Closed Bug 891891 Opened 11 years ago Closed 6 years ago

Intermittent test_bug594730.html, test_hc_attrnextsiblingnull.html, test_secondPlugin.html, test_bug593174.html, test_pluginstream_err.html, test_bug73586.html, browser_webconsole_network_panel.js | application crashed [@ GetContextFromObjectOrDefault]

Categories

(Core :: XPConnect, defect, P5)

Product:
Core
Component:
XPConnect
Platform:
x86_64
macOS
Type:
defect

Tracking

()

Status:
RESOLVED WONTFIX

People

(Reporter: RyanVM, Unassigned)

References

Details

(Keywords: crash, intermittent-failure, Whiteboard: [leave open])

Crash Data

Attachments

(1 file)

Print diagnostics when an XPCWrappedJS is QI-ed off-main-thread. v1
1.89 KB, patch
mrbkap
: review+
Details | Diff | Splinter Review
https://tbpl.mozilla.org/php/getParsedLog.php?id=25094632&tree=Mozilla-Inbound Rev4 MacOSX Lion 10.7 mozilla-inbound opt test mochitest-5 on 2013-07-09 17:10:05 PDT for push a65599fc1dd4 slave: talos-r4-lion-086 17:15:04 INFO - 169603 INFO TEST-START | /tests/parser/htmlparser/tests/mochitest/test_bug594730.html 17:15:04 INFO - 169604 INFO TEST-PASS | /tests/parser/htmlparser/tests/mochitest/test_bug594730.html | Decoded bytes should have matched the Unicode escape. 17:15:04 INFO - 169605 INFO TEST-PASS | /tests/parser/htmlparser/tests/mochitest/test_bug594730.html | Decoded bytes should have matched the Unicode escape. 17:15:04 INFO - 169606 INFO TEST-PASS | /tests/parser/htmlparser/tests/mochitest/test_bug594730.html | Decoded bytes should have matched the Unicode escape. 17:15:04 INFO - 169607 INFO TEST-PASS | /tests/parser/htmlparser/tests/mochitest/test_bug594730.html | Decoded bytes should have matched the Unicode escape. 17:15:04 INFO - 169608 INFO TEST-PASS | /tests/parser/htmlparser/tests/mochitest/test_bug594730.html | Decoded bytes should have matched the Unicode escape. 17:15:04 INFO - 169609 INFO TEST-PASS | /tests/parser/htmlparser/tests/mochitest/test_bug594730.html | Decoded bytes should have matched the Unicode escape. 17:15:04 INFO - 169610 INFO TEST-PASS | /tests/parser/htmlparser/tests/mochitest/test_bug594730.html | Decoded bytes should have matched the Unicode escape. 17:15:04 INFO - 169611 INFO TEST-PASS | /tests/parser/htmlparser/tests/mochitest/test_bug594730.html | Decoded bytes should have matched the Unicode escape. 17:15:04 INFO - 169612 INFO TEST-PASS | /tests/parser/htmlparser/tests/mochitest/test_bug594730.html | Decoded bytes should not have matched the Unicode escape. 17:15:04 INFO - 169613 INFO TEST-END | /tests/parser/htmlparser/tests/mochitest/test_bug594730.html | finished in 179ms 17:15:06 WARNING - TEST-UNEXPECTED-FAIL | /tests/parser/htmlparser/tests/mochitest/test_bug594730.html | Exited with code 1 during test run 17:15:06 INFO - INFO | automation.py | Application ran for: 0:03:18.155683 17:15:06 INFO - INFO | zombiecheck | Reading PID log: /var/folders/qd/srwd5f710sj0fcl9z464lkj00000gn/T/tmp4G9_nipidlog 17:15:06 INFO - mozcrash INFO | Downloading symbols from: http://ftp.mozilla.org/pub/mozilla.org/firefox/tinderbox-builds/mozilla-inbound-macosx64/1373408158/firefox-25.0a1.en-US.mac.crashreporter-symbols.zip 17:15:06 INFO - Downloading symbols from: http://ftp.mozilla.org/pub/mozilla.org/firefox/tinderbox-builds/mozilla-inbound-macosx64/1373408158/firefox-25.0a1.en-US.mac.crashreporter-symbols.zip 17:15:38 WARNING - PROCESS-CRASH | /tests/parser/htmlparser/tests/mochitest/test_bug594730.html | application crashed [@ GetContextFromObjectOrDefault] 17:15:38 INFO - Crash dump filename: /var/folders/qd/srwd5f710sj0fcl9z464lkj00000gn/T/tmpBg62Eo/minidumps/6D2DCFB2-C709-42B6-B748-D458F1C5E1B2.dmp 17:15:38 INFO - Operating system: Mac OS X 17:15:38 INFO - 10.7.2 11C74 17:15:38 INFO - CPU: amd64 17:15:38 INFO - family 6 model 23 stepping 10 17:15:38 INFO - 2 CPUs 17:15:38 INFO - Crash reason: EXC_BAD_ACCESS / KERN_INVALID_ADDRESS 17:15:38 INFO - Crash address: 0x0 17:15:38 INFO - Thread 17 (crashed) 17:15:38 INFO - 0 XUL!GetContextFromObjectOrDefault [xpcprivate.h:a65599fc1dd4 : 458 + 0x0] 17:15:38 INFO - rbx = 0x000000010f609aa0 r12 = 0x00000001177feb80 17:15:38 INFO - r13 = 0x00000001031e1ffc r14 = 0x00000001177feb80 17:15:38 INFO - r15 = 0x00007fff74ebdf60 rip = 0x00000001023ad8e7 17:15:38 INFO - rsp = 0x000000010f609800 rbp = 0x000000010f609910 17:15:38 INFO - Found by: given as instruction pointer in context 17:15:38 INFO - 1 XUL!nsXPCWrappedJSClass::DelegatedQueryInterface(nsXPCWrappedJS*, nsID const&, void**) [XPCWrappedJSClass.cpp:a65599fc1dd4 : 624 + 0x7] 17:15:38 INFO - rbx = 0x000000010f609aa0 r12 = 0x00000001177feb80 17:15:38 INFO - r13 = 0x00000001031e1ffc r14 = 0x000000010f609ae8 17:15:38 INFO - r15 = 0x000000010f609aa0 rip = 0x00000001023ad3c1 17:15:38 INFO - rsp = 0x000000010f609920 rbp = 0x000000010f609a70 17:15:38 INFO - Found by: call frame info 17:15:38 INFO - 2 XUL!nsQueryReferent::operator()(nsID const&, void**) const [nsWeakReference.cpp:a65599fc1dd4 : 56 + 0x8] 17:15:38 INFO - rbx = 0x000000010f609aa0 r12 = 0x00000001177f2400 17:15:38 INFO - r13 = 0x0000000000000001 r14 = 0x000000010f609ae8 17:15:38 INFO - r15 = 0x0000000103df33f0 rip = 0x0000000101659fc4 17:15:38 INFO - rsp = 0x000000010f609a80 rbp = 0x000000010f609a90 17:15:38 INFO - Found by: call frame info 17:15:38 INFO - 3 XUL!nsCOMPtr_base::assign_from_helper(nsCOMPtr_helper const&, nsID const&) [nsCOMPtr.cpp:a65599fc1dd4 : 110 + 0x11] 17:15:38 INFO - rbx = 0x000000010f609b00 r12 = 0x00000001177f2400 17:15:38 INFO - r13 = 0x0000000000000001 r14 = 0x0000000000000001 17:15:38 INFO - r15 = 0x0000000103df33f0 rip = 0x000000010165679b 17:15:38 INFO - rsp = 0x000000010f609aa0 rbp = 0x000000010f609ab0 17:15:38 INFO - Found by: call frame info 17:15:38 INFO - 4 XUL!nsDocLoader::DoFireOnStateChange(nsIWebProgress*, nsIRequest*, int&, tag_nsresult) [nsCOMPtr.h:a65599fc1dd4 : 747 + 0xb] 17:15:38 INFO - rbx = 0x0000000100370048 r12 = 0x00000001177f2400 17:15:38 INFO - r13 = 0x0000000000000001 r14 = 0x0000000000000001 17:15:38 INFO - r15 = 0x0000000103df33f0 rip = 0x0000000102414e77 17:15:38 INFO - rsp = 0x000000010f609ac0 rbp = 0x000000010f609b30 17:15:38 INFO - Found by: call frame info
Blocks: 895024
Bobby, we're seeing this and bug 895024 intermittently. Any chance you could take a look if you get a chance?
Flags: needinfo?(bobbyholley+bmo)
Whoa, this is bad - for some reason, thread 17 is touching XPCWrappedJS stuff, which is verboten (and causes XPConnect to abort). It would be very helpful to know what XPCWrappedJS this is. How do I go about getting my hands on that minidump?
Flags: needinfo?(bobbyholley+bmo) → needinfo?(ryanvm)
(In reply to Bobby Holley (:bholley) from comment #7) > Whoa, this is bad - for some reason, thread 17 is touching XPCWrappedJS > stuff, which is verboten (and causes XPConnect to abort). > > It would be very helpful to know what XPCWrappedJS this is. How do I go > about getting my hands on that minidump? You can ping #releng to can get it for you (though they don't stick around on the slave forever, so you'll want to do so reasonably soon after it happens).
Flags: needinfo?(ryanvm)
Crash Signature: [@ GetContextFromObjectOrDefault]
Depends on: 904474
Ok, looks like we just hit this. Callek, can you grab me the minidump from comment 12?
Flags: needinfo?(bugspam.Callek)
wow this dissapeared fast: [cltbld@talos-r4-lion-012.build.scl1.mozilla.com ~]$ ls /var/folders/s5/b_ptq5b51zv6wvdzs2g4jg7400000w/T/tmpEr8Gok/minidumps/9CA64D0A-79CE-4B97-A90D-F0273784779D.dmp ls: /var/folders/s5/b_ptq5b51zv6wvdzs2g4jg7400000w/T/tmpEr8Gok/minidumps/9CA64D0A-79CE-4B97-A90D-F0273784779D.dmp : No such file or directory [cltbld@talos-r4-lion-012.build.scl1.mozilla.com ~]$ ls /var/folders/s5/b_ptq5b51zv6wvdzs2g4jg7400000w/T/tmpEr8Gok/minidumps ls: /var/folders/s5/b_ptq5b51zv6wvdzs2g4jg7400000w/T/tmpEr8Gok/minidumps: No such file or directory [cltbld@talos-r4-lion-012.build.scl1.mozilla.com ~]$ Connection to talos-r4-lion-012.build.mozilla.org closed by remote host. Connection to talos-r4-lion-012.build.mozilla.org closed.
Flags: needinfo?(bugspam.Callek)
Summary: Intermittent test_bug594730.html | Exited with code 1 during test run | application crashed [@ GetContextFromObjectOrDefault] → Intermittent test_bug594730.html, test_hc_attrnextsiblingnull.html | application crashed [@ GetContextFromObjectOrDefault]
Hey Callek, can I get a minidump?
Flags: needinfo?(bugspam.Callek)
(In reply to Bobby Holley (:bholley) from comment #17) > Hey Callek, can I get a minidump? http://people.mozilla.org/~jwood/tmp/1C7214B6-7F15-4453-BF98-719FC8C61A87.dmp
Flags: needinfo?(bugspam.Callek)
(In reply to Justin Wood (:Callek) from comment #18) > (In reply to Bobby Holley (:bholley) from comment #17) > > Hey Callek, can I get a minidump? > > http://people.mozilla.org/~jwood/tmp/1C7214B6-7F15-4453-BF98-719FC8C61A87.dmp 403 forbidden.
Flags: needinfo?(bugspam.Callek)
(In reply to Bobby Holley (:bholley) from comment #20) > (In reply to Justin Wood (:Callek) from comment #18) > > (In reply to Bobby Holley (:bholley) from comment #17) > > > Hey Callek, can I get a minidump? > > > > http://people.mozilla.org/~jwood/tmp/1C7214B6-7F15-4453-BF98-719FC8C61A87.dmp > > 403 forbidden. sorry people's default umask bit me again. should be fixed
Flags: needinfo?(bugspam.Callek)
Let's try to figure out which XPCWrappedJS is causing this. I've filed a reminder for myself to back this out by sept 3.
Attachment #793321 - Flags: review?(mrbkap)
From what I can glean from googling, I can't actually do anything with a mac minidump other than get the stack (which is already in the logs). Ted, can you confirm that this is correct? Assuming so, I'll land the diagnostic patch in comment 23.
Flags: needinfo?(ted)
Summary: Intermittent test_bug594730.html, test_hc_attrnextsiblingnull.html | application crashed [@ GetContextFromObjectOrDefault] → Intermittent test_bug594730.html, test_hc_attrnextsiblingnull.html, test_secondPlugin.html, test_bug593174.html, test_pluginstream_err.html, test_bug73586.html, browser_webconsole_network_panel.js | application crashed [@ GetContextFromObjectOrDefault]
No longer blocks: 895024
Correct, there are no tools to do anything more useful with a minidump on Mac than get the stack out. (Also, vaguely related, RelEng is on track to fix bug 642167 in the near future so that minidumps from test slaves will be simple to get.)
Flags: needinfo?(ted)
(In reply to Ted Mielczarek [:ted.mielczarek] from comment #29) > Correct, there are no tools to do anything more useful with a minidump on > Mac than get the stack out. > > (Also, vaguely related, RelEng is on track to fix bug 642167 in the near > future so that minidumps from test slaves will be simple to get.) Ok. Waiting on Blake's review for the diagnostic patch then.
Comment on attachment 793321 [details] [diff] [review] Print diagnostics when an XPCWrappedJS is QI-ed off-main-thread. v1 Review of attachment 793321 [details] [diff] [review]: ----------------------------------------------------------------- ::: js/xpconnect/src/XPCWrappedJSClass.cpp @@ +573,5 @@ > void** aInstancePtr) > { > + if (MOZ_UNLIKELY(!NS_IsMainThread())) { > + printf("Uh oh! DelegatedQueryInterface called off-main-thread!\n"); > + printf("Name: %s\n", GetInterfaceName()); Is it worth printing the IID as well? @@ +584,5 @@ > + } else { > + printf("Unable to get origin from principal :-(\n"); > + } > + nsAutoCString loc(EnsureCompartmentPrivate(c)->GetLocation()); > + printf("Global's Location: %s\n", loc.get()); I don't think you actually need the explicit local autostring here, right?
Attachment #793321 - Flags: review?(mrbkap) → review+
(In reply to Blake Kaplan (:mrbkap) from comment #31) > Comment on attachment 793321 [details] [diff] [review] > Is it worth printing the IID as well? I'm not sure what information that would give us on top of the interface name, and I don't know how to format an IID for printing offhand. > > + nsAutoCString loc(EnsureCompartmentPrivate(c)->GetLocation()); > > + printf("Global's Location: %s\n", loc.get()); > > I don't think you actually need the explicit local autostring here, right? GetLocation returns an nsACString, which isn't null-terminated, so it doesn't have a get. So I can't pass it to printf, though maybe there's some other way.
The diagnostics morphed the crash into bug 909613, which in turn pointed us to bug 904474. The diagnostics have served their purpose, so I've backed them out: https://hg.mozilla.org/integration/mozilla-inbound/rev/7f683b01fb4f
Bulk assigning P3 to all open intermittent bugs without a priority set in Firefox components per bug 1298978.
Priority: -- → P3
Priority: P3 → P5
Closing because no crash reported since 12 weeks.
Status: NEW → RESOLVED
Closed: 6 years ago
Resolution: --- → WONTFIX
You need to log in before you can comment on or make changes to this bug.

Attachment

General

Creator:
Created:
Updated:
Size: