Another suggestion, crude I must admit:
Since gcc will not let you put /usr/include first,
To get asetkey to compile, I added a symlink to the
source aklog source directory  for com_err.h -> /usr/include
So the Kerberos com_err.h is found first in aklog and asetkey.


It also looks like this patch to the aklog Makefile.in
is needed to get "make dest"  or "make install" to work.

--- ,Makefile.in        Tue Oct 25 14:08:02 2005
+++ Makefile.in Mon Nov 21 16:22:09 2005
@@ -28,7 +28,7 @@
 # Installation targets
 #
 install: \
-       ${DESTDIR}${bindir}/aklog
+       ${DESTDIR}${bindir}/aklog \
        ${DESTDIR}${afssrvbindir}/asetkey

 ${DESTDIR}${bindir}/aklog: aklog
@@ -38,7 +38,7 @@
        ${INSTALL} $? $@

 dest: \
-       ${DEST}/bin/aklog
+       ${DEST}/bin/aklog \
        ${DEST}/root.server/usr/afs/bin/asetkey

 ${DEST}/bin/aklog: aklog




Douglas E. Engert wrote:




Ken Hornstein wrote:

More on aklog using krb5-config, yes MacOS 10.4 has one, but the order of the includes is different using the krb5-config vs the --with-krb5=yes and
KRB5CFLAGS= and KRB5LIBS=.  This caues problems with asetkey.c



Sigh.  This is a mess with two different packages wanting two different
com_err libraries and include files. I have NO idea what the right answer
is here.  I glossed over it with aklog, but it never really got solved
properly.  Suggestions from the peanut gallery?


Are there any other systems where the kerberos headers are in /usr/include?
Is this a gcc bug? Why won't gcc let me put /usr/include first if I want to?
(gcc-3.3.2 on Solaris will also drop /usr/include)

On the Mac at least uisng the Kerberos framwork will put the Kerberos headers
before the AFS.



--Ken
_______________________________________________
OpenAFS-devel mailing list
[email protected]
https://lists.openafs.org/mailman/listinfo/openafs-devel




--

 Douglas E. Engert  <[EMAIL PROTECTED]>
 Argonne National Laboratory
 9700 South Cass Avenue
 Argonne, Illinois  60439
 (630) 252-5444
_______________________________________________
OpenAFS-devel mailing list
[email protected]
https://lists.openafs.org/mailman/listinfo/openafs-devel

Reply via email to