The current code relies on the tmpfs mount defaulting to 1777
permissions. If it doesn't, things break badly.
This can come up when tmpfs is disabled in the kernel and ramfs
is being used instead (the kernel will alias tmpfs to ramfs).
The default permissions for ramfs is 0755.
All Linux systems need /dev to be set up,so add code to devfs to do
this. The process devfs follows is below.
1. If static_dev is yes, nothing is done.
2. if /dev is an entry in fstab it is mounted or remounted based on that
entry.
3. If /dev is not in fstab, it attempts to mount /dev as a devtmpfs or
tmpfs depending on which is defined in the kernel; devtmpfs is
preferred.
4. If neither devtmpfs nor tmpfs is defined, it assumes the user wants
static /dev and prints a warning.
X-Gentoo-Bug: 492694
X-Gentoo-Bug-URL: https://bugs.gentoo.org/show_bug.cgi?id=492694
This was requested by Debian, because the minicom software, which is
available on Debian and other distros, has a binary named runscript. We
are keeping a backward compatibility symlink for now, but this allows
Debian or any other distro to safely remove the symlink.
X-Gentoo-Bug: 494220
X-Gentoo-Bug-URL: https://bugs.gentoo.org/show_bug.cgi?id=494220
Bootmisc was running before the root file system was remounted rw in
some situations. This fixes that issue.
X-Gentoo-Bug: 493442
X-Gentoo-Bug-URL: https://bugs.gentoo.org/show_bug.cgi?id=493442
The loopback interface is supposed to be handled by the loopback
service, but sys_interfaces includes it. This causes network to try to
start it and means that network provides net even if lo is the only
interface configured.
Signed-off-by: Andrew Gregory <andrew.gregory.8@gmail.com>
The posix equivalent of the type command is "command -v", so now we use
that. Thanks to Jonathan Callen <jcallen@gentoo.org> for informing me
wrt the fix.
The localmount and mount-ro scripts were flushing pending disk writes by
calling sync twice in succession. This is no longer necessary; see the
bug report and blog post for reasons we were still doing this.
Reported-by: Patrick Lauer <patrick@gentoo.org>
X-Gentoo-Bug: 487382
X-Gentoo-Bug-URL: https://bugs.gentoo.org/show_bug.cgi?id=487382
It has been determined that it will be best for gentoo's net.* scripts
to be in a separate package to allow independent development.
This package will be called netifrc and maintained by Gentoo.
In the 3.10 kernel, EFI variables are now provided by a dedicated
filesystem that needs to be mounted.
Signed-off-by: Robin H. Johnson <robbat2@gentoo.org>
Free BSD's ifconfig outputs a netmask in the form 0xffffff00, which
was translating to 0xff.0xff.0xff.0x00,. Now we convert this to decimal
numbers before we convert it to cidr notation.
Reported-by: 4glitch@gmail.com
X-Gentoo-Bug: 460268
X-Gentoo-Bug-URL: https://bugs.gentoo.org/show_bug.cgi?id=460268
SBINDIR and BINDIR can be set independently of PREFIX. This fixes
broken shebangs in service files when SBINDIR is set to something other
than PREFIX/sbin
Signed-off-by: Andrew Gregory <andrew.gregory.8@gmail.com>
This makes binfmt.d consistent with tmpfiles.d and systemd which uses
hard coded paths for both.
Signed-off-by: Andrew Gregory <andrew.gregory.8@gmail.com>
Using setfont directly to save the font prevents breakage
when a distro stores consolefonts in a location other than
/usr/share/consolefonts such as Arch which stores them in
/usr/share/kbd/consolefonts
Signed-off-by: Andrew Gregory <andrew.gregory.8@gmail.com>
The loadkeys man page states that if you don't pass a filename loadkeys
will read from stdin. However, this is not correct, so we now pass "-"
as the filename to explicitly request stdin.
Reported-by: andi@grois.info
X-Gentoo-Bug: 457524
X-Gentoo-Bug-URL: https://bugs.gentoo.org/show_bug.cgi?id=457524
We were telling users that setting shutdown_network=YES would shut down
the network interfaces during shutdown, but this was exactly the
opposite of what we were doing. The default was YES, which was keeping
the interfaces active.
This keeps the default behavior, but renames the setting to keep_network
which more accurately describes its function, and instructs users to set
it to NO if they want the network interfaces to go down.
This setting, already in use in newnet, allows the user to control
whether network interfaces are stopped when the system shuts down. By
default, under newnet, they are not, so I am making oldnet have the same
default.
A side-affect of this is that in the default configuration this fixes
bugs like the one below.
Reported-by: jerryfleming2006@gmail.com
X-Gentoo-Bug: 259183
X-Gentoo-Bug-URL: https://bugs.gentoo.org/show_bug.cgi?id=259183
rkhunter thinks OpenRC is a rootkit because of the hidefirstrout
variable. This has been renamed to hideFirstroute in order to get past
rkhunter.
I realize this is not an openrc bug. In this case though I do not have a
problem renaming the variable.
Reported-by: ago@gentoo.org
X-Gentoo-Bug: 339714
X-Gentoo-Bug-URL: https://bugs.gentoo.org/show_bug.cgi?id=339714
The /run directory is a mount point for a tmpfs and should not contain
any files or directories. This cleans out the /run/openrc
symlink and any other files which were incorrectly placed in /run.
Thanks to Ian Stakenvicius for pointing out this solution.
For devices that are always connected (e.g. ethernet cards), the current
carrier always wastes time by sleeping for 1 second. This is because the
code sleeps first, then checks for carrier. Invert the order so that we
return quickly for devices already active. For devices which are not yet
up, there shouldn't be any real difference.
Signed-off-by: Mike Frysinger <vapier@gentoo.org>
Add a test when localmount is started to determine if /usr is mounted
from inside an initramfs for Linux systems. If it is not, we can unmount it when
localmount stops.
On *bsd systems, we always unmount /usr if it is separate.
Reported-by: ryao@gentoo.org
In a pathname expansion, specifically single-character match, the pure
POSIX specification uses '!' as the Negation character where a regular
expression would normally be '^'.
Regular expression: "a[^a]a"
Pathname expansion pattern: "a[!a]a"
Reference:
IEEE Std 1003.1, 2004 Edition
2. Shell Command Language
2.13 Pattern Matching Notation
2.13.1 Patterns Matching a Single Character
> The description of basic regular expression bracket expressions in the
> Base Definitions volume of IEEE Std 1003.1-2001, Section 9.3.5, RE
> Bracket Expression shall also apply to the pattern bracket expression,
> except that the exclamation mark character ( '!' ) shall replace the
> circumflex character ( '^' ) in its role in a "non-matching list" in
> the regular expression notation. A bracket expression starting with an
> unquoted circumflex character produces unspecified results.
Signed-off-by: Robin H. Johnson <robbat2@gentoo.org>
Initially, we were creating tmpfiles entries in the sysinit runlevel and
again in the boot runlevel. Systemd runs the --create and --remove
options in one service called systemd-tmpfiles-setup after the local
file systems are mounted. Now we have a service called tmpfiles.setup
which emulates this.
This also closes the bug mentioned below, since we were originally
writing to files that were on read-only file systems and that were not
available.
Reported-by: <devurandom@gmx.net>
X-Gentoo-Bug: 439012
X-Gentoo-Bug-URL: https://bugs.gentoo.org/show_bug.cgi?id=439012
Some types of interfaces do not have a carrier, so it doesn't make sense
to automatically wait for one.
Reported-by: <rose@rz.uni-potsdam.de>
X-Gentoo-Bug: 438970
X-Gentoo-Bug-URL: https://bugs.gentoo.org/show_bug.cgi?id=438970
There were references in the devfs script to mdev, udev and
udev-mount. These all provide the virtuals dev and dev-mount; that is
how we should refer to them.
I believe in the discussion I had with Tony and Robin about this, we
were going to change the "use" line to "need". However, after thinking
that over, I'm not comfortable doing so because someone could be running
a static /dev with no device manager.
Reported-by: <tokiclover@gmail.com>
X-Gentoo-Bug: 438932
X-Gentoo-Bug-URL: https://bugs.gentoo.org/show_bug.cgi?id=438932
Now that the tmpfiles.d code is more tested, actually call it from
init.d. It assumes that /run is already available when it runs.
Please note it runs TWICE.
- During sysinit, ideally just after /dev/shm is created, but before
udev has started. After udev is also acceptable, but not ideal.
- During boot, ideally just after localmount has completed.
Signed-off-by: Robin H. Johnson <robbat2@gentoo.org>
Using the new dev-mount virtual, with udev-mount included until new udev
version is rolled out, we run devfs earlier now, before udev/mdev.
It only needs (u)dev-mount before it, so that /dev is mounted.
This opens the way for tmpfiles.d, which needs to be sandwiched in the
middle.
Signed-off-by: Robin H. Johnson <robbat2@gentoo.org>
Urandom should not run in lxc containers since it is provided by the
host.
Reported-by: <walter@pratyeka.org>
X-Gentoo-Bug: 436270
X-Gentoo-Bug-URL: http://bugs.gentoo.org/436270
Since nfs and nfs4 file systems require extra daemons to be running on
the client to function properly, netmount should not try to handle these
file systems.
Reported-by: <devurandom@gmx.net>
X-Gentoo-Bug: 427996
X-Gentoo-Bug-URL: http://bugs.gentoo.org/show_bug.cgi?id=427996
This is needed in case of a read-only root filesystem such as a diskless
client.
Reported-by: <walter@pratyeka.org>
X-Gentoo-Bug: 430382
X-Gentoo-Bug-URL: http://bugs.gentoo.org/show_bug.cgi?id=430382
If CONFIG_SWAP is turned off in the kernel, this file may not exist. In
that case, we should not try to read from it.
reported-by: <walter@pratyeka.org>
X-Gentoo-Bug: 430378
X-Gentoo-Bug-URL: https://bugs.gentoo.org/show_bug.cgi?id=430378
The OpenRC upstream default network stack was changed, but there was no
reason to change it. Now since we have the MKNET build switch, it is
easy for the gentoo ebuild to install oldnet by default.
The upstream default is newnet.
The MKNET variable can be used to select the network stack you want to
build and install with OpenRC.
The current default is the gentoo "oldnet" stack. If you want to install
the OpenRC newnet stack, use MKNET=newnet on the make command line.
This reverts commit 5994e55937.
There are situations where these scripts can be useful, so I am bringing
them back. Also, I want to start discussions about simplifying the
OpenRC network stack.
On linux systems, fsck was not taking into account which filesystems
were local or remote. This commit adds the -t option, with an
appropriate value, to the fsck call so that remote file systems are not
checked.
reported-by: Vladimir Berezhnoy <non7top@gmail.com>
X-Gentoo-Bug: 408363
X-Gentoo-Bug-URL: http://bugs.gentoo.org/show_bug.cgi?id=408363
Make the stop function in localmount only unmount file systems when the
system is going down.
reported-by: Alexey Prokopchuk <alexpro@homelan.lg.ua>
X-Gentoo-Bug: 407167
X-Gentoo-Bug-URL: https://bugs.gentoo.org/show_bug.cgi?id=407167
The -q option is not implemented by BusyBox, so instead of using that, make
it so that the standard error is caught, but standard output is thrown
away.
Note: the ordered behaviour of redirection is part of POSIX so we should be
on the safe side with this change, as first we duplicate the output
descriptor to be used as stderr, then we change the output descriptor to
point to NULL.
Signed-off-by: Diego Elio Pettenò <flameeyes@flameeyes.eu>
This reverts commit 06d6701785.
After researching this further, we do need this commit. The other issue
is a bug in fstabinfo which will also need to be fixed.
There are several reasons for going back to this:
- fstabinfo doesn't work with labels in fstab
- when a device is passed to swapon individually, the swap priority from
fstab is not honored.
- We also add the -e option to only activate available swap devices.
reported-by: Duncan <1i5t5.duncan@cox.net>
X-Gentoo-Bug: 405021
X-Gentoo-Bug-URL: https://bugs.gentoo.org/show_bug.cgi?id=405021
I was informed that "provide net" should mean that there is network
communication outside the local computer. In this case, the loopback
interface can't "provide net", but there needs to be a way for other
processes to know that the loopback interface is active.
To this end, this commit makes the loopback "provide lo" and all other
interfaces "provide net".
Previously, the default on linux systems was to not set the hardware
clock to match the system clock during shutdown.
This changes that default to be consistent with *bsd and swclock.
