Thanks - that seems to have fixed it - although I've now hit a problem
with repquota endlessly looping - but I notice that this has already
been reported.
James Pearson
Nathan Scott wrote:
>
> hi James,
>
> On May 3, 3:36pm, James Pearson wrote:
> > Subject: Problem setting up quotas ...
> > ...
> > # edquota james
> > No filesystems with quota detected.
> >
>
> OK, this is a bug in the way the Redhat 7.1 kernel rpms
> were integrated with the XFS code. Here is a patch which
> will fix it - Eric has tested this for me & it works.
>
> It _only_ affects the kernel rpms with Redhat 7.1 changes
> - the base 2.4.2 patches, cvs trees, etc. are not affected
> and _do not_ need the patch..
>
> --- fs/dquot.c.orig Fri May 4 10:17:40 2001
> +++ fs/dquot.c Fri May 4 11:02:13 2001
> @@ -2001,7 +2001,7 @@
> type = cmd & SUBCMDMASK;
>
>
> - if ((uint) type >= MAXQUOTAS || cmds > 0x0F00 || cmds < 0x100 || cmds
> == 0x0300 ||
> + if ((uint) type >= MAXQUOTAS || cmds < 0x100 || cmds == 0x0300 ||
> cmds == 0x0400 || cmds == 0x0500)
> goto out;
>
> @@ -2049,9 +2049,6 @@
> unlock_kernel();
> return sb->s_op->quotactl(sb, cmds, type, id, addr);
> }
> -
> - if (id & ~0xFFFF)
> - goto out;
>
> ret = -EINVAL;
> switch (cmds) {
>
> > ...
> > Incidentally, the examples in /usr/share/doc/xfsprogs/README.quota and
> > the quotaon man page don't agree with the actual commands e.g.
> >
> > in README.quota the syntax to the setquota example is incorrect.
> >
> > in quotaon(8), the examples talk about a -s option to repquota and a -o
> > option to quotaoff which don't seem to exist ...
> >
>
> That's been updated in the development tree now -- thanks.
>
> cheers.
>
> --
> Nathan
|