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

Issue 619978 link

Starred by 7 users

Issue metadata

Status: WontFix
Owner:
Closed: Aug 2017
Cc:
Components:
EstimatedDays: ----
NextAction: ----
OS: Windows
Pri: 3
Type: Bug-Regression



Sign in to add a comment

LayoutTest crash : fast/css/giant-stylesheet-crash.html

Project Member Reported by dmazz...@chromium.org, Jun 14 2016

Issue description

Failing on Win7, Win10, Win7 (dbg)
fast/css/giant-stylesheet-crash.html

Suspecting V8 roll.

fast/css/giant-stylesheet-crash-crash-log.txt
crash log for renderer (pid <unknown>):
STDOUT: #CRASHED - renderer
STDERR: Backtrace:
STDERR:         WTF::Partitions::initialize [0x06C5E318+344]
STDERR:         WTF::Partitions::handleOutOfMemory [0x06C5E0D1+49]
STDERR:         WTF::partitionFreeSlowPath [0x06C5C540+1056]
STDERR:         WTF::partitionAllocSlowPath [0x06C5A5BA+1050]
STDERR:         WTF::partitionBucketAlloc [0x06C49E6C+252]
STDERR:         WTF::partitionAllocGenericFlags [0x06C49D18+136]
STDERR:         WTF::partitionAllocGeneric [0x06C49C76+22]
STDERR:         WTF::Partitions::bufferMalloc [0x06C49426+22]
STDERR:         WTF::StringImpl::createUninitialized [0x06C8011F+63]
STDERR:         WTF::String::createUninitialized [0x06C50824+20]
STDERR:         (No symbol) [0x1454F8DF]
STDERR:         (No symbol) [0x1454FF35]
STDERR:         (No symbol) [0x14455B17]
STDERR:         (No symbol) [0x14457943]
STDERR:         (No symbol) [0x14701937]
STDERR:         (No symbol) [0x14701A3C]
STDERR:         std::`dynamic initializer for 'nothrow'' [0x088BFF2A+4555418]
STDERR:         std::_Atomic_int::store [0x06C4AE98+24]
STDERR:         WTF::SpinLock::unlock [0x06C4AF23+19]
STDERR:         WTF::partitionFreeGeneric [0x06C4A348+200]


Revision range is 399411 ... 399413

commit ee5a916c898a0831596336e0c7e3eff87dcba8c9
Author: Xianzhu Wang <wangxianzhu@chromium.org>
Date:   Mon Jun 13 00:19:31 2016 -0700

    Force FrameView::checkDoesNotNeedLayout() not inlined in FrameView.cpp
    
    Some new crash reports just report the line calling
    checkDoesNotNeedLayout() but not which line in the function
    crashes.
    
    BUG=590856
    TBR=chrishtr@chromium.org
    
    Review URL: https://codereview.chromium.org/2062693002 .
    
    Cr-Commit-Position: refs/heads/master@{#399413}

commit 9e6a14d7d4c03bc038c775c1aac4a6ffd8c9e2b4
Author: v8-autoroll <v8-autoroll@chromium.org>
Date:   Mon Jun 13 00:16:06 2016 -0700

    Update V8 to version 5.3.214.
    
    Summary of changes available at:
    https://chromium.googlesource.com/v8/v8/+log/245b66c2..580a3b99
    
    Please follow these instructions for assigning/CC'ing issues:
    https://github.com/v8/v8/wiki/Triaging%20issues
    
    Please close rolling in case of a roll revert:
    https://v8-roll.appspot.com/
    This only works with a Google account.
    
    CQ_INCLUDE_TRYBOTS=tryserver.blink:linux_blink_rel
    
    TBR=hablich@chromium.org,machenbach@chromium.org,yangguo@chromium.org,vogelheim@chromium.org
    
    Review-Url: https://codereview.chromium.org/2059083002
    Cr-Commit-Position: refs/heads/master@{#399412}

commit c364ba363241859f8fb92069293e30feee583e8b
Author: yabinh <yabinh@chromium.org>
Date:   Mon Jun 13 00:11:45 2016 -0700

    Add events tests for inputText in IME test
    
    BUG= 614937 
    
    Review-Url: https://codereview.chromium.org/2010803005
    Cr-Commit-Position: refs/heads/master@{#399411}




 
Project Member

Comment 1 by bugdroid1@chromium.org, Jun 14 2016

The following revision refers to this bug:
  https://chromium.googlesource.com/chromium/src.git/+/cd259d4d6fccd45d7febe940625f002b4fb45c3a

commit cd259d4d6fccd45d7febe940625f002b4fb45c3a
Author: Dominic Mazzoni <dmazzoni@chromium.org>
Date: Tue Jun 14 17:31:00 2016

Disable crashing test: fast/css/giant-stylesheet-crash.html

BUG= 619978 
TBR=hcarmona@chromium.org

Review URL: https://codereview.chromium.org/2061253002 .

Cr-Commit-Position: refs/heads/master@{#399736}

[modify] https://crrev.com/cd259d4d6fccd45d7febe940625f002b4fb45c3a/third_party/WebKit/LayoutTests/TestExpectations

Project Member

Comment 2 by bugdroid1@chromium.org, Jun 15 2016

The following revision refers to this bug:
  https://chromium.googlesource.com/chromium/src.git/+/cd259d4d6fccd45d7febe940625f002b4fb45c3a

commit cd259d4d6fccd45d7febe940625f002b4fb45c3a
Author: Dominic Mazzoni <dmazzoni@chromium.org>
Date: Tue Jun 14 17:31:00 2016

Disable crashing test: fast/css/giant-stylesheet-crash.html

BUG= 619978 
TBR=hcarmona@chromium.org

Review URL: https://codereview.chromium.org/2061253002 .

Cr-Commit-Position: refs/heads/master@{#399736}

[modify] https://crrev.com/cd259d4d6fccd45d7febe940625f002b4fb45c3a/third_party/WebKit/LayoutTests/TestExpectations

Labels: Needs-Bisect
Chrome-TE, please bisect down to the Chromium commit.
Components: -Blink>JavaScript Blink
Components: -Blink Blink>CSS

Comment 6 by ajha@chromium.org, Jun 17 2016

Cc: ajha@chromium.org
Labels: Needs-Feedback
Looks like the test has been disabled. If the bisect is still needed, please attach the test file as we(Chrome-TE) are unable to access the giant-stylesheet-crash.html file.

Comment 7 by ajha@chromium.org, Jul 25 2016

Labels: -Needs-Feedback -Needs-Bisect
Removing the Needs-Bisect label and requesting dmazzoni@ to close the issue if there is no further work to be done on this.
Labels: Needs-Bisect
The test is there to make sure that enormous inline style sheets don't crash. They seem to be crashing, so a bisect would be appreciated please. Test file attached.
giant-stylesheet-crash.html
788 bytes View Download

Comment 9 by ajha@chromium.org, Jul 26 2016

Labels: TE-NeedsfurtherTriage
I was unable to reproduce this on the official build without the Fix(53.0.2765.0(Official Build) (64-bit) master@{#399323}) with the attached 'giant-stylesheet-crash.html' file on Windows-7.

This is specific to Debug build and don't have set up to do bisect on the same. Labeling accordingly for help in bisecting this further.
  
Labels: -Needs-Bisect -TE-NeedsfurtherTriage TE-NeedsTriageHelp
Owner: rjwright@chromium.org
Assigning to me for debug bisect.
Status: Assigned (was: Untriaged)
Labels: -Build-Tests Test-Layout
Owner: ----
Status: Untriaged (was: Assigned)
#11 did not happen.
Labels: -Pri-1 Pri-2
Status: Available (was: Untriaged)
Cc: -hablich@chromium.org
Labels: Update-Weekly

Comment 18 by suzyh@chromium.org, Mar 27 2017

Labels: -TE-NeedsTriageHelp
Status: Unconfirmed (was: Available)
I don't think this ever got fully triaged, and it's now quite old. Bumping back to Unconfirmed. Can we get this reconfirmed and if appropriate bisected please?
Labels: OS-Windows
Status: Available (was: Unconfirmed)
The test is still disabled. Someone needs to create a patch to re-enable it.

Comment 20 by suzyh@chromium.org, Mar 28 2017

Owner: suzyh@chromium.org
Status: Assigned (was: Available)

Comment 21 by suzyh@chromium.org, Mar 28 2017

Created https://codereview.chromium.org/2781813002 to re-enable the test and verified that crash occurred on win_chromium_rel_ng. Next step is to bisect on a Windows machine.
I could repro the crash (OOM) on a 32 bit Canary on Windows 10 

59.0.3056.0 (Official Build) canary (32-bit)
Components: -Blink>CSS Blink>JavaScript
Owner: ----
Status: Untriaged (was: Assigned)
I ran a bisect on this just now, from the last time that test file was touched to today. This test has been reproducibly crashing since June 2016.

python tools\bisect-builds.py -a win -g 160683 --verify-range --use-local-cache -- --no-first-run file:///C:/src/chromium-fresh/src/third_party/WebKit/LayoutTests/fast/css/giant-stylesheet-crash.html

Note that you need to use the 32-bit Windows builds, as this does not repro on 64 bit builds.

And the results were exactly the same as in comment #1 (there are only 3 commits in it).

I strongly suspect the V8 roll in commit 399412. I don't know how to bisect within a V8 roll, so assigning to the V8 team.


Cc: mikelawther@chromium.org
Labels: Needs-Bisect
Hi test team - can you please do a per-commit bisect on this to confirm the suspicion that this is the V8 roll?

The test file is attached in comment #8. See my notes in comment #23 about reproducing it.

Comment 25 by ajha@chromium.org, Apr 3 2017

Labels: -Type-Bug -Pri-2 -Needs-Bisect M-59 hasbisect Pri-1 Type-Bug-Regression
Owner: peria@chromium.org
Status: Assigned (was: Untriaged)
Able to reproduce the issue on the latest Windows canary version: 59.0.3060.0(32 bit), works fine on 64 bit with the same chrome version.

Crash id: bf070f1640000000
Stack trace:
============
Thread 0 CRASHED [Out of Memory @ 0x7714a6f2 ] MAGIC SIGNATURE THREAD
Stack Quality92%Show frame trust levels
0x7714a6f2	(KERNELBASE.dll + 0x000da6f2 )	RaiseException
0x66575c83	(chrome_child.dll -partitions.cpp:119 )	WTF::partitionsOutOfMemoryUsing2G
0x66575b1e	(chrome_child.dll -partitions.cpp:177 )	WTF::Partitions::handleOutOfMemory()
0x6614c767	(chrome_child.dll -partition_alloc.cc:266 )	base::PartitionOutOfMemory
0x65a7599c	(chrome_child.dll -partition_alloc.cc:818 )	base::PartitionAllocSlowPath(base::PartitionRootBase *,int,unsigned int,base::PartitionBucket *)
0x65432b3a	(chrome_child.dll -v8stringresource.cpp:81 )	blink::StringTraits<WTF::String>::fromV8String<blink::V8StringOneByteTrait>(v8::Local<v8::String>,int)
0x65432902	(chrome_child.dll -v8stringresource.cpp:127 )	blink::v8StringToWebCoreString<WTF::String>(v8::Local<v8::String>,blink::ExternalMode)
0x65432881	(chrome_child.dll -v8stringresource.h:207 )	blink::V8StringResource<0>::operator WTF::String()
0x65432788	(chrome_child.dll -v8document.cpp:2756 )	blink::DocumentV8Internal::createTextNodeMethod
0x266060ca		
0x266063b3		
0x0fa0fc7d		
0x0b796657		
0x653014ba	(chrome_child.dll -execution.cc:145 )	v8::internal::`anonymous namespace'::Invoke
0x653026c2	(chrome_child.dll -execution.cc:191 )	v8::internal::Execution::Call(v8::internal::Isolate *,v8::internal::Handle<v8::internal::Object>,v8::internal::Handle<v8::internal::Object>,int,v8::internal::Handle<v8::internal::Object> * const)
0x65428228	(chrome_child.dll -api.cc:2023 )	v8::Script::Run(v8::Local<v8::Context>)
0x654280e6	(chrome_child.dll -v8scriptrunner.cpp:544 )	blink::V8ScriptRunner::runCompiledScript(v8::Isolate *,v8::Local<v8::Script>,blink::ExecutionContext *)
0x65427fe1	(chrome_child.dll -scriptcontroller.cpp:135 )	blink::ScriptController::executeScriptAndReturnValue(v8::Local<v8::Context>,blink::ScriptSourceCode const &,blink::AccessControlStatus)
0x65427e19	(chrome_child.dll -scriptcontroller.cpp:325 )	blink::ScriptController::evaluateScriptInMainWorld(blink::ScriptSourceCode const &,blink::AccessControlStatus,blink::ScriptController::ExecuteScriptPolicy)
0x65427d5e	(chrome_child.dll -scriptcontroller.cpp:296 )	blink::ScriptController::executeScriptInMainWorld(blink::ScriptSourceCode const &,blink::AccessControlStatus)
0x65426bd2	(chrome_child.dll -scriptloader.cpp:773 )	blink::ScriptLoader::doExecuteScript(blink::ScriptSourceCode const &)
0x65427c91	(chrome_child.dll -scriptloader.cpp:648 )	blink::ScriptLoader::executeScript(blink::ScriptSourceCode const &)
0x654267ce	(chrome_child.dll -scriptloader.cpp:500 )	blink::ScriptLoader::prepareScript(WTF::TextPosition const &,blink::ScriptLoader::LegacyTypeSupport)
0x654206d5	(chrome_child.dll -htmlparserscriptrunner.cpp:642 )	blink::HTMLParserScriptRunner::processScriptElementInternal(blink::Element *,WTF::TextPosition const &)
0x654205b3	(chrome_child.dll -htmlparserscriptrunner.cpp:406 )	blink::HTMLParserScriptRunner::processScriptElement(blink::Element *,WTF::TextPosition const &)
0x65420584	(chrome_child.dll -htmldocumentparser.cpp:291 )	blink::HTMLDocumentParser::runScriptsForPausedTreeBuilder()
0x65413a9e	(chrome_child.dll -htmldocumentparser.cpp:567 )	blink::HTMLDocumentParser::processTokenizedChunkFromBackgroundParser(std::unique_ptr<blink::HTMLDocumentParser::TokenizedChunk,std::default_delete<blink::HTMLDocumentParser::TokenizedChunk> >)
0x65421b6b	(chrome_child.dll -htmldocumentparser.cpp:625 )	blink::HTMLDocumentParser::pumpPendingSpeculations()
0x65234970	(chrome_child.dll -callback.h:80 )	?Run@?$Callback@$$A6AXXZ$00$00@base@@QGBEXXZ
0x6559894d	(chrome_child.dll -webtaskrunner.cpp:75 )	blink::TaskHandle::Runner::run(blink::TaskHandle const &)
0x65598921	(chrome_child.dll -bind_internal.h:339 )	base::internal::Invoker<base::internal::BindState<void ( blink::TaskHandle::Runner::*)(blink::TaskHandle const &),base::WeakPtr<blink::TaskHandle::Runner>,blink::TaskHandle>,void >::Run(base::internal::BindStateBase *)
0x652df196	(chrome_child.dll -task_annotator.cc:59 )	base::debug::TaskAnnotator::RunTask(char const *,base::PendingTask *)
0x65256145	(chrome_child.dll -task_queue_manager.cc:539 )	blink::scheduler::TaskQueueManager::ProcessTaskFromWorkQueue(blink::scheduler::internal::WorkQueue *,bool,blink::scheduler::LazyNow,base::TimeTicks *)
0x65257297	(chrome_child.dll -task_queue_manager.cc:337 )	blink::scheduler::TaskQueueManager::DoWork(bool)
0x6576703c	(chrome_child.dll -bind_internal.h:339 )	base::internal::Invoker<base::internal::BindState<void ( blink::scheduler::TaskQueueManager::*)(bool),base::WeakPtr<blink::scheduler::TaskQueueManager>,bool>,void >::Run(base::internal::BindStateBase *)
0x652df196	(chrome_child.dll -task_annotator.cc:59 )	base::debug::TaskAnnotator::RunTask(char const *,base::PendingTask *)
0x65255e0d	(chrome_child.dll -message_loop.cc:423 )	base::MessageLoop::RunTask(base::PendingTask *)
0x652dedc3	(chrome_child.dll -message_loop.cc:527 )	base::MessageLoop::DoWork()
0x652de985	(chrome_child.dll -message_pump_default.cc:33 )	base::MessagePumpDefault::Run(base::MessagePump::Delegate *)
0x6536e825	(chrome_child.dll -run_loop.cc:37 )	base::RunLoop::Run()
0x6579165b	(chrome_child.dll -renderer_main.cc:200 )	content::RendererMain(content::MainFunctionParams const &)
0x65369506	(chrome_child.dll -content_main_runner.cc:437 )	content::RunNamedProcessTypeMain(std::basic_string<char,std::char_traits<char>,std::allocator<char> > const &,content::MainFunctionParams const &,content::ContentMainDelegate *)
0x65369488	(chrome_child.dll -content_main_runner.cc:729 )	content::ContentMainRunnerImpl::Run()
0x6536879a	(chrome_child.dll -main.cc:179 )	service_manager::Main(service_manager::MainParams const &)
0x6536843d	(chrome_child.dll -content_main.cc:19 )	content::ContentMain(content::ContentMainParams const &)
0x65368357	(chrome_child.dll -chrome_main.cc:123 )	ChromeMain
0x008d59a8	(chrome.exe -main_dll_loader_win.cc:202 )	MainDllLoader::Launch(HINSTANCE__ *,base::TimeTicks)
0x008d21ca	(chrome.exe -chrome_exe_main_win.cc:271 )	wWinMain
0x00943587	(chrome.exe -exe_common.inl:253 )	__scrt_common_main_seh
0x778262c3	(KERNEL32.DLL + 0x000162c3 )	BaseThreadInitThunk
0x77a90718	(ntdll.dll + 0x00060718 )	__RtlUserThreadStart
0x77a906e3	(ntdll.dll + 0x000606e3 )	_RtlUserThreadStart


Regressed in M-53.
==================
Last good build: 53.0.2766.0
First bad build: 53.0.2767.0

Changelog:
==========
https://chromium.googlesource.com/chromium/src/+log/8e411d16171d27612776a2f05356b0ed9f06b848..ee5a916c898a0831596336e0c7e3eff87dcba8c9

V8 changelog:
=============
https://chromium.googlesource.com/v8/v8/+log/245b66c2..580a3b99

Suspecting: https://codereview.chromium.org/2046933002 from the above V8 changelog.

peria@: Could this be related to the above change.

Note: Unable to provide per revision bisect as Win32 bit bisect is not supported in that range. 
Project Member

Comment 26 by sheriffbot@chromium.org, Apr 3 2017

Labels: FoundIn-M-59 Fracas
Users experienced this crash on the following builds:

Win Canary 59.0.3060.0 -  8.60 CPM, 121 reports, 97 clients (signature base::PartitionOutOfMemory)

If this update was incorrect, please add "Fracas-Wrong" label to prevent future updates.

- Go/Fracas
Labels: -Pri-1 Pri-2
Status: Started (was: Assigned)
I did not reproduce this issue yet, but I agree my change can trigger it.
Any updates?

Comment 29 by peria@chromium.org, Apr 14 2017

Labels: -Pri-2 -Fracas Fracas-Wrong Pri-3
I confirmed reverting the change http://crrev.com/2046933002 fixes the crash,
but the behavior seems to depend on parameters.

As far as I searched, we have no crash reports based on this issue (#26 is wrong),
and the change improved performance, so I think this issue is not so critical
and would like to hold it as-is for a while.
Issue 714366 has been merged into this issue.
Cc: brajkumar@chromium.org
Just to update the latest behavior, Still crashes observed on latest beta channel. Last crash is observed on #60.0.3080.5 with 1 instance. Currently this crash is ranked as number #4 for Windows platform under renderer process. Below information provides the comparison between previous and latest channels including total number of instances.
+--------------------------------------------------+      
|Latest Channel        |    Previous Channel       |
+--------------------------------------------------+  
|59.0.3071.36   55     |  59.0.3071.29	  1568	   |--> Beta
+--------------------------------------------------+
	
Link to the list of the builds getting crash:
---------------------------------------------
https://crash.corp.google.com/browse?q=product.name%3D%27Chrome%27%20%20AND%20custom_data.ChromeCrashProto.channel%3D%27beta%27%20AND%20custom_data.ChromeCrashProto.ptype%3D%27renderer%27%20AND%20custom_data.ChromeCrashProto.magic_signature_1.name%3D%27base%3A%3APartitionOutOfMemory%27&ignore_case=false&enable_rewrite=true&omit_field_name=&omit_field_value=&omit_field_opt=%3D

peria@ Since the crash instances are high, could you please let us know is there any latest update available on this issue.

Thanks!
In cases that top of the stack is "base::PartitionOutOfMemory", it means OOM, and it doesn't mean this issue is the root cause.

This issue cares only the case OOMs happen through "	blink::V8StringResource::operator WTF::AtomicString()", and as far as I see some reports, I couldn't find such example.

Comment 33 by peria@chromium.org, May 10 2017

Labels: -M-59 -Update-Weekly
Status: Assigned (was: Started)
Summary: LayoutTest crash : fast/css/giant-stylesheet-crash.html (was: LayoutTest crash (suspecting v8 roll): fast/css/giant-stylesheet-crash.html)
Cc: hablich@chromium.org ahaas@chromium.org
Labels: ReleaseBlock-Stable M-59
Cc: neis@chromium.org
Cc: ranjitkan@chromium.org
@ neis: Request you to please take a look into it and help us triage it. Issue is tagged with M59 and has a stable blocker with it.

Thanks.!

Comment 37 by peria@chromium.org, May 23 2017

Labels: -ReleaseBlock-Stable -FoundIn-M-59
I can't agree to set ReleaseBlock-Stable on this issue, because we already released Stable Chrome for >1 year, and I can't find suitable crash reports.


Please add ReleaseBlock label again if you feel this is a release blocker.

Comment 38 by neis@chromium.org, May 24 2017

I agree that this is not a release block. I'm not aware of any problems caused by the mentioned CL other than that one extreme test sometimes running out of memory. From what I can tell, none of the chrome-crashes reported here have anything to do with it.
Project Member

Comment 39 by sheriffbot@chromium.org, May 29 2017

Labels: ReleaseBlock-Dev
This crash has high impact on Chrome's stability.
Signature: base::PartitionOutOfMemory.
Channel: dev. Platform: win.
Labeling  issue 619978  with ReleaseBlock-Dev.


If this update was incorrect, please add "Fracas-Wrong" label to prevent future updates.

- Go/Fracas

Comment 40 by peria@chromium.org, May 29 2017

Labels: -ReleaseBlock-Dev
Let me repeat saying this issue is not the reason.
Status: WontFix (was: Assigned)

Sign in to add a comment