This question relates, at least vaguely, to recent postings
from me concerning my attempts to build R-1.5.1 under Solaris 2.7.
I've more or less given up on this, and I and a colleague are
attempting to build R-1.5.1 on another machine in our Department,
and are getting a ***different*** problem.
Version details:
platform sparc-sun-solaris2.9
arch sparc
os solaris2.9
system sparc, solaris2.9
status
major 1
minor 5.1
year 2002
month 06
day 17
language R
The ``other'' machine is relatively new and is running Solaris 2.9.
It has just had gcc version 3.2 installed on it. It also has just
had the readline package installed on it.
But when we do ./configure, we get errors of the following form.
(These lines are copied and pasted from config.log):
===+===+===+===+===+===+===+===+===+===+===+===+===+===+===+===+===+===+==configure:10914:
checking for rl_callback_read_char in -lreadline
configure:10947: gcc -o conftest -g -O2 -I/usr/local/include -L/usr/local/lib
conftest.c -lreadline -ldl -ltermcap -lm >&5
cc1: warning: changing search order for system directory
"/usr/local/include"
cc1: warning: as it has already been specified as a non-system directory
ld: fatal: symbol `__dso_handle' is multiply-defined:
(file /usr/local/lib/gcc-lib/sparc-sun-solaris2.9/3.2/crtbegin.o and
file /usr/local/lib/libreadline.so);
ld: fatal: symbol `_init' is multiply-defined:
(file /usr/local/lib/gcc-lib/sparc-sun-solaris2.9/3.2/crti.o and file
/usr/local/lib/libreadline.so);
ld: fatal: symbol `_start' is multiply-defined:
(file /usr/local/lib/gcc-lib/sparc-sun-solaris2.9/3.2/crt1.o and file
/usr/local/lib/libreadline.so);
ld: fatal: symbol `_fini' is multiply-defined:
(file /usr/local/lib/gcc-lib/sparc-sun-solaris2.9/3.2/crti.o and file
/usr/local/lib/libreadline.so);
ld: fatal: symbol `_lib_version' is multiply-defined:
(file /usr/ccs/lib/values-Xa.o and file /usr/local/lib/libreadline.so);
ld: warning: symbol `_siguhandler' has differing sizes:
(file /usr/local/lib/libreadline.so value=0xb8; file /usr/lib/libc.so
value=0xbc);
/usr/local/lib/libreadline.so definition taken
ld: fatal: File processing errors. No output written to conftest
collect2: ld returned 1 exit status
===+===+===+===+===+===+===+===+===+===+===+===+===+===+===+===+===+===+==
If we proceed on to ``make'' and ``make check'' everything
appears to
go OK, and R-1.5.1 gets made, except that command line editing is not
available. Naturally, since the readline facility is not working.
(Since readline/command line editing is not available, savehistory()
doesn't work either.)
Can anyone give me any idea as to why we are getting these ``multiply
defined'' symbols, and what we can do about it? Please be as
simplistic as you can! The intricacies of compilers and ld are
pretty mysterious to us.
As far as we can discern, gcc and readline were installed in a
completely standard way, so we have no idea why these hiccups
are occurring.
Thanks for any advice/wisdom/insight that anyone can give.
cheers,
Rolf Turner
-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-
r-help mailing list -- Read http://www.ci.tuwien.ac.at/~hornik/R/R-FAQ.html
Send "info", "help", or "[un]subscribe"
(in the "body", not the subject !) To: r-help-request at
stat.math.ethz.ch
_._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._
Rolf Turner <rolf at math.unb.ca> writes:> This question relates, at least vaguely, to recent postings > from me concerning my attempts to build R-1.5.1 under Solaris 2.7....> But when we do ./configure, we get errors of the following form. > (These lines are copied and pasted from config.log): > > ===+===+===+===+===+===+===+===+===+===+===+===+===+===+===+===+===+===+==> configure:10914: checking for rl_callback_read_char in -lreadline > configure:10947: gcc -o conftest -g -O2 -I/usr/local/include -L/usr/local/lib conftest.c -lreadline -ldl -ltermcap -lm >&5 > cc1: warning: changing search order for system directory "/usr/local/include" > cc1: warning: as it has already been specified as a non-system directory > ld: fatal: symbol `__dso_handle' is multiply-defined: > (file /usr/local/lib/gcc-lib/sparc-sun-solaris2.9/3.2/crtbegin.o and file /usr/local/lib/libreadline.so); > ld: fatal: symbol `_init' is multiply-defined: > (file /usr/local/lib/gcc-lib/sparc-sun-solaris2.9/3.2/crti.o > and file /usr/local/lib/libreadline.so);Hm? I wouldn't expect to find those symbols in a dynamic library, and readline hasn't come in dynamic form previously (I have only libreadline.a). Where did you get it from? -- O__ ---- Peter Dalgaard Blegdamsvej 3 c/ /'_ --- Dept. of Biostatistics 2200 Cph. N (*) \(*) -- University of Copenhagen Denmark Ph: (+45) 35327918 ~~~~~~~~~~ - (p.dalgaard at biostat.ku.dk) FAX: (+45) 35327907 -.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.- r-help mailing list -- Read http://www.ci.tuwien.ac.at/~hornik/R/R-FAQ.html Send "info", "help", or "[un]subscribe" (in the "body", not the subject !) To: r-help-request at stat.math.ethz.ch _._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._
Martin Maechler <maechler@stat.math.ethz.ch> writes:> Ehhm, your last note is not true in general (I know you talked > about Solaris, but still): > > On Redhat 7.1 and 7.3 (here), > readline is available both as .so and .a > and R uses the .so one :...> PD> Hm? I wouldn't expect to find those symbols in a dynamic library, and > PD> readline hasn't come in dynamic form previously (I have only > PD> libreadline.a). Where did you get it from?Actually it's not even true for Solaris. I got a sense of deja vu, looked through my install logs, and found that I had *deleted* libreadline.so because it conflicted with gcc 3.0.1 when building R (this was almost exactly a year ago on Sep 10, 2001; the events of the following day might have dislodged things from my memory!) That is of course still a possible workaround... Or downgrading to gcc 2.95.3, which seems more and more attractive given all the trouble with gcc 3.x. -- O__ ---- Peter Dalgaard Blegdamsvej 3 c/ /'_ --- Dept. of Biostatistics 2200 Cph. N (*) \(*) -- University of Copenhagen Denmark Ph: (+45) 35327918 ~~~~~~~~~~ - (p.dalgaard@biostat.ku.dk) FAX: (+45) 35327907 -.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.- r-devel mailing list -- Read http://www.ci.tuwien.ac.at/~hornik/R/R-FAQ.html Send "info", "help", or "[un]subscribe" (in the "body", not the subject !) To: r-devel-request@stat.math.ethz.ch _._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._