From: Allison Henderson <allison.henderson@oracle.com>
To: linux-xfs@vger.kernel.org
Subject: [PATCH v22 02/16] xfs: clear log incompat feature bits when the log is idle
Date: Mon, 26 Jul 2021 23:20:39 -0700 [thread overview]
Message-ID: <20210727062053.11129-3-allison.henderson@oracle.com> (raw)
In-Reply-To: <20210727062053.11129-1-allison.henderson@oracle.com>
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>
Reviewed-by: Allison Henderson <allison.henderson@oracle.com>
Signed-off-by: Allison Henderson <allison.henderson@oracle.com>
---
fs/xfs/xfs_log.c | 49 +++++++++++++++++++++++++++++++++++++++++++++++++
fs/xfs/xfs_log.h | 3 +++
fs/xfs/xfs_log_priv.h | 3 +++
3 files changed, 55 insertions(+)
diff --git a/fs/xfs/xfs_log.c b/fs/xfs/xfs_log.c
index 9254405..c58a0d7 100644
--- a/fs/xfs/xfs_log.c
+++ b/fs/xfs/xfs_log.c
@@ -1338,6 +1338,32 @@ xfs_log_work_queue(
}
/*
+ * 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
* indicate that the filesystem is idle.
@@ -1363,6 +1389,7 @@ 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);
@@ -1450,6 +1477,8 @@ 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);
@@ -3895,3 +3924,23 @@ 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 813b972..b274fb9 100644
--- a/fs/xfs/xfs_log.h
+++ b/fs/xfs/xfs_log.h
@@ -142,4 +142,7 @@ 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 4c41bbfa..c507041 100644
--- a/fs/xfs/xfs_log_priv.h
+++ b/fs/xfs/xfs_log_priv.h
@@ -449,6 +449,9 @@ struct xlog {
xfs_lsn_t l_recovery_lsn;
uint32_t l_iclog_roundoff;/* padding roundoff */
+
+ /* Users of log incompat features should take a read lock. */
+ struct rw_semaphore l_incompat_users;
};
#define XLOG_BUF_CANCEL_BUCKET(log, blkno) \
--
2.7.4
next prev parent reply other threads:[~2021-07-27 6:21 UTC|newest]
Thread overview: 67+ messages / expand[flat|nested] mbox.gz Atom feed top
2021-07-27 6:20 [PATCH v22 00/16] Delayed Attributes Allison Henderson
2021-07-27 6:20 ` [PATCH v22 01/16] xfs: allow setting and clearing of log incompat feature flags Allison Henderson
2021-07-27 12:24 ` Chandan Babu R
2021-07-28 9:01 ` Allison Henderson
2021-07-27 6:20 ` Allison Henderson [this message]
2021-07-27 12:46 ` [PATCH v22 02/16] xfs: clear log incompat feature bits when the log is idle Chandan Babu R
2021-07-28 9:02 ` Allison Henderson
2021-07-27 6:20 ` [PATCH v22 03/16] xfs: refactor xfs_iget calls from log intent recovery Allison Henderson
2021-07-28 11:54 ` Chandan Babu R
2021-07-30 9:17 ` Allison Henderson
2021-07-27 6:20 ` [PATCH v22 04/16] xfs: Return from xfs_attr_set_iter if there are no more rmtblks to process Allison Henderson
2021-07-27 23:30 ` Darrick J. Wong
2021-07-28 9:01 ` Allison Henderson
2021-07-28 12:18 ` Chandan Babu R
2021-07-30 9:17 ` Allison Henderson
2021-08-09 17:24 ` Darrick J. Wong
2021-07-27 6:20 ` [PATCH v22 05/16] xfs: Add state machine tracepoints Allison Henderson
2021-07-28 13:42 ` Chandan Babu R
2021-07-30 9:17 ` Allison Henderson
2021-07-27 6:20 ` [PATCH v22 06/16] xfs: Rename __xfs_attr_rmtval_remove Allison Henderson
2021-07-29 7:56 ` Chandan Babu R
2021-07-30 9:17 ` Allison Henderson
2021-07-27 6:20 ` [PATCH v22 07/16] xfs: Handle krealloc errors in xlog_recover_add_to_cont_trans Allison Henderson
2021-07-29 8:27 ` Chandan Babu R
2021-07-30 9:17 ` Allison Henderson
2021-07-27 6:20 ` [PATCH v22 08/16] xfs: Set up infrastructure for deferred attribute operations Allison Henderson
2021-07-28 0:56 ` Darrick J. Wong
2021-07-28 9:04 ` Allison Henderson
2021-07-30 4:46 ` Chandan Babu R
2021-07-30 9:17 ` Allison Henderson
2021-07-27 6:20 ` [PATCH v22 09/16] xfs: Implement attr logging and replay Allison Henderson
2021-07-27 9:38 ` Chandan Babu R
2021-07-28 9:01 ` Allison Henderson
2021-07-28 0:39 ` Darrick J. Wong
2021-07-28 9:05 ` Allison Henderson
2021-07-30 12:21 ` Chandan Babu R
2021-08-02 8:33 ` Allison Henderson
2021-07-27 6:20 ` [PATCH v22 10/16] RFC xfs: Skip flip flags for delayed attrs Allison Henderson
2021-07-28 19:18 ` Darrick J. Wong
2021-07-31 5:11 ` Allison Henderson
2021-08-02 7:47 ` Allison Henderson
2021-07-30 14:40 ` Chandan Babu R
2021-08-02 8:33 ` Allison Henderson
2021-07-27 6:20 ` [PATCH v22 11/16] xfs: Add xfs_attr_set_deferred and xfs_attr_remove_deferred Allison Henderson
2021-07-28 19:24 ` Darrick J. Wong
2021-08-02 8:18 ` Allison Henderson
2021-07-30 14:58 ` Chandan Babu R
2021-08-02 8:33 ` Allison Henderson
2021-07-27 6:20 ` [PATCH v22 12/16] xfs: Remove unused xfs_attr_*_args Allison Henderson
2021-07-28 19:31 ` Darrick J. Wong
2021-08-02 8:11 ` Allison Henderson
2021-08-02 3:26 ` Chandan Babu R
2021-08-02 8:33 ` Allison Henderson
2021-07-27 6:20 ` [PATCH v22 13/16] xfs: Add delayed attributes error tag Allison Henderson
2021-08-02 3:27 ` Chandan Babu R
2021-08-02 8:33 ` Allison Henderson
2021-07-27 6:20 ` [PATCH v22 14/16] xfs: Add delattr mount option Allison Henderson
2021-07-28 0:47 ` Darrick J. Wong
2021-07-28 2:13 ` Dave Chinner
2021-07-28 9:05 ` Allison Henderson
2021-07-28 9:02 ` Allison Henderson
2021-07-27 6:20 ` [PATCH v22 15/16] xfs: Merge xfs_delattr_context into xfs_attr_item Allison Henderson
2021-08-02 3:27 ` Chandan Babu R
2021-08-02 8:33 ` Allison Henderson
2021-07-27 6:20 ` [PATCH v22 16/16] xfs: Add helper function xfs_attr_leaf_addname Allison Henderson
2021-07-28 19:52 ` Darrick J. Wong
2021-08-02 8:18 ` Allison Henderson
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=20210727062053.11129-3-allison.henderson@oracle.com \
--to=allison.henderson@oracle.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).