Hi,
On Fri, 26 Feb 2010 14:33:32 +0900, Jiro SEKIBA <jir@xxxxxxxxx> wrote:
> Hi,
>
> This is a trivial patch to protect super_block flags.
>
> super_block flags must be protected by the proper semaphore
> when modifying it.
>
> Signed-off-by: Jiro SEKIBA <jir@xxxxxxxxx>
> ---
> fs/nilfs2/super.c | 2 ++
> 1 files changed, 2 insertions(+), 0 deletions(-)
>
> diff --git a/fs/nilfs2/super.c b/fs/nilfs2/super.c
> index 8173fae..3ee02b8 100644
> --- a/fs/nilfs2/super.c
> +++ b/fs/nilfs2/super.c
> @@ -110,7 +110,9 @@ void nilfs_error(struct super_block *sb, const char *function,
>
> if (nilfs_test_opt(sbi, ERRORS_RO)) {
> printk(KERN_CRIT "Remounting filesystem read-only\n");
> + down_write(&nilfs->ns_super_sem);
> sb->s_flags |= MS_RDONLY;
> + up_write(&nilfs->ns_super_sem);
> }
> }
>
This has a potential for recursive locking since nilfs_error() can be
called from within nilfs_remount() that locks ns_super_sem.
It seems unfavorable though the above sb->s_flags operation is not
protected. I hope we can find a simple and better workaround.
Regards,
Ryusuke Konishi
--
To unsubscribe from this list: send the line "unsubscribe linux-nilfs" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
