On Mon, Jan 29, 2018 at 03:00:04PM -0800, Darrick J. Wong wrote:
> On Mon, Jan 29, 2018 at 07:26:38AM -0500, Brian Foster wrote:
> > On Fri, Jan 26, 2018 at 05:10:34PM -0800, Darrick J. Wong wrote:
> > > From: Darrick J. Wong <darrick.wong@xxxxxxxxxx>
> > >
> > > Since the CoW fork only exists in memory, it is incorrect to update the
> > > on-disk quota block counts when we modify the CoW fork. Unlike the data
> > > fork, even real extents in the CoW fork are only delalloc-style
> > > reservations (on-disk they're owned by the refcountbt) so they must not
> > > be tracked in the on disk quota info. Ensure the i_delayed_blks
> > > accounting reflects this too.
> > >
> > > Signed-off-by: Darrick J. Wong <darrick.wong@xxxxxxxxxx>
> > > ---
> >
> > Patch doesn't apply.. I'm guessing because xfs_bmap_btalloc_accounting()
> > doesn't exist. Is this series incomplete, or perhaps not consistent with
> > for-next..?
>
> Yeah, sorry, I think I forgot to send the middle patch when I sent this
> series....
>
> ...no, wait, "xfs: refactor accounting updates out of xfs_bmap_btalloc"
> was already reviewed, so these last few patches were supposed to apply
> atop that, and that went on top of for-next.
>
Hmm.. I think I saw that a day or two ago, but then for-next moved
forward/backward and when I tried to apply this series in full, the head
of for-next was commit 75d4a13b ("xfs: fix non-debug build compiler
warnings"), which is behind the original refactoring patch. :/
Annnyways.. it looks like these are all merged now as of this morning,
so moving on..
Brian
> --D
>
> >
> > Brian
> >
> > > fs/xfs/libxfs/xfs_bmap.c | 32 ++++++++++++++++++++++++++++++--
> > > fs/xfs/xfs_reflink.c | 15 +++++++++++----
> > > 2 files changed, 41 insertions(+), 6 deletions(-)
> > >
> > >
> > > diff --git a/fs/xfs/libxfs/xfs_bmap.c b/fs/xfs/libxfs/xfs_bmap.c
> > > index 4582f55..cad21fd 100644
> > > --- a/fs/xfs/libxfs/xfs_bmap.c
> > > +++ b/fs/xfs/libxfs/xfs_bmap.c
> > > @@ -3343,8 +3343,35 @@ xfs_bmap_btalloc_accounting(
> > > struct xfs_bmalloca *ap,
> > > struct xfs_alloc_arg *args)
> > > {
> > > - if (!(ap->flags & XFS_BMAPI_COWFORK))
> > > - ap->ip->i_d.di_nblocks += args->len;
> > > + if (ap->flags & XFS_BMAPI_COWFORK) {
> > > + /*
> > > + * COW fork blocks are in-core only and thus are treated as
> > > + * in-core quota reservation (like delalloc blocks) even when
> > > + * converted to real blocks. The quota reservation is not
> > > + * accounted to disk until blocks are remapped to the data
> > > + * fork. So if these blocks were previously delalloc, we
> > > + * already have quota reservation and there's nothing to do
> > > + * yet.
> > > + */
> > > + if (ap->wasdel)
> > > + return;
> > > +
> > > + /*
> > > + * Otherwise, we've allocated blocks in a hole. The transaction
> > > + * has acquired in-core quota reservation for this extent.
> > > + * Rather than account these as real blocks, however, we reduce
> > > + * the transaction quota reservation based on the allocation.
> > > + * This essentially transfers the transaction quota reservation
> > > + * to that of a delalloc extent.
> > > + */
> > > + ap->ip->i_delayed_blks += args->len;
> > > + xfs_trans_mod_dquot_byino(ap->tp, ap->ip, XFS_TRANS_DQ_RES_BLKS,
> > > + -(long)args->len);
> > > + return;
> > > + }
> > > +
> > > + /* data/attr fork only */
> > > + ap->ip->i_d.di_nblocks += args->len;
> > > xfs_trans_log_inode(ap->tp, ap->ip, XFS_ILOG_CORE);
> > > if (ap->wasdel)
> > > ap->ip->i_delayed_blks -= args->len;
> > > @@ -4820,6 +4847,7 @@ xfs_bmap_del_extent_cow(
> > > xfs_iext_insert(ip, icur, &new, state);
> > > break;
> > > }
> > > + ip->i_delayed_blks -= del->br_blockcount;
> > > }
> > >
> > > /*
> > > diff --git a/fs/xfs/xfs_reflink.c b/fs/xfs/xfs_reflink.c
> > > index 85a119e..c4f0aff 100644
> > > --- a/fs/xfs/xfs_reflink.c
> > > +++ b/fs/xfs/xfs_reflink.c
> > > @@ -599,10 +599,6 @@ xfs_reflink_cancel_cow_blocks(
> > > del.br_startblock, del.br_blockcount,
> > > NULL);
> > >
> > > - /* Update quota accounting */
> > > - xfs_trans_mod_dquot_byino(*tpp, ip, XFS_TRANS_DQ_BCOUNT,
> > > - -(long)del.br_blockcount);
> > > -
> > > /* Roll the transaction */
> > > xfs_defer_ijoin(&dfops, ip);
> > > error = xfs_defer_finish(tpp, &dfops);
> > > @@ -613,6 +609,13 @@ xfs_reflink_cancel_cow_blocks(
> > >
> > > /* Remove the mapping from the CoW fork. */
> > > xfs_bmap_del_extent_cow(ip, &icur, &got, &del);
> > > +
> > > + /* Remove the quota reservation */
> > > + error = xfs_trans_reserve_quota_nblks(NULL, ip,
> > > + -(long)del.br_blockcount, 0,
> > > + XFS_QMOPT_RES_REGBLKS);
> > > + if (error)
> > > + break;
> > > } else {
> > > /* Didn't do anything, push cursor back. */
> > > xfs_iext_prev(ifp, &icur);
> > > @@ -795,6 +798,10 @@ xfs_reflink_end_cow(
> > > if (error)
> > > goto out_defer;
> > >
> > > + /* Charge this new data fork mapping to the on-disk quota. */
> > > + xfs_trans_mod_dquot_byino(tp, ip, XFS_TRANS_DQ_DELBCOUNT,
> > > + (long)del.br_blockcount);
> > > +
> > > /* Remove the mapping from the CoW fork. */
> > > xfs_bmap_del_extent_cow(ip, &icur, &got, &del);
> > >
> > >
> > > --
> > > To unsubscribe from this list: send the line "unsubscribe linux-xfs" in
> > > the body of a message to majordomo@xxxxxxxxxxxxxxx
> > > More majordomo info at http://vger.kernel.org/majordomo-info.html
> > --
> > To unsubscribe from this list: send the line "unsubscribe linux-xfs" in
> > the body of a message to majordomo@xxxxxxxxxxxxxxx
> > More majordomo info at http://vger.kernel.org/majordomo-info.html
> --
> To unsubscribe from this list: send the line "unsubscribe linux-xfs" in
> the body of a message to majordomo@xxxxxxxxxxxxxxx
> More majordomo info at http://vger.kernel.org/majordomo-info.html
--
To unsubscribe from this list: send the line "unsubscribe linux-xfs" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
