Re: Can't get fs recognized as ext3
On Wed, 7 Aug 2002 13:38:06 -0500
Lance Simmons <lance@lsimmons.net> wrote:
> On Wed, Aug 07, 2002 at 01:22:17PM -0500, Jamin W. Collins wrote:
> >
> > This has been covered repeated on this list, and repeated someone
> > spouts this gibberish (no offense, but it's the equivalent of an
> > urban-legend). Once again to set the record straight: Compiling ext3
> > support as a module will work for all partitions, even root.
>
> Ext3 as a module does not work for root partitions on systems that don't
> have initrd-tools (or something similar) installed, so it's not true to
> say that it works on _all_partitions.
Please quote correctly. Leaving out relavent portions of the previous
post such as:
<omited_statement>
>The catch is that you need to use an initrd image to supply the module to
>the kernel prior to mounting the root partition.
</omited_statement>
can be misleading and incorrect. I clearly indicated in my posting that
for it to work on the root partition, one would need to use an initrd
image. Leaving this out completely changes the meaning of my post and the
accuracy of it.
Now, to claim that this requirement means that ext3 (when compiled as a
module) will not work for _all_ partitions is like saying that the kernel
doesn't work on a system because it needs a boot loader. It (ext3 as a
module) _does_ work for all partitions, it's just a question of how.
> Nor it is gibberish or an urban legend to point this out. There are
> people who want to use ext3 who have no interest in using initrd.
It is gibberish to claim that you _must_ compile it into the kernel. It's
simply not true. Will it work, yes. Is it the _only_ way it will work,
no. Give them the correct information. Which is, if they have ext3
compiled as a module and want it to support the root partition, they will
have to either use an initrd image to supply it to the kernel _or_
alternatively they can recompile it directly into the kernel.
--
Jamin W. Collins
Reply to: