Hey Mikel,
On May 31, 2009, at 10:06 AM, Mikel Maron wrote:
Thanks Dane, that got it moving.
Great.
Now in the linking, getting this error. What needs configuring?
Scons still puzzles me.
/usr/bin/ld: /usr/lib64/libboost_system-mt.a(error_code.o):
relocation R_X86_64_32 against `a local symbol' can not be used when
making a shared object; recompile with -fPIC
/usr/lib64/libboost_system-mt.a: could not read symbols: Bad value
Hmm, that's not an error I've ever seen. The '.a' library denotes a
statically compiled library, which must need to be compiled with the -
fPIC flag to get the right symbols exposed. How did you build/install
boost? This seems like a boost install problem rather than a Mapnik/
Scons issue. Normally the linker will choose the .so over the
static .a lib, so I would try to reinstall boost (via source or
packages to make the 'so' dynamic lib available), then clean your
mapnik builds ('python scons/scons.py -c'), and try again.
Note: if you do a source install of Boost then you'll likely need to
point SCons at the right library name with the BOOST_TOOLKIT=gcc42
(for example) flag.
Also, perhaps you have boost installed in /usr/lib already instead of /
usr/lib64. If you do then you can point SCons there with
BOOST_INCLUDE=/usr/include and BOOST_LIBS=/usr/lib
- Dane
-Mikel
From: Dane Springmeyer <[email protected]>
To: Mikel Maron <[email protected]>
Cc: Mapnik users mailing-list <[email protected]>
Sent: Sunday, May 31, 2009 9:56:16 AM
Subject: Re: [Mapnik-users] memory error compiling mapnik .. help?
Hey mikel. In the previous case I stopped apache which freed up the
needed men to get past that one particularly expensive compile.
Scons may have a low mem setting - I'll look around for that when I
get a chance. A longer term solution may be to spilt apart some of
that code file. Also perhaps making the pickle support an optional
compile which might reduce the memory burden a bit as well.
Dane
On May 30, 2009, at 11:42 PM, Mikel Maron <[email protected]>
wrote:
Hi
I'm running into a problem when the compile comes to mapnik_rule.cpp.
It's basically this problem, as reported by Dane.
http://trac.mapnik.org/ticket/235
It's not a mapnik problem per se .. but wondering if anyone has
advice on how to get through this compile step, without having the
OS kill off process.
Thanks
Mikel
_______________________________________________
Mapnik-users mailing list
[email protected]
https://lists.berlios.de/mailman/listinfo/mapnik-users
_______________________________________________
Mapnik-users mailing list
[email protected]
https://lists.berlios.de/mailman/listinfo/mapnik-users
