Linux NILFS development
 help / color / mirror / Atom feed
From: Ryusuke Konishi <konishi.ryusuke@gmail.com>
To: Andrew Morton <akpm@linux-foundation.org>
Cc: linux-nilfs@vger.kernel.org, linux-kernel@vger.kernel.org
Subject: [PATCH 1/5] nilfs2: treat missing sufile header block as metadata corruption
Date: Thu, 22 Aug 2024 00:46:23 +0900	[thread overview]
Message-ID: <20240821154627.11848-2-konishi.ryusuke@gmail.com> (raw)
In-Reply-To: <20240821154627.11848-1-konishi.ryusuke@gmail.com>

The sufile, a metadata file that holds metadata for segment management,
has statistical information in its first block, but if reading this
block fails, it receives the internal code -ENOENT and returns it
unchanged to the callers.

To prevent this -ENOENT from being propagated to system calls, if
reading the header block fails, return -EIO (or -EINVAL depending on
the context) instead.

Signed-off-by: Ryusuke Konishi <konishi.ryusuke@gmail.com>
---
 fs/nilfs2/sufile.c | 23 ++++++++++++++++++-----
 1 file changed, 18 insertions(+), 5 deletions(-)

diff --git a/fs/nilfs2/sufile.c b/fs/nilfs2/sufile.c
index 7bfc0860acee..f071eba48163 100644
--- a/fs/nilfs2/sufile.c
+++ b/fs/nilfs2/sufile.c
@@ -79,10 +79,17 @@ nilfs_sufile_block_get_segment_usage(const struct inode *sufile, __u64 segnum,
 		NILFS_MDT(sufile)->mi_entry_size;
 }
 
-static inline int nilfs_sufile_get_header_block(struct inode *sufile,
-						struct buffer_head **bhp)
+static int nilfs_sufile_get_header_block(struct inode *sufile,
+					 struct buffer_head **bhp)
 {
-	return nilfs_mdt_get_block(sufile, 0, 0, NULL, bhp);
+	int err = nilfs_mdt_get_block(sufile, 0, 0, NULL, bhp);
+
+	if (unlikely(err == -ENOENT)) {
+		nilfs_error(sufile->i_sb,
+			    "missing header block in segment usage metadata");
+		err = -EIO;
+	}
+	return err;
 }
 
 static inline int
@@ -1237,9 +1244,15 @@ int nilfs_sufile_read(struct super_block *sb, size_t susize,
 	if (err)
 		goto failed;
 
-	err = nilfs_sufile_get_header_block(sufile, &header_bh);
-	if (err)
+	err = nilfs_mdt_get_block(sufile, 0, 0, NULL, &header_bh);
+	if (unlikely(err)) {
+		if (err == -ENOENT) {
+			nilfs_err(sb,
+				  "missing header block in segment usage metadata");
+			err = -EINVAL;
+		}
 		goto failed;
+	}
 
 	sui = NILFS_SUI(sufile);
 	kaddr = kmap_local_page(header_bh->b_page);
-- 
2.34.1


  reply	other threads:[~2024-08-21 15:46 UTC|newest]

Thread overview: 6+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2024-08-21 15:46 [PATCH 0/5] nilfs2: prevent unexpected ENOENT propagation Ryusuke Konishi
2024-08-21 15:46 ` Ryusuke Konishi [this message]
2024-08-21 15:46 ` [PATCH 2/5] nilfs2: treat missing cpfile header block as metadata corruption Ryusuke Konishi
2024-08-21 15:46 ` [PATCH 3/5] nilfs2: do not propagate ENOENT error from sufile during recovery Ryusuke Konishi
2024-08-21 15:46 ` [PATCH 4/5] nilfs2: do not propagate ENOENT error from sufile during GC Ryusuke Konishi
2024-08-21 15:46 ` [PATCH 5/5] nilfs2: do not propagate ENOENT error from nilfs_sufile_mark_dirty() Ryusuke Konishi

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=20240821154627.11848-2-konishi.ryusuke@gmail.com \
    --to=konishi.ryusuke@gmail.com \
    --cc=akpm@linux-foundation.org \
    --cc=linux-kernel@vger.kernel.org \
    --cc=linux-nilfs@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