Hi,

This problem was corrected in the CVS-HEAD, however for the older
versions (including 4.8.4) it should be added manually for java.

Best Regards,

Tamas


2006/7/18, Mario Basa <[EMAIL PROTECTED]>:
Hello,

Yes, this works for me. Adding the lines below to the
mapscript/java/Makefile.in stopped the SIGSEGV whenever I create a
shapeObj class:

# Optional GEOS Support.
# See http://geos.refractions.net/
GEOS=     @GEOS_ENABLED@
GEOS_LIB=  @GEOS_LIB@
GEOS_INC=  @GEOS_INC@

CCFLAGS= @CFLAGS@ @DEBUG_FLAGS@ \
        $(IGNORE_MISSING_DATA) $(STRINGS) $(EPPL) $(PROJ) $(OWS) $(MING) \
        $(ORACLESPATIAL) $(TIFF) $(JPEG) $(GD) $(PDF) $(SDE) $(OGR) $(GDAL) \
        $(POSTGIS) $(MYGIS) $(MPATROL) $(THREAD) $(GD_INC) $(PDF_INC) \
        $(PROJ_INC) $(TIFF_INC) $(JPEG_INC) $(EGIS_INC) $(SDE_INC) \
        $(REGEX_INC) $(MPATROL_INC) $(GDAL_INC) $(POSTGIS_INC) $(MYGIS_INC) \
        $(CURL_INC) $(MING_INC) $(ORACLESPATIAL_INC) $(GEOS)
                                                     ^^^^^^^
# Link flags and shared libs only
SUP_LIBS =  $(GD_LIB) $(PDF_LIB) $(TIFF_LIB) $(PROJ_LIB) \
          $(JPEG_LIB) $(SDE_LIB) $(GDAL_LIB) $(MING_LIB) $(POSTGIS_LIB) \
          $(MYGIS_LIB) $(CURL_LIB) $(ORACLESPATIAL_LIB) \
          $(MPATROL_LIB) $(THREAD_LIB) $(XTRALIBS) $(GEOS_LIB)
                                                   ^^^^^^^^^^^



Thanks.

mario.


On 6/24/06, Tamas Szekeres <[EMAIL PROTECTED]> wrote:
> Hi,
>
> Enabling GEOS support for Makefile.in should fix this problem.
> Otherwise the shapeObj size is different for mapscript and mapserver
> causing heap corruption.
>
> Tamas
>
>
>
> 2006/6/12, Rodriguez, Heraldo <[EMAIL PROTECTED]>:
> > Hi guys,
> >          I have started  developing  with java mapscript and I have the same
> > problems with the shapeObj class.
> >      Every time I want to create a new shape I get the message
> > > #
> > > # An unexpected error has been detected by HotSpot Virtual Machine:
> > > #
> > > #  SIGSEGV (0xb) at pc=0x00002aaaaae43d5b, pid=23120, tid=46912501734144
> > > #
> > > # Java VM: Java HotSpot(TM) Client VM (1.4.2_11-b06 mixed mode)
> > > # Problematic frame:
> > ...
> >
> > I am using mapserver  4.8.1  with java 1.4.2_11 version under Debian 3.1.
> >
> > Mapserver Compiling options
> > ./configure --with-freetype=/usr/local/bin/freetype-config
> > --with-zlib=/usr/lib --with-png=/usr/local/lib --with-jpeg=/usr/lib
> > --with-libiconv=/usr/local/lib --with-gd=/usr/local/bin/gdlib-config
> > --with-pdf=/usr/local/lib --with-proj=/usr/local/lib --with-threads
> > --with-geos=/usr/local/bin/geos-config --with-ogr=/usr/local/bin/gdal-config
> > --with-gdal=/usr/local/bin/gdal-config
> > --with-postgis=/usr/local/pgsql/bin/pg_config --with-wfs --with-wcs
> > --with-wmsclient --with-curl-config=/usr/local/bin/curl-config
> >
> > Steps to compile java mapscript
> >          Run make
> >        Make interface (swig version 1.3.29)
> >        Make all
> >
> > I looked for  a solution in google and mapserver documentation without
> > success.   Could you please give me  a clue to solve this? Thanks in
> > advance.
> >   Best regards.
> >
> >
> >
> > Heraldo G. Rodriguez
> >   [EMAIL PROTECTED]
> >    TE 4326-4002 int 255
> >  Dpto. Desarrollo
> >   Sistemas Catastrales S.A
> >
> >
> >
> > -----Mensaje original-----
> > De: Andre Sachs [mailto:[EMAIL PROTECTED]
> > Enviado el: miƩrcoles, 31 de mayo de 2006 5:16
> > Para: [email protected]
> > Asunto: Re: [UMN_MAPSERVER-USERS] JavaMapscript SIGSEGV
> >
> > On 31 May 2006, at 8:37 AM, Umberto Nicoletti wrote:
> >
> > > Andre, could you report whether the patch attached to gentoo bugzilla
> > > works or not?
> >
> > Hi Umberto,
> >
> > Sorry for not replying earlier.
> >
> > The patch works in a 32 intel environment, thanks !
> >
> > Due to time constraints on my current project I ditched my AMD 64,
> > Java 1.5 64 bit environment in favour of Intel 32 bit and this has
> > resolved my problem.
> >
> > The problems I experienced are somewhere in the application stack,
> > either in :
> > - Sun's 64 bit 1.5 jdk
> > - Gentoo's glibc 2.3.5-r2
> > - Linux kernel 2.6.16
> >
> > Once my current project is over (+- 3 weeks) I will be have some
> > spare time to hunt down the culprit.
> >
> > Regards,
> >     Andre
> > --
> > Andre Sachs
> > IT Architect, Clue Technologies PTY (LTD)
> > email: [EMAIL PROTECTED]
> >
>

Reply via email to