build-sys: release++ (v2.14.1-rc1)

Signed-off-by: Karel Zak <kzak@redhat.com>

docs: update v2.14.1 ReleaseNotes

Signed-off-by: Karel Zak <kzak@redhat.com>

docs: update AUTHORS file

Signed-off-by: Karel Zak <kzak@redhat.com>

po: merge changes

Signed-off-by: Karel Zak <kzak@redhat.com>

po: update cs.po (from translationproject.org)

po: update nl.po (from translationproject.org)

scriptreplay: new implementation is out-of-sync

Reproduction steps:

1. Run "script -t 2>timing".
2. When the prompt comes up, hit return.
3. New prompt. Wait several seconds, then type "true" and hit return.
4. Repeat (3) as desired, then exit scripted shell.
5. Run "scriptreplay timing".

You'll find on the replay that, rather than waiting several seconds
between the prompt and true, the wait will actually occur after the
first "t" of true.

Reported-by: Micah Cowan <micah@cowan.name>
Signed-off-by: Karel Zak <kzak@redhat.com>

mount: add docs about utf8=0 for vfat

Addresses-Red-Hat-Bugzilla: #454354
Signed-off-by: Karel Zak <kzak@redhat.com>

mount: add fallback for versionsort()

I was trying to build util-linux-ng on uclibc and it failed

| lomount.c: In function 'looplist_next':
| lomount.c:223: error: 'versionsort' undeclared (first use in this function)
| lomount.c:223: error: (Each undeclared identifier is reported only once
| lomount.c:223: error: for each function it appears in.)
| make[2]: *** [mount-lomount.o] Error 1
|

which is understood because versionsort is GNU extension and uclibc does
not implement it.

Reported-by: Khem Raj <raj.khem@gmail.com>
Signed-off-by: Karel Zak <kzak@redhat.com>

build-sys: tgets is not in ncurses but in tinfo

ncurses library can be build with terminal related functions
landing in separate libtinfo library. Check for ncurses function
when testing ncurses library existence.

Signed-off-by: Arkadiusz Miśkiewicz <arekm@maven.pl>

build-sys: fix dmesg.1 installation

Signed-off-by: Alexey Gladkov <legion@altlinux.org>
Signed-off-by: Karel Zak <kzak@redhat.com>

build-sys: cleanup sys-utils/Makefile.am

Signed-off-by: Karel Zak <kzak@redhat.com>

hwclock: unshadow a diagnostic printf

- Bogus if test means one message is never produced.

- Avoid needless passing of a global variable (debug).

The --test option flag ought to be a global as well (and perhaps -n/--dry-run).

Signed-off-by: Kalev Soikonen <ksop@hot.ee>

mount: add info about tz=UTC option for FAT to mount.8

CC: Joe Peterson <joe@skyrush.com>
Signed-off-by: Karel Zak <kzak@redhat.com>

sfdisk: print version should end with a newline

Signed-off-by: Denis ChengRq <crquan@gmail.com>

rtcwake: cleanup return codes

Signed-off-by: Karel Zak <kzak@redhat.com>

rtcwake: prefer RTC_WKALM_SET over RTC_ALM_SET

rtcwake: Prefer RTC_WKALM_SET over RTC_ALM_SET, fixing bug with not
waking up with new RTCs.  Also, return error if unable to set the alarm.

Signed-off-by: Gabriel Burt <gburt@novell.com>
Signed-off-by: Karel Zak <kzak@redhat.com>

write: doesn't check for tty group

write(1) selects a wrong tty, because there is not a proper
check of tty group ownership:

$ write kzak
write: kzak is logged in more than once; writing to tty7
write: /dev/tty7: Permission denied

$ ls -la /dev/tty7
crw--w---- 1 root root 4, 7 2008-07-04 00:32 /dev/tty7
                  ^^^^

$ ls -la /usr/bin/write
-rwxr-sr-x 1 root tty 11864 2008-04-02 16:24 /usr/bin/write
      ^           ^^^

We have to check for tty group owner, because we don't have
permissions to write to arbitrary tty.

Fixed version:

$ write kzak
write: kzak is logged in more than once; writing to pts/6
                                                    ^^^^
Message from test@nb on pts/7 at 15:22 ...

^C

$ ls -la /dev/pts/6
crw--w---- 1 kzak tty 136, 6 2008-07-07 15:35 /dev/pts/6
                  ^^^

Addresses-Red-Hat-Bugzilla: #454252
Signed-off-by: Karel Zak <kzak@redhat.com>

umount: improve "-d" option for autoclear loops

The new loop auto-destruct feature detaches automatically loop devices
when no longer used. This means they are detached with the umount()
call. But when we call umount with -d, del_loop() is called and fails
because the ioctl() returns ENXIO. We have to check for autoclear
loop devices rather than blindly call del_loop().

Reported-by: Matthias Koenig <mkoenig@suse.de>
Signed-off-by: Karel Zak <kzak@redhat.com>

selinux: is_selinux_enabled() returns 0, 1 and -1

Unfortunately, the current libselinux implementation of
is_selinux_enabled() returns -1 on error. This behavior is
undocumented.

The proper solution is to use "if (is_selinux_enabled() > 0)".

Signed-off-by: Karel Zak <kzak@redhat.com>

mount: add norealtime to mount.8

Signed-off-by: Karel Zak <kzak@redhat.com>

hwclock: omit warning about drift if --noadjfile given

Currently, if hwclock is given the --noadjfile option it will
nevertheless display information about the drift rate when invoked with
the --debug option.

Signed-off-by: Matthias Koenig <mkoenig@suse.de>

build-sys: release++ (v2.14)

Signed-off-by: Karel Zak <kzak@redhat.com>

docs: update v2.14 ReleaseNotes

Signed-off-by: Karel Zak <kzak@redhat.com>

docs: update AUTHORS file

Signed-off-by: Karel Zak <kzak@redhat.com>

po: merge changes

Signed-off-by: Karel Zak <kzak@redhat.com>

po: update hu.po (from translationproject.org)

ddate: 11th, 12th and 13th of month

the Discordian date utility ddate gives the 11th, 12th and 13th of the month as
the "11st", "12nd" and "13rd".  Unless this is a religious thing, please apply
the patch below.

Signed-off-by: Volker Schatz <oss@volkerschatz.com>

rtcwake: fix the default mode to "standby"

Writing "suspend" to /sys/power/state does nothing.
Even "man rtcwake" says that default should be "standby" :)

Signed-off-by: Paulius Zaleckas <paulius.zaleckas@teltonika.lt>
Signed-off-by: Karel Zak <kzak@redhat.com>

mount: fix a small typo in mount.8

While working on french translation of the Linux Man Pages, I've found a
small typo in mount.8.

Only one wrong letter : the option "osyncis_o_sync" for XFS filesystem
is erroneously replaced by "osyncis_d_sync" (the previous option).

Signed-off-by: Christophe Blaess <Christophe@Blaess.fr>
Signed-off-by: Karel Zak <kzak@redhat.com>

build-sys: release++

Signed-off-by: Karel Zak <kzak@redhat.com>

docs: update 2.14 ReleaseNotes

Signed-off-by: Karel Zak <kzak@redhat.com>

po: merge changes

Signed-off-by: Karel Zak <kzak@redhat.com>

po: update vi.po (from translationproject.org)

po: update nl.po (from translationproject.org)

po: update it.po (from translationproject.org)

po: update cs.po (from translationproject.org)

docs: we already rewrote the scriptreplay script; remove that TODO entry

Signed-off-by: James Youngman <jay@gnu.org>

mount: remount doesn't care about loop=

The command

   # mount -oremount <spec> <dir>

doesn't read fstab or mtab. This is expected behaviour. Unfortunately,
we have to care about the internal loop= option which is generated and
maintained by mount(8)/umount(8). The loop= option has to be persistent.

How to reproduce this bug:

 # mount -o loop       /home/images/vfat.img /mnt/img; grep vfat /etc/mtab; \
   mount -o remount,ro /home/images/vfat.img /mnt/img; grep vfat /etc/mtab;

 /home/images/vfat.img /mnt/img vfat rw,loop=/dev/loop0 0 0
 /home/images/vfat.img /mnt/img vfat ro 0 0

Reported-By: David Chinner <dgc@sgi.com>
Signed-off-by: Karel Zak <kzak@redhat.com>

build-sys: add support ionice for Super-H architecture

Addresses-Red-Hat-Bugzilla: #446560
Signed-off-by: Karel Zak <kzak@redhat.com>

fdisk: doesn't recognize the VMware ESX partitions

The fdisk programs do not recognize the partition types used by VMware
ESX.  They show up as "unknown".

Addresses-Red-Hat-Bugzilla: #447023
Signed-off-by: Karel Zak <kzak@redhat.com>

setarch: add fallback for linux/personality

setarch.c:248: error: 'ADDR_NO_RANDOMIZE' undeclared (first use in this function)
setarch.c:248: error: (Each undeclared identifier is reported only once
setarch.c:248: error: for each function it appears in.)
setarch.c:251: error: 'FDPIC_FUNCPTRS' undeclared (first use in this function)
setarch.c:257: error: 'ADDR_COMPAT_LAYOUT' undeclared (first use in this function)
setarch.c:260: error: 'READ_IMPLIES_EXEC' undeclared (first use in this function)

Linux gzp1 2.4.36.1-gzp1 #1 SMP Tue Feb 19 10:23:48 CET 2008 i686 GNU/Linux

Reported-By: Gabor Z. Papp <gzp@papp.hu>
Signed-off-by: Karel Zak <kzak@redhat.com>

build-sys: release++

Signed-off-by: Karel Zak <kzak@redhat.com>

docs: update v2.14 ReleaseNotes

Signed-off-by: Karel Zak <kzak@redhat.com>

po: merge changes

Signed-off-by: Karel Zak <kzak@redhat.com>

scriptreplay: gettextize a forgotten messages

Signed-off-by: Karel Zak <kzak@redhat.com>

docs: update v2.14 ReleaseNotes

Signed-off-by: Karel Zak <kzak@redhat.com>

ionice: update man page to reflect IDLE class change in  2.6.25

The idle class is safe for non-root users since 2.6.25.
http://lwn.net/Articles/266256/

Addresses-Red-Hat-Bugzilla: #443823
Signed-off-by: Karel Zak <kzak@redhat.com>

login: audit log injection attack via login

A while back I found a couple audit log injection attacks which became
CVE-2007-3102. I forgot to look at login to see if its vulnerable and Mirek
found that it is. To verify the problem, type:

root addr=xyz.com

for the account name while logging in. It will look like root logged in with
an address of xyz.com.

Signed-off-by: Steve Grubb <sgrubb@redhat.com>

docs: remove date from ReleasNotes

Signed-off-by: Karel Zak <kzak@redhat.com>

docs: update v2.14-ReleaseNotes

Signed-off-by: Karel Zak <kzak@redhat.com>

docs: update AUTHORS file

Signed-off-by: Karel Zak <kzak@redhat.com>

cal: use HAVE_LIB{NCURSES,NCURSESW} instead HAVE_NCURSES

The HAVE_NCURSES is obsolete macro. The autotool generates HAVE_LIBxxx macros.

Signed-off-by: Karel Zak <kzak@redhat.com>

more: use HAVE_WIDECHAR instead ENABLE_WIDECHAR

This problem has been detected by tools/codecheck-config.

Signed-off-by: Karel Zak <kzak@redhat.com>

po: merge changes

Signed-off-by: Karel Zak <kzak@redhat.com>

po: update nl.po (from translationproject.org)

build-sys: remove unnecessary check-local target from login-utils/

We have all tests in the tests/ directory.

Signed-off-by: Karel Zak <kzak@redhat.com>

mkswap: set errno=0 in write_all()

Signed-off-by: Karel Zak <kzak@redhat.com>

setarch: provide backwards compatibility

The getopt_long(3) use introduced by commit
8df90dec2bfd8c876ddd4cea010f40550d4e24aa
broke backwards compatibility.

Old behavior:
$ i386 uname -m
i686

New behavior:
$ i386 uname -m
i386: invalid option -- m
Linux

Traditional method to fix it is to disable program arguments reordering
by prefixing getopt_long's optstring with '+' character.

Signed-off-by: Dmitry V. Levin <ldv@altlinux.org>
Signed-off-by: Karel Zak <kzak@redhat.com>

losetup: fix typo in losetup.8

Signed-off-by: Karel Zak <kzak@redhat.com>

docs: add note about losetup --sizelimit to ReleaseNotes

Signed-off-by: Karel Zak <kzak@redhat.com>

chfn, chsh, login: collapsing three similar messages into a single one

Signed-off-by: Benno Schulenberg <bensberg@justemail.net>

losetup: split help message into two smaller parts

As it has already changed since the previous release, take the opportunity
to cut it into two more manageable chunks for translators.

Signed-off-by: Benno Schulenberg <bensberg@justemail.net>
Signed-off-by: Karel Zak <kzak@redhat.com>

swapon: fix swsuspend detection

Welcome to hell where S1SUSPEND/S2SUSPEND is

  "swsuspend" in libblkid

and

  "suspend" in libvolume_id

Signed-off-by: Karel Zak <kzak@redhat.com>

mount: fix typo in mount.8

Signed-off-by: Karel Zak <kzak@redhat.com>

losetup: mark the option -s as deprecated

Signed-off-by: Karel Zak <kzak@redhat.com>

mount: add support for sizelimit= mount option (for loop mounts)

[kzak@redhat.com: split the original patch to small patches]
Signed-off-by: Shachar Shemesh <shachar@lingnu.com>
Signed-off-by: Karel Zak <kzak@redhat.com>

losetup: add --sizelimit option

[kzak@redhat.com: split the original patch to small patches]
Signed-off-by: Shachar Shemesh <shachar@debian.org>
Signed-off-by: Karel Zak <kzak@redhat.com>

mkswap: when writing the signature page, handle EINTR returns

If the signature page write bumps into EINTR, it should finish the
write, instead of dying.

Addresses-Ubuntu-Bug: 206113
Signed-off-by: LaMont Jones <lamont@canonical.com>
Signed-off-by: Karel Zak <kzak@redhat.com>

build-sys: update .gitignore files

Signed-off-by: Karel Zak <kzak@redhat.com>

build-sys: ignore a bunch of generated files, mostly binaries

Signed-off-by: James Youngman <jay@gnu.org>
Signed-off-by: Karel Zak <kzak@redhat.com>

login: rewrite is_local() to remove limits on line length

Signed-off-by: James Youngman <jay@gnu.org>
Signed-off-by: Karel Zak <kzak@redhat.com>

docs: cleanup README.devel, add note about coding style and Signed-off-by

Signed-off-by: Karel Zak <kzak@redhat.com>

fdisk: message tweak

Signed-off-by: Karel Zak <kzak@redhat.com>

scriptreplay: rewrite in C

The dependence on Perl sucks...

Co-Author: James Youngman <jay@gnu.org>
Signed-off-by: Karel Zak <kzak@redhat.com>
Signed-off-by: James Youngman <jay@gnu.org>

docs: update v2.14-ReleaseNotes

Signed-off-by: Karel Zak <kzak@redhat.com>

mount: remove set_proc_name()

This function has been used in background mount code.

Signed-off-by: Karel Zak <kzak@redhat.com>

hwclock: add --adjfile=path option

Signed-off-by: Karel Zak <kzak@redhat.com>

ldattach: use glibc termios

Signed-off-by: Karel Zak <kzak@redhat.com>

fsck.minix: reset the terminal state if we are killed by a fatal signal

[kzak@redhat.com: remove unnecessary volatile type qualifier]
Signed-off-by: James Youngman <jay@gnu.org>
Signed-off-by: Karel Zak <kzak@redhat.com>

mkfs.cramfs: switch on localization.

Signed-off-by: Benno Schulenberg <bensberg@justemail.net>

script: read returns a size_t

Signed-off-by: James Youngman <jay@gnu.org>
Signed-off-by: Karel Zak <kzak@redhat.com>

login-utils: cleanup strlen() and fgets() usage

The strlen() returns size_t, not int. The fgets() returns
pointer, not int.

Signed-off-by: James Youngman <jay@gnu.org>
Signed-off-by: Karel Zak <kzak@redhat.com>

fsck.minix: correct the error message given when we can't open the device

Don't actually print %s in the error message, print the device name,
as was obviously intended.  Also, print the error message
corresponding to the errno value.

[kzak@redhat.com: add __attribute__, coding style cleanups]
Signed-off-by: James Youngman <jay@gnu.org>
Signed-off-by: Karel Zak <kzak@redhat.com>

mount: remove built-in support for background mounts

EX_BG is archaism from old integrated NFS code. The built-in NFS code
has been removed in the previous version and all "bg" stuff is handled
by /sbin/mount.nfs. It seems we can remove all "bg" stuff from
mount(8).

Signed-off-by: Karel Zak <kzak@redhat.com>

Revert "script: improve quiet mode"

This reverts commit 7baec9a5612358559d1556113657223a64e373d1.

Note, scriptreplay(8) expects the "Script started on.." line. We cannot remove
this message...

docs: cleanup DEPRECATED file

Signed-off-by: Karel Zak <kzak@redhat.com>

docs: update AUTHORS file

Signed-off-by: Karel Zak <kzak@redhat.com>

docs: add v2.14 ReleaseNotes

Signed-off-by: Karel Zak <kzak@redhat.com>

po: merge changes

Signed-off-by: Karel Zak <kzak@redhat.com>

ldattach: add NLS support

Signed-off-by: Karel Zak <kzak@redhat.com>

po: add pl.po (from translationproject.org)

po: add eu.po (from translationproject.org)

po: update vi.po (from translationproject.org)

po: update uk.po (from translationproject.org)

po: update tr.po (from translationproject.org)

po: update sv.po (from translationproject.org)

po: update sl.po (from translationproject.org)

po: update ru.po (from translationproject.org)

po: update pt_BR.po (from translationproject.org)