Okay, I sent this a while back, and the current source still exhibits all
the previous errors. Is there a chance this can be fixed? Even fixing the
check for dn_expand in resolv.so would help ...
Tony
>Date: Tue, 10 Apr 2001 15:29:39 -0700
>To: "jeremy@valinux.com" <jeremy@valinux.com>,
samba@samba.org
>From: Anthony Brock <abrock@georgefox.edu>
>Subject: Re: 2.2 release imminent.
>
>Okay,
>
>After hacking around a little, I got it to compile. Basically, Solaris
>requires the resolv.so library be added to the list of libraries for
>dn_expand, and the library libcrypto.a does not exist on my
>system. Instead, I found the library libk5crypto.a, which appears to be
>the default installed by MIT v5 1.2.2.
>
>Let me know if this helps (I have a diff of the source/configure script
>listed below):
>
># diff configure.orig configure
>10788c10788,10836
>< LIBS="$LIBS -ldes425 -lkrb5 -lcrypto -lcom_err"
>---
> > echo $ac_n "checking for dn_expand in -lresolv""...
$ac_c" 1>&6
> > echo "configure:10725: checking for dn_expand in -lresolv"
>&5
> > ac_lib_var=`echo resolv'_'dn_expand | sed
'y%./+-%__p_%'`
> > if eval "test \"`echo
'$''{'ac_cv_lib_$ac_lib_var'+set}'`\" = set";
then
> > echo $ac_n "(cached) $ac_c" 1>&6
> > else
> > ac_save_LIBS="$LIBS"
> > LIBS="-lresolv $LIBS"
> > cat > conftest.$ac_ext <<EOF
> > #line 10733 "configure"
> > #include "confdefs.h"
> > /* Override any gcc2 internal prototype to avoid an error. */
> > /* We use char because int might match the return type of a gcc2
> > builtin and then its argument prototype would still apply. */
> > char dn_expand();
> >
> > int main() {
> > dn_expand()
> > ; return 0; }
> > EOF
> > if { (eval echo configure:10744: \"$ac_link\") 1>&5;
(eval $ac_link)
> 2>&5; } &&
> > test -s conftest${ac_exeext}; then
> > rm -rf conftest*
> > eval "ac_cv_lib_$ac_lib_var=yes"
> > else
> > echo "configure: failed program was:" >&5
> > cat conftest.$ac_ext >&5
> > rm -rf conftest*
> > eval "ac_cv_lib_$ac_lib_var=no"
> > fi
> > rm -f conftest*
> > LIBS="$ac_save_LIBS"
> >
> > fi
> > if eval "test \"`echo
'$ac_cv_lib_'$ac_lib_var`\" = yes"; then
> > echo "$ac_t""yes" 1>&6
> > ac_tr_lib=HAVE_LIB`echo resolv | sed -e
's/[^a-zA-Z0-9_]/_/g' \
> > -e
'y/abcdefghijklmnopqrstuvwxyz/ABCDEFGHIJKLMNOPQRSTUVWXYZ/'`
> > cat >> confdefs.h <<EOF
> > #define $ac_tr_lib 1
> > EOF
> >
> > LIBS="-lresolv $LIBS"
> >
> > else
> > echo "$ac_t""no" 1>&6
> > fi
> >
> > LIBS="$LIBS -ldes425 -lkrb5 -lk5crypto -lcom_err"
>
>Lastly, when compiling, I see the following warnings:
>
>...
>Compiling lib/readline.c
>Compiling tdb/tdb.c
>"tdb/tdb.c", line 1110: warning: integer overflow detected: op
"<<"
>Compiling tdb/spinlock.c
>Compiling tdb/tdbutil.c
>Compiling printing/printing.c
>Compiling printing/nt_printing.c
>Compiling smbd/quotas.c
>"smbd/quotas.c", line 297: warning: argument #2 is incompatible
with
>prototype:
> prototype: pointer to int : "/usr/include/rpc/xdr.h",
line 340
> argument : pointer to uint
>"smbd/quotas.c", line 301: warning: argument #2 is incompatible
with
>prototype:
> prototype: pointer to int : "/usr/include/rpc/xdr.h",
line 340
> argument : pointer to uint
>"smbd/quotas.c", line 305: warning: argument #2 is incompatible
with
>prototype:
> prototype: pointer to int : "/usr/include/rpc/xdr.h",
line 340
> argument : pointer to uint
>Compiling smbd/oplock.c
>Compiling smbd/oplock_irix.c
>...
>
>Tony
>
>At 11:44 AM 04/10/2001 -0700, jeremy@valinux.com wrote:
>>Anthony Brock wrote:
>> >>
>> >> When installing the new version of Samba from CVS (updated
yesterday), I
>> >> was impressed with some of the additional functionality and
support for
>> >> services it now offers.
>> >>
>> >> However, I was disappointed when I was unable to compile using
the
>> >> '--with-krb5' option. Any attempt to use this option
ends with a
>> >> 'configuration summary' error. As we are a mostly
Kerberos 5 shop (MIT
>> >v5
>> >> 1.2.2), this would be ideal for unifying people's
passwords.
>>
>>I just did this with
>>
>>./configure --with-krb5=/path/to/kerberos/base/
>>
>>in the current 2.2 CVS and it compiled and linked fine.
>>
>>Can you give more details ?
>>
>>Jeremy.
>
>******************************************************************************
>* Anthony Brock abrock@georgefox.edu
*
>* Director of Network Services George Fox University
*
>******************************************************************************
******************************************************************************
* Anthony Brock abrock@georgefox.edu *
* Director of Network Services George Fox University *
******************************************************************************
