On Tue, 2015-03-10 at 00:13 +0100, Ivo De Decker wrote:
> Hi,
>
> On Mon, Mar 09, 2015 at 09:22:13PM +0100, Cyril Brulebois wrote:
> > here's another list of unblock and/or unblock-udeb for d-i. Please note
> > that I'm only not-objecting to lvm2, I didn't extensively review it.
>
> I unblocked it. Ben, is there a specific reason you didn't include the fix for
> #778828?
No, I just didn't think to look for other RC bugs.
> > Ben also mentioned initramfs-tools past week, which you might want to
> > consider. I've added Ben to Cc so that he can comment on both as
> > appropriate.
>
> It looks like this version of initramfs-tools should fix all the pending
> issues, so I unblocked it.
[...]
Just to make sure you're clear about what's changing: most of the bugs
fixed in 0.119 were regressions in 0.117 that do not currently affect
testing. So, ignoring those, the key changes from 0.116 to 0.119 are:
* Mount /usr if present in the /etc/fstab on the mounted rootfs
(Closes: #652459)
* init: Only mount /usr if the real init is systemd
* Check filesystems prior to mounting (Closes: #708000)
* local: Call local-block boot scripts to prepare additional block devices
All the above changes should reduce the risk of regressions when
switching to systemd and using a separate /usr. (Not that we don't have
any problems with initscripts and a separate /usr, but I think systemd
manages to be more aggressive in parallelising boot.)
* local: Invoke local-block scripts from the loop in local_device_setup
(Closes: #678696)
Together with the lvm2 change and earlier changes to cryptsetup, this
fixes a longstanding limitation in device discovery that made boot
unreliable for configurations such as LVM or LUKS on USB storage.
* hook-functions: Add modules for various important device types
(Closes: #762042)
This is needed for booting on some ARM platforms.
Ben.
--
Ben Hutchings
Life is what happens to you while you're busy making other plans.
- John Lennon
Attachment:
signature.asc
Description: This is a digitally signed message part