Closed Bug 724617 Opened 14 years ago Closed 8 years ago

[meta] Firefox 11.0 Crash Report [@ hang | NtCreateFile ]

Categories

(Core Graveyard :: Plug-ins, defect)

Product:
Core Graveyard
Component:
Plug-ins
Platform:
x86
Windows 7
Type:
defect
Priority:
Not set
Severity:
critical

Tracking

(firefox11+)

Status:
RESOLVED INCOMPLETE
Tracking Flags:
Tracking Status
firefox11 + ---

People

(Reporter: marcia, Assigned: smadayag)

References

Details

(Keywords: hang, meta)

Crash Data

Noticed this while looking at Beta crash stats. Filing bug so we have something to track the issue since it is in the top 10 of early beta data and we want bugs associated with all the top crashes. https://crash-stats.mozilla.com/report/list?signature=hang%20|%20NtCreateFile links to the crashes. Some correlations: hang | NtCreateFile|EXCEPTION_BREAKPOINT (121 crashes) 53% (64/121) vs. 45% (7284/16009) firefox-hotfix@mozilla.org The user below has a file LVPrcInj.dll which seems to be related to a logitech webcam, but I don't see an overall module correlation to that dll in the correlation report. https://crash-stats.mozilla.com/report/index/db1ac3f6-4e57-4ce3-9c4c-01b8b2120206 Frame Module Signature Source 0 ntdll.dll KiFastSystemCallRet 1 ntdll.dll NtCreateFile 2 LVPrcInj.dll LVPrcInj.dll@0x2e62 3 @0x7fffffff 4 NPSWF32.dll F132471607_____________________________________________________________________________ F648544825____________________________________________________________:2409 5 NPSWF32.dll F_1692058121_________________________________________________________ F_1228312162___________________________________________:85 6 NPSWF32.dll F795362988___________________________________________________ F447597432___________________________________________:4232 7 NPSWF32.dll F_143720307_____________________________________ F447597432___________________________________________:3789 8 NPSWF32.dll F_590358957_______________ F_1897618207________________________________________:46 9 NPSWF32.dll F_143720307_____________________________________ F447597432___________________________________________:3454 10 NPSWF32.dll F1786948743_______________________________ F_1406208219______________________________________________________:189 11 NPSWF32.dll F1786948743_______________________________ F_1406208219______________________________________________________:189 12 NPSWF32.dll F1479269175_______________________________________________________________________________________________________________________ 13 NPSWF32.dll free_frame F968498621______________________________________________________:436 14 NPSWF32.dll match F968498621______________________________________________________:4467 15 NPSWF32.dll F_66105703______________________________ F_1055812748____________________________________:6429 16 NPSWF32.dll F_66105703______________________________ F_1055812748____________________________________:6438 17 NPSWF32.dll F1939775799________________________________ F1792260806___________________________________________:253 18 NPSWF32.dll F1587269073______________________________________________________________________ F_1055812748____________________________________:28531 19 NPSWF32.dll F1678868676________________________________ F_1055812748____________________________________:28500 20 @0x661e8bf 21 NPSWF32.dll F_1865229927____________________________________________________________________________________________________________________________________________________________________________________________ F1618650697__________________________________________:408 22 NPSWF32.dll F_1427589130_________________ F447597432___________________________________________:1191 23 NPSWF32.dll F1391759860_________________ F_600042814_____________________________________:438 24 NPSWF32.dll F1059397351_______________________________________________________________________________________________________________________ F_1790492338_______________________________________:69 25 NPSWF32.dll F_1427589130_________________ F447597432___________________________________________:1191 26 NPSWF32.dll F1540094443_____________________________________ F_838025704________________________________________________________:1177 27 NPSWF32.dll F894294227_______________________________________________________________________ F918497155_______________________________________________________________________________:17353 28 NPSWF32.dll F1437472147_____________________ F_1897618207________________________________________:244 29 NPSWF32.dll F_653180670__________________________________________________ F_1530258674__________________________________________________:296 30 NPSWF32.dll F_966691389_____________________________ F_2059477748________________________________________________________:188 31 NPSWF32.dll F_118520524_________________________________________________________ F185401610________________________________________________________:116 32 NPSWF32.dll F860743057__________ F_1942443180__________________________________:539 33 @0x7dcc3a4 34 NPSWF32.dll F_797141883___________________________________________________________ F_1530258674__________________________________________________:142 35 NPSWF32.dll F_501710443___________________________ F1756712150____________________________________________________________:106 36 NPSWF32.dll F_2122742607________________________________________ F1756712150____________________________________________________________:242 37 NPSWF32.dll F_1238009141___________________________________ F1970998866________________________________________________________:319 38 NPSWF32.dll F_693444022__________________________________________________
(In reply to Marcia Knous [:marcia] from comment #0) > hang | NtCreateFile|EXCEPTION_BREAKPOINT (121 crashes) > 53% (64/121) vs. 45% (7284/16009) firefox-hotfix@mozilla.org I don't think this is particularly out of the ordinary, but definitely a good data point. The percentage of users with the firefox-hotfix isn't too much higher than for other a random sampling of crashes.
tracking-firefox11: --- → +
Depends on: 724802
Keywords: meta
Summary: Firefox 11.0 Crash Report [@ hang | NtCreateFile ] → [meta] Firefox 11.0 Crash Report [@ hang | NtCreateFile ]
Depends on: 724803
Josh - is there any more information that we can collect for you to help make this bug actionable? Were there any code changes in FF11 that may be related to this hang? Thanks!
Assignee: nobody → joshmoz
Sal, this seems like an issue with the Flash plugin. See the two "Depends on" bugs for this bug. Do you have any information? This does suspiciously show up in Firefox 11 and not earlier, but I still think it is likely a Flash issue because of the stack.
Assignee: joshmoz → smadayag
A partial list of plugin patches that landed for Firefox 11 (I think, hg dates are frustrating): Bug 700835 - Restore plugin caching to work around bad crash bug in latest Apple Java updates. Bug 687610 - QuickTime can't play media from the network cache when running OOP. Bug 683967 - Disable child side aborts on parent hang Bug 90268: Part 1, improve support for re-parenting native widgets. Bug 429592 part whatever - disable the hang monitor during all tests which also disable the DOM script timeout, since the hang monitor relies on the DOM script timeout to avoid firing on long content script Bug 429592 - Hook up more correct hang monitoring with Cocoa widgets to avoid false positives and negatives Bug 429592 disable hang monitor during GLib main loop poll Bug 705154 Temporarily disable the hang monitor on Mac because it is probably triggering false positives Bug 705365 - shorten the plugin hang timeout so that it is shorter than the main process hang timeout Bug 705748 - Disable the hang monitor by preference until existing found issues are resolved Bug 705154 - Hang monitor incorrectly killing the app on mac when the user is idle: unlike other platforms, the main loop on Mac doesn't call processNextEvent in a loop: instead it calls into [NSApplication run] and we check whether there are pending XPCOM events during processing. Suspend the hang monitor when we pause for incoming events by subclassing [NSApplication nextEventMatchingMask]
All of the "hang monitor" patches are irrelevant, it is disabled. Do *all* of these stacks go through the logitech webcam driver, or is that just one particular one which was picked out? NtCreateFile is a pretty generic signature, and I want to make sure we're not lumping too much together under it.
is it possible to get a specific model number of the logitech? i can run it through our webcam tests, but it would be great to have STRs associated with the hang...
(In reply to Benjamin Smedberg [:bsmedberg] from comment #5) > Do *all* of these stacks go through the logitech webcam driver From what I see in correlation reports, no: Correlations for 10.0.2 Windows NT: 100% (65/65) vs. 6% (3354/60364) plugin-container.exe 97% (63/65) vs. 5% (3018/60364) NPSWF32.dll 97% (63/65) vs. 6% (3460/60364) dsound.dll 97% (63/65) vs. 7% (4396/60364) schannel.dll 97% (63/65) vs. 10% (6092/60364) mlang.dll 95% (62/65) vs. 48% (28772/60364) wdmaud.drv 92% (60/65) vs. 47% (28601/60364) midimap.dll 92% (60/65) vs. 48% (28908/60364) msacm32.dll 91% (59/65) vs. 47% (28516/60364) msacm32.drv 100% (65/65) vs. 58% (35198/60364) urlmon.dll 95% (62/65) vs. 61% (36786/60364) icm32.dll 91% (59/65) vs. 60% (35992/60364) iertutil.dll 100% (65/65) vs. 70% (42044/60364) secur32.dll 82% (53/65) vs. 54% (32309/60364) normaliz.dll 32% (21/65) vs. 5% (3060/60364) credssp.dll 100% (65/65) vs. 74% (44511/60364) wininet.dll 43% (28/65) vs. 19% (11687/60364) msctfime.ime 32% (21/65) vs. 12% (7152/60364) ksuser.dll 78% (51/65) vs. 59% (35794/60364) msctf.dll 32% (21/65) vs. 14% (8334/60364) avrt.dll 71% (46/65) vs. 58% (34795/60364) netapi32.dll 68% (44/65) vs. 56% (34058/60364) comres.dll 100% (65/65) vs. 89% (53698/60364) mscms.dll 68% (44/65) vs. 58% (34716/60364) ws2help.dll 100% (65/65) vs. 90% (54589/60364) userenv.dll 9% (6/65) vs. 1% (801/60364) asoehook.dll 100% (65/65) vs. 94% (56896/60364) crypt32.dll 100% (65/65) vs. 94% (56910/60364) msasn1.dll Correlations for 11.0 Windows NT: 100% (70/70) vs. 27% (8311/31249) plugin-container.exe 93% (65/70) vs. 23% (7338/31249) dsound.dll 94% (66/70) vs. 25% (7841/31249) mlang.dll 93% (65/70) vs. 25% (7889/31249) schannel.dll 80% (56/70) vs. 21% (6491/31249) NPSWF32.dll 93% (65/70) vs. 56% (17582/31249) wdmaud.drv 90% (63/70) vs. 56% (17516/31249) midimap.dll 90% (63/70) vs. 57% (17682/31249) msacm32.dll 89% (62/70) vs. 56% (17526/31249) msacm32.drv 47% (33/70) vs. 15% (4583/31249) credssp.dll 97% (68/70) vs. 66% (20489/31249) urlmon.dll 97% (68/70) vs. 69% (21444/31249) secur32.dll 46% (32/70) vs. 20% (6234/31249) ksuser.dll 86% (60/70) vs. 61% (18991/31249) iertutil.dll 44% (31/70) vs. 21% (6496/31249) avrt.dll 77% (54/70) vs. 54% (16781/31249) normaliz.dll 89% (62/70) vs. 67% (20806/31249) icm32.dll 21% (15/70) vs. 2% (577/31249) rpnpshimswf.dll 97% (68/70) vs. 79% (24614/31249) wininet.dll 30% (21/70) vs. 13% (4106/31249) msctfime.ime 19% (13/70) vs. 2% (636/31249) rpcommon150browserrecordplugin.dll 26% (18/70) vs. 9% (2899/31249) GdiPlus.dll 73% (51/70) vs. 59% (18318/31249) msctf.dll 19% (13/70) vs. 5% (1602/31249) rpchrome150browserrecordhelper.dll 43% (30/70) vs. 30% (9490/31249) sspicli.dll 11% (8/70) vs. 1% (440/31249) NPSWF32_11_2_202_160.dll 13% (9/70) vs. 3% (958/31249) serwvdrv.dll 13% (9/70) vs. 3% (958/31249) umdmxfrm.dll 46% (32/70) vs. 39% (12168/31249) profapi.dll 9% (6/70) vs. 2% (733/31249) quartz.dll 33% (23/70) vs. 27% (8531/31249) msvcr90.dll 9% (6/70) vs. 3% (984/31249) BtMmHook.dll Those are from yesterday, and it looks like that driver wasn't loaded in any of those hangs that day.
It's #39 top crasher and #13 top plugin crasher in 11.0.
Keywords: topcrash
Keywords: topcrash
Resolving old bugs which are likely not relevant any more, since NPAPI plugins are deprecated.
Status: NEW → RESOLVED
Closed: 8 years ago
Resolution: --- → INCOMPLETE
Product: Core → Core Graveyard
You need to log in before you can comment on or make changes to this bug.