linux-xfs.vger.kernel.org archive mirror
 help / color / mirror / Atom feed
From: "Darrick J. Wong" <darrick.wong@oracle.com>
To: Brian Foster <bfoster@redhat.com>
Cc: david@fromorbit.com, linux-xfs@vger.kernel.org,
	Christoph Hellwig <hch@lst.de>
Subject: Re: [PATCH 08/63] xfs: account for the refcount btree in the alloc/free log reservation
Date: Thu, 29 Sep 2016 14:18:23 -0700	[thread overview]
Message-ID: <20160929211823.GH14092@birch.djwong.org> (raw)
In-Reply-To: <20160928162005.GC8852@bfoster.bfoster>

On Wed, Sep 28, 2016 at 12:20:05PM -0400, Brian Foster wrote:
> On Tue, Sep 27, 2016 at 07:54:20PM -0700, Darrick J. Wong wrote:
> > Every time we allocate or free an extent, we might need to split the
> > refcount btree.  Reserve some blocks in the transaction to handle
> > this possibility.
> > 
> 
> I'm probably not far along enough to grok how this all works, but I'm a
> bit confused by the requirement here. Why is generic block
> allocation/free a refcountbt matter? E.g., the function below is used
> for things like inode chunk allocation and whatnot, aren't those
> irrelevant to the refcountbt?

Overnight tests seem to run ok, so I'll remove this patch.  If it
becomes a problem we can always put it back in.

--D

> 
> Brian
> 
> > (Reproduced by generic/167 over NFS atop XFS)
> > 
> > Signed-off-by: Christoph Hellwig <hch@lst.de>
> > [darrick.wong@oracle.com: add commit message]
> > Signed-off-by: Darrick J. Wong <darrick.wong@oracle.com>
> > ---
> >  fs/xfs/libxfs/xfs_trans_resv.c |    5 ++++-
> >  1 file changed, 4 insertions(+), 1 deletion(-)
> > 
> > 
> > diff --git a/fs/xfs/libxfs/xfs_trans_resv.c b/fs/xfs/libxfs/xfs_trans_resv.c
> > index 7c840e1..a59838f 100644
> > --- a/fs/xfs/libxfs/xfs_trans_resv.c
> > +++ b/fs/xfs/libxfs/xfs_trans_resv.c
> > @@ -67,7 +67,8 @@ xfs_calc_buf_res(
> >   * Per-extent log reservation for the btree changes involved in freeing or
> >   * allocating an extent.  In classic XFS there were two trees that will be
> >   * modified (bnobt + cntbt).  With rmap enabled, there are three trees
> > - * (rmapbt).  The number of blocks reserved is based on the formula:
> > + * (rmapbt).  With reflink, there are four trees (refcountbt).  The number of
> > + * blocks reserved is based on the formula:
> >   *
> >   * num trees * ((2 blocks/level * max depth) - 1)
> >   *
> > @@ -83,6 +84,8 @@ xfs_allocfree_log_count(
> >  	blocks = num_ops * 2 * (2 * mp->m_ag_maxlevels - 1);
> >  	if (xfs_sb_version_hasrmapbt(&mp->m_sb))
> >  		blocks += num_ops * (2 * mp->m_rmap_maxlevels - 1);
> > +	if (xfs_sb_version_hasreflink(&mp->m_sb))
> > +		blocks += num_ops * (2 * mp->m_refc_maxlevels - 1);
> >  
> >  	return blocks;
> >  }
> > 
> > --
> > To unsubscribe from this list: send the line "unsubscribe linux-xfs" in
> > the body of a message to majordomo@vger.kernel.org
> > More majordomo info at  http://vger.kernel.org/majordomo-info.html

  parent reply	other threads:[~2016-09-29 21:18 UTC|newest]

Thread overview: 105+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2016-09-28  2:53 [PATCH v9 00/63] xfs: add reflink and dedupe support Darrick J. Wong
2016-09-28  2:53 ` [PATCH 01/63] vfs: support FS_XFLAG_COWEXTSIZE and get/set of CoW extent size hint Darrick J. Wong
2016-09-29 16:48   ` Christoph Hellwig
2016-09-28  2:53 ` [PATCH 02/63] xfs: return an error when an inline directory is too small Darrick J. Wong
2016-09-28 16:19   ` Brian Foster
2016-09-29 16:48   ` Christoph Hellwig
2016-09-28  2:53 ` [PATCH 03/63] xfs: define tracepoints for refcount btree activities Darrick J. Wong
2016-09-29 16:49   ` Christoph Hellwig
2016-09-28  2:53 ` [PATCH 04/63] xfs: introduce refcount btree definitions Darrick J. Wong
2016-09-28  2:54 ` [PATCH 05/63] xfs: refcount btree add more reserved blocks Darrick J. Wong
2016-09-28  2:54 ` [PATCH 06/63] xfs: define the on-disk refcount btree format Darrick J. Wong
2016-09-28 16:20   ` Brian Foster
2016-09-28 18:35     ` Darrick J. Wong
2016-09-28  2:54 ` [PATCH 07/63] xfs: add refcount btree support to growfs Darrick J. Wong
2016-09-28  2:54 ` [PATCH 08/63] xfs: account for the refcount btree in the alloc/free log reservation Darrick J. Wong
2016-09-28 16:20   ` Brian Foster
2016-09-28 19:45     ` Darrick J. Wong
2016-09-29 21:18     ` Darrick J. Wong [this message]
2016-09-29 23:13       ` Darrick J. Wong
2016-09-28  2:54 ` [PATCH 09/63] xfs: add refcount btree operations Darrick J. Wong
2016-09-28 16:20   ` Brian Foster
2016-09-28 18:46     ` Darrick J. Wong
2016-09-28  2:54 ` [PATCH 10/63] xfs: create refcount update intent log items Darrick J. Wong
2016-09-28 16:20   ` Brian Foster
2016-09-28 18:47     ` Darrick J. Wong
2016-09-29 16:52   ` Christoph Hellwig
2016-09-29 17:44     ` Darrick J. Wong
2016-09-28  2:54 ` [PATCH 11/63] xfs: log refcount intent items Darrick J. Wong
2016-09-29 16:56   ` Christoph Hellwig
2016-09-29 20:48     ` Darrick J. Wong
2016-09-28  2:54 ` [PATCH 12/63] xfs: adjust refcount of an extent of blocks in refcount btree Darrick J. Wong
2016-09-29 14:44   ` Brian Foster
2016-09-29 19:03     ` Darrick J. Wong
2016-09-30 11:59       ` Brian Foster
2016-09-30 18:27         ` Darrick J. Wong
2016-09-30 19:23           ` Brian Foster
2016-09-28  2:54 ` [PATCH 13/63] xfs: connect refcount adjust functions to upper layers Darrick J. Wong
2016-09-28  2:55 ` [PATCH 14/63] xfs: adjust refcount when unmapping file blocks Darrick J. Wong
2016-09-28  2:55 ` [PATCH 15/63] xfs: add refcount btree block detection to log recovery Darrick J. Wong
2016-09-28  2:55 ` [PATCH 16/63] xfs: refcount btree requires more reserved space Darrick J. Wong
2016-09-28  2:55 ` [PATCH 17/63] xfs: introduce reflink utility functions Darrick J. Wong
2016-09-28  2:55 ` [PATCH 18/63] xfs: create bmbt update intent log items Darrick J. Wong
2016-09-28  2:55 ` [PATCH 19/63] xfs: log bmap intent items Darrick J. Wong
2016-09-28  2:55 ` [PATCH 20/63] xfs: map an inode's offset to an exact physical block Darrick J. Wong
2016-09-28  2:55 ` [PATCH 21/63] xfs: pass bmapi flags through to bmap_del_extent Darrick J. Wong
2016-09-28  2:55 ` [PATCH 22/63] xfs: implement deferred bmbt map/unmap operations Darrick J. Wong
2016-09-28  2:56 ` [PATCH 23/63] xfs: when replaying bmap operations, don't let unlinked inodes get reaped Darrick J. Wong
2016-09-28  2:56 ` [PATCH 24/63] xfs: return work remaining at the end of a bunmapi operation Darrick J. Wong
2016-09-28  2:56 ` [PATCH 25/63] xfs: define tracepoints for reflink activities Darrick J. Wong
2016-09-28  2:56 ` [PATCH 26/63] xfs: add reflink feature flag to geometry Darrick J. Wong
2016-09-28  2:56 ` [PATCH 27/63] xfs: don't allow reflinked dir/dev/fifo/socket/pipe files Darrick J. Wong
2016-09-28  2:56 ` [PATCH 28/63] xfs: introduce the CoW fork Darrick J. Wong
2016-09-28  2:56 ` [PATCH 29/63] xfs: support bmapping delalloc extents in " Darrick J. Wong
2016-09-28  2:56 ` [PATCH 30/63] xfs: create delalloc extents in " Darrick J. Wong
2016-09-28  2:56 ` [PATCH 31/63] xfs: support allocating delayed " Darrick J. Wong
2016-09-28  2:57 ` [PATCH 32/63] xfs: allocate " Darrick J. Wong
2016-09-28  2:57 ` [PATCH 33/63] xfs: support removing extents from " Darrick J. Wong
2016-09-28  2:57 ` [PATCH 34/63] xfs: move mappings from cow fork to data fork after copy-write Darrick J. Wong
2016-09-28  2:57 ` [PATCH 35/63] xfs: report shared extent mappings to userspace correctly Darrick J. Wong
2016-09-28  2:57 ` [PATCH 36/63] xfs: implement CoW for directio writes Darrick J. Wong
2016-09-28  2:57 ` [PATCH 37/63] xfs: cancel CoW reservations and clear inode reflink flag when freeing blocks Darrick J. Wong
2016-09-29 17:01   ` Christoph Hellwig
2016-09-29 20:23     ` Darrick J. Wong
2016-09-28  2:57 ` [PATCH 38/63] xfs: cancel pending CoW reservations when destroying inodes Darrick J. Wong
2016-09-28  2:57 ` [PATCH 39/63] xfs: store in-progress CoW allocations in the refcount btree Darrick J. Wong
2016-09-28  2:57 ` [PATCH 40/63] xfs: reflink extents from one file to another Darrick J. Wong
2016-09-28  2:58 ` [PATCH 41/63] xfs: add clone file and clone range vfs functions Darrick J. Wong
2016-09-29 17:03   ` Christoph Hellwig
2016-09-28  2:58 ` [PATCH 42/63] xfs: add dedupe range vfs function Darrick J. Wong
2016-09-29 17:03   ` Christoph Hellwig
2016-09-29 17:49     ` Darrick J. Wong
2016-09-28  2:58 ` [PATCH 43/63] xfs: teach get_bmapx about shared extents and the CoW fork Darrick J. Wong
2016-09-29 17:05   ` Christoph Hellwig
2016-09-29 17:40     ` Darrick J. Wong
2016-09-29 19:51       ` Christoph Hellwig
2016-09-30  0:18         ` Dave Chinner
2016-09-30  1:50           ` Darrick J. Wong
2016-09-28  2:58 ` [PATCH 44/63] xfs: swap inode reflink flags when swapping inode extents Darrick J. Wong
2016-09-28  2:58 ` [PATCH 45/63] xfs: unshare a range of blocks via fallocate Darrick J. Wong
2016-09-29 17:07   ` Christoph Hellwig
2016-09-29 19:45     ` Darrick J. Wong
2016-09-28  2:58 ` [PATCH 46/63] xfs: CoW shared EOF block when truncating file Darrick J. Wong
2016-09-29 17:29   ` Christoph Hellwig
2016-09-29 20:13     ` Darrick J. Wong
2016-09-29 20:22       ` Christoph Hellwig
2016-09-29 21:23         ` Darrick J. Wong
2016-09-28  2:58 ` [PATCH 47/63] xfs: create a separate cow extent size hint for the allocator Darrick J. Wong
2016-09-28  2:58 ` [PATCH 48/63] xfs: preallocate blocks for worst-case btree expansion Darrick J. Wong
2016-09-28  2:58 ` [PATCH 49/63] xfs: don't allow reflink when the AG is low on space Darrick J. Wong
2016-09-28  2:58 ` [PATCH 50/63] xfs: try other AGs to allocate a BMBT block Darrick J. Wong
2016-09-28  2:59 ` [PATCH 51/63] xfs: garbage collect old cowextsz reservations Darrick J. Wong
2016-09-28  2:59 ` [PATCH 52/63] xfs: increase log reservations for reflink Darrick J. Wong
2016-09-28  2:59 ` [PATCH 53/63] xfs: add shared rmap map/unmap/convert log item types Darrick J. Wong
2016-09-28  2:59 ` [PATCH 54/63] xfs: use interval query for rmap alloc operations on shared files Darrick J. Wong
2016-09-28  2:59 ` [PATCH 55/63] xfs: convert unwritten status of reverse mappings for " Darrick J. Wong
2016-09-28  2:59 ` [PATCH 56/63] xfs: set a default CoW extent size of 32 blocks Darrick J. Wong
2016-09-28  2:59 ` [PATCH 57/63] xfs: check for invalid inode reflink flags Darrick J. Wong
2016-09-28  2:59 ` [PATCH 58/63] xfs: don't mix reflink and DAX mode for now Darrick J. Wong
2016-09-28  2:59 ` [PATCH 59/63] xfs: simulate per-AG reservations being critically low Darrick J. Wong
2016-09-28  3:00 ` [PATCH 60/63] xfs: recognize the reflink feature bit Darrick J. Wong
2016-09-28  3:00 ` [PATCH 61/63] xfs: various swapext cleanups Darrick J. Wong
2016-09-28  3:00 ` [PATCH 62/63] xfs: refactor swapext code Darrick J. Wong
2016-09-28  3:00 ` [PATCH 63/63] xfs: implement swapext for rmap filesystems Darrick J. Wong
2016-09-29 13:46 ` [PATCH v9 00/63] xfs: add reflink and dedupe support Christoph Hellwig
2016-09-29 17:23   ` Darrick J. Wong

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=20160929211823.GH14092@birch.djwong.org \
    --to=darrick.wong@oracle.com \
    --cc=bfoster@redhat.com \
    --cc=david@fromorbit.com \
    --cc=hch@lst.de \
    --cc=linux-xfs@vger.kernel.org \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).