[<prev] [next>] [<thread-prev] [day] [month] [year] [list]
Message-ID: <68fe75c0-852b-4e81-ab39-1c10a3af506b@wdc.com>
Date: Mon, 28 Apr 2025 12:27:37 +0000
From: Hans Holmberg <Hans.Holmberg@....com>
To: "Darrick J. Wong" <djwong@...nel.org>
CC: Carlos Maiolino <cem@...nel.org>, Dave Chinner <david@...morbit.com>, hch
<hch@....de>, "linux-xfs@...r.kernel.org" <linux-xfs@...r.kernel.org>,
"linux-kernel@...r.kernel.org" <linux-kernel@...r.kernel.org>
Subject: Re: [PATCH] xfs: allow ro mounts if rtdev or logdev are read-only
On 25/04/2025 16:54, Darrick J. Wong wrote:
> On Fri, Apr 25, 2025 at 08:52:53AM +0000, Hans Holmberg wrote:
>> Allow read-only mounts on rtdevs and logdevs that are marked as
>> read-only and make sure those mounts can't be remounted read-write.
>
> If the log device is readonly, does that mean the filesystem gets
> mounted norecovery too? Your test might want to check that a dirty log
> is not recovered even if the filesystem mounts.
>
Read only-mounts do not mean that norecovery is set, but xfs will fail
to mount if recovery is needed and the logdev/rtdev/.. is read-only.
To mount in this state, the user needs to pass in norecovery.
> --D
>
>> Signed-off-by: Hans Holmberg <hans.holmberg@....com>
>> ---
>>
>> I will post a couple of xfstests to add coverage for these cases.
>>
>> fs/xfs/xfs_super.c | 24 +++++++++++++++++++++---
>> 1 file changed, 21 insertions(+), 3 deletions(-)
>>
>> diff --git a/fs/xfs/xfs_super.c b/fs/xfs/xfs_super.c
>> index b2dd0c0bf509..d7ac1654bc80 100644
>> --- a/fs/xfs/xfs_super.c
>> +++ b/fs/xfs/xfs_super.c
>> @@ -380,10 +380,14 @@ xfs_blkdev_get(
>> struct file **bdev_filep)
>> {
>> int error = 0;
>> + blk_mode_t mode;
>>
>> - *bdev_filep = bdev_file_open_by_path(name,
>> - BLK_OPEN_READ | BLK_OPEN_WRITE | BLK_OPEN_RESTRICT_WRITES,
>> - mp->m_super, &fs_holder_ops);
>> + mode = BLK_OPEN_READ | BLK_OPEN_RESTRICT_WRITES;
>> + if (!xfs_is_readonly(mp))
>> + mode |= BLK_OPEN_WRITE;
>> +
>> + *bdev_filep = bdev_file_open_by_path(name, mode,
>> + mp->m_super, &fs_holder_ops);
>> if (IS_ERR(*bdev_filep)) {
>> error = PTR_ERR(*bdev_filep);
>> *bdev_filep = NULL;
>> @@ -1969,6 +1973,20 @@ xfs_remount_rw(
>> struct xfs_sb *sbp = &mp->m_sb;
>> int error;
>>
>> + if (mp->m_logdev_targp && mp->m_logdev_targp != mp->m_ddev_targp &&
>> + bdev_read_only(mp->m_logdev_targp->bt_bdev)) {
>> + xfs_warn(mp,
>> + "ro->rw transition prohibited by read-only logdev");
>> + return -EACCES;
>> + }
>> +
>> + if (mp->m_rtdev_targp &&
>> + bdev_read_only(mp->m_rtdev_targp->bt_bdev)) {
>> + xfs_warn(mp,
>> + "ro->rw transition prohibited by read-only rtdev");
>> + return -EACCES;
>> + }
>> +
>> if (xfs_has_norecovery(mp)) {
>> xfs_warn(mp,
>> "ro->rw transition prohibited on norecovery mount");
>> --
>> 2.34.1
>>
>
>
Powered by blists - more mailing lists