On Fri, Sep 16, 2005 at 08:49:54AM +0200, Geert Hendrickx wrote:
> Is this normal output for etcupdate? 
> 
> # etcupdate -a -l -s /cvs/src
> *** Creating /tmp/temproot
> *** Populating /tmp/temproot from /cvs/src
> umount: /cvs/obj/regress/sys/kern/umount/umount-f_mount: not currently mounted
> vnconfig: /dev/rvnd0d: VNDIOCCLR: Device not configured
> In file included from /cvs/src/lib/libc/gen/pw_scan.c:34:
> /cvs/src/tools/compat/compat_pwd.h:45:1: warning: "user_from_uid" redefined
> In file included from nbtool_config.h:133,
>                  from /cvs/src/lib/libc/gen/pw_scan.c:33:
> /cvs/src/tools/compat/compat_defs.h:312:1: warning: this is the location of the previous definition
> In file included from /cvs/src/lib/libc/gen/pw_scan.c:34:
> /cvs/src/tools/compat/compat_pwd.h:46:1: warning: "group_from_gid" redefined
> In file included from nbtool_config.h:133,
>                  from /cvs/src/lib/libc/gen/pw_scan.c:33:
> /cvs/src/tools/compat/compat_defs.h:314:1: warning: this is the location of the previous definition
> /cvs/src/lib/libc/sys/_getcontext.c:36:2: warning: suggest hiding #error from traditional C with an indented #
> /cvs/src/lib/libc/stdlib/getopt_long.c:74:2: warning: suggest not using #elif in traditional C
> In file included from /cvs/src/lib/libc/gen/nlist.c:92:
> /cvs/src/lib/libc/gen/nlist_private.h:40:2: warning: suggest not using #elif in traditional C
> /cvs/src/lib/libc/gen/nlist_private.h:43:2: warning: suggest not using #elif in traditional C
> /cvs/src/lib/libc/gen/nlist_private.h:47:2: warning: suggest not using #elif in traditional C
> /cvs/src/lib/libc/gen/nlist_private.h:51:2: warning: suggest not using #elif in traditional C
> /cvs/src/lib/libc/gen/nlist_private.h:55:2: warning: suggest not using #elif in traditional C
> /cvs/src/lib/libc/gen/nlist_private.h:58:2: warning: suggest not using #elif in traditional C
> /cvs/src/lib/libc/gen/nlist_private.h:61:2: warning: suggest not using #elif in traditional C
> In file included from /cvs/src/lib/libc/gen/nlist_aout.c:87:

Heh, my mail wasn't finished, but it had lines with single dots on them :-)
(to indicate that the output went on like this).  

Anyway, etcupdate starts recompiling stuff in /cvs/src...  I guess that is
not intended.  

When I re-mount /cvs (containing both src/ and obj/) read-only, etcupdate
aborts with: 

# etcupdate -a -l -s /cvs/src
*** Creating /tmp/temproot
*** Populating /tmp/temproot from /cvs/src
umount: /cvs/src/regress/sys/kern/umount/umount-f_mount: not currently
mounted
vnconfig: /dev/rvnd0d: VNDIOCCLR: Device not configured
rpcgen: unable to open rpcb_prot.h: Read-only file system

Did I do something wrong, or is this a bug?  (it wasn't there in RC3...)  

GH