*** For details on how to be removed from this list visit the ***
*** CCP4 home page http://www.ccp4.ac.uk ***
Yes
That is exactly what has happened. And there are few more routine to
deal with external restraints, new more complex ncs restraints etc.
All files are in makefile_ibmmac in the refmac5.3 source file. I have
not managed to use gfortran properly yet (I guess it is my fault and/or
my g compiler installation). If anybody will have gfortran
compilation makefile please let me know.
Garib
On 5 Oct 2006, at 22:38, Peter Adrian Meyer wrote:
*** For details on how to be removed from this list visit the ***
*** CCP4 home page http://www.ccp4.ac.uk ***
Hi,
Is it possible that subroutines have moved into difference source
files,
or that some of the new source files haven't been incorporated into
the
makefile targets? Or that the compilations flags have changed?
_errwrt_ seems like it should be defined as subroutine errwrt in
hkon_secder_tch.f (and that object file is listed in your log). So
possibly different fortran name-mangling schemes were used during
compilation (although make clean before starting should've taken
care of
this, so it's probably not the problem). _read_psuedorest_ seems
to be a
new subroutine (at least grep -nir psuedorest in $CCP4 comes up
empty on
my system).
Hopefully one of the refmac (or os x) people will have a better
idea of
what's going on (I'm slightly confused as to why the c++ standard
library
is being linked into a fortran program, but I'm sure there's a
reason for
it).
Good luck,
Pete
Hi folks:
I thought I would try compiling refmac version 5.3 in situ in
ccp4, so I
added the new files, replacing the old. This strategy seemed to work
until linking, when I encountered this:
eval /sw/lib/gcc4/bin/gfortran ${refmac5_FLAGS-"-O0 -fno-second-
underscore"} -o refmac5 make_CIF.o make_subr.o make_lib.o make_crd.o
make_new.o make_PDB.o refmac_newdic.o solvent_contr_mask.o
hkon_secder_tch.o fast_hessian_tabulation.o subvag.o lsq_symm_atoms.o
oppro_tor1.o rcard_tor1.o subcf_scale1.o ncs_const.o
subag_scale_hessian.o extra_eigen.o rigid_body_newdic.o
aniso_ref_eigen.o linalgebra.o make_subr_refmac.o tls_newdic_non.o
vdw_symmetry.o make_rst.o makecif.o make_refmac_dummy.o
ls_scaling_double.o interpolate_mod.o rharvest.o newentry_idealise.o
ml_scaling_double.o gaussian_int.o subnumrec.o -L../lib/src -lccp4f -
lccp4c -lmmdb -L/sw/src/fink.build/ccp4-6.0.1-1005/ccp4-6.0.1/lib/
ccif
-lccif -framework vecLib -L/sw/lib/gcc4/lib/gcc/i386-apple-
darwin8/4.2.0 -L/sw/lib/gcc4/lib/gcc/i386-apple-
darwin8/4.2.0/../../.. -lstdc++ -lgcc_s.10.4 -lSystem -lm
/sw/lib/odcctools/bin/ld: Undefined symbols:
_errwrt_
_read_psuedorest_
collect2: ld returned 1 exit status
make: *** [refmac5] Error 1
Any idea what I am missing?
Thanks.
Bill Scott
Pete Meyer
Fu Lab
BMCB grad student
Cornell University