On Aug 27, 2006, at 10:31 AM, Justin C. Walker wrote:
I am attempting to build a chunk of code from the open source
world (Singular, a computer algebra system). In the course of the
build, I get this failure (both PowerPC and Intel):
libtool -dynamic -twolevel_namespace -weak_reference_mismatches
weak -undefined dynamic_lookup -o mpsr.so mpsr_Put.dl_o
mpsr_PutPoly.dl_o mpsr_GetPoly.dl_o mpsr_sl.dl_o mpsr_Get.dl_o
mpsr_GetMisc.dl_o mpsr_Error.dl_o -L/lib -lMPT -lMP
ld: for architecture ppc
ld: common symbols not allowed with MH_DYLIB output format with
the -multi_module option
/lib/libMP.a(MP_Util.o) definition of common _fix_log_msg (size 128)
/lib/libMP.a(MP_Util.o) definition of common _AnnotFlagsStr (size 32)
libtool: internal link edit command failed
After R'ing TFM, I tried adding -single_module to LDFLAGS. That
worked, AFAICT, but I'm not sure this is the correct fix.
That's an excellent fix.
It seemed like one, and I must have gotten lucky; after I "cleaned
up" and rebuilt, I now get this complaint:
g++ -g -pipe -fpic -DPIC -dynamic -single_module -twolevel_namespace -
weak_reference_mismatches weak -undefined dynamic_lookup
p_Procs_Lib_FieldZp.dl_og -o p_Procs_FieldZp.sog
/usr/bin/ld: -single_module or -multi_module flags can only be used
when -dylib is also specified
at a different spot in the build. Any thoughts? Should -dynamic be -
dylib?
Changing configure.in is great, but to get the fix to be used in
the build you'll also have to re-run autoconf to regenerate the
configure file.
This, I think, will be the easy part :-}
Justin
--
Justin C. Walker, Curmudgeon at Large
Institute for the Absorption of Federal Funds
-----------
If it weren't for carbon-14, I wouldn't date at all.
-----------
_______________________________________________
Do not post admin requests to the list. They will be ignored.
Unix-porting mailing list (email@hidden)
Help/Unsubscribe/Update your Subscription:
http://lists.apple.com/mailman/options/unix-porting/email@hidden
