>> However, I'm not sure whether this cap_dirty thing is generally
>> useful, or whether all that is needed is a special-case for
>> CAP_RAW_IO. Generalizing it through cap_dirty, however is
>> probably as simple as a patch implementing a single PF_RAWIO flag.
Someone actually uses raw IO when they actually use it...
That is, use of ioperm() does not imply use of IO privs.
The IO bitmap could just remain empty until IO is actually
used. Then the process gets a fault and the process is marked
as having used raw IO. The same goes for mmap() of /dev/mem.
Let the process take a fault on the first access, then mark
the process as having used raw IO.
> Exporting it in /proc probably _is_ a good idea.
It is, though perhaps in a different file.
> You can just disallow access to other users and
> return zero when p->euid!=p->uid.
That might be a job for Mandatory Access Control.
(why is that always capitalized?)
-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.rutgers.edu