linux-xfs.vger.kernel.org archive mirror
 help / color / mirror / Atom feed
From: "Darrick J. Wong" <darrick.wong@oracle.com>
To: Allison Henderson <allison.henderson@oracle.com>
Cc: linux-xfs@vger.kernel.org, Dave Chinner <dchinner@redhat.com>
Subject: Re: [PATCH v3 10/17] xfs: extent transaction reservations for parent attributes
Date: Tue, 28 Nov 2017 10:58:02 -0800	[thread overview]
Message-ID: <20171128185802.GP21412@magnolia> (raw)
In-Reply-To: <1510942905-12897-11-git-send-email-allison.henderson@oracle.com>

On Fri, Nov 17, 2017 at 11:21:38AM -0700, Allison Henderson wrote:
> From: Dave Chinner <dchinner@redhat.com>
> 
> We need to add, remove or modify parent pointer attributes during
> create/link/unlink/rename operations atomically with the dirents in the parent
> directories being modified. This means they need to be modified in the same
> transaction as the parent directories, and so we need to add the required
> space for the attribute modifications to the transaction reservations.
> 
> [achender: rebased, added xfs_sb_version_hasparent stub]
> 
> Signed-off-by: Dave Chinner <dchinner@redhat.com>
> Signed-off-by: Allison Henderson <allison.henderson@oracle.com>
> ---
>  fs/xfs/libxfs/xfs_format.h     |   5 ++
>  fs/xfs/libxfs/xfs_trans_resv.c | 103 ++++++++++++++++++++++++++++++++---------
>  2 files changed, 85 insertions(+), 23 deletions(-)
> 
> diff --git a/fs/xfs/libxfs/xfs_format.h b/fs/xfs/libxfs/xfs_format.h
> index b9ea5bf..121862a 100644
> --- a/fs/xfs/libxfs/xfs_format.h
> +++ b/fs/xfs/libxfs/xfs_format.h
> @@ -556,6 +556,11 @@ static inline bool xfs_sb_version_hasreflink(struct xfs_sb *sbp)
>  		(sbp->sb_features_ro_compat & XFS_SB_FEAT_RO_COMPAT_REFLINK);
>  }
>  
> +static inline bool xfs_sb_version_hasparent(struct xfs_sb *sbp)
> +{
> +	return false; /* We'll enable this at the end of the set */
> +}
> +
>  /*
>   * end of superblock version macros
>   */
> diff --git a/fs/xfs/libxfs/xfs_trans_resv.c b/fs/xfs/libxfs/xfs_trans_resv.c
> index 6bd916b..54399e2 100644
> --- a/fs/xfs/libxfs/xfs_trans_resv.c
> +++ b/fs/xfs/libxfs/xfs_trans_resv.c
> @@ -802,29 +802,30 @@ xfs_calc_sb_reservation(
>  	return xfs_calc_buf_res(1, mp->m_sb.sb_sectsize);
>  }
>  
> +/*
> + * Namespace reservations.
> + *
> + * These get tricky when parent pointers are enabled as we have attribute
> + * modifications occurring from within these transactions. Rather than confuse
> + * each of these reservation calculations with the conditional attribute
> + * reservations, add them here in a clear and concise manner. This assumes that
> + * the attribute reservations have already been calculated.
> + *
> + * Note that we only include the static attribute reservation here; the runtime
> + * reservation will have to be modified by the size of the attributes being
> + * added/removed/modified. See the comments on the attribute reservation
> + * calculations for more details.
> + *
> + * Note for rename: rename will vastly overestimate requirements. This will be
> + * addressed later when modifications are made to ensure parent attribute
> + * modifications can be done atomically with the rename operation.
> + */
>  void
> -xfs_trans_resv_calc(
> +xfs_calc_namespace_reservations(
>  	struct xfs_mount	*mp,
>  	struct xfs_trans_resv	*resp)
>  {
> -	/*
> -	 * The following transactions are logged in physical format and
> -	 * require a permanent reservation on space.
> -	 */
> -	resp->tr_write.tr_logres = xfs_calc_write_reservation(mp);
> -	if (xfs_sb_version_hasreflink(&mp->m_sb))
> -		resp->tr_write.tr_logcount = XFS_WRITE_LOG_COUNT_REFLINK;
> -	else
> -		resp->tr_write.tr_logcount = XFS_WRITE_LOG_COUNT;
> -	resp->tr_write.tr_logflags |= XFS_TRANS_PERM_LOG_RES;
> -
> -	resp->tr_itruncate.tr_logres = xfs_calc_itruncate_reservation(mp);
> -	if (xfs_sb_version_hasreflink(&mp->m_sb))
> -		resp->tr_itruncate.tr_logcount =
> -				XFS_ITRUNCATE_LOG_COUNT_REFLINK;
> -	else
> -		resp->tr_itruncate.tr_logcount = XFS_ITRUNCATE_LOG_COUNT;
> -	resp->tr_itruncate.tr_logflags |= XFS_TRANS_PERM_LOG_RES;
> +	ASSERT(resp->tr_attrsetm.tr_logres > 0);
>  
>  	resp->tr_rename.tr_logres = xfs_calc_rename_reservation(mp);
>  	resp->tr_rename.tr_logcount = XFS_RENAME_LOG_COUNT;
> @@ -846,15 +847,69 @@ xfs_trans_resv_calc(
>  	resp->tr_create.tr_logcount = XFS_CREATE_LOG_COUNT;
>  	resp->tr_create.tr_logflags |= XFS_TRANS_PERM_LOG_RES;
>  
> +	resp->tr_mkdir.tr_logres = xfs_calc_mkdir_reservation(mp);
> +	resp->tr_mkdir.tr_logcount = XFS_MKDIR_LOG_COUNT;
> +	resp->tr_mkdir.tr_logflags |= XFS_TRANS_PERM_LOG_RES;
> +
> +	if (!xfs_sb_version_hasparent(&mp->m_sb))
> +		return;
> +
> +	/* rename can add/remove/modify 2 parent attributes */
> +	resp->tr_rename.tr_logres += 2 * max(resp->tr_attrsetm.tr_logres,
> +					     resp->tr_attrrm.tr_logres);
> +	resp->tr_rename.tr_logcount += 2 * max(resp->tr_attrsetm.tr_logcount,
> +					       resp->tr_attrrm.tr_logcount);
> +
> +	/* create will add 1 parent attribute */
> +	resp->tr_create.tr_logres += resp->tr_attrsetm.tr_logres;
> +	resp->tr_create.tr_logcount += resp->tr_attrsetm.tr_logcount;
> +
> +	/* mkdir will add 1 parent attribute */
> +	resp->tr_mkdir.tr_logres += resp->tr_attrsetm.tr_logres;
> +	resp->tr_mkdir.tr_logcount += resp->tr_attrsetm.tr_logcount;
> +
> +	/* link will add 1 parent attribute */
> +	resp->tr_link.tr_logres += resp->tr_attrsetm.tr_logres;
> +	resp->tr_link.tr_logcount += resp->tr_attrsetm.tr_logcount;
> +
> +	/* symlink will add 1 parent attribute */
> +	resp->tr_symlink.tr_logres += resp->tr_attrsetm.tr_logres;
> +	resp->tr_symlink.tr_logcount += resp->tr_attrsetm.tr_logcount;
> +
> +	/* remove will remove 1 parent attribute */
> +	resp->tr_remove.tr_logres += resp->tr_attrrm.tr_logres;
> +	resp->tr_remove.tr_logcount = resp->tr_attrrm.tr_logcount;

djwong, clearly reviewing patches in reverse order:

This looks pretty similar to the function in the next patch; can they
be combined into a single helper?

> +}
> +
> +void
> +xfs_trans_resv_calc(
> +	struct xfs_mount	*mp,
> +	struct xfs_trans_resv	*resp)
> +{
> +	/*
> +	 * The following transactions are logged in physical format and
> +	 * require a permanent reservation on space.
> +	 */
> +	resp->tr_write.tr_logres = xfs_calc_write_reservation(mp);
> +	if (xfs_sb_version_hasreflink(&mp->m_sb))
> +		resp->tr_write.tr_logcount = XFS_WRITE_LOG_COUNT_REFLINK;
> +	else
> +		resp->tr_write.tr_logcount = XFS_WRITE_LOG_COUNT;
> +	resp->tr_write.tr_logflags |= XFS_TRANS_PERM_LOG_RES;
> +
> +	resp->tr_itruncate.tr_logres = xfs_calc_itruncate_reservation(mp);
> +	if (xfs_sb_version_hasreflink(&mp->m_sb))
> +		resp->tr_itruncate.tr_logcount =
> +				XFS_ITRUNCATE_LOG_COUNT_REFLINK;
> +	else
> +		resp->tr_itruncate.tr_logcount = XFS_ITRUNCATE_LOG_COUNT;
> +	resp->tr_itruncate.tr_logflags |= XFS_TRANS_PERM_LOG_RES;
> +
>  	resp->tr_create_tmpfile.tr_logres =
>  			xfs_calc_create_tmpfile_reservation(mp);
>  	resp->tr_create_tmpfile.tr_logcount = XFS_CREATE_TMPFILE_LOG_COUNT;
>  	resp->tr_create_tmpfile.tr_logflags |= XFS_TRANS_PERM_LOG_RES;
>  
> -	resp->tr_mkdir.tr_logres = xfs_calc_mkdir_reservation(mp);
> -	resp->tr_mkdir.tr_logcount = XFS_MKDIR_LOG_COUNT;
> -	resp->tr_mkdir.tr_logflags |= XFS_TRANS_PERM_LOG_RES;
> -
>  	resp->tr_ifree.tr_logres = xfs_calc_ifree_reservation(mp);
>  	resp->tr_ifree.tr_logcount = XFS_INACTIVE_LOG_COUNT;
>  	resp->tr_ifree.tr_logflags |= XFS_TRANS_PERM_LOG_RES;
> @@ -886,6 +941,8 @@ xfs_trans_resv_calc(
>  		resp->tr_qm_dqalloc.tr_logcount = XFS_WRITE_LOG_COUNT;
>  	resp->tr_qm_dqalloc.tr_logflags |= XFS_TRANS_PERM_LOG_RES;
>  
> +	xfs_calc_namespace_reservations(mp, resp);
> +
>  	/*
>  	 * The following transactions are logged in logical format with
>  	 * a default log count.
> -- 
> 2.7.4
> 
> --
> 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

  reply	other threads:[~2017-11-28 18:58 UTC|newest]

Thread overview: 69+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2017-11-17 18:21 [PATCH v3 00/17] Parent Pointers v4 Allison Henderson
2017-11-17 18:21 ` [PATCH v3 01/17] Add helper functions xfs_attr_set_args and xfs_attr_remove_args Allison Henderson
2017-11-28 19:54   ` Darrick J. Wong
2017-11-29  1:02     ` Dave Chinner
2017-11-29 18:52     ` Allison Henderson
2017-11-29 22:34       ` Allison Henderson
2017-11-17 18:21 ` [PATCH v3 02/17] Set up infastructure for deferred attribute operations Allison Henderson
2017-11-28 19:45   ` Darrick J. Wong
2017-11-29  1:19     ` Dave Chinner
2017-11-29 18:52       ` Allison Henderson
2017-11-29 18:51     ` Allison Henderson
2017-11-17 18:21 ` [PATCH v3 03/17] Add xfs_attr_set_defered and xfs_attr_remove_defered Allison Henderson
2017-11-28 19:19   ` Darrick J. Wong
2017-11-29 18:50     ` Allison Henderson
2017-11-17 18:21 ` [PATCH v3 04/17] Remove all strlen calls in all xfs_attr_* functions for attr names Allison Henderson
2017-11-28 19:10   ` Darrick J. Wong
2017-11-29 18:50     ` Allison Henderson
2017-11-17 18:21 ` [PATCH v3 05/17] xfs: get directory offset when adding directory name Allison Henderson
2017-11-28 19:07   ` Darrick J. Wong
2017-11-29 18:50     ` Allison Henderson
2017-11-17 18:21 ` [PATCH v3 06/17] xfs: get directory offset when removing " Allison Henderson
2017-11-28 19:05   ` Darrick J. Wong
2017-11-29 18:49     ` Allison Henderson
2017-11-17 18:21 ` [PATCH v3 07/17] xfs: get directory offset when replacing a " Allison Henderson
2017-11-28 19:04   ` Darrick J. Wong
2017-11-29 18:49     ` Allison Henderson
2017-11-17 18:21 ` [PATCH v3 08/17] xfs: add parent pointer support to attribute code Allison Henderson
2017-11-28 19:01   ` Darrick J. Wong
2017-11-29 18:48     ` Allison Henderson
2017-11-17 18:21 ` [PATCH v3 09/17] xfs: define parent pointer xattr format Allison Henderson
2017-11-28 18:59   ` Darrick J. Wong
2017-11-29 18:48     ` Allison Henderson
2017-11-17 18:21 ` [PATCH v3 10/17] xfs: extent transaction reservations for parent attributes Allison Henderson
2017-11-28 18:58   ` Darrick J. Wong [this message]
2017-11-29 18:48     ` Allison Henderson
2017-11-17 18:21 ` [PATCH v3 11/17] Add the extra space requirements for parent pointer attributes when calculating the minimum log size during mkfs Allison Henderson
2017-11-28 18:51   ` Darrick J. Wong
2017-11-29 18:47     ` Allison Henderson
2017-11-29 20:18       ` Darrick J. Wong
2017-11-17 18:21 ` [PATCH v3 12/17] xfs: parent pointer attribute creation Allison Henderson
2017-11-28 18:49   ` Darrick J. Wong
2017-11-28 18:54     ` Darrick J. Wong
2017-11-29 18:46       ` Allison Henderson
2017-11-17 18:21 ` [PATCH v3 13/17] xfs: add parent attributes to link Allison Henderson
2017-11-28 18:37   ` Darrick J. Wong
2017-11-29 18:45     ` Allison Henderson
2017-11-17 18:21 ` [PATCH v3 14/17] xfs: remove parent pointers in unlink Allison Henderson
2017-11-28 18:24   ` Darrick J. Wong
2017-11-29 18:44     ` Allison Henderson
2017-11-17 18:21 ` [PATCH v3 15/17] Add parent pointers to rename Allison Henderson
2017-11-28 18:20   ` Darrick J. Wong
2017-11-29 18:43     ` Allison Henderson
2017-11-17 18:21 ` [PATCH v3 16/17] Add the parent pointer support to the superblock version 5 Allison Henderson
2017-11-28 18:08   ` Darrick J. Wong
2017-11-29 18:41     ` Allison Henderson
2017-11-17 18:21 ` [PATCH v3 17/17] Add parent pointer ioctl Allison Henderson
2017-11-22 19:54   ` Allison Henderson
2017-11-22 21:07     ` Dave Chinner
2017-11-22 22:49       ` Allison Henderson
2017-11-22 21:13     ` Darrick J. Wong
2017-11-22 22:49       ` Allison Henderson
2017-11-28 20:35   ` Darrick J. Wong
2017-11-29 18:52     ` Allison Henderson
2017-11-29 21:37     ` Dave Chinner
2017-11-29 22:48       ` Allison Henderson
2017-11-30  0:02         ` Dave Chinner
2017-11-30  1:52           ` Allison Henderson
2017-11-30 21:11           ` Darrick J. Wong
2017-12-01  2:58             ` Dave Chinner

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=20171128185802.GP21412@magnolia \
    --to=darrick.wong@oracle.com \
    --cc=allison.henderson@oracle.com \
    --cc=dchinner@redhat.com \
    --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).