linux-ext4 - Re: [BUG REPORT] ext4: “errors=remount-ro” has become “errors=shutdown”?

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives
Hash Suite for Android: free password hash cracker in your pocket
[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <20250106234956.GM6174@frogsfrogsfrogs>
Date: Mon, 6 Jan 2025 15:49:56 -0800
From: "Darrick J. Wong" <djwong@...nel.org>
To: Baokun Li <libaokun1@...wei.com>
Cc: Theodore Ts'o <tytso@....edu>, Jan Kara <jack@...e.cz>,
	"linux-ext4@...r.kernel.org" <linux-ext4@...r.kernel.org>,
	Christian Brauner <brauner@...nel.org>, sunyongjian1@...wei.com,
	Yang Erkun <yangerkun@...wei.com>, linux-fsdevel@...r.kernel.org,
	linux-xfs@...r.kernel.org
Subject: Re: [BUG REPORT] ext4: “errors=remount-ro” has become “errors=shutdown”?

On Sat, Jan 04, 2025 at 10:41:28AM +0800, Baokun Li wrote:
> Hi Ted,
> 
> On 2025/1/3 23:54, Theodore Ts'o wrote:
> > On Fri, Jan 03, 2025 at 10:35:17AM -0500, Theodore Ts'o wrote:
> > > I don't see how setting the shutdown flag causes reads to fail.  That
> > > was true in an early version of the ext4 patch which implemented
> > > shutdown support, but one of the XFS developers (I don't remember if
> > > it was Dave or Cristoph) objected because XFS did not cause the
> > > read_pages function to fail.  Are you seeing this with an upstream
> > > kernel, or with a patched kernel?  The upstream kernel does *not* have
> > > the check in ext4_readpages() or ext4_read_folio() (post folio
> > > conversion).
> > OK, that's weird.  Testing on 6.13-rc4, I don't see the problem simulating an ext4 error:
> > 
> > root@...-xfstests:~# mke2fs -t ext4 -Fq /dev/vdc
> > /dev/vdc contains a ext4 file system
> > 	last mounted on /vdc on Fri Jan  3 10:38:21 2025
> > root@...-xfstests:~# mount -t ext4 -o errors=continue /dev/vdc /vdc
> We are discussing "errors=remount-ro," as the title states, not the
> continue mode. The key code leading to the behavior change is as follows,
> therefore the continue mode is not affected.

Hmm.  On the one hand, XFS has generally returned EIO (or ESHUTDOWN in a
couple of specialty cases) when the fs has been shut down.

OTOH XFS also doesn't have errors=remount-ro; it just dies, which I
think has been its behavior for a long time.

To me, it doesn't sound unreasonable for ext* to allow reads after a
shutdown when errors=remount-ro since it's always had that behavior.

Bonus Q: do you want an errors=fail variant to shut things down fast?

--D

> --- a/fs/ext4/super.c
> +++ b/fs/ext4/super.c
> @@ -657,7 +657,7 @@ static void ext4_handle_error(struct super_block *sb,
> bool force_ro, int error,
>                 WARN_ON_ONCE(1);
> 
>         if (!continue_fs && !sb_rdonly(sb)) {
> -               ext4_set_mount_flag(sb, EXT4_MF_FS_ABORTED);
> +               set_bit(EXT4_FLAGS_SHUTDOWN, &EXT4_SB(sb)->s_ext4_flags);
>                 if (journal)
>                         jbd2_journal_abort(journal, -EIO);
>         }
> 
> See the end for problem reproduction.
> 
> > [   24.780982] EXT4-fs (vdc): mounted filesystem f8595206-fe57-486c-80dd-48b03d41ebdb r/w with ordered data mode. Quota mode: none.
> > root@...-xfstests:~# cp /etc/motd /vdc/motd
> > root@...-xfstests:~# echo testing > /sys/fs/ext4/vdc/trigger_fs_error
> > [   42.943141] EXT4-fs error (device vdc): trigger_test_error:129: comm bash: testing
> > root@...-xfstests:~# cat /vdc/motd
> > 
> > The programs included with the Debian GNU/Linux system are free software;
> > the exact distribution terms for each program are described in the
> > individual files in /usr/share/doc/*/copyright.
> > 
> > Debian GNU/Linux comes with ABSOLUTELY NO WARRANTY, to the extent
> > permitted by applicable law.
> > root@...-xfstests:~#
> > 
> > 
> > HOWEVER, testing with shutdown ioctl, both ext4 and xfs are failing with EIO:
> Yes, this is as expected.
> > root@...-xfstests:~# mount /dev/vdc /vdc
> > [    7.969168] XFS (vdc): Mounting V5 Filesystem 7834ea96-eab0-46c5-9b18-c8f054fa9cf4
> > [    7.978539] XFS (vdc): Ending clean mount
> > root@...-xfstests:~# cp /etc/motd /vdc
> > root@...-xfstests:~# /root/xfstests/src/godown -v /vdc
> > Opening "/vdc"
> > Calling XFS_IOC_GOINGDOWN
> > [   29.354609] XFS (vdc): User initiated shutdown received.
> > [   29.356123] XFS (vdc): Log I/O Error (0x6) detected at xfs_fs_goingdown+0x55/0xb0 (fs/xfs/xfs_fsops.c:452).  Shutting down filesystem.
> > [   29.357092] XFS (vdc): Please unmount the filesystem and rectify the problem(s)
> > root@...-xfstests:~# cat /vdc/motd
> > cat: /vdc/motd: Input/output error
> > root@...-xfstests:~#
> > 
> > So I take back what I said earlier, but I am a bit confused why it
> > worked after simulating an file system error using "echo testing >
> > /sys/fs/ext4/vdc/trigger_fs_error".
> > 
> It's because "errors=remount-ro" wasn't used when mounting...
> 
> Here's a replication:
> 
> root@...-xfstests:~# mount -o errors=remount-ro /dev/vdc /mnt/test
> [  115.731007] EXT4-fs (vdc): mounted filesystem
> 0838f08f-c04e-440c-a9a5-417677efb03e r/w with ordered data mode. Quota mode:
> none.
> root@...-xfstests:~# echo test > /mnt/test/file
> root@...-xfstests:~# cat /mnt/test/file
> test
> root@...-xfstests:~# echo 1 > /sys/fs/ext4/vdc/trigger_fs_error
> [  131.537649] EXT4-fs error (device vdc): trigger_test_error:129: comm
> bash: 1
> [  131.538226] Aborting journal on device vdc-8.
> [  131.538844] EXT4-fs (vdc): Remounting filesystem read-only
> root@...-xfstests:~# cat /mnt/test/file
> cat: /mnt/test/file: Input/output error
> root@...-xfstests:~# uname -a
> Linux kvm-xfstests 6.13.0-rc4-xfstests-g6cfe3548f8f5-dirty #284 SMP
> PREEMPT_DYNAMIC Fri Dec 27 10:39:02 CST 2024 x86_64 GNU/Linux
> 
> 
> Regards,
> Baokun
> 
>