shadow-4.0.11 breaks shadow group support
Greg Schafer <[email protected]> Tue, 19 Jul 2005 20:53:57 +1000
| Newsgroups | gmane.linux.pld.shadow.general |
|---|---|
| Message-ID | <[email protected]> |
Hi
This latest release breaks shadow group suport. For example, trying to run
grpconv results in:
$ grpconv
grpconv: not configured for shadow group support.
This worked fine in shadow-4.0.10
It seems the cause is related to the definition of `SHADOWGRP' ie: it longer
gets defined.
--------------------------------------
>From configure script of shadow-4.0.10
if test "$enable_shadowgrp" != "no"; then
cat >>confdefs.h <<\_ACEOF
#define SHADOWGRP 1
_ACEOF
fi
--------------------------------------
>From configure script of shadow-4.0.11
if test "$enable_shadowgrp" = "yes"; then
cat >>confdefs.h <<\_ACEOF
#define SHADOWGRP 1
_ACEOF
fi
--------------------------------------
It appears `enable_shadowgrp' is not defined anywhere so in actual fact,
this has been broken all along and only worked in the last release by
chance.
Additionally, there is a separate autoconf check for shadow group support
but it defines something else (HAVE_SHADOWGRP) but this check appears to be
broken anyway because it always fails:
checking for working shadow group support... no
config.log shows this:
configure:22999: checking for working shadow group support
configure:23025: gcc -o conftest -O2 -pipe -s conftest.c >&5
conftest.c: In function `main':
conftest.c:99: warning: initialization makes pointer from integer without a cast
conftest.c:101: error: dereferencing pointer to incomplete type
conftest.c:101: error: dereferencing pointer to incomplete type
configure:23028: $? = 1
configure: program exited with status 1
configure: failed program was:
In summary, it appears shadow group support is very broken and the
autoconfigury is confused. I'm not sure of the best way to fix it.
Regards
Greg