| Commit message (Collapse) | Author | Age | Files | Lines |
|
|
|
| |
minor fix that makes it work with bash
|
|
|
|
|
|
|
|
| |
on lxc we dont have the boot device available so extlinux will fail.
We cannot check for `openrc --sys` because apk run the trigger in chroot
and grsecurity will prevent openrc access /proc/1/environment.
So we simply add a --warn-only option and use that from trigger.
|
| |
|
| |
|
|
|
|
| |
ref #3273
|
| |
|
|
|
|
|
| |
the vanilla kernel does not have any -* suffix so it needs special
handling.
|
|
|
|
| |
fixes #2761
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
On Sun, Jun 30, 2013 at 02:13:29PM +0200, Natanael Copa wrote:
> On Sat, 29 Jun 2013 11:49:05 -0400
> Dubiousjim <dubiousjim@gmail.com> wrote:
>
> > While testing that, I discovered also that HDT is no longer working on
> > my machine.
>
> copy libgpl.c32 and libmenu.c32 in addition to hdt.c32 to /boot.
>
> (syslinux now uses elf format so you can use readelf or scanelf -n to
> troubleshoot this kind of issues)
Oh great! I've verified now that this works, and also verified that syslinux properly detects the sha256 and sha512 passwords.
This patch updates the usage message for HDT. Based on current master, ignoring patch I sent a few minutes ago.
|
|
|
|
| |
and trivial fix for comment in conf
|
|
|
|
|
|
| |
Don't bother mention salt or other crypto algorithms.
Also wrap the lines with comments.
|
| |
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
We allow a password to be set in /etc/update-extlinux.conf. Instructions for
generating this are in /etc/update-extlinux.conf. For reference, here is another
(equivalent) way to generate the MD5 password: openssl passwd -1 -salt yy pass
If one sets a password, one will presumably want to make
/etc/update-extlinux.conf world-unreadable. We don't do that for you; however
we do make sure when a password is present to make the /boot/extlinux.conf
files we generate be world-unreadable.
Of the auto-generated entries, only HDT (if this is generated) is now
configured to respect the password; however, you can include "MENU PASSWD" in
any entries you put in /etc/update-extlinux.d/.
For example, I configure my BIOS to only boot from the internal drive, but I
have an entry in /etc/update-extlinux.d that permits chain-booting from a USB
key, and I have this entry configured to also require the password. (The BIOS
is also passworded, so that these settings can't be changed willy-nilly.)
Conflicts:
main/syslinux/update-extlinux.conf
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
HDT provides a curses-like interface to display lots of hardware info
about your machine at bootloader time.
We don't install /boot/hdt.c32, but if it's present (it can be copied from
/usr/share/syslinux/hdt.c32), we add a menu entry for it---in preference to,
rather than in addition to, memtest, since HDT has a menu entry which invokes
memtest.
Using HDT to its full capacity requires finding or generating modules.pcimap
and pci.ids files for your machine, and installing them in /boot. We might want
to document this, which I don't here (but the online docs for HDT do). These
aren't required to use other functionality of HDT; and it's pretty useful
already without those.
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
We don't install /boot/reboot.c32, but if it's present (it can be copied from
/usr/share/syslinux/reboot.c32), we add a menu entry for it---just like
with memtest.
Add a comment to /etc/update-extlinux.conf stating this.
It'd already be possible to get a reboot entry using the
/etc/update-extlinux.d/ folder, but this patch provides a more intelligent
framework with nicer layout.
The syslinux sourceball also provides a shutdown module, but I couldn't get this to
work and looking at the sources reveals it to be for machines with APM enabled.
Not sure how many machines that applies to anymore.
|
|
|
|
|
|
| |
Add a blank line between any entries added from /etc/update-extlinux.d/*.
Move the memtest entry (if any) after a separator line, below these.
|
| |
|
| |
|
|
|
|
|
|
|
|
|
|
| |
If xen.gz is present, there are multiple kernels, and later ones lack an
initramfs, they'll currently be (wrongly) configured to use a initramfs anyway.
Fixed.
We also refactor the check for /boot/initramfs-$tag, and configure this using a
separate INITRD line, instead of adding it to the APPEND line. In passing, this
also closes a second might-use-a-stale-initramfs bug.
|
| |
|
| |
|
| |
|
| |
|
| |
|
| |
|
| |
|
| |
|
| |
|
| |
|
|
|
|
| |
needs util-linux-dev
|
| |
|
|
|
|
| |
so extlinux.conf don't end up as world writeable
|
| |
|
| |
|
| |
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
- rename /etc/extlinux-conf to /etc/update-extlinux.conf
- Do not exit with fail if update-extlinux.conf is missing
- Always warn if root= is not defined in update-extlinux.conf
- Try harder to detect the root device by parsing /proc/mounts
- Exit with error if we cannot detect the Root device
- Rename fancy_menu to vesa_menu
- Unifiy the code generating vesa/standard menus
- Keep a backup of old extlinux.conf
- add "overwrite" config option so it is possible to skip overwriting
the extlinux.conf.
- Added a post-upgrade that imports current extlinux.conf to
update-extlinux.conf
|
|
|
|
|
| |
It has now destroyed my extlinux.conf and caused my box unbootable
atleast twice.
|
| |
|
| |
|
| |
|
| |
|
| |
|
|
|
|
| |
this is needed for apk-tools-2.1 migration
|
| |
|
| |
|
| |
|