xfs
[Top] [All Lists]

Re: [patch 04/19] xfs: untange SYNC_WAIT and SYNC_TRYLOCK meanings for x

To: Christoph Hellwig <hch@xxxxxxxxxxxxx>
Subject: Re: [patch 04/19] xfs: untange SYNC_WAIT and SYNC_TRYLOCK meanings for xfs_qm_dqflush
From: Ben Myers <bpm@xxxxxxx>
Date: Thu, 8 Dec 2011 15:10:30 -0600
Cc: xfs@xxxxxxxxxxx
In-reply-to: <20111206215853.264745959@xxxxxxxxxxxxxxxxxxxxxx>
References: <20111206215806.844405397@xxxxxxxxxxxxxxxxxxxxxx> <20111206215853.264745959@xxxxxxxxxxxxxxxxxxxxxx>
User-agent: Mutt/1.5.18 (2008-05-17)
On Tue, Dec 06, 2011 at 04:58:10PM -0500, Christoph Hellwig wrote:
> Only skip pinned dquots if SYNC_TRYLOCK is specified, and adjust the callers
> to keep the behaviour unchanged.
> 
> Signed-off-by: Christoph Hellwig <hch@xxxxxx>
> Reviewed-by: Dave Chinner <dchinner@xxxxxxxxxx>

s/untange/untangle  in the subject line.

> ---
>  fs/xfs/xfs_dquot.c      |    2 +-
>  fs/xfs/xfs_dquot_item.c |    2 +-
>  fs/xfs/xfs_qm.c         |    2 +-
>  3 files changed, 3 insertions(+), 3 deletions(-)
> 
> Index: xfs/fs/xfs/xfs_dquot.c
> ===================================================================
> --- xfs.orig/fs/xfs/xfs_dquot.c       2011-11-05 08:54:01.729993938 +0100
> +++ xfs/fs/xfs/xfs_dquot.c    2011-11-25 11:45:32.668742260 +0100
> @@ -1169,7 +1169,7 @@ xfs_qm_dqflush(
>        * If not dirty, or it's pinned and we are not supposed to block, nada.
>        */
>       if (!XFS_DQ_IS_DIRTY(dqp) ||
> -         (!(flags & SYNC_WAIT) && atomic_read(&dqp->q_pincount) > 0)) {
> +         ((flags & SYNC_TRYLOCK) && atomic_read(&dqp->q_pincount) > 0)) {

There is a SYNC_WAIT at the bottom of xfs_qm_dqflush.  Should it be
modified too?

>               xfs_dqfunlock(dqp);
>               return 0;
>       }
> Index: xfs/fs/xfs/xfs_dquot_item.c
> ===================================================================
> --- xfs.orig/fs/xfs/xfs_dquot_item.c  2011-11-25 11:43:25.269432441 +0100
> +++ xfs/fs/xfs/xfs_dquot_item.c       2011-11-25 11:45:32.668742260 +0100
> @@ -133,7 +133,7 @@ xfs_qm_dquot_logitem_push(
>        * lock without sleeping, then there must not have been
>        * anyone in the process of flushing the dquot.
>        */
> -     error = xfs_qm_dqflush(dqp, 0);
> +     error = xfs_qm_dqflush(dqp, SYNC_TRYLOCK);
>       if (error)
>               xfs_warn(dqp->q_mount, "%s: push error %d on dqp %p",
>                       __func__, error, dqp);
> Index: xfs/fs/xfs/xfs_qm.c
> ===================================================================
> --- xfs.orig/fs/xfs/xfs_qm.c  2011-11-19 20:14:00.400421363 +0100
> +++ xfs/fs/xfs/xfs_qm.c       2011-11-25 11:45:32.672075575 +0100
> @@ -1661,7 +1661,7 @@ xfs_qm_quotacheck(
>        * successfully.
>        */
>       if (!error)
> -             error = xfs_qm_dqflush_all(mp, 0);
> +             error = xfs_qm_dqflush_all(mp, SYNC_TRYLOCK);
>  
>       /*
>        * We can get this error if we couldn't do a dquot allocation inside

from xfs_qm_dqreclaim_one:
        /*
         * We flush it delayed write, so don't bother
         * releasing the freelist lock.
         */
        error = xfs_qm_dqflush(dqp, 0); 

Should that also be changed to SYNC_TRYLOCK?

SYNC_TRYLOCK also has meaning for xfs_qm_sync().  Maybe the intention
was that SYNC_TRYLOCK would be used with xfs_qm_sync, and SYNC_WAIT with
xfs_qm_dqflush?

-Ben

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