Re: vfs_altfsid & dm_fsid

[Aurelien Degremont - Stagiaire <degremont@xxxxxxxxxxx>]

Dean Roehrich a écrit:
> If your filesystem of choice doesn't have an fsid, then you could just
> generate one that is valid while the filesystem is mounted and is not written
> to the filesystem, or you could come up with something else.  Whatever you
> choose for an fsid should fit into a 64-bit type.


> 1)      The fsid should be a parameter to dm_send_mount_event() and
>         dm_send_namesp_event() and dm_send_unmount_event().
>         The get_fsid op in struct filesystem_dmapi_operations should be
>         dropped.

Ok, so we need to modify :
 xfs_dm_mount()
 xfs_dm_send_namesp_event()
 xfs_dm_send_unmount_event()

The fsid could be taken in vfsp.

> 2)      The dm_fsys_vector_t should be folded into struct
>         filesystem_dmapi_operations.  The new ops vector should be
>         a parameter to dm_send_mount_event().
>         The part where today we copy the ops vector from
>         fsys_function_vector_t to dm_fsys_vector_t seems cumbersome and
>         just makes things hard to understand, so maybe
>         that copy should be skipped and some of these datatypes could
>         be removed.
>         The get_fsys_vector op in struct filesystem_dmapi_operations should be
>         dropped.

Remove
 dm_query_fsys_for_vector()
 dm_fsys_vector()

Move the fsys_vector pointer to dmapiops.
Add a bitmask that informs which functions id set when registering.

Change all the dm_sys_vector calls in dmapi.

Maybe the dm_vector_map should be replace by a list_head ?

> 3)      It would be nice to keep in mind distributed filesystems and stackable
>         filesystems; try to make it easier, not harder, to move in those
>         directions.  Unfortunately, I have no experience with the Linux
>         view of stackable filesystems and I don't quite know how to approach
>         that problem.

Ok, I'm not specialist neither. I'll try...

> 4)      The new, unified, ops vector should be kept on a
>         per-filesystem-instance basis, rather than one ops vector to be
>         shared by all filesystems.

So, all the by_fstype should be changed to by_fsid ?
So, we're back to the issue : "how fetch the fsid only knowing the 
superblock ?"
Maybe, we can store a mapping between the superblock and the fsid in the 
dm_fsys_map ? But I don't like using the sb pointer address as id 
(especially concerning distributed fs)
If you have a better solution...

> Does that sum up the things that interest you?

Yes, It did :).
I'd like to discuss all of this.
As soon as we agree, I could start to patch this.


Aurelien