[Date Prev][Date Next]   [Thread Prev][Thread Next]   [Thread Index] [Date Index] [Author Index]

Re: Ext3 partition not appearing in df output

----- Original Message -----
From: "Eugene Crosser" <crosser online ru>
To: "Andreas Dilger" <adilger turbolabs com>
Cc: <tytso mit edu>; <Andries Brouwer cwi nl>; <akpm zip com au>;
<ext3-users redhat com>
Sent: Tuesday, October 30, 2001 7:45 PM
Subject: Re: Ext3 partition not appearing in df output

> > > (You can make the huersitcs approach of identifying filesystems work
> > > but it is ugly.  However, it is possible to make its reliability
> > > asymptotically reach 100% by putting in more sophisticated filesystem
> > > checks into a library like libblkid.)
> Gentelemen, please excuse my ignorance, but I don't quite get this
pro-heuristics and counter-heuristics stance.  Am I right that the kernel
_already_ makes use of heuristics anyway when mounting root (if I understand
the code right, it just tries to mount all compiled-in filesystems one by
one).  Does not mount(8) _already_ use magic when it is not given `-t'
> I think that what we want is make mount fill /etc/mtab with the result of
_previous_ _successful_ use of heuristics.  After all, when you run
`mount -a' at boot, root is already mounted and it is known with the highest
level of accuracy which exactly filesystem it is.

Yes. Something like this is also what I think also! When using "auto" on any
other filesystem than root, it gets the correct ext3 entry in /etc/mtab.
Problem is propably as you say, the kernel mount process is too limited, and
will not update mtab to the correct value; and later on, when remounting,
mount sees that this filesystem is already happily mounted and doesn't
update the fs type entry in mtab (?).

So I believe we have two possibilities:
1) Fix kernel mounting
2) Fix remounting

Could maybe all this spawn from the unwillingness of "mount" to remount with
a different filesystem? Thus not caring about updating the filesystem type
in /etc/mtab when a remount occurs?

Maybe I am just far off and awayyy....

|  Martin Eriksson <nitrax giron wox org>
|  MSc CSE student, department of Computing Science
|  Umeå University, Sweden

[Date Prev][Date Next]   [Thread Prev][Thread Next]   [Thread Index] [Date Index] [Author Index]