xfs
[Top] [All Lists]

Re: [PATCH v2] xfs: store rmapbt block count in the AGF

To: Christoph Hellwig <hch@xxxxxxxxxxxxx>
Subject: Re: [PATCH v2] xfs: store rmapbt block count in the AGF
From: "Darrick J. Wong" <darrick.wong@xxxxxxxxxx>
Date: Thu, 4 Aug 2016 08:50:42 -0700
Cc: david@xxxxxxxxxxxxx, xfs@xxxxxxxxxxx
Delivered-to: xfs@xxxxxxxxxxx
In-reply-to: <20160804154439.GA15656@xxxxxxxxxxxxx>
References: <20160803214613.GF8593@xxxxxxxxxxxxxxxx> <20160804154439.GA15656@xxxxxxxxxxxxx>
User-agent: Mutt/1.5.24 (2015-08-30)
On Thu, Aug 04, 2016 at 08:44:39AM -0700, Christoph Hellwig wrote:
> On Wed, Aug 03, 2016 at 02:46:13PM -0700, Darrick J. Wong wrote:
> > Track the number of blocks used for the rmapbt in the AGF.  When we
> > get to the AG reservation code we need this counter to quickly
> > make our reservation during mount.
> > 
> > v2: Actually include growfs support.
> > 
> > Signed-off-by: Darrick J. Wong <darrick.wong@xxxxxxxxxx>
> > ---
> >  fs/xfs/libxfs/xfs_alloc.c      |    1 +
> >  fs/xfs/libxfs/xfs_format.h     |   11 ++++++++---
> >  fs/xfs/libxfs/xfs_rmap_btree.c |    6 ++++++
> >  fs/xfs/xfs_fsops.c             |    1 +
> >  4 files changed, 16 insertions(+), 3 deletions(-)
> > 
> > diff --git a/fs/xfs/libxfs/xfs_alloc.c b/fs/xfs/libxfs/xfs_alloc.c
> > index 776ae2f..af1a72e 100644
> > --- a/fs/xfs/libxfs/xfs_alloc.c
> > +++ b/fs/xfs/libxfs/xfs_alloc.c
> > @@ -2264,6 +2264,7 @@ xfs_alloc_log_agf(
> >             offsetof(xfs_agf_t, agf_longest),
> >             offsetof(xfs_agf_t, agf_btreeblks),
> >             offsetof(xfs_agf_t, agf_uuid),
> > +           offsetof(xfs_agf_t, agf_rmap_blocks),
> >             sizeof(xfs_agf_t)
> >     };
> >  
> > diff --git a/fs/xfs/libxfs/xfs_format.h b/fs/xfs/libxfs/xfs_format.h
> > index f814d42..e6a8bea 100644
> > --- a/fs/xfs/libxfs/xfs_format.h
> > +++ b/fs/xfs/libxfs/xfs_format.h
> > @@ -640,12 +640,15 @@ typedef struct xfs_agf {
> >     __be32          agf_btreeblks;  /* # of blocks held in AGF btrees */
> >     uuid_t          agf_uuid;       /* uuid of filesystem */
> >  
> > +   __be32          agf_rmap_blocks;        /* rmapbt blocks used */
> > +   __be32          agf_padding;            /* padding */
> 
> Now _spare?  It could easily be used by the next new field.

The reflink patches will (almost immediately) consume agf_padding
and the next spare for refcount_{root,level,blocks}.

--D

> Otherwise this looks fine to me:
> 
> Reviewed-by: Christoph Hellwig <hch@xxxxxx>

<Prev in Thread] Current Thread [Next in Thread>