On Fri, Jan 30, 2009, Paul Jakma wrote:
What would be the best way to have some kind of support for retaining
old kernel/initrds and booting those? (And is anyone working on this?)
We were looking into this in Ubuntu; actually the general problem was
recovering from a breakage after a
hi,
as loic pointed out, i wrote a little (very ugly yet ... dont look to
deep at the sed commands, its really just a proof of concept weekend
hackery) script that can parse a file and display a menu on a
framebuffer or serial console ...
the script is based on the assumption that you have an
On Fri, 30 Jan 2009, Paul Jakma wrote:
Possible methods:
- APEX support for N2100 (??) (as a 2nd stage)
- A small, static Linux env in flash to act as bootloader via kexec ?
(suggested by suihkulokki on #debian-arm)
- custom N2100 bootloader
- ?
The 2nd one seems the easiest to hack together
(PS: Obviously I meant the 1st one in my original mail and the
followup-incorrection).
On Sat, 31 Jan 2009, Loïc Minier wrote:
There are many solutions to the first part, perhaps our preferred
is when the bootloader is clever enough to e.g. boot from USB or SD
card first, or even clever
hi,
Am Samstag, den 31.01.2009, 13:25 + schrieb Paul Jakma:
...
https://lists.ubuntu.com/archives/ubuntu-devel/2009-January/027229.html
https://wiki.ubuntu.com/Specs/ARMSoftbootLoader
Why build a new environment? Why not use the standard Debian initram?
What modifications are needed
On Sat, 31 Jan 2009, Oliver Grawert wrote:
i think michaels inital drive was bling with a cool gui :)
Aha. Not one of my goals :)
though the problem might be size here, you would probably need to
strip the defaults a bit to get an initramfs that matches the usual
size constraints you have
hi,
Am Samstag, den 31.01.2009, 16:12 + schrieb Paul Jakma:
though the problem might be size here, you would probably need to
strip the defaults a bit to get an initramfs that matches the usual
size constraints you have on MTD devices ... (an initramfs of the
current ubuntu
hi,
Am Samstag, den 31.01.2009, 16:12 + schrieb Paul Jakma:
One thing though, there seems to be no kexec-tools package available
- is kexec not supported on ARM? That'd be a problem..
seems its just not being built:
http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=489175
ciao
oli
On Sat, Jan 31, 2009, Paul Jakma wrote:
- have a fallback option (though, this would require modifying some
state shared between loader and booted environment which, ideally,
would be in NVRAM - is that possible with redboot?)
Not sure what you mean here; if your kernel panics, it just
On Sat, 31 Jan 2009, Loïc Minier wrote:
Not sure what you mean here; if your kernel panics, it just sits here
and doesn't give back control to the boot loader or previous kernel
AFAIK; I don't know whether you can get control back in the kexec case,
but I suspect not in all cases.
Hence the
Hi,
Are there any plans to add some kind of support for a two-stage boot
to Thecus N2100?
At present kernel+initrd are stored in flash on this platform, and so
obviously have to be updated on every change. Redboot then boots
these directly. The root is stored in the initrd (this avoids
On Fri, Jan 30, 2009 at 04:06:49PM +, Paul Jakma wrote:
Hi,
Are there any plans to add some kind of support for a two-stage boot to
Thecus N2100?
At present kernel+initrd are stored in flash on this platform, and so
obviously have to be updated on every change. Redboot then boots
12 matches
Mail list logo
