New issue
Advanced search Search tips

Issue 691607 link

Starred by 9 users

Issue metadata

Status: Fixed
Owner:
Closed: Feb 2017
Cc:
Components:
EstimatedDays: ----
NextAction: ----
OS: Windows
Pri: 2
Type: Bug



Sign in to add a comment

IndexedDB Memory Leak

Reported by joe.ju...@gmail.com, Feb 13 2017

Issue description

UserAgent: Mozilla/5.0 (Windows NT 10.0; WOW64; rv:51.0) Gecko/20100101 Firefox/51.0

Steps to reproduce the problem:
1. Open attached file
2. Let it run for a few minutes
3. Aww, snap

What is the expected behavior?
Page should run indefinitely. It is essentially a stress test of IndexedDB. It creates a large string and repeatedly stores and retrieves the string from an IndexedDB store.

What went wrong?
I suspect BlinkGC is involved somehow. It seems like garbage collection isn't kicking in at the right moment. This test fails on Windows 10 Chrome 55 32/64 bit, Windows 10 Chrome 56 32/64 bit, and Windows 7 Chrome 56 32 bit. It ran for several hours successfully without failing on a fresh Windows 7 Pro VM running Chrome 56 32 bit. So memory fragmentation may be involved or this bug may not affect fresh installs of Chrome. 

Firefox can run the test indefinitely. Safari on macOS can run it for a long time, but the memory usage does slowly climb so it looks like it has a leak as well.

This may be related to https://bugs.chromium.org/p/chromium/issues/detail?id=533648 or https://bugs.chromium.org/p/chromium/issues/detail?id=612525.

Did this work before? N/A 

Does this work in other browsers? N/A

Chrome version: 56.0.2924.87 (Official Build) (32-bit)  Channel: stable
OS Version: 10.0
Flash Version: 

Profiling JS heap memory usage with the Developer Tools shows that there is no increase in memory usage over the course of the test

 

Comment 1 by jsb...@chromium.org, Feb 13 2017

Cc: pwnall@chromium.org cmumford@chromium.org dmu...@chromium.org
Components: -Blink>Storage Blink>Storage>IndexedDB

Comment 2 by dmu...@chromium.org, Feb 13 2017

Can you please attach the file?

Comment 3 by joe.ju...@gmail.com, Feb 13 2017

Here's the test.
idbMemoryLeak.html
2.6 KB View Download
Received signal 4 ILL_ILLOPN 7f7eda7729e0
#0 0x7f7eded58df7 base::debug::StackTrace::StackTrace()
#1 0x7f7eded5897f base::debug::(anonymous namespace)::StackDumpSignalHandler()
#2 0x7f7edea85330 <unknown>
#3 0x7f7eda7729e0 WTF::partitionsOutOfMemoryUsing2G()
#4 0x7f7eda7727e9 WTF::Partitions::handleOutOfMemory()
#5 0x7f7edee0da02 base::partitionOutOfMemory()
#6 0x7f7edee0d9d1 base::PartitionAllocSlowPath()
#7 0x7f7eda772659 WTF::PartitionAllocator::allocateBacking()
#8 0x7f7ed955558c WTF::Vector<>::expandCapacity()
#9 0x7f7ed9552865 blink::SerializedScriptValue::toWireBytes()
#10 0x7f7ed8b67b94 blink::IDBObjectStore::put()
#11 0x7f7ed8b67214 blink::IDBObjectStore::put()
#12 0x7f7ed8963f19 blink::V8IDBObjectStore::putMethodCallback()
#13 0x3e365e609b2b <unknown>
  r8: 00000000ffffffff  r9: 0000000000000000 r10: 0000000000000022 r11: 0000000000000246
 r12: 0000000000801000 r13: 007ffffffffff000 r14: 0000000000801000 r15: 00007f7eda7a9d18
  di: 00007ffd89db7728  si: 0000000000805000  bp: 0000000000000000  bx: 0000000000000000
  dx: 0000000000000003  ax: 0000000080000000  cx: 00000003b80a9000  sp: 00007ffd89db7720
  ip: 00007f7eda7729e0 efl: 0000000000010202 cgf: 0000000000000033 erf: 0000000000000000
 trp: 0000000000000006 msk: 0000000000000000 cr2: 0000000000000000
[end of stack trace]

Comment 5 by ajha@chromium.org, Feb 14 2017

Labels: Needs-Triage-M56
Labels: M-56 M-57 Merge-Merged M-58
Owner: cmumford@chromium.org
Status: Fixed (was: Unconfirmed)
Labels: -Needs-Triage-M56 TE-Verified-57.0.2987.54 TE-Verified-M57
Verified the issue on Win 10 using 57.0.2987.54 and its working fine now.
Previously it used to be crashed when the counter reached 23+ or so, now checked it running more than 10 mins and did not observed any crashes.
Hence adding respective TE-Verified labels for the same.
691607_Feb_15.mp4
413 KB View Download
:D Woooo! Glad it worked.
Owner: dmu...@chromium.org
switching owner for my accounting
Issue 698028 has been merged into this issue.
Issue 636105 has been merged into this issue.
Issue 700044 has been merged into this issue.
Issue 586620 has been merged into this issue.

Sign in to add a comment