Re: [PATCH] xfs: group quota is still enforced if mount v5 xfs with gqnoenforce and pqnoenforce

[Zorro Lang <zorro.lang@xxxxxxxxx>]

2016å4æ14æ 02:30ï"Eric Sandeen" <sandeen@xxxxxxxxxxx>åéï
>
> This looks fine, but I think Eryu already sent this patch on 3/31:
>
> [PATCH] xfs: add missing break in xfs_parseargs()

Hi,

Sorry for this duplicate patch... I think I missed that patch email from mail list. Please ignore this DUP patch.

Thanks,
Zorro

>
> -Eric
>
> On 4/13/16 1:22 PM, Zorro Lang wrote:
> > This's a regression bug caused by:
> >
> >Â Â2e74af0 xfs: convert mount option parsing to tokens
> >
> > xfstests xfs/299 can reproduce this bug, it will fail when do
> > gpnoenforce test, the error output as below(diff with 299.out):
> >
> >Â *** push past the soft block limit
> >Â Â[ROOT] 0 0 0 00 [--------] 3 0 0 00 [--------] 0 0 0 00 [--------]
> >Â Â-[NAME] 140 100 500 00 [--------] 4 4 10 00 [--------] 0 0 0 00 [--------]
> >Â Â+[NAME] 140 100 500 00 [7 days] 4 4 10 00 [--------] 0 0 0 00 [--------]
> >
> >Â *** push past the hard inode limit (expect EDQUOT)
> >Â Â[ROOT] 0 0 0 00 [--------] 3 0 0 00 [--------] 0 0 0 00 [--------]
> >Â Â-[NAME] 140 100 500 00 [--------] 16 4 10 00 [--none--] 0 0 0 00 [--------]
> >Â Â+[NAME] 140 100 500 00 [7 days] 10 4 10 00 [7 days] 0 0 0 00 [--------]
> >
> >Â *** push past the hard block limit (expect EDQUOT)
> >Â Â[ROOT] 0 0 0 00 [--------] 3 0 0 00 [--------] 0 0 0 00 [--------]
> >Â Â-[NAME] 540 100 500 00 [--none--] 16 4 10 00 [--none--] 0 0 0 00 [--------]
> >Â Â+[NAME] 500 100 500 00 [7 days] 10 4 10 00 [7 days] 0 0 0 00 [--------]
> >
> > It shows that group quota is still enforced. But this bug only can
> > be reproduced when mount v5 xfs with gqnoenforce and pqnoenforce
> > options. The reason as I found is a missed "break" in xfs_parseargs()
> > function.
> >
> > After deal with "pqnoenforce", it should break. But the code keep
> > running into "gquota/grpquota" lines, and make gquota enforced.
> >
> > Signed-off-by: Zorro Lang <zlang@xxxxxxxxxx>
> > ---
> >
> > Hi,
> >
> > This's a tiny bug(only one line patch). But it's belong to
> > regression bug and take me a few hours to find the problem.
> > So I hope it can be fixed:)
> >
> > Thanks,
> > Zorro
> >
> >Â fs/xfs/xfs_super.c | 1 +
> >Â 1 file changed, 1 insertion(+)
> >
> > diff --git a/fs/xfs/xfs_super.c b/fs/xfs/xfs_super.c
> > index 187e14b..f3c0ffc 100644
> > --- a/fs/xfs/xfs_super.c
> > +++ b/fs/xfs/xfs_super.c
> > @@ -350,6 +350,7 @@ xfs_parseargs(
> >Â Â Â Â Â Â Â Âcase Opt_pqnoenforce:
> >Â Â Â Â Â Â Â Â Â Â Â Âmp->m_qflags |= (XFS_PQUOTA_ACCT | XFS_PQUOTA_ACTIVE);
> >Â Â Â Â Â Â Â Â Â Â Â Âmp->m_qflags &= ~XFS_PQUOTA_ENFD;
> > +Â Â Â Â Â Â Â Â Â Â Âbreak;
> >Â Â Â Â Â Â Â Âcase Opt_gquota:
> >Â Â Â Â Â Â Â Âcase Opt_grpquota:
> >Â Â Â Â Â Â Â Â Â Â Â Âmp->m_qflags |= (XFS_GQUOTA_ACCT | XFS_GQUOTA_ACTIVE |
> >
>
> _______________________________________________
> xfs mailing list
> xfs@xxxxxxxxxxx
> http://oss.sgi.com/mailman/listinfo/xfs