chiark / gitweb /
elogind.git
7 years agoudevd: fix bogus mkdir invocation
Dave Reisner [Mon, 2 Jul 2012 01:09:18 +0000 (03:09 +0200)]
udevd: fix bogus mkdir invocation

The filename parameter passed to mkdir can't contain anything but a
garbage value at this point. This was meant to be the full pathname to
the new udev DB, as the mkdir_parents() call before it won't create the
trailing child directory.

[replace mkdir_parents() + mkdir() with mkdir_p() -- kay]

7 years agoudevd: use ROOTPREFIX to create path to modules.devname
Dave Reisner [Sun, 1 Jul 2012 23:37:52 +0000 (19:37 -0400)]
udevd: use ROOTPREFIX to create path to modules.devname

7 years agomanager: serialize/deserialize job counters across reexec/reload
Lennart Poettering [Fri, 29 Jun 2012 17:47:38 +0000 (19:47 +0200)]
manager: serialize/deserialize job counters across reexec/reload

7 years agologind: add PreparingForShutdown/PreparingForSleep properties
Lennart Poettering [Fri, 29 Jun 2012 17:38:35 +0000 (19:38 +0200)]
logind: add PreparingForShutdown/PreparingForSleep properties

7 years agoupdate TODO
Lennart Poettering [Fri, 29 Jun 2012 17:15:12 +0000 (19:15 +0200)]
update TODO

7 years agoremount-fs: also remount /usr according to fstab
Lennart Poettering [Fri, 29 Jun 2012 14:14:01 +0000 (16:14 +0200)]
remount-fs: also remount /usr according to fstab

7 years agoman: add references to the new inhibitor lock documentation
Lennart Poettering [Fri, 29 Jun 2012 00:07:40 +0000 (02:07 +0200)]
man: add references to the new inhibitor lock documentation

7 years agomount: load only if we there's mountinfo or fragment
Michal Schmidt [Thu, 28 Jun 2012 23:50:31 +0000 (01:50 +0200)]
mount: load only if we there's mountinfo or fragment

Having information from /proc/self/mountinfo is sufficient to consider a
mount unit loaded.

When there's no mountinfo, the loading of the fragment for the mount
unit is not optional. No extra dependency links must be added when the
loading fails.

https://bugzilla.redhat.com/show_bug.cgi?id=835848

7 years agomount: split adding of extras from mount_load()
Michal Schmidt [Thu, 28 Jun 2012 23:47:24 +0000 (01:47 +0200)]
mount: split adding of extras from mount_load()

7 years agologind: add reference to new logind dbus api documentation
Lennart Poettering [Thu, 28 Jun 2012 21:51:47 +0000 (23:51 +0200)]
logind: add reference to new logind dbus api documentation

7 years agoman: add various links from man pages to appropriate wiki pages
Lennart Poettering [Thu, 28 Jun 2012 16:58:56 +0000 (18:58 +0200)]
man: add various links from man pages to appropriate wiki pages

7 years agoman: minor accuracy corrections
Lennart Poettering [Thu, 28 Jun 2012 14:34:00 +0000 (16:34 +0200)]
man: minor accuracy corrections

7 years agomain: print detected virtualiaztion at boot
Lennart Poettering [Thu, 28 Jun 2012 11:46:45 +0000 (13:46 +0200)]
main: print detected virtualiaztion at boot

7 years agonspawn: introduce new --capabilities= flag and make use of it in the nspawn test...
Lennart Poettering [Thu, 28 Jun 2012 11:44:39 +0000 (13:44 +0200)]
nspawn: introduce new --capabilities= flag and make use of it in the nspawn test case

7 years agoman: fix reference to systemd-tty-ask-password-agent
Tom Gundersen [Wed, 27 Jun 2012 23:05:08 +0000 (01:05 +0200)]
man: fix reference to systemd-tty-ask-password-agent

7 years agoupdate TODO
Lennart Poettering [Thu, 28 Jun 2012 11:20:28 +0000 (13:20 +0200)]
update TODO

7 years agotest: reenable nspawn, even if it fails for now
Lennart Poettering [Thu, 28 Jun 2012 10:14:18 +0000 (12:14 +0200)]
test: reenable nspawn, even if it fails for now

7 years agounits: automatically respawn the core services
Lennart Poettering [Thu, 28 Jun 2012 10:13:52 +0000 (12:13 +0200)]
units: automatically respawn the core services

7 years agomodules-load: use correct va_list logging function
Michal Schmidt [Wed, 27 Jun 2012 22:50:46 +0000 (00:50 +0200)]
modules-load: use correct va_list logging function

7 years agoman: extend watchdog docs a bit
Lennart Poettering [Wed, 27 Jun 2012 22:24:36 +0000 (00:24 +0200)]
man: extend watchdog docs a bit

7 years agomodules-load: fix return value
Michal Schmidt [Wed, 27 Jun 2012 22:12:47 +0000 (00:12 +0200)]
modules-load: fix return value

7 years agoupdate TODO
Lennart Poettering [Wed, 27 Jun 2012 20:38:35 +0000 (22:38 +0200)]
update TODO

7 years agoman: fix incorrectly named manpage for udev service
Dave Reisner [Wed, 27 Jun 2012 18:56:42 +0000 (14:56 -0400)]
man: fix incorrectly named manpage for udev service

The service is systemd-udev.service, not systemd-udevd.service.

7 years agounits: add explicit dependency on display-manager.service to graphical.target
Lennart Poettering [Wed, 27 Jun 2012 20:21:35 +0000 (22:21 +0200)]
units: add explicit dependency on display-manager.service to graphical.target

This replaces the symlink based dependency by an explicit one in the
unit file so that we avoid the dangling symlink when no display manager
is installed.

7 years agomodules-load: don't fail on builtin modules, better prints
Michal Schmidt [Wed, 27 Jun 2012 19:44:49 +0000 (21:44 +0200)]
modules-load: don't fail on builtin modules, better prints

Distinguish between non-existing modules, builtin modules, already
loaded modules, and modules we load.
Only the non-existing ones are treated as errors.

https://bugzilla.redhat.com/show_bug.cgi?id=817760

7 years agoman: add reference to systemd-delta to tmpfiles
Lennart Poettering [Wed, 27 Jun 2012 14:30:57 +0000 (16:30 +0200)]
man: add reference to systemd-delta to tmpfiles

7 years agoman: document preset files
Lennart Poettering [Wed, 27 Jun 2012 14:29:08 +0000 (16:29 +0200)]
man: document preset files

7 years agoman: document unit names in systemd-tmpfiles(8)
Lennart Poettering [Wed, 27 Jun 2012 13:39:51 +0000 (15:39 +0200)]
man: document unit names in systemd-tmpfiles(8)

7 years agoupdate TODO
Lennart Poettering [Wed, 27 Jun 2012 13:35:05 +0000 (15:35 +0200)]
update TODO

7 years agoudev: use the matching binary name "systemd-udevd" in log messages
Kay Sievers [Wed, 27 Jun 2012 13:27:19 +0000 (15:27 +0200)]
udev: use the matching binary name "systemd-udevd" in log messages

7 years agoman: document fstab generator
Lennart Poettering [Wed, 27 Jun 2012 13:27:29 +0000 (15:27 +0200)]
man: document fstab generator

7 years agoman: document systemd-update-generator
Lennart Poettering [Wed, 27 Jun 2012 13:12:29 +0000 (15:12 +0200)]
man: document systemd-update-generator

7 years agoman: document getty-generator
Lennart Poettering [Wed, 27 Jun 2012 13:03:02 +0000 (15:03 +0200)]
man: document getty-generator

7 years agoman: document systemd-cryptsetup-generator
Lennart Poettering [Wed, 27 Jun 2012 12:51:47 +0000 (14:51 +0200)]
man: document systemd-cryptsetup-generator

7 years agocore: rename system.preset to system-presets to follow naming scheme of other dirs...
Lennart Poettering [Wed, 27 Jun 2012 12:34:24 +0000 (14:34 +0200)]
core: rename system.preset to system-presets to follow naming scheme of other dirs in /usr/lib/systemd/

Thankfully nobody is using this yet, and presets aren't documented yet,
hence take the liberty to rename this.

7 years agoman: this time, fix Documentation= properly
Lennart Poettering [Wed, 27 Jun 2012 12:28:25 +0000 (14:28 +0200)]
man: this time, fix Documentation= properly

7 years agoman: add reference to crypttab(5) from cryptsetup units
Lennart Poettering [Wed, 27 Jun 2012 11:24:13 +0000 (13:24 +0200)]
man: add reference to crypttab(5) from cryptsetup units

7 years agoman: document /etc/crypttab
Lennart Poettering [Wed, 27 Jun 2012 11:23:12 +0000 (13:23 +0200)]
man: document /etc/crypttab

7 years agoman: document systemd-cryptsetup
Lennart Poettering [Wed, 27 Jun 2012 10:19:35 +0000 (12:19 +0200)]
man: document systemd-cryptsetup

7 years agounits: fix Documentation= tag in fsck units
Lennart Poettering [Wed, 27 Jun 2012 09:51:38 +0000 (11:51 +0200)]
units: fix Documentation= tag in fsck units

7 years agoman: fix --h vs. -h typos
Lennart Poettering [Wed, 27 Jun 2012 08:48:13 +0000 (10:48 +0200)]
man: fix --h vs. -h typos

7 years agoman: typo in systemd-analyze.xml
Shawn Landden [Wed, 27 Jun 2012 01:58:10 +0000 (18:58 -0700)]
man: typo in systemd-analyze.xml

7 years agoman: document tty-ask-password-agent
Lennart Poettering [Wed, 27 Jun 2012 00:12:39 +0000 (02:12 +0200)]
man: document tty-ask-password-agent

7 years agoman: write man page for systemd-analyze
Lennart Poettering [Tue, 26 Jun 2012 23:53:07 +0000 (01:53 +0200)]
man: write man page for systemd-analyze

7 years agoman: document systemd-ask-password-console.service and friends
Lennart Poettering [Tue, 26 Jun 2012 23:32:15 +0000 (01:32 +0200)]
man: document systemd-ask-password-console.service and friends

7 years agoman: document quotacheck
Lennart Poettering [Tue, 26 Jun 2012 23:17:49 +0000 (01:17 +0200)]
man: document quotacheck

7 years agoman: add Documentation= tag to tmp.mount
Lennart Poettering [Tue, 26 Jun 2012 23:09:51 +0000 (01:09 +0200)]
man: add Documentation= tag to tmp.mount

7 years agoman: add Documentation= tags to all kinds of missing units
Lennart Poettering [Tue, 26 Jun 2012 23:07:04 +0000 (01:07 +0200)]
man: add Documentation= tags to all kinds of missing units

7 years agoman: properly document .socket units in man page
Lennart Poettering [Tue, 26 Jun 2012 23:06:35 +0000 (01:06 +0200)]
man: properly document .socket units in man page

7 years agoman: document systemd-fsck
Lennart Poettering [Tue, 26 Jun 2012 22:37:36 +0000 (00:37 +0200)]
man: document systemd-fsck

7 years agoman: document the readahead services
Lennart Poettering [Tue, 26 Jun 2012 16:16:00 +0000 (18:16 +0200)]
man: document the readahead services

7 years agoman: document systemd-halt.service and friends
Lennart Poettering [Tue, 26 Jun 2012 15:50:29 +0000 (17:50 +0200)]
man: document systemd-halt.service and friends

7 years agoman: document systemd-suspend.service
Lennart Poettering [Tue, 26 Jun 2012 15:33:11 +0000 (17:33 +0200)]
man: document systemd-suspend.service

7 years agoman/bootup.xml: Fix grammar by adding missing verb »is«
Paul Menzel [Mon, 25 Jun 2012 21:00:38 +0000 (23:00 +0200)]
man/bootup.xml: Fix grammar by adding missing verb »is«

7 years agojournal: typo fix in sd-journal.h
David Strauss [Tue, 26 Jun 2012 10:25:11 +0000 (12:25 +0200)]
journal: typo fix in sd-journal.h

7 years agocore: make systemd.confirm_spawn=1 actually work
Lennart Poettering [Tue, 26 Jun 2012 10:16:18 +0000 (12:16 +0200)]
core: make systemd.confirm_spawn=1 actually work

This adds a timeout if the TTY cannot be acquired and makes sure we
always output the question to the console, never to the TTY of the
respective service.

7 years agoservice: pass via FAILED/DEAD before going to AUTO_RESTART
Michal Schmidt [Tue, 26 Jun 2012 09:42:52 +0000 (11:42 +0200)]
service: pass via FAILED/DEAD before going to AUTO_RESTART

With misconfigured mysql, which uses Restart=always, the following two
messages would loop indefinitely and the "systemctl start" would never
finish:
  Job pending for unit, delaying automatic restart.
  mysqld.service holdoff time over, scheduling restart.

In service_enter_dead() always set the state to SERVICE_FAILED/DEAD first
before setting SERVICE_AUTO_RESTART. This is to allow running jobs to
complete. OnFailure will be also triggered at this point, so there's no
need to do it again from service_stop() (where it was added in commit
f0c7b229).

Note that OnFailure units should better trigger only after giving up
auto-restarting, but that's for another patch to solve.

https://bugzilla.redhat.com/show_bug.cgi?id=832039

7 years agoservice: fix incorrect argument
Michal Schmidt [Tue, 26 Jun 2012 09:32:30 +0000 (11:32 +0200)]
service: fix incorrect argument

The last argument to service_enter_dead() is bool allow_restart, not
a service result.

7 years agoman: document systemd-shutdownd
Lennart Poettering [Mon, 25 Jun 2012 22:21:13 +0000 (00:21 +0200)]
man: document systemd-shutdownd

7 years agoman: document systemd-initctl
Lennart Poettering [Mon, 25 Jun 2012 22:15:59 +0000 (00:15 +0200)]
man: document systemd-initctl

7 years agoman: various minor fixes
Lennart Poettering [Mon, 25 Jun 2012 22:12:00 +0000 (00:12 +0200)]
man: various minor fixes

7 years agoman: document systemd-hostnamed
Lennart Poettering [Mon, 25 Jun 2012 22:04:50 +0000 (00:04 +0200)]
man: document systemd-hostnamed

7 years agobus: drop HaveWatchdog option
Lennart Poettering [Mon, 25 Jun 2012 22:03:43 +0000 (00:03 +0200)]
bus: drop HaveWatchdog option

Clients should always watch /dev/watchdog directly, instead of going
indirect. Let's keep our stacks small.

7 years agoman: fix typo in sysntemd-user-sessions.service page
Tom Gundersen [Mon, 25 Jun 2012 20:50:41 +0000 (22:50 +0200)]
man: fix typo in sysntemd-user-sessions.service page

7 years agoman: document localed
Lennart Poettering [Mon, 25 Jun 2012 21:55:35 +0000 (23:55 +0200)]
man: document localed

7 years agosystemctl: make sure we print URLs we can't show in systemctl in a nicely clickable way
Lennart Poettering [Mon, 25 Jun 2012 20:57:20 +0000 (22:57 +0200)]
systemctl: make sure we print URLs we can't show in systemctl in a nicely clickable way

7 years agoman: document systemd-timedated.service
Lennart Poettering [Mon, 25 Jun 2012 20:56:45 +0000 (22:56 +0200)]
man: document systemd-timedated.service

7 years agoman: document systemd-update-utmp-runlevel
Lennart Poettering [Mon, 25 Jun 2012 20:07:58 +0000 (22:07 +0200)]
man: document systemd-update-utmp-runlevel

7 years agocryptsetup: fix escaping when generating cryptsetup units
Lennart Poettering [Mon, 25 Jun 2012 18:16:15 +0000 (20:16 +0200)]
cryptsetup: fix escaping when generating cryptsetup units

7 years agotmpfiles: write /run/nologin during early boot to disallow too early user logins
Lennart Poettering [Mon, 25 Jun 2012 15:35:13 +0000 (17:35 +0200)]
tmpfiles: write /run/nologin during early boot to disallow too early user logins

systemd-user-sessoins.service will later on remove the flag file, thus
permitting user logins when the time has come.

7 years agoman: document systemd-user-sessions.service
Lennart Poettering [Mon, 25 Jun 2012 15:34:50 +0000 (17:34 +0200)]
man: document systemd-user-sessions.service

7 years agobuild-sys: make distcheck pass
Lennart Poettering [Mon, 25 Jun 2012 13:49:05 +0000 (15:49 +0200)]
build-sys: make distcheck pass

7 years agounits: rename cryptsetup@.service to systemd-cryptsetup@.service
Lennart Poettering [Mon, 25 Jun 2012 12:27:58 +0000 (14:27 +0200)]
units: rename cryptsetup@.service to systemd-cryptsetup@.service

It's also our own code, hence should have the prefix.

7 years agounits: rename halt/hibernate/kexec/poweroff/reboot/suspend to systed-xxx
Lennart Poettering [Mon, 25 Jun 2012 12:24:56 +0000 (14:24 +0200)]
units: rename halt/hibernate/kexec/poweroff/reboot/suspend to systed-xxx

These service units also execute our own code, hence rename the
accordingly and prefix them with systemd-

7 years agounits: rename fsck@.service to systemd-fsck@.service
Lennart Poettering [Mon, 25 Jun 2012 11:47:45 +0000 (13:47 +0200)]
units: rename fsck@.service to systemd-fsck@.service

The rule is that units that encapsulate our own code are prefixed with
"systemd-". Since the fsck units invoke our own code, hence add the
missing prefix. Since a long long time the fsck units didn't invoke the
naked fsck binaries anymore, and it is unlikely that this well ever
change. On the opposite: the code in systemd-fsck will probably get more
complex over time to handle fsck progress to plymouth forwarding.

Same for quotacheck (but not quotaon!)

7 years agoman: document systemd-remount-fs.service
Lennart Poettering [Mon, 25 Jun 2012 11:08:23 +0000 (13:08 +0200)]
man: document systemd-remount-fs.service

7 years agoman: document systemd-random-seed-load.service
Lennart Poettering [Mon, 25 Jun 2012 10:13:17 +0000 (12:13 +0200)]
man: document systemd-random-seed-load.service

7 years agounits: stop sleep.target when it has done its job
Lennart Poettering [Mon, 25 Jun 2012 10:01:09 +0000 (12:01 +0200)]
units: stop sleep.target when it has done its job

7 years agoman: document systemd-vconsole-setup.service
Lennart Poettering [Mon, 25 Jun 2012 10:00:43 +0000 (12:00 +0200)]
man: document systemd-vconsole-setup.service

7 years agoman: introduce bootup(7)
Lennart Poettering [Fri, 22 Jun 2012 21:14:19 +0000 (23:14 +0200)]
man: introduce bootup(7)

7 years agoman: add more references to k-c-l(7)
Lennart Poettering [Fri, 22 Jun 2012 14:59:32 +0000 (16:59 +0200)]
man: add more references to k-c-l(7)

7 years agoman: fix references to udev man page
Lennart Poettering [Fri, 22 Jun 2012 14:57:13 +0000 (16:57 +0200)]
man: fix references to udev man page

7 years agoman: add reference to bootparam(7) even though it is terribly out of date
Lennart Poettering [Fri, 22 Jun 2012 14:57:01 +0000 (16:57 +0200)]
man: add reference to bootparam(7) even though it is terribly out of date

7 years agoman: mention other initrd implementations in kernel-command-line(7)
Lennart Poettering [Fri, 22 Jun 2012 14:29:26 +0000 (16:29 +0200)]
man: mention other initrd implementations in kernel-command-line(7)

7 years agounit: drop the Names= option
Lennart Poettering [Fri, 22 Jun 2012 14:24:57 +0000 (16:24 +0200)]
unit: drop the Names= option

Names= is a source of errors, simply because alias names specified like
this only become relevant after a unit has been loaded but cannot be
used to load a unit.

Let's get rid of the confusion and drop this field. To establish alias
names peope should use symlinks, which have the the benefit of being
useful as key to load a unit, even though they are not taken into
account if unit names are listed but they haven't been explicitly
referenced before.

7 years agodbus: don't expose SysVInitPath and SysVRcndPath on the bus
Lennart Poettering [Fri, 22 Jun 2012 14:18:47 +0000 (16:18 +0200)]
dbus: don't expose SysVInitPath and SysVRcndPath on the bus

People should use systemd.pc if anything at all to determine these
directories, and people should not assume that the bus fields are part
of the supported API, so let's just drop this.

7 years agocore: remove sysv_console option
Lennart Poettering [Fri, 22 Jun 2012 11:16:28 +0000 (13:16 +0200)]
core: remove sysv_console option

This option never made much sense. It was originally intended to make
sure that the usual startup output of sysv scripts goes to the terminal.
However, since SysV scripts started from a terminal would not output to
that terminal, but rather /dev/console this effect was more often than
not actually taking place. Nowadays systemd has much nicer boot time
status output than SysV which makes the sysv output redundant. Finally,
all output of services goes to the journal anyway, and is not lost.

Hence, let's drop this option, and simplify things a bit.

7 years agosystemctl: automatically turn paths and unescaped unit names into proper unit names
Lennart Poettering [Fri, 22 Jun 2012 11:08:48 +0000 (13:08 +0200)]
systemctl: automatically turn paths and unescaped unit names into proper unit names

This makes sure that

  systemctl status /home

is implicitly translated to:

  systemctl status /home.mount

Similar, /dev/foobar becomes dev-foobar.device.

Also, all characters that cannot be part of a unit name are implicitly
escaped.

7 years agomodules-load: parse driver=/rd.driver= kernel command line option
Lennart Poettering [Fri, 22 Jun 2012 09:39:04 +0000 (11:39 +0200)]
modules-load: parse driver=/rd.driver= kernel command line option

This generalizes logic that already has been available in dracut before.

7 years agoman: refer to udevd by its service name not binary name
Lennart Poettering [Fri, 22 Jun 2012 08:38:40 +0000 (10:38 +0200)]
man: refer to udevd by its service name not binary name

Since the binary name is now hidden away in /usr/lib/ the primary user
handle for the udev service is the unit name, hence change the man page
to be available under the unit name, and make the binary name an alias
for it.

7 years agofstab-generator: support fstab=/rd.fstab= kernel cmdline
Lennart Poettering [Fri, 22 Jun 2012 08:27:05 +0000 (10:27 +0200)]
fstab-generator: support fstab=/rd.fstab= kernel cmdline

This generalizes functionality already available in dracut.

7 years agoman: add more links to kernel-command-line(7)
Lennart Poettering [Fri, 22 Jun 2012 08:13:50 +0000 (10:13 +0200)]
man: add more links to kernel-command-line(7)

7 years agocryptsetup: allow configuration of LUKS disks via the kernel cmdline
Lennart Poettering [Fri, 22 Jun 2012 08:11:06 +0000 (10:11 +0200)]
cryptsetup: allow configuration of LUKS disks via the kernel cmdline

This generalizes a bit of the functionality already available in dracut.

7 years agoman: prettify logind man page
Lennart Poettering [Thu, 21 Jun 2012 23:47:43 +0000 (01:47 +0200)]
man: prettify logind man page

7 years agoman: minor typo fixes
Lennart Poettering [Thu, 21 Jun 2012 23:38:42 +0000 (01:38 +0200)]
man: minor typo fixes

7 years agoman: document all kernel command line options we understand
Lennart Poettering [Thu, 21 Jun 2012 23:35:52 +0000 (01:35 +0200)]
man: document all kernel command line options we understand

7 years agojournal: fix sd_journal_stream_fd()
Lennart Poettering [Thu, 21 Jun 2012 22:19:13 +0000 (00:19 +0200)]
journal: fix sd_journal_stream_fd()

7 years agojournal: set the _SYSTEMD_UNIT field for messages from terminated processes
Eelco Dolstra [Thu, 21 Jun 2012 20:40:47 +0000 (16:40 -0400)]
journal: set the _SYSTEMD_UNIT field for messages from terminated processes

As described in

  https://bugs.freedesktop.org/show_bug.cgi?id=50184

the journal currently doesn't set fields such as _SYSTEMD_UNIT
properly for messages coming from processes that have already
terminated.  This means among other things that "systemctl status" may
not show some of the output of services that wrote messages just
before they exited.

This patch fixes this by having processes that log to the journal
write their unit identifier to journald when the connection to
/run/systemd/journal/stdout is opened.  Journald stores the unit ID
and uses it to fill in _SYSTEMD_UNIT when it cannot be obtained
normally (i.e. from the cgroup).  To prevent impersonating another
unit, this information is only used when the caller is root.

This doesn't fix the general problem of getting metadata about
messages from terminated processes (which requires some kernel
support), but it allows "systemctl status" and similar queries to do
the Right Thing for units that log via stdout/stderr.

7 years agoreadahead: minor code style fixes
Lennart Poettering [Thu, 21 Jun 2012 22:09:23 +0000 (00:09 +0200)]
readahead: minor code style fixes

7 years agoreadahead: make sure to close pack file before exiting, to be valgrind clean
Lennart Poettering [Thu, 21 Jun 2012 22:03:25 +0000 (00:03 +0200)]
readahead: make sure to close pack file before exiting, to be valgrind clean