linux-xfs.vger.kernel.org archive mirror
 help / color / mirror / Atom feed
From: Dave Chinner <david@fromorbit.com>
To: linux-xfs@vger.kernel.org
Subject: [PATCH 4/7] xfs: add tracepoints for stale pinned inode state debug
Date: Thu, 26 Jun 2025 08:48:57 +1000	[thread overview]
Message-ID: <20250625224957.1436116-5-david@fromorbit.com> (raw)
In-Reply-To: <20250625224957.1436116-1-david@fromorbit.com>

From: Dave Chinner <dchinner@redhat.com>

I needed more insight into how stale inodes were getting stuck on
the AIL after a forced shutdown when running fsstress. These are the
tracepoints I added for that purpose.

Signed-off-by: Dave Chinner <dchinner@redhat.com>
---
 fs/xfs/xfs_inode_item.c | 5 ++++-
 fs/xfs/xfs_log_cil.c    | 4 +++-
 fs/xfs/xfs_trace.h      | 9 ++++++++-
 fs/xfs/xfs_trans.c      | 4 +++-
 4 files changed, 18 insertions(+), 4 deletions(-)

diff --git a/fs/xfs/xfs_inode_item.c b/fs/xfs/xfs_inode_item.c
index c6cb0b6b9e46..285e27ff89e2 100644
--- a/fs/xfs/xfs_inode_item.c
+++ b/fs/xfs/xfs_inode_item.c
@@ -758,11 +758,14 @@ xfs_inode_item_push(
 		 * completed and items removed from the AIL before the next push
 		 * attempt.
 		 */
+		trace_xfs_inode_push_stale(ip, _RET_IP_);
 		return XFS_ITEM_PINNED;
 	}
 
-	if (xfs_ipincount(ip) > 0 || xfs_buf_ispinned(bp))
+	if (xfs_ipincount(ip) > 0 || xfs_buf_ispinned(bp)) {
+		trace_xfs_inode_push_pinned(ip, _RET_IP_);
 		return XFS_ITEM_PINNED;
+	}
 
 	if (xfs_iflags_test(ip, XFS_IFLUSHING))
 		return XFS_ITEM_FLUSHING;
diff --git a/fs/xfs/xfs_log_cil.c b/fs/xfs/xfs_log_cil.c
index f66d2d430e4f..a80cb6b9969a 100644
--- a/fs/xfs/xfs_log_cil.c
+++ b/fs/xfs/xfs_log_cil.c
@@ -793,8 +793,10 @@ xlog_cil_ail_insert(
 		struct xfs_log_item	*lip = lv->lv_item;
 		xfs_lsn_t		item_lsn;
 
-		if (aborted)
+		if (aborted) {
+			trace_xlog_ail_insert_abort(lip);
 			set_bit(XFS_LI_ABORTED, &lip->li_flags);
+		}
 
 		if (lip->li_ops->flags & XFS_ITEM_RELEASE_WHEN_COMMITTED) {
 			lip->li_ops->iop_release(lip);
diff --git a/fs/xfs/xfs_trace.h b/fs/xfs/xfs_trace.h
index 9f0d6bc966b7..ba45d801df1c 100644
--- a/fs/xfs/xfs_trace.h
+++ b/fs/xfs/xfs_trace.h
@@ -1146,6 +1146,7 @@ DECLARE_EVENT_CLASS(xfs_iref_class,
 		__field(xfs_ino_t, ino)
 		__field(int, count)
 		__field(int, pincount)
+		__field(unsigned long, iflags)
 		__field(unsigned long, caller_ip)
 	),
 	TP_fast_assign(
@@ -1153,13 +1154,15 @@ DECLARE_EVENT_CLASS(xfs_iref_class,
 		__entry->ino = ip->i_ino;
 		__entry->count = atomic_read(&VFS_I(ip)->i_count);
 		__entry->pincount = atomic_read(&ip->i_pincount);
+		__entry->iflags = ip->i_flags;
 		__entry->caller_ip = caller_ip;
 	),
-	TP_printk("dev %d:%d ino 0x%llx count %d pincount %d caller %pS",
+	TP_printk("dev %d:%d ino 0x%llx count %d pincount %d iflags 0x%lx caller %pS",
 		  MAJOR(__entry->dev), MINOR(__entry->dev),
 		  __entry->ino,
 		  __entry->count,
 		  __entry->pincount,
+		  __entry->iflags,
 		  (char *)__entry->caller_ip)
 )
 
@@ -1249,6 +1252,8 @@ DEFINE_IREF_EVENT(xfs_irele);
 DEFINE_IREF_EVENT(xfs_inode_pin);
 DEFINE_IREF_EVENT(xfs_inode_unpin);
 DEFINE_IREF_EVENT(xfs_inode_unpin_nowait);
+DEFINE_IREF_EVENT(xfs_inode_push_pinned);
+DEFINE_IREF_EVENT(xfs_inode_push_stale);
 
 DECLARE_EVENT_CLASS(xfs_namespace_class,
 	TP_PROTO(struct xfs_inode *dp, const struct xfs_name *name),
@@ -1653,6 +1658,8 @@ DEFINE_LOG_ITEM_EVENT(xfs_ail_flushing);
 DEFINE_LOG_ITEM_EVENT(xfs_cil_whiteout_mark);
 DEFINE_LOG_ITEM_EVENT(xfs_cil_whiteout_skip);
 DEFINE_LOG_ITEM_EVENT(xfs_cil_whiteout_unpin);
+DEFINE_LOG_ITEM_EVENT(xlog_ail_insert_abort);
+DEFINE_LOG_ITEM_EVENT(xfs_trans_free_abort);
 
 DECLARE_EVENT_CLASS(xfs_ail_class,
 	TP_PROTO(struct xfs_log_item *lip, xfs_lsn_t old_lsn, xfs_lsn_t new_lsn),
diff --git a/fs/xfs/xfs_trans.c b/fs/xfs/xfs_trans.c
index c6657072361a..b4a07af513ba 100644
--- a/fs/xfs/xfs_trans.c
+++ b/fs/xfs/xfs_trans.c
@@ -742,8 +742,10 @@ xfs_trans_free_items(
 
 	list_for_each_entry_safe(lip, next, &tp->t_items, li_trans) {
 		xfs_trans_del_item(lip);
-		if (abort)
+		if (abort) {
+			trace_xfs_trans_free_abort(lip);
 			set_bit(XFS_LI_ABORTED, &lip->li_flags);
+		}
 		if (lip->li_ops->iop_release)
 			lip->li_ops->iop_release(lip);
 	}
-- 
2.45.2


  parent reply	other threads:[~2025-06-25 22:50 UTC|newest]

Thread overview: 16+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2025-06-25 22:48 Dave Chinner
2025-06-25 22:48 ` [PATCH 1/7] xfs: xfs_ifree_cluster vs xfs_iflush_shutdown_abort deadlock Dave Chinner
2025-06-26  9:30   ` Carlos Maiolino
2025-07-03  7:27   ` Carlos Maiolino
2025-06-25 22:48 ` [PATCH 2/7] xfs: catch stale AGF/AGF metadata Dave Chinner
2025-06-26 10:09   ` Carlos Maiolino
2025-06-25 22:48 ` [PATCH 3/7] xfs: avoid dquot buffer pin deadlock Dave Chinner
2025-06-26 11:29   ` Carlos Maiolino
2025-06-25 22:48 ` Dave Chinner [this message]
2025-06-26 11:50   ` [PATCH 4/7] xfs: add tracepoints for stale pinned inode state debug Carlos Maiolino
2025-06-25 22:48 ` [PATCH 5/7] xfs: rearrange code in xfs_buf_item.c Dave Chinner
2025-06-26 11:57   ` Carlos Maiolino
2025-06-25 22:48 ` [PATCH 6/7] xfs: factor out stale buffer item completion Dave Chinner
2025-06-26 11:59   ` Carlos Maiolino
2025-06-25 22:49 ` [PATCH 7/7] xfs: fix unmount hang with unflushable inodes stuck in the AIL Dave Chinner
2025-06-27 10:53   ` Carlos Maiolino

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=20250625224957.1436116-5-david@fromorbit.com \
    --to=david@fromorbit.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).