Re: Compiling LibreOffice 4.1.2.3 on Solaris

2013-11-13 Thread Michael Stahl
On 13/11/13 12:32, Jan Holesovsky wrote: > Eric Bautsch píše v Pá 01. 11. 2013 v 19:57 +: > >> And the output suggests that the versions are OK: >> bautsche@cressida $ pvs solver/unxsogi.pro/bin/idlc >> libm.so.2 (SUNW_1.1); >> libgcc_s.so.1 (GCC_3.0);

Re: Compiling LibreOffice 4.1.2.3 on Solaris

2013-11-13 Thread Eric Bautsch
No, I haven't played with this any further. I'm intending to, but I was going to move the compile over to Solaris 11.1, that being obviously a more stable platform. I have no idea if the lack of brackets or UDK_3_0_0 on its own is significant, but doing a quick pvs on other files (/opt/openoffi

Re: Compiling LibreOffice 4.1.2.3 on Solaris

2013-11-13 Thread Jan Holesovsky
Hi Eric, Eric Bautsch píše v Pá 01. 11. 2013 v 19:57 +: > Yes, the command is pvs: > http://docs.oracle.com/cd/E19253-01/816-5165/pvs-1/ Did you get anywhere, or still stuck, please? :-) > And the output suggests that the versions are OK: > bautsche@cressida $ pvs solver/unxsogi.pro

Re: Compiling LibreOffice 4.1.2.3 on Solaris

2013-11-01 Thread Eric Bautsch
Yes, the command is pvs: http://docs.oracle.com/cd/E19253-01/816-5165/pvs-1/ And the output suggests that the versions are OK: bautsche@cressida $ pvs solver/unxsogi.pro/bin/idlc libm.so.2 (SUNW_1.1); libgcc_s.so.1 (GCC_3.0); libc.so.1 (SUNWprivate_1.1, SUNW_1.1, SUNW_0.9, SUNW_0.7,

Re: Compiling LibreOffice 4.1.2.3 on Solaris

2013-11-01 Thread Stephan Bergmann
On 11/01/2013 03:48 PM, Eric Bautsch wrote: Output from LD_DEBUG=all ldd -r solver/unxsogi.pro/bin/idlc >log 2>&1 attached. Unfortunately doesn't give a clue why the loader refuses to pick the symobls from libuno_sal.so.3. The last idea I have is that there's something wrong with the symbol

Re: Compiling LibreOffice 4.1.2.3 on Solaris

2013-11-01 Thread Stephan Bergmann
On 11/01/2013 03:24 PM, Eric Bautsch wrote: On 01/11/2013 13:55, Stephan Bergmann wrote: Try "ldd -r solver/unxsogi.pro/bin/idlc" to force symbol resolution. bautsche@cressida $ ldd -r solver/unxsogi.pro/bin/idlc [...] libuno_sal.so.3 => /export/home/bautsche/libre-off

Re: Compiling LibreOffice 4.1.2.3 on Solaris

2013-11-01 Thread Stephan Bergmann
On 11/01/2013 02:12 PM, Eric Bautsch wrote: bautsche@cressida $ ldd solver/unxsogi.pro/bin/idlc libnsl.so.1 => /lib/libnsl.so.1 libsocket.so.1 =>/lib/libsocket.so.1 libreglo.so => /export/home/bautsche/libre-office/libreoffice-4.1.2.3/solv

Re: Compiling LibreOffice 4.1.2.3 on Solaris

2013-11-01 Thread Eric Bautsch
Hi Norbert and Stephan. The LD_LIBRARY_PATH looks fine to me: [build IDL] udkapi/com/sun/star/idl mkdir -p /export/home/bautsche/libre-office/libreoffice-4.1.2.3/workdir/unxsogi.pro/UnoApiPartTarget/udkapi/com/sun/star/ && RESPONSEFILE=/tmp/gbuild.LYKziC && SOLARBINDIR=/export/h

Re: Compiling LibreOffice 4.1.2.3 on Solaris

2013-11-01 Thread Stephan Bergmann
On 10/31/2013 01:25 PM, Eric Bautsch wrote: I'm trying to compile LibreOffice 4.1.2.3 on Solaris 12 (build 26, currently). Note: don't get excited, I'm doing this in my spare time and Oracle are in no way endorsing, supporting, or anything else'ing my efforts. I've been using this threa

Re: Compiling LibreOffice 4.1.2.3 on Solaris

2013-10-31 Thread Norbert Thiebaud
On Thu, Oct 31, 2013 at 7:25 AM, Eric Bautsch wrote: > > I think rtl_string_new is defined in include/rtl/string.h, but I can't see > where the function is actually coming from (running a find -exec grep across > the code). see sal/rtl/strtmpl.cxx This is a hacky way to implement both ustring an