Tigran Aivazian writes:
> On Thu, 29 Jun 2000, Richard Gooch wrote:
> > What happens when you try (user-space) mounting a FS_SINGLE filesystem
> > without calling kern_mount() first?
>
> I get an oops at the line of code I mentioned - it wasn't a theoretical
> oops but a practical one :)
>
> Basically, I was documenting file_system_type structure and wrote:
>
> <item><bf>read_super</bf> - a pointer to the function that reads the super
> block during mount operation. This function is required - if it is
> not
> provided, mount operation (whether from userspace or inkernel) will
> always fail except in FS_SINGLE case where it will Oops in
> get_sb_single() trying to dereference a NULL pointer in
> fs_type->kern_mnt->mnt_sb with (fs_type->kern_mnt = NULL) if the
> module did not call kern_mount() in the initialisation routine after
> filesystem was successfully registered by register_filesystem().
>
> Now, it would sound much nicer if I could just say instead:
>
>
> <item><bf>read_super</bf> - a pointer to the function that reads the super
> block during mount operation. This function is required - if it is
> not provided, mount operation (whether from userspace or
> inkernel) will always fail.
>
> Don't you agree? So, to test it I wrote a trivial filesystem that
> sets FS_SINGLE and yet provides no read_super in fs_type and
> discovered the oops. Then I added a dummy (always failing)
> read_super() and it oops'd exactly the same - so it doesn't matter
> if read_super is NULL or not for this thing (which reading
> read_super() function suggests anyway).
Hm. I agree that the documentation for read_super() would best not
have to mention FS_SINGLE/kern_mount() issues.
However, these issues should be discussed. I think you just need to
document that at a higher level. That will be "clean".
Regards,
Richard....
Permanent: [EMAIL PROTECTED]
Current: [EMAIL PROTECTED]
