525725 - xulrunner -help crashes [@ LoadPlatformDirectory]

Status: RESOLVED FIXED

(Toolkit :: Startup and Profile System, defect)

Description

[[e-mail]]

Access violation - code c0000005
  nsresult rv = aBundleDirectory->Clone(getter_AddRefs(platformDir));
aBundleDirectory is null.

Call stack:
xul!LoadPlatformDirectory
xul!nsXREDirProvider::LoadBundleDirectories
xul!nsXREDirProvider::GetFilesInternal
xul!nsXREDirProvider::GetFiles
xpcom_core!FindProviderFile
xpcom_core!nsSupportsArray::EnumerateBackwards
xpcom_core!nsDirectoryService::Get
xpcom_core!AppendFromDirServiceList
xpcom_core!xptiInterfaceInfoManager::BuildFileSearchPath
xpcom_core!xptiInterfaceInfoManager::GetInterfaceInfoManagerNoAddRef
xpcom_core!NS_InitXPCOM3_P
xul!ScopedXPCOMStartup::Initialize
xul!DumpArbitraryHelp
xul!DumpHelp
xul!XRE_main
seamonkey!NS_internal_main
seamonkey!wmain
seamonkey!__tmainCRTStartup
seamonkey!wmainCRTStartup

So I'm assuming -help doesn't set mXULAppDir, but it should.

Comment 1

[[e-mail]]

Attachment: Proposed patch

Comment 3

[[e-mail]]

Pushed changeset ac5673902abd to mozilla-central.

Comment 4

[Mike Beltzner [:beltzner, not reading bugmail]]

Not blocking, but we should take the patch.

Comment 5

[[e-mail]]

Pushed changeset eca0dce6296c to releases/mozilla-1.9.2

Comment 6

[chris hofmann]

turns out that lots of people seem to do do this.  its top crash in early 3.6b3 data.  re-asking for blocking so it gets a good flag an back on the radar if a regression happens with the landing.

see Bug 529482 for crash analysis

Comment 7

[chris hofmann]

no reports of this on beta3 or 4 over the last couple of weeks.

there are reports on trunk and 3.5.5
http://crash-stats.mozilla.com/report/index/10dfc0f8-f1c0-4385-aee5-68cce2091203

and more at
http://crash-stats.mozilla.com/report/list?product=Firefox&query_search=signature&query_type=contains&query=LoadPlatformDirectory&date=&range_value=2&range_unit=weeks&do_query=1&signature=LoadPlatformDirectory

does the patch need to go to other branches?

Comment 9

[[e-mail]]

(In reply to comment #7)
> does the patch need to go to other branches?
No, this particular crash should just be a regression from bug 521262.