[GRASS-dev] Re: [GRASS GIS] #1131: Global LFS for wingrass

GRASS GIS trac at osgeo.org
Sun Aug 15 07:22:33 EDT 2010


#1131: Global LFS for wingrass
--------------------------+-------------------------------------------------
 Reporter:  mmetz         |       Owner:  grass-dev@…              
     Type:  defect        |      Status:  new                      
 Priority:  critical      |   Milestone:  7.0.0                    
Component:  libgis        |     Version:  svn-trunk                
 Keywords:  LFS,wingrass  |    Platform:  MSWindows 7              
      Cpu:  All           |  
--------------------------+-------------------------------------------------

Comment(by mmetz):

 Replying to [comment:8 glynn]:
 > >
 > > a quick compiling test with r43116
 > >
 {{{
 > syringia at NADA /c/osgeo4w/usr/src/grass_trunk/lib/vector/diglib
 > $ make
 >
 > Files OBJ.i686-pc-mingw32/test.tmp and test64.ok differ
 }}}
 >
 The diglib test is passed in r43119
 >
 > 1. The config.h.in changes are conditionalised upon
 {{{
 > #ifdef USE_LARGEFILES
 }}}
 Hmph. We discussed that in May, I forgot, sorry! Fixed in r43119.
 >
 > 2. The config.h.in changes are also conditionalised upon
 > {{{
 > #if defined(__MINGW32__) && defined(_FILE_OFFSET_BITS) &&
 _FILE_OFFSET_BITS == 64
 > }}}
 > The configure checks will determine that _FILE_OFFSET_BITS don't have
 any effect (upon the test cases) on Windows, and so won't define it.

 On Windows, the configure checks do define _FILE_OFFSET_BITS=64, at least
 something defines _FILE_OFFSET_BITS=64 on Windows, that's why the diglib
 test did not pass previously.

 I had to add some more conditionalised defines to config.h.in because we
 redefine some internal types and functions. I could not get grass7 to
 compile when using old names, so I defined _NO_OLD_NAMES_, but pulled
 those in that grass needs. This way I did not have to modify the source
 code for compilation on Linux and Mac, everything is conditionalised for
 mingw32 in config.h.in.

 Needs more testing, I can compile grass7, errors are ctypes, but I can't
 get it running for various reasons, a bit arbitrary. Grass7 can't start
 the shell using /C/OSGeo4w/apps/msys/bin/sh.exe although that file
 exists??? Sometimes grass7 can't find the .gislock file in the mapset I
 want to use. Mysterious. Configuration and compilation with cairo worked
 fine BTW.

 Markus M

-- 
Ticket URL: <http://trac.osgeo.org/grass/ticket/1131#comment:9>
GRASS GIS <http://grass.osgeo.org>



More information about the grass-dev mailing list