[PATCH] do not stat autofs mounts(was Re: hall and autofs)

[John (J5) Palmieri]

On Sat, 2006-05-20 at 11:19 +0800, Ian Kent wrote:
> On Wed, 2006-05-17 at 18:55 -0400, John (J5) Palmieri wrote:
> ? .blockdev.c.swp
> ? hal-donot-check-created-by-hal-on-autofs-mounts-2.patch
> ? hal-donot-check-created-by-hal-on-autofs-mounts-3.patch
> ? hal-donot-check-created-by-hal-on-autofs-mounts.patch
> ? hal-donot-stat-autofs-mounts.patch
> Index: blockdev.c
> ===================================================================
> RCS file: /cvs/hal/hal/hald/linux2/blockdev.c,v
> retrieving revision 1.41
> diff -u -r1.41 blockdev.c
> --- blockdev.c	9 May 2006 20:13:42 -0000	1.41
> +++ blockdev.c	17 May 2006 22:53:13 -0000
> @@ -185,6 +185,7 @@
>  	dev_t devt = makedev(0, 0);
>  	GSList *volumes = NULL;
>  	GSList *volume;
> +        GSList *autofs_mounts = NULL;
>  
>  	/* open /proc/mounts */
>  	g_snprintf (buf, sizeof (buf), "%s/mounts", get_hal_proc_path ());
> @@ -213,6 +214,28 @@
>  		if (strcmp(mnt.mnt_type, "nfs") == 0)
>  			continue;
>  
> +		/* If this is an autofs mount (fstype == 'autofs') 
> +		 * store the mount in a list for later use. 
> +		 * On mounts managed by autofs accessing files below the mount
> +		 * point cause the mount point to be remounted after an 
> +		 * unmount.  We keep the list so we do not check for
> +		 * the .created-by-hal file on mounts under autofs mount points
> +		 */
> +		if (strcmp(mnt.mnt_type, "autofs") == 0) {
> +			char *mnt_dir;
> +
> +			if (mnt.mnt_dir[strlen (mnt.mnt_dir) - 1] != '/')
> +				mnt_dir = g_strdup_printf ("%s/", mnt.mnt_dir);
> 
> Is this right?
> 
> The slash won't appear in /proc/mounts.
> Is it needed at all as the fact that it's an autofs filesystem should
> exclude other media that's present in the fstab.

It is used for comparison later on so if /m is an autofs mount we don't
exclude /media.  Of course if the kernel ABI is stable here *faint
chuckle* We could just add the slash to the end of the directory without
the comparison.  Personally I don't think it hurts and is more robust to
check for the slash.

-- 
John (J5) Palmieri <johnp at redhat.com>