Warren Togami [Wed, 10 Jun 2009 14:21:11 +0000 (10:21 -0400)]
dhclient.* files become net.* only after successful netboot.
This is an attempt to avoid confusing post-boot scripts by having them
consider only net.* files.
Warren Togami [Mon, 8 Jun 2009 20:51:13 +0000 (16:51 -0400)]
Copy netboot state into /dev/.initramfs/ before switch_root
If cmdline "rdcopystate", then copy all of /tmp/* into /dev/.initramfs/ for later debugging examination.
David Dillow [Sun, 7 Jun 2009 21:17:48 +0000 (17:17 -0400)]
NFS root: add support for default root path and %s substitution
This implements a default path of /tftpboot/%s if no path is provided,
and adds host name substitution for %s, with a fall back to the IP address,
as provided by the kernel's nfsroot handling.
The test suite is updated to test this functionality.
David Dillow [Sun, 7 Jun 2009 19:48:07 +0000 (15:48 -0400)]
NFS test: separate NFSv3 and NFSv4 tests into their own functions
The full tests take over 6 minutes to run, and commenting out unneeded
ones is time consuming -- this change lets me comment out one line to
temporarily avoid an entire class of tests.
David Dillow [Sat, 6 Jun 2009 09:09:43 +0000 (05:09 -0400)]
NFS test: additional option tests
Test additional combinations of command line and DHCP option formats. This
is by no means the complete list, but gets us started with some common ones.
David Dillow [Sat, 6 Jun 2009 05:13:47 +0000 (01:13 -0400)]
NFS test: re-add the fix for the terminal
When debugging, this is still needed as we won't be redirecting out
output to /dev/null or a file. However, we need to guard against not
having a terminal.
David Dillow [Mon, 1 Jun 2009 23:59:22 +0000 (19:59 -0400)]
Allow NFS and NBD to be built into the kernel
Probe for NFS and NBD capability before trying to load their modules
in case they are built into the kernel. Ugly use of flag files, but
avoids the need for grep to be on the image.
Warren Togami [Mon, 1 Jun 2009 22:21:40 +0000 (18:21 -0400)]
Include only wired ethernet drivers, explicitly exclude wireless.
(If you REALLY want wireless boot, you need to write your own module to configure it properly.)
David Dillow [Mon, 1 Jun 2009 04:53:52 +0000 (00:53 -0400)]
nfsroot: move to netroot framework
Parse and convert commandline options in the cmdline hook, and
fill in the missing pieces in the netroot hook. This also allows
root=dhcp to work as expected.
David Dillow [Mon, 1 Jun 2009 04:42:43 +0000 (00:42 -0400)]
netroot: add common handler for network root devices
/sbin/netroot is a jumping off point to allow various network
root devices to share infrastructure. It will loop over scriptlets
in the netroot handler, looking for a handler to run for this type
of netroot. Handlers can do choose to act based on command line
options to the kernel, or via DHCP options received on this interface.
They should massage root= into a form suitable for their handler.
David Dillow [Mon, 1 Jun 2009 04:25:09 +0000 (00:25 -0400)]
resume: add support for labeled resume devices
I noticed when doing the block root= parser that resume
didn't support labeled resume devices. Fix that up and
get rid of a dependency on /bin/echo in the process.
David Dillow [Mon, 1 Jun 2009 04:21:53 +0000 (00:21 -0400)]
cmdline: cleanup handling and block: handler
Now init checks that we understood the root= option, and can
act upon it, presenting and error if not. Also, clean up the
generation of udev rules such that we don't require /bin/echo.
Victor Lowther [Sun, 31 May 2009 02:23:55 +0000 (21:23 -0500)]
Make root on local block device test completly automated.
We no longer require any user intervention when testing dracut on
a local block device in qemu, assuming everything passes. If things fail,
we still might need to manually kill things.
Victor Lowther [Sat, 30 May 2009 21:16:34 +0000 (16:16 -0500)]
Several updates to make dracut a bit more robust.
First, add a check script to 99base to ensure that it will load its
prerequisites.
Second, disable the udev magic dracut normally uses when generating
test images -- it was causing random failures when creating the test
root filesystem, presumably due to race conditions between the
rootfs creation scripts and udev.
Third, consolidate the rootfs creation scripts into one script.
Victor Lowther [Sat, 30 May 2009 16:45:47 +0000 (11:45 -0500)]
Have udevd never try to resolve user or group ownership of devices.
This gets rid of the need to have potentially security-sensitive files
like /etc/passwd and /etc/group in the initramfs, and makes the initramfs
a bit smaller. We trust the final userspace to do The Right Thing when
assigning user/group ownership to devices.
Victor Lowther [Sat, 30 May 2009 15:52:12 +0000 (10:52 -0500)]
Fix bug in kernel module loading
We were not loading all the required kernel modules due to a bug in the
dependency checking code. This was causing us to load only the first
dependent module and ignore the rest.
Victor Lowther [Sat, 30 May 2009 03:53:11 +0000 (22:53 -0500)]
Modify kernel module installation routine to respect --hostonly.
If dracut was run with --hostonly, instmods will only load a module
into the initramfs if it is already loaded on the host machine.
This really trims the fat out of a --hostonly generated initramfs, and
eliminates the need for the kernel-modules-loaded hook.
This patch also allows a module to flag that it should only load as
a dependency by exiting 255 instead of 0. Currently, only the network module
uses this functionality.
Victor Lowther [Fri, 29 May 2009 22:55:31 +0000 (17:55 -0500)]
Don't create fully generic initramfs'es in the test framework
There is no point loading every driver available when we are not testing
network functionality and qemu does not even support the stuff we were
loading. It just wastes time and space.