Pim van Riezen wrote:
>
> On Tue, 29 May 2001, David Douthitt wrote:
>
> > Here it is:
> >
> > COMPILE TIME ERRORS FOR TCP WRAPPERS AGAINST GLIBC 2.0
> >
> > make[1]: Entering directory
> > `/pub/oxygen/src/net/security/tcpwrappers/tcp_wrappers_7.6'
> > i386-glibc20-linux-gcc -O2 -g -DFACILITY=LOG_DAEMON
> > -DHOSTS_ACCESS -DDAEMON_UMASK=022 -DREAL_DAEMON_DIR=\"/usr
> > /sbin\" -DPROCESS_OPTIONS -DKILL_IP_OPTIONS -DSEVERITY=LOG_INFO
> > -DRFC931_TIMEOUT=10 -DHOSTS_DENY=\"/etc/hosts.deny\"
> > -DHOSTS_ALLOW=\"/etc/hosts.allow\" -DSYS_ERRLIST_DEFINED
> > -DHAVE_WEAKSYMS -D_REENTRANT backadm.cuna.com -o tcpd.o -c
> ^^^^^^^^^^^^^^^^
> > tcpd.c
>
> I dunno where that came from, but I bet that gcc thinks it is a
> sourcefile, which is why it is moaning about the "multiple compiles".
> Grep the makefile/makeinclude for it I'd say.
That turned out to be what it was, but the reason was rather
insidious.....
My compile time environment is Red Hat 6.x with the glibc 2.0
compatability libraries - thus I've gotten to appreciate make -e quite
well. In the case of tcpwrappers, make -e was required to make ${CC}
equal to i386-glic20-linux-gcc; however, it ALSO made ${HOSTNAME}
equal to the compiler's system hostname - which is NOT what was
intended. Cutting out $(HOSTNAME) from the make file made it work...
Thanks Pim!
_______________________________________________
Leaf-devel mailing list
[EMAIL PROTECTED]
http://lists.sourceforge.net/lists/listinfo/leaf-devel
