/+/cdd6db93665914e9a55d2a69e1b0e0bdca3ca41a

commit cdd6db93665914e9a55d2a69e1b0e0bdca3ca41a
Author: Ryan Cui <rcui@chromium.org>
Date: Fri Sep 07 21:16:02 2018

Move SingleProcessMash tests to main waterfall

The tests have been running for a while now and look stable.  Move them over
to the main waterfall so they start being run by the CQ.

Bug:  874090 
Bug:  847585 
Change-Id: I8b013c568d0abbfb2697d5166dd6edda31bbbf88
Reviewed-on: https://chromium-review.googlesource.com/1212093
Commit-Queue: Ryan Cui <rcui@chromium.org>
Reviewed-by: Scott Violet <sky@chromium.org>
Cr-Commit-Position: refs/heads/master@{#589655}
[modify] https://crrev.com/cdd6db93665914e9a55d2a69e1b0e0bdca3ca41a/testing/buildbot/chromium.chromiumos.json
[modify] https://crrev.com/cdd6db93665914e9a55d2a69e1b0e0bdca3ca41a/testing/buildbot/chromium.fyi.json
[modify] https://crrev.com/cdd6db93665914e9a55d2a69e1b0e0bdca3ca41a/testing/buildbot/chromium.memory.json
[modify] https://crrev.com/cdd6db93665914e9a55d2a69e1b0e0bdca3ca41a/testing/buildbot/filters/BUILD.gn
[rename] https://crrev.com/cdd6db93665914e9a55d2a69e1b0e0bdca3ca41a/testing/buildbot/filters/chromeos.single_process_mash.ash_unittests.filter
[modify] https://crrev.com/cdd6db93665914e9a55d2a69e1b0e0bdca3ca41a/testing/buildbot/filters/chromeos.single_process_mash.browser_tests.filter
[rename] https://crrev.com/cdd6db93665914e9a55d2a69e1b0e0bdca3ca41a/testing/buildbot/filters/chromeos.single_process_mash.exo_unittests.filter
[delete] https://crrev.com/4372d199af1f0d555d7bcc2241e3fc8d46dd06c2/testing/buildbot/filters/chromeos.single_process_mash.fyi.browser_tests.filter
[modify] https://crrev.com/cdd6db93665914e9a55d2a69e1b0e0bdca3ca41a/testing/buildbot/test_suites.pyl

Project Member

Comment 24 by bugdroid1@chromium.org, Sep 10

The following revision refers to this bug:
  https://chromium.googlesource.com/chromium/src.git/+/39254f021e7b580cf2ae546fa1f8644e912b53f9

commit 39254f021e7b580cf2ae546fa1f8644e912b53f9
Author: Dominic Battre <battre@chromium.org>
Date: Mon Sep 10 09:08:05 2018

Double timeout of single_process_mash_browser_tests

It may make sense to increase the number of shards instead but I am not
sufficiently knowledgable of capacity planning to do this change.

The reason for increasing the timeout is that this bot times out consistently.

TBR=rcui@chromium.org,sky@chromium.org

Bug:  874090 
BUG:  847585 
Change-Id: I14f9d2eccbead7ef403e346deda5442a82d5ea27
Reviewed-on: https://chromium-review.googlesource.com/1215283
Reviewed-by: Dominic Battré <battre@chromium.org>
Reviewed-by: Yutaka Hirano <yhirano@chromium.org>
Commit-Queue: Dominic Battré <battre@chromium.org>
Cr-Commit-Position: refs/heads/master@{#589864}
[add] https://crrev.com/39254f021e7b580cf2ae546fa1f8644e912b53f9/internal_gyp
[modify] https://crrev.com/39254f021e7b580cf2ae546fa1f8644e912b53f9/testing/buildbot/chromium.chromiumos.json
[modify] https://crrev.com/39254f021e7b580cf2ae546fa1f8644e912b53f9/testing/buildbot/chromium.memory.json
[modify] https://crrev.com/39254f021e7b580cf2ae546fa1f8644e912b53f9/testing/buildbot/test_suites.pyl

Cc: battre@google.com
ExtensionWebRequestApiTest.WebRequestTypes are failing consistently on https://ci.chromium.org/p/chromium/builders/luci.chromium.ci/Linux%20Chromium%20OS%20ASan%20LSan%20Tests%20%281%29 since battre@ stopped them from outright timing out through r589864.

Example failure : https://logs.chromium.org/logs/chromium/buildbucket/cr-buildbucket.appspot.com/8935751851391670800/+/steps/single_process_mash_browser_tests/0/logs/ExtensionWebRequestApiTest.WebRequestTypes/0

I have no idea how to parse these logs :(
Cc: karandeepb@chromium.org
The CM API tests have a list of events that are expected and here it complains about an unexpected event. It is unexpected because the actual event does not have a valid frameId, parentFrameId and tabId. It is expected that the "onResponseStarted" event is associated with the tab in which the request was triggered.

The data is populated in WebRequestEventDetails::SetFrameData. +karandeepb worked last on extensions/browser/extension_api_frame_id_map.*.
Is there supposed to be something special about this test config? This seems to run ok on the other trybots - https://test-results.appspot.com/dashboards/flakiness_dashboard.html#testType=browser_tests&tests=ExtensionWebRequestApiTest.WebRequestTypes.

It looks like the failure is particular to the ASAN bot.  The test passes in single_process_mash_browser_tests on other bots - linux-chromeos-rel, linux-chromeos-dbg: https://test-results.appspot.com/dashboards/flakiness_dashboard.html#testType=single_process_mash_browser_tests&tests=ExtensionWebRequestApiTest.WebRequestTypes
Not sure on what the reason is. You can file a bug and add it to the filter file for now, I think.
Filed bug  crbug.com/882631 
Project Member

Comment 31 by bugdroid1@chromium.org, Sep 11

The following revision refers to this bug:
  https://chromium.googlesource.com/chromium/src.git/+/036dd350148a7c4be0f3c41f5d8ea860db6ec1a9

commit 036dd350148a7c4be0f3c41f5d8ea860db6ec1a9
Author: Ryan Cui <rcui@chromium.org>
Date: Tue Sep 11 17:18:47 2018

Filter file update for single_process_mash_browser_test

Add blacklist for CrSettingsResetPageTest.All (flaky) and
ExtensionWebRequestApiTest.WebRequestTypes (failing on ASAN bot).

Bug:  874090 
Bug:  882631 
Bug:  880584 
Change-Id: Ie2a5bee8de4497a470a064ea853805a0e74b685a
Reviewed-on: https://chromium-review.googlesource.com/1217710
Reviewed-by: Ben Pastene <bpastene@chromium.org>
Commit-Queue: Ryan Cui <rcui@chromium.org>
Cr-Commit-Position: refs/heads/master@{#590366}
[modify] https://crrev.com/036dd350148a7c4be0f3c41f5d8ea860db6ec1a9/testing/buildbot/filters/chromeos.single_process_mash.browser_tests.filter

Project Member

Comment 32 by bugdroid1@chromium.org, Sep 11

The following revision refers to this bug:
  https://chromium.googlesource.com/chromium/src.git/+/cd3c168d3ea3c503db461c458abc4ccb376437d9

commit cd3c168d3ea3c503db461c458abc4ccb376437d9
Author: Ryan Cui <rcui@chromium.org>
Date: Tue Sep 11 19:32:23 2018

Increase shard count for single_process_mash_browser_tests

Increase shard count used for the debug and msan/asan builders to prevent
the tests from timing out.  Repurposes shards previously used for
mash_browser_tests, as crrev.com/c/1175222 reduced the number of tests running
as part of that suite to just 1.

Also sets the timeout for tests back to 1800.  It was bumped up to 3600
by crrev.com/c/1217124.

Bug:  874090 
Change-Id: Ia2b0ea4d3264e240dadde04f8ff049ff9a8b5735
Reviewed-on: https://chromium-review.googlesource.com/1217124
Reviewed-by: James Cook <jamescook@chromium.org>
Reviewed-by: Dirk Pranke <dpranke@chromium.org>
Commit-Queue: Ryan Cui <rcui@chromium.org>
Cr-Commit-Position: refs/heads/master@{#590437}
[modify] https://crrev.com/cd3c168d3ea3c503db461c458abc4ccb376437d9/testing/buildbot/chromium.chromiumos.json
[modify] https://crrev.com/cd3c168d3ea3c503db461c458abc4ccb376437d9/testing/buildbot/chromium.memory.json
[modify] https://crrev.com/cd3c168d3ea3c503db461c458abc4ccb376437d9/testing/buildbot/test_suite_exceptions.pyl
[modify] https://crrev.com/cd3c168d3ea3c503db461c458abc4ccb376437d9/testing/buildbot/test_suites.pyl

Cc: -karandeepb@chromium.org
Status: Fixed (was: Assigned)
With the last two patches, the Debug, ASAN and MSAN tests are cycling green now (or failing for other reasons).  Marking this bug as fixed.  Adding single_process_mash_content_browsertests on main waterfall is tracked in  crbug.com/874089 .

Sign in to add a comment