Re: LibreOffice for FreeBSD on PowerPC

On 11/18/2016 05:26, Michael Stahl wrote:
> On 17.11.2016 17:54, Curtis Hamilton wrote:
> > This might be a lost cause and I’m sure this subject has been previously
> > addresses, however, I’d like to know if any has successfully ported
> > LibreOffice to FreeBSD on PowerPC.
> >
> > I’ve been able to successfully build version 4.3.7, from the FreeBSD
> > ports, using the Linux PowerPC code as a guide.  However, the resulting
> > build does not display any icons in the menus or toolbars.
> well... 4.3 is really ancient and EOL so i guess investigating this
> isn't going to be very productive, i suggest first trying master or 5.2,
> it's quite possible that the problem is gone but you get a different one
> instead :)
realize this is an old version, i've tried building 5.2 which looks for 
firebird, which does not build on freebsd powerpc64. appears to be 
endian and/or 32-bit related. 4.3 requires less patching than 5.2.

was checking to see if anyone had already done work with this code 
base.  was not looking to for any time investment, other that pointers, 
which you provided :)

i'm taking another shot at 5.2 from the freebsd ports.  have made an 
initial set of patches, including disabling firebird support. will 
provide feedback on results.
> My build environment:
>
> OS: FreeBSD 10.2-STABLE powerpc (64-bit)
> Compiler: GCC version 4.8.5
> Java Version:  OpenJDK Runtime Environment (build 1.8.0_102-b14)
>                       OpenJDK 64-Bit Server VM (build 25.102-b14, mixed mode)
>
> I’m hoping that someone with knowledge of the workings of the user
> interface can provide some insight on what to look at.  I’m not sure if
> this is a LibreOffice issue or FreeBSD on PowerPC issue.
> it could be either, in particular if it's big-endian PPC there might be
> an endian issue in some image handling code, of which there is quite a
> lot for various different formats.
that was my initial assumption, but could not figure out how 
OSL_BIGENDIAN was being defined, since "osl/endian.h" is not being 
included in all files that tests for OSL_BIGENDIAN.  trying to define it 
globally, fixes some test errors, but causes additional errors.