[Open Office] regcomp crashes during make install

[Weichuan Lin]

hi, Bernhard.
which platform do you use? on 64 bit system, the registering of java
component still has some weik prolem.

On 2/19/06, Bernhard Rosenkraenzer <bero at arklinux.org> wrote:
> Any ideas on this one?
>
> ... creating preregistered services.rdb ...
> sh: line 1: 13225 Segmentation
> fault      /usr/src/ark/BUILD/ooo-build/build/oob680-m1/solver/680/unxlngi6.pro/bin/regcomp
> -register
> -br /usr/src/ark/BUILD/ooo-build/build/oob680-m1/instsetoo_native/util/OpenOffice//services.rdb/3a50f2e5f55c2428360ca4db9042f066_inprogress_3/regcomp.rdb
> -r /usr/src/ark/BUILD/ooo-build/build/oob680-m1/instsetoo_native/util/OpenOffice//services.rdb/3a50f2e5f55c2428360ca4db9042f066_inprogress_3/services.rdb
> -c 'vnd.sun.star.expand:
> $UNO_JAVA_COMPONENT_PATH/java_uno_accessbridge.jar;vnd.sun.star.expand:
> $UNO_JAVA_COMPONENT_PATH/report.jar;vnd.sun.star.expand:
> $UNO_JAVA_COMPONENT_PATH/table.jar;vnd.sun.star.expand:
> $UNO_JAVA_COMPONENT_PATH/letter.jar;vnd.sun.star.expand:
> $UNO_JAVA_COMPONENT_PATH/form.jar;vnd.sun.star.expand:
> $UNO_JAVA_COMPONENT_PATH/fax.jar;vnd.sun.star.expand:
> $UNO_JAVA_COMPONENT_PATH/agenda.jar;vnd.sun.star.expand:
> $UNO_JAVA_COMPONENT_PATH/query.jar;vnd.sun.star.expand:
> $UNO_JAVA_COMPONENT_PATH/web.jar;vnd.sun.star.expand:
> $UNO_JAVA_COMPONENT_PATH/ScriptFramework.jar;vnd.sun.star.expand:
> $UNO_JAVA_COMPONENT_PATH/ScriptProviderForBeanShell.jar;vnd.sun.star.expand:
> $UNO_JAVA_COMPONENT_PATH/ScriptProviderForJava.jar;vnd.sun.star.expand:
> $UNO_JAVA_COMPONENT_PATH/ScriptProviderForJavaScript.jar;vnd.sun.star.expand:
> $UNO_JAVA_COMPONENT_PATH/XSLTFilter.jar;vnd.sun.star.expand:
> $UNO_JAVA_COMPONENT_PATH/XSLTValidate.jar' -l com.sun.star.loader.Java2
> -env:UNO_JAVA_COMPONENT_PATH='file:///usr/src/ark/BUILD/ooo-build/build/oob680-m1/solver/680/unxlngi6.pro/bin'
> -env:URE_INTERNAL_JAVA_DIR='file:///usr/src/ark/BUILD/ooo-build/build/oob680-m1/solver/680/unxlngi6.pro/bin'
> 2>&1
> ... cleaning the output tree ...
>
>
> The gdb backtrace doesn't look very useful...
>
> #0  0x4202876f in sigprocmask () from /lib/libc.so.6
> #1  0x3be33e66 in _Jv_SetInitialHeapSize ()
>    from /usr/lib/libgcj.so.7
> #2  0x3be347a0 in _Jv_ThrowBadArrayIndex ()
>    from /usr/lib/libgcj.so.7
> #3  <signal handler called>
> #4  0x3c4703b2 in GC_local_gcj_malloc () from /usr/lib/libgcj.so.7
> #5  0x3be34570 in _Jv_AllocObjectNoFinalizer ()
>    from /usr/lib/libgcj.so.7
> #6  0x3be343f6 in _Jv_AllocObject () from /usr/lib/libgcj.so.7
> #7  0x3be347b9 in _Jv_ThrowBadArrayIndex ()
>    from /usr/lib/libgcj.so.7
> #8  <signal handler called>
> (at this point, everything starting from #4 to #8 repeat over and over again -
> apparently somethis is trying to catch a SIGSEGV)
>
> Unfortunately I can't claim I fully understand OOo's component system
> (especially not the Java stuff that seems to be the root of all evil these
> days)
>
> Best regards,
> bero
> _______________________________________________
> Openoffice mailing list
> Openoffice at lists.ximian.com
> http://lists.ximian.com/mailman/listinfo/openoffice
>


--
Best Regards
Weichuan