The new option ‘skins’ allows specifying a list of directories
providing skins to be added to the MediaWiki installation. The
‘defaultSkin’ option just sets the default.
resolvconf prefers a locally running BIND resolver over the forwarders; we just
have to tell it whether we have one or not. We use 'config.services.bind.enable'
to make that decision, assuming that people are not going to configure a local
BIND that won't respond to queries on 127.0.0.1. If we run into such a (weird)
case, then we'll need to introduce another variable for that purpose which can
be set independently from 'config.services.bind.enable'.
BusyBox doesn't handle the "auto" filesystem type very well: fsck will
just ignore such filesystems, and mount will only work properly if the
required kernel module is already loaded. Therefore, use blkid to
determine the filesystem type.
Also generate an /etc/fstab in the initrd rootfs on the fly. This is
useful if you're dropped into an emergency shell since it allows you
to say "fsck /dev/sda1" or "mount /dev/sda" and have the right thing
happen.
So it turns out that BusyBox doesn't auto-load modules for filesystems
that have type "auto", e.g. it doesn't figure out that it should load
the "iso9660" module should be loaded when mounting the CD-ROM. We
don't want to give the root FS on the CD type "iso9660", because that
breaks booting from a USB stick created by unetbootin. So make sure
the "iso9660" module is loaded.
Using BusyBox instead of Bash plus a bunch of other tools gives us a
much more feature-full, yet smaller initrd. In particular, BusyBox
contains networking commands such as ip and a DHCP client, useful for
NFS boots. It's also much more convenient for rescue situations
because the shell has builtin readline support and there are many more
tools (including vi).
Upstart requires /dev/ptmx since its 1.4, and will lock up in case of it missing.
I was hitting this in the fuloong, where I don't use the nixos initrd.
svn path=/nixos/trunk/; revision=34429
I remember the 'named' log was giving annoying messages on systems not ipv6
capable (I can't recall if lacking the kernel ipv6 code or unconfigured ipv6
addresses).
svn path=/nixos/trunk/; revision=34419
are included in the manual, so this causes a different manual to be
built for each machine.
* Clean up indentation of cntlm module.
svn path=/nixos/trunk/; revision=34387
probably lots of others). The $PATH used to invoke the filter
didn't contain Ghostscript and Perl, so it silently fails. (A nice
property of CUPS is that it will just silently discard the job when
that happens, so you need to set LogLevel to "debug" to see this.)
Fortunately, CUPS now has a "SetEnv" option to set $PATH explicitly.
Also, remove config.system.path from the PATH of CUPS' Upstart job.
It seems to serve no purpose.
svn path=/nixos/trunk/; revision=34244
It complained that grub device wasn't defined, and I guess it wanted some
fileSystems too. I copied those details from the sheevaplug expression, that I
made evaluate some minutes ago.
svn path=/nixos/trunk/; revision=34238
the device has no filesystem yet. Useful in Charon deployments.
The check for an uninitialised filesystem is kind of shaky now.
svn path=/nixos/trunk/; revision=34133
sources should be stored. E.g. "nixos-checkout /home/eelco/srcs"
will check out to /home/eelco/srcs/{nixos,nixpkgs}. Probably the
/etc/nixos default should be removed eventually because /etc is a
pretty weird place to keep large source trees.
svn path=/nixos/trunk/; revision=34086
interfaces black-listed for dhcpcd via configuration.nix. I use this option to
disable DHCP for "veth*" interfaces, which are created by LXC for use inside of
virtual machines.
svn path=/nixos/trunk/; revision=34018
smart shutdown mode, Postgres waits until all
active connections have closed, which can take an
unbounded amount of time.
svn path=/nixos/trunk/; revision=33959
lines below a certain marker. This is undesirable because commands
like "ssh-copy-id" add keys to the end of the file. Instead mark
all automatically added lines individually.
svn path=/nixos/trunk/; revision=33918
were obtained from the NixOS channel. "nixos-install" copies this
to the installed system as well.
* In the installation CD, set GC_INITIAL_HEAP_SIZE to a low value for
the benefit of memory-constrained environments.
svn path=/nixos/trunk/; revision=33887
Note that for attribute-based nix-env installations, the NixOS
channel contains Nixpkgs as its "pkgs" attribute, so
$ nix-env -iA nixos.pkgs.foo
will work instead of the old
$ nix-env -iA nixpkgs_sys.foo
* Fix the path to root's channels installed in ~/.nix-defexpr.
svn path=/nixos/trunk/; revision=33823
* Change the name of the top-level derivation from "system" to
"nixos-<version>".
* Show the NixOS version in the GRUB boot menu instead of the kernel
version.
svn path=/nixos/trunk/; revision=33785
I set the grub entry for memtest to read "Memtest86", as suggested by Eelco.
I run:
svn merge -c -33692 ^/nixos/trunk
And edited the change a bit.
svn path=/nixos/trunk/; revision=33734
currently activated NixOS. This is "0.1pre-svn" when built from a
SVN tree, but contains the actual revision when installed from the
NixOS channel or from the ISO.
svn path=/nixos/trunk/; revision=33730
rather than root. Also copy all manifests into the chroot so that
any nix-pull done outside the chroot is "inherited".
svn path=/nixos/trunk/; revision=33725
dhcpcd to segfault randomly or give corrupt output. See e.g.
http://hydra.nixos.org/build/2380666/nixlog/1/raw
The problem seems to be that wicd sends a "-k" command to dhcpcd to
release the interface, which doesn't work well with a dhcpcd that
handles all interfaces in one process.
svn path=/nixos/trunk/; revision=33720
the same as the usual nixos. I think this trivial change works;
I checked the grub.cfg output generated in the iso_minimal.
svn path=/nixos/trunk/; revision=33692
because it can be overriden choosing another memtest86.
As an effect of a change in nixpgks, the isos will include
memtest86 4.0a instead of memtest86+ 4.20, only because the
former is released later, and I deduce it should work better.
svn path=/nixos/trunk/; revision=33691
* Load scsi_wait_scan after all other kernel modules to ensure that
all SCSI device nodes have been created.
* Increase the timeout for the appearance of the root device to 20
seconds.
* Do a "udevadm settle" just after the root device has appeared to
make sure that udev isn't accessing the device anymore (hopefully).
On EC2 (Xen), I've seen fsck on the root fail randomly with "device
in use" errors.
svn path=/nixos/trunk/; revision=33650
challenge-response is an authentication method that does not need the
plain text password to be emitted over the (encrypted) connection.
This is nice if you don't fully trust the server.
It is enabled (upstream) by default.
To the end user, it still looks like normal password authentication,
but instead of sending it, it is used to hash some challenge.
This means that if you don't want passwords to be used ever at all,
and just stick to public key authentication, you probably want to
disable this option too.
svn path=/nixos/trunk/; revision=33513
wpa_gui or wpa_cli.
Comes with a default wpa_supplicant.conf, which gets updated through
aforementioned utilities.
svn path=/nixos/trunk/; revision=33510
You can now set the forwardX11 config option for the ssh client and server separately.
For server, the option means "allow clients to request X11 forwarding".
For client, the option means "request X11 forwarding by default on all connections".
I don't think it made sense to couple them. I might not even run the server on some machines.
Also, I ssh to a lot of machines, and rarely want X11 forwarding. The times I want it,
I use the -X/-Y option, or set it in my ~/.ssh/config.
I also decoupled the 'XAuthLocation' logic from forwardX11.
For my case where ssh client doesn't want forwarding by default, it still wants to set the path for the cases I do need it.
As this flag is the one that pulls in X11 dependencies, I changed the minimal profile and the no-x-libs config to check that instead now.
svn path=/nixos/trunk/; revision=33407
delete routes and addresses when it quits. This causes those routes
and addresses to stick around forever, since dhcpcd won't delete
them when it runs next (even if it acquires a new lease on the same
interface). This is bad; in particular the stale (default) routes
can break networking.
The downside to removing "persistent" is that you should never ever
do "stop dhcpcd" on a remote machine configured by dhcpcd.
svn path=/nixos/trunk/; revision=33388
* Add a slash to the end of $MODULE_DIR, as expected by depmod. (Not
that running depmod from the command line is all that useful, since
you can't use it to update the tree in the Nix store. But at least
commands like "depmod -n" work now.) Reported by Kirill Elagin on
IRC.
svn path=/nixos/trunk/; revision=33312
The VirtualBox build in Nixpkgs is insecure because it uses the
"--disable-hardened" flag, which disables some checks in the
VirtualBox kernel module. Since getting rid of that flag looks like
too much work, it's better to ensure that only explicitly permitted
users have access to VirtualBox.
* Drop the 666 permission on "sonypi" because it's not clear why that
device should be world-writable.
svn path=/nixos/trunk/; revision=33301
USR1 signal before it has forked into the background (because it
will be in the start/running state immediately).
svn path=/nixos/trunk/; revision=33288
monitor the postgres process directly (so that it can be restarted
if necessary), let Upstart send SIGTERM to postgres to shut it down
gracefully. Also drop the Mediawiki references.
svn path=/nixos/trunk/; revision=33262
warning
-bash: warning: setlocale: LC_TIME: cannot change locale (en_GB.UTF8): No such file or directory
when $LC_TIME is set in environment.shellInit.
svn path=/nixos/trunk/; revision=33248
wasn't sourced in a parent shell (as determined by the environment
variable __ETC_PROFILE_DONE). This prevents overriden values of
environment variables such as $PATH from being clobbered in
subshells.
* Move all aliases to /etc/bashrc (since those are for interactive
use).
svn path=/nixos/trunk/; revision=33246
slow: calling basename in a loop somewhere has a noticable impact on
performance. We really shouldn't use bash scripts.
svn path=/nixos/trunk/; revision=33242
were redirecting output to /var/log/upstart/<job>, so it didn't work
properly.
* mountall-ip-up: send the USR1 signal to the mountall process by
looking up its PID, rather than doing "pkill -USR1 mountall". This
prevents a very subtle race condition where USR1 is delivered to a
child process of mountall (such as fsck), if pkill sees the child
just before its execve(). There is actually still a race condition
because mountall installs its USR1 handler *after* daemonising, so
mountall-ip-up could accidentally kill mountall. Should report this
to upstream.
svn path=/nixos/trunk/; revision=33236
The patch is currently being discussed on LKML and hopefully will be included
in mainline in some form in the future. Note that booting from the livecd has
to do a lot of work before anything is output to the console, so if the drive
is still busy don't assume the boot has hanged
svn path=/nixos/trunk/; revision=33235
It needs udevd to be running because the modules may require
firmware. Thanks to Mathijs and Arie for pointing this out.
svn path=/nixos/trunk/; revision=33234
longer compares the current configuration to the previous
configuration, but instead compares the current Upstart state to the
intended state. Thus, if the switch script is interrupted, running
nixos-rebuild again will resume starting/stopping Upstart jobs where
the previous run left off.
We determine if an Upstart job has changed by having the pre-start
script of each Upstart job put a symlink to its .conf file in
/var/run/upstart-jobs. So if this symlink differs from the target
of /etc/init/<job>.conf, then the job has changed. This also
prevents multiple restarts of dependent jobs. E.g., if job B has
"start on started A" and "stop on stopping A", then restarting A
will cause B to be restarted, so B shouldn't B restarted a second
time.
We only start jobs that are not running if 1) they're tasks that
have been previously run (like mountall); or 2) they're jobs that
have a "start on" condition. This seems a reasonable heuristic.
svn path=/nixos/trunk/; revision=33222
running. The user won't see it, and the "console owner" stanza
breaks VT switching and causes the X server to go to 100% CPU time.
svn path=/nixos/trunk/; revision=33221
starts the given job and waits until it's running; "stop_check"
checks that the current job hasn't been asked to stop.
svn path=/nixos/trunk/; revision=33214
modprobe.
* Move the implementation of boot.kernelModules from the udev job to
the activation script. This prevents races with the udev job.
* Drop references to the "capability" kernel module, which no longer
exists.
svn path=/nixos/trunk/; revision=33208
JOB", but it does kill the job's main process. So if the post-start
script if waiting for the job's main process to reach some state, it
may hang forever. Thus, the post-start script should monitor
whether its job has been requested to stop and exit in that case.
svn path=/nixos/trunk/; revision=33176
