reassigned to daniel.

I cannot reproduce this one, can you please provide a stack trace?

some folk told me that he has seen such crash in regcomp some time ago
(exactly solaris sparc non-pro): happens when regcomp gets false
parameters; regcomp then crashes due to a compiler bug.  Can you
please verify whether
/oo/BuildDir/ooo_11beta2_src/solver/644/unxsols4.pro/bin/java_uno_accessbridge.jar
exists?

I have a variation on this problem building setup2/util on sparc64
linux debian/testing gcc-3.3 (see full listing attached):

Trace Message: javaloader - could not init javaloader cause of
javaloader error - no C++ environment available
register component
'vnd.sun.star.expand:$UNO_JAVA_COMPPATH/java_uno_accessbridge.jar' in
registry '../unxlngs.pro/b in/setup_services.tmp' failed!
error (RuntimeException): javaloader error - no C++ environment available
dmake:  Error code 1, while making '../unxlngs.pro/bin/setup_services.rdb'

Created attachment 6399 [details]
setup2/util build error messages

@sparcmoz: Please verify whether you have a libgcc3_uno.so in your
LD_LIBRARY_PATH, and do a ldd -r on it to verify that it can be
dlopen()ed.
The error message you are facing seems to be that the gcc3 C++ UNO
environment cannot be instantiated, because the bridge init code is
not available.

@Pavel: Supply a stack trace and verify whether
/oo/BuildDir/ooo_11beta2_src/solver/644/unxsols4.pro/bin/java_uno_accessbridge.jar
exists.
I assume that you receive notifications for this issue, but I cannot
help you, if you don't supply the above information.

I do not know how to provide stack trace - could you help me?

that file exists:

oo@ultra:~/BuildDir/ooo_11beta2_src> ls -al
/oo/BuildDir/ooo_11beta2_src/solver/644/unxsols4.pro/bin/java_uno_accessbridge.jar
-rw-r--r--   1 oo       staff     152239 kvě 20 19:40
/oo/BuildDir/ooo_11beta2_src/solver/644/unxsols4.pro/bin/java_uno_accessbridge.jar
oo@ultra:~/BuildDir/ooo_11beta2_src> 

and could be successfully unpacked:

Archive: 
/oo/BuildDir/ooo_11beta2_src/solver/644/unxsols4.pro/bin/java_uno_accessbridge.jar
  inflating: META-INF/MANIFEST.MF    
   creating: org/openoffice/accessibility/
   creating: org/openoffice/accessibility/internal/
  inflating:
org/openoffice/accessibility/internal/AccessibleContextInfo.class  
  inflating:
org/openoffice/accessibility/internal/AccessibleImageInfo.class  
  inflating: org/openoffice/accessibility/internal/AccessibleState.class  

@Pavel: extract the regcomp command from your makefile using full path
(=> which regcomp) and run it solely: see if it crashes the same way.
 Then run the same command within the dbx debugger (just prepend dbx).
 I assume you have a dbx coming with your CC.  When it crashes, do a
"where" in dbx.  That hopefully dumps stack.

RE: @sparcmoz: Please verify whether you have a libgcc3_uno.so in your
LD_LIBRARY_PATH - i can not find any file like libgcc3_uno* anywhere
(where should it be build?)  

@sparcmoz: libgcc3_uno.so is build in module bridges

source files exist for gcc3_linux_sparc but this is not named in the
bridges/prj/build.lst (see attachment). 

Created attachment 6491 [details]
bridges build.lst missing gcc3_linux_sparc

Created attachment 6492 [details]
gcc3_linux_sparc patch to build bridges (and thus setup2)

please note this patch did allow me to build bridges and setup2 - but
i have not build everything yet so can not be sure, (using gcc3 linux
sparc)

dbx output attached - the full dump from script(1).

Created attachment 6504 [details]
The log from dbx session

hmm, seems that throwing exceptions does not work.  Nevertheless this
exception ought not to be thrown.  Meaning that maybe some types from
the rdb files given with regcomp are missing:

1. please check whether all given rdb files for regcomp are existing

2. please log which missing type causes the exception (that cannot be
thrown...): please change bridges/source/jni_uno/jni_info.cxx line 96
the following:

        OUString msg( buf.makeStringAndClear() );
        OString cstr( OUStringToOString( msg,
RTL_TEXTENCODING_ASCII_US ) );
        fprintf( stderr, cstr.getStr() );
        throw BridgeRuntimeError( msg );

committed bridges/prj/build.lst on ooo11beta2 to include
gcc3_linux_sparc directory.

regcomp.rdb is there:

-rw-r--r--   1 oo       staff     688128 čen  3 19:53
../unxsols4.pro/bin/regcomp.rdb

I changed that function to:

JNI_type_info::JNI_type_info(
    JNI_context const & jni, ::com::sun::star::uno::TypeDescription
const & td )
    : m_base( 0 ),
      m_td( td ), 
      m_class( 0 )
{
    fprintf(stderr, "PJ: entering JNI_type_info::JNI_type_info\n");
    m_td.makeComplete();
    if (! m_td.get()->bComplete)
    {
        fprintf(stderr, "PJ: inside if(! m_td.get()->bComplete)\n");
        OUStringBuffer buf( 128 );
        buf.appendAscii(
            RTL_CONSTASCII_STRINGPARAM("cannot make type incomplete: ") );
        buf.append(
            *reinterpret_cast< OUString const * >(
&m_td.get()->pTypeName ) );
        buf.append( jni.get_stack_trace() );
        OUString msg( buf.makeStringAndClear() );
        OString cstr( OUStringToOString( msg,
RTL_TEXTENCODING_ASCII_US ) );
        fprintf(stderr, "PJ: Here!\n");  
        fprintf( stderr, cstr.getStr() );
        fprintf(stderr, "PJ: Here!\n"); 
        throw BridgeRuntimeError( msg );
    }
    fprintf(stderr, "PJ: leaving JNI_type_info::JNI_type_info\n");
}


ie. fprintf when entering, when leaving and also inside if. The result:

PJ: entering JNI_type_info::JNI_type_info
PJ: leaving JNI_type_info::JNI_type_info
PJ: entering JNI_type_info::JNI_type_info
t@1 (l@1) signal SEGV (no mapping at the fault address) in
jni_uno::JNI_type_info::JNI_type_info at line 105 in file "jni_info.cxx"
  105   }

I use:

CC: Sun WorkShop 6 update 2 C++ 5.3 Patch 111685-10 2002/09/16
cc: Sun WorkShop 6 update 2 C 5.3 2001/05/15

java version "1.4.1_02"
Java(TM) 2 Runtime Environment, Standard Edition (build 1.4.1_02-b06)
Java HotSpot(TM) Client VM (build 1.4.1_02-b06, mixed mode)

It is quite strange but maybe you know...

#i9598 is quite similar to this.

Martin, does this tells you something?

@Pavel: try another trace right after the m_td.makeComplete(),
m_td.makeComplete();
fprintf(stderr, "m_td.get() returns %p\n", m_td.get());
OString cstr( OUStringToOString( m_td.get()->pTypeName,
RTL_TEXTENCODING_ASCII_US ) );
if (!...

additionally build libcppu.so with debug and see whether it crashes in
typelib_typedescription_complete()

OK, I added that printf and recompiled cppu project with debug=true.
The result is:

Reading libstlport_sunpro.so
Reading libm.so.1
Reading shlibloader.uno.so
Reading simplereg.uno.so
Reading nestedreg.uno.so
Reading typemgr.uno.so
Reading implreg.uno.so
Reading security.uno.so
Reading libsalhelperC52.so.3.1.0
Reading libreg.so.3.1.0
Reading libstore.so.3.1.0
Reading regtypeprov.uno.so
Reading javaloader.uno.so
Reading libjvmaccessC52.so.3.1.0
Reading libsocket.so.1
Reading libnsl.so.1
Reading libmp.so.2
Reading javavm.uno.so
Reading libjvm.so
Reading libhpi.so
Reading libverify.so
Reading libjava.so
Reading libzip.so
Reading libjava_uno.so
Reading libsunpro5_uno.so
Trace Message: > inserting new mapping: ;uno[1ca8f8];sunpro5[1ca868]
PJ: entering JNI_type_info::JNI_type_info
PJ: m_td.get() returns 333c8
PJ: leaving JNI_type_info::JNI_type_info
Trace Message: > inserting new mapping: ;java[1ca1b8];uno[1ca8f8]
Trace Message: > inserting new mapping: ;java[1ca1b8];sunpro5[1ca868]
PJ: entering JNI_type_info::JNI_type_info
PJ: m_td.get() returns fe5859a8
PJ: leaving JNI_type_info::JNI_type_info
t@1 (l@1) signal BUS (invalid address alignment) in
jni_uno::JNI_info::get_type_info at line 459 in file "jni_info.cxx"
  459   }
(/opt/SUNWspro/bin/../WS6U2/bin/sparcv9/dbx) where
current thread: t@1
=>[1] jni_uno::JNI_info::get_type_info(this = ???, jni = CLASS, td =
???) (optimized), at 0xfd7b3d84 (line ~459) in "jni_info.cxx"
  [2] jni_uno::JNI_interface_type_info::JNI_interface_type_info(this =
???, jni = CLASS, td_ = CLASS) (optimized), at 0xfd7b2b34 (line ~280)
in "jni_info.cxx"
  [3] jni_uno::JNI_info::create_type_info(this = ???, jni = CLASS, td
= ???) (optimized), at 0xfd7b3b48 (line ~432) in "jni_info.cxx"
  [4] jni_uno::JNI_info::get_type_info(this = ???, jni = CLASS, td =
???) (optimized), at 0xfd7b3e34 (line ~459) in "jni_info.cxx"
  [5] Mapping_map_to_uno(mapping = ???, ppOut = ???, pIn = ???, td =
???) (optimized), at 0xfd7c4d64 (line ~157) in "jni_bridge.cxx"
  [6] mediate_mapInterface(pMapping = 0x1c9e78, ppOut = 0xffbede68,
pInterface = 0xcea5c, pInterfaceTypeDescr = 0x330c8), line 308 in
"lbmap.cxx"
  [7] stoc_javaloader::JavaComponentLoader::JavaComponentLoader(this =
???, xCtx = CLASS) (optimized), at 0xfe1a2ae0 (line ~306) in
"javaloader.cxx"
  [8] stoc_javaloader::JavaComponentLoader_CreateInstance(xCtx =
CLASS) (optimized), at 0xfe1a37b8 (line ~401) in "javaloader.cxx"
  [9] cppu::OSingleFactoryHelper::createInstanceEveryTime(this = ???,
xContext = CLASS) (optimized), at 0xff106328 (line ~233) in "factory.cxx"
  [10] cppu::OSingleFactoryHelper::createInstanceWithContext(this =
???, xContext = CLASS) (optimized), at 0xff106464 (line ~258) in
"factory.cxx"
  [11] cppu::OFactoryComponentHelper::createInstanceWithContext(this =
???, xContext = CLASS) (optimized), at 0xff106f78 (line ~535) in
"factory.cxx"
  [12] cppu::ORegistryFactoryHelper::createInstanceEveryTime(this =
???, xContext = CLASS) (optimized), at 0xff1074f8 (line ~668) in
"factory.cxx"
  [13] cppu::OSingleFactoryHelper::createInstanceWithContext(this =
???, xContext = CLASS) (optimized), at 0xff106464 (line ~258) in
"factory.cxx"
  [14] cppu::OFactoryComponentHelper::createInstanceWithContext(this =
???, xContext = CLASS) (optimized), at 0xff106f78 (line ~535) in
"factory.cxx"
  [15] stoc_smgr::OServiceManager::createInstanceWithContext(this =
???, rServiceSpecifier = CLASS, xContext = CLASS) (optimized), at
0xff09a874 (line ~1337) in "servicemanager.cxx"
  [16]
stoc_impreg::ImplementationRegistration::registerImplementation(this =
???, implementationLoaderUrl = CLASS, locationUrl = CLASS, xReg =
CLASS) (optimized), at 0xfe4aa8d8 (line ~1648) in "implreg.cxx"
  [17] DoIt::operator()(this = ???, url = CLASS) (optimized), at
0x14a2c (line ~611) in "registercomponent.cxx"
  [18] _STL::for_each<rtl::OUString*,DoIt>(__first = ???, __last =
???, __f = STRUCT) (optimized), at 0x18008 (line ~108) in "_algo.c"
  [19] main(argc = ???, argv = ???) (optimized), at 0x16158 (line
~869) in "registercomponent.cxx"

Thank you for your patience :-)

When cppu is compiled as normally, ie. without debug, it will go further :

                regcomp -br ../unxsols4.pro/bin/regcomp.rdb          
          \
                        -register                                       \
                        -r ../unxsols4.pro/bin/setup_services.tmp    
          \
                        -l com.sun.star.loader.Java2                    \
                       
-env:UNO_JAVA_COMPPATH=file:///oo/BuildDir/ooo_11beta2_src/solver/644/unxsols4.pro/bin
 \
                        -c
vnd.sun.star.expand:\$UNO_JAVA_COMPPATH/java_uno_accessbridge.jar
PJ: entering JNI_type_info::JNI_type_info
PJ: m_td.get() returns 333c8
PJ: leaving JNI_type_info::JNI_type_info
PJ: entering JNI_type_info::JNI_type_info
PJ: m_td.get() returns 0

An unexpected exception has been detected in native code outside the VM.
Unexpected Signal : 11 occurred at PC=0xF9B928A0
Function=[Unknown. Nearest: UDK_3.1+0x28A0]
Library=/oo/BuildDir/ooo_11beta2_src/solver/644/unxsols4.pro/lib/libjava_uno.so



I'll be online for about 8 hours from now, so I'm ready to try almost
everything you'll tell me.

@Pavel: very strange, but give it some more...

try to change the following (appr. line 384)

    case typelib_TypeClass_INTERFACE:
        new_info = new JNI_interface_type_info(
                jni,
                css::uno::TypeDescription(td) );
        break;

additionally, try to print out the type name as first line of
JNI_type_info::JNI_type_info()
{
OString cstr(OUStringToOString(
            td.get()->pTypeName,
            RTL_TEXTENCODING_ASCII_US ) );
fprintf(stderr, "td name=%s", cstr.getStr());
cstr = OUStringToOString(
            m_td.get()->pTypeName,
            RTL_TEXTENCODING_ASCII_US ) );
fprintf(stderr, "m_td name=%s", cstr.getStr());


...
}

After this change, it finished:

                regcomp -br ../unxsols4.pro/bin/regcomp.rdb          
          \
                        -register                                       \
                        -r ../unxsols4.pro/bin/setup_services.tmp    
          \
                        -l com.sun.star.loader.Java2                    \
                       
-env:UNO_JAVA_COMPPATH=file:///oo/BuildDir/ooo_11beta2_src/solver/644/unxsols4.pro/bin
 \
                        -c
vnd.sun.star.expand:\$UNO_JAVA_COMPPATH/java_uno_accessbridge.jar
td name=com.sun.star.uno.XInterfacem_td
name=com.sun.star.uno.XInterfacePJ: entering JNI_type_info::JNI_type_info
PJ: m_td.get() returns 333c8
PJ: leaving JNI_type_info::JNI_type_info
td name=com.sun.star.loader.XImplementationLoaderm_td
name=com.sun.star.loader.XImplementationLoaderPJ: entering
JNI_type_info::JNI_type_info
PJ: m_td.get() returns 330c8
PJ: leaving JNI_type_info::JNI_type_info
td name=com.sun.star.lang.XInitializationm_td
name=com.sun.star.lang.XInitializationPJ: entering
JNI_type_info::JNI_type_info
PJ: m_td.get() returns 34fd0
PJ: leaving JNI_type_info::JNI_type_info
td name=com.sun.star.lang.XMultiComponentFactorym_td
name=com.sun.star.lang.XMultiComponentFactoryPJ: entering
JNI_type_info::JNI_type_info
PJ: m_td.get() returns 34cf8
PJ: leaving JNI_type_info::JNI_type_info
td name=com.sun.star.lang.XMultiServiceFactorym_td
name=com.sun.star.lang.XMultiServiceFactoryPJ: entering
JNI_type_info::JNI_type_info
PJ: m_td.get() returns 350a0
PJ: leaving JNI_type_info::JNI_type_info
td name=com.sun.star.registry.XRegistryKeym_td
name=com.sun.star.registry.XRegistryKeyPJ: entering
JNI_type_info::JNI_type_info
PJ: m_td.get() returns 33140
PJ: leaving JNI_type_info::JNI_type_info
td name=com.sun.star.beans.XPropertySetm_td
name=com.sun.star.beans.XPropertySetPJ: entering
JNI_type_info::JNI_type_info
PJ: m_td.get() returns 34d60
PJ: leaving JNI_type_info::JNI_type_info
td name=com.sun.star.uno.XComponentContextm_td
name=com.sun.star.uno.XComponentContextPJ: entering
JNI_type_info::JNI_type_info
PJ: m_td.get() returns 331e8
PJ: leaving JNI_type_info::JNI_type_info
td name=com.sun.star.util.XMacroExpanderm_td
name=com.sun.star.util.XMacroExpanderPJ: entering
JNI_type_info::JNI_type_info
PJ: m_td.get() returns 1da838
PJ: leaving JNI_type_info::JNI_type_info
register component
'vnd.sun.star.expand:$UNO_JAVA_COMPPATH/java_uno_accessbridge.jar' in
registry '../unxsols4.pro/bin/setup_services.tmp' succesful!
mv ../unxsols4.pro/bin/setup_services.tmp
../unxsols4.pro/bin/setup_services.rdb
mkdir -p ../unxsols4.pro/bin/remote
rm -f ../unxsols4.pro/bin/remote/setup_services.rdb
regcomp -register -r ../unxsols4.pro/bin/remote/setup_services.rdb -c
"libi18n644ss.so libi18npool644ss.so libmcnttype.so libdtransX11644ss.so"
register component 'libi18n644ss.so' in registry
'../unxsols4.pro/bin/remote/setup_services.rdb' succesful!
register component 'libi18npool644ss.so' in registry
'../unxsols4.pro/bin/remote/setup_services.rdb' succesful!
register component 'libmcnttype.so' in registry
'../unxsols4.pro/bin/remote/setup_services.rdb' succesful!
register component 'libdtransX11644ss.so' in registry
'../unxsols4.pro/bin/remote/setup_services.rdb' succesful!


So the cause is reinterpret_cast?

@Pavel: Yes, I think so... maybe this occurs due to different
optimization settings or different versions of CC.  You see, Sander's
sparc builds don't have this problem.  I will review the code
corresponding to this.  Nevertheless I think these fixes won't go into
1.1.

@Martin: what is your opinion?

fixed in OOo 2.0

No problem for me, I can apply that fix to my builds, but maybe other
builders could reach it too.

Closing.

confirmed as P2 and thus it should go into 1.1.1.

@Pavel: can you please verify the fix I made in jni_info.cxx on cws
ooo111fix1?  thank you.

This did not fixed it. It is somewhat shorter than the patch you sent
me to test. Do you remeber those brackets around the code? I do not
see them in 

===================================================================
File: jni_info.cxx      Status: Up-to-date

   Working revision:    1.8.6.3
   Repository revision: 1.8.6.3
/cvs/udk/bridges/source/jni_uno/jni_info.cxx,v
   Sticky Tag:          cws_srx645_ooo111fix1 (branch: 1.8.6)
   Sticky Date:         (none)
   Sticky Options:      (none)



#
# The exception above was detected in native code outside the VM
#
# Java VM: Java HotSpot(TM) Client VM (1.4.1_02-b06 mixed mode)
#
# An error report file has been saved as hs_err_pid25546.log.
# Please refer to the file for further information.
#
Abort (core dumped)
dmake:  Error code 134, while making
'../unxsols4.pro/bin/setup_services.rdb'
---* TG_SLO.MK *---

@Pavel: please give jni_info.cxx another try.  Works for you?

I fetched the latest version on fix1 (1.8.6.4) and recompiled and
delivered bridges. Then build --from setup2 in setup finished with:

setenv CLASSPATH
/ultra/BuildDir/ooo_1.1.0_src/solver/645/unxsols4.pro/bin/unoil.jar:/ultra/BuildDir/ooo_1.1.0_src/solver/645/unxsols4.pro/bin/java_uno.jar:/ultra/BuildDir/ooo_1.1.0_src/solver/645/unxsols4.pro/bin/ridl.jar:/ultra/BuildDir/ooo_1.1.0_src/solver/645/unxsols4.pro/bin/sandbox.jar:/ultra/BuildDir/ooo_1.1.0_src/solver/645/unxsols4.pro/bin/jurt.jar:/ultra/BuildDir/ooo_1.1.0_src/solver/645/unxsols4.pro/lib:/usr/java/jre/lib/rt.jar:.
&&         setenv LD_LIBRARY_PATH
.:/usr/dt/lib:/usr/openwin/lib:/usr/local/lib:../lib:/usr/java/jre/lib/sparc:/usr/java/jre/lib/sparc/motif21:/usr/java/jre/lib/sparc/native_threads:/ultra/BuildDir/ooo_1.1.0_src/solver/645/unxsols4.pro/lib:
&& \
                regcomp -br ../unxsols4.pro/bin/regcomp.rdb          
          \
                        -register                                       \
                        -r ../unxsols4.pro/bin/setup_services.tmp    
          \
                        -l com.sun.star.loader.Java2                    \
                       
-env:UNO_JAVA_COMPPATH=file:///ultra/BuildDir/ooo_1.1.0_src/solver/645/unxsols4.pro/bin
\
                        -c
vnd.sun.star.expand:\$UNO_JAVA_COMPPATH/java_uno_accessbridge.jar

An unexpected exception has been detected in native code outside the VM.
Unexpected Signal : 11 occurred at PC=0xFE61AD18
Function=[Unknown. Nearest: osl_decrementInterlockedCount+0xC0]
Library=/ultra/BuildDir/ooo_1.1.0_src/solver/645/unxsols4.pro/lib/libsal.so.3.1.0

Current Java thread:

Dynamic libraries:
0x10000         regcomp
0xff370000      /usr/lib/libpthread.so.1
0xff330000      /usr/lib/libthread.so.1
0xff390000      /usr/lib/libdl.so.1
0xff200000      /usr/lib/libc.so.1
0xff310000      /usr/platform/SUNW,Ultra-5_10/lib/libc_psr.so.1
0xff1d0000      /usr/lib/libCrun.so.1
0xff2f0000      /usr/lib/libw.so.1
0xfe600000     
/ultra/BuildDir/ooo_1.1.0_src/solver/645/unxsols4.pro/lib/libsal.so.3
0xff180000      /usr/lib/librt.so.1
0xff160000      /usr/lib/libaio.so.1
0xff140000      /usr/lib/locale/cs_CZ.ISO8859-2/cs_CZ.ISO8859-2.so.2
0xff0d0000     
/ultra/BuildDir/ooo_1.1.0_src/solver/645/unxsols4.pro/lib/libcppuhelperC52.so.3
0xff080000     
/ultra/BuildDir/ooo_1.1.0_src/solver/645/unxsols4.pro/lib/servicemgr.uno.so
0xff030000     
/ultra/BuildDir/ooo_1.1.0_src/solver/645/unxsols4.pro/lib/libcppu.so.3
0xfe480000     
/ultra/BuildDir/ooo_1.1.0_src/solver/645/unxsols4.pro/lib/libstlport_sunpro.so
0xfe5d0000      /usr/lib/libm.so.1
0xfe7e0000     
/ultra/BuildDir/ooo_1.1.0_src/solver/645/unxsols4.pro/lib/shlibloader.uno.so
0xfe5a0000     
/ultra/BuildDir/ooo_1.1.0_src/solver/645/unxsols4.pro/lib/simplereg.uno.so
0xfe450000     
/ultra/BuildDir/ooo_1.1.0_src/solver/645/unxsols4.pro/lib/nestedreg.uno.so
0xfe420000     
/ultra/BuildDir/ooo_1.1.0_src/solver/645/unxsols4.pro/lib/typemgr.uno.so
0xfe3f0000     
/ultra/BuildDir/ooo_1.1.0_src/solver/645/unxsols4.pro/lib/implreg.uno.so
0xfe3b0000     
/ultra/BuildDir/ooo_1.1.0_src/solver/645/unxsols4.pro/lib/security.uno.so
0xfe390000     
/ultra/BuildDir/ooo_1.1.0_src/solver/645/unxsols4.pro/lib/libsalhelperC52.so.3
0xfe360000     
/ultra/BuildDir/ooo_1.1.0_src/solver/645/unxsols4.pro/lib/libreg.so.3
0xfe320000     
/ultra/BuildDir/ooo_1.1.0_src/solver/645/unxsols4.pro/lib/libstore.so.3
0xfe2e0000     
/ultra/BuildDir/ooo_1.1.0_src/solver/645/unxsols4.pro/lib/regtypeprov.uno.so
0xfe2c0000     
/ultra/BuildDir/ooo_1.1.0_src/solver/645/unxsols4.pro/lib/javaloader.uno.so
0xfe1d0000     
/ultra/BuildDir/ooo_1.1.0_src/solver/645/unxsols4.pro/lib/libjvmaccessC52.so.3
0xfe1b0000      /usr/lib/libsocket.so.1
0xfe100000      /usr/lib/libnsl.so.1
0xfe0e0000      /usr/lib/libmp.so.2
0xfe080000     
/ultra/BuildDir/ooo_1.1.0_src/solver/645/unxsols4.pro/lib/javavm.uno.so
0xfdc00000      /usr/java/jre/lib/sparc/libjvm.so
0xfdbd0000      /usr/j2se/jre/lib/sparc/native_threads/libhpi.so
0xfdba0000      /usr/j2se/jre/lib/sparc/libverify.so
0xfdb60000      /usr/j2se/jre/lib/sparc/libjava.so
0xfdb40000      /usr/j2se/jre/lib/sparc/libzip.so
0xf9b90000     
/ultra/BuildDir/ooo_1.1.0_src/solver/645/unxsols4.pro/lib/libjava_uno.so
0xf9ae0000     
/ultra/BuildDir/ooo_1.1.0_src/solver/645/unxsols4.pro/lib/libsunpro5_uno.so

Local Time = Thu Nov 13 12:37:43 2003
Elapsed Time = 2
#
# The exception above was detected in native code outside the VM
#
# Java VM: Java HotSpot(TM) Client VM (1.4.1_02-b06 mixed mode)
#
# An error report file has been saved as hs_err_pid273.log.
# Please refer to the file for further information.
#
Abort (core dumped)
dmake:  Error code 134, while making
'../unxsols4.pro/bin/setup_services.rdb'
---* TG_SLO.MK *---

ERROR: Error 65280 occurred while making
/ultra/BuildDir/ooo_1.1.0_src/setup2/util
oo@ultra:/ultra/BuildDir/ooo_1.1.0_src/setup2> 


Do I need to recompile more than just bridges?

@Pavel: no, bridges ought to be sufficient.  I am wondering, because
now you have exactly the version that was running previously (i.e. the
one with the braces)... Please have another look at the source and
assure that the setup zip has all files updated.

OK, I did some experiments with jni_info.cxx. I pul all files to

http://tmp.janik.cz/jni_uno/jni_info.cxx.1.1.0

- original file with tag OpenOffice_1_1

http://tmp.janik.cz/jni_uno/jni_info.cxx.1.1.0.patched

- the same file patched by
ftp://ftp.linux.cz/pub/localization/OpenOffice.org/devel/1.1.0/build-6/build/Patches/OOo_1.1.0_source-SPARC-only-fix-regcomp-coredump.diff

http://tmp.janik.cz/jni_uno/jni_info.cxx.fix1

- current fix1 version

The only working version is *patched. Rebuilding bridges is really
enough :-)

All in all, this is strange ;-)

Maybe we met another CC bug?

BTW - shouldn't closing } be just before the break or is it OK to have
break inside the block?

@Pavel: have another try, I ifdef'ed for SOLARIS.  No, break inside
braces is right.

@Pavel: please assure you are using a similar version that Sander
uses.  I suspect that your compiler misses some patches... so this
workaround has been necessary.  Makes more sense to fix the compiler
bug / get the patch.  There can be more (not so obvious) code in OOo.

It is OK right now in fix1.

My CC is:

CC: Sun WorkShop 6 update 2 C++ 5.3 Patch 111685-10 2002/09/16

I'm trying to patch it only when there are issues with it. This patch
was the last one the day I reported this issus. Maybe there is a new
version right now.

Thank you for the help. No back to FreeBSD issues ;-)

dbo->pjanik: please verify, then close when reintegrated to master branch.

Of course. Thanks!

Verified in fix2. Thanks!

close issue.