Syslinux - Apr 2006 - The 'Unknown keyword in config file' problem

Hello,

the topic was already discussed in (for example)
http://syslinux.zytor.com/archives/2006-March/006586.html
but I'd like to f'up as there does not seem to be a fix yet.

I'm maintainer of a Live-CD and several users reported the problem:

| Unknown keyword in config file
| Could not find kernel image: grml

to me. The problem wasn't present with older isolinux versions (IIRC
until 3.08), but with newer versions (including 3.11) it's present
on several boxes. :-(

I don't own such a problematic box, but a workaround on most (all?)
systems seems to be just running "linux26", the name of the kernel
shipped with the ISO.

I mention the problem and the workaround in the docs of course and
people who read the docs or already know the problem can work around
the issue of course. But it's very annoying for new users because
they think either the distribution or their CD is broken. :-/
(And upgrading/fixing the BIOS isn't an option always.)

I'm very interested in finding a solution for the problem, as a
fallback to an older version of isolinux is just a workaround and
doesn't solve the problem itself. And (AFAICS) the more newer
versions of isolinux are around, the more users/distributions are
stumbling on this problem.

Please let me know if you need any further information or how I can
help you tracking down the problem.

thx && regards,
-mika-
-- 
 ,'"`.         http://www.michael-prokop.at/
(  grml.org -? Linux Live-CD for texttool-users and sysadmins
 `._,'         http://www.grml.org/

On 4/12/06, Michael Prokop <news at michael-prokop.at>
wrote:
> Hello,
>
> the topic was already discussed in (for example)
> http://syslinux.zytor.com/archives/2006-March/006586.html
> but I'd like to f'up as there does not seem to be a fix yet.
>
> I'm maintainer of a Live-CD and several users reported the problem:
>
> | Unknown keyword in config file
> | Could not find kernel image: grml
>
> to me. The problem wasn't present with older isolinux versions (IIRC
> until 3.08), but with newer versions (including 3.11) it's present
> on several boxes. :-(
>
> I don't own such a problematic box, but a workaround on most (all?)
> systems seems to be just running "linux26", the name of the
kernel
> shipped with the ISO.
>
> I mention the problem and the workaround in the docs of course and
> people who read the docs or already know the problem can work around
> the issue of course. But it's very annoying for new users because
> they think either the distribution or their CD is broken. :-/
> (And upgrading/fixing the BIOS isn't an option always.)
>
> I'm very interested in finding a solution for the problem, as a
> fallback to an older version of isolinux is just a workaround and
> doesn't solve the problem itself. And (AFAICS) the more newer
> versions of isolinux are around, the more users/distributions are
> stumbling on this problem.
>
> Please let me know if you need any further information or how I can
> help you tracking down the problem.
>
> thx && regards,
> -mika-
> --
>  ,'"`.         http://www.michael-prokop.at/
> (  grml.org -? Linux Live-CD for texttool-users and sysadmins
>  `._,'         http://www.grml.org/
>
> _______________________________________________
> SYSLINUX mailing list
> Submissions to SYSLINUX at zytor.com
> Unsubscribe or set options at:
> http://www.zytor.com/mailman/listinfo/syslinux
> Please do not send private replies to mailing list traffic.
>
>
Weird.  I remember there was troubles with certain versions where the
default statement couldn't access labels.  Might have been that one.
Either way, 3.11 is kind of old.  It seems to me like it could be
rather handy to try a newer version.  I know it's the latest
"stable"
version, but, the only problems you'll find in a beta usually tend to
be minor annoyances rather than fatal errors.

You can find the beta versions here:
http://www.kernel.org/pub/linux/utils/boot/syslinux/Testing/

If the problem still occurs in them we'll know for certain something
is up.  If it doesn't and you have patience enough to keep writing to
a rewritable and testing the various versions, it'd probably narrow
things down more to find out just where the problem stops if it does
stop so you'd know just what changed.