249420 - firefox 0.9.1 crashes on startup

Status: RESOLVED WORKSFORME

(Core :: XPCOM, defect)

Description

[James F. Hranicky]

User-Agent:       Mozilla/5.0 (X11; U; SunOS sun4u; en-US; rv:1.7) Gecko/20040618 Firefox/0.8.0+
Build Identifier: ftp://ftp.cise.ufl.edu/pub/mirrors/mozilla/firefox/releases/0.9.1/firefox-0.9.1-source.tar.bz2

firefox 0.9.1 on Solaris 8 crashes upon startup. firefox was configured
like so:

------------

FLAGS="-g"
CC="cc $FLAGS" CXX="CC $FLAGS" LDFLAGS="-L/usr/local/lib -R/usr/local/lib"

MOZ_PHOENIX=1
export FLAGS CC CXX LDFLAGS MOZ_PHOENIX

echo "C Compiler: $CC"
echo "C++ Compiler: $CXX"

    ./configure \
        --with-ft-prefix=/usr/local \
        --with-pthreads \
        --with-system-jpeg=/usr/local \
        --with-system-zlib=/usr/local \
        --with-system-png=/usr/local \
        --enable-default-toolkit=gtk2 \
        --with-gtk-prefix=/usr/local \
        --disable-calendar \
        --disable-mailnews \
        --disable-composer \
        --enable-xft \
        --disable-xprint \
        --enable-crypto \
        --enable-js-ultrasparc \
        --enable-necko-protocols=all \
        --with-gssapi=/usr/local \
        --enable-debug \
        --disable-optimize

Also, the patch from bug 215112 has been applied as well.

Reproducible: Always
Steps to Reproduce:
1. Start firefox
2. 
3.

Actual Results:  
firefox crashed

Expected Results:  
firefox runs

When firefox runs, it cores at this point:

--------------
*** Registering mozMySpellModule components (all right -- a generic module!)
*** Registering mozSpellCheckerModule components (all right -- a generic module!)
*** Registering nsGnomeVFSModule components (all right -- a generic module!)
*** Registering nsHttpGssapiAuthModule components (all right -- a generic module!)
*** Registering nsBrowserCompsModule components (all right -- a generic module!)
nsNativeComponentLoader: autoregistering succeeded
WARNING: NS_ENSURE_TRUE(NS_SUCCEEDED( rv)) failed, file nsINIParser.cpp, line 51
Segmentation Fault - core dumped
-------------

A dbx backtrace shows this:

-------------

=>[1] PL_HashString(key = (nil)), line 523 in "plhash.c"
  [2] RDFServiceImpl::RegisterDataSource(0x13a688, 0x13b018, 0x1, 0x0, 0x0,
0x0), at 0xfc95bf14 
  [3] nsChromeUIDataSource::nsChromeUIDataSource(this = 0x13b018, aComposite =
0x13be90), line 75 in "nsChromeUIDataSource.cpp"
  [4] NS_NewChromeUIDataSource(aComposite = 0x13be90, aResult = 0x13a614), line
390 in "nsChromeUIDataSource.cpp"
  [5] nsChromeRegistry::AddToCompositeDataSource(this = 0x13a5d0, aUseProfile =
0), line 3158 in "nsChromeRegistry.cpp"
  [6] nsChromeRegistry::LoadInstallDataSource(this = 0x13a5d0), line 3197 in
"nsChromeRegistry.cpp"
  [7] nsChromeRegistry::CheckForNewChrome(this = 0x13a5d0), line 3291 in
"nsChromeRegistry.cpp"
  [8] nsChromeRegistry::Init(this = 0x13a5d0), line 310 in "nsChromeRegistry.cpp"
  [9] nsChromeRegistryConstructor(aOuter = (nil), aIID = STRUCT, aResult =
0xffbed940), line 50 in "nsChromeFactory.cpp"
  [10] nsGenericFactory::CreateInstance(0x137088, 0x0, 0x6c8f8, 0xffbed940,
0x81010100, 0xff00), at 0xff11b358 
  [11] nsComponentManagerImpl::CreateInstanceByContractID(0xbd0e0, 0xf9cf72b9,
0x0, 0x6c8f8, 0xffbed940, 0xff3c7890), at 0xff04dec8 
  [12] nsComponentManagerImpl::GetServiceByContractID(0xbd0e0, 0xf9cf72b9,
0x6c8f8, 0xffbeda60, 0x0, 0x5), at 0xff04f840 
  [13] nsGetServiceByContractID::operator()(0xffbedb4c, 0x6c8f8, 0xffbeda60,
0x13a41c, 0x0, 0x0), at 0xff119574 
  [14] nsCOMPtr_base::assign_from_helper(0xffbedb70, 0xffbedb4c, 0x6c8f8,
0xffbedb98, 0x0, 0x0), at 0xff118c58 
  [15] nsCOMPtr<nsIChromeRegistry>::operator=(this = 0xffbedb70, rhs = CLASS),
line 642 in "nsCOMPtr.h"
  [16] nsChromeProtocolHandler::NewURI(this = 0x136fc8, aSpec = CLASS, aCharset
= (nil), aBaseURI = (nil), result = 0xffbee37c), line 604 in
"nsChromeProtocolHandler.cpp"
  [17] nsIOService::NewURI(this = 0x125f40, aSpec = CLASS, aCharset = (nil),
aBaseURI = (nil), result = 0xffbee37c), line 421 in "nsIOService.cpp"
  [18] NS_NewURI(0xffbee37c, 0xffbedd74, 0x0, 0x0, 0x0, 0x1), at 0xfa4bcfe4 
  [19] NS_NewURI(0xffbee37c, 0x6c85b, 0x0, 0x0, 0x0, 0x0), at 0xfa4b3f34 
  [20] nsWindowWatcher::URIfromURL(0x6c85b, 0x0, 0xffbee37c, 0x0, 0x0, 0x0), at
0xfa4a984c 
  [21] nsWindowWatcher::OpenWindowJS(0xdde68, 0x0, 0x6c85b, 0x5580a, 0x55811,
0x1), at 0xfa4a5698 
  [22] nsWindowWatcher::OpenWindow(0x1, 0x9b070, 0xfa4a5448, 0x5580a, 0x55811,
0xddef8), at 0xfa4a53f0 
  [23] ShowProfileManager(aProfileSvc = 0xb6d20, aNative = 0xa65b8), line 1325
in "nsAppRunner.cpp"
  [24] SelectProfile(aResult = 0xffbef054, aNative = 0xa65b8, aStartOffline =
0xffbef050), line 1543 in "nsAppRunner.cpp"
  [25] xre_main(argc = 1, argv = 0xffbef16c, aAppData = 0x6c7d8), line 1783 in
"nsAppRunner.cpp"
  [26] main(argc = 1, argv = 0xffbef16c), line 58 in "nsBrowserApp.cpp"
(dbx) 
--------------

The key passed into PL_HashString is null, causing a segault on dereference.

Comment 1

[Ryuichi KUBUKI]

Do you still have the same issue with a newer version of Firefox?

Comment 2

[Mark Trompell]

I have similiar problems with ff-1.0.4 on rpath-linux for amd64:
$ firefox
WARNING: NS_ENSURE_TRUE(NS_SUCCEEDED(rv)) failed, file nsINIParser.cpp, line 51
/usr/lib64/firefox/run-mozilla.sh: line 451: 13885 Segmentation fault     
"$prog" ${1+"$@"}

Comment 3

[Charles C. Van Tilburg]

Solaris 10 x86, Firefox 1.0.6 build using gcc 4.0.1 (gas, sun ld), 
patches snarfed from redhat's SRPM; all applied except their 
customization patches (solaris 10 does not include a compatible
pango), attempted run after make install gives:

# WARNING: Patch checker not supported for SunOS.
File descriptors set to 512
No Persistent Registry Found.
WARNING: NS_ENSURE_TRUE(NS_SUCCEEDED(rv)) failed, file nsINIParser.cpp, line 51
Type Manifest File: /users/ctilburg/.mozilla/firefox/p1nk1tpk.default/xpti.dat
nsNativeComponentLoader: autoregistering begins.
*** Registering xpcomObsoleteModule components (all right -- a generic module!)
*** Registering xpconnect components (all right -- a generic module!)
*** Registering nsUConvModule components (all right -- a generic module!)
*** Registering nsUCvMathModule components (all right -- a generic module!)
*** Registering nsI18nModule components (all right -- a generic module!)
*** Registering nsCtlLEModule components (all right -- a generic module!)
*** Registering nsJarModule components (all right -- a generic module!)
*** Registering nsCJVMManagerModule components (all right -- a generic module!)
*** Registering necko_core_and_primary_protocols components (all right -- a
generic module!)
*** Registering necko_secondary_protocols components (all right -- a generic
module!)
*** Registering nsPrefModule components (all right -- a generic module!)
*** Registering nsSecurityManagerModule components (all right -- a generic module!)
*** Registering nsRDFModule components (all right -- a generic module!)
*** Registering nsParserModule components (all right -- a generic module!)
*** Registering nsGfxPSModule components (all right -- a generic module!)
*** Registering nsGfxXprintModule components (all right -- a generic module!)
*** Registering nsGfxGTKModule components (all right -- a generic module!)
*** Registering nsImageLib2Module components (all right -- a generic module!)
*** Registering nsPluginModule components (all right -- a generic module!)
*** Registering nsWidgetGtk2Module components (all right -- a generic module!)
*** Registering XRemoteClientModule components (all right -- a generic module!)
*** Registering nsLayoutModule components (all right -- a generic module!)
*** Registering nsMorkModule components (all right -- a generic module!)
*** Registering docshell_provider components (all right -- a generic module!)
*** Registering embedcomponents components (all right -- a generic module!)
*** Registering Browser_Embedding_Module components (all right -- a generic module!)
*** Registering nsEditorModule components (all right -- a generic module!)
*** Registering nsTransactionManagerModule components (all right -- a generic
module!)
*** Registering nsComposerModule components (all right -- a generic module!)
*** Registering appshell components (all right -- a generic module!)
*** Registering nsChromeModule components (all right -- a generic module!)
*** Registering nsAccessibilityModule components (all right -- a generic module!)
*** Registering mozgnome components (all right -- a generic module!)
*** Registering BOOT components (all right -- a generic module!)
*** Registering NSS components (all right -- a generic module!)
*** Registering PKI components (all right -- a generic module!)
*** Registering nsFileViewModule components (all right -- a generic module!)
*** Registering nsFindComponent components (all right -- a generic module!)
*** Registering XRemoteServiceModule components (all right -- a generic module!)
*** Registering application components (all right -- a generic module!)
*** Registering nsToolkitCompsModule components (all right -- a generic module!)
*** Registering nsSoftwareUpdate components (all right -- a generic module!)
*** Registering JavaScript_Debugger components (all right -- a generic module!)
*** Registering nsCookieModule components (all right -- a generic module!)
*** Registering nsXMLExtrasModule components (all right -- a generic module!)
*** Registering nsAutoConfigModule components (all right -- a generic module!)
*** Registering nsSystemPrefModule components (all right -- a generic module!)
*** Registering TransformiixModule components (all right -- a generic module!)
*** Registering nsUniversalCharDetModule components (all right -- a generic module!)
*** Registering nsWebServicesModule components (all right -- a generic module!)
*** Registering nsInspectorModule components (all right -- a generic module!)
*** Registering nsP3PModule components (all right -- a generic module!)
*** Registering nsGnomeVFSModule components (all right -- a generic module!)
*** Registering nsNegotiateAuthModule components (all right -- a generic module!)
*** Registering nsBrowserCompsModule components (all right -- a generic module!)
nsNativeComponentLoader: autoregistering succeeded
WARNING: NS_ENSURE_TRUE(NS_SUCCEEDED(rv)) failed, file nsINIParser.cpp, line 51
Segmentation Fault

Comment 4

[Charles C. Van Tilburg]

(In reply to comment #3)

GOT IT: clue from Sun engineer regarding build with Studio 10 failure gave
me the probable location, but using gcc 4.0.1, the problem is missing logic
to set the correct low level interface routines:

In my case, the logic that sets these routines in:

mozilla/xpcom/reflect/xptcall/src/md/unix

incorrectly builds:

xptcstubs_x86_solaris.o
xptcinvoke_x86_solaris.o

when it should build:

xptcstubs_gcc_x86_unix.o
xptcinvoke_gcc_x86_unix.o

I hotwired this by compiling the latter source files with output to 
the former object files, re-running the make and make install and
things seem to work.

Hope this helps others.

Comment 5

[Charles C. Van Tilburg]

(In reply to comment #4)

Further notes of probable interest:

1. nsprpub/pr/src/md/unix has a build problem on os_SunOS_x86.s where
        the following flag is included erroneously: -Wa,-P

2. in layout/base/src/nsBidi.cpp, one must change CS->CSS and ES->ESS

3. in config/autoconf.mk, -mt needs to be removed at all locations.

4. there is a strange step, which seems to be a build order glitch, whereby 
   I had to add /usr/lib/mps to my LD_LIBRARY_PATH to gain a shared library,
   and then subsequently, the firefox build built that very library... ?!

5. NOTE: I have applied all the non-customization patches from the Redhat
   FC4 SRPM as well.  Included there is a gcc4 patch.  Be careful to use
   'gpatch' and NOT Sun's vanilla 'patch' or you will only get one diff 
   applied per file!!

my build environment setup script (gcc4 is in /usr/local/bin, bash):

 export MOZ_PHOENIX=1
 export MOZILLA_OFFICIAL=1
 export BUILD_OFFICIAL=1
 export LD_LIBRARY_PATH="/usr/local/lib"
 export 'RPM_OPT_FLAGS=-Os -g -pipe -Wp,-D_FORTIFY_SOURCE=2 -fexceptions -m32
-march=i386 -mtune=pentium4 -fasynchronous-unwind-tables'
 export PATH=/users/ctilburg/bin:/usr/local/bin:/usr/ccs/bin:/usr/bin
 export LDFLAGS='-lnsl'

my .mozconfig (this results in a considerably larger install than static):

    ac_add_options --enable-optimize="$RPM_OPT_FLAGS"

    ac_add_options --disable-ldap
    ac_add_options --disable-mailnews
    ac_add_options
--enable-extensions="cookie,xml-rpc,xmlextras,pref,transformiix,universalchardet,webservices,inspector,p3p,gnomevfs,negotiateauth"
    ac_add_options --enable-crypto
    ac_add_options --disable-composer
    ac_add_options --enable-single-profile
    ac_add_options --disable-profilesharing

    ac_add_options --disable-debug
    ac_add_options --disable-tests
    ac_add_options --disable-static
    ac_add_options --enable-shared
    ac_add_options --with-xprint
    ac_add_options --enable-xinerama
    ac_add_options --enable-x11-shm
    ac_add_options --enable-ctl
    ac_add_options --enable-default-toolkit=gtk2
    ac_add_options --disable-freetype2
    ac_add_options --enable-xft
    ac_add_options --without-system-nspr
    ac_add_options --without-system-zlib
    ac_add_options --without-system-jpeg
    ac_add_options --without-system-png
    ac_add_options --without-system-mng
    ac_add_options --x-includes=/usr/openwin/include
    ac_add_options --x-libraries=/usr/openwin/lib
    mk_add_options MOZ_PHOENIX=1
    mk_add_options BUILD_OFFICIAL=1
    mk_add_options MOZILLA_OFFICIAL=1

Comment 6

[Charles C. Van Tilburg]

(In reply to comment #5)

Be aware that a gcc 4.0.1 compile of FF will refuse to use any java
plugin on this platform, gcc specific or not.  It remains to be seen
whether Sun will provide gcc4 compatible plugin .so files.

If, as Leon Sha suggested, one copies over *solaris* and *SUNW* from 
the current mozilla trunk CVS checkout are to the 1.0.6 area: 

mozilla/xpcom/reflect/xptcall/src/md/unix

Then, the Sun Studio 10 compiler will succeed in building a working 
binary, and the java plugin will also work.

Leon used Studio 8 to generate his binaries.  My binaries also include
all non-customization patches from the RedHat FC4 SRPM.  I have 
offered to upload them, but have not received any word on interest.

Comment 7

[Frank Wein [:mcsmurf]]

Leon: I wonder if you could add support (attach a patch) for the gcc compiler
under Solaris like you did in Bug 295306 for the Sun compiler? The issue is
explained in Comment 4 and following.

Comment 8

[Charles C. Van Tilburg]

Attachment: Here are the patches I used from the Redhat FC4 SRPM

Comment 9

[emily chen]

works for me on Gecko/2007012914 Firefox/2.0.0.1 on Solaris 10, reporter: feel free to reopen if you can reproduce this bug with a latest release build

Comment 10

[Carsten Book [:Tomcat]]

also wfm on Gecko/2007012914 Firefox/2.0.0.1 on Solaris 10 and per comment#9 from emily chen