[Cluster-devel] [PATCH v3 2/2] fs: remove mandatory file locking support

[Al Viro]

On Fri, Aug 20, 2021 at 12:39:19PM -0400, Jeff Layton wrote:

> diff --git a/fs/locks.c b/fs/locks.c

> @@ -2857,8 +2744,7 @@ static void lock_get_status(struct seq_file *f, struct file_lock *fl,
>  			seq_puts(f, "POSIX ");
>  
>  		seq_printf(f, " %s ",
> -			     (inode == NULL) ? "*NOINODE*" :
> -			     mandatory_lock(inode) ? "MANDATORY" : "ADVISORY ");
> +			     (inode == NULL) ? "*NOINODE*" : "ADVISORY ");

Huh?

<looks>
        if (fl->fl_file != NULL)
		inode = locks_inode(fl->fl_file);

So basically that's fl->fl_file ? "ADVISORY" : "*NOINODE*"?

How could that trigger, though?  With locks_mandatory_area() gone, I don't
see how FL_POSIX file_lock with NULL ->fl_file could be ever found...
Confused...

Why is locks_copy_conflock() exported (hell, non-static), BTW?

> diff --git a/fs/namespace.c b/fs/namespace.c


> -#ifdef	CONFIG_MANDATORY_FILE_LOCKING
> -static bool may_mandlock(void)
> +static void warn_mandlock(void)
>  {
> -	pr_warn_once("======================================================\n"
> -		     "WARNING: the mand mount option is being deprecated and\n"
> -		     "         will be removed in v5.15!\n"
> -		     "======================================================\n");
> -	return capable(CAP_SYS_ADMIN);
> +	pr_warn_once("=======================================================\n"
> +		     "WARNING: The mand mount option has been deprecated and\n"
> +		     "         and is ignored by this kernel. Remove the mand\n"
> +		     "         option from the mount to silence this warning.\n"
> +		     "=======================================================\n");
>  }
> -#else
> -static inline bool may_mandlock(void)
> -{
> -	pr_warn("VFS: \"mand\" mount option not supported");
> -	return false;
> -}
> -#endif

Is there any point in having the previous patch not folded into this one?