From: Allison Henderson <allison.henderson@oracle.com>
To: "Darrick J. Wong" <djwong@kernel.org>
Cc: linux-xfs@vger.kernel.org, linux-fsdevel@vger.kernel.org,
linux-api@vger.kernel.org
Subject: Re: [PATCH 04/18] xfs: clear log incompat feature bits when the log is idle
Date: Fri, 2 Apr 2021 16:20:54 -0700 [thread overview]
Message-ID: <fc3f091d-fced-b342-6d58-f31a6552eccb@oracle.com> (raw)
In-Reply-To: <161723934912.3149451.16053630119296453937.stgit@magnolia>
On 3/31/21 6:09 PM, Darrick J. Wong wrote:
> From: Darrick J. Wong <djwong@kernel.org>
>
> When there are no ongoing transactions and the log contents have been
> checkpointed back into the filesystem, the log performs 'covering',
> which is to say that it log a dummy transaction to record the fact that
> the tail has caught up with the head. This is a good time to clear log
> incompat feature flags, because they are flags that are temporarily set
> to limit the range of kernels that can replay a dirty log.
>
> Since it's possible that some other higher level thread is about to
> start logging items protected by a log incompat flag, we create a rwsem
> so that upper level threads can coordinate this with the log. It would
> probably be more performant to use a percpu rwsem, but the ability to
> /try/ taking the write lock during covering is critical, and percpu
> rwsems do not provide that.
>
> Signed-off-by: Darrick J. Wong <djwong@kernel.org>
ok, makes sense
Reviewed-by: Allison Henderson <allison.henderson@oracle.com>
> ---
> fs/xfs/libxfs/xfs_shared.h | 6 +++++
> fs/xfs/xfs_log.c | 49 ++++++++++++++++++++++++++++++++++++++++++++
> fs/xfs/xfs_log.h | 3 +++
> fs/xfs/xfs_log_priv.h | 3 +++
> fs/xfs/xfs_trans.c | 14 +++++++++----
> 5 files changed, 71 insertions(+), 4 deletions(-)
>
>
> diff --git a/fs/xfs/libxfs/xfs_shared.h b/fs/xfs/libxfs/xfs_shared.h
> index 8c61a461bf7b..c7c9a0cebb04 100644
> --- a/fs/xfs/libxfs/xfs_shared.h
> +++ b/fs/xfs/libxfs/xfs_shared.h
> @@ -62,6 +62,12 @@ void xfs_log_get_max_trans_res(struct xfs_mount *mp,
> #define XFS_TRANS_SB_DIRTY 0x02 /* superblock is modified */
> #define XFS_TRANS_PERM_LOG_RES 0x04 /* xact took a permanent log res */
> #define XFS_TRANS_SYNC 0x08 /* make commit synchronous */
> +/*
> + * This transaction uses a log incompat feature, which means that we must tell
> + * the log that we've finished using it at the transaction commit or cancel.
> + * Callers must call xlog_use_incompat_feat before setting this flag.
> + */
> +#define XFS_TRANS_LOG_INCOMPAT 0x10
> #define XFS_TRANS_RESERVE 0x20 /* OK to use reserved data blocks */
> #define XFS_TRANS_NO_WRITECOUNT 0x40 /* do not elevate SB writecount */
> #define XFS_TRANS_RES_FDBLKS 0x80 /* reserve newly freed blocks */
> diff --git a/fs/xfs/xfs_log.c b/fs/xfs/xfs_log.c
> index cf73bc9f4d18..cb72be62da3e 100644
> --- a/fs/xfs/xfs_log.c
> +++ b/fs/xfs/xfs_log.c
> @@ -1335,6 +1335,32 @@ xfs_log_work_queue(
> msecs_to_jiffies(xfs_syncd_centisecs * 10));
> }
>
> +/*
> + * Clear the log incompat flags if we have the opportunity.
> + *
> + * This only happens if we're about to log the second dummy transaction as part
> + * of covering the log and we can get the log incompat feature usage lock.
> + */
> +static inline void
> +xlog_clear_incompat(
> + struct xlog *log)
> +{
> + struct xfs_mount *mp = log->l_mp;
> +
> + if (!xfs_sb_has_incompat_log_feature(&mp->m_sb,
> + XFS_SB_FEAT_INCOMPAT_LOG_ALL))
> + return;
> +
> + if (log->l_covered_state != XLOG_STATE_COVER_DONE2)
> + return;
> +
> + if (!down_write_trylock(&log->l_incompat_users))
> + return;
> +
> + xfs_clear_incompat_log_features(mp);
> + up_write(&log->l_incompat_users);
> +}
> +
> /*
> * Every sync period we need to unpin all items in the AIL and push them to
> * disk. If there is nothing dirty, then we might need to cover the log to
> @@ -1361,6 +1387,7 @@ xfs_log_worker(
> * synchronously log the superblock instead to ensure the
> * superblock is immediately unpinned and can be written back.
> */
> + xlog_clear_incompat(log);
> xfs_sync_sb(mp, true);
> } else
> xfs_log_force(mp, 0);
> @@ -1443,6 +1470,8 @@ xlog_alloc_log(
> }
> log->l_sectBBsize = 1 << log2_size;
>
> + init_rwsem(&log->l_incompat_users);
> +
> xlog_get_iclog_buffer_size(mp, log);
>
> spin_lock_init(&log->l_icloglock);
> @@ -3933,3 +3962,23 @@ xfs_log_in_recovery(
>
> return log->l_flags & XLOG_ACTIVE_RECOVERY;
> }
> +
> +/*
> + * Notify the log that we're about to start using a feature that is protected
> + * by a log incompat feature flag. This will prevent log covering from
> + * clearing those flags.
> + */
> +void
> +xlog_use_incompat_feat(
> + struct xlog *log)
> +{
> + down_read(&log->l_incompat_users);
> +}
> +
> +/* Notify the log that we've finished using log incompat features. */
> +void
> +xlog_drop_incompat_feat(
> + struct xlog *log)
> +{
> + up_read(&log->l_incompat_users);
> +}
> diff --git a/fs/xfs/xfs_log.h b/fs/xfs/xfs_log.h
> index 044e02cb8921..8b7d0a56cbf1 100644
> --- a/fs/xfs/xfs_log.h
> +++ b/fs/xfs/xfs_log.h
> @@ -145,4 +145,7 @@ bool xfs_log_in_recovery(struct xfs_mount *);
>
> xfs_lsn_t xlog_grant_push_threshold(struct xlog *log, int need_bytes);
>
> +void xlog_use_incompat_feat(struct xlog *log);
> +void xlog_drop_incompat_feat(struct xlog *log);
> +
> #endif /* __XFS_LOG_H__ */
> diff --git a/fs/xfs/xfs_log_priv.h b/fs/xfs/xfs_log_priv.h
> index 1c6fdbf3d506..75702c4fa69c 100644
> --- a/fs/xfs/xfs_log_priv.h
> +++ b/fs/xfs/xfs_log_priv.h
> @@ -436,6 +436,9 @@ struct xlog {
> #endif
> /* log recovery lsn tracking (for buffer submission */
> xfs_lsn_t l_recovery_lsn;
> +
> + /* Users of log incompat features should take a read lock. */
> + struct rw_semaphore l_incompat_users;
> };
>
> #define XLOG_BUF_CANCEL_BUCKET(log, blkno) \
> diff --git a/fs/xfs/xfs_trans.c b/fs/xfs/xfs_trans.c
> index eb2d8e2e5db6..e548d53c2091 100644
> --- a/fs/xfs/xfs_trans.c
> +++ b/fs/xfs/xfs_trans.c
> @@ -71,6 +71,9 @@ xfs_trans_free(
> xfs_extent_busy_sort(&tp->t_busy);
> xfs_extent_busy_clear(tp->t_mountp, &tp->t_busy, false);
>
> + if (tp->t_flags & XFS_TRANS_LOG_INCOMPAT)
> + xlog_drop_incompat_feat(tp->t_mountp->m_log);
> +
> trace_xfs_trans_free(tp, _RET_IP_);
> xfs_trans_clear_context(tp);
> if (!(tp->t_flags & XFS_TRANS_NO_WRITECOUNT))
> @@ -110,10 +113,13 @@ xfs_trans_dup(
> ASSERT(tp->t_flags & XFS_TRANS_PERM_LOG_RES);
> ASSERT(tp->t_ticket != NULL);
>
> - ntp->t_flags = XFS_TRANS_PERM_LOG_RES |
> - (tp->t_flags & XFS_TRANS_RESERVE) |
> - (tp->t_flags & XFS_TRANS_NO_WRITECOUNT) |
> - (tp->t_flags & XFS_TRANS_RES_FDBLKS);
> + ntp->t_flags = tp->t_flags & (XFS_TRANS_PERM_LOG_RES |
> + XFS_TRANS_RESERVE |
> + XFS_TRANS_NO_WRITECOUNT |
> + XFS_TRANS_RES_FDBLKS |
> + XFS_TRANS_LOG_INCOMPAT);
> + /* Give our LOG_INCOMPAT reference to the new transaction. */
> + tp->t_flags &= ~XFS_TRANS_LOG_INCOMPAT;
> /* We gave our writer reference to the new transaction */
> tp->t_flags |= XFS_TRANS_NO_WRITECOUNT;
> ntp->t_ticket = xfs_log_ticket_get(tp->t_ticket);
>
next prev parent reply other threads:[~2021-04-02 23:21 UTC|newest]
Thread overview: 30+ messages / expand[flat|nested] mbox.gz Atom feed top
2021-04-01 1:08 [PATCHSET RFC v3 00/18] xfs: atomic file updates Darrick J. Wong
2021-04-01 1:08 ` [PATCH 01/18] vfs: introduce new file range exchange ioctl Darrick J. Wong
2021-04-01 1:44 ` Al Viro
2021-04-01 21:18 ` Darrick J. Wong
2021-04-01 3:32 ` Amir Goldstein
2021-04-02 0:37 ` Darrick J. Wong
2021-04-01 1:08 ` [PATCH 02/18] xfs: support two inodes in the defer capture structure Darrick J. Wong
2021-04-02 23:20 ` Allison Henderson
2021-04-01 1:09 ` [PATCH 03/18] xfs: allow setting and clearing of log incompat feature flags Darrick J. Wong
2021-04-02 23:20 ` Allison Henderson
2021-04-01 1:09 ` [PATCH 04/18] xfs: clear log incompat feature bits when the log is idle Darrick J. Wong
2021-04-02 23:20 ` Allison Henderson [this message]
2021-04-01 1:09 ` [PATCH 05/18] xfs: create a log incompat flag for atomic extent swapping Darrick J. Wong
2021-04-02 23:21 ` Allison Henderson
2021-04-01 1:09 ` [PATCH 06/18] xfs: introduce a swap-extent log intent item Darrick J. Wong
2021-04-05 23:08 ` Allison Henderson
2021-04-01 1:09 ` [PATCH 07/18] xfs: create deferred log items for extent swapping Darrick J. Wong
2021-04-01 1:09 ` [PATCH 08/18] xfs: add a ->xchg_file_range handler Darrick J. Wong
2021-04-01 1:09 ` [PATCH 09/18] xfs: add error injection to test swapext recovery Darrick J. Wong
2021-04-01 1:09 ` [PATCH 10/18] xfs: port xfs_swap_extents_rmap to our new code Darrick J. Wong
2021-04-01 1:09 ` [PATCH 11/18] xfs: consolidate all of the xfs_swap_extent_forks code Darrick J. Wong
2021-04-01 1:09 ` [PATCH 12/18] xfs: refactor reflink flag handling in xfs_swap_extent_forks Darrick J. Wong
2021-04-01 1:09 ` [PATCH 13/18] xfs: allow xfs_swap_range to use older extent swap algorithms Darrick J. Wong
2021-04-01 1:10 ` [PATCH 14/18] xfs: remove old swap extents implementation Darrick J. Wong
2021-04-01 1:10 ` [PATCH 15/18] xfs: condense extended attributes after an atomic swap Darrick J. Wong
2021-04-01 1:10 ` [PATCH 16/18] xfs: condense directories " Darrick J. Wong
2021-04-01 1:10 ` [PATCH 17/18] xfs: make atomic extent swapping support realtime files Darrick J. Wong
2021-04-01 1:10 ` [PATCH 18/18] xfs: enable atomic swapext feature Darrick J. Wong
2021-04-01 3:56 ` [PATCHSET RFC v3 00/18] xfs: atomic file updates Amir Goldstein
2021-04-02 0:22 ` 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=fc3f091d-fced-b342-6d58-f31a6552eccb@oracle.com \
--to=allison.henderson@oracle.com \
--cc=djwong@kernel.org \
--cc=linux-api@vger.kernel.org \
--cc=linux-fsdevel@vger.kernel.org \
--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).